TOTHECLOUD

TOTHECLOUDBIGDATAINATURBULENTWORLD

VINCENTMOSCO

Firstpublished2014byParadigmPublishers

Published2016byRoutledge2ParkSquare,MiltonPark,Abingdon,OxonOX144RN711ThirdAvenue,NewYork,NY10017

RoutledgeisanimprintoftheTaylor&FrancisGroup,aninformabusiness

Copyright©2014VincentMosco

Allrightsreserved.Nopartofthisbookmaybereprintedorreproducedorutilisedinanyformorbyanyelectronic,mechanical,orothermeans,nowknownorhereafterinvented,includingphotocopyingandrecording,orinanyinformationstorageorretrievalsystem,withoutpermissioninwritingfromthepublishers.

Notice:Productorcorporatenamesmaybetrademarksorregisteredtrademarks,andareusedonlyforidentificationandexplanationwithoutintenttoinfringe.

LibraryofCongressCataloging-in-PublicationData

Mosco,Vincent.Tothecloud:bigdatainaturbulentworld/VincentMosco.pagescm.Includesbibliographicalreferencesandindex.ISBN978-1-61205-615-9(hardcover:alk.paper)ISBN978-1-61205-616-6(pbk.:alk.paper)ISBN978-1-315-63155-4(e-book)1.Cloudcomputing—Socialaspects.2.Bigdata—Socialaspects.3.Privacy,Rightof.I.Title.QA76.9.C66M6632014004.67′82—dc3

2013046088

DesignedandTypesetbyStraightCreekBookmakers.

ISBN13:978-1-61205-615-9(hbk)ISBN13:978-1-61205-616-6(pbk)

WiththankstoLouise

Chapter1

Chapter2

Chapter3

Chapter4

Chapter5

CONTENTSAcknowledgments

TheCloudAteMyHomework

FromtheComputerUtilitytoCloudComputing

SellingtheCloudSublime

DarkClouds

BigDataandCloudCulture

NotesReferencesIndexAbouttheAuthor

ACKNOWLEDGMENTS

Asalways,Iamgratefulfortheideas,criticism,andsuggestionsofCatherineMcKercher,mypartnerinlifeandinresearch.ShehelpedmakethisabetterbookinmorewaysthanIcan count. Most importantly, Cathy’s unconditional support and love are a continuingsourceof inspiration.DanSchillerhasbeenadear friend for thirty-fiveyears.Wehavemetallovertheworldtoprovidethebesthelpafriendcanoffer:constructivebutcandidassessmentsofeachother’swork.Shortlybeforestartingtowritethisbook,ImetupwithDaninSantaFe,NewMexico,and,afterpatientlyhearingwhatIhadinmind,headvisedthatIwriteachapteronmarketingthecloud.IamnotsurehowhewillfindtheexecutionofhisrecommendationinChapter3,butIamverygratefulforhiscreativesuggestionandforsharingresearchmaterialthathasstrengthenedthebook.ThanksalsotoDerekMortonforaccompanyingmeonmysearchforthecloudinNewYork’sMetropolitanMuseumofArt,where his photography helpedme to reflect onTomásSaraceno’sCloudCity longafterwe left theexhibition.Derekwasalsokind to remembermewhenhecameacrossmaterial useful for the book. When a Microsoft advertisement for cloud computingseemedtovanishfromtheInternet,MadelineMoscotrackeditdownforme.Thankyou,Madeline.

Iamalwaysespeciallygratefulwhen formerstudents learnaboutanewbookprojectand take the time to sendme useful reports or just share their views.Thank you,RickEmrich,PatMazepa,IanNagy,andAlexSavulescu.LikewiseforcurrentstudentsImeetwhen lecturing on a topic like cloud computing. In this case, special thanks to AdeelKhamisaofCarletonUniversity,aswellasLaimaJanciuteandEmmaAgusita,whomImetattheUniversityofWestminster,London.

Lecturingprovidesagreatopportunity to tryout ideasandencounternewones. IamgratefultoDanielParé,whohostedatalkoncloudcomputingattheUniversityofOttawa;toChristianFuchs,whoorganizedmylectureatLondon’sUniversityofWestminster;andto Aliaa Dakoury, who kindly invited me to deliver a keynote address to the annualconferenceoftheGlobalCommunicationAssociation.

IwouldalsoliketothanktheCanadaResearchChairsProgramandQueen’sUniversityforprovidingresearchfunding.Finally,Iamgratefultothemanypeoplewhomadetheirmarkonmythinkingoverthepastfortyyearsofresearchandteachingoncommunicationandinformationtechnology.

CHAPTER1THECLOUDATEMYHOMEWORK

Likewaterorelectricity,cloudcomputingshouldnowbeconsideredakeyutilityandthereforeshouldbeavailabletoall.

(Groucutt2013)

TheInternethadbeenaroundforawhilewhenonJuly5,1993,theNewYorkermagazinefeaturedacartoonthat, in themindsofsome,markeditsrealarrival.“OntheInternet,”saysthedogatthecomputerscreentohiscaninefriend,“nobodyknowsyou’readog.”IknewitwastimetowritethisbookwhenIwokeuponemorning,downloadedmydigitaleditionof theOctober8,2012,NewYorker,andcameacrossanewversionofaclassiccartoon.Alittleboylooksupathisteacherand,withhopeandtrepidation,pleadshiscase:“Thecloudatemyhomework.”Okay,perhapsnoteveryonegotthejoke,butmostreaderswouldhavesomeconceptionofthecloudastheplacewheredatalivesuntilitiscalleduponthecomputer,tablet,orsmartphone—or,inthecaseofamalfunction,theplacewheredata goes to die. This book explains what little Johnny is talking about and why it isimportant. For better or for worse, the cloud has arrived. The cloud that ate Johnny’shomework is a key force in the changing international political economy. The globalexpansion of networked data centers controlled by a handful of companies continues aprocessofbuildingaglobalinformationeconomy,oncecharacterizedbyBillGates(1995)as “friction-free capitalism.” Companies that once housed an information-technologydepartmentwithitscrafttraditioncannowmovemostofitsworktothecloud,whereITfunctions and its labor are centralized in an industrialmode of production, processing,storage,anddistribution.Furthermore, thecloudtakesthenextstepinalongprocessofcreatingaglobalcultureofknowing,capturedinthetermbigdata,orwhatmightbetterbecalleddigitalpositivism.Hereinformationproductionacceleratesinnetworksthatlinkdatacenters,devices,organizations,and individualsappearing tocreate, in thewordsofoneguru, “aglobal superintelligence” (Wolf2010).Thecloudandbigdata are enginesthatpowerinformationalcapitalismevenastheyenableanincreasinglydominantwayofknowing. These interlinked processes and the challenges to them comprise the majorthemesofTotheCloud.Ihavebeenthinkingaboutcloudcomputingsince2010,whenitbegantoenterpublic

consciousness,particularlyafteracoupleofsplashySuperBowladsairedduringthe2011game.ThenApplegotintotheactwheniturgeduserstomovetheirphotos,music,mail,andfilestoitsiCloud.Notwantingtogiveupcontrolovermystashoffamilyphotosandworriedabout the securityofmymail, I resisteddoinganythingmore thanuploadingafewincidentals(althoughforsomereasonIdidnotmindsendingmyphotosintothecloudknownasFlickr).Likemanypeople, Iwas aware that someofmy thingswere finding

their way from my computer to remote servers, but this left me feeling a bituncomfortable. Stories about cloud security breaches, disappearing data, andenvironmentalrisksatclouddatacentersweremakingpeoplefeelthatnotallcloudswerebrightandonlyafewweregreen.But themigrationoforganizationalandpersonaldatacontinued,asdidthemarketing.

Idecidedtotakeacloserlookwhenreferencestocloudsofallsortsbegantoappear,partlypromptedbythearrivalofcloudcomputingandpartlyowingtomygrowingcloud-consciousness.Firstitwasmediaattentiontoanobscuremedievaltreatise,TheCloudofUnknowing, that led me to wonder about the philosophical assumptions embedded incloudcomputing.ThentherewasDavidMitchell’sstrangelytitlednovelCloudAtlasandthe announcement of a blockbuster film based on the book’smystical account of soulsmigrating like clouds across time and space. I began collecting images of cloud datacenters as they continued to spring up around the world, and was struck by the clashbetween the banality of their form—low-rise, endlessly bland warehouses—and thesublimityof real clouds.There is nothing ethereal about thesebuildings.Moreover,myreadingandconversationspointedtogrowingtensionsinthepolitical,economic,social,andaestheticdimensionsofcloudcomputing.Butat thisearlystageofitsdevelopment,mostextendedtreatmentsremainedlimitedtotechnicaldescriptions.

Although cloud computing did notmake an appearance onmypersonal radar screenuntil2010,Ihavebeenresearching,writing,andspeakingaboutcomputercommunicationforfortyyears,includingworkingonandaroundpredecessorstocloudcomputing.Intheearly1970s,asagraduatestudentinsociologyatHarvard,Ihandedovermypunchcardsto thecentralcomputerfacilityandhopedtoreceiveapaperprintoutofresearchresultsusingmyprofessor’spioneeringGeneral Inquirersoftware that, remarkably for its time,analyzedthecontentoftext.Atthat time,wewereall inthecloudbecausethepersonalcomputer,withitsbuilt-instoragedevice,wasyearsaway.Allthatwecoulddowasfindtimetoenterdatainacomputerterminal,appropriatelyreferredtoasdumb,andwaitforthe mainframe to provide results. Ten years later I wrote about the cloud of its time,videotex, which promised, and in rudimentary ways delivered, text and images fromcentralcomputerstoenhancedscreens(Mosco1982).MovingtoCanadain1984,Itriedout Telidon, which Canadian technologists and policy makers insisted was the mostadvancedofthenewinteractivetelecommunicationsservices.Moreimportantly,IlearnedabouttheresearchofCanadianDouglasParkhill,whosework,particularlyTheChallengeoftheComputerUtility(1966),iswidelyrecognizedasaforerunnertocloudcomputing.Overthattime,inadditiontoaddressingmanyoftheissuesthatarenowemergingin

cloud computing, I began to understand the importance of recognizing problems thatinevitablyarisefromnewsystemsforstoring,processing,andexchanginginformation.Itis tempting to applywhat appear to be the lessons of history to new technologies and,while it iscertainlywisetosituatenewtechnologiesintheirhistoricalcontext, it isalsoessential to recognize thatchanging technologiesandachangingworldalsobringaboutdisruptions,disjunctions,and,sometimes,revolutionsinhistoricalpatterns.

Therearenownumeroustechnicalguidesandprimersthatofferusefuloverviewsofthesubject,andmybookiscertainlyindebtedtothese(Erl,Puttini,andMahmood2013).Butmypurposeistopromotethediscussionofcloudcomputingbeyondwhatthesetextshave

tosaybytakingupitspolitical,economic,social,andculturalsignificance.Inordertodothis, the book draws from the transdisciplinary contributions to be found in technologystudies,sociology,culturalstudies,andpoliticaleconomy.Myaimistounsettletraditionalwaysofthinkingwithacriticalinterrogation.Sendingdataintothecloudisadecisiontoengage with one or another data center, say Amazon’s orMicrosoft’s. But it is also achoicethathasimplicationsthatareeconomic(whopaysforit?),political(whocontrolsit?), social (how private is it?), environmental (what is its impact on the land and onenergyuse?), and cultural (whatvaluesdoes it embody?).Akeygoal of thebook is toadvanceaconversationbetweentheprofessionalswhoworkinthefield,thoseresponsiblefor promoting it, and the researchers, policy makers, and activists who study cloudcomputingandthinkaboutitsimpact,implications,andchallenges.

Whyisitnecessarytoplacecloudcomputinginthebiggerpictureofpoliticaleconomy,society,andculture? Is itnotsufficient tosimplydescribewhatcloudcomputinghas toofferabusinessandweighitscostsandbenefits?Itakeupsomeofthepracticalproblemsinvolved in adopting and implementing cloud systems in the next chapter. However,limiting discussion to this point alone does not give sufficient credit to the cloudcomputing movement as a force in society. Notwithstanding the hyperbole thataccompaniesnewcommunicationtechnologiesandsystems,fromthetelegraphthatwouldbringtogethernationsinpeacefulharmonytothepromiseofmasseducationontelevision,cloud computing is having an enormous impact across societies. This extends fromcompanies that aremoving theirdataandbusiness-process software to thecloud, to themilitarythatplansandexecutesbattlestrategiesinthecloud, toschoolsanduniversitiesthat are using the cloud to transform education, and to individualswho are storing thetracesoftheiridentitiesinthecloud.Italsoencompasseswhatsomeconsiderbottom-upversionsofcloudcomputing,suchascommunitygridprojectsthatharnessthecombinedpowerofpersonalcomputers tocarryoutpublic-interest research.Thecloud iscreditedwith catapulting companies likeApple into the corporate stratosphere.Amazon’s cloudwasoneof themost important instrumentsbehindBarackObama’s2012victory.Whiletheseare importantdevelopments, theyarebenigncompared to theclaim that thecloudcan save capitalism by powering it to renewed heights of productivity, or the oppositeexpectationthatitwillopenthedoortocarefullyplannedhackerattacksthatwilldisrupttheworldeconomy.AreChinaandIrantryingtobringAmerica’sfinancialsystemtothedigital brink? Or, as China claims, is the United States becoming a major “hackingempire”?

Sinceexaggeratedpromisestypicallyaccompanytheriseofnewtechnicalsystems,itiseasytodismisstoday’shypeaboutcloudcomputing,butthatwouldbewrong.Thisisnotbecausethestoriesaboutacloud-computingandbig-datarevolution,withtheirvisionsofboundlesseconomicprosperity,areanymoreaccuratethanpromisesofworldpeaceintheage of radio. Rather, the marketing hype supports myths that are taken seriously asstorylines for our time. If successful, they become common sense, the bedrock ofseemingly unchallengeable beliefs that influence not only how we think about cloudcomputing,butabouttechnologyingeneralandourrelationshiptoit.Thedecisiontogiveup your own or your organization’s data to a cloud company is a significant one andcompaniespromotingthetechnologywouldunderstandablyhaveusfocusonitsbenefits.Moreover,itisimportanttotakethehypeseriouslyasthemythicembodimentofwhat,in

an earlier book, I called the digital sublime, the tendency of technology, in this casecomputercommunication,totakeonatranscendentroleintheworldbeyondthebanalityof its role ineveryday life (Mosco2004). It is time togivecloudcomputing itsduebystartingaconversationaboutitsplaceinsocietyandculture.

Cloud computing is a significant development in its own right and a prism throughwhich to viewproblems facing societies confronting the turbulentworld of informationtechnology.Thecloudhasdeephistoricalrootsanditisimportanttoconsiderthem,butitalsohasnewfeaturesthatrequireacloselookatwhatmakescloudsystemsquantitativelyandqualitativelydifferent.Moreover,cloudcomputingservesasaprismthatreflectsandrefractseverymajorissueinthefieldofinformationtechnologyandsociety,includingthefragile environment, ownership and control, security and privacy, work and labor, thestrugglesamongnationsfordominanceintheglobalpoliticaleconomy,andhowwemakesenseofthisworldindiscourseandinculturalexpression.

Chapter2tellsthestoryofcloudcomputing,fromitsoriginsinthe1950sconceptofthecomputerutilitytothepresent-daygiantdatacentersthatfillvastopenspaceseverywherein the world. Back in the 1950s, as even most casual histories of cloud computingdescribe,debatesovertheneedfora“computerutility”anticipatedtoday’sdebatesaboutthecloud.Atthattime,peoplewhowerefamiliarwithutilitiesthatprovidedroads,water,and electricity wondered whether there was need for a public or regulated utility forcomputercommunication.Wasnotinformationasessentialaresourceasroads,water,andpower? With widespread agreement that it was both a resource and essential, someconcludedthatahandfulofcentralizedcomputerfacilitiesstrategicallylocatedaroundtheworld and connected by telecommunications networks to keyboards and screenswouldsatisfytheworld’sneedforinformation.Today,therearefarmorethanahandfuloflargedata centersworldwide, but the principle of the utility is inscribed in cloud computingsystems to the point that interest is returning to this venerable idea.Questions are alsoemergingaboutwhethercomputerutilities shouldbegovernmententerprises,or at leastpubliclyregulatedeveniftheyremaincommercialenterprises.

Chapter 2 examines a variety of the cloud’s predecessors from when the computerutilitywasyoung.TheSovietUnionstakedmuchofitseconomicstrategyinthe1950sonthe ability to build large-scale “cybernetic” systems to carry out thework of a plannedeconomy.Inthe1970stheChileangovernmentexperimentedonademocraticversionofsuchastrategy,withworkersonthegroundcontributingtotheeconomic-planningprocessthrough computer systems. The 1980s saw the development of government andcommercial systems for providing information on demand through what were calledteletext and videotex systems. Their full potential was not realized until the InternetappearedondesktopcomputersandinNewYorkercartoonsinthe1990s.Chapter 2 proceeds to define cloud computing and take up its diverse forms and

characteristics.Cloudcomputinghasbeendefined inmanyways,butmostwouldagreethat it is a powerful system for producing, storing, analyzing, and distributing data,information, applications, and services to organizations and individuals. If youcommunicatewithGmail,downloadmusicfromiCloud,buyKindlebooksfromAmazon,orifyourcompanyusesSalesforcetomanageitscustomerdatabase,thenyouknowaboutandusethecloud.Amongitsmajorcharacteristics,cloudcomputingenableson-demand

self-serviceaccesstoinformationandservicesdeliveredoverglobalnetworks—including,butnotlimitedto,thepublicnetworksoftheInternet.Informationandapplicationscanbepooled to meet user needs, provided and withdrawn on demand, and paid throughmeasuredservicebilling.Thechapterdescribestherangeofcloudcomputingformsfromthe simple provision of an infrastructure, such as a data storage center, to services thatincludeapplications,software,andanalyticsthataddvaluetodata.Italsoconsiderstypesofcloudcomputingfrompubliccloudsthatareavailabletoallpayingcustomers,aratherlimitedmeaningoftheterm“public,”toprivatecloudsthatsellstorageandservicesonlytoaselectsetofcustomerswhoprefertheirdatagatedandsecure,andhybridcloudsthatoffercombinationsofthetwo.

Thechapterexaminestheleadingcloudcompanies,includingthewell-knownfirmsthatgrewupintheInternetera,helpedtocreatesocialmedia,andarenowservingcompaniesand individuals in thecloud.Amazon isarguably the leadingcloud-computingprovider,but the listof familiarnamesalso includesMicrosoft,Google,Apple,andFacebook. Inaddition,legacyfirmssuchasIBM,Oracle,andCiscoaretryingtomakethetransitiontothecloudafteryearsofsuccessservicingcorporateandgovernmentITdepartments.Thentherearethecompaniesbornincloud,suchasRackspace,Salesforce,andVMware,thatprovidegeneralandspecializedcloud-computingandbig-dataservices.Chapter2coversthebattlesamongkeycompetitorsandthegrowingconcentrationofpowerat the topoftheindustry.Privatefirmsdominatethecloud,buttheU.S.governmentishelpingtoshapeits expansion primarily through partnerships with leading companies, mainly in themilitaryand intelligence sectorsbut also in education, including thehumanities.This isleadingsometowonderabouttheriseofamilitaryinformationcomplexthatpromotesthepowerofahandfulofcompaniesandtheexpansionofthesurveillancestate,besttypifiedby the National Security Agency. The U.S. cloud industry is powerful, but it isincreasinglychallengedbyforeigncompetitors—especiallyChina,which isconstructingentirecloudcitiestoclosethegapwiththeUnitedStates.

Thereisamassive,worldwidemovementtopromotecloudcomputing,andChapter3examines its many forms. The campaign includes advertising, blogs, the reports ofcorporate research and consulting firms, international economic-policy organizations,lobbying campaigns, conferences, and trade fairs. Having begun in the banality of atechnicaldiagramandinthehazyvisionsofcomputerpioneers,theimageofthecloudhastaken on a richer aesthetic in the hands of today’s Mad Men, the advertising gurusmarketingthenextnewthing.Inthisrespect,thematerialityofthecloudisnotlimitedtobuildings,computers,software,anddata.Itisalsoembodiedincampaignstoremaketheprosaicstuffofengineeringintothecompellingimageofthecloud.Therewasnomagicinhow this happened. To bring the cloud into widespread awareness it took marketingcampaigns thatdevelopedfromSalesforce’s twoveryexpensiveadvertisements featuredin the 2011SuperBowl game; they highlighted the singerWill.i.amof theBlackEyedPeasandtheanimatedcharacterChatty“themagicalcloud.”Layingthegroundworkforthis big splash was IBM’s foray into cloud marketing with its 2010 “smart cloud”campaign pitched to corporate decision makers, and Microsoft’s “To the Cloud”advertisementsaimedatsmallbusinessandconsumers.Applejoinedthechorusinabigway by changing the name of its online service,which began as “.mac,” shifted to thepersonal(and,somewouldsay,self-absorbed)“.me,”andthensettledoniCloud.

Commercial advertising is important to reach both institutional and individualcustomers.However,itisonlyonepartofacircuitofpromotionthatalsoincludesblogs,newsletters, and social-media sites that provide information about the industry with anemphasis on how to sell cloud computing by countering its critics and advancing itsbenefits.Oneof theirmost important functions is toserveasa transmissionbelt for thefindings of more legitimate outlets like the reports of private research and consultingfirms, includingGartner,McKinsey,Deloitte,andForrester.Eachof these leaders in thefield has produced one or more reports on cloud computing and big data. With theexceptionofone,whichappearedearly (andwasnullifiedbya later reportby thesamecompany), they are all massively optimistic in their forecasts about the cloud. Themessageissimple:movetothecloud.Althoughtheirreportsareexpensive,theessentialfindingsandtheenthusiasm,asChapter3demonstrates,circulatethroughthehundredsofblogs and newsletters that share the enthusiasm. The circuit of promotion expandsinternationallywithreportsthatbringtogetherglobalplayersinbusinessandgovernmenttopromotethecloud.Chapter3concentratesonareportproducedbytheWorldEconomicForum, best known for the annual Davos conference, that documents the unassailablesignificanceofinformationtechnology,cloudcomputing,andbig-dataanalytics.Withthestamp of global legitimacy and the blessing of national and international governmentagencies, as well as corporate participants, the World Economic Forum adds to thelegitimacyofthecloudastheleading-edgeforcefortheexpansionoftheworldeconomy.Thechapterconcludesbyexaminingtwomorevitalelementsinthiscircuitofpromotion,lobbying,andtradeshows.Formostofitshistory,especiallysincethedevelopmentoftheInternet, the information-technology industry has not invested significant resources tolobbyWashington.Inrecentyears,butespeciallywiththegrowthofsocialmediaandthecloud,allofthathaschanged,andChapter3demonstratestheimportanceoflobbyingatthelocal,national,andinternationallevelsofpower.Finally,tradeshowsandconferencesbringthemajorcloudandbig-dataplayerstogethertopromotetheirproducts,theindustry,and themyth of the cloud as a transcendent force to solve theworld’s problems. Thissectiondrawsfrommyparticipationinthelargestannualcloud-computingconferenceandsalesevent,CloudComputingExpo2013inNewYorkCity.

Chapter 4 explains why a massive promotional effort is essential. Cloud computingfacesseriousproblemsbecauseitputsgreatstressontheenvironment,requiressignificantpower supplies, threatens privacy, is difficult to secure, and challenges the future of ITwork.Theseproblems,understandably,receivelittleattentioninthepromotionalaccountsaddressed in Chapter 3. When discussed, they are typically dismissed out of hand orframedinthecontextofhowtocounterargumentsagainstmovingtothecloudbecauseofthese problems. Chapter 4 demonstrates why, contrary to the claims made in thepromotionalculture,itisimportanttogivethemcarefulattention.

Cloud companies promise, and their customers expect, that data centerswill operatewithnodown time.Thisalonemakesenormousdemandson theelectricalgrid,but thedemand increases substantially because servers require a constant source of cooling toavoid overheating.Moreover, 24/7 operationmakes it necessary to build backup powersourceslikedieselgeneratorsandchemicalbatteriesthatcreatesignificantenvironmentalproblems for the communities that host data centers. Moving to the cloud is far fromentering the ethereal,weightless, and green environment that the image of the physical

cloudandthemythologyofcloudcomputingsuggest.Thenextdarkcloud toappear, inChapter4,isthethreattoprivacyandsecurity.Afterexaminingarangeofwaystothinkabout privacy and security, it takes up three major problems, starting with themultiplicationofhackingattacksagainstcloudcomputingsystemsemanatingfromwithinandoutsidethebordersofcompaniesofferingcloudservices.Cyber-attackshavebecomeaninstrumentofgovernmentpolicy.Furthermore,privacyandsecurityarechallengedbythenature ofwhat I call surveillancecapitalism.A significant source of revenue in thecloud and big data is the opportunity to market information about subscribers andcustomers to advertisers. For example, Facebook could not survive as a commercialenterprise without the ability to exercise close surveillance on its 1.3 billion users.Alongsidesurveillancecapitalismisthesurveillancestate,which,astherevelationsabouttheNationalSecurityAgencyrevealed,hasalmostcompleteaccess todatastored in thecloudanddeliveredovertheInternetandotherelectronicnetworks.It isnowonderthatinstitutionsofallsorts,aswellasindividualconsumers,areincreasinglyworriedaboutthesecurity implications of moving to the cloud, whether the data centers are located inChina,Europe,ortheUnitedStates.

One of the primary reasons formoving to the cloud is to streamline, if not entirelyeliminate, an organization’s IT department, amounting to an emerging dark cloud forprofessional labor. But the issue is not limited to IT. Specialized cloud companies likeSalesforcecantakeover themanagementofcustomerrelations, therebyfreeingfirmstocut back on their in-house sales and marketing activities. Moreover, since thepreponderanceofknowledge labor increasingly involves ITwork,whether ineducation,journalism, or health care, this dark cloud now hovers over a large segment of theoccupationalworld.Chapter4documentsthesedevelopmentsandsituatesthemwithinadynamic international division of labor in the information-technology industries wherechainsofaccumulationmeetchainsofresistance,fromFoxconninShenzhentoAppleinCupertino.Asmoreorganizationsandindividualsdecidetoenterthecloud,willtheglobalsystemthatsupportsitremainintact?Whathappensifitruptures?

Chapter 5 concludes the book by shifting to the cultural significance of cloudcomputing.Itisguidedbytheviewthatcultureresistsessentialismsofalltypes,includingthetendencyinthedigitalworld,nowembodiedincloudcomputing,toreducethecloudto an information repository and the foundation for the digital positivism of big-dataanalysis. It starts to pursue this theme by considering what we can learn from themovement touse thecloudfor large-scaledataanalysis—whathasbeencalledbigdata.Thechapterassessestheassumptionsandcomponentsofbigdata,includingarelianceonquantitative, correlational analysis, free from theoretical considerations and aiming topredictevents.Manyofbigdata’sproponentsferventlybelievethatthedatawillspeakforitself,enablingresearcherstoeschewqualitativedata(ortrytorenderitinquantities)andend relianceon causality, theory, andhistory, the traditionalbedrockof social scientificanalysis.Concludingthatatechnicalcritique,howeveruseful,isinsufficienttoaddressthephilosophicalgroundingofwhatisprimarilyadigitalpositivism,thechapterdrawsfromthe culture of clouds to take up the specific way of knowing that underlies big-dataanalysis.Thismattersbecauseeverytechnologycontainsanaesthetic,awayofseeingandfeeling, that is drawn from the machine’s design—as well as from its discursiveassociations. Cloud computing is no exception. The simple schematic diagram of a

networkofcloudsthatgaverisetothetermpresentsawell-ordered,natural,andbenignwaytothinkaboutitthatischallengedbythecultureofclouds,asubterraneanstreamofthoughtthatprovidesapowerfulcounterweighttodigitalpositivism.

FromtheearlydaysoftheInternet,supporterswerenotshyaboutdressingitupinthelanguage of philosophy and even mysticism. For example, many big names, includingsuch luminaries asAlGore andTomWolfe, praised the Jesuit priestPierreTeilharddeChardin(whowasalsoaphilosopher,paleontologist,andpoet)asacyberspacevisionary.Heneverusedacomputeranddiedin1955,butWiredmagazineexclaimed,“TeilhardsawtheNetcomingmorethanhalfacenturybeforeitarrived”(Kreisberg1995).Althoughhepredictednothing about computers andwrote in the impenetrable languageof amystic,the Jesuit priest appealed to cyber-gurus and others because he saw information as theleading force in cosmic evolution. For Teilhard, the growth of information literallyproduced an atmosphere of thought,what he called thenoosphere,which encircled theglobe,putting increasingpressureon theplanet.Eventually, thepressureof informationwould create a massive explosion, taking humankind into the next phase of cosmicevolution.Howeverbizarre the image andhowever it clasheswith everythingweknowaboutphysics, there are fewmoredramaticways tomythologize theburgeoningdigitalworldthanwithacloudofknowingpointingthewaytoprogress.However,othervoicesinthecultureofcloudsanswer,“notsofast.”

There is more to the metaphor of the cloud than capturing the sublimity of cloudcomputing. In its rich history, thatmetaphor contains a critique that challenges utopianvisions finding transcendence, if not the divine, in new technology. Considering itsubiquitouspresenceandpersistencethroughouttime,itisnosurprisetofindthecloudinmanyexpressionsofthehumanimagination.Thewrittenword,music,andthevisualartswouldbemuchpoorerwithoutthemetaphoricalcloud.Fromthebroadsweepofthecloudinculture,IhavechosenthreeexemplarsfromvastlydifferentperiodsinWesternsocietytodocumentcontrastsbetween themetaphorand the information technology thatwouldadoptit.

ItbeginswithTheClouds,acomedywrittenbyAristophanesthatsatirizedintellectuallifeinfifth-century-BCGreece.Itraisesaclear,andhumorous,challengetotheadamantlyrationalmodelofthoughtthatthecloudandbigdataembody,andquestionstheinherentsuperiority of the seemingly apolitical philosopher-technician. Its chorus of cloudsremindsaudiences to thisday thateven themostseeminglyobjectiveof intellectuals, inthiscasethegreatphilosopherSocrates,isembeddedinapoliticalworldwherepracticalexperience often trumps technical knowledge. For the Greek playwright, the way ofknowingestablished2,500yearsagocomesnotintheformoftheintellectuallivingalifeof contemplation in the clouds of abstraction. That was little more than a Platonicaspiration. Rather it is the philosopher-trickster, the intellectual spin doctor, whodominateswith rhetoric and propaganda seasonedwith just enough information. In theWesternwayofknowing,thereisnopuretruthstoredandprocessedinthecloud—justtheongoing struggle between reason and rhetoric. It is amessage that today’s philosopherkings, the computer gurus and data scientists that live in our new cloud,would benefitfromhearing.

Next,wemoveahead to the lasthalfof the fourteenthcenturyADandTheCloudof

Unknowing, theworkofanEnglishmonkwhoadvisesayoungmonkonhow to liveagood,morallife.AlthoughwrittenintheMiddleEnglishofthetime,itisnotanobscurework today.Therearenumerouscontemporary translationsand ithas receivedattentionfromsuchliterarygiantsasDonDeLillo,whousesitinhismagisterialnovelUnderworld.Whatmakesthisbookmostinterestingisitsuseofthecloudasasymbolofwhatgetsinthewayandblockspeoplefromknowingthemselvesandrealizingtheirdestiny.Asonewouldexpect,TheCloudofUnknowingiswritteninareligiousidiom.Fortheunknownwriterof thisspiritualguide, thegoal is tocomeascloseaspossible togod.But justasonedoesnothavetoacceptTeilhard’sgod,thegodofperfectinformation,wedonothavetobelieveinthemonk’sgodinordertoappreciatethepointthatthecloudofinformationthat increasingly saturates our world can get in the way of fulfillment, spiritual orotherwise. Writing in strong yet conversational language, the elder monk advises theneophyte to emptyhimself of information in order to growas a person.The cloud thatappearssoattractiveisactuallyadeterrenttowisdom,acloudofunknowing.

The Cloud of Unknowing bears the imprint of Eastern philosophy, making it all themoreremarkable that itcomesfromtheworkofamedievalEnglishmonkwhoseworldhadbeenshakenbytheBlackPlague.Theviewthatweneedtoemptyourselvesofwhatpasses for knowledge in order to achieve true wisdom and fulfillment is increasinglypopularintheWest,wherepeopleappeartobeoverwhelmedbydata,evenastheyworkto figureout the latest device that promises instant connection to thedigitalworld.Myreasonforanalyzingitinthefinalsectionofthisbookistoaddresstheconflictednatureofourthinkingandfeelingaboutthecloud.Cloudcultureisacontestedterrainfeaturingdifferentviewsaboutepistemology(whatitmeanstoknow),metaphysics(whatitmeanstobe),andmoralphilosophy(whatitmeanstoliveethically).

One of themost interesting cultural expressions of uncertainty is contained inDavidMitchell’s novel Cloud Atlas, which became a feature film directed by the teamresponsible for theMatrix trilogy. The title itself presents a jarring clash because thetraditionalatlasismeanttochartfixedgeographicalformssuchasoceansandlandmasses,nottheconstantlychangingmistsofwatervapor.Thecloudisanythingbutafixedentityanddefiesconventionalmapping,somethingthatisborneoutintheplotofCloudAtlasaswefollowthesixseparatestoriesthattakebothbookandfilmoverseveralcenturies.ForMitchell and the film’s trio of producers, the cloud represents neither the certainty ofinformationnorthebarriertoperfection,butthewispyandvaporousconnectionsthatlinkpeopleovergenerations.Thevarietyofstructuredandrandomactionsthatpropelpeoplethrough life touches those who come after them, so here mapping the cloud becomestellingthestoriesoftheirconnectionsnotinthenetworkdiagramofcloudcomputing,butinthemuchlooserbutno-less-powerfulimageofthematerialcloud.Thisatlasofcloudsrethinkstheconventionalatlasbymappingconnectionsintimeandnotjustinspace.Forthesereasons,CloudAtlasoffersonealternativeforhowtothinkaboutcloudculturethatdoesnotsimplyrequireachoicebetweenthecloudofknowingandofunknowing.

Thisbookconcludesbytakingupartisticmanifestationsoftheseideasincloudculture,oneofwhose icons isRenéMagritte’sTheEmpireofLight, apainting that features thebrightblueofadaytimeskyfilledwithpuffywhitecloudsthatoverseearowofhousesinnighttimedarkness.Somethingisawryinthecloudsandontheground.Takingadifferentperspectiveisacontemporarywork,TomásSaraceno’sremarkableinstallationCloudCity,

an assemblage of large, interconnected modules built with transparent and reflectivematerialsthatoccupiedtheroofgardenoftheMetropolitanMuseumofArtforsixmonthsin2012.WecallonMagrittetoquestiontheseemingharmonyofcloudnetworksandonSaraceno to see ourselves in the reflecting glass of his cloud.Where are we in cloudcomputing?Someartists arebeginning toaddress this issuedirectlybyproducingworkaboutcloudcomputing.For thatweconsiderTamikoThiel,whose installationCloudingGreendepictsdifferentlycoloredcloudsthathoverovereightmajorSiliconValleycloud-computingproviderstodescribetheirenvironmentalrecord.Thesesurrealrepresentationsdraw from and add aesthetic power to a 2012 Greenpeace environmental assessment,“HowCleanIsYourCloud?”

TotheCloudrecognizesthatitistimetomovebeyondtechnicaldescriptionsofcloudcomputing by producing a critical assessment. To begin the process, the next chapterexplorestheoriginsofcloudcomputinginvisionsofthecomputerutility.Itproceedstoexaminetheprinciplesthatdistinguishcloudcomputing,describeswhatcloudcomputingactuallydoes,andmapsthestateofthecloud-computingindustry.

CHAPTER2FROMTHECOMPUTERUTILITYTO

CLOUDCOMPUTING

Weareonashiftthatisasmomentousandasfundamentalastheshifttotheelectricalgrid.It’shappeningalotfasterthananyofusthought.

—ArthurR.Jassy,headofAmazonWebServices(Hardy2012a)

Most general accounts of cloud computing attribute the use of the cloud image to itsappearanceindiagramsthatidentifykeyelementsinatelecommunicationsnetwork.Theterm cloud computing emerged in 1996when technology leaderswith Compaq, then amajordesktop-computercompany,mettodiscussthefutureofcomputingandespeciallytheInternet.Specifically,theyhopedthat“cloudcomputing–enabledapplications”wouldboost sales.Althoughnotentirelyclearabout this, theyconcluded thatonlineconsumerfile storage would likely be among the successful applications. Their prescience wasrewarding for the company because it contributed toCompaq’s decision to start sellingservers to Internet serviceproviders,whichbecamea$2billionannualbusiness for thecompany.HoweverbeneficialforCompaq,whichHPboughtin2002,theserverdecisionwasnotassuccessfulforoneofthemeeting’sparticipants,SeanO’Sullivan,whowentontostartalessthansuccessfulfirmsellingfilestorageandvideo-on-demandtoindividualcustomers. It was just too early for this cloud to rain dollars, even on innovators withforesight.Thegenuinegrowthofthecloudawaitedtheexpansionincomputerprocessingpower and in telecommunications networks, as well as a general economic recoveryfollowing the dot-com collapse of the early 2000s. It was not until 2006 that the termcloud computing came into more general use as companies, led by Google, Dell, andAmazon,startedusingthetermtodescribeanewsystemforaccessingfiles,software,andcomputerpowerover theInternet insteadof fromacomputer’sownharddriveorsomeotherportablestoragemechanism(Regalado2011).

DefiningCloudComputingTherearethosewhobelievethatthefirstuseoftheterminthetwenty-firstcenturywasbyEricSchmidt,Google’sCEO,whenhedescribedthecloudatanAugust9,2006,industryconference: “What’s interesting [now] is that there is an emergent newmodel. I don’tthinkpeoplehave reallyunderstoodhowbig thisopportunity really is. It startswith thepremise that the data services and architecture should be on servers. We call it cloudcomputing—theyshouldbeina‘cloudsomewhere.’”ThePCmakerDellsawmarketingvalue in the term, and in 2008 the company tried to secure a trademark for “cloudcomputing.”Thatattempt,whichupsetmanyintheindustry,ultimatelyfailed.Asaresult,anyonewasfreetousethetermandmanycompaniesdecidedthatthecloudwasagreatwaytocapturethenextstageinthedevelopmentofonlineservices(Regalado2011).

There is no generally accepted definition of cloud computing. Indeed, one overviewsuggeststhattwenty-fivecloudpunditswouldlikelydefineitintwenty-fivedifferentways(McFedries 2012). An entrepreneur who teaches programmers how to use the clouddescribesitas“ametaphorfortheInternet.It’sarebrandingoftheInternet.Thatiswhythere is a raging debate. By virtue of being a metaphor, it’s open to differentinterpretations.”But the debate continues because “it’sworthmoney” (Regalado 2011).Most cloud analysts do not equate the Internet with cloud computing. Although cloudsystems use the network of networks we know as the Internet to transmit data andapplications,theyalsomakeuseofprivatenetworksthatmaybelinkedtotheInternetbutare separate from it and accessible to only a fraction of users. Moreover, since cloudcomputing also involves the customized provision of applications and services, it isgenerally considered to bemore than a network of networks. Although the cloud as adefiningconceptmayeventuallywithdrawintothepowerfulbanalityoftechnologieslikeelectricity,most agree that it has not yet reached the sweet spot of generic universality(Linthicum2013e).

Asof2013,yearsaftercloudcomputingbegantocirculateinpublicdiscourseandwellafter the first mass advertising, including two commercials that aired during the 2011SuperBowl,Americansremainedunclearaboutwhatitmeans.Asurveyof1,000adultscarriedoutinAugust2012suggestedthatfewpeoplehadevenaroughideaofwhatcloudcomputingmeans. Nevertheless,most indicated that they expect to beworking “in thecloud”inthefutureand,whentheyhaditexplained,demonstratedsavvyinunderstandingitspotentialproblems—primarilyprice,security,andprivacy(Forbes2012).WhentheU.S.governmentdecidedthatcloudcomputingmightbeacost-effectiveway

todeliverservices,itpusheddepartmentstoconsideramovetothecloud.However,whendepartmentheadsexpressedlittleknowledgeofcloudcomputing,thegovernment’schiefinformationofficer asked theNational Institute ofStandards andTechnology (NIST) tocomeupwithadefinitionanddescription(Regalado2011).So theclosestwehave toagenerally accepted formal definition is, in the words of a NIST report, “a model forenabling ubiquitous, convenient, on demand network access to a shared pool ofconfigurable computing resources (e.g., networks, servers, storage, applications, andservices)thatcanberapidlyprovisionedandreleasedwithminimalmanagementeffortorserviceproviderinteraction”(MellandGrance2011).Toputitinplainerlanguage,cloudcomputing involves the storage, processing, and distribution of data, applications, and

servicesforindividualsandorganizations.Itisgenerallyviewedasthefastest-growing,ornearthefastest-growing,segmentoftheITsector,eventhoughin2012itrepresentedonly3 percent of all IT spending (Butler 2012b).NIST’s definition of cloud computing hasbeenwidelyaccepted throughout the industryasanobjectivedescriptionof theservice.Butitisimportanttounderstandthatcloud-computingdescriptions,howeverobjectiveinappearance,aretypicallyconflatedwithpromotion.Whetheritisthefederalgovernment’schief information officer, NIST, or the National Science Foundation, which in 2012announceditsowncommitmenttofundcloud-computingresearch,thegoalistopromotethecloudandnotjusttounderstandit.Soalongwiththecleardefinition,NISTproclaims,“TheCloudComputingmodeloffersthepromiseofmassivecostsavingscombinedwithincreasedITagility.Itisconsideredcriticalthatgovernmentandindustrybeginadoptionofthistechnologyinresponsetodifficulteconomicconstraints”(NIST2013).

TheEarlyCloud:TheComputerUtilityandVideotexTodeepenunderstandingofwhatcloudcomputingmeans,itisusefultoconsiderhowitisbothanextensionofearlierformsofcomputercommunicationand,atleastinscale,anewdevelopment in the use of information technology. In the 1950s, the computer scientistHerbGroschforecastaworldthatwouldsharecomputingresourcessothatnomorethanfifteendatacenterswouldbeneededtomeettheworld’sinformationneeds.Inthe1960s,the concept of the computer utility emerged when Stanford IT expert John McCarthyimagined“computationasapublicutility”(C.Ross2012).Thiswasformalizedin1966with the publication of Douglas Parkhill’s widely read book The Challenge of theComputerUtility.Why is ituseful to thinkofcloudcomputingasautility? Inpart it isbecausesomespecialistssee thecloudas littlemore thananextensionof thecomputer-utility concept, once referred to as “time-sharing,” because usage time on a centralcomputer was shared bymultiple users. For example, according to Linthicum, “If youthinkyou’veseenthismoviebefore,youareright.Cloudcomputingisbasedonthetime-sharingmodelwe leveragedyears agobeforewe could affordour own computers.Theidea is tosharecomputingpoweramongmanycompaniesandpeople, therebyreducingthecostofthatcomputingpowertothosewholeverageit.Thevalueoftimeshareandthecorevalueofcloudcomputingareprettymuchthesame,onlytheresourcesthesedaysaremuchbetterandmorecosteffective”(citedinMcKendrick2013a).

Most people are familiar with public utilities for resources like roads, water, andelectricity,whichprovideservicestothepublicoveraninfrastructurethatutilitiesmanageandoperate.Theycanbeownedbygovernmentorbyprivateenterprisebutwhenitisthelatter, utilities are typically subject to some formof local (city, community) or regional(state, county, province) regulation. Without entering the dense thicket of debate overwhether they provide a net public benefit over a competitive market arrangement orwhether the government-ownedor private utility is best, it is sufficient to state that theutilityarrangement is typicallychosenbecause it isexpensive tobuild the infrastructurefor water and power. When governments conclude that duplicating infrastructure sonumerous competitors can enter the market will likely waste resources, they declare a“naturalmonopoly”andestablishapublicutility.

As the concepts associated with computer technology, among them cybernetics,informationprocessing,andcommunicationflows,attractedtheattentionofawidercircleofscholarsandpolicymakersinthe1950sand’60s,somebegantothinkofinformationas a resource not unlikewater and power. The shift from analog to digitalmethods ofprocessinginformationprovidedatangibleormaterialoutputthatmadeiteasiertothinkof information in resource terms. The mathematicians Claude Shannon and WarrenWeaver (1949) built awidely acceptedmodel of communication flows that emphasizedthematerialityof communicationover theabstract sendersand receivers throughwhichcommunicationflowed.Theywerelessconcernedwiththesocialforcesthatmadesomepeople senders and some receivers than theywerewith identifying communication as atangible flow.When the economists Dallas Smythe and Herbert Schiller began to turntheir attention tocommunication in the1950sand ’60s, theydrewconnectionsbetweentheir new field of study and the resources, like agriculture and oil, that had occupiedeconomistsformanyyears(Mosco2009,82–89).Aroundthistimethecomputerscientistturnedpublic-policyanalystAnthonyOettingerdevelopedageneralresourcetheorythat

1.

2.

3.

4.

5.

linkedenergyandmaterials to information,anditbecametheconceptual foundationforthe Harvard University Program on Information Resources Policy, which Oettingerchaired for several decades. When the communication scholar Marc Uri Porat (1977)publishedhisinfluentialmapoftheshifttoaneconomypoweredbyinformationworkers,itbecametimetothinkaboutaninformationeconomy.

These developments gave renewed force to a view that had been debated since theemergence of postal communication and extended to electronic communicationtechnologies, starting with the telegraph and repeated with the telephone, radio, andtelevision.Isitappropriateandusefultoemploytheconceptofaresourcetoidentifytheproduct of these devices and, if so, should this resource be organized in the form of autility?Overtheyears,differentconstellationsofpoliticalforcesproduceddifferentpolicyresponsestothesequestions.Butwiththefoundationofthinking,forexample,abouttheprovisionof telephoneserviceasa“naturalmonopoly,”expertsexaminingtheoutputofcomputer technologybegan towonderwhether the resourcespropelling the informationeconomywerecreatingtheneedforanewutility.

Advancingthisdiscussionofhowtoorganizeinformationresources,DouglasParkhillwroteaboutthechallengesfacingwhatheforesawasthecomingcomputerutility.FromthestartParkhillrecognizedthattheideaoforganizingcomputersystemsasautilitywasintheair:“Evennowthesubjectofcomputerutilitiesisverymuchinthepubliceye,asevidencedbymanyarticles inboth thepopularand technicalpress,prognosticationsbyleading industrial and scientific figures and growing signs of interest on the part ofgovernmentseverywhere”(1966,v).Parkhilltookthispopularideaandgaveitthecleardefinition and specificity required to move it forward. For him, there were five keycomponentstothecomputerorinformationutility:

Essentiallysimultaneoususeofthesystembymanyremoteusers

Concurrentrunningofmultipleprograms

Availability of at least the same range of facilities and capabilities at the remotestationsastheuserwouldexpectfromaprivatecomputer

Asystemofpricingbaseduponaflatservicechargeandavariablechargebasedonusage

Capacityforindefinitegrowth,sothatasthecustomerloadincreases,thesystemcanbeexpandedwithoutlimitbyvariousmeans

Parkhillenvisionedthecomputerutilitytobeapublicserviceinthesensethatitwouldmakeavailable to anyone,wherever located, awide rangeof information resources andservicesinanonlineform.Withthatsaid,hedidnotmakeacommitmenttoanyspecificmanagementform,butratheraddressedthemeritsofpublic,private,andmixedsystemsbecause“itisnecessarytoconsidereachapplicationofcomputerutilityseparatelyon itsmeritsandbalanceoffineachcasethegainsandlossesresultingfromtheadoptionoftheutility concept” (1966, 125). Elements changed as yesterday’s computer utility becametoday’scloud-computingsystem,butitisworthwhiletoreflectonhowmuchofParkhill’sthoughtisrepeatedintoday’sdiscussionsofcloudservices.Wearenowmorelikelytoaskifasystemisscalableratherthanifithasthe“capacityforinfinitegrowth,”butnewtermsshould not mask the striking conceptual similarities. Parker would go on to play an

importantroleinimplementinghisvisionofthecomputerutilitythroughthecreationofwhatbore thediscernibleyetoddnameofvideotex.Thiswas a computer-based servicethatdelivered informationfromacentral facility tousersat terminals in theirhomes, inpublicplaces,and,toalesserdegree, inbusinesses.Userswereabletointeractwiththeservice by making specific information requests. Parker helped bring about the mostadvancedof these systems inaCanadiangovernment-sponsoredprojectnamedTelidon.Becauseitsuseofcolorimagesanditsprocessingdemandsoutstrippedthecapacityoftheexisting telecommunicationsnetwork, the systemdidnotadvance faroutof thestartinggate. Nevertheless, simpler systems featuring more manageable services were widelydistributed. The best known of these, France’s Minitel service, brought terminals tolibraries,postoffices,andotherpublicplaces,providinguserswithbasicinformationlikethetelephonedirectory,trainschedules,informationongovernmentservices,stockquotes,and the opportunity to chat with fellow users and have messages delivered to a “mailbox.”Theserviceprovidedmillionsofconnectionseachmonthandwasnotretireduntil2012 (Sayare 2012).Videotex held great promise as report after report predictedmajortransformationsineveryaspectoflife,withcomparisonsmadetotheautomobileandthetelevision(Tydemanetal.1982).

Videotex was only one of many cloudlike services that emerged in the pre-Internetdecades.Infact,whatisveryinterestingtoobserve,andoftenlostinthelinearhistoriesthat see the past as simple precursor to the present, are the vast arrays of differentapplications that arose under the resource/utility umbrella. Consider the atlas of cloudsrepresentedbytheSovietUnion’scyberneticsystemsofthe1960s,Chile’sexperimenttobringaboutcomputerizedworkplacedemocracyandeconomicplanninginthe1970s,andthe Pentagon’s development of a research computer network that helped to create theInternetfromthe1970stotheearly1990s.

CyberneticsintheSovietUnionInspiteofWorldWar II’sdevastating impact, theSovietUnionproduced leaders in theburgeoning field of cybernetics, formally the science of communication and control inmachinesandanimals.IntheWest,thecomputerscientistNorbertWienerledthefieldofluminaries, with a stellar group that in 1953 included John von Neumann, ClaudeShannon, William Ross Ashby, Gregory Bateson, and Roman Jakobson, who metregularlyundertheauspicesoftheMacyFoundationfrom1946to1953.Rebellingagainstestablished approaches to theory and applied science, they transformed establisheddisciplinesandhelpedtocreatenewones.Littlewasleftuntouchedinfieldsasdiverseasbiology, communication studies, computer science, linguistics, andpsychology. ItmightonlybethegentlestofoverstatementstoconcludethatcyberneticsbecameaHolyGrailofgeneraltheorythatmanybelievedwouldrevolutionizehumanthought(Parkman1972).

These ideas slowly simmered inSoviet science, permitting quiet questioning of rigidtheoryenshrinedintheworkofTrofimLysenkoinbiologyandIvanPavlovinpsychologywhile Joseph Stalin retained his iron grip on power. But when Nikita KrushchevconsolidatedhiscontrolasPremier in1958,changeacceleratedand thecybernetics thathadbeenofficiallydenouncedas“notonlyanideologicalweaponofimperialistreactionbutalsoatoolforaccomplishingitsaggressivemilitaryplans”wasby1961hailedastheprimarytechnicalmeanstorealizetheCommunistideal(Gerovitch2010).InthatyeartheSoviet Academy of Sciences published Cybernetics in the Service of Communism, adetailed examination of how cybernetics would transform practically every field ofknowledgeandapplication,butespecially, to thepleasureof therepresentativesmeetingthat year in the Twenty-Second Congress of the Communist Party, the modern Sovieteconomy.

For its supporters, economic cybernetics would demonstrate the superiority of theSovietsystembyapplyingthenewsciencetothenewtechnologyofpowerfulcomputerstopreciselyplanfortheproductionanddistributionofgoodsandservicesthroughouttheSovietUnion. In1962 thechairmanof theU.S.S.R.’sAcademyCouncilonCyberneticsmade the importance of the marriage between cybernetics and economic planningabsolutely clear when he declared that “However unusual this may sound to someconservatives who do not wish to comprehend elementary truths,we will be buildingcommunism on the basis of the most broad use of electronic machines, capable ofprocessingenormousamountsof technological,economic,andbiological information inthe shortest time. These machines, aptly called ‘cybernetic machines’, will solve theproblem of continuous optimal planning and control” (ibid.). In effect, these wordsannouncedthebirthoftheSovietcomputerutility.Anetworkofcomputercenterswouldbe built across the vast expanse of theU.S.S.R., throughwhich a continuous streamofdatawouldflowfromshops,factories,andoffices.Plannerswouldusethedatatoassessthesuccessorfailureofpoliciesandtoplan, inthemostminutedetail,futureeconomicactivity. Regional computer centers would link up in a nationwide network under theauspices of the Central EconomicMathematical Institute, giving the country “a singleautomatedsystemofcontrolofthenationaleconomy”(ibid.;Spufford2010).Thiswasaplanforstate-directedcloudcomputingin theserviceofcentraleconomicplanning,andU.S. intelligenceservices—alreadyworriedabout thegrowthofSovietmilitarymight—fearedwhatmightresult.

TheCIArespondedin1962bysettingupaspecialunittostudythethreatposedbytheSovietcyberneticsinitiative.Oneofthemostremarkableconclusionsdrawnfromthespyagency’sinvestigationwastheexpectation,andconsequentuneasewiththeidea,thattheSoviet plan would actually succeed. According to its task force report, “tremendousincrements in economic productivity as the result of cybernetization of productionmaypermitdisruptionofworldmarkets”(Gerovitch2010).TheCIAconcludedthateconomicsuccesswouldbringanadditionalthreat:“Thecreationofamodelsocietyandthesocio-economic demoralization of theWestwill be the added ideologicalweapon” (ibid.). SoconcernedwastheintelligenceagencythatitcontinuedtodiscusstheissuewithKennedyadministrationofficialsintheperiodleadinguptoandthroughoutthe1962CubanMissileCrisis. The president’s people were equally worried. In a memo to Attorney GeneralRobertKennedy,Arthur Schlesinger Jr., historian and special assistant to the president,concludedthatthe“all-outSovietcommitmenttocybernetics”wouldgivetheSoviets“atremendousadvantage”andthat“by1970theUSSRmayhavearadicallynewproductiontechnology, involving total enterprises or complexes of industries, managed by closed-loop, feedback control employing self-teaching computers.” Pulling no punches, heconcluded that if theUnited States continued to neglect cybernetics, “we are finished”(ibid.).

Evendiscountingforthehyperbolethatoftenaccompaniestheefforttoconvincethosein power to take action, Schlesinger’s statement and those of the CIA amount to adeclarationthat theSoviets’earlyversionof thecloud,with itscentralplanningthroughcybernetics, would work and might very well defeat the United States. The furorcontinued as President Kennedy set up a task force to examine the threat of SovietcyberneticsandtheCIAcontinuedtosoundthealarm.TheU.S.militarygotintotheact,too,withthecommanderoftheAirForceForeignTechnologydivisionalarmedthat“thesystemcouldbe imposeduponus fromanauthoritarian,centralized,cybernated,world-powerfulcommandandcontrolcenterinMoscow”(ibid.).

AswithmanyU.S. assessments of theSoviet threat, these fears proved exaggerated.Only a small fraction of the Soviet programwas implemented because the governmentdivertedavailableresourcestothemilitary,whichsteadfastlyrefusedtosharethemwithwhat top commanders believedwas the useless project of the economic cyberneticians.Thisclouddidnotvaporizeovernight,however.TheSovietUnion’scyberneticsteamwasable to patch together a semblance of a computer system for planning and allocatingresources,producinglessthanarobustnetwork,moremistthancloud.Moreover,ittookanationalnetworkofhuman“fixers”whosejobitwastousewhatevermeansnecessarytokeepchainsofproductionanddistributionworking,or,atleast,keepthemfromseizingupentirely, so that the façade of central planning through cybernetics and what FrancisSpufford(2010)calledthebeliefin“RedPlenty”couldbemaintained.

TheSovietUnion’sdalliancewith anearlyversionof cloudcomputingdemonstratedboththepotentialandthepitfallsofusingitfornationaleconomicplanning.Mostanalystshave understandably focused on negative lessons, including some combination of theinherent difficulty of developing a cloudmodel for amassively complex economy, thestructuralproblemsbuiltintotheSovietsystem,andtherecognitionthatcomputerswerenot nearly advanced enough to carry the load.Scholars are just beginning to assess theactual potential of the Soviet cybernetics program to meet the government’s economic

goals (Dyer-Witheford2013). Itwould also be interesting to consider the impact of thecybernetics program on the ultimate opening of Soviet life.We know that it permittedscientistsandintellectuals toconsideralternatives toStalinistabsolutes.Perhapsifmorethan one generation had continued towork on the program, cybernetic planningmighthavenudgedopenmoredoorsintheSovietUnion.Wedoknowthatonealternativeearlycomputer utility or cloud experiment, Chile’s Project Cybersyn, was influenced by theSovietcyberneticsproject,but itdepartedfromtheSovietproject insignificantwaysaswell.

TheComputerUtilityComestoChile(Almost)After the people of Chile elected Salvador Allende to the presidency in 1970, heproceeded to carryout socialdemocratic reforms that included increasing theminimumwage and expanding education, public housing, and food programs for the poor.Morecontroversial was the government’s decision to nationalize Chile’s lucrative copperindustry, which had been largely under the control of U.S.-based multinationalcorporations. In 1973, with the assent and support of the United States, the ChileanmilitaryoverthrewAllendeinacoupresultinginthousandsofdeathsandimprisonments.Themilitaryruledforthenextfifteenyears.

DuringAllende’spresidencyandwith theassistanceofanAmericancomputerexpertStaffordBeer,Chile experimentedwith computer-assisted economicplanning.Arguablythefirstofthecyberneticianstoachievebusinesssuccess,BeerwasdubbedbynoneotherthanNorbertWienerhimselfas“thefatherofmanagementcybernetics”(Miller2002,3).SoonafterAllende’selection,BeeracceptedtheinvitationofFernandoFlores,anengineerworking in the Chilean State Development Corporation, to establish Project Cybersyn(ProyectoSyncoinSpanish),aprogramtobuildacomputercommunicationsnetworkthatwouldhelpruntheChileaneconomy.LiketheSovietsystem,itwouldprocess,organize,and display information on economic activity in real time. But unlike the U.S.S.R.’ssystem, Cybersyn would use the information to enable workers and local managers toparticipate by providing information and making decisions. Specifically, the project’sdevelopersplannedtohaveworkersparticipateinthedevelopmentofproductionmodels,inthedesignandimplementationoftechnology,andineconomicmanagementatthelocalandnationallevels(Medina2011).

Inthe1970stheconceptofworkerdemocracywaspopularasameansoftappingintothe tacit knowledge of skilled workers; as one way to combat what was viewed aspervasive workplace alienation, especially among young workers; and as a means ofextendingparticipationfromtheelectoralarenaintothemodernworkplace.Experimentsinworkplace democracy andworker controlwere takingplace at the time in numerouslocations, including prominently in the United States, Israel, and in what was thenYugoslavia(Hunnius,Garson,andCase1973).Withworkerdemocracyintheair,expertsin the new technology of computer communication thought about how to apply theirtechnical skills to what was becoming a global movement. As Beer said in 1972, “InChile,IknowthatIammakingthemaximumefforttowardsthedevolutionofpower.Thegovernmentmadetheirrevolutionaboutit;Ifinditgoodcybernetics”(Medina2011,3).Allende and his government agreed that cybernetics would enable them to build acomputersystemthatwouldhelp“tocreateanewpoliticalandtechnologicalreality…,one that broke with the strategic ambitions of both the United States and the SovietUnion”(ibid.,3).

LimitedcomputerresourcesandtheshortlifespanoftheAllendegovernmentdidnotpermit implementation of Project Cybersyn, but it remains important in the history ofcloudcomputingforseveralreasons.ItdemonstratedthatthehistoryofthecloudcontainsanimportantchapterfromoutsidetheUnitedStates,theSovietUnion,andothercentersof world power. Audacious as it was, Project Cybersyn was proposed and designedprimarilybyengineersandplannersinwhatwasthencalledathird-worldcountry—inthe

mindsofsome,abackwardnationthatshouldhavebeenconcentratingonminingcopperfor transnationalcorporations insteadofexperimentingwithcomputer-assistedplanning.Moreover, Cybersyn was consciously designed as an alternative to standard models ofeconomicdevelopmentonofferfromtheUnitedStatesandtheSovietUnion.Beersoughtabalancebetweencentralizedanddecentralizedcontrol,andbetweentheoverallneedsofafirmandtheautonomyofitscomponentparts.Hisworktappedintoalineofthinkingthathasfounditswayintodiscussionsofthecloud.Howcanwecreatecomputersystemsthat bring about efficiencies through centralizationwithout sacrificing local autonomy?Willbigdatainthecloudfacilitatedemocracyoroverwhelmit?Beer’sthinkinglinedupwell with the Popular Unity government’s interest in promoting national developmentwithoutsacrificingcivilliberties,afreeandopenmedia,andindividualautonomy.Finally,the proposal for the Chilean version of a computer utility demonstrates the need toconsider the social relations of technology in any discussion of cloud computing. ForChile, the Cybersyn network was important because it would advance nationaldevelopment,butalsobecause itwouldpromotepublicparticipation in thepoliticalandeconomiclifeofthenation.Toovaluabletobekeptunderprivatecontrol,itwouldservesocietyasawhole.

It is easy to question whether Allende’s government moved too fast to nationalizeresourceindustriesandpromoteworkplacedemocracywithnewinformationtechnology.Or perhaps it proceeded too slowly, because the government refused to arm supportersundermilitantattackfromU.S.-backedsectorsofthesociety.ItisalsoeasytobrandBeerasaneccentricwhogotinoverhisheadinaplacehedidnotunderstand.Butbeforedoingso,itisworthwhiletocompareChile’sambitiousplanstouseanewtechnologytobringaboutathoroughdemocratizationofsocietywithtwoexamplesfromthepoliticalusesoftoday’scloud.Thefirst isgenerallyviewedasanunalloyedsuccessbecauseit iswidelyseenasamajorcontributortoreturningBarackObamatotheWhiteHouse.Iamreferringtohiscampaign’suseofcloudcomputingandbig-dataanalysisprovidedbyAmazonWebServices (AWS), a division of the online retail giant, to identify potential voters andsuccessfully deliver enough of them to the polls to exceedmany pundits’ expectations.Thecampaignbuiltmorethan200appsthatraninAWS,makingsuchheavyusethatthecompany’schieftechnologyofficertweetedhispersonalcongratulationstohiscounterpartin the Obama campaign once victory was certain. The campaign utilized the Amazoncloudinmanyways,buttheskilleddeploymentofdatabasesinmodeling,analytics,andintegrationwaskey.Specifically,“Thisarrayofdatabasesallowedcampaignworkers totargetandsegmentprospectivevoters,shiftmarketingresourcesbasedonnearreal-timefeedbackon the effectivenessof certain ads, anddrive a donation system that collectedoveronebilliondollars(makingitthe30thlargestecommercesiteintheworld)”(Cohen2012). Another key was a set of tools that helped the campaign determine the mostefficient television advertising buys (dubbed the Optimizer) and targeted messages toTwitterandFacebookusers(calledblasters)(Hoover2012).

There is nothing especially unusual about these and other strategies in the Obamacampaign’spartnershipwithAmazon.Itappearsthatthecampaignsimplymadebetteruseof its data-management resources thandid theopposition.What is striking, however, ishow little this has to do with practicing democracy, with civic participation, or withactivism at any level. In place of democracy, including anything envisioned in the

Cybersynproject,wehavepopulationmanagementandcontrol.

ThesecondexamplecomesfromGreatBritain,wherePrimeMinisterDavidCameron,abigfanoftheiPadandespeciallythegameFruitNinja,orderedthecreationofanappthatwouldenablehimandhisinnercircletomonitortheBritisheconomy.DubbedNo.10Dashboard, according to the website of the government’s cabinet office, it provides asummary view of national and international information, including housing andemployment data and stock prices, as well as data on the performance of governmentdepartments.Inaddition,thereis“politicalcontext”datadrawnfrompolls,commentary,andasamplingfromTwitter.Proudoftheapp,theprimeministershoweditofftonewlyreelectedPresidentObamaataG8summitmeeting.

Itwould be easy to draw the conclusion thatwithObama’s use of the largest cloud-computingcompanyandCameron’sNo.10Dashboard,wearenowlightyearsaheadofChile’sCybersyn.Afterall,roomsfullof1970sequipmentandsoftwarecannowfitonahandhelddevice.Butoncloser inspection,somethingsubstantialhasalsobeenlost.ThefruitsofCybersynweretobesharedwiththeentirenationinatransparentprocessofdataproduction, modeling, display, and distribution. The goal was to advance the Chileannational economy even as it promoted democracy in the workplace and in society.Cameron’s app, like Obama’s use of AWS, is intended to better manage a population.Neitherhasmuchtodowithpublicparticipationinpoliticaldecisionmaking.Respondingto just this type of criticism, the data director of Obama’s campaign felt compelled todeclare,“IamnotBigBrother.”Heinsistedthat“campaignsdon’tknowanymoreaboutyour online behavior than any retailer, news outlet or savvy blogger” (Roeder 2012).Althoughitismorethanabitdisingenuoustocompareacampaignorganizationthatspentover $11 million on technology services with the resources of a savvy blogger, it isaccurate to comparewhat both campaigns knewabout online andoffline behaviorwithwhatWalmart, Target, or any other large, global retailer knows (Gallagher 2012). ButwhatkindofdefenseisittomaintainthatapresidentialcampaignisnoworsethanagiantretailerlikeWalmartwhenitcomestosurveillance?Obama’sdatadirectormaynotbeBigBrother,butdoesthisjustifytheconclusionthat“newtechnologiesandanabundanceofdatamayrattlethesenses,buttheyarealsobringingafreshappreciationofthevalueofthe individual toAmericanpolitics” (Roeder 2012)?Whatwouldwe think if this camefromthedatadirectorofTargetonlywith“theAmericaneconomy”replacing“Americanpolitics”? The same holds for No. 10 Dashboard. Indeed, as one commentator noted,Cameron’s“appcould…beanaptmetaphorforpoliticiansreducedtospectatorsbythesurgesandshocksoftheglobalizedworld”(Wiles2012).Itdoesnotreallyempowertheinnercircleofpeople forwhom itwasmade. In that respect, it isnotdissimilar fromaspecial-purpose iPadappmade for the team responsible for restructuringGreece’sdebt.Butthisconclusionmissesamoreimportantpoint.Politicianswhobuildappsthattakeasnapshotoftheeconomymayormaynotbepowerlesstodoanything.Butthereislittle,ifany,considerationforhowsuchdatamightempowercitizens,norforhowcitizensmightparticipateinitscreationasworkers,voters,orcustomers.Thatiswhyitisimportanttorevisittheprecursorsofcloudcomputing,likeProjectCybersyn,whatevertheiroutcomes.Moreover,weneedtodomorethanmarvelattheadvanceintechnologyoverthedecadesbecause history suggests that technological progress does not necessarily bring aboutadvancesinthepracticeofdemocracy,andsometimescanresultingenuineregression.

ThePentagonandtheInternetAlthough they left behind important legacies and lessons, videotex, Soviet cybernetics,andProjectCybersynarenolongeraround.TheworkoftheDefenseAdvancedResearchProjects Agency (DARPA), on the other hand, is not only important for understandingwhere cloud computing comes from; it is a significant participant in current militarycloud-computing projects.When theSovietUnion successfully placedSputnik, the firstoperationalsatellite,intoorbitaroundtheearthin1957,itcaughttheU.S.governmentbysuchsurprisethatPresidentEisenhowercreatedanagencywithinthePentagonwhosejobitwastokeepthesesurprisesfromhappeningagain.

Starting in 1958 the agency, then known asARPA,was responsible for carrying outresearchanddevelopmentonprojectsatthecuttingedgeofscienceandtechnology.Whilethesetypicallydealtwithnationalsecurity–relatedmatters,theagencyneverfeltboundbymilitary projects alone. One outcome of this view was significant work on generalinformationtechnologyandcomputersystems,startingwithpioneeringresearchonwhatwascalledtime-sharing.Thefirstcomputersworkedonaoneuser–onesystemprinciple,butbecauseindividualsusecomputers intermittently, thiswastedresources.Researchonbatch processing helped tomake computersmore efficient because it permitted jobs toqueue up over time and thereby shrunk nonusage time. Time-sharing expanded this byenablingmultipleuserstoworkonthesamesystematthesametime.DARPAkick-startedtime-sharingwithagranttofundanMIT-basedprojectthat,undertheleadershipofJ.C.R.Licklider,broughttogetherpeoplefromBellLabs,GeneralElectric,andMIT(Waldrop2002).Withtime-sharingwasborntheprincipleofonesystemservingmultipleusers,oneof the foundations of cloud computing. The thirty or so companies that sold access totime-sharingcomputers,includingsuchbignamesasIBMandGeneralElectric,thrivedinthe 1960s and 1970s. The primary operating system for time-sharing wasMultics (forMultiplexed Information and Computing Service), which was designed to operate as acomputerutilitymodeledaftertelephoneandelectricalutilities.Specifically,hardwareandsoftwarewereorganizedinmodulessothatthesystemcouldgrowbyaddingmoreofeachrequired resource, suchascorememoryanddisk storage.Thismodel forwhatwenowcallscalabilitywouldreturninafarmoresophisticatedformwiththebirthofthecloud-computing concept in the1990s, and thenwith the arrival of cloud systems in thenextdecade.Oneofthekeysimilarities,albeitatamoreprimitivelevel,betweentime-sharingsystemsandcloudcomputingisthattheybothoffercompleteoperatingenvironmentstousers.Time-sharingsystemstypicallyincludedseveralprogramming-languageprocessors,software packages, bulk printing, and storage for files on- and offline. Users typicallyrented terminalsandpaid fees forconnect time, forCPU(centralprocessingunit) time,andfordiskstorage.Thegrowthofthemicroprocessorandthenthepersonalcomputerledto the end of time-sharing as a profitable business because these devices increasinglysubstituted,farmoreconveniently,fortheworkperformedbycompaniesthatsoldaccesstomainframecomputers.

DARPA is even better known for the creation of the Advanced Research ProjectsAgency Network—ARPANET—the first wide-area network using packet-switchingtechnology.Packetswitchingbreaksupdataintoblocksorpackets,whichseekthemostefficientnetwork routing.Theblocksare reassembledat theendpointand,unless thereare major network problems, appear to an end user as a unified data, voice, or video

transmission. The network was created to link secure military installations and majorresearchfacilitiesandbecameadirectprecursoroftoday’sInternet.Infact,somedatethebirthoftheInternettoJanuary1,1983,whenforonedayARPANETcompletelyshutoffservicetothe400hoststhesystemservedinordertoreplacetheNCPprotocolwiththeTCP/IP network protocol that has defined the Internet ever since (Kerner 2013). Thegrowthof the Internet released the brake on cloud computing that the expansion of thefirstmicrocomputers and then personal computers had applied. In addition to requiringsignificant expansion of distribution capacity in wireline, wireless, and switchingcapabilities, the Internet’sacceleratingdemand fordata storageandprocessinghastenedthearrivalofcloudsystems.

Theprecursorsofcloudcomputingdemonstratethatwhatwenowcallthecloudcamefromvariousplaces thatusedcomputing fordifferentgoals.Videotex systemsaimed tolink terminals and television receivers to remote computers that, in practice, providedbasicinformationtopeopleinahandfulofnations.TheSovietUnionapplieditsleadingrole incybernetics todevelopanational systemofeconomicplanning.Notwithstandingstrong fears in theKennedy administration, including theCIA, that the programwouldenabletheSovieteconomytoovertakeitscompetitorsintheWest,itwasatbestapartialsuccess.ItfellvictimtothelimitedcapacityofcomputersystemsandtothepoweroftheSoviet military, which resisted investing technology resources to build the domesticeconomy.Chile’sCybersynsoughttobringaboutasocial-democraticversionofnationaldevelopment planning by connecting central computers to terminals throughout thecountry, primarily to establish an interactive system of economic decisionmaking. Theshort-lived rule of the Popular Unity government of Salvador Allende meant thatCybersyn never made it out of the planning phase. Nevertheless, it demonstrated thatcloudcomputinghashistoricallinkstotheGlobalSouth,wheredemocraticvaluesexistedside-by-side with technical visions. Finally, DARPA made use of big military budgetsduring the Cold War to help bring about time-sharing and the Internet. Perhaps mostimportantly,unliketheSovietmilitary,whichwashostile tocivilian-sectorparticipation,DARPAworkedwithcorporationsthatdevelopedbusinessapplicationsthateventuallyledtocloudcomputing.DARPAcontinuestobeveryactiveinthedevelopmentofamilitarycloud.

AnatomyoftheCloudToday’s cloud computing deepens and extends key tendencies established by these andother predecessors. The rise of data centers controlled by a handful of companiescontinues a process of creating global networks of informational capitalism (Schiller2014).CompaniesthatoncecontainedanITdepartment,withitscrafttradition,cannowmovetothecloud,whereITanditslaborarecentralizedandstreamlinedinanindustrialmode of production, processing, distribution, and storage. Furthermore, the cloud takesone more step in a long process of building a global culture of knowing in whichinformation production accelerates through networks that connect data centers, devices,organizations, and individuals.The cloudmakes up both a new industrial infrastructureandacultureofknowing,basedondigitalpositivism.

It is easy to lose sight of the significance of cloud computing for informationalcapitalismandforbuildingacultureofknowingbecausetimeandtimeagainintheearlyyears of a technology, there is a tendency to concentrate on those flashy utopian ordystopian visions that make up what has been called the technological sublime (Nye1994).This isunderstandable. Justas itwashard to resist the feelingofmagic the firsttimeawebpagescrolleddownahomecomputerscreen,sotoowasitmagicalwhen,forthefirst time,street lightsbrightened thenightwithelectricity’s illuminationandvoicesemanatedfromthemusicalboxthatcametobecalledradio.Cloudcomputingcurrentlyresides in thismagical sublimephasewhere transcendentvisionsofending space, time,andsocialdivisionstendtodiluteourappreciationofthemoregrounded,long-term,butbanal consequences of implementing cloud systems. The experience with electricity isespeciallyrelevantbecauseitsearlydayswerefocusedonthecapacitytobringlightandpower, an admittedly significant, if not revolutionary, development. But electricity’ssublimeallureworeoffwhenpeoplegotused touniversal lighting,especiallywhen thepromisedendtocrimeonthestreetsdidnotpanout.Thesublimebecamebanal.Butthegenuine revolutionarypowerofelectricityawaited itswithdrawal into thewoodworkofbanality.Itwasnotuntilelectricalgenerationwasorganizedintoutilitiesandsentouttopower industrial and household applications (yesterday’s apps) that one could safelyconclude that electrification was a principal participant in an economic and socialtransformation.Frompoweringautomobileassemblylinestoturningonvacuumcleaners,electricity’smanyapplicationswerenotterriblysublime,butcertainlyweretransformative(Nye 1990). Indeed, some economists argue that electrification, including centralizedpower generation and near universal distribution, has been the most significanttechnologicalforceforeconomicgrowthinthemodernera(Gordon2000).

Cloudcomputing ismoving from the sublime stageof infinitepromises towhatmayamounttoasimilarbanality.Inthisrespect,thecloudisagatheringofutilities,certainlynot the same as the electrical-power generators that enabled a leap in the industrialrevolution, but not so different that it is inappropriate to consider a similar process atwork. The sublime cloud is entering a banal phasewhere there is less focus on it as adiscrete entity and more on the transformative applications that it is enabling. As oneanalystputsit,“Inthemid19thCentury,centralisedgenerationallowedelectricitytobeprovided as a utility, meaning that consumers only had to pay for what they used.Consumption could be scaled up or down tomeet demandwithout the need for capitalexpenditure.Acenturyandahalfon,thisispreciselytheemancipatingeffectthatcloud

computing is now having on the enterprise. Organizations no longer need to build,maintainandrenewcumbersomeITinfrastructureinordertoconsumeasmuch,oraslittlecomputingresourceastheyneed”(John2013).

Cloud computing builds on its predecessors, but there are sufficiently significantdifferences thatmark itsdeparture fromearliermodels. It isuseful to consider someofthesedifferences,beginningwith theextraordinarygrowthin thesheersizeandscaleofcloudfacilities.Itisnooverstatementtoarguethatcloudcentersrequireamajorstretchinour conceptual vision to begin to understand their enormity.Consider the plans for thelargest data center (in cost, size, and processing power) now under construction. InSeptember 2012 China’s major social-networking firm Baidu, a Chinese version ofGoogle andFacebook combined, announced that itwould spend $1.6 billion to build anewcloudcenter inYangquan,ShanxiProvince,covering120,000squaremeters(aboutthirtyacres),roughlythesizeoftheU.S.Pentagon,oneofthelargeststandalonebuildingsintheworld.TheYangquanfacilitywillcontainthecapacitytostore4,000petabytes(PB)of data (1 PB equals 1million gigabytes; see the following table).When completed in2016, itwilldeploy700,000centralprocessingunits.Drawingcomparisonsandmakingestimatesisalwaysperilous,butithasbeenestimatedthatdigitizingtheentirecollectionofprint,audio,andvideostoredinthecollectionoftheLibraryofCongresswouldamountto roughly 15 terabytes of data. The storage capacity of the Baidu cloud center wouldthereforeenableittohousethedataequivalentof268,000LibrariesofCongress.

FROMMEGABYTESTOZETTABYTES

1,000megabytes=1gigabyte(GB)

1,000GB=1Terabyte(TB)

1,000TB=1Petabyte(PB)

1,000PB=1Exabyte(EB)

1,000EB=1Zettabyte

Whenitopensin2016,theBaiducenterwillsetanewstandardfordatafacilities,butthoseoperatingnowarefarfromsmall.AsofDecember2013,thelargestexistingclouddata center was a 400,000-square-foot structure, part of a 2.2-million-square-footinterconnectedcollectionofdatacentersoperatedbytheSwitchcorporationinLasVegas,where the absence of natural disasters provides a margin of safety. Admittedly, datacentersofthissizeareat theouteredgeoftypicalclouddatacenters,butthetrendistobuild ever-largeronesbecause sizeprovides efficiencies that areneededasdata storageandprocessingdemandcontinuestogrow.Infact,China,inajointventurewithIBM,isintheprocessofbuildingitsown“cloudcity”inLangfang,anoldindustrialdistrictnearBeijing,thatwillcoverover6millionsquarefeetoffacilities,includingagiantdatacenterandofficestohouseITdevelopmentcompanies(Zhu2013).

The corporationCisco, amajor participant in the cloud industry, has put together anindexofglobaldata-centertraffic.Theseareestimates,buttheyprovideageneralsenseofthegrowthinthesheeramountofdatainthecloud,andonceagainrequireastretchoftheimagination.Ciscoestimates thatby theendof2017,69percentofall Internetprotocol(IP)trafficwillbeprocessedinthecloudasopposedtoinfacilitiesoperatedbyaspecific

organization, likeacorporationorgovernmentunit,orby individualconsumers.AnnualglobalcloudIPtrafficisforecasttoreach5.3zettabytes(ZB)(asinglezettabyteisequaltoonebillionterabytesor,inmoreconcreteterms,250billionstandardDVDsor36millionyearsofHDvideo)bytheendof2017.Globalcloudtrafficisexpectedtogrowsixfoldbythat year (Cisco 2013).This has led some toworry about a cloud “plumbing problem”becausetheamountofdatastoredisgrowingmuchfasterthanthebandwidthofnetworkconnectionsneededtoprocessandanalyzedata(Wegener2013).

Statisticson the industryarenoteasilyobtainedbecauseclouddatacentersareeitherunder private control or operated by governments not inclined to share information.Estimatesvary,butonecensusproducedatotalof509,000datacentersworldwideattheend of 2011, occupying close to 300 million square feet. Cloud centers are locatedeverywhereintheworldbuttendtobeconcentratedinplaceswherelandisplentifulbutnotfarfromcommunicationandpowerfacilities.Thisincludeswhatwasonceagriculturelandontheoutskirtsofpopulationcenters,wherecompaniescanbenefitfromlowlaborcosts.TheseconsiderationshaveledAppletolocateitsclouddatacentersinruralNorthCarolina and in Oregon. The North Carolina location is especially interesting for bothAppleandGooglebecauselowlaborcostsarematchedwithlowenergycosts—30percentlowerinNorthCarolinathanthenationalaverage.Moreover,NorthCarolinapossessesanincreasingly valuable commodity that one would not naturally associate with cloudcomputing:pigmanure,or,asitisreferredtomoreeuphemistically,blackgold.Thestateholds14percentoftheswinepopulationintheUnitedStatesandpigmanurecanproducemethanegasenergy tohelpmeet themassivepower-consumptionneedsofdatacenters.AppleandGooglearenotonlycompetingforclicksandcustomers;theyareinaracetodeterminewhocanbestexploitthisunlikelyNorthCarolinaresource(Wolonick2012).

Securityisagrowingconcern,especiallyasthesizeandthereforethevalueoffacilitiesand data have grown. This has led some cloud companies to locate their facilities inmountainous regions that, while quite far from urban areas, offer added protection.Increasingly, the propensity for earthquakes and severe climate events is taken intoaccountinthechoiceoflocation.Energycostsfora24/7operationareakeyconsiderationand this is leading some cloud companies to explore the novel solution of buryingfacilitiesinsidemountainsclosetosuppliesofcoolwatertolessentherequirementforairconditioning. For example, Norway’s Green Mountain Data Centre is located on theshoresof the islandofRennesøy,close toa largefjord.Thecenter itself iscontained inconcretebuildingswithincavesbuilt into themountain.Racksof servers fill hallsonceused to store ammunition for NATO forces, but what makes the location especiallyattractive is proximity to a fjord that provides a constant supply of coolwater to keepsensitive systems from overheating. Locations like Rennesøy provide both enhancedsecurityandlowerenergycosts.

It is interesting to observe, and not a little bit ironic, that a technology promisingfreedom from locational constraints is itself constrained by the need to maximize theability to house enormous amounts of data andguarantee system reliability.Companiesincreasingly aim for the sweet spot: cold climate, access to low-cost power, abundantwater supply, high-bandwidth Internet connections, political stability, and financialincentives.Severalcountriesmeettherequirements,butnonemoresothanCanada,whichisincreasinglyadata-centerdestinationofchoice(Perkins2013).FacilitiesinCanadatake

advantageofa technologyknownas“freecooling”thatreducesenergyrequirementsbyabouthalfthroughtheuseofacoolingcircuitthatdrawsonoutdoorairtosupplementadatacenter’senergy-intensiveneeds.Aspecializedheatexchangerusesoutdoorairtocoolwater and glycol that circulate to the server racks, thereby reducing the load oncompressorsandpumps,whicharethebigenergyhogsindatacenters.IBMopeneda$90milliondatacenterinasmallOntariocommunitypartlybecausethecompanycancoolthefacility for 210 days a year without running energy-consuming chillers. While exoticlocations likemountainsand fjordsattractattention,Canadaworks formanycompaniesbecause practically the entire country is in a cold climate, which means there arenumerouslocationsnearpowerandwatersuppliesandclosetolargecities.AccordingtotheheadofoneITresearchcompany,“TheadvantageCanadahasisit’sfarcheaperandeasiertobringdatatopowersources,andviceversa.It’smuchcheapertostickyourdatanext to a hydro dam” (Stoller 2012). The town of Barrie, Ontario, which houses theaforementionedIBMfacilityaswellas facilitiesofmajorbanks,hasabundant, reliable,and inexpensive supplies of water and power, and benefits from proximity to Toronto,which provides it with excellent Internet connections. Canadian cloud-data-centercompanieshavealsopioneered theuseofenergy-savingsystems.OVH.com, aQuebec-based company, uses a unique heat-dissipation and cooling system that has completelyeliminatedtheneedforair-conditioningserversinitsCanadianlocations,andreduceditby98percentinitsworldwidelocations.

Canada,liketheScandinaviannationswithwhichitviesfordata-centerbusiness,alsobenefits from political stability and strong data security. Additionally, Canada benefitsfrom proximity to the United States and the additional incentive that data located inCanada isnotsubject to theUSAPATRIOTAct,whichpermits theU.S.government tointerceptandanalyzedatastoredwithinitsborderswithoutasearchwarrant.InadditiontoCanadian and Scandinavian locations, Switzerland, with its long-standing politicalneutrality, isan increasingly favoredchoice fordatacenters,but it isexpensive.All thediscussionofsizeandproximitytoresourcesmakesclearthatcloudcomputingisaverymaterialindustrywithlocationalrequirementsthatbelietheimageofanephemeralcloud.Cloud-computing data centers are the communication version of those industrialtransportationhubsofthepastwhere,forexample,thecityofChicagoplayedalargerolein America’s industrial expansion. It should not be surprising that, until recently, thelargestclouddatacenterintheworldwaslocatedinthatcity.Ofcourse,datacentersarenot rail yards, but just as transportation centerswere key nodes in the global industrialgrid, cloud data centers are material hubs for global information and communicationtraffic. Imagesof invisibledatamoving throughcloudshelpconveyasenseofwhat thesociologist ZygmuntBauman (2000) describes as our era’s “liquidmodernity.” Today’siconicproductsaredata,information,andmessages,whichflowaroundtheworldthroughthin wires or just through the air. But they are rooted in physical structures thatmakesignificantmaterialdemandsonresourcesandthatcalltomindthefactoriesofanearlierera.Understandingcloudcomputingabsolutelyrequiresanappreciationofitsmateriality,ofitssubstantialphysicalityanditsextraordinarydemandsontheenvironment.

Therearemanyotherwaystodescribethisdanceofpetabytesandzettabytes,andwewill certainly explore some of these, but suffice it to say that nothing in the history ofcommunication and information processing approximates in scale the levels of storage,

processing,anddistributionthatthecloudmakespossible.Withthatsaid,itisimportanttogiveattentiontosomethingmissingfromcloudcomputing,butinordertodosoweneedtoaddressmoreofitscharacteristics.

On-demand self-service. Cloud computing allows users to choose their storagerequirementsandservertimeautomaticallywithoutrequiringhumaninteractionwiththeproviderofeachservice.

Broad network access. Users can access the cloud in standardized ways through anyplatform,suchasatablet,smartphone,orpersonalcomputer.

Resourcepooling.Resources likestorage,processing,memory,bandwidth,network,andvirtual machines can be brought together by the provider to serve multiple users withdifferent physical and virtual resources assigned and rapidly redeployed to meet userdemand.Thisenablestheprovidertoengageuserswithoutregardtolocation,unlessusersdemandthattheproviderspecifyalocationbynation,region,ordatacenter.Forexample,users in theUnited Statesmay notwant to be served by a data center inChinawhose“Great Firewall” of censorship limits access to the online world. Or, fearing theapplicationof thePATRIOTAct,users inEuropeorCanadamaynotwant tobeservedfromtheUnitedStates.

Rapid elasticity. Cloud resources can be expanded and contracted quickly based oncustomerneeds.UsersarenotlockedintoITinvestments,butcanmakeuseofjustwhattheyneed.However,italsomeansthattheymustrelyonaproviderthatistypicallynotasfamiliar as an internal IT department with the history and culture of the organization.SincemovingtothecloudincreasesthelikelihoodthatanorganizationwillshrinkitsITdepartment, that leaves the organization with less inside technical expertise or tacitknowledgetohelpdetermineitsinformation-technologyrequirements.

Measured service. Cloud companies can provide and control services efficiently byemployingameasurementbasedononeormorespecificservices,suchasamountofdatastored, bandwidthused, or quantity of processing. If theprovider is using ametric thatreasonablyreflectstheserviceprovided,thenthereistransparencyforbothprovideranduser.

TypesofCloudComputingInadditiontothesecharacteristicsofcloudservices,onecanidentifythreedifferenttypesof cloud servicemodels that focus on infrastructure, platform, and software,with eachmodel providing the customer with different levels of control. These are not inviolatecategoriesbut,inspiteoflargegrayareasatthemargins,theyareneverthelessusefulinaddingasenseofwhatdifferent typesofcloudmodelsaim toaccomplish, fromsimplyprovidingastorageservicetoofferingadditionalapplicationsandsoftwareforcustomerstouse.

IaaS:InfrastructureasaService.Withthismodel,thecloud-serviceprovidermanagesastorageinfrastructureforcustomerdata,leavingthecustomertodeployitsownsoftware,including operating systems and applications. Furthermore, under thismodel customerscancontrolcertainnetworkcomponents, likefirewalls.It is idealforrepetitiveusesthatrequire elasticity or the capacity to expand or contract quickly depending on use.Examples include online gaming sites, online advertising networks, video-sharing sites,andsocial-mediaapplications.

PaaS: Platform as a Service. Here, in addition to offering storage facilities, the cloudproviderdeploysonto thecloudinfrastructureapplications that thecustomerhascreatedor acquired using programming languages and tools that the provider supports. Onceagain, the customer does not manage the infrastructure; all of that is left to the cloudprovider.Rather,thecustomergetstocontrolthedeployedapplications.Forexample,thecityofEdmonton,Alberta,contractedwithacloudprovidertocreateitsowntool,calledOpen Data Catalogue, which made information about city services accessible to thepublic.TheU.S.DepartmentofDefense(DOD)usedacloudproviderwhenitneededtoemulate battlefield conditions. In the latter case, DOD technical staff developed anapplicationontheMicrosoftAzureplatform.

SaaS: Software as a Service. Under this model, cloud companies offer their ownapplicationsforcustomers touseonthecloudinfrastructure.Customerstypicallyaccesstheseapplicationsthroughwhatiscalledathinclientinterface,suchasawebbrowserthatmightprovidethecustomerwithdocumentprocessingorweb-basedemail.Thecustomerleaves to the provider control over such infrastructure items as operating systems,networks, server storage, and application capabilities. For example, instead of buying acopyofMicrosoftWord,acustomerrents theuseof theword-processingsoftwareforafixedchargepermonthorpaysaper-usefee.Tousethesoftware,thecustomerlogsintothecloudcompany’ssystem.Similarly,asmallbusinessmightrentasophisticatedsalesdatabasefromacloudcompanylikeSalesforcebecauseitwouldnotmakeeconomicsensetopurchasesuchadatabase.Successdependsonthequalityoftherentedsoftwareandthereliability of the cloud provider, especially when the software involves multiple toolsresponsible for running several different units of a business (sales, accounting,administration, and so on). A primary benefit of SaaS is that it minimizes or entirelyeliminates the requirement for in-house IT professionals. Companies selling softwarethrough thecloudgain froma regular flowof income,especiallywhen theyareable toshift popular software from a purchase to amonthly subscriptionmodel, asAdobe didwithitspopularPhotoshop(Pogue2013).Itisalsoimportanttodistinguishamongdifferentmodelsofdeployingcloudsystems,

includingprivate,public,hybrid,andcommunityclouds.

Private cloud. Under this model, the cloud is customized and deployed for a singleorganization. It may exist on or off the organization’s premises, but when it is offpremises,theprivatecloudisprotectedbytheorganization’sfirewall.Privatecloudstendto be chosen by organizations, like banks, that have security and regulatory concernsprohibitingthemfromusingcloudservicesthatarewidelyavailabletothegeneralpublic.Inessence,theprivatecloudisagatedcommunitysetasideforthosewillingtopayforanextradegreeofsecurity.Inthisrespect,itisamanifestationofatrend,troublingtosome,thatwoulddissolve the Internet intoasetofprivately runnetworks (Moses2012). Thismodelistemptingbecauseprivatecloudscanserveasvaultstosecuredatafromsnoopingeyes,theessenceofthebusinessmodelofcompanieslikeReputation.com(Singer2012).Publiccloud.Thismodelistypicallyprovidedbylargecloud-servicesbusinesses,suchasAmazonWebServices, and offers software, platforms, and infrastructure to the generalpublicortoanindustryassociation.Inessencethepubliccloudisavailabletoanyonewhocan pay for it and it is expected to grow five times faster than overall IT-industryexpansion through 2016 (Lee 2013b). Public cloud SaaS configurations are the mostwidelyknownbecausetheyincludefamiliarserviceslikeGoogle’sGmail,Apple’siCloud,andthemarketingservicesprovidedbySalesforce.OrganizationsthatrequirethegreatercontrolbutprefertostayinthepubliccloudmightoptforaPaaSsystemlikeMicrosoft’sAzureorGoogle’sAppEngine.ThoseneedingstillmorecontrolturntoIaaSpublic-cloudserviceslikethoseprovidedbyAmazonandTerremark.

Hybrid cloud. When the cloud infrastructure is composed of both public and privateclouds that remainuniqueentitiesbutare linkedby technology thatallows fordataandapplication portability, we refer to a hybrid cloud service. Many organizations havedividedrequirementsthatmightleadthemtoseekoutthepubliccloudformostoftheirneedsandaprivatecloudconfigurationtomaintainthesecurityofsensitivedata.Hybrid-cloudproviderswhoshareownershipandmanagementwiththeircustomerorganizationsenable them to enjoy the benefits of both types of deployments. While hybrid cloudsappear tobeanexcellent choicebecause theycanbeall things to their customers, theyalso require careful management to balance the component cloud formations. ThecompanyRackspacehasbecomealeaderinthehybridmodel.

Community cloud. Thismodel brings together several organizations that have commoninterests,suchasasimilarorganizationalmission,similarsetofregulatoryrequirements,securityneeds,complianceexpectations,orpolicies.Oneormoreof theseorganizationsmightmanage the cloud or,what ismore frequently the case, theymay together hire athirdpartywhorunsthecloudinthedatacenterofoneoftheorganizationsorhousesitoff-site. For example, a group of airlines might build a community cloud to house acommon reservation system. Community clouds are chosen because they can becustomizedtomeetthespecificneedsofanorganizationalgroup,suchasacollectionofmediafirmsinterestedinsharingfile-baseddigitalmediacontent.Communitycloudsarealso interesting because they have kept alive the early cloud-computing discussion ofbuilding systems that are not primarily under vendor control and operate in a moreenvironmentallysustainablefashion(BriscoeandMarinos2009).

What’sMissing?Although the words public and community are used in cloud computing, every cloudmodel is presumed to be a private service operated by a business with the goal ofmaximizingprofit.Governmentsystems,whichoftenuseprivateprovisioning (even theCIAwillbeusingAmazonWebServices for$600millionworthofcloudprojects),areprimarily employed formanagement, control, and surveillance (Babcock 2013a). In thecontextofcloudcomputing,“public”simplymeans thatvendorswill sell to thegeneralpublicratherthantoasinglepreferredcustomer,and“community”referstothecommoncommercial interests shared by users of that cloud model—for instance, they are allairlines.These areverynarrowuses, if not outrightdistortions, of the termspublic andcommunity.Thepublictraditionallyreferstocitizenswhoparticipateinthedecisionsthataffecttheirlives,andacommunityisacollectionofactivecitizenswithcommoninterests.The history of computing has included extensive debates about public and communityparticipation in the construction of networks and in the provision of services.Unless acloudsystemisspecificallysetuptoprovideinformationtoapublicortoacommunityofcitizens, then thevastmajorityofpeopledonotparticipate in thecloudascitizens,butratherasconsumerswhoarevaluednot for theirparticipation indecisionmakingaboutthecloud,butratherfortheirpropensitytopurchaseservicesandtoprovideinformationtocompaniesabouttheirconsumptionpatterns.

Inaddition tobeinganextraordinary leap inprocessingandstoragepoweroverearlycloud-like systems, cloud computing, unlike computer systems that preceded it, is asingularlymarket-drivenprojectwith little considerationof alternatives to themodel ofmanagementandcontrol thatgovernsit.Wherearethedebatesaboutusingthecloudtoexpand economic or political democracy?How aboutworker participation in corporatedecisionmakingorgreatercitizenparticipationinnationalorcommunitylife?Whataboutpublicparticipationindecisionsaboutclouddatacentersorcloudsystems?Unlikeearliercommunication systems that,whatever the outcome, sparked vigorous encounters abouttheirpotentialtoexpandcitizenshipanddemocracy,thecloudisessentiallysilentontheseissues. There appears to be an enormous gap between the prodigious sublimity of thecloud’spowertoprocess,store,anddistributeinformationandthebanalityofitscurrentapplications,howeverpracticalandprofitable.

Whilealmostallcloudsystemsoperateaccording toacommercialmodel, thereareafewexceptions.Forexample,gridcomputingisameansofcreatingacloudfrombelowbyharnessingthecombinedpowerofmillionsofpersonalcomputerstocarryoutprojects.Buteven theseare typicallyorganizedbycommercialenterprises.Since2004, IBMhassponsored theWorldCommunityGrid,which takes the principle of using the availablespaceonmainlyhouseholdPCs toaddressavarietyofpublic-healthandenvironmentalresearchprojects.Specifically,itmakesthiscombinedcomputerpoweravailabletopublicand not-for-profit organizations for use in humanitarian research. All results are in thepublic domain open to the global research community. Research projects cover cleanwaterandenergy;thedevelopmentofdrugstocombatmalariaanddenguefever;aswellas research on muscular dystrophy, cancer in the young, and AIDS. For example,advancedcomputationalmethodshelptoidentifycandidatedrugsthathavetherightshapeand chemical characteristics to block HIV development. Commercial projects arebeginning to take advantage of this distributed processingmodel, including harnessing

idlePCsinhomes(Novet2013).Theseopenadoortoanalternativeformoflarge-scalecomputingthatdoesnotrequireatop-downcloudcomputermodel.

Cloudcomputingthereforedistinguishesitselffromearliermodelsintwofundamentalways.First, receiving themost attention is the capacity to store, process, anddistributedatabeyondanythingthatprecededit.Whatwereoncetheexceptional“supercomputers”arenowstandardinthehalfmillionorsodatacentersworldwide.Second,evenasithasexceeded its predecessors, cloud computing operates from a diminished vision that isalmost entirely driven by the twin goals of profit and control. There is little interest inusing the cloud to bring democracy, citizen-driven design and implementation, workercontrol, or evenworker involvement in decisionmaking.While all of these ideas havebeenraisedinthecourseofcomputing’shistory,theyarenotpartofdebatesabouttoday’sortomorrow’scloudcomputing.

•

•

•

•

•

•

•

IstheCloudaUtility?This may change before long because there is a contradiction at the center of cloudcomputing thatwill likelyheatup thedebate.Put simply, someforecasts forcomputingarecomingtofruitionandthecloudistakingonmoreofthecharacteristicsofagenuineutility(Clark2012a).Itisnotjusttheacademicandpolicycommunitiesthatarebeginningto thinkof today’s ITenvironment inpublic-utility language.Whenaskedwhathis twocompanies, Twitter and the e-payment firm Square, have in common, Jack Dorseyanswers, “They’rebothutilities.”Moreover,FacebookheadMarkZuckerberghas spentyears referring tohiscompanynotasa socialnetwork,butasa socialutility.However,when asked if his utility should be regulated, the Facebook founder backed off:“Somethingthat’scoolcanfade.Butsomethingthat’susefulwon’t.That’swhatImeantbyutility.”Ofcourse,mostofthecoolthingswethinkofthatwilllastarenotreferredtoas utilities. But whatever the definition or the reaction, the concept of the utility isincreasingly part of the ongoing debate about the developing structure of the computeruniverse(Fox2013).Cloudcomputinghasmadeitamorefrequentlyusedconcept.

Taking into consideration the experience of earlier utilities, such as water, gas, andelectricity,oneenergyexpertdefines the requirementsof autilitymarket as comprisingthefollowing:

Asourceofenergygeneration

Atransportationnetwork

Atransmissionanddistributioncapability

Ameteringcapability

Apricingmechanism

Aregulatortoensureadherencetorules

Acustomer(JamesConstant,citedinClark2012a)

Thisconfigurationofcharacteristicscanbedebated,butmostwouldagreethattheyareamongthemajoronesdefiningautility.AccordingtoClark,cloudcomputingmeetsmostofthesecriteria.Itisasourceofenergygenerationinitsabilitytocomputeandstoredata.TheInternetandthetelecommunicationssystemsconnectedtoitformthetransportationnetwork. Data centers handle the transmission and distribution capability because theyhouse the storage and processing capabilities. Cloud services, especially “public” ones,canmeter precisely howmuch storage and processing are being used at any particulartime, albeit with different providers applying different pricing methods. Pricing isdetermined by the cost to receive, process, and respond to a request for storage,processing,anddistribution.Althoughawiderangeoffactorsisinvolved,cloudprovidersdirectly control the costs of hardware and software because they engineer their ownsystems,andthecostsofotherfactors,suchasfacilities,staffing,andelectricity,dependon the particular market within which they operate. Finally, there is no shortage ofcustomers.Indeedthemarketmostlikelywillgrowtooneinwhichahandfulofprovidersservepracticallyeveryone, justas, forexample,waterandenergymarketsdo.AsClarkconcludes,“Allthat’slackingisaregulator.Whetherthecloudcomputingindustryshould

beregulatedisacomplexissuethatwillundoubtedlybecomeamajordebatebeforetoolong”(Clark2012a;M.O’Connor2013).

Onecandebatewhetheragovernmentregulatorwilleverbecomeessentialtothecloudcomputingindustry.Whatcannotbedebatediswhethercloudcomputingwillbesubjecttogovernance; that is, to theneedforgeneralmanagement,coordination,andoversight.Thiscanbeaccomplishedbyagenciesofgovernment,ashasbeenthecasehistoricallyformost utilities, or it can be accomplished by those with market power. These are bothgovernance structures, notwithstanding themythology of themarket’s “invisible hand.”Themyth tends to focuson themagicof invisible coordinationmore thanon thehand,which,inreality,isquitevisible.Itisapparenttomostobserversofcloudcomputingthat,howevertheymightfeelaboutgovernmentregulation,thereisagrowingconcentrationofpower among a handful of cloud providers, most of which are also key players in theproduction and distribution of software and content. Utility markets often becomegovernmentregulatedbecauseoneorafewproducers,whousetheirpositiontoexercisesignificantpoweroverservicesandpricing,cometodominate.1Historically,thishasbeenthecase throughout thehistoryof communicationmedia thatwasmarked in theUnitedStates,forexample,byWesternUnion’scontrolovertelegraphy,AT&T’sovertelephony,and the broadcasting networks’ domination over radio and then television. In each caseregulation was called on to temper the threats of monopoly or oligopoly control. Thispatternwasfollowedinothernations,butsomeofthesealsoturnedtopublicownershiptoguaranteewidespread, if not universal, access to an essential service. So it is not verysurprising that alongwith the term utility, the concept of regulation has entered publicdebate in today’s computer and social-media world (Marshall 2013). The demise ofspecificcloudservices,suchasGoogleReader,becausecompaniescannotrecoverfixedcosts from their provision, has led some economists to wonder whether governmentownership or regulation through public utility status is inevitable for essential butunprofitable services like search (Kaminska 2013). Even as businesses in the lessdevelopedworldbegin toembrace thecloud, they fear that itmightdomoreharmthangoodwithoutthestabilityprovidedbygovernmentregulation(Hanna2013).

Oneofthekeyreasonswhyexpertattentionisreturningtotheconceptoftheutilityincloud computing is that the industry is rapidly becoming dominated by a handful ofcompanies.ThepowerofAmazon,Apple,Google,Facebook,andMicrosoftistroublingenoughtoleadsometodoubtthatthe“invisiblehand”willproveadequatetorestraintheirabilitytodominatecloudmarkets(McKendrick2013b).Consequently,theymaintain,weshould begin to think about broader national or even international oversight by electedrepresentatives.Asoneconcernedanalystputit,“TheInternethastakentheplaceofthetelephone as the world’s basic, general-purpose, two-way communication medium. AllAmericansneedhigh-speedaccess,justastheyneedcleanwater,cleanair,andelectricity.Buttheyhaveallowedanaivebeliefinthepowerandbeneficenceofthefreemarkettocloudtheirvision.Asthingsstand,theU.S.hastheworstofbothworlds:nocompetitionand no regulation” (Crawford 2012). According to Crawford, when it comes to theInternet, theUnitedStatesshould follow thehistoricalexampleofotherutilities.When,forexample,electricitycameunderthecontrolofahandfuloffirmsthatprovidedserviceonlytothosewhopaidtopdollar,publicpressureledtothecreationofregulatedutilitiesandpubliccorporations.OpponentsofthisviewarguethattheInternetandthecloudare

fundamentallydifferentfromroads,water,andelectricityandthatgovernmentregulationwould stifle the incentive to risk-taking innovation. In 2013, the divide between cloudcomputingandelectricalutilitiesblurredwhenresearchfoundthatagrowingnumberofcloudcompaniesweremaking significantprofitsby resellingelectricity tocustomers inaddition to providing space to house data. This practice, what has been dubbed thecreationof“wildcatelectricalutilities,”hasledtomorecallsforgovernmentregulationofthecloud(Glanz2013).

Inadditiontotheseconcerns,thereistheissueofdatapreservation.Absentsomeformof regulation ormutual agreementwithin the IT industry, and specifically among thosewhoaremajor cloud-servicesproviders, there is no requirement topreserve thephotos,email, videos, postings, data, and flies that individuals and organizations believe aresecurelystoredindatacentersaroundtheworld.Asaresult,muchofthedigitalevidencefrom thedaily livesof individuals and thedecisions andactivitiesoforganizationswillvaporize, irrespectiveof howmanyclouddata centers fill theworld.Asone concernedtechwriterargued,“We’rereallygoodatmakingthingsfaster,smaller,andcheaper.Andevery step along the way makes for great headlines. But we’re not nearly so good atmigratingourdigitalstufffromonegenerationoftechtothenext.Andwe’rehorribleatcomingupwithbusinessmodels that assure its longevity and continuity” (Udell 2012).Anotherpersonwhohasbeenactiveintheonlineworldforyears,hostingnumeroussitesandarchives,worried,“Nottobedramaticoranything,butnomorethanfortydaysafterIdie,andprobablymuchsooner,allthecontentIamhostingwilldisappear”(Winer,quotedin ibid.).Todate, theonlyreasonmostof thismaterialhasbeenpreserved isdue to theheroiceffortsofindividualswhopersonallyportarchiveswhentechnologyandstandardschange. Referring to several archives dating from the turn of this century, Udellcommented inaWiredcolumn,“If Ihadn’tmigrated them, they’dalreadybegone.Notbecause somebodydied, it’s just thatbusinesses turnedoveror lost interest and thebitsfellofftheweb.Gettingpublished,itturnsout,isalousywaytostaypublished.Withalldue respect to wired.com, I’ll be amazed if this column survives to 2022 without myintervention” (ibid.). There are some efforts, primarily by governments, to archive andpreserveflies.ThemostnotableofthesemaybeattheU.S.LibraryofCongress,which,amongother things, isarchiving themassivedatabaseofTwitterpostings.Theseareallimportant activities, but they are isolated and much more data disappears than ispreserved.Of course, one can argue, there is a great deal of digital content that is notworthpayingtopreserve.Societyhassurvivedinthepastwithoutcarryingforwardfromgeneration to generation the entire weight of the historical record. Nevertheless, sincemostofthatrecordisnowdigital,isitnotworthwhiletodevelopstrategiestopreserveatleastsomeofitinasystematicfashion?

Now it is important to turn to an overview of major participants in the cloudmarketplace, starting with the five companies generally considered dominant on theInternetandinthecloud.

MappingtheCloudIndustry:LeadersandChallengersArguably the leading force in theU.S. cloud computing industry, and a global giant aswell,Amazonbeganby applying computer power to transformpublishing and then thegeneralretailindustry.Asprodigiousasthisaccomplishmenthasbeen,onecommentatorconcluded that theseachievements“maybe footnotes to thecompany’s largerandmoresecretivegoal:givinganyoneontheplanetaccess toanalmostunimaginableamountofcomputingpower”(Hardy2012a).By2013,accordingtomostaccounts,itssubdivision,AmazonWeb Services, was the leader in U.S. cloud computing. As an analyst for theconsulting firm Forrester described it, “Almost every major consultancy supportsAmazon; almost every advertising agency runs onAmazon; if I need to hire 10 peopletomorrow to help me build my application, it’s super easy to find people who haveAmazonexperience”(MillerandHardy2013).WhileAmazondoesnotbreakoutrevenueforcloudcomputing,2012estimatesrangefrom$800milliontoasmuchas$2.4billion(ibid.;Mims2013).ThecompanyoperatesitscloudservicesthroughtheaforementionedAWS, which achieved widespread public attention in 2012 because the Obamapresidential campaign used AWS to organize its successful voter analysis and voter-turnoutdrive.Bythemiddleof2013,onetypicallymodestindustryobserverconcluded,followingthecompany’sthirty-seventhcutofitscloudprices,whichsenttremorsthroughtheindustry,“Theproofisin:Amazonfullycontrolsthecloud”(Linthicum2013c).Thisconclusionmaybeprematureandatadoverstated,butitdoescorrectlyidentifyAmazonasanincreasinglydominantforceinthecloudbusiness.

AWSwas created in 2004with about forty employees, andwas the first company torent its data storage and computing power to other companies. Although it is highlysecretiveaboutmostofitsoperation,by2012Amazonwasregularlylistingmorethan600jobopeningsonthecompany’swebsite.ItoperatesseverallargedatacentersintheUnitedStates,eachofwhichcontainsmultiplebuildingswith thousandsofservers. ItalsorunsdatacentersoutsidetheUnitedStatesandhasseveralunderconstruction.AWSisnotthelargestcloudprovider in theUnitedStatesbyquantitativemeasuressuchassizeofdatacentersortotalnumberofservers,butitisarguablythemostpowerfulbecauseitispartoftheAmazoncorporateempireandtherelationshipmarksoneof thefewtimeswhentheoften-usedbuzzword“synergy”isanunderstatement.AWSbenefitsfromthesheersizeofitsparent’scomputingpower.Forexample,whiletheparentAmazondoesnotrevealthesizeofitsoperations,anexecutivewhoknowsAmazonwellmaintainsthatjustoneofthecompany’s data facilities in the eastern half of theUnited States containsmore serversdedicatedtocloudcomputingthandoes theentireoperationofoneof themajorhybrid-cloud companies, Rackspace, which in 2013 served 200,000 clients, mainly businesscustomers,withabout100,000servers inninedatacenters.AWSalsobenefits fromthedata that Amazon gathers on its millions of customers whose purchases of books,homeware,clothing,andsoonprovideinformationthatAWSusestoforecastconsumerbehavior,aboostforboththeparentandthefirmsthatpurchaseAWS’sservices.AmongitsmajorcustomersarepopularmediafirmslikeNetfix,Pinterest,Shazam,andSpotify.Amazon has been so successful in the cloud that company management expects it tobecome the leading revenue producer for Amazon, topping even its renowned retaildivision, with sustained growth estimated at 45 percent per year through 2017 (Finkle2012).

Market power gives Amazon considerable leverage over its competition, large andsmall.AstheheadofAWSputitwhenaskedaboutastepped-upchallengefromGoogle,“We’ve always been very good at making everything as low-cost as possible, then weloweritsomemore”(MillerandHardy2013).Thecompanyisabletopriceitsservices,particularly the storage and data-analysis capacity of its servers, so inexpensively thatneithermanyestablishednorstart-upcompaniesanylongerbotherinvestingintheirown.Instagram,forexample,thehighlysuccessfulwebphotocompany,whichisnowapartofFacebook,didnotbotherinvestinginitsowncomputers.Thestart-upCue,whichadmitsto spending $100,000 amonth onAWS services, uses them to scanmillions of emails,Facebookpostings, andcorporate records toprovideenhanceddata that subscribers canuseinalloftheironlineactivity.Over185federalgovernmentagenciesalsorunsomepartof their services throughAWSandAmazonhaswona$600millioncontract toprovidecloud services for theCIA (Babcock 2013a). The company is active internationally; inaddition to having data centers located in Asia, Europe, and Latin America, it hostsnumerous corporate and government clients outside the United States. For example, aGermancompanyusedAWStomakedigitalcopiesof20,000televisionshows,ajobthatcostthefirmlessthanitwouldhavespentontheelectricityaloneifithaddonetheworkinhouse.AWSserverslocatedinCaliforniaandIrelandprovidepeopleinAfricawiththeability to comparison-shop cars using smartphones connected to AWS. There is nogainsayingAmazon’srichdatabaseofcustomersearchesandpurchases,whichaddsvaluetoAWS’sofferings.Asone customer commented, “You cannow test a product againstmillionsofusersforjustafewthousanddollars,orstartacompanywithjustoneortwopeople”(Hardy2012a).

To multiply these success stories, Amazon has to successfully deal with two majorchallenges:providing continuous reliable service and fendingoff the competition.AWShas been a generally reliable cloud provider, but a handful of notable outages havedamaged the company’s reputation. One of the most significant took place over theChristmas holidays in 2012, when Netfix customers lost access for the better part ofChristmasEveandAmazonitselflostserviceforitsowncustomersonChristmasDay.In2013Netfix reliedonAmazonfor95percentof itsdata-centerneedsand, in thehighlycompetitive video-streaming marketplace, the company cannot tolerate significantdowntime.Asoneindependentanalystconcluded,“NetfixandotherorganizationswhichrelyonAWSwillhavetoreexaminehowtheyconfiguretheirservicesandallocatetheirservice requirements across multiple providers to mitigate over-dependency and risks”(Finkle 2012). Amazon is not alone in experiencing outages. They affect the entireindustry, are primarily caused by power problems, and, on average, last for 7. 5 hours(Talbot 2013). They also lead to unanticipated consequences and hidden costs (Franck2013).

Reliability also requires guarantees of security, another problem for public cloudcompanies,andAmazonisnoexception.In2013,asinglesecurityresearchermanagedtouncover126billionfliesthatwereleftopentothepublic.Fromasampleof40,000flies,he found exposed data belonging to a medium-sized social-media service, the salesrecordsofacardealership,employeespreadsheets,andvideogamesourcecodefromamobile-games developer. The shockingly exposed flies also included unsecuredpasswords.Amazon tookmeasures to secure the data andwarn customers, but this one

eventleftitsclientsunderstandablyworriedthatpublic-clouddatawasfarmoreexposedthananyonethought(Brian2013).

Amazon also needs to overcome competitive pressures, especially from a handful ofcompanies thatcanalso leverage their leadership innew-mediahardware, software,andmedia services. Someof these, likeMicrosoft, IBM, andOracle, havemore experiencethanAmazoninthemarketforlargecorporateclients.OnewayforAWStosucceedisbyheavilydiscountingcloudservices,then,oncethecompetitionisdrivenoutofthemarket,raisingpricesoncemore,a tactic thatprovedsuccessful inAmazon’sretailbook-sellingoperation(Streitfeld2013).Itisnotanexaggerationtosaythateveninthisearlystageofdevelopment, thestruggle forcompetitivedominance incloudcomputing, justasacrosstheInternet,isnarrowingtoahandfulofcorporationsthatcanmarshalasimilardegreeofleverage (McChesney 2013). These include familiar names: Apple, Google, andMicrosoft. Of these leaders, Microsoft is probably the most committed to providinggeneralcloudservices,especiallytobusinesses,whichhavehelpedthecompanymaintainitselevatedpositionevenastheothershavesuccessfullychallengeditsconsumer-servicesmarket. Businesses and government agencies have long been committed to Microsoftsoftwareandthecompanynowaimstomovetheseandnewcustomersfromrelianceonphysicalprograms toonlineservices fora fee.So far ithasbeen reasonablysuccessful,with over 100,000 businesses using the company’s cloud services. It is important toemphasizethispointbecausemuchoftheday-to-dayattentioninthepopularpressgoestotheothers,primarilybecauseGoogleisthemajorgatewaytosearch,Appletomusic,andFacebook to socialmedia.EvenTwitter, amuch smaller company, garnersmore noticethanMicrosoft.ButthecompanyBillGatesstartedin1975hasaverystrongfoundationin business software and, with softwaremigrating to the cloud,Microsoft has investedheavilyincloudplatforms.Overthelastfewyears, thecompanyhasquietlybuiltupitsServerandToolsdivisionanditnowgenerates$18billionayearinrevenues,withsixofitssubdivisionstoppingthe$1billionmark.

Microsoft is counting on the cloud platform offering its Azure service to enablecustomers to develop applications and otherwise make profitable use of their owninformation.AzureprovidesbothPlatformandInfrastructureasaServiceandonceagaindemonstratesthevalueofproximitytoservicesandsystemswithinalargecompanylikeMicrosoft,whichdevelopedAzurebyusing someof theelementsof its successfulwebbrowser Bing (Wilhelm 2012). In recent yearsMicrosoft has not been as successful inconsumerservices,butitisalsomakingabigpushtotakeindividualsandfamilies,asitsadvertisingsloganrepeats,“tothecloud.”TheseincludeWindowsLive,asuiteofcloudservicesthatincludesfilestorage,image,video,email,messaging,theBingsearchengine(now the secondmost popular in theUnitedStates), andXboxLive.Finally,Microsoftexpects that thecloudversionof itsverypopular suiteofword-processing, spreadsheet,and related programs will succeed in the cloud, as what it calls Office 365 begins todeliverthemonasubscriptionbasis.

Google’sconcentrationonconsumerservicespioneeredinitssearchenginehasledthecompany to focus on that side of the cloud market. It has expanded the company’sconsumercloudbeyond searchwithdocument storage (GoogleDrive),wordprocessing(Google Docs), and entertainment (Google Music) applications. Furthermore, howevermuch itworries techobservers, thecompanyalso sells itsowndevices that areentirely

dependenton thecloudfordatastorageandapplications(Gilmoor2013).These includethefamiliarChromebooksaswellasGoogleGlass,whichGooglehopestousetosellpay-per-gaze, forwhich it holds a patent, to advertisers (Bilton andMiller 2013).Butwithcompetitive threats fromAWS andMicrosoft,Google has begun amajor push into thebusiness market with Google Compute Engine (GCE), its IaaS unit. Again, as withAmazon,built-in leveragemattersagreatdeal. In thiscase,Googleruns its IaaSon thesame technology that powersGoogle search,which leads the company to claimgreaterreliability thanAWS,especiallybecauseof thenotableoutages the latterexperienced in2012 (Chen2012). In2013,Google tiedGCE to theGoogleAppEngineand itsglobalnetworkofappdevelopersinthehopeofbeatingthecompetitionbyprovidingcustomerswithacloudservicethatincludesprivilegedaccesstothelargestsetofappsincyberspace(Hardy2013d).ThisiswhyGoogleisnotreluctanttoboast:“Forthemostpart,GCEispositioned as a way for customers to benefit from years and years of infrastructureinvestments, which span everything from our datacenter design to our operationalpractices, our hardware design and software design, [and] includes the software stack”(Clark 2012b). Reassurances aside, breakdowns lead users to worry that they are notkeeping a close enough eye on their own data. Indeed, one of the key challenges forcompanieslikeAmazonandGoogleistobalancethecostsofmeetingworriedcompanies’demandsforgeographicalproximitytodata,evenastheymakeuseofaglobalnetworkofdatacenterstoensuresufficientnetworkredundancytosupporttheirclaimsofprotectionagainstoutages.

Likeitsrivalgiantsintheindustry,Googleiscomfortablemovingintonewterritory,inthiscase thebusinessapplicationsmarket, longdominatedbyMicrosoft. IndeedGooglehas been so committed to innovative product development that it has been dubbed theGeneralElectricofthetwenty-firstcentury(Gapper2013b).Foryears,GoogleAppswaspitchedmainly tosmall firmsandstart-upsbecauseMicrosoftdominated themarket forlarge businesses. But Google has begun to cut into this lucrative segment with majorprivate-sector clients like the pharmaceutical giant Hoffmann-La Roche, where 80,000employeesusethepackage,andpublic-sectorclientssuchastheU.S.DepartmentoftheInterior, where 90,000 use Google Apps as their staple business-productivity software.BorrowingapagefromAmazon’splaybook,GooglereliesonconsistentlowpricingthatMicrosoft has difficulty matching (Hardy 2012b). Microsoft fights back, but does notappeartotakeGoogleveryseriouslyasacontenderinthismarket.Somemightconsiderthis amistake, butMicrosoft is clear that Google is not a threat in the business cloudmarket because, according to the general manager of Microsoft’s business division,Google“hasnotyetshowntheyaretrulyserious”aboutenterpriseapplications.“Fromtheoutside, theyare an advertising company” (Kerr2012).There is some substance to thisview.Afterall,in2011only4percentofGooglerevenuecamefromitsbusinessservices,whereas96percentcamefromadvertising.Microsoft’scloud-basedOffice365isintendedto keep Google’s business market share from growing, but Microsoft has yet todemonstratewidespreaduptakeoftheservicebecausebusinesses,worriedaboutsecurityandoutages, stillpreferMicrosoft’smore familiarOfficesoftware (ibid.).Early in2013Google accelerated a push to challenge Amazon and Microsoft in cloud services. ItdoubledthesizeofitsofficespaceintheSeattlearea,neartheheadquartersofbothrivals,andbeganlarge-scalehiringofcloud-computingexperts.Inadditiontoopeninganotherinthe many revenue streams that Google enjoys, the company expects it will have the

multipliereffectofluringappdevelopersandothercompaniestouseGoogleproductsandtolaunchfromtheGoogleplatform.

It ishard tocontendwith theview thatApplehas succeeded increatinga successfulconsumer cloud. With iCloud and iTunes Match, Apple has the largest share of theconsumer cloud-services market in the United States, substantially ahead of Dropbox,AmazonCloudDrive,andGoogleDrive.Moreover,thesheersizeofApple’sdatacentersintheUnitedStates(itsNorthCarolinafacilityaloneisoneofthelargestintheworld)andits seemingly constant process of expansion demonstrate the company’s continuingpopularity. So do the sales of its line of computers, tablets, and smartphones (Fingas2013).Muchof this success can be traced to the vision of its founderSteve Jobs,whorecognized the importance of the cloud in 2008 and committed to it in 2011 when,althoughillwith thecancer thatwouldsoontakehis life,heannounced toaWorldwideDevelopersConferencethecompany’s“nextbiginsight”:“WearegoingtodemotethePCandtheMactobejustadeviceandwearegoingtomovethedigitalhubintothecloud”(Isaacson 2011, 533). While Google, Facebook, and Twitter garner attention as mediadisrupters, Apple has become one of the world’s largest media companies by creatingcloudversionsof traditionalmedia.Apple’s iTunesStoreandAppStore, throughwhichpeople purchasemusic, video, and e-publications, earnmoremoney than the combinedrevenueoftheNewYorkTimes,theSimon&Schusterpublishingcompany(whichputoutthebest-sellingbiographyofApple’sfounder),WarnerBros.filmstudios,andTimeInc.(thelargestmagazinepublisherintheUnitedStates).ForthefiscalyearendingSeptember2012Apple’smediacloudservicesearnedabout$8.5billion,or$300millionmorethanthecombinedrevenuesoftheothers(Lee2012).BecauseAppledoesnotclearlybreakoutitspuremediasales fromthose, forexample,of itsnonmediaapps,notallof its iTunesearningscomesolelyfrommedia.Furthermore,Apple’scontentdivision isstilldwarfedby conglomerates like News Corp. and Disney. Nevertheless, Apple’s cloud media isincreasing at a 35 percent annual rate,making it the fastest-growing commercialmediaoperationintheworld.

Assuccessfulasithasbeeninconsumerservices,Applehasbarelymadeadentinthebusinessmarketforcloudservices.Whenithastried—forexample,withiWeb,awebsite-publishing service—the company has failed to win over customers from its business-servicescompetition.AsApplebackedofffromiWeb,itscustomersneedingapplicationsto designwebsites and a host to serve themwere left out of the cloud and in the cold.UnlikethatofAmazon,Google,andMicrosoft,Apple’sbusinesspresenceisfeltonlyinhardwaresales.Theseareadmittedlysubstantial,buttherehasbeenlittlecrossoverfromhardware into platforms, applications, and services. As one reviewmaintained, “WhileiCloud,again,isawesomeforpersonaluse,businesseswillfindthemselvesbetterservedbyaterminalserverparkedinasecuredatacenter,VPN[virtualprivatenetwork]accessto a corporate server, or another cloud-based file sharing solution that ensures onlyauthorizeduserssecurelyaccesscorporatedata”(Eckel2012).Inotherwords,customerswill continue to shop the cloud at AWS,Google,Microsoft, or one of the other cloudbusiness-servicecompanieslikeRackspace.

Facebookisalsoamajorplayerinthecloudcomputingindustrybut,likeApple,itusesthecloudtoservicethegargantuanneedsofitsownsite,whichincludesabout1.3billionusers. The company learned about cloud computing the hard way when in 2006 its

computers came close to literally melting down. At that time Facebook was renting asmall space in SantaClara,California, and filled itwith the racks of servers needed tostore and process activity on its members’ accounts. When electricity powering thegrowingsystemoverheatedcriticalcomponents,thechiefengineerandafewstaffheadedto a local pharmacy and bought every electric fan in the store. The fans worked, theservers were saved, and the rest, as they say, is history. The company had 10 millionsubscribers at the time and would not have reached anything close to the billion-plusmembers who upload 300million photos a day if it failed to master the cloud (Glanz2012b).Today, all thosephotosamount to7petabytesofdata eachmonth, anda cloudserversystemthatcalibratesstorageconditions,includingtemperature,bycalculatingthelikelihoodthatmemberswillaccessinformationandphotos.Forexample,colderstorageslows retrieval time, but that works fine for the billion photos a day uploaded aroundHalloweenthatmembersareunlikelytowanttoretrieveafterthecostumesareputawayforanotheryear.Theseissuesarechallenging,butFacebookbenefitsfromkeepingallofitsdataneedsinhouse.Asaresult, thekeypressuresfacinganycloudprovideroruser,such as sharing, securing, and syncing, aremore easily addressedbyFacebook thanbycompaniesthatareinthebusinessofservingthousandsofdifferentbusinesses.

Amazon, Microsoft, and, to a lesser degree, Google, demonstrate their marketdominancetocloudcustomersthroughongoingpricecutsthatbenefitthegeneraluseranddrive smaller competitorsoutofbusiness. It is certainly aproblem forolder companieslikeOracle,HP, and IBM,which have significant costs associatedwith legacy systemsthat are not as scalable as the latest technology.As a result, these firms are starting tochange,eitherbyjoininginpartnershipswithcloudcompaniesorbyacquiringpromisingsmaller firms, as all three did in mid-2013 (Hardy 2013b, 2013e, 2013h; Kolakowski2013). Moreover, IBM, which operates twenty-six data centers around the world, hasbegun to transform itself into a company that resembles marketing giants like WPP,Omnicom, and Publicis (Waters 2013c). All of this is taking place even as these sameadvertisingfirmsare transforming themselves intoonesdrivenby theuseofbigdata inthecloud.ThemergerofOmnicomandPublicistoformthelargestadvertisingbusinessintheworldisgroundedintheneedtotakeonthenewcompetitionfromintegratedcloud-basedinformation-technologycompanies(Vega2013).

Pricecutswouldappear tobeanunqualifiedbenefit to thecloudcomputing industryandespeciallytoitsusers,whoareincreasinglydependentontheservice.However,whencarriedoutbyindustryleaderslikeAmazonorGoogle,theyarealsoclassicstrategiestoconcentrate power over a market. This has been demonstrated throughout economichistory, including in the communication industry fromWestern Union in telegraphy toAT&T in telephony. For years, AT&T initiated price cuts in telecommunications at themerewhiff of a competitive threat, only to raise them againwhen the competitionwaserased.ThatthecompanywasabletoaccomplishthisevenundertheregulatorynoseoftheFederalCommunicationsCommissionisevidenceofitspowerandofthecontinuingfailureofgovernment tocarryoutregulatoryresponsibilities. Itwasnotuntil the largestcorporate users of telecommunication services organized collectively to fight back thatAT&T’sgriponthemarketwasbroken.Today,analystswonderifcloudcomputingwillgodownthatsamepath.Accordingtooneanalyst,“Thereisaracetothebottomwhenitcomestocloudpricing,asthelargerproviderstrytocaptureasmuchshareastheycanof

thisexplodingmarket.Thedownsideisthatthesmallerproviderswithouthugewarchestsofcash,butwithimpatientinvestors,won’tbeabletomakemoneyatthepricesthatthelargernamescharge.Manyofthemwillstruggletohanginthroughthedaysoflowornocloud computing profits—andmany of themwill have to toss in the towel or have thetoweltossedinforthem.”2Theonlylong-termupsideisforthelargestproviders:“Oncethesmallerprovidersarepushedout,youcanbegintoraiseyourprices.Hmm,itsoundssuspiciously like a page from the big-box stores’ playbook—and a warning for cloudadoptersnottocountonlow,lowpricesasthenorm”(Linthicum2012).

Oneofthekeystocreatingandmaintainingmarketcontrolistoexercisepowerupanddown the chain of production. A handful of companies are doing this in one directionthrough price cuts and in another direction through their relationships with key ITproducers—particularlythegiantinthismarket,Intel,theworld’slargestandmosthighlyvaluedsemiconductorproducer.Intelworriesthatthehardwareworlditdominated,ledbythevenerablePC, is indecline.Accordingtooneanalyst,“Intelstillhasa lotofdough,but their old world is cracking” (Hardy 2013f). As a result, the company is especiallyconcernedwithpleasingwhat it refers to as theBigFour:Google,Microsoft,Amazon,andFacebook(ApplepurchasesitschipsmainlyfromSamsung)notonlybecauseoftheirsize but also because they lead a critical and growing market. Intel has been losingrevenueinthepersonal-computermarketthatmadeitahistoricleader.The$25.8billionitearned from itsPCclient group in2012 remains enormous, but that figure represents adeclineof2.25percentfromthefirstthreequartersofthepreviousyear,largelybecauseoftheshifttotabletsandsmartphonesfromstandardpersonalcomputersandlaptops.Ontheother hand, the company’s revenue shot up by 6.7 percent in its data-center business,whereitearned$7.9billion.Thathastriggeredaseriousmakeoveratthecompany,whichnow views itself as more of a cloud-computing company than a client-server business(Hardy2013f,2013g).

TheheadofIntel’sdata-centergrouprealizesthatthecompanyhastochangedirection,but believes that if it does so successfully, it could boost data-center revenues to $20billion by2016.But in order to accomplish this, Intel needs to listen and at times takedirectionfromlarge, influentialcompanies,something it isnotused todoing.AsIntel’sdata-centerdirectordescribedthesituation,“TheBigFouroperateataverydifferentbeatrate, and they are very tech savvy, so they don’t want a lot of input. They all get adedicatedsalesperson, thesameas theothers inourTop40customers,but thereisa lotmoredirectinnovationfromthem,andalotofsharingofideas”(Hardy2012c).TheBigFourarenowactiveinengineering,innovating,andtestingnewsemiconductors,includingone Intel installed in September 2011 but did not introduce to the general public untilMarch2012.Inteladmitsthatitswillingnesstoabsorbthepotentialproductionproblemsassociatedwithanewchipthathasnotyetbeenreleasedtothegeneralpublicinordertohave the latest semiconductor “was a new thing” for the company (ibid.). MeanwhileApple, which has been dependent on Samsung for the bulk of its chips, is seriouslycontemplating manufacturing more of its own, partly because of the Koreanmanufacturer’sannouncement inNovember2012 that itwouldboostchippricessold toAppleby20percent.ButthisisalsobecauseApplesimplywantstocontrolmoreoftheproductionprocess(Whittaker2012).PatentbattleswithSamsungarecertainlyanissue,buttheneedforcontrolandtheabilitytocarryitoutareevenbigger.

Largecloudcompanies are challenging firmsat all points in the chainofproduction,fromsmallcloudcompetitorstochipmanufacturers.Theyarealsogoingaftercompaniesthatmanufacturecomputers.Amazon,Apple,Google,Microsoft, andFacebookallnowbuildtheirownandchallengecompanieslikeIntelandHPtomeetorexceedperformancespecifications.PerhapsthemostsurprisingforitsactivityinthisareaisFacebook,becauseithasnotbeenamongthoseidentifiedwithdevices.ThecompanyhasjoinedwithbothHPandIntelinthepublicannouncementofanewchip.Googlehasevendevelopeditsownsemiconductor but has not patented it because the company is concerned that doing somight reveal too much about its plans (Hardy 2012c). Amazon is building a globalcomputer system including its own customized computers, data storage systems,networkingsystems,andpowerstations(Hardy2013a).

These examples demonstrate some of the ways that large cloud companies areexpandingtocontrolthemarket.Theyareintegratinginternallytorationalizeproductionfromhardware to software,applications,andpricing.Thesemovesenablecompanies toextendtheircontrolovercloudcomputingmarketsand,fromthere,establishkeypositionsinthedevelopmentofinformationalcapitalism.Onewaytolookatthisprocessistoseeitasaseriesofstepsonthewaytothecomputerutility.Thatwouldbeaccuratebut,aswasnotedearlier,withnoregulatoryapparatusinplaceoronthehorizon,itisalsoreasonabletoseethemasstepsonthewaytoaglobalcartel,differentfrombutalsosimilartotheoilcartelthatinfluencedglobalenergy-resourcemarketsformanyyears.Beforelong,itmaybe time to think seriously about the implications of a global cartel in informationresources. As in oil, such a cartel would provide for the needs of organizations andindividuals, using control overvarious stages in theproduction anddistributionprocessthatpowersglobalcapitalism toexpandprofitandcontrol. Justas inoilorotherglobalcommoditymarkets, therewillbesmall-andmedium-sizedproducerswho, fromtime totime,disruptthesystem.Geopoliticalupheavalsandtechnologicalchangewillalsohavean impact. Inshort,cloudcomputing is rapidlybecomingapowerful force in theworldbecause of the quantitative and qualitative leap in information production, processing,storage, and distribution, and because of the way the cloud is evolving into a global,private oligopoly,well on theway to becoming a global cartel. It is also interesting toobservethewaysthatsomeofthecompaniesmakingupwhatmightbecomeacartelarebeginningtointernalizetheappropriateidentityforthisnewrole.ConsiderGoogle,whosefounder,EricSchmidt,nowtalksabout theneedfor thecompanyand itscompetitors tostart thinking of themselves as nations, especiallywhen it comes to dispute resolution:“Theadultwaytorunabusinessistorunitmorelikeacountry.Theyhavedisputes,yetthey’veactuallybeenabletohavehugetradewitheachother.They’renotsendingbombsateachother….IthinkbothTim[Cook,Apple’sCEO]andLarry[Page,Google’sCEO],the sortof successors toSteve [Jobs]andme ifyouwill,haveanunderstandingof thisstatemodel”(Lessin2012).

Schmidtmaytakethisviewmoreseriouslythanpeoplethink.InJanuary2013hecameundersomeprettyharshcriticismfromtheU.S.StateDepartmentfor travelingtoNorthKorea tomeetwith its leadership in a round of private diplomacy unsanctioned by theU.S. government.CitingU.S. concerns about aNorthKorean rocket launch onemonthearlier,aStateDepartmentspokespersoncommented,“Frankly,wedon’tthinkthetimingof this is particularly helpful.”Moreover, “They are traveling in an unofficial capacity.

Theyarenotgoing tobeaccompaniedbyanyU.S.officials.Theyarenot carryinganymessages fromus.They are private citizens and they aremaking their own decisions.”Coming from the agency responsible for American diplomacy, these are pretty strongwordsaboutaprominentU.S.citizen(Gordon2013;seealsoSchmidtandCohen2013).

Developmentsliketheseleadsometowonderwhetherwearesoontofacetheproblemofmonopolymarket domination that once led the government to intervene against thepowerofStandardOil, IBM,andAT&T.Somehavemaintained that itwasgovernmentpressure on IBM, even as it dropped the thirteen-year-old case in 1982, that led thecompanytounbundleitssoftwarefromthehardwareportionofthebusinessandtherebyadvancethemassivegrowthoftheU.S.information-technologyindustry.Furthermore,itwas likely that the breakup of AT&T around that same time helpedmake the Internetpossible. In addition, the government’s 1990s case against Microsoft, which hadsuffocated innovative companies likeNetscape,made it considerably easier forGoogleandFacebooktoappear(Fox2013).

Not everyone agreeswith the view that an oligopoly or a cartel is about to be born.Somemaintainthat,evenwithcontinuouspricecuts,Amazonwillfacestiffcompetitionfrom within and outside the major cloud providers, including from small innovativecompanies.TherearealsoconcernsaboutApple’sabilitytoenjoyelitestatusinthecloud.AnalystspointtothedifficultythecompanyhasexperiencedinmakingitsbedrockiTunesservice meet the promise of seamless integration and synchronicity across platforms.Moreover, the company has not expanded its services with offerings that have earnedGoogle and Microsoft the reputation of general cloud-server companies. Also, whileeveryone agrees that Microsoft has succeeded in building on its success in businessservicesasithasmovedtothecloud,doubterswonderwhetherWindows8andSkyDrivewillsucceedincreatingamajorcloud-computingpresenceintheconsumermarket(CloudTweaks2012).Somealso insist thatmanycompanies,seeminglybeatenby thenewBigFour(orFive,ifyouincludeApple),havethecapacitytofightbackandarebeginningtodo so. These include big broadcasters who have seen their audiences diminish in theexpansion of digital social media. According to one analyst, “But as more and moreInternet-connectedsmarttelevisionsfindtheirwaysintopeople’shomes,broadcastersseeanewopportunitytoremainatthecenteroftheglobaladindustry”(Steel2012b).Theycan do so partly because the new wave of Internet-connected televisions permitsbroadcasters likeCBS to sell new forms of advertising to directmarketerswho do nottypicallypurchasecommercialadvertisingbecausetheyfocusoncoupons,searchads,anddirect marketing. Internet-enabled television receivers permit broadcasters to add webadvertiserstothebrandadvertisingthatbuilttheindustry.Broadcastersnowcaptureonly$10billionofthe$60billionspentannuallyondirectmarketing.ButtheshifttoInternettelevision has the potential to enable broadcasters to expand that share and enter newmarkets. So while it is likely that there is some hyperbole in the statement by a CBSresearcherthatthiswillusherin“anewgoldenageofnetworktelevision,”itdoesindicatethat“legacy”companieslikeNBC,CBS,andABCwillhavesomethingtosayabouttheemergingconsumercloudcartel(ibid.).

Three of the most important challengers to Amazon and other major players in thecloudshouldbefamiliartoanyonewhohaspurchasedacomputerorprinteroverthelasttwenty years: IBM,HP, andDell. These companies hope to profit by building on their

establishedbaseindataprocessingandstoragetoprovideservicestocloudcustomersandbyservingothercloud-computingcompanies. Itshouldcomeasnosurprise that IBMisinvolved in the cloud; the company has had its fingerprints on just about every deviceassociatedwith thehistoryofcomputing.Inadditionto thestandardbusinessofhostingprovidersofferingapplicationsovertheInternet,IBMiswellonitswayto,inthewordsofoneanalyst,“becomingasortofarmsproviderforthecloud,sellingcustomizedhardwareandsoftwarethathelpsgovernments,largeandmid-sizedcompanies,orWebdevelopers”(Ante 2012). The company is involved in every facet of cloud services, but in 2012 itmadeamajormovetopromoteitscloudtomid-sizedbusinesses,whichmeanttakingonmarket leaders AWS and Salesforce. The company was initially successful, postingdouble-digit gains in its cloud business.However, aswith other firmswhose history ofprovidingsoftwareandotherITservicesprecededthedevelopmentofcloudcomputing,success in thecloudmaycomeat theexpenseof itscorebusiness.Thismajor riskwascaptured in continuing revenue declines in IBM’sGlobal Services unit and in softwaresales.TheproblemforcompanieslikeIBM,aswellasforHP,Dell,andMicrosoft,isthatcloud services can cannibalize their ownkeybusinesses, including selling software andoffering consulting services to help companies run their own IT-linked supply chains.With more and more of IT bumped to the cloud, companies are less likely to requiresoftware and services that maintain their own individual IT silos. According to oneinvestmentanalyst,“Wecouldbeseeingthetipoftheicebergonanimportantdeflationaryforce for traditional packaged applications services” (Ante 2012). That just happens torepresent themajority of IBM’s global-services business. Compounding the problem isthataslongascloudservicesliveuptotheirpromiseofloweredITcostsforcompanies,andsofartheyhave,cloudrevenueforfirmswithalonghistorycannotpossiblykeepupwith what they enjoyed in the past when they sold software and services to a host ofindividualbusinesses.This is not aproblem for companies likeAmazon (with itsAWSoffering), which does not have a legacy business to protect. How IBM, HP, Dell,Microsoft,andnowApplehandlethisclassiccaseofthe“innovator’sdilemma”willgoalongwaytodeterminingwhethertheyhaveafutureofanyconsequenceinoroutsidethecloud(Bradshaw2012).3

Rackspacerepresentsasetofcloudcompanies that,unlikeIBM,doesnothaveeithertheadvantagesordisadvantagesoflegacysystemstoworryaboutandhasmovedfullboreinto providing cloud services. The company, which began in 1998 as a small InternetserviceproviderinfounderRichardYoo’sgarage,quicklygrewtobecomeanestablishedhost for customized applications, providing private, public, and hybrid cloud services.Widely recognized as one of the leading cloud companies and with more than 4,000employees,RackspacereliesonwhatiscalledtheOpenStack,softwarethatisuniversallyavailable based on open source principles.4 In 2012 it approached 200,000 customersusingclose to100,000servers inabout250,000square feetofdata-centerspacearoundtheworld.Demonstrating that it canplaywith theheavyweights, the company’s annualrevenuessurpassed$1.5billion.Nevertheless,withlong-establishedfirmspouringmoneyinto cloud offerings, Rackspace faces an uncertain future. Consider that Dell aloneinvested$1billionintoitscloudin2012.Howdoesafirmthattakesinnotmuchmoreinannualrevenueskeeppace?Additionally,Rackspacebenefitedfromcomplicatedpricingforcompaniesunsureofthetechnologyandthemarketandunabletogaugepricingwell.

Now,asthecloudapproachescommoditizedutilitystatus,withstandardizedpricingbasedonhourlyuseforallcustomers,RackspacewillhaveamoredifficulttimedistinguishingitselffromlargefirmslikeDellandAWS.

Unlike Rackspace, which has grown to become a leader in general cloud services,companies likeSalesforce,whichuses thecloudformanagingcustomers,andVMware,which provides cloud services through virtualized servers, are leaders among thespecialists.5Thegeneral publicbecameacquaintedwithSalesforcewhen it ran twoadscosting $3million during the 2011SuperBowl.MarcBenioff founded the company in1999asoneofthefirsttoofferSoftwareasaService,andthecompanyhassinceaddedPlatformasaService to itsofferings. Its specialty iscustomer-relationshipmanagement(CRM),asystemformanaginginteractionswithclientsandprospectiveclients,primarilytoexpandsalesbutalsotomanagecustomerserviceandtechnicalsupport.CRMhasbeeninusefortwodecadesandisnowexpandingintothecloud.Itoperatesthroughsoftwarethatenablescompanies tomanage their salesandcustomer-serviceprocessesandassesssuccesses and failures. Rather than house CRM internally, companies contract withSalesforce,which provides software and services from its cloud servers. These includestorageforalldataassociatedwithmarketingandsalesforaspecificcompanyandaccessto20millionorsofilesonbusinesscontacts.CompaniescanalsoworkwithSalesforcetodevelop their own applications and tools in the Salesforce cloud. In April 2012, thecompanyemployedcloseto8,000peopleandgenerated$2.25billioninannualrevenues.In2013 it joinedawaveofmerger andacquisition activity in the industryby spending$2.5billiononExactTarget,acompanythatspecializesinmanagingsalescampaigns.Ascloud leaders likeAWSbulkupwith takeover activity,Salesforce felt theneed tokeeppace.Theupsideof specialization is that it enables a company toconcentrate resourcesandexpertise,butthedownsideisvulnerability.Thecompanyfacedthisin2007whenitfell victim to a phishing attack that enabled hackers to lure an employee into revealingcredentialsthatwereusedtogathercustomercontactdata.TheattackerswentontosendfurtherattackstocustomersthroughfakeSalesforceinvoices.Somecustomersfellforthescam and coughed up more information. For a company specializing in the securemanagementofcustomerrelations, thiswasanespeciallydifficultandalmostcompany-destroying failure. Larger firms like Amazon have faced similar challenges, but highlydiversifiedcompanieslikeAmazonarebetterabletoweathersuchstorms.

Theotherchallengeforaspecialistcompanyisfacinggenuinecompetitionfromoneofthe giants that can bankroll a major initiative and keep it going in the absence of animmediateboostinprofit.SuchachallengecamefromMicrosoft,whichmovedintoCRMafter Salesforce but has begun to catch up in customers, markets, and offerings.Moreimportantly,MicrosoftDynamicsCRMcandraw fromusers’ familiaritywithMicrosoftproductslikeOfficeandOutlooktomakethemfeelmoresecureabouttakingtheleapintocloud-basedCRM.Furthermore,becauseMicrosofthasyearsof experience in servicingon-premises IT departments, it can offer clients a mix of cloud and on-premises data-center services. The key point here is that challenges to leading cloud companies likeAWSandMicrosoftdocomefromthediversesetoffirmsinthecloudmarketplace,butbigplayerscanalsorespondpowerfullytoevensubstantialinroadsfromspecialistfirms(CRMSoftwareBlogEditors2011).Undaunted,managementatSalesforceisrethinkingits future by preparing for what it calls Cloud 2, or use of the cloud in social media,

especiallyinmobilecommunication.In2012ittookastepinthisdirectionwiththe$212millionpurchaseofHeroku,aleadingPlatformasaServiceproviderthathelpscompaniesdevelopcloud-basedapplications.

ItishardtodeterminewhetherSalesforcecanwithstandthecompetitivepushfromoneofthegiantsandmoveintonewlinesofbusiness.TheoutcomewillalsodependonhowwellSalesforcefendsoffpressurefromothercompaniesmakingsoftwarecloudservicesakey part of their business. One of the firms to contend with is Oracle, a majorbusinesssoftwareproviderthatuntil2012eschewedthecloudinfavorofsellingsoftwaredirectly to its business clientele. In fact, its CEO, Larry Ellison, is known to havedismissedcloudcomputingasafad.ThesuccessofSalesforceandsimilarcompanieshaschangedthisviewand,afteryearsoffootdragging,thecompanywentonabuyingspreethataddedelevennewcompaniestotheOraclestable,allbutoneofwhichsellssoftwareapplicationsthroughthecloud.6In2013thecompanyextendeditsreachintothecloudbylaunchingasetofpartnerships,includingdealswithMicrosoftandwithSalesforce.Thesedrewalotofattention,especiallyamongthoseconcernedaboutgrowingconcentrationinthecloudindustry(Hardy2013h).AnotherchallengertoSalesforce,theGermansoftwarecompanySAP,hasbeenevenmoreaggressivethanOracle,spending$8billiononcloudsoftwarecompanies.SAPandOracleareespeciallyconcernedthatthecloudwilldisrupttheirtraditionalmodelofprovidingsoftwaretobusinessclients(Waters2013d).Allofthisamounts to both intense competitive pressure in the growing market to sell softwarethroughthecloudandgrowingconsolidationinthecloudsoftwaremarketplace.Althoughanumberofsmallfirmsremain,mostarefacingamalgamationbychoiceornecessity.Asoneindustryexpertexplained,“awaveofdealsislikelytoleaveonlyasmallhandfulofbiggerandmorediversifiedcompaniesstanding”(Waters2012).

TelecommunicationsCompaniesTaketotheCloudFor several reasons, telecommunications companies have an enormous stake in cloudcomputing and they are well positioned to battle the leaders in the industry (Babcock2013b).Itisimportanttounderstandthatthesebusinesses,especiallylargecompanieslikeAT&TandVerizon,arenotjustconduitsforotherfirms’data.Throughtheirsubsidiaries,they are well integrated into the entire digital economy, including content provision.Consequently, the cloud challenges the entire telecommunications industry because itprovides new ways to offer services that have been part of the telecommunicationsindustry foryears.The challengedeepens as ahandfulof integrated conglomerates, thedigital giantsGoogle, Apple, Amazon, Facebook, andMicrosoft, solidify their hold oncloud services. As these firms build towering silos of their own, once-dominanttelecommunications companies are wondering about their place in the cloud economy.Ratherthansitbackandwaitfortheindustrytosettle,firmslikeAT&TandVerizonhavemovedquicklytosecureastake.Verizon,inparticular,hasbecomeamajorleaderamongcloud-telecommunicationsfirmsbyemployingastrategythathasbeenusedoverandoveragainintheindustry’shistory:whenthenextnewthingcomesalong,buyit.Verizondidso in 2011 by spending $1.4 billion on the major cloud company Terremark, and byacquiring the cloud-application firm CloudSwitch to make the total of the company’scloudinvestmentsfortheyearmorethan$2billion.ThesedealstookVerizontothetopofagrowingfieldoftelecommunicationscompaniesthathavemovedintothecloudand,inthe words of one industry analyst “are prepping Verizon for massive future growth”(Hickey2012).Asimportantasitwastopurchasetheseassets,Verizon’smoreimportantchallengewas to integrate them into its other lines of business, especiallywireless andFIOS, its bundled Internet access, telephone, and cable service delivered by fiber-opticcable.

For Verizon, the cloud is a key component of a media, telecommunication, andinformationconvergencestrategythatwillallowthecompanytocontrolpracticallyallkeynodes in the networks that produce, store, process, and distribute services to individualand organizational customers. Moreover, Terremark gives Verizon a significantinternational presence, something that the company has lacked, particularly in LatinAmerica.It isuncertainwhetherVerizoncanmakethisstrategywork.Manycompanies,with AOL Time-Warner the most celebrated, have run aground with “can’t miss”convergencedeals.TheoutcomewillgoalongwaytodeterminingwhetherVerizoncanjoin the leaders in the cloud-based communications industry. Complicating matters forVerizon is the expansion of competitive pressures that threaten its comfortable duopolywithAT&TintheUnitedStates.TheacquisitionsofSprintandofClearwirehavemadeSoftBank,inthewordsofoneanalyst,“abetter-fundednumberthreewiththespectrumtolaunch low-pricedwireless data products.”Moreover, the T-Mobile–Metro PCSmergercreateda fourthbigplayer in theU.S.market and theabilityof the spectrum-richDishNetworkpromisestofurtherdisruptthecomfortablecontrolofthemarketthatVerizonhasenjoyed(GlobeInvestor2012;Taylor2013b).

TheU.S.Government:TrustingtheCloudandCommercialProvidersNot all cloud computing is controlled by private organizations. But it is interesting toobserve the extent to which the U.S. government depends on the private sector for itscloud computing needs, including relationships based on no-bid, sole-supplier contractswith the largest cloud providers.This is significant for several reasons, not the least ofwhichistheamountofmoneyinvolved.Accordingtoonereport,thegovernmentspends$80billionannuallyoninformationtechnologyandplanstomoveabout25percentofitsIT budget to the cloud. An example of themovement to cloud services provided on asingle-source,no-bidbasisistheNavalSupplySystemsCommand’splantouseAmazonWebServicestostoreanddistributedigitalphotographyandvideo.TheNavy’sargumentis that AWS offers a single, integrated package that ismore reliable and less prone toattack than other cloud services (Foley 2012). Furthermore, NASA, which helped todevelopOpenStack,theopensourcestandardthatIBMusesforitscloud,alsocontractedwith AWS (Thibodeau 2013). Even the CIA planned to tender AWS a $600 millioncontractuntilIBMblewthewhistle,raisingquestionsabouthowthefederalgovernmenthandlescloudcontracts,andareviewoftheagreementwithAWS(Woodall2013).WhilewaitingtolearnwhetheritsbidfortheCIA’scloudbusinesswouldsucceed,IBMwonthelargest government cloud-computing contract, worth $1 billion, from the InteriorDepartment (Miller and Strohm 2013). That helped cushion the blow for IBM whenAmazonwasofficiallyawardedtheCIAcontract(Babcock2013a).

These moves are not very surprising, particularly in light of the history of the U.S.government’s relationship to large communication companies (Mazzucato 2013). Foryears, government agencies, including the Department of Defense, had a very closerelationship with IBM for computing and an even closer one with AT&T fortelecommunicationsservices.EvenasbusinessconsumerslineduptosupportbreakingupAT&T and deregulating the telecommunications industry in order to lower prices, theDODarguedthatnationalsecurityrequiredtheend-to-endservicethatAT&Tprovided.ItwasnotuntilthePentagonwasassuredthatsecurityneedswouldbemetthatitdroppeditsopposition to breaking up the telecommunications giant (Schiller 1981). Given thispreferenceforlarge,stablecompanies,itisnotsurprisingthatthegovernmentwouldturntoAWStomeetsomeofitscloud-computingneeds.

TheU.S.government’scurrentmovetothecloudispropelledbythebeliefthatcloudcomputingmustbecomeacentralmeansofmeetingitsinformation-technologyneeds.InDecember2010thefederalChiefInformationOfficersCouncilreleasedaplantoreformgovernmentinformationtechnology,whichincludedrequiringagenciestoadopta“cloud-first”policyfornewITdeployments.Accordingtotheplan,cloud-firstisdrivenbythreeinterrelatedforces.First,largedatacentersprovideeconomiesofscalethatarenecessarytomeetthegrowingneedsofthefederalgovernment’s“computationinfrastructure.”Forfederal ITplanners, it is less expensive to centralizedata in a few large centers than toretain it in local offices. Second, cloud systems are able to provide almost any type ofcomputation on demand. It is difficult to predict the type and speed of processing andanalysis that will be needed and the planners side with those who believe that cloudsystemsareagileenoughtomeettheirneeds,includingthosetheycannotnowanticipate.Finally,thecloudunleashesunprecedentedanalyticscapabilityonlargedatacollections.ItisclearfromthisviewthatfederalITplannersrankbigdataamongthemajorattractions

of cloud computing. Data centers are intended to be not only storage warehouses thatagenciescancallonwhen theyneeddata,but alsoactiveproducersof information thatdraw on stored data sets (Page 2011). In 2011NIST released its report defining cloudcomputingandcarefullydescribingthecloud’sspecificcharacteristicstoenablemanagersandstaffoperatingwithinagenciestohaveabetterideaor,insomecases,theirfirstclearideaofwhatitwastheywerebeingorderedtoimplement.In2012theNationalScienceFoundation (NSF 2012) produced a short report supporting NIST’s conclusions andcommitting thegovernment to fund research into cloud computing.The combinationofstrong affirmations from the federal government’s CIO, fromNIST, and from theNSFprovidedthegroundingforstrongstatesupportforthecloud.

Therearealsomajorimplicationsinanumberofgovernmentdemonstrationprojectsineducation and research. One of the most significant is a program operated out of theNational Endowment for the Humanities Office of Digital Humanities. It demonstrateshow government’s use of the cloud and big data is contributing to the restructuring ofeducation, and not just in the areas where wewould expect change, such as computerscienceandthedisciplinesassociatedwiththesciences.Itisalsoreachingintothesocialsciencesandeventhehumanities.Onecanlearnalotaboutthedirectionofchangefromthesizeofaforcecreatingit,butonecanalsolearnagreatdealfromitsreach,aswhengovernment projects extend to fields traditionally kept outside the scope ofcomputerization.Chapter5examinesthedigitalhumanitiesinthecontextofassessingbigdata in the cloud.Suffice it to sayhere that thedigitalhumanitiesproject represents animportant initiative that is often lost in the understandable focus on largermilitary andcivilianprojects.Itssignificanceforthefutureofeducationandresearchfaroutweighsthesizeofitsbudget(Gold2012).

Inspiteoftheenthusiasmforthecloudingovernment,thereremainseveralissuesthathave the military and intelligence sectors especially concerned about moving data tocorporate-ownedcloudsystems.Arguablythemostimportantissecurity.Attheveryleast,there is concern aboutmoving classifieddata and computer power essential for combatmissions tooff-site locations.Formal concernshavealreadybeen raisedwith respect tothe security of data in NASA’s cloud systems (Kerr 2013). Furthermore, the size andcomplexityofgovernmentandespeciallymilitarycomputersystemsmaketheprospectofmovingtothecloudveryexpensive.Itwouldnotbeamatterofsimplyrelyingsolelyonavailabletechnologiesbecausemanygovernmentdepartments,andespeciallythemilitaryandintelligencesectors,requirecustomizedsystemsthatareintegratedwithinandacrossunits.Finally,government, andespeciallydefense, requiresaveryhigh levelof supportand, while some of the major providers have developed excellent backup for theircustomers, it isuncertainwhether thenecessarysupport isavailable in thecurrentcloudindustry(Gangireddy2012).

Eveninthefaceoftheseworries,thegovernmentisshowingaleveloffaithinprivatecloud companies that has surprised some experts. This extends to using private cloudfirms to provide security for the government’s systems. For example, the Naval WarCollegeawardedasingle-sourcecontracttotheSaaSvendorCloudLocktosafeguardtheimplementation of online tools likeGoogleDocs andGoogleDrive.Given the concernwithsecurity,oneanalyst responds to thisuseof thecloud toprotect thecloudwith theconclusion that “it’s remarkable that agencies are defying conventional wisdom in this

way”(Foley2012).Inamoresignificantstep,intelligenceagenciesarebeginningtomakeuse of commercial cloud computing, including the public cloud, which serves allcustomers. Furthermore, according to one IT leader in the intelligence community,agencies now have enough confidence in the public cloud “to bring some commercialcloudcapabilitiesinsideourfencelines”(ibid.).

ThealternativetothisuseofcommercialcloudservicesistoretainITactivityon-siteorto develop a government, military, or intelligence-agency cloud capability. This iscertainly taking place too. In 2011, Los Alamos National Lab began providing IaaSservices from its own data center and has joined with the National Nuclear SecurityAdministration to develop a community cloud that extends to the entireDepartment ofEnergy(ibid.).OfgreaterstrategicsignificanceistheDepartmentofDefensedecisiontocreateamilitarycloudasameanstofendoffcyber-attacksthathavebeenproliferatinginrecentyears.TheseincludetheApril2010attackemanatingfromChinathatredirected15percent of Internet traffic throughChina’s networks for eighteenminutes and the 2011virus attack onU.S. droneweapons. The latter usedmalware to record keystrokes andrequired continuous deletion and rebuilding of hard drives. To avoid these attacks,DARPA set up Cloud to the Edge (COE) in 2011, which began by opening a set ofhotspots for secure communication. According to one analyst, COE looks a lot likeGoogle’ssuiteofonlineservices,minusthesearchengine(Tanaka2012).Itishostedonasecure systemof servers by theDefense Information SystemsAgency,which has itselfgivenouta$45millionsole-sourcecontracttotheAllianceTechnologyGroupforadata-storagefacilitytoprovidefourexabytesofstoragecapacity(Hoover2013).Tobackupitscloudinitiative, theDepartmentofDefensecommittedanother$5milliontoadvanceitscyber-battleground project, with the auspicious title of Plan X, that would allow theagency“torehearseandmanagewhatofficialscall‘cyberwarfareinreal-time,large-scale,and dynamic network environments’” (Nextgov 2013).7 To implement its plan thePentagonwill hire and deploy 4,000military and civilian technology specialists to theU.S.CyberCommand,butthatisnotlikelytobeenough(Brannen2013).Thispromptssometoanticipateanear-termshortfallincloudexperts(Weisinger2013).

Itisnotjustsecuritythatpromptsinterestinthecloud.TheDODalsowantstobettermanageitsITbudgetandhopestheturntocloudcomputingwillgoalongwaytosaving30 percent by 2016. Already engaged in the consolidation and modernization of datacenters, theDODhaseliminatedmanyandcutthenumberoftechnicalsupportdesksinhalf.Overall,itwouldliketoreducethenumberofnetworks,datacenters,andhelpdesksby 80 percent (Tanaka 2012). Storing everything from unclassified to top-secretinformation,themilitarycloudbeganwithatestcaseledbytheNationalSecurityAgency,whichgathers,stores,processes,andanalyzeshugeamountsofdata.TypicallyshelteredfromthepublicattentionthatismoretypicallydirectedattheCIAandtheFBI,theNSA,which is three times the size of the CIA and has one-third of total U.S. intelligencespending,burstonto the frontpagesofnewspapersworldwide in the springof2013.Aseries of leaks and newspaper accounts revealed that, contrary to previous claims, theagencyworked closelywithU.S. telecommunications providers and the largest Internetcompanies to gather data on Americans and foreigners by scooping up and analyzingtelephone conversations, emails, social-media postings, and other electroniccommunication.With the$20millionPrismprogramthat includedmajorSiliconValley

and telecommunications companies that shared information on userswith theNSA, thespyagencyhopedtobettertargetthreatstotheUnitedStatesbyanalyzingmetadata—thatis,whowascontactingwhom,aswellascontentwhosekeywordsbig-dataanalystscoulduse to root out suspected terrorists (Luckerson 2013). Nevertheless, many critics tookissuewithwhatappearedtobeanunprecedentedand,untiltheleaks,secretiveattackontheprivacyofusers (WilsonandWilson2013).Controversies aside,governmentpolicymakers hope that cloud computing will enable the NSA to meet its goals with greatersecurityandat lowercost, therebydemonstratingthevalueofmovingothergovernmentagenciestothecloud.Nevertheless,expertsworrythatconcentratingmilitaryinformationin one large cloud system, howeverwell secured, provides an inviting target for cyber-attackersaroundtheworld.Oneexpertworriesthatthemovetothecloudistheequivalentof“paintingacyberbullseye”ontheNSAandthemilitary:“Cloudcomputing,inmilitaryterms, fosters a target-rich environment because the very things that make the cloudappealingalsomake it a temptingmark.Becauseof this and thehighprobability that avast amount of datawill be stored on a cloud, attackers only need to be lucky once ascomparedtohavingtobeluckymultipletimeswhenattackingalegacysystem.Withthisinmind, amore appropriate question for theNSAwould be ‘what kind of informationwouldyourorganizationrefusetoplaceonaCloud?’”(Tanaka2012)

It is not as if military planners are unaware of the security problems of cloudcomputing.AccordingtoDARPA,“Cloudcomputinginfrastructures,inparticular,tightlyintegratelargenumbersofhostsusinghighspeedinterconnectionfabricsthatcanservetopropagateattacksevenmorerapidlythanconventionalnetworkedsystems.Today’shosts,ofcourse,arehighlyvulnerable,butevenifthehostswithinacloudarereasonablysecure,any residual vulnerability in the hosts will be amplified dramatically” (ibid.).Nevertheless, like many other agencies, it is convinced that, with appropriate securitymeasures, themilitarybenefitsofcloudsystemsoutweightherisksbecause“cloudsanddistributed computing environments can: provide redundant hosts, correlate attackinformation from across the ensemble, and provide for diversity across the network”(ibid.).Whatmatters for themilitary is whether it can developwhat it calls “mission-orientedresilientclouds”thatcanbedeployedeffectivelyincombat.

CloudsoverChinaCloud computing systems have a firm foothold in the United States, where about 40percentoftheworld’sdatacentersarelocated,buttheyarealsospreadinginternationally.Outside the United States, Scandinavia has become a major data-center venue and thecloudisnostrangertotheMiddleEast,butChinahasmadethemostsignificantprogressintheoveralldevelopmentofcloudcomputing(Horn2011;Glover2013).Bytheendof2012Chinarepresentedabout3percentoftheglobalcloudmarketplace,butitisexpectedtogrowata40percentannualrate,reaching$18.6billioninannualrevenuebytheendof2013.Led byChina, theAsia region is expected to lead theworld in cloud traffic andworkloadsby2016(Ong2012).China’sburgeoningcloudindustrybenefitsfromminimalcompetitionwith themajorU.S. providers.Amazon is not there andMicrosoft has justbegun to introduce itsAzure cloud service inChina. This has left lots of room for thedevelopmentof indigenouscloudservices, includingtheAlibabaGroup,whichprovidesboth cloud infrastructure and services to a variety of national andmultinational clientsoveritsAliyunnetwork.Inaddition,Baidu,knownintheWestasthe“GoogleofChina”for itsprowess in searchservices,has investedheavily incloudstorageandprocessing,evidencedina2012investmentof$1.6billioninanewdatacenterandadealtoofferfreepersonal cloud storage onAndroid phones. Baidu’smajor competitive challenge comesfromTencent,aninstant-messagingandonline-gamingcompanywith400millionusers,making itoneof the largest consumer-applicationcloudcompanies in theworld,withavaluation in 2012 of $60 billion. In 2013 Tencent took a major leap in the cloudmarketplacewhenitannouncedthatitwouldbethefirsttobuildacenterinthewesternChina city of Chongqing, where planners expect significant new growth in the cloud(People’s Daily Online 2013). In 2012, the world’s leader in telecommunicationsequipmentproduction,Huawei,alsomovedintocloudcomputingandstorage,adecisionthat led to a significant growth in company profit (Reuters 2013a). China’s clouddevelopmentishelpedbythepresenceofAsianfirmslikePacnetthatbenefitfromhavingdeveloped network and data-center services in theAsian region, includingHongKong,Singapore,andAustralia(Powell2013).

In2013,BaidudemonstratedthatitdoesfarmorethanprovideservicetoChinawhenitsignedadealwithFranceTelecomtoofferitsmobilebrowserthroughoutAfricaandtheMiddleEaston theFrenchcompany’ssmartphones (Thomas2013). Inaddition to thesenetwork-drivencloudproviders,companieshaveemergedthatprovidestorageservices.Aleader in this area is MeePo, a storage service similar to Dropbox. The company hasexperiencedremarkablegrowth,withcapacity in2012reliablyestimatedat50 terabytes(Chou2012).

OneofthemostambitiouscloudprojectsintheworldisChina’scommitmenttobuildcloud cities. The goal is to construct giant data centers connected to firms that providevalue-addedservices,aswellasresearchanddevelopmentfordomesticandinternationalmarkets. Some of these involveworkingwithmajor international partnerswho providecapital and expertise, even as local companies control the project. For example,China-basedRangeTechnology is teamingupwith IBM toconstruct a6.6-million-square-footcloud-computing center in Langfang, near Beijing. It will provide cloud services togovernment and private-sector organizations, aswell as host cloud systems andmobiledevices(BundyandHaley2012).Inadditiontolinkingupcomputer-serviceproviderslike

BaiduandcomputercompanieslikeLenovo,cloudcentersalsowelcometheinvolvementof China’s large telecommunications companies. For example, in 2011 China Telecomformedapartnershipwiththeglobalcloud-servicescompanySAPtooffercloudservicesto small and medium-size businesses in China. In 2012 the country’s three gianttelecommunications firms,ChinaTelecom,ChinaMobile,andChinaUnicom,agreed toinvest $47 billion to develop data centers, including one of theworld’s largest, to helpcreateaneconomichubinChengdu,acityinChina’ssouthwesternprovinceofSichuan.Chengdualreadybuildsone-fifthoftheworld’scomputersandtheplanistoexpandtheTianfuSoftwareParkaroundtheclouddatacenters.InthiswayChengduwillmoveupthevalueladderfromcomputermanufacturingtodatastorage,processing,andtransmission,on theway tobecominga center for researchanddevelopment (Evans-Pritchard2012).With fifty-one universities graduating 200,000 scientists and engineers each year,Chengduhasthefoundationtotakethesestepstohigher-valueproduction.

Chinacertainlyappearstobepoisedtobecomeaworldleaderincloudcomputing.Itisbuildingenormousclouddatacenters,includingsomeoftheworld’slargest,atafeverishpace. Not satisfied with the construction of cloud facilities, it is creating entire cloudcities.Of equal significance, China is carrying out a detailed cloud-computing strategythat is most significant for integrating all the major participants, including hardwaremanufacturerswhoarebecomingleadersinserverproductionfortheglobalmarketplace,software designers, application developers, business-service providers, andtelecommunicationscompanies.Butthereisanothersidetothissuccessstory.Chinafacestechnical challenges, including connectivity problems and the absence of certificationprograms for cloud companies and their staff, something that has been institutionalizedamong leading companies like Amazon. Moreover, as Chapter 4 describes, cloudcomputing faces numerous environmental, social, and labor challenges. These are allgreatlyheightenedbythesizeandspeedofclouddevelopmentinChina,aswellasbytheunsettlednatureofitspoliticalandlegalinfrastructure(Qian2013).

Cloud computing creates significant environmental problems associated with itsmassiveenergyrequirementsand,secondarily,withconstructionanddisposalofmaterialsandequipment.TheseareallexacerbatedinChinabecausethecountryisalreadyplaguedbywidespreadairpollutionasenergyneedshavespikedacrossthecountry,andrelianceon coal-fired power plants deepens the problem. Building the world’s largest cloudfacilities, includingentirecloudcities,willonlyadd toanalreadycriticalproblem.Thesame holds for security, surveillance, and privacy issues. These pose challengeseverywhere, but nowheremore prominently than inChina,where there is no guaranteethatiftheybuildit,theworldwillcome.Chinahaslongbeenmiredincontroversiesaboutthesecurityofpersonalandorganizationaldata.WillWesterncompaniesandgovernmentsthat have complained about the theft of data store their information in China’s datacenters?A society that practicesmassive surveillance of its own citizens and routinelycensors information can hardly be surprised to find very low trust in the security of itscloud systems. It is not only foreign businesses thatworry about surveillance issues.A2013ForresterResearchreportdocumentedconcernsamongChineseentrepreneurswhoare reluctant to take to thecloud.Someof this results from the lackofexperiencewithoutsourcing or externally managed services. With little to prepare them for the cloud,companiesareunderstandablycautious.Butsecurityworriesloomlargeandthisaccounts

for a distinct preference for private-cloud services as the less risky cloud option (Qing2013). Finally, China’s hyper-accelerated industrialization has created massive laborproblems that include but extendwell beyond the notorious practices of the electronicsmanufacturerFoxconn.Annuallyproducing200,000scientistsandengineers inonecityaloneisanoutstandingachievement,butmanagingthemandthemillionsofnewworkerswho constitute China’s army of knowledge workers is an entirely different challenge.From construction to operation, from maintenance to support, cloud computing makesenormousdemandsonlabormarketsandworkplacepractices.Toaddthesedemandstoasociety already in the throes of labor upheavals across the country will certainly taxChina’sleadershipforyearstocome.

Thisoverviewofcloudcomputinghascoveredkeyfeaturesofitsgenealogy,definingelements, key characteristics, and major exemplars. The next chapter builds on thisfoundationbyexamininghowcloudcomputingispromotedinmarketingandmyth,anddescribeswhyitisimportantforsupporterstofashionthiscomplex,butnonethelessbanal,technologyintothetechnologicalsublime.

CHAPTER3SELLINGTHECLOUDSUBLIME

Windowsgivesmethefamilynaturenevercould.

—TelevisioncommercialfortheMicrosoftcloud

Howdoesamassivedatafactorygiverisetotheimageofacloud?Althoughthemetaphorofthecloudcameupfromtimetotimeinearlydiscussionsofcomputingatadistance,theimmediatereasoncanbefoundinmost technicalprimerson thesubject: the imageofacloud was used in diagrams to describe the interconnected elements of a computercommunicationsnetwork.Withitsstartinthebanalityofatechnicaldiagram,theimageofthecloudhasgrowntotakeonaricheraestheticthatcorporatemarketinghastakentheleadinbuilding.Toappreciatethesignificanceofcloudcomputing,it isimportanttogobeyondwhatthemanytechnicalbooksdescribetounderstandhowitisbeingconstructedin discourse and sold to business, government, and individual consumers because thesetoohelptoshapewhatcloudcomputingmeans.Thematerialityofthecloudisnotlimitedto data centers, computers, software, applications, and data. It is also embodied incampaignstoremaketheprosaicstuffofengineeringintothecompellingaestheticofthecloud. Just as it was important to describe the technical, political, and economicdimensionsofcloudcomputinginthelastchapter,itisalsoessentialtoexaminehowitisbeing sold in advertising, social media, private-think-tank reports, intergovernmentalreports,lobbying,andtradeshows.Discourse,myth,andmagichavealargeroletoplayincreatingthecloud.

Itdoes sometimes feelas if technologiesappear likemagic,notdeusexmachina butmorelikemachinaexdeo,asmachinesemergingfromthegeniusofinventors(preferablyworkingintheirparents’garages).Evenwell-regardedbiographieslikeWalterIsaacson’s(2011)onSteveJobscannothelpbutbuildashrine,evenastheytellagoodstory.Indeed,whenitcomestotechnology,theshrineappearstobeanessentialpartofthestory.Mythscelebratethismagicanditisimportanttotakethisprocessseriouslybecauseithelpsustounderstandhowwethinkandfeelaboutthecloud.Butitisalsoimportanttodrawbackthecurtainonthisversionof“thegreatandgloriousOz”andrevealtheprocessthatgiveslife to themagic.AsChapter2described, cloud computing ismadeupofdata centers,servers,software,applications,anddata,allofwhicharedesigned,built,andoperatedbythousandsofworkers,rangingfromhighlyskilledengineerstounskilledlaborers.Theseprovidethefamiliarfoundationsforsuccessfulcloudsystems.Butthecloudisalsomadeupofwords,startingwiththenamecloud,aswellastheimagesanddiscoursesthatgiveshapeandformtohowwethinkaboutcloudcomputing.Putanotherway,technologyisnotonlycomposedofthematerialthatentersitscreation;itisalsodefinedbythelaborofthose who design, build, and operate it and by the language we use to describe and

imagine it. More formally, technology results from the mutual constitution of objects,labor,andlanguage.Thischapterfocusesonhowcloudcomputingiscreatedinlanguageanddiscoursebyconstructing,withaneyetoselling,thecloudsublime.

Assessing the effort to sell cloud computing is important because companies in thecloudbusinesshaveasteepclifftoclimbiftheyaretoconvincecompanies,governmentagencies, and individual consumers to signup.That is because selling the cloudmeansconvincing a potential client to give up its data on employees, customers, products,services, and competitors and trust that it will be available when needed. This raisesquestionsaboutdatasecurity,theprivacyoftransactions,systemreliability,andthefutureof the client’s IT unit. Businesses and government agencies know that when a cloudcompanytriestosellthemontheideaofendingdependenceonseparatedatasilos,theyarenotjusttalkingaboutcostsavingsandefficiency.Italsomeansafundamentalchangein how the business or agency works, how it is organized, and how power flowsthroughoutitsstructure.

ThereisalsonothingsimpleaboutwinningoverindividualconsumerstousethecloudforanythingmorethanthemostbasictaskslikeGmail.Whystorefiles,audio,andvideoinanunknownlocationwhenyoucanleaveitallonyourowndeviceorbackituptoaportable external drive?Aside from the fees charged for cloud storage, people wonderaboutthewisdomofgivingacompany,evenonewithagoodreputation,photographsofyour family, your treasured music collection, personal email, and sensitive files.Companiesmaypromisethatyourfileswillbesecureandavailabletheinstantyouwantthem, but just how secure are they, how reliable is the service, how private are yourcommunications, and what will your cloud provider do when a government agencydemandstoaccessyourfiles?Whathappenstoyourdataifthecloudcompanyyoudealwithgoesoutofbusiness?Thedecisiontoenter thecloud,forbusinessesaswellasforindividuals, isfarfromautomaticandcertainlynotsimple,andsoithastobepromotedvigorously.

AdvertisingtheCloudOnFebruary 6, 2011, aworldwide audience in the hundreds ofmillions, including 111millionAmericanswatchingonFox, to that date the largest audience inU.S. televisionhistory,settledinfortheannualspectacleoftheSuperBowl.InadditiontothemanykeentoseewhetherthePittsburghSteelersortheGreenBayPackerswouldwintheLombardiTrophy that goes to the annual winner, there were many whose primary interest waswatchingandassessingthecommercials.Giventhesizeoftheaudienceandtheintensityofthespectacle,sponsorssavetheirbestadsforthebigevent.Amongtherecognizedbestthat yearwere aVolkswagen commercial featuring a youngDarthVader practicing theForce,anadfortheChevyCamaroaboutthedemurebutdangerousMissEvelyn,andaplugforCoca-Cola thatbrought togetheradesertborderguardandadragon.Given thepriceofbuyingtheattentionofall thoseviewers, it isnotsurprisingthatbigcompaniesdominatetheads,paying$3millionforaspot.

Sandwiched among these tall treeswas a lesser-known company that sold somethingmanyAmericansusedbutknewverylittleabout.ForitwasatSuperBowlXLVthatthecloud-computingcompanySalesforcedebutedtwoadvertisementsforitsChatterservice.Bothcommercialswereslickanimationsmeanttointroducetheaudiencetoafreeprivatenetwork for businesses that could use the cloud to help internal communication andcollaboration,aswellasexpandtheirreach.TheadsfeaturesWill.i.am,leadsingerofthemusical group Black Eyed Peas, who poses the question, “What do you think of thecloud?”whichleadstoatourofChatter.ThespotfocusesonChattytheCloud,whohelpskeep the band “in line… and on the same page,”with bandmembers communicatingabouttourupdates,“flyshoedesigns,”andnewDJgigs.Allofthisisdoneincompleteprivacyandsafety.Thesloganthatendsthespot,“Doimpossiblethingsasateam,”marksthedifference thecloudmakes(Chatter .com2011a).ThesecondspotfeaturesWillandthePeasdemonstratingsomeofthoseimpossiblethings, includinggettingajob, turninganoldfactoryintoaSiliconValleyworkplacepalace,andfindinggreatclothes.Itendsonthemost impossiblenoteof all, asChattybrings together thewarringRepublicanPartyelephantandtheDemocraticPartydonkeyinaconciliatoryembrace(Chatter.com2011b).EachadpointedtoawebsitethatprovidedmoredetailsabouttheSalesforcecloud.

These ads were unusual because they promoted a specialized business product, asopposed to the more typical consumer goods and services aired during most mass-audience events. Chatty the Cloud was not the stuff of hot cars and beer, nor even ofGoDaddy,whoseadsapproach—andsomesaycross—theboundaryofpermissiblesexualcontent.NorwasChattyacriticalsuccess.Mostanalystsdidnotrelegatethecommercialsto the trash bin of Super Bowl failures, but the lack of punch did not make themmemorable. If anything, theSalesforce ads succeeded in letting the audienceknow thattherewas this new thing called the cloud,whichmattered enough thatwell-knownpopmusiciansgaveitaride.

Arguablythead’smost importantpoint,whichoccupiesthecritical lastframes, is theproclamation“Doimpossible thingsasa team,”because it turnscloudcomputing intoamyth.Inthiscaseamythdoesnotrefertosomethingthatcanbejudgedbywhetheritistrueorfalse.1Afterall,doingtheimpossibleisbydefinitionafalseproposition.Rather,myths are judged by their resonance: not by their truthfulness, but bywhether theyare

aliveordead in thepopular imagination.Theyare thestorieswe telleachother tohelpdealwithlife’sunanswerablequestions,andwhenitcomestotechnology,theyhelpraiseourlatest“nextnewthing”totherealmofthetranscendent.Mythsprovideballastforthesublime but fleeting visions contained in the promise of universal knowledge, virtualworlds,andunlimitedcommunicationthatwereonceembodiedinreligionandnaturebutarenowmorelikelysupportedbydigitaltechnologies.Theassertionthatcloudcomputingenablesagrouptodotheimpossibleissimilartotheclaimthatthetelegraphwouldbringworldpeaceorthatlightingupthestreetswithelectricitywouldendcrime(Nye1994).Itis not an exaggeration to suggest that we make myths whenever we make newtechnologies,andfurthermorethattechnology,especiallycommunicationtechnology,hasbecome,likereligionandthenaturalworldbeforeit,asourceofthesublime.Today,manypeoplehavebecomecloudworshipers(Lohr2013c).

Specifically,technologybecomessublimewhenweattributetoitsuperhumanpowers,eitherheavenlyorhellish,thatwereoncereservedforreligionandthesupernatural,orfortreasurednaturalwonders.Priortothedevelopmentofthetechnologiesthatpropelledthemodern age, such as the railroad, the telegraph, and electrification, the sublime wasassociatedwithimagesoftranscendencelocatedinreligionandinnature.First,onlythegodsorgodcouldachieve the levelof transcendence that transportedpeoplebeyondalllanguage and certainly beyond the banality of everyday life. The very name Yahweh,accordingtoonecommentator,“alsobespeakstheuttertranscendenceofGod.InHimself,God is beyond all ‘predications’ or attributes of language: He is the Source andFoundation of all possibility of utterance and thus is beyond all definite descriptions”(Parsons2013).The“unutterablename”conjuresboththerapturousaweandtheterrifyingshockofthesublime.Formany,thereligioussublimeincreasinglycametobemetbythenatural sublime as wonders of nature like the Grand Canyon, natural eruptions likeearthquakesandvolcanoes,orthecelestialmagicofasolareclipseconjuredsomeofthesame awe-inspiring and fearsome feelings. While certainly not eclipsing the religioussublime,whichremainsapowerfulforcearoundtheworld,thenaturalsublimehasgrowninsignificanceandmaycontinuetodosoastheterrifyingresultsofclimatechangeleavemore and more people speechless. But as Rebecca Solnit (2010) has described in herremarkablebookabouthowpeoplerespondtounspeakabledisasters,thereisconsiderableevidencethat,justasanextraordinarilyresilientcommunityaroseinthetimeofYahweh,weobservesublimeactsofcommunity-buildinginanageofnaturaldisasters.

Itisdifficulttocapturethesublimeinwords,sowerefertoimageslikethelocomotiverippingthroughaprairiefieldinthe1870sthatterrifiedonlookersintoasublimestupor,ortheGrandCanyonresplendentinthemorninglightthatrenderssublimeitsseeminglyinfinite layers of contrasting color. Some of the great modernist writers could create asublime riff through a stream of consciousness such as the one described by VirginiaWoolfinthecharacterofClarissaDalloway,whotranscendstimeandspaceinamorningwalkthroughLondon.IndeedoneofMrs.Dalloway’smostsublimescenesinvolveswhathastobeoneofthefirstactsofcloudcreation,asasky-writingairplaneliftstheeyesandthentheheartsofobserversonthegrounduntiltheairplanerevealsacommercialpurposein a banal advertisement for candy. Echoes of this cloud return in the work ofphotographer Sergio de la Torre, whose 2003 digitally constructed skywriting cloudexclaims against an azure sky, “Thinking About Expansion.” Today, great popular

filmmakers likeStevenSpielbergconjure thesublime inwhathasbeendescribedashissignature visual technique,what the filmwriterMattPatchesdubs “theSpielberg face”(thinkoftheyoungDakotaFanninggazingbackthroughtherearwindowofherdad’scarasthecarnageeruptsinthefilmWaroftheWorlds),andisbestdescribedinavideoessaybyKevinB.Lee:“Eyesopen,staringinwordlesswonderinamomentwheretimestandsstill.Butaboveall,achildlikesurrenderintheactofwatching”(Scott2012).

Yes,doing the impossible is amarketingexaggeration—somewould sayamarketingconvention.But it is also the foundation for amyth that asserts superhumanor sublimeprowess.ThehistorianofcultureLeoMarxputitbestwhenheassertedthat“therhetoricofthetechnologicalsublime”involveshymnstoprogressthatrise“likefrothonatideofexuberantself-regardsweepingoverallmisgivings,problems,andcontradictions”(1964,207; see alsoMosco 2004).No longer locked intowhat EdmundBurke 250 years agocalled the “stale, unaffecting familiarity” of the banal (Burke 1998, 79), the sublimetechnologybecomestranscendent.

The Salesforce ads had the widest reach of any cloud-computing spots, but thatcompany was not the only one to hitch its wagon to the National Football League’s(NFL’s)star.Infact,strangeasthismaysound,oneofitsmajorcompetitors,SAP,paidforthe right to be the “official cloud solutions software sponsor” of the NFL. Nor wasSalesforce the first to try to sell the cloud to amass audience. In earlier adsMicrosoftsought to reach its business and consumer audiences separately. The software giant’scampaignreflectsoneofthechiefmarketingchallengesthatcloudcompaniesface:howtosellaserviceaimedatbothcorporateandindividualconsumers.Microsoft’sanswerwasseparatetracks,bothofwhichattempttoadvancethemythoftheall-powerfulcloudbut,aswithmostcampaignsthattrytoselltechnology,withmoremagicinthesalespitchtoconsumers. In 2010 with the unveiling of its new business cloud services, Microsoftdevelopedcommercialsaround“cloudpower.”ThetypicalonefeaturesmanagersandITprofessionalsboastingaboutallthattheycandowiththenewforceofthecloud:

Icanchangehoweveryoneworks…withoutchanginghoweveryoneworks.

Icanturnaspikeindemandintoajoyride.

Icanexpandoverseas,overnight.

Icantakeappsliveinfifteenminutes.

Iammasterandcommanderofmyownprivatecloud.

Iamthechampionofacorporatecultureofyes.

Ihavecloudpower.Ihavecloudpower.Ihavecloudpower.

Themostcomprehensivesolutionsforthecloud,onearth.Microsoft.

Theadissimple,withthesublimityrestrainedbywinkingreferencestobringingaboutchangewithouttheunsettlingexperiencethatchangeofteninvolves.Theemphasisisonpleasure.TheequallyunsettlingspikeindemandisnolongeranITnightmare;itisajoyride. Global expansion takes place overnight and software that would inevitably bringbug-infestedturmoilgoesoutinjustfifteenminutes.Executivesarenow(Jedi?)masters

and thebosswhoseworkday is typically filledwithpainful repetitionsof“no”cannowsay“yes”overandoveragain.Thesepromisesarethestuffofcorporatemyth,butnotsoutopian as to undermine a sales pitch to knowing executives. They remain within therealm of the real, the heavenly cloud on earth, delivered by Microsoft. The productsthemselves—Windows Azure, Microsoft Office 365, and Windows Hyper-V Server—makethebriefestofappearances,withtheemphasisplacedonthevalueofthecompanyproviding the service. Microsoft itself is mythologized in the faces of the presenters,whosenationalitiesandspeechdescribeacompanythatcoverstheentireglobe,fromtheBritishexecutivewhoturnschangeintoajoyridetotheIndianbusinessmanwhocangoglobalovernight.Itmaybethecloudonearth,buttheearthiswellcoveredbyMicrosoft(Warren2010).

A bit later Microsoft released a very different advertisement that, understandably,aroused controversy. This campaign directs individual consumers who learn that thesolution to problems involves a trip “To the Cloud.” The ad features aMom carefullyexaminingthefamilyphotoshewantstopreserve.Andwhatanunrulyfamilyitis!Withafrustrated lookonher face,Momcomplains because her daughter is texting, her son issticking an action figure inhis brother’s ear, andDad is trying to remove it (the actionfigure, not the ear).The solution is found in taking a trip “To theCloud,”whereMom“cantakealltheseunrulyshotsandswapinsomesmiles.”Thismeansdownloadingfromthecloudimagesofeachmemberofherfamilyinhappyreposeandplacingthemintheappropriatespotstoperfectherfamilyphoto.Momisthrilledbecause,thankstothecloud,“finallyaphotoIcansharewithoutridicule.”Asif thiswerenotenough,Momturnstoher“real”family,whichislookingbored,exceptforDad,whoappearsdepressinglysad,ashebowshisheadinshame.ThespotendswithMom’sthankfulconclusion:“Windowsgivesmethefamilynaturenevercould.”2

Thisadstirredsomecontroversyforacoupleofreasons—primarilybecauseitfeaturesamomwhoappearstopreferadigitallyalteredversionofherfamilytoher“unruly”realone.Whoneedsthefamilynatureprovidedwhenthecloudcansendyouabetterone,oratleastonefreefromtheall-too-humanpropensitytoactautonomously?AddedtothisistheadmittedlygeekyconcernthatMomhasnotreallygonetothecloudatall.Bothconcernsare captured in the comments of one apparent cloud enthusiast: “Wrong, bad andmisleading.Thishasnothingwhatsoever todowithCloudServices.A touchscreenPCandWindows7runningLiveEssentialsLOCALLYtomashupafamilypicture.Where’sthe‘Cloud’inthat?Theonlyon-lineexperienceispushingtheresultingpieceoffakerytoFacebook.TheseadvertsaretheworstpieceofCloudwashingaroundatthemoment.”3Itishard toknowwhat ismorephony,Mom’snewWindowsfamilyorMicrosoft’sclaimthat she has taken to the cloud. One might reasonably argue that the charge ofcloudwashing,thetermappliedtoidentifyingasacloudservicewhatisreallysomethingelse, isaquibble.Momappears tobegoing toacloudservice to findmoreappropriateportraitsofherfamily.OnemightalsomakethecasethattheclouddoesexpandMom’soptions.Shecouldusethe“unruly”photoortheonethatthecloudmakespossible.Whatmakesthisadvertisementparticularlyinterestingisthatitexplicitlyrepresentsthetriumphofthetechnologicaloverthenaturalsublime.

Itwouldbeeasy tooverinterpret thisad.Afterall, it’s justacommercialand, likesomany that came before, it uses provocation, simplification, and exaggeration to send a

message,keepitintheviewer’shead,andsellaproduct.Butitisalsosafetosaythatthisparticularspotisdoingmorethanjustsellingviewersonthebenefitsofcloudcomputing.Italsoreflectsandadvancesastreamofthoughtgenerallydescribedaspost-humanism,aphilosophical perspective that questions the human-centered values that emerged in theRenaissance and became foundational in Enlightenment thought (Hayles 1999). Thehumanistwouldfinditbeyondthepaleforanyone,letaloneamother,toreplaceaflesh-and-bloodfamilywithatechnologicalsubstitute.Forthepost-humanist,itisnothingmorethanacceptingtherealityofourtimeandusingittohumanadvantage.Ratherthanfeelingguilt for appearing antihuman, Mom is justified in demonstrating the pride ofaccomplishment for acceptingwhat technology has to offer. Post-humanist thought hashelped to shape howmany people, and not just professional philosophers, think abouttechnology, whether it has to do with biological or informational systems. Some havetaken serious issue with the position, arguing that it is little more than surrenderingprogressivevaluestoawealthmachinemasqueradingasanewphilosophyoftechnology(Winner2004).Thedebatearoundpost-humanismhasilluminatedmanyissuesthatarefarmore serious than the ethics of image substitutes. These include using technology toextendlifeorendit,tobeginlifeortoterminatelifebeforeitemerges.Theyalsoaddressrobotics, automation, and the opportunities and threats posed by thinking and feelingmachines.The“Mom’sFamily”spotis justanad,butitalsoprovidescloudcomputing,andMicrosoft,withtheopportunitytotakesidesinagrowingdebate.Clearlytheychoosethe technological over the natural sublime and, in doing so, advance an increasinglypowerfulmythorstorylineinglobalculture:thesuperiorityoftechnologyoverhumanityasaninstrumentoftranscendence.

Microsoftchosetopursuebothbusinessandthehomeconsumerinitscloud-computingmarketingcampaign.TwootherITgiantschoseadifferentstrategy,withIBMfocusingonbusinesscustomersandAppleconcentratingontheindividualconsumermarket.Theyarebothsignificantforunderstandingthediscursiveconstructionofthecloud.Fromitsfirstcommercials in 2009, IBM has had the longest run among major corporations ofadvertisingcloudcomputing.Despiteasimplemessageanda relianceon talkingheads,IBM’sadsareamongthemostlavishlyproduced.The2012ad“AllintheClouds”usedthirty-two animators, designers, illustrators, andmodelers to bring to life an imaginaryworldthatexistsonlyinthecloud(IBM2012a).Accordingtothecompany,“Everythingin the spot was painted by hand and then mapped onto 3D wireframes to create thecompletelybespoke look.Eachcharacterhas abackstory,which sparked the animators’imaginations.Every ‘location’was extensively researched tomake sure the transformedworldlookedliketherealone”(Marshall2012).Thecommercialbeginswithavoiceoverwelcomingviewers“tobusinessasusual,”thanksto“theIBMsmartcloud.”Aswesee,however,businessasusualisanythingbut.ThefirstanimationswoopsdownonasmalllaboratoryinBerlinthatusescloudcomputingtofightcancer.ThesecondlandsinChina,wherethecloudismakingitpossibleforanindustrialcitytobecomeahigh-techhubinlessthanfouryears.Thenit’sbacktotheWest,whereBritainisbuildingasmartgridtohelp cut emissions by up to 80 percent. Finally, we return to Asia, where “even anindependentstudioinMalaysiacanproducebig-timeblockbusters.”Thenweseetheonenonanimated human in the commercial, a woman who, behind studious-lookingeyeglasses, announces, “Transforming business through the cloud. That’s what I’mworkingon.I’manIBMer.Let’sbuildasmarterplanet.”

Smart is the operativeword in these IBM commercials and, even though the visualshavegottenmoresophisticatedandexpensive,themessageremainswhatithasbeensincetheseriesbeganin2009:computers,includingthecloud,willconstructasmarterplanet.IBMdepartsfrommostcloudcompanies,whichhavebeenreluctant toadvertisewidelyfortheirbusinessservices.Indeed,oneindustryobserverhaschidedcloudexecutiveswho“think their totallyawesome solution canmarket itself.Oncepeople try it and see howgreat it is, they’ll be sure to tell their business associates—right?Well, actually…no”(Shaw2013).IBMhascertainlynotbeenamongthereluctantand,inastepunusualformost of today’s advertisers, whatever the business, it has deployed a major printadvertisingcampaignaroundthecloud.Thisincludesaprintcampaign,whichproclaims,incaseyou forgot theoperative term,“SmarterTechnology for aSmarterPlanet.”Likemostofthecompany’sads,itdoesnotshyawayfromcleverhyperbole:“Foratechnologythat’sbuilttobeinvisible,cloudcomputingismakingsweepingchangeseverywhereyoulook.” It then demonstrates this with examples from “themainstream” to “the revenuestream”(IBM2012b).

For IBM,by2012 the cloud, ormore specifically, the IBMSmartCloud, had alreadymadeitintothemainstream,bringing“achangeintheatmosphere.”Companiescannowsellseafood“freshoffthehook,”engineerscancreatenewmedicinesfromgenomics,andtennistournamentscanserve“dynamictensionoutsidethevenue.”Whilethesedoappearto be far from the examples that conclusively demonstrate “how businesses arereinventing themselves with IBM SmartCloud,” IBM appears to be more interested indemonstratinghowcompaniesarebeginningtomovefromthesemainstreamexamplestothe “revenue stream.”This represents for IBMa kind of cloud 2.0whereby companiesmovefromacloud“takenatfacevalue”—thatis,as“aconduitforincreasingflexibilityandreducingcomplexity.”Now,“forward-lookingbusinessesarerethinkingthecloud”bytaking profitable advantage of new mobile, social, and big-data analytical capabilities.IBMpromises that itscustomerswillbeable tochangemodels forhow todobusiness,disrupt whole industries, and speed up the process of getting products and services tomarket. While some might mourn the demise of the IT department, the SmartCloudenablesconversations thatwereonce limited to the techexperts to takeplaceacross thecompany.Theadisshortonspecifics,singlingoutonlyonecompany,3M,whichusesthecloud to analyze eye movement so that graphic designers can more effectively “grabviewers’ attention.” Onemight think that improving the delivery of eyeballs is not themost spectacular example to lure people to the company’s Smarter Planet strategy.Nevertheless, it is clear from the ad that thepurposeof the cloud is tomake everyone,everywhere,smarter.

Perhapsbecauseitisdirectedatthebusinesscustomer,IBM’scampaignneverreachesthedramaofMicrosoft’s“TotheCloud.”Yes,thehyperboleisinescapable,withphraseslike “sweeping changes” and “perfect storm” and words like “reinvention” and“transformation”usedagainandagain.Buttheobjectistosellintelligenceandrationalityrather than to create the family that nature never could. Yet the message is just asprofound.IBM’scloudisnotaboutemotionorempathy;rather, itsSmartCloudisaboutknowledge and rationality. IBM’s is clearly a cloud of knowing and other large cloudproviders,likeVerizonwithits2013“PowerfulAnswers”campaign,havefollowedIBM’slead (VerizonWireless 2013).We will consider the deeper importance of this view in

Chapter 5 when we contrast it with “the cloud of unknowing,” a stream of influentialthoughtderivedinpartfromalatefourteenth-centurybookwiththattitle.

Advertising isaboutmanythings,butoneof themost important isaspecificformofperfection.ForMicrosoftbusinesscustomers,itisperfectpower;forconsumers,itistheperfectfamily.ForIBM,itisperfectknowledgeandforAppleperfectiontakestheformof synchronized harmony. Like IBM, Apple pitches mainly to one side of thebusiness/consumerdivide—totheindividualcustomer.Applecustomershavebeeninthecloudforsometimenow,butitwasnotuntilOctober12,2011,thatthecompanyformallyinviteditscustomerstojoiniCloud.Priortothattime,Applesubscriberscouldenterthecloud with an iTools account, which launched a primitive cloud service in 2000.Improvementsledtothe2004creationof.macandthatgavewaytoMobileMein2008.Many of us who used the early Apple cloud found MobileMe, with the “.me” suffixattached to everything, a bit hard to accept. After all, Apple had a reputation fornarcissismandwhatmoremirror-gazingservicecanyouthinkofthanonecelebratingme.Perhapsmoreimportantly,MobileMewasplaguedwithglitchesthatpromptedevenSteveJobstowonderaloudwhetherpeoplewouldevertrustApplecloudservices.Accordingtothe company’s founder, people were justified in saying, “Why should I believe them?They’retheonesthatbroughtmeMobileMe!”(Sutter2011).Oneoftheservice’sbiggestproblemswassynchronizingcustomerfiles,music,videos,mail,contacts,andcalendarsacross multiple devices. The source of many complaints, the world of MobileMe wasanythingbutharmonious.Thatisprobablywhy,whenthecloudsymbolbecametheiconforthecompany’sonlineservices,Applewaskeentofocusonharmony.

The first iCloud ads predated the launch and provided a simple explanation of theservicethatdemonstratesitscapacitytoseamlesslyintegratecustomerdevices.Aftertheframefeaturingtheimageofacloud,thevoiceoverexplains,“WithiCloud,whenyoubuyasongononedevice,itinstantlydownloadstoallofyourothers.Takeapicturehere,itshowsupthere.Startaprojectinoneplaceandpickuprightwhereyouleftoffinanother.Capture themomenthere, and it’swaiting foryou there.Makea changeon this, and itupdatesonthat.AndwithiClouditallworksinstantlyandwirelessly.Soyoualwayshavethe things youwant, exactlywhere youwant them” (Apple 2011).Having educated itsusersiniCloudbasics,lateradvertisingcamewithnovoiceoveratall(Apple2012).Theads were composed of purely visual images of music, photos, books, and appsdownloaded fromanduploaded to the cloudand instantly synchronizedacross iPhones,iPads, and laptop computers. No voiceover, and one frame of script: “Automatic,Everywhere,iCloud,”adescriptionofpureharmony.

With or without a voiceover, Apple ads are distinctive because the discourse onperfectionisembodiedinthecommercialaesthetic.Theparticipantsstartperfectandraisethelevelofperfectionthroughthecloud.Thereisnodramatictension,justanewlevelofsublimeharmony.ThisdifferssharplyfromtheMicrosoftads,which,whetherdirectedatbusiness or consumers, acknowledge theworld’s imperfections anddemonstrate how tousetechnologytocorrectthem.Thesourceofdifferenceisuncertain,butitisclearfrominterviews with the founders of these companies, Steve Jobs and Bill Gates, that theformerwas a perfectionist whose goal was to control the entire experience. Gates is asupporter of open systems that risk flaws in order to expand the number of users, apositionthatGooglehasfollowedwithitsAndroid-baseddevices(Isaacson2011,534).

Commercials like Apple’s for iCloud help to construct the discourse around cloudcomputing.Theexampleswehaveexploredareparticularlysignificantfortheirattempttosurroundthecloudwithvisionsofperfectionamountingtovariationsonthetechnologicalsublime.Theydiffer only inhowperfection is imagined.ForSalesforce, the sublime isrendered in businesses enabled by the cloud to do the impossible. In Microsoft adsdirectedatabusinessaudience,perfectionmeans technologicalchangewithoutnegativedisruptions. Inadifferent formofperfection,Microsoft’scloudbrings toconsumers thetechnologicalkeytocreatingtheperfectfamily.ForIBM,progresscomesfrombuildingasmarterplanetthroughacultureofknowingthatextendsrationalthoughtandpracticestoall areas of social life. Finally, there is Apple, deepening and extending the perfectionalreadypresentinlifethroughharmonioussynchronizationofallthedevicesthatfillourlives.

Therearenoguaranteesthatthediscourseembeddedinadvertisementsisthesameaswhat viewers, readers, and listeners take away. After all, a quick review of commentsreveals that some people were offended by Microsoft’s perfect-family ad and somethought thecompanyhadnobusinessclaiming that theads revealedanythingabout thereal meaning of the cloud. It is hard to imagine anyone convinced by the companyannouncingwithout ironyorsatire,“TotheCloud.”But thepoint isnotaboutassessingtheextentofanad’sinfluence.Thatismuchtoodifficulttoaccomplishwithanyscientificrigor.After all, therearemanyothervariables to consider.Does someone’s adoptionofiCloud or some business’s decision to join the IBM SmartCloud have any directrelationshiptoanadvertisingcampaign?Advertisingmayormaynothavesomesmallorlargeimpact.Whatthesecampaignsundoubtedlydoiscreateandbuildcloudcomputingindiscourse.Theydefineandteachindividualsandorganizationswhatcloudcomputingisandwhatgood it canaccomplish.This is important to fill avoid in the absenceof anyclearunderstanding,andespecially significantasameansofcountering journalisticandresearchaccountsthatidentifysignificantproblemswithcloudcomputing.Byassociatingcloudcomputingwithperfectingindividuals,families,andorganizations,promotersofthecloud construct an alternative to stories about environmental risks, power outages,pervasivesurveillance,andthreatstojobsintheITfield.

BloggingtheCloudCommercial advertisementswere once the overwhelmingly dominantway to promote aproduct.While they remain significant, there are nowmanymore ways to construct adiscoursearoundthecloudand, throughthis, tosellcloudcomputing.Amongthemanysignsofthischange,whereoncewetalkedaboutadvertisingagencies,todaytheyarejustagencies whose creative talent designs many forms of communication, including, butcertainly not limited to, advertising. The expansion of media forms in the twenty-firstcenturyhascertainlyhelpedtoopennewpromotionalopportunities.Forexample,asonewould expect, all of the major and minor cloud-computing providers promote theirserviceson theirownwebsitesandblogs.Theirsitesaregenerally informative,but it iseasytodismiss themascorporateself-promotion.Nevertheless, theydoserveadditionalfunctions.WhentheMicrosoftsitereportedthatitsresearchfoundthattwo-thirdsofsmalland medium-sized businesses lacked a marketing strategy, a widely circulating privateblog picked up the item in its lead to a story on how cloud-computing companies canimprovemarketing(CloudTweaks2013).Hereacorporatewebsiteprovidesinformationto a site that has a greater claim to objectivity, thereby conveying legitimacy for theMicrosoftfigures.Inthisrespect,companysitesprovidenourishmentforthegrowthofthepromotionalfoodchain.

Clouddiscourseisalsobuiltbythemanyonlinesites,includingblogs,newsletters,andreports on research, that promote cloud computingwithout a clear associationwith anyparticularcloudenterprise.Thisenablesthemtoenjoyasenseofobjectivityevenastheyadvanceapartisanview.Mostofthesebearthecloudlabel:CloudTweaks,Talkin’Cloud,andAsiaCloudForumareamongthemany.Someareconnectedtolargercompaniesthatdo IT research and sell cloud products. Others are just the product of an enterprisingindividualorsmallbusiness.Anumberofthesesitesaredirectlylinkedtoasaleseffort.For example, to download a white paper on overcoming challenges facing cloudcomputing,Iwasaskedtoprovideastreetaddressandphonenumber.Ididso,foundthepaperuseful,andreceivedaphonecallthenextdayinquiringaboutmyinterestinbuyingacloudservice formycompany.Anotherblog followedasimilarprocessbutwasevenmoreclearlyaimedathelpingITpeopleconvincetheirbossestomovetothecloud.Titled“HowtoBeataCloudSkeptic,” thepaperIdownloadedfromthatsite“detailsfourkeystepstodispelskeptics’fearssoyourorganizationcantakeadvantageofthecloud’smanybenefits” (Shields 2013). The article “Five Different Ways to Sell Cloud Computing”conveysthesalesmessageonCloudTweaksbutwithahumoroustouch(Kenealy2013).Some of what they do involves general consciousness-raising about cloud-computingcompanies:whatarethetoptenortoponehundredcloudcompaniesorwhatarethefiveortencompaniestowatchinthecomingyear(Panattieri2012)?

TheblogsiteCloudTweaksisagoodexampleofaninformationalblogwhosegoalistopromoteinterestinthecloudandinsalesofcloudservices.Establishedin2009,itisoneof the older sites. Its readership is made up of IT professionals, government workers,financial institutions, and corporate executives who subscribe for free by providingidentifyinginformation.For this theyreceive informationonjobs,vendors,conferences,courses, andwhite papers that contain research on the cloud industry.Cloud Tweaks issupportedby advertising,which is primarily placedby cloud-computingbusinesses andthecompaniesthatservicethem.AnissuepostedonJanuary8,2013,providesinsightinto

howsiteslikethisbringtogetherinformationalandpromotionalcharacteristicstoadvancetheconstructionofacloud-computingdiscourse.Itstartsbyraisingconcernsabouthowcloudcompanies,especiallysmallandmedium-sizedfirms,marketorfailtomarkettheirproduct. Many cloud companies, the article maintains, believe that the cloud is soextraordinarythatitwillsellitselfandsotheyrelyonasinglepersonorasmallconsultingcompanytopromotethesaleseffort.Thisisviewedasamistakeandasetofremediesissuggested.Firstonthelistissecuringaserious“channel”program.Achannelislingoforhow a seller communicates with potential customers, typically by opening an onlinepresencesuchasawebsiteorblog.Furthermore,whilebrandingthechannelisimportant,companiesneedtobecautiousaboutusingthetermcloudinanonspecificwaysincemostcompanies,especiallysmallones,likelyknowmoreaboutthespecificservicetheyneedthan about the concept of the cloud.Next, in a recognition that cloud promotion takesplace inmanydifferentways, the site recommends involvement in thecloud-computingcommunity by posting on cloud blogs, contributing guest articles, and participating inonlinediscussions.Alloftheseareformsofcompanypromotion.Finally,itisessentialtoparticipate in trade showsandconferences that focusoncloudcomputingbecause they,too,arevitalpromotionalopportunities(Kenealy2013).

Manyotheronlinesitesaredirectlyinvolvedinprovidingpromotionalinformationonsuch topics as how tomarket cloud computing. It is especially important to pay closeattention to these because they offer concrete insight into the ongoing process ofconstructing the cloud in discourse, including protecting the image of the cloud fromcriticalaccountsthatmightdamagetheindustry.Forexample,CloudComputingJournaloffersanarticleonhowto“AvoidFailureWhenMarketingCloudComputing.”Thepiecefirst declares just how essential it is to learn how to market the cloud: “Researchorganizations are predicting that ‘the cloud’ will dominate every facet of the softwareindustry;nomatterhowconcernedcustomersarewithsecurity,access,andcustomization,theSoftwareasaService(SaaS)marketisguaranteedtogrow”(Wilson2012).Theappealto research delivers a measure of legitimacy for what follows and the admission ofproblemsisimportantbecauseitprovideswhattheliteraryscholarRolandBarthes(1979)referredtoasaninoculationthatisimportant,ifnotessential,formaintainingthemythicstatusofanobject. In thiscase, theadmissionof thesecurity,access,andcustomizationproblemswiththecloudinoculatesthemythoftheperfectcloudwiththerecognitionofproblems,whichinmostcasesstrengthenstheargumentforitsessentialimportance.Thegentle nod of recognition gives greater weight to the primary point that the cloudwilldominatethesoftwareindustryandthatitsmarketswillgrow.Sogetonboard.Buthow?

Whenitcomestospecifics,onecanseehowthisformofcommunicationdepartsfromthe pure promotion of a commercial advertisement. Maintaining that the cloud givescustomers more power, cloud marketers are advised to stay ahead of the process byassuming that any prospective customer has done the necessary research prior to thepersonalsalespitch.Itisessentialforcloudcompaniestodeveloptheonlinepresencethatmakesiteasiestforpotentialcustomerstodeterminewhat’srightwiththeircloudservicesbeforecompaniesandcustomers talk.This includeswhitepapers,blogposts,anddemovideos. In fact, the piece recommends that cloudproviders avoid spendingmore than aminimal marketing budget on offline advertising such as print ads, direct mailing, orpresenceat trade shows. Inaddition tocommunicatingonline, cloudcompaniesneed to

usetheirwebsitestolaunchfreesoftwaretrialsforcustomersbecausecustomerswanttoknowhowsoftwareworks(Wilson2012).Anothersite,BusinessSolutions,offerstipsonhowtosellcloudcomputingtobusiness.Itadvisesthatfirst,ifthecustomerappearstobeenvironmentallyconscious,aprovidershouldsellthecloudasagreentechnologythatwillcut the corporate electricity bill. Second, for a company worrying about the potentialdisruptiveeffectsofnaturaldisasters,sellthecloudasasystemthatpracticallyguaranteescontinuity of service. Finally, sell those concerned about dependence on a particularplatformsuchasAppleorMicrosoftontheabilitytouseanyplatformatanytime,asteponthewaytofullvirtualization(McCall2012).

Other cloud sites offer advice on how to market to specific constituencies. Onedescribes how themarketing pitches to chief information officers (CIOs) should differfromthosedirectedatchieffinancialofficers(CFOs).Theformerareprimarilyconcernedwith security, followed by the ability to scale cloud resources, and then with theavailabilityofapplications.Alongtheway,thepostingidentifiesthecorrectanswerstothekindsofquestionsthatCIOstypicallyask.Bydefinition,CFOsareconcernedaboutcosts,specificallywithhowmuchthecompanywillsavebymovingtothecloud.Buttheyarealsoworriedaboutregulatoryissuesandtheoverallimpactofshiftingtothecloudonthecompany’s business model. Given potential clashes between the interests of CIOs andCFOs, the piece concludes that finding common ground is critical tomaking both feelcomfortablewiththetransition(Ko2012).

Stillothersitesconcentrateonhowtosell specifickindsofcloudcomputingservices.Forexample,GigaomofferssuggestionstobothsellersandbuyersonhowtomarketIaaS,which, as described in Chapter 2, is a form of cloud servicewhere the cloud providermanages servers that customers use to store and process their data. For sellers, the siterecommends eliminating the fits and starts that often come with human contact by“ensuring a seamless and human-free process to try your service.” Moreover, becausesellingIaaSoranycloudserviceinvolvesabigfinancialcommitmentfrommorethanoneexecutiveattheclientcompany,“Don’texpectthataself-servicetrialprocessaloneleadstosales.”Finally,itrecommendsthatcloudsellersbringinateamofspecialistsinareaslikesystemsintegrationandtelecommunications,evenifthatmeanspartneringwithothercompanies. As for buyers, the guide recommends choosing applications that minimizedependencies,somethingthatisdifficulttodowhenaclientispurchasingacloudservicethat requires using a cloud provider’s proprietary software and applications. It alsorecommends that buyers actively convince others in their organizations that the IaaSsolution is best because it is sometimes necessary to take small steps, if these can beadvancedasexemplarsofsuccess.Lastly,buyersneedtodemonstratethattheservicetheywant theircompany tobuy isone that is soldandusedbymany firms.Defending theirchoice as part of a widely known set of such choices eases concerns within theorganization (Orenstein 2010). This advice is particularly interesting because itacknowledges that constructing the cloud as a general business solution requiressalesmanship frombothsellersandbuyers.Thenotion that“weareall in this together”overcomesstructuraldivisions(buyers/sellers;CIOs/CFOs)andattemptstocreateanauraofcommonsense,community,andconsensusaroundthedecisiontomovetothecloud.

PrivateThinkTanksPromotetheCloudOnline newsletters and blog sites help to build the promotional culture for cloudcomputing without the commercial appearance of advertising. They are informative,educational, and service-oriented, while at the same time demonstrating keycharacteristics thatare found,albeitmoreexplicitly, incommercials.Thesesitespresentthe cloud as a technological breakthrough that will have widespread influence on allbusinessesandthroughoutsociety.Whilenotingtheoccasionalproblem,suchassecurity,theyarequick topoint to solutions, suchaspurchasingstrongencryption services fromcloudsecurityfirms.Otherwisetheyareoverwhelminglyupbeataboutthecloudandfocusondirecting readers to follow their leadbypointingout jobs in the industry, suggestingtrainingopportunities,and identifyingkeyplayers (fromtop-tencompanies toattention-grabbing start-ups).More than anything, they demonstrate how to sell cloud-computingservices with general advice and specific suggestions for different cloud constituenciesand segments of the cloudmarketplace. These sites are dependent on cloud-computingfirms for much of their information, but they also depend on another category of keyplayers in the cloud-computing arena: private research and consulting firms. Bycomparison to newsletter and blog sites, companies like Deloitte, Forrester Research,Gartner,andMcKinsey&Companyareasteportwoclosertotheperceptionofprovidingwhatareperceived tobeobjectiveaccounts.Althoughtheydonotalwaysofferpositivereports,privateresearchandconsultingfirmstendtoadvanceasupportiveandgenerallypromotionalstorythathighlightsgrowthandpositiveinfluence.

Private research firms are careful to define themselves as independent sources ofobjective information that businesses and governments should find valuable enough topurchase,evenwhenthepriceissteep.In2009Deloittefocuseditsresearchattentiononcloud computing with a report whose cover sported various types of clouds, each indifferent weather, appropriate to a document subtitled “Forecasting Change.” From thestart, it distinguished the report from promotionalmaterial, even as itmade an explicitpromotional pitch for the cloud: “The goal of this brochure is to enable ‘hype-free’discussion on cloud computing and align actors around a common understanding. Wehopethat,likeus,youwillbeconvincedofthecompellingpowerofcloudcomputing,notjustbecauseofitsadvantages,butalsobyunderstandingtherisksitentails,andwhatcanbedonetoaddressthese”(Deloitte2009,3).Evenasitaimstoavoidhyperbole,Deloittewishestoconvincereadersofthecloud’scompellingpower.Thereportquicklyleavesonewonderingaboutthedefinitionof“hype-free”becausetwopageslateritdeclareswithoutqualificationthat“Cloudcomputingwillbethenexttechnologicaldisruptiontotransformenterprise IT delivery and services” (ibid., 5). The report itself is more nuanced, butneverthelessisconsistentwhenitcomestoonecentralthemethatunitesmostreportslikethis:inevitablegrowth.Hereisarepresentativeassessment:“Manyexpertsstatethatthecloud market will drastically expand in the coming years. For the 2008–2013 period,Gartnerpredicts an impressivegrowthof theCloudcomputingmarket from9.1 to26.6billionUSD,which represents a CAGR (compound annual growth rate) of 24% (thesenumbersexcluderevenuesderivedfromCloud-basedadvertising)”(ibid.,29).Thesearestrong numbers for an industry that is just getting started, but using the authority ofanother private research firm, Deloitte is convinced that the cloud will expandsignificantly.Andifweneedareminder,thereportsummarizesitskeypoints(ibid.,34):

•

•

•

•

Economic,technological,andsocialfactorsfavorcloudcomputinggrowth

Industrytrendsshowsignificantfive-year,worldwidegrowth

Customer surveys indicate a high level of interest in cloud computing by ITstakeholders

Withmanyorganizationsstarting tobenefit fromthecloud,companiesofall sizesshouldevaluateitspotentialfit

Thesepointsaresignificantbeyondanyspecifictestoftheiraccuracybecause,whetheror not the report accuratelypredicts the industry’s future, it succeedsunambiguously inadvancing the promotional discourse. In addition to providing a singular form oflegitimacyforthecloud,thereport’sinfluenceismultipliedwhen,asisalmostinevitablythe case, it is highlighted in newsletters, blogs, andother promotional literature. In thiscase,blogs, includingSoftwareStrategiesResearch (Columbus2012a) andThe StorageEffect, a blog produced by the storage manufacturer Seagate (Wojtakiak 2012), and aresearchreport(Dalwadi2012)areamongthenumerousexamples.Theresultisacircleofaffirmationwherereportswithlegitimacygetreferenced,amplified,andreconstitutedbyotherpointsinwhatamountstoaglobalchainofdiscursiveproduction.Inthiscase,theproductisanarrativepromotingtheinevitablegrowthofthecloudcomputingindustry.

Forrester Research, which describes itself as a research and advisory company, isanother excellent exampleof a firm thatuses its legitimacy to advance thepromotionalculturearoundthecloud.Thecompany,whichfocusesonprovidingproprietaryresearchto the IT industry, has given considerable attention to cloud computing, including in a2011 report called “Sizing the Cloud.” Forrester uses the report to restate the growthmantrabypredictingthatthemarketforthecloudwillexpandfromunder$41billionin2010toover$240billionin2020.Whilethereportitselfrequiresabudgetbeyondthatofmostreaders($2,495acopy),growthappearstobeakeytheme(ReidandKisker2011).Like the Deloitte report, this one has been picked up by bloggers who single out thegrowththemeandseeitaspartofthegeneralviewamongexperts:“Thus,the“SizingtheCloud”reportsupportsacommonviewamonganalyststhatthecloudcomputingmarketwillwitnesstremendousgrowthintheforeseeablefuture.Themarketwillgrowsixtimeswithin a decade, according to Forrester, which is typical only for new and relativelyunderdevelopedmarkets”(Kirilov2011).Forresterreiteratedthisviewattheendof2011,when one of its researchers used a blog to make this prediction for 2012: “All cloudmarkets will continue to grow, and the total cloud market (including private, virtualprivate,andpubliccloudmarkets)willreachabout$61billionbytheendof2012.Byfar,thelargestindividualcloudmarketcontinuestobethepublicSaaSmarket,whichwillhit$33billionbytheendof2012”(ibid.).Aswithothersuchreports,thereisnuance—somecloudmarketswillgrowfasterthanothers,andmuchwilldependontheoverallstateoftheworldeconomy.Butonbalance,cloudcomputingwillcontinuetoadvanceasacentralforce in theglobal ITeconomy. It isdifficult todeterminewhether this forecastprovedaccurate because there is no clear measuring stick and many companies, includingindustry leaders likeAmazon,donotseparately identifycloud-computingrevenues.Theestimateofthetotalcloudmarketappearsreasonable,butthatforthepublicSaaSmarketmost likely overstates its actual growth. The point is that the specific forecast is lessimportant than its trajectory, whose arrow almost universally supports the promotional

discourse.

Once again, the report and its optimistic forecast circulated widely. This wasparticularlyimportantbecauseAmazonWebServiceshadsufferedamajoroutageshortlybefore it appeared and commentatorswerepleased to see thatForrester’s findingswereabletorelievesomeoftheunderstandableanxietiesaboutthecloudmarketplace.Ablogthat serves CIOs headlined its coverage “Forrester: Public Cloud to Surge, EspeciallySaaS.”MoreimportantisitssummaryconnectedtotheAWSfailure:“Longafterthebuzzabout Amazon’s two-day cloud outage dies down, the public cloud will be a growthtrajectory”(O’Neill2011).ThearticlegoesontorepeatForrester’sgrowthprojections,allheadingupward,totheyear2020.TheForresterreportmadeiteasiertoviewthefailureatAmazon,inspiteofsignificantcoveragebyjournalists,asanisolatedeventratherthanasa portent of disasters to come (Miller 2011). There certainly was no guarantee thatpotential cloud customerswould quickly come back to the cloud given thewidespreadnegativereactiontotheAmazonevent.ConsiderthisfromoneITexecutive:“‘Wedon’tthinkthecloudisenterprise-ready,’saidJimmyTam,generalmanagerofPeerSoftware,whichprovidesdatabackupforbusinesses.‘Areyoureallygoingtotrustyourcorporatejewels to these cloud providers?’”(ibid.). This was certainly no isolated comment, asothersalsochimedin:“‘Clearlyyou’renotincontrolofyourdata,yourinformation,’saidCampbellMcKellar,founderofLoosecubes,aWebsiteforfindingtemporaryworkspacethat was among those that lost service. ‘It’s a major business interruption. I’m gettingbusinessinterruptioninsurancetomorrow,believeme,andmaybewegetadifferentcloudproviderasabackup’”(ibid.).ItisimpossibletosaywhetherForrester’saffirmationofthesurging cloud succeeded in calming fears, but it is important to contrast the Forresterpredictionswiththenewsofthedaybecausedoingsodemonstratestheimportancetotheindustry of having a discursive apparatus at work to offset critical concerns raised byjournalistsandresearchers.

Gartner describes itself as “the world’s leading information technology research andadvisorycompany”andboasts,“Wedeliver the technologyrelated insightnecessary forour clients tomake the right decisions, every day. FromCIOs and senior IT leaders incorporations and government agencies, to business leaders in high-tech and telecomenterprisesandprofessional services firms, to technology investors,weare thevaluablepartner toclients in12,400distinctorganizations” (Gartner2013).Evendiscounting forthe hyperbole that often accompanies such self-descriptions, there is general agreementthatthecompanyexertsconsiderableinfluenceintheITindustrythroughitsresearchandforecasting.Asaresult,itcancommandtopdollarforitsassessments.AGartnerreport,described shortly, forecasts the development of the cloud through 2016, runs to ninepages, and costs $9,995. Like its counterparts at Deloitte and Forrester, Gartner’spredictionsaboutcloudcomputinghavecontributedsubstantiallytopromotingthevisionof unrelenting expansion. For example, in July 2012, in an article headlined “Gartner:Cloud putting crimp in traditional software, hardware sales” the company’s “cloudforecaster”predicted that thesectorwouldgrowby19percent in2012,goingfrom$91billionin2011to$109billion.By2016Gartnerexpectsit tobea$207billionindustry,which, while still representing a small percentage of the total IT sector, neverthelessmeans that itwillbegrowingconsiderably faster than theoverall sector (Butler2012a).This rosy forecast came a few weeks after even rosier predictions about consumer

adoptionofthecloud.Whileonly7percentofconsumerdatawasstoredinthecloudatthetimeoftheforecast,itconcludesthatby2016thecloudwillcontain36percentofallsuchdata.Thiswill result in increaseddemandthroughout the industryfordatacenters,for synchronization services, and for flawless uploading and downloading capacity.AccordingtoGartner,“Cloudstoragewillgrowwiththeemergenceofthepersonalcloud,whichinturnwillsimplifythedirect-to-cloudmodel,allowinguserstodirectlystoreuser-generatedcontentinthecloud”(ibid.).Asidefromsomeminorpotentialproblemssuchasthethreatofcommoditizationaspersonalstorageexpands,theimmediatefutureisclearlypositive for the expanding cloud. The Gartner results spread widely among thoseproducingcloudnewslettersandwebsitesandmade it into thebusinesspress, includingsuch influential publications as Forbes, which ran a compilation of upbeat marketforecaststhatincludedtheGartnerstudy(Columbus2012b).Accordingtoanotherreport,cloud computing has penetrated every facet of the global corporate supply chain.As aresult,itconcludes,Gartner’sforecastthattheIaaScloudwillgrowby42percentby2016shouldnotbeshocking(SmartDataCollective2013).

The final example of a private research organization that is helping to create apromotionaldiscoursearoundthecloudisMcKinsey&Company,whichdescribesitselfas“the trustedadvisor to theworld’s leadingbusinesses,governments, and institutions”(McKinsey&Company2013).Foundedin1926, thecompanyboasts that itworkswithtwo-thirds of Fortune magazine’s top 1,000 corporations. McKinsey’s relationship tocloudcomputingbeganwithsomecontroversywhenin2009itdefiedtheearlyboostersandarguedthat,especiallyforlargecompanies,movingtothecloudwasnotnecessarilythe best choice. Its report “Clearing the Air on Cloud Computing” concluded that theservicewasoverhyped,particularlyasacost-saver,becausecloudservices likeAmazonWebServiceschargedmorethanitwouldcostcompaniestokeeptheirdataprocessinginhouse by using their own data centers and servers. Ideally, McKinsey recommendedkeepingitall inhousebutvirtualizingtheserversor, inessence,carvingupserversintomultiplevirtualmachines, enabling software tomaximizepower fromonemachine andadding the ability to scale according to the company’s changing needs. Even theserecommendationswerequalified, asMcKinsey recognized that small andmedium-sizedfirmswouldnotbeabletoenjoythesamescaleeconomiesforin-housesystemsastheirlargercounterparts(Rao2009).Thisearlyresearchcontinuestoresonate,asthereportofone independent study concluded: “Large enterprises with highly optimized IT shopstailoredtotheirbusiness’needsmayfindcloudcomputingtobemoreexpensive.But,ifacompanyhasworkloadsthatebbandflowintheiruseofcomputepower,thenthecloudcanyieldsubstantialsavings”(Butler2013a).

Thesuggestionthatlargefirmsshouldshunthecloudwasmetwithconsternationandcriticism from cloud-computing supporters. Most were dismayed that such a reputableresearchfirmwouldrush to judgmentandcharged that thereport“neglects toaddressafewkeytrendsthatareoccurringincloudserverservices.Innovationisrapidlychangingin thecloud.The space is still verymuch awork in progress and big cloud computingservices, likeAWS,Google,SunMicrosystemsandMicrosoft,are regularlycomingoutwith different products.As these companies throw their hats into the ‘cloud computingring,’AWSwill face increased competition in themarket and could cause prices to godown to fight formarket share” (Butler 2013a). For some analysts, toomuch attention

paid to current prices (“the report seems to hype the cloud costs”) and too little to theprospectsforinnovationdoomedthereport(ibid.).

The first McKinsey report does not sound promotional at all. But what makes thisexampleparticularlyinterestingisthatthecompanyhascompletelychangeditstune.Ata2012 conference, a senior partner with the company delivered an altogether differentoutlook.InaninterviewBertilChappuisdescribed“anentrepreneurialgroundswellforthecloud.”His point is that the cloud has not just been good for companies selling cloudservices;ithasbeengoodforallbusinessandforentrepreneurshipaswell.Regardingthelatter,Chappuismakesitclear thatheisnot just talkingaboutaSiliconValleybrandofentrepreneurialbusinessformation,butaboutall formsofbusinessactivity, large,small,and individual. What changed from 2009 to 2012? For this Chappuis concentrated onthreekeydevelopments.Inareversalfromtheexpectationcontainedinthefirstreport,it’s“cheapcomputing.”Specifically,hecitesa threefolddifference incostbetween runningyourownserversystemandshippingittothecloud.Infact,owingto“massivelyscaledandefficientdatacenters” thecostofacompletecloudservice is lower thanthecostofprovidingthepowerforin-houseservers.Furthermore,cloudservicesarefarmoreagilefor provisioning infrastructure.Whereas it takes anywhere from 60 to 150 days for anenterprisetoprovideforaserversystem,accesstothecloudispracticallyinstantaneous.Infact,hecitescasesofpeople“buyingcomputepowerontheircreditcards.”Ontopofthishenotesthecapacityofthecloudtoenablenewexperiences:“social, local,mobile,big data.” These require rapid, agile development to satisfy the requirements acrossmultiple device platforms that have what he calls very “bursty” processing profiles.Puttingtogethercost,agility,andthepossibilityfornewexperiencescreates“areinforcingcyclethatwillenablethesecloudenvironmentstopropagateinallsortsofenvironments.”As a result, cloud computing actually becomes farmore significant than evenwhat itsearlyboosterspredicted.Inadditiontoservingoreventransformingbusiness,itbecomesacriticalforceincreatingentirelynewlinesofbusiness(Chappuis2012).

Chappuis supports this view with several examples, including a pharmaceuticalcompany that was motivated to revamp its entire customerrelationship management(CRM) system when it decided to incorporate detailed molecular information into itsexistingsystem.Sincethecompanydidnothavetheresourcestodothejobinhouse, itcontractedwithacloudprovider,whichprovidedanapp thatdid the job sowell that itconvincedthefirmtorethinkitsentireCRMstrategy.Inanotherexample,anHRmanagerwantedtoapplyanalyticstohisemployeedatabase,andthecloudproviderthatsolvedtheproblem convinced the manager to restructure all of its HR systems. Next, a smallbusinesswith twenty-five employees,which did everything in-house, contractedwith acloud company to host its email. Thisworked sowell that the firm decided to port itsonlinevideotothecloud.Whenthat,too,succeeded,thecompanyshiftedallofitsITtothe cloud, saved 55 percent of its IT costs, andwas able to focus on its core business.Finally,Chappuis turns toAWS,which receivedheavy criticism in the2009McKinseyreport for its high prices. Now the story is about an IT manager who, facing longprovisioningtimestostageanapp,savedmonthsbyturningtoAWS,whichcompletedthejob in minutes. Problems associated with locked-in contracts and endless subscriberpaymentsdisappearfromconsiderationasoutsourcingtothecloudbecomes,astheclichégoes,awin-winsituation(R.Cohen2013).

To paraphrase a familiar line, we can be assured of three things: death, taxes, andchanging weather. So it should come as no surprise that McKinsey’s forecast wouldchangefrom“partlycloudy”in2009toabsolutelysunnyin2012.Butthefirm’schangingforecastalsooffersanimportantlessoninthedevelopmentofapromotionalculture.TheevolvingagreementinmuchoftheITworldthatcloudcomputingis“thenextbigthing,”guaranteedtogrowwellintothefutureandtotransformbusinessintheglobaleconomy,does not automatically become common sense orwhat scholars callhegemony. Rather,hegemonytakestimetogrowandinevitablychangesinthefaceofbothinternaltensionssuchasthedifferencesinearlyforecastsbetweencloudsupporters,andexternaltensionssuchasthedisagreementsbetweencloudboostersandjournalistswhohavechallengedthecloud because of environmental, security, and labor concerns. The development of ahegemonic promotional culture is not amechanical process that arises simply from thebalance of societal forces, but an organic one that emerges, changes, and can wither,disappear,orthrivedependingontheextenttowhichkeyparticipantscontinuetoactivelyaffirmits importance.McKinsey’schangeofviewmayrepresentsimply therecognitionthatitoncemisreadthecloudordidnotappreciatetheextenttowhichitcouldimproveina short time. But one can also see it as a key turn in the development of the cloud’spromotional culture because a major participant in one of its key sectors, the privateresearch and consulting community, overwhelmingly affirmed the dominant view afterhavingraisedsignificantconcernsthreeyearsearlier.

PromotingtheCloudtotheWorldInaddition toadvertisements,websites,and thereportsandforecastsofprivateresearchandconsultingfirms,itisimportanttoconsidertheworkofglobalresearchorganizationsthat take a further step in building hegemony in support of the cloud. An excellentexemplar is theWorldEconomicForum (WEF),whoseGlobal InformationTechnologyReport2012(DuttaandBilbao-Osorio2012)focusedoncloudcomputingastheessentialnew ingredient in a networked world. The forum describes itself as “an independentinternational organization committed to improving the state of the world by engagingbusiness, political, academic and other leaders of society to shape global, regional andindustryagendas”(WorldEconomicForum2013).ItisbestknownfortheannualDavosconference, which brings these leaders together to discuss global issues and buildconsensus in support of policy initiatives. In the last two years, cloud computing hasattracted the forum’s attention and the report is its first effort tomobilize internationalsupport for a common approach to the cloud and give it the stamp of approval from amajorglobaleconomicorganization.Theforumreportcontainsanumberofindividuallyauthored chapters written by people who work for some of the most importantcorporations in the information-technology and telecommunications industries; privateresearch organizations (a group of researchers with McKinsey wrote one chapter);international bodies, including theWEF itself, the UN’s Unesco, and the InternationalTelecommunications Union; and universities. There are two chapters about the cloud,including “The Wisdom of the Cloud: Hyperconnectivity, Big-Data, and Real-TimeAnalytics,”writtenbytwoexecutiveswiththesoftwarecompanySAS,and“Harnessingthe Power of Big Data in Real Time through In-Memory Technology and Analytics,”producedbythesoftwareandcloud-computingfirmSAP.

Although the reportcontainsworkbypeople fromdifferentprofessions, there is littledoubtthatitspeakswithanoverwhelminglycorporatevoice.Withnoworkfromthelargecommunityofnon-governmentalorganizationsin theglobalITsector, there isalsolittledoubtaboutwhosevoice issilent.Accentuating thecorporatestampis thecollaborationbetween the Forum and INSEAD, a global business school with campuses in France,Singapore, and Abu Dhabi. It is arguably even more interesting that the full report issponsoredbytheChinesefirmHuawei,aworldleaderinelectronicsand,somewouldsay,acorporateleaderincontroversy.In2012thecompanysurpassedEricssonastheworld’slargesttelecommunicationsequipmentmakerandleapedoverNokiaandRIMtobecome,after Apple and Samsung, the third largest producer of smartphones in the world. ThecompanymanufacturesformarketsaroundtheworldbuthasbenefitedfromtheexplosioninsmartphoneuseacrossChinaparticularlybecause,unlikeAppleandSamsung,Huaweiproducesinexpensivedevices.Butmakenomistakeaboutit:Huawei’sreputationforlow-costdevicesdoesnotmakeitalow-endfirm.Infact,halfofitsworldwidelaborforceisinvolved in research and development, employed at some twenty research anddevelopment institutes around theworld. Partly because of its commanding position inglobal electronics production and partly because the company has rapidly become adominant force in leading-edgeresearch,Huaweihasattractedwidespreadattention,butnotallofitisgood.

In 2012 the U.S. House Permanent Select Committee on Intelligence charged thatHuawei, alongwith anotherChinese telecommunications firm, servedas an intelligence

front for the Chinese government and its military “that could undermine core U.S.national-security interests” (Rogers and Ruppersberger 2012, vi). Not everyone in theUnited States agreed with the House report, citing the lack of strong, direct evidence(Mathias 2012).Nevertheless, the charges spread and other governments, including theAustralianandCanadian,raisedseriousconcernsaboutHuaweiandbannedthecompanyfrom bidding on critical government infrastructure projects (Marlow 2013). In thiscontext,theWEFreportgainsfurtherimportancebecauseitenabledHuaweitolaunderitsreputationasaglobalsecurityriskandmitigate,atleastinasmallway,thefearsthat,asonecommentatorexplained,“ifChinastopsplayingbyDavosrules,thenthegoldenyearsof theWorldEconomicForumwill beover” (Rachman2013).Huawei’sworkwith theWEFhelpedthecompanybuildlegitimacyasattacksonitsactionscontinued.In2013ittook another step by becoming a partner with CERN (the European Organization forNuclearResearch),providingcloudstoragefortheworld’smajorparticle-physicscenter,nominortasksincethelabrequirestwenty-fivepetabytesofdataeachyear.Asananalystconcluded,“CERNhasnowputthecompanybackonthe‘nicelist’”(Harpreet2013).

The content of the report is important because it gives concentrated and repeatedattentiontothreethemes.First,itpromotesthevisionofinformationtechnologyinallofitsformsasthekeytoeconomicgrowthandtotheoverallsuccessoftheglobaleconomy.Second, it identifies cloud computing as the leading edge of IT development and theessential ingredient for organizational success, especially in business. Third, the reportinsiststhattheprimarychallengetotheeffectiveuseofcloudcomputingistheadoptionoftechnicalstandardsthatwouldenabletheseamlessconvergenceofmachinesanddevicesresponsibleforstorage,processing,distribution,anduse.Almostasimportantistheformthatthecontentofreportslikethistakes.Toreachasbroadanaudienceofdecisionandopinionmakersaspossible,thereportiswritteninaclearstylewithpracticallynojargon.Moreover, it is replete with the kinds of summary text, figures, and tables that bothsimplify arguments and add the legitimating weight of quantitative data. Finally, itcontains numerous lists that rank order nations according to how well they embraceinformation technology—for instance, readiness toenter theworldofhyperconnectivity.Theseappeal to thosewhomightonlyhave the time to flip through thereport,butwhowould be interested to locate and compare where their own country appears on a list.Although certainlymore nuanced than a commercial ad or a short blog post, andwithmorelegitimacythanacommissionedprivateresearchreport, thedocumentiscarefultooffertheclarityandsimplicitythatadvancesthepromotionalproject.Unlikeajournalisticorscholarlyaccount,whichcanoftenreadlikeacontestedterrainofclashingviews,thereport is singular in its positive, promotional message about IT and the cloud. Wherenuanceexists,itisonlytohighlightthetechnicalhurdlesthatleavesomequestionmarksalongtheroadtofullconvergencebetweenthecloudandthose“pipes”anddevicesthatdeliveranddisplaycontentfromthecloud.

Fromthestart,thereportresemblesthosethat,intheearlydaysoftheInternet,createdmythicalvisionsof adigital sublime (Mosco2004).Wedonot just live in a connectedworld, the cover subtitle tells us; ours is “hyperconnected.”Thepreface,written by thechiefbusinessofficerof theWEF,does little to temper thehyperbolic enthusiasm for aworld shaped by information and communication technology (ICT). It describes thedocument as a detailed analysis of “the main drivers and impacts of this ICT-enabled

hyperconnected world and contributes to the work of the World Economic Forum’sHyperconnectedWorldInitiative,whichestablishesaholisticmeansofunderstandingthesystemicnatureofchangeinahyperconnectedworld”(DuttaandBilbao-Osorio2012,v).The chairwoman of Huawei chimes in with her iteration on the theme of digitalenthusiasm: “Ubiquitous super-broadbandwillmake almost everything faster andbetterwhiledeliveringanimproveduserexperience”(ibid.,ix).Theechoesofhyperconnectivitycontinue through the executive summary, where representatives of both the WEF andINSEAD mix hyperconnectivity and social transformation to create a rich stew oftechnologicaleuphoria:“WeliveinanenvironmentwheretheInternetanditsassociatedservices are accessible and immediate, where people and businesses can communicatewitheachotherinstantly,andwheremachinesareequallyinterconnectedwitheachother.Theexponentialgrowthofmobiledevices,bigdata,andsocialmediaarealldriversofthisprocess of hyperconnectivity. Consequently, we are beginning to see fundamentaltransformations in society” (ibid., xi). This establishes the model for the document:technologyiscreatingahyperconnectedworldthatis,withafewminordisturbances,anunalloyed blessing for theworld. The only reasonable response of governments to thisinevitable development is to figure out how best to adapt. Consequently, the reportproducesa“worldreadinessframework”whoseprimaryindexmeasures“thefriendlinessofacountry’smarketandregulatoryframeworkinsupportinghighlevelsofICTuptake”(ibid.,xii).Evenbeforeweenterthebodyofthedocument,itisclearthatweareenteringa mythic universe filled with the reification of a technology that drives the world toprogress,providedthatpeoplefigureouthowtoproperlyadjusttoitsrequirements.Itismythicbecauseittellsastoryofalarger-than-lifecharacter,InformationTechnology,thatofferstheworldthemagicofhyperconnectivity.Themythturnsonthedramaofwhetherwe will adapt ourselves and our societies sufficiently to the needs of technology, bycreating,forexample,thosebusiness-friendlypoliciesthatencourage,asthereportputsit,“highlevelsofICTuptake.”

The body of the WEF document details the promise of information technology todeliverevergreater levelsofprogress to theworld’speople.Likemostmyths,howevercomplexthestoryappears, it isfundamentallyasimplenarrative: themoreIT, themoreprogress.ButtheWEFreportismorethanapromotionalblurbfor“thenextnewthing.”Thisisdemonstratedinitswillingnesstoadmittochallengesthatcangetinthewayof,andperhapsevenslowdown,thearrivaloftheinevitableprogressthatITdelivers.Sinceitisintendedforaknowledgeablereadership,thereportcannotsimplydismissproblems.Rather,itredefinestheminawaythatdeflatestheirpowerandtheirsignificance.

Considerprivacy,acentralissueindebatesoverinformationtechnologyandthecloud.Whilenotgoingsofarastoseeitinquitethisway,thereportdoesnameprivacyamongthe issues facing a hyperconnected world (ibid., 4). However, from the standpoint ofpromotingIT,thekeyistorethinkprivacytominimize,ifnotcompletelyeviscerate,itasanissueworthyofcarefulpolicyattention.Interestingly,thisismadeclearinadiscussionof what some would see as the central place where privacy matters—the collection,storage,processing,anduseofhealthdata:“Isprivacyaconcern? Itcertainlyhas tobefrontandcenterwithrespecttovirtuallyanyeffortconnectedtohealthcaredata.However,someexpertsaregraduallyadoptingasomewhatcontrarianviewonthistopic,believingthatoursocietymustmovepastthefearofdataandprivacybreaches.Manytechnological

innovations that have revolutionized medicine might not have been possible withoutsharing data. Any data—electronic or paper-based—are vulnerable. But here, too,hyperconnectivitywill enablenew tools to fight crime, fraud, andabuse” (ibid., 99). Inessence,privacyisaconcern,butnotreally.Specifically,first,getoverit.Second,ifyouwantmedical progress, thenyour datamust be shared.Third, all data, includingpaper-based,arevulnerable.Finally,technology,intheformofahyperconnectedITworld,willfind solutions to problems presented by privacy breaches. This is a version of whatEvgenyMorozov(2013b)calls“solutionism,” theviewthatproblemsandsolutionswillbedefinedandsolvedwithintheparametersofthetechnologiesthatmajorITcompaniesidentify.Witheachreasonnottoworryaboutprivacy,oneisledtowonderwhyitshouldbeaconcernatall.Themythofadigitalsublimeisstrengthenedbyinoculatingitwiththeidentification and subsequent dismissal of what many see as a major limitation on itspowertobringuniversalprogress.

Cloud computing makes up a second major theme of the World Economic Forumreport,withonechapteronthecloudandanotheronbigdata.Itisparticularlyinterestingthatthecloud-computingchapterreturnstotheissueofprivacyand,evenbeforegettingtothe specific details that make the cloud important, deflates fears about privacy andsecurity:aboutconcernsover“infringementofprivacy…wecannotescapethefactthatbig data offer meaningful social and economic benefits that mitigate these legitimateconcernsbecauseofthehugelyfavorablesocialand/oreconomicimpacttheyimpart—onprivate commerce, international economies, and economic development. Certainly datasecurityissuesareimportant,butifbigdataaretobecomethecurrencyofthefuture,weneedgovernance,transparency,andsecurity,asopposedtoreactionaryplanstolockupthedataandthrowawaythekey.Aswithanycurrency,suppressionisnotasustainablewayforward” (Dutta and Bilbao-Osorio 2012, 97). In essence, the report concludes thateconomicbenefits trumpprivacyworries and,more importantly, it sets up a dichotomybetweenprogressivepoliciesthatunleashthepowerofbigdataandretrogradeapproachesthatlockitup.Thereisonlyoneintelligentchoice,onewaytomoveforward.

Withprivacyessentiallyoutof theway,weare free tounleash thepowerofbigdataand,more specifically, its power to benefit business in a very bigway.Aside from thefinancialbenefitsofusing the information“hidden in theworld’sexistingdata sources”(ibid.,98),suchas theestimated$600billionrevenuegainfromusingpersonal locationdatagloballyand the60percentpotentialgain in retailoperatingmargins, thereare theenormous qualitative benefits to companies, particularly those that mine social-mediadata.Among thebenefitsare theability to“protectabrand,engage themost influentialvoices in a market, understand what trends lead to sales, identify an untappedmarket,enhancemarket research,understand the impactof industrychanges,gathercompetitiveintelligence,improvewarrantyanalysis,createabettercustomerexperience,andmanageacrisis”(ibid.,99).Acknowledgingwhatitviewsasthe“irony”thatsocial-mediadataaregenerated not by businesses but by individual userswho are linked through Facebook,Twitter, and other social-media sites, the chapter simply assumes that all of the datageneratedshouldbefullyavailabletobusinessesseekingtoturnuseractionsintorevenuestreams.Suchis,asthechaptertitlesuggestswithoutirony,“thewisdomofthecloud.”Asanotherchapternotes,“socialmediapresentnewopportunitiesforsavvyorganizationstocapture‘thewisdomofthecloud’andleveragethefloodofunstructureddatathatisbeing

created”(ibid.,xvi).

Byapproachinguser-generateddataasfreelyavailabletobusinessestouseinwhateverways generate profit, the report demonstrates the difference between promotionalliteratureand research.Asexemplified in theWEFreport,promotionaffirmsapositionthat itscreatorswishtoadvanceinorder toaccomplishthegoalofconvincingothers tofollow their lead. On the other hand, research raises questions about positions thatgenerate thoughtanddebate rather thansimplyassent.Forexample, in2013 theprivatetech-analysis firmOvum reportedon the results of a surveyof11,000people in elevencountriesoncorporateuseofpersonaldata.Amongitsmanyinterestingfindingswerethat68percentofrespondentswoulduse“donottrack”softwareifitwerereadilyavailabletothemon awebsite.More troublingwas the finding that only14percent of respondentsbelieved that Internet companies were honest about their use of personal data (Gross2013).Ovum’s research demonstrated a profound lack of trust in online commerce thatfindsnoplaceinpromotionalliterature.Basedonitsresults,theOvumreportconcluded,“MoreandmoreconsumersaredecidingtoeffectivelybecomeinvisibleindatatermsontheInternet.ItwillshaketheInterneteconomyasmoreandmoreusersdecidetheydon’twant to be tracked” (ibid.). Facing the reality of that decision, the report raised seriousquestionsforitsbusinessclients:“Unfortunately,inthegoldrushthatisbigdata,takingthe supplyof littledata—personaldata—forgranted seems tobeanaccidentwaiting tohappen” (ibid.).But that iswhatpromotional literature like theWEFreport aims todo:take for granted pliable users who will ignore dataprotection opportunities. Genuineresearch does not, even if itmeans facing hard questions andmaking difficult choices,suchasdevelopingabusinessstrategythataddressestherealitythat,asatechnicalanalystforOvumconcluded,“Youaregettingthissqueezebetweenahardeningconsumerattitudeandtighterregulation”(ibid.).

TheWEFreport ispromotional inpartbecause itcompletely ignoresresults like this.Instead,notunlikethecommercialadvertisementsdescribedearlier,itchoosestofocusonthe cloud as a source of intelligence that turns inefficient businesses into smartorganizationalmachines.Thekeyistheabilityofcloudcomputingtoperfecttheprocessofconvergence,which,over thehistoryofcommunicationtechnology,hasadvancedtheconnectionsbetweentheproduction,dissemination,anduseofinformation.FortheWEF,this is the heart of the matter: “Cloud computing services provide a catalyst for ICTconvergence. Telecommunications carrierswill graduallymove IT systems and Internetdata centers into the cloud, and telecommunications and IT industries will developuniformstandardstofacilitaterapidclouddevelopment”(DuttaandBilbao-Osorio2012,xiv).Thecloudisimportantnotonlybecauseofitssuperiorstorageandprocessingpower,sufficient to absorb the Internet and all of today’s IT, but also because it provides themissinglinkenablingtelecommunicationsproviderstoservetheentireworldfaster,morecheaply,andmoreefficientlythanever.Asaresultofjoiningthe“pipe”andthe“device”inwhatisclosetoaliteralcloudofconvergence,“thecloudhasreshapedtheITindustry”(ibid.,38).Butthis,asthereportrecognizes,istoosimple.Itmaybepromotional,butthisisnotaslickcommercialduringwhichsomeoneannounces,“TotheCloud,”and,withthesnap of the fingers, transports us to a world of seamless integration and sublimeconvergence.Insteaditisthereportofawell-respectedinternationalorganization,whichneeds to avoid the appearance of the myth-making that is taken for granted in thirty-

secondcommercials.

SoinadditiontopromotingthewondersofITingeneralandthecloudinparticular,thedocumentacknowledgesthat“thereareobstaclestothisintegration,includinginsufficientopenness in the ICT industry; a lack of unified technical standards; and a lack ofconnectionamongcloudcomputing, telecommunicationsnetworks (thepipe), andsmartdevices. Overcoming these obstacles and unifying ICT’s technical standards is a toppriorityifwearetoimproveinteroperabilitywithintheindustry”(ibid.,ix).FortheWEF,themajorproblemfacingthefutureofITandcloudcomputingisnottheenvironmentalconsequences of building enormous data centers around the world and powering themwith several levels of backup, including banks of spinning flywheels and thousands oflead-acid batteries. It is not the potential to violate privacy built into a system thatgenerates revenue precisely by scooping up and analyzing personal information. It iscertainlynotthesecurityproblemsofstoringdatainnationsthatwillnotprotect it,but,instead,willuseittomeettheirownneeds.GivenHuawei’sownsecurityproblems,oneshouldnotfinditsurprisingthatnomentionismadeoftheproblemsposedbystoringdatabeyondanation’sborders.Nor is it themassivechanges in theglobaldivisionof laborresulting from transferring the ITdepartmentsof theworld’sorganizations to thecloud.Rather, theprimary issueofsignificance to thecloud-computing industry isdeterminingthebestwaytocreateaglobalsystemofuniformstandardsthatwillguaranteethesmoothperformanceofacloud-basedglobalgrid.Giventheheavytelecommunications-industryinvolvement in the report’s creation, including sponsorship by the world’s leadingelectronics-equipment company, it is not a great surprise that the document wouldconcentrate on technical standards. Indeed the chapters that focus most heavily ontechnical convergence are the ones written by Huawei and representatives from theInternational Telecommunications Union. Moreover, it is an issue that thetelecommunications industry hasworried about andworked on for generations and onethat private research organizations insist needs careful attention in order to properlymaintaincloud-computingnetworks(Bernnatetal.2012).But there ismoreto this thanpromotingamajorissuefortheindustry.

Thereportrepresentsthetechnicismthatiscommoninmostpromotionaldocuments.Itis constructed to represent the general public interest, but is written from a particularindustryinterest.Toavoidtensionsbetweenthepublicinterestandtheneedsofbusiness,promotional reports avoid social and political issues and focus instead on technicalproblems like standards and convergence that are both real andunlikely to threaten thegoal of equating a specific industry interest with the general public interest. For thereport’swriters, there isnoquestioning thegeneralvalueand legitimacyof informationtechnologyand thecloud.Any thoughtof restricting theirdevelopment, forexample, toprotect theenvironment,secureprivacy,orsave jobs, is foolishand irrationalbecause itmeansgivingupthebenefits.Itis,however,legitimatetoraisetechnicalissuesthatstandin the way of their full development. Technicism, a focus not just on how technologydeterminesthingsbutonhowitbecomesthesingularsourceofsolutionstoproblems,isamajormeansofunitingthespecificinterestoftheindustryandthegeneralinterestoftheworld’sITandcloudusers.

LobbyingfortheCloudTwo additional forms of promotionalism are important to consider: the expansion oflobbying by firms involved in cloud computing and the proliferation of corporate tradeshows dedicated to the cloud. There are certainly overlaps among the various forms ofbuildingavisionofthesublimecloud.WhileitmightnottakeplaceasdirectlyatDavos,hostoftheWorldEconomicForum,asitdoesinWashington,D.C.,lobbyingisintrinsictopolitical activity in both places. Nevertheless, there is enough difference to warrantdistinguishing among the contributionsmade by the advertising spots that promote thecloud’sabilitytocreatetheperfectfamily,thesponsoredblogsthatchroniclethecloud’sseeminglyunstoppablegrowth, thereport thatdocumentsahyperconnectedworldinthecloud,andthehand-to-handnetworksthatlobbyingandconferencingbuild.

PerhapsbecausethefirstwavesofITentrepreneursbelievedthatthetechnologywouldsell itself to decision makers in Washington and other world capitals, there was littleorganized lobbying until recent years. This is particularly surprising because thetelecommunicationsandelectronicsindustriesarelegendaryfortheirlobbyingprowess.IntheUnitedStates,AT&TandGeneralElectricwereatthetopofalonglistoffirmsthatwereprominent in thecorridorsofpower.ScholarsattributemuchofAT&T’sability tomaintain itsmonopolycontrolover the telephone industry to itsarmyof lobbyists,whomadethecasethatMaBellembodiedtheneedsoflocalsubscribers,amassiveworkforce,and the millions of shareholders, all of whom held America together in one seamlessnetwork(Tunstall1986).Totamperwiththenetworkinanyway,whetherwithcompaniesthatmightwanttocompetebybuildingabetterorcheaperservice,orjustthosewhowantto sell a pink telephone, would be harmful if not downright un-American. AT&T’slobbying clout built a particularly cozy relationshipwith the Pentagon,which could becountedontodefendthetelecommunicationsmonopolyasamatterofnationalsecurity.Accordingtothecompanyline,multipleproviderswouldendangersecurenetworksthatwereessentialfornationaldefense.ItwasnotuntilAT&Tmetitslobbyingmatchthatitlost its monopoly control over the telecommunications marketplace. That could onlyhappenwhenthebanks,insurancecompanies,retailers,andotherswhopaidapremiumtosustain a telecommunications monopoly decided to form user associations whosecombined lobbying power exceeded that of AT&T (Schiller 1981). Even then, AT&TalmostsidesteppedthemovetocompetitionwhenitslobbyistsconvincedanearmajorityinCongress to support legislationcementingBell’smonopoly.But thatmove fell a fewvotes short andwhen theDepartment ofDefense recognized that lobbyists representingmajoruserswouldwintheday,itwithdrewsupportforthemonopoly,openingthedoortomarketcompetition.

Inspiteofthisandmanyothermodelsoflobbyingpower,theburgeoningITindustryofthe1990schosetomaintainthebareminimumofalobbyingpresenceinWashington.Asoneaccountdescribed,“Untilthemid-1990’s,politicswasaforeignsubjecttoexecutivesatmosttechnologycompanies—justassoftware,hardwareandtheInternetwereforeignconceptstomostmembersofCongress”(Rivlin2004).Abusinessuserexplained,“Therewas benign neglect on both sides,Washington andSiliconValley. The valley generallytooktheattitude,‘Aslongastheyain’tinmyface,justignorethem’”(ibid.).WhatcontactexistedmainlytooktheformofpoliticiansmakingthepilgrimagetoSiliconValleyforagenerous sprinkling of the gold dust that turned politicians into visionaries. This view

begantosubsideoncethenewcenturyarrivedandthedot-combustrockedtheindustry.First, since theyno longerhad theMidas touch,SiliconValley lobbyistshad to lineupalong with those from other industries and make sure to bring their checkbooks.Politicians,manyofwhomlostalotofmoneyinthecrash,werenolongertherejustforaphoto op. As one lobbyist for the IT industry said in a 2004 article, “Back in the late1990’s,SiliconValleyassumed thatall theyhad todo isshowupandpoliticianswouldfall at their feet, and for awhile theywere right.Now it takes a checkbook to get thatmeeting” (ibid.).Second, thepolicy issues that seemed lowon thepriority listwhen ITexecutives were rolling in venture capital, like tax rules on stock options and visaprogramsforforeignworkers,nowgrewinsignificance.

AlthoughlobbyinggrewinthewakeofthedownturnandespeciallywhensocialmediaandcloudcomputingstartednewwavesofITexpansion,itdidnotreallypickupsteamuntilsocialmediaandthecloudbegantoraisesignificantconcerns,includingtheneedtopromotegovernment’suseofcloudservices.Whileotherissuesmightattractmorepressattention, cloud companies’ success in getting the U.S. government, both civilian andmilitary, to fully commit to shifting services to the cloud has been one of the greatvictories for the industry. In addition, IT and cloud companies have used lobbying toactively resist efforts to tighten privacy protections in theUnitedStates andEurope, todemandhighercapsonimmigrationvisasforskilledforeigntechworkersorremovethemaltogether,tostopeffortstotightencontrolsoveronlineadvertising,andtopreventreformoftaxlawsthathaveenabledcompaniestoperfectthedarkartoftaxavoidance(NelsonandDuhigg2013;Houlder2013).

Googleledthewaywithamajorboostinitslobbyingoutlayin2010,justasconcernswere growing aboutwhat some chargedwere the company’s anti-competitive practices(Rao 2010). As the Federal Trade Commission (FTC) continued to look closely atpotential antitrust violations, Google, fearing a repeat of earlier rulings that severelydamagedMicrosoft,intensifieditslobbyingactivities.Asonereportsummarized,“insteadof ignoringWashington—as rivalMicrosoftdidbefore itscostlymonopolization trial inthe1990s—Googlespentabout$25millioninlobbying,madeanefforttocozyuptotheObama administration and hired influential Republicans and former regulators. Thecompany even consultedwith theHeritageFoundation andmetwith senators like JohnKerrytomakeitscase.Inotherwords,thesetraditionaloutsidersworkedthesystemfromthe inside” (Romm2013a). In 2012 aloneGoogle spent $16millionon lobbying,morethan twice that of any other tech company, and, with twelve different lobbying firmsworking on its behalf, it succeeded in forestalling anymajor restrictions on its marketcontrol(T.Lee2013).

Learning from Google’s success and concerned about its post-IPO bottom line,Facebook significantly increased its lobbying outlays from $1.34million in 2011 to $4millionin2012(Dembosky2013b).Thelastthingthecompanyneededwasstifferprivacylegislation thatwould cut into its plans to boost revenues byproviding companieswithinformation about its one billion users. So when the U.S. Federal Trade Commissionbegan an investigation into nine data brokers that do business with Facebook, thecompanyboosted itsWashington lobbyingsignificantly.Facebookstated,“OurpresenceandgrowthinWashingtonreflectourcommitmenttoexplaininghowourserviceworks,theactionswetaketoprotectthebillionpluspeoplewhouseourservice,theimportance

of preserving an open internet, and the value of innovation to our economy” (ibid.). In2013Facebook set up its own lobbying coalition, FWD.us, to engage in a broad-basedlobbying effort primarily promoting itsmembers’ support for expanding the number ofvisas for foreign workers (Wallsten, Yang, and Timberg 2013). However, its activitiescreatedturmoilandanadvertisingboycottonFacebookitselfwhenthecoalitionlobbiedonbehalfofoilcompaniesandforRepublicanPartycausesintheSouth(Edwards2013).Other companies also boosted their lobbying budgets, but sometimes, as in the case ofMicrosoft’s effort to rein in Google, they increased their spending to better positionthemselvesagainsteachother.AnothergoodexampleisSamsung,which,asitfacedoffagainstAppleinasetofpatent-infringementcases,wentfrompracticallynoWashingtonlobbying presence to spending $800,000 on lobbying the American capital in 2012,including$480,000inthefourthquarteralone(Quinn2013).

Notalllobbyingtakesplaceinnationalcapitals,though.Becausecloudcomputingdatacentersrequirelocationsthatoffercheapland,lowutilityrates,andtaxbreaks,companiesthatrunthemspendtimelobbyinglocalofficials,powerauthorities,andstatelegislaturesfor the best possible deal. For example,Microsoft’s decision to build a data center onseventy-five acres of bean fields in centralWashington required considerable corporatefinesse to convince the state and the local government to provide taxbreaks andutilityratesthatwerelessthanhalftheU.S.nationalaverage.Thelobbyingpaidoff,atleastuntilthecompany’suseofpollutingdieselgeneratorsledtoaseriesofconflicts,astorytakenup inChapter4.Lobbying likeMicrosoft’s incentralWashington is seenoverandoveragain across the United States and around the world. In North Carolina, for instance,lobbyingprovidedenormousbenefitstoApplewhenthecompanyproposedtobuilddatacentersinthestate,partlytotakeadvantageoflowlaborcostsandlow-pricedpower.Toattract the company, the state legislature approved $46million in tax breaks, and localgovernmentscutApple’sreal-estatetaxbillby50percentanditspersonal-propertytaxesby 85 percent (Greenpeace International 2011, 19).4 Additionally, North Carolinarewarded Google’s efforts with tax breaks, infrastructure upgrades, and other benefitsworth$212millionoverthirtyyearsandFacebookreceivedasimilarpayoff(GreenpeaceInternational2011).Whencloud-computingcompaniesinBoise,Idaho,foundthemselveswithaheftytaxbillleviedbyastateauthoritythatdeterminedcloudcomputingtobethetaxablesaleofsoftware,theyenlistedthelocalChamberofCommercetohelprollbackthetax(Glanz2012a;Moeller2013).

Lobbying helps organizations representing companies promote the common industryinterest.Butthisissometimesamixedblessing.BuoyedbythesuccessoftheirlobbyingintheUnitedStatesandwithabiggerstakeintheglobaleconomy,techfirms,includingcloud providers, began lobbying the European Union (EU) for favorable treatment,includingmorebusiness-friendlyprivacypolicies.Brussels(theEU’sdefactocapital) isnotas soldoncloudcomputingas isWashington,D.C., inpartdue to threats thecloudposestoprivacyandsecurity,anditsproposedpolicies,especiallyondataprotection,takea stronger position than those advanced by theU.S. government.Brusselswas also nothappyabouthearingfromthelobbyistsofmajorU.S.techfirmsandmadeitsviewsclear.Specifically, theheadofan industrycoalition that isworking todevelopEU-widedata-protection rules criticizedU.S. tech giants, especiallyGoogle and Facebook, for hiringlobbyiststopressuretheEUtoweakenitsprivacylaws.Settingasidediplomaticniceties,

Jacob Kohnstamm, chairman of the Dutch Data Protection Authority, declared thatEuropean officials were “fed up” with U.S. businesses putting their corporate interestsaheadofwhatEuropeansseeastheirfundamentalrightstodatasecurity.CallingouttheU.S.governmentaswellasitsbigtechfirms,hemaintainedthatCongresswouldnotbeastolerantifthetableswereturned:“IfsuchalobbyfromtheEuropeansidewereorganizedtowardsCongress,wewouldbekickedoutof there.”Americans,he insisted,simplydonot understand that for theUnited States privacy is a consumer protection, whereas inEuropeitisconsideredafundamentalhumanright.AGermanpoliticiansummarizedtheextent of the lobbying pressure: “Throughout the last year there has been a massivecampaign from the side of AmCham [the American Chamber of Commerce], whichorganisedeventsthroughoutEuropeandmetwithmanyMEPs[membersoftheEuropeanParliament] in Brussels and Strasbourg. But now, since January when my report waspublished, lobbyists, especially from SiliconValley, have stepped up their campaign towaterdowntheEUprivacyregulation”(DemboskyandFontanella-Khan2013).WhiletheObama administration, the American Chamber of Commerce, and lobbyists for the ITindustrymaysucceedinforcingEuropetosynchronizeitsdata-protectionlawsbyparingthembacktowherethoseoftheUnitedStatesstand,thisisonecasewherelobbyingcaneasilybackfire.

For some observers, major IT and cloud providers face an even more significantchallenge thanoppositionfromtheEU.Lobbying, theymaintain, turns firmsknownfortheir inventions, innovations, and entrepreneurship into ordinary companies that wouldratherfocusoninfluencingWashingtontoprotectwhattheyhavethanondeveloping“thenext new thing.” One business publication harkened back to a 1999 speech by NobelPrize–winningeconomistMiltonFriedman,whoreferredtolobbyingastheITindustry’s“suicideimpulse”(Crovitz2013).WhenGooglelobbiedtheFederalTradeCommission—successfully, it turns out—to forestall an antitrust investigation, the commission’schairmanquestioned the lobbying strategy: “Stop! Invest yourmoney in expansion andinnovation.Google’slobbyingexpenseshadnoeffectonthecare,diligenceoranalysisoftheagency’s incrediblyhard-workingstaffor thedecisionsreachedbyanyof theFTC’sfive commissioners” (ibid.). But even if its lobbying was successful, there is still anargumenttobemadethatlobbyingdistractscompaniesfromtheircoremission.“Insteadofthe‘suicideimpulse’oflobbyingformoreregulation,”oneanalystconcluded,“SiliconValley should seek deregulation and a long-overdue freedom to return to itsentrepreneurialroots”(ibid.).Theseobservationsareunderstandable.DidFacebookreallyneedthirty-eightlobbyistsin2013,anincreaseoffifteenover2011?DoApple,Google,andMicrosoft reallyneed topad theirhigh-paidrankswithformerFTCstaffers?Is thisnot “spinning the revolving door that fuels the growth of lobbying” (ibid.)? AndwhataboutAmazon,whoseownerappeared to trumphis lobbyingcompetitorsbypurchasingtheprimarynewspaper in theAmericancapital, therebygivinghim,andpresumablyhiscompany,privilegedaccesstothecorridorsofpower(Cassidy2013)?

Wellfoundedastheyare,thesecriticismsalsorevealasimplisticviewofgovernmentasacompletelynegativeinfluenceonbusiness,especiallyinnewindustries,suchasthosethat took root in Silicon Valley starting in the 1950s. It is simplistic because, whilegovernmentcanslowthegrowthofinnovationthroughexcessiveregulation,itisalsothecasethatbusinesseshavehistoricallydependedongovernmentforinfrastructuresupport,

formaintaininga stable intellectual-propertyenvironment, and foramarket in theearlydaysofexperimentation.Governmentwasallof thesethingsforSiliconValleyandit isreasonabletomaintainthatSiliconValleywouldnothavesucceededwithoutgovernmentsupport(Mazzucato2013).Thisisnotjustbecausegovernmentfundedearlyresearchoninformation technology through its own research labs such as at theDefenseAdvancedResearch Projects Agency (DARPA); it also provided a market when no private firmssteppedupinthefirstroundsofsemiconductorproduction.Lobbyingcanbeessentialtomakingsuregovernmentprovidesastableenvironmentforgrowingbusinesses.Moreover,criticismspresumethatlobbyingisonlyaboutachievingaspecificgoal.Googlelobbiestostaveoffantitrustregulation,Facebooktoavoidprivacycontrols,andMicrosofttoweakenenvironmentalrulesandwinlow-costpowerforitsclouddatacenters.But,importantastheseare,lobbyingmeansmorethanaccomplishingshort-termgoals.Lobbyingalsohelpscompanies promote thegeneral interest of the industry, including selling its products togovernment,which often helps tomake amarket and towin government support for afavorable business climate abroad. Looked at in this way, lobbying is every bit aspromotionalascommercialadvertisements,blogpostings,andhigh-levelbusinessreports.

CloudExpo:PromotingCloudComputingthroughTradeShowsLobbyingisalsointerestingbecause,inanerathattoutsthewondersofsocialmediaandmovingeverythingtothecloud,itremainsadecidedlyinterpersonal,real-timeactivity.SoaretradeshowsandconferencesthataimtoadvancebothknowledgeandsupportforITandthecloud.IntheITsectorthereareendlessroundsoftheseevents,butovertheyearsarguablythemostimportanthavebeentheCOMDEX(ComputerDealerExhibition)tradeshows,whichtookplacefrom1979to2003,andtheConsumerElectronicsShow(CES),whichbringstogethercompaniesaimingtohavetheirnewproductsnamed“thenextnewthing.” CES began meeting in 1967 and continues as an annual event in Las Vegas.COMDEXwasthemajorITeventuntil1999,whenittriedtorestrictmediacoveragetowriters accredited with a handful of the leading trade publications. Competitioncontributedtoadrop-offinattendancefromapeakof200,000attendeesand,whenmajorcompanies decided to make big product announcements at CES or other venues,COMDEXdiscontinuedtheevent.CESpickeduptheslack,topping150,000attendeesin2012andagainin2013(Takahashi2013).

Tradeshowsareimportantbecausetheycirculatetechnicalandmarketinginformationaboutproductsandbecause theybuildnetworksofpromoterswhosharethewondersofinformationtechnology.Itisonlyaslightexaggerationtosaythattradeshowsaresimilarto religious events that bring together believers in a magical setting full of icons andsymbols that affirm their mutual faith. On a more practical note, they provideopportunities forwidespread coverage inmainstream and socialmedia that amounts tofreeadvertisingofnewproducts.

Nevertheless,attendanceattheseshowsislevelingoff,asignthatthedaysofthegrandtrade show that aimed to be all things to all participants are nearing an end.The sheernumber of participants as well as the diversity of interests (or faiths) they representappears to be overwhelming the goal of offering anything resembling a clear focus oncommon themes. The mass trade show is suffering some of the same effects as thereligious pilgrimages, such as the Camino de Santiago de Compostela in Spain, whichbecamesopopular that it ismoreandmoredifficult tomaintain theconditionsofquietcontemplation and austerity so attractive to its supporters over the years. “Pilgrims”deckedoutinthelatesthikinggearfromREIandcarryingiPhonesupdatedwiththelatestpilgrimagesapps(eachroutehasitsown)donotexactlyconveythespiritofsacrificeandpovertybeforeGodthatthethousand-year-oldeventwasmeanttoinstill.WhileLasVegasis not Santiago de Compostela, the variety of pilgrimsmaking their way to CES is sooverwhelming thatmanyof thebigcompanies, suchasAppleandMicrosoft,no longershowuporappearonlythroughtheirpartners’products,choosingtofocusontheirownorspecializedeventswithmuchlessclutterthanthebigtradeshow.Suchisincreasinglythecase for cloud computing, which holds specialized events throughout the year. In June2013 I attended the leading cloud-computing and big-data conference and exhibition,CloudExpo,inNewYorkCity.OverfourdaysIheardspeakersfromacross-sectionofcloud companies; participated in cloud bootcamp, a set of sessions spanning thetechnologies that comprise cloud computing and data analytics; and spent hours on theexhibitionfloorobservingandspeakingtoasmanyofthe500orsovendorsasIcould.

The show’s website announcement should dispel any doubt about its promotional

nature: “Recent IDC [International Data Corporation] research shows that worldwidespending on cloud serviceswill grow almost threefold, reaching $44.2 billion by 2013.And a recent Gartner report predicts that the volume of enterprise data overall willincrease by a phenomenal 650% over the next five years. These two unstoppableenterpriseITtrends,CloudComputingandBigData,willconvergeinNewYorkCityatthe12thCloudExpo—beingheld June10–13, 2013, at the JavitsCenter inNewYork,NY.” Moreover, the website proclaimed, “In the most transformative technology shiftsincethepersonalcomputerandtheInternet, it’sapparent thatmigratingbusiness to thecloud has reached a tipping point in 2012, where it is no longer a trend but rather anabsolute business requirement.” And if we needed an exclamation point: “Join us as amediapartner—togetherwecanrocktheITworld!”(CloudExpo2013).Allpilgrimagesexact a price; even las peregrinas who walk the Camino have to pay for equipment,accommodations,andthemuchencourageddonations.ButthepilgrimagetothecloudintheBigApple costs considerablymore.To simply attend all of the conference sessionsoverfourdaysruns$2,500.So,unliketheCamino,thecloudpilgrimage,whethertoNewYorkortoanyofanumberofcloud-trade-showvenues,islimitedtothosewhocanaffordthehighentryfee.

Trade shows build community in several different ways. The registration fee itselfmakescertain thatonlypeoplewhoarestronglymotivated tobepartof thecommunityparticipate. The content ranges across every dimension of promotionalism. Registrantswhoneedbasictraininginthewondersofthecloudcanjoinacloud-computingbootcampandtakeacloud-essentialscourse.Allparticipantshaveaccesstoexhibitorsrepresentingevery type of cloud-computing and big-data company. The exhibition hall is amassivemarketing and sales space. As in any promotional event, whether people are sellingspiritualityorcomputerservices,somearesingledoutasespeciallygiftedinthefield,andthese takeup rolesaskeynote speakerswhosell thecloudandbigdata from theirownpositionswithintheindustry.Whethertheyarecoveringthetrade-offsbetweenthecloudandon-premisescomputing, thepotentialofbigdata to identifycustomersorvoters,orthe transformation of the IT profession from operations to service delivery, there is apattern to the keynotes and the breakout sessions. They tend to begin with a broadoverview that praises the cloud as a general and profitable business tool. This mightinvolvecostcomparisonsbetweendifferenttypesofcloudarrangements:public,private,andhybrid.Next, they identifyaproblemthatbusinessesface,suchasmaintainingdatasecurity or entering the Asian market. Finally, they conclude with a pitch on how theproducts and services of the speaker’s company, whether Rackspace’s hybrid cloud orPacnet’sexperienceintheAsianmarket,willsolvetheproblem.Whateverthesubject,theoutcomeisthesame:followourlead,buyourproduct,andwatchyourbusinesstakeoff.

Despitethebesteffortsoftheself-proclaimedcloudevangelistwhochairedCloudExpoandintroducedthekeynoteandgeneralsessions,occasionaldiscordantnotesreverberatedthroughout the event.At a lunch panel discussion, big-data expertswere asked to statewhatcomestomindwhentheyhearthetermbigdata.Followingtheunwrittenscript,theexpertschirpedtheexpected—“opportunity,challenge.”One,however,refusedtofollowtheirleadandinsteadproclaimedit“abullsh*tmarketingterm.”Asthesayinggoes,youcould hear a pin drop.5 But soon thereafter, the evangelist MC returned to the upbeatmessagethatmightconvincetheaudiencetobuyabig-dataanalyticsservicefromHadoop

orTeradata.Thiseventwasnoexceptiontothewidespreaduseofpropsandinducementstospurattendeestobuythecloud.Asanacademicunusedtothespecialeffectsthatfilltheseevents,Iwasabitsurprisedtohearloudrockmusic,includingheavymetal,blaringin therun-uptoageneralsession.Alsounexpectedwasthepresenceofmodels inshortshorts,thigh-highboots,andsparingnomakeupopportunity,walkingtheconferencefloorandchattingupdelegates.The spokesmodelpresencewas right outof anold-fashionedautoshowexceptforthehigh-techtooleachusedtoscanattendeeconferencebadgesforinformation useful to the company that hired her. In addition, there were the cheesyfreebiessuchasbuttons(I“heart” thecloud;DoITinthecloud),yo-yos,wind-uptoys,and T-shirts (mine supports the hybrid cloud). To trade on the icons of techwork, theexhibition hall featured bean-bag seats for plopping, aswell as foosball and air-hockeygames for unwinding. Exhibitors offered more serious enticements to attract shoppers,such as lottery drawings for tech equipment. One enterprising speaker, in what wasactuallyaninterestingsessiononcloudsecurity,kepttheaudienceintheroombyrafflingtwo state-of-the-art, high-capacity Intel solid-state drives at the end of the session. Inaddition to equipping their spokesmodels with scanners, the conference made use ofanother modern conference add-on by live-streaming the entire event to a worldwideaudience of paying viewers. High-tech gear aside, one of the most remarkable, andremarkably ironic, points in the conference arose when a massive line snaked its waythroughtheexhibitionhall.Itwasbyfarthelongestqueueofthefour-dayevent,withathousand or so people waiting patiently for a very low-tech reward: free copies of ahardcover book on how cloud computing will change everything (Erl, Puttini, andMahmood2013).

CloudExpohelpedadvancemyunderstandingofcloud-computingtechnology,big-datamethodology, and the leading companies that produceboth.But it alsounderscored therole of large conventions in the promotion of cloud computing and big data. Theconferenceandotherslikeitarepromotionalbecausetheyinsistontheabsolutenecessityof adopting cloud computing. They are also promotional forwhat they do not address,primarily the pressures that the cloud imposes on the built environment and on theelectrical grid, the tendency to concentrate power in a few large companies, and thechallenge toemploymentarisingfrombigchanges in the internationaldivisionof labor.Data security and privacy attract a bit of attention, but largely as a threat to cloudadoption.

The forms of cloud promotion that this chapter has considered—commercialadvertising,blogpostsandsocialmedia,promotionalresearchreports,lobbying,andtradeconferences—donotexhaustthemajorexamples.Theycoveragreatdealofground,butthere are other topic areas, including government promotion. In the United States, the2010federalgovernmentchiefinformationofficer’sreporthailingthecloudandorderingagencies to adopt cloud computing was one of the first in a series of governmentpromotional steps. In addition, there was a 2011 National Institute of Standards andTechnology (NIST) report that promised major cost savings for government agenciesmoving their information technology functions to the cloud (NIST2011).Then in2012the National Science Foundation joined the chorus supporting the NIST report andcommitted the government to carry out research on all aspects of cloud computing(NationalScienceFoundation2012).

Allofthepromotionandthehyperboleareimportanttomobilizesupport,which,asthehistoryofcommunicationtechnologydemonstrates,canbefickle,aspeoplecontinuouslyflocktothenextnewthing.Soitisessentialforthosewhoenvisionthecloudasanengineto drive informational capitalism to continually promote its revolutionary capabilities.6Promotion is also essential to protect the cloud from criticisms about its challenges,problems,andevendangers.Thenext twochaptersaddresstheseand, indoingso,raisequestionsaboutthewisdomofmovingtothecloud.

CHAPTER4DARKCLOUDS

The inherent nature of themobile Internet, a key feature of the emergentCloud architecture, requires far more energy than do wired networks….Trends now promise faster, not slower, growth in ICT energy use. (Mills2013,fromthereport“TheCloudBeginswithCoal”)

SECRETSARELIES

SHARINGISCARING

PRIVACYISTHEFT(Eggers2013,303)

CloudcomputingisnothingmorethanthenextstepinoutsourcingyourIToperations.(McKendrick2013c)

There isnoquickerwaytodescendfromthecloud than to look inonanold-fashioned,down-to-earthdisputeaboutmoneyandpower.SuchwasthecasewhentheveteranNewYorkTimesreporterJamesGlanz,knownforhisworkasBaghdadbureauchiefandforaninvestigativehistoryof theWorldTradeCenter (GlanzandLipton2004), arrived in thetown of Quincy in centralWashington to do a story on cloud-computing data centers.Therehefoundadisputebetweenacomputergiantandasmallpowercompany.Now,thiswasnoordinarybigcomputercompany—itwasMicrosoft,thebusinessthat,inthemindsofmany,savedthestateofWashingtonfromthefateofotherdecliningindustrialregionsby setting up its headquarters there rather than in Silicon Valley. In 2006 Microsoftdecidedtoexpandbybuyingseventy-fiveacresofanoldbeanfarmandbuildingadatacenter to support its cloud services. The company was drawn by the abundance ofhydroelectricpowerproducedbygeneratorsoperatedfromthenearbyColumbiaRiver.Itwasalsoattractedbyutilityratespriced,thankstoitseffectivelobbying,atlessthanhalfthenationalaverage,whichbroughtareliableflowofpowermadepossiblebydamsalongtheriver, including twooperatedby the localpowercompany.Finally,Microsoftsoughtandreceivedgeneroustaxbreaksfromthestatebecauseitpaidpropertytaxestothetown,helpingtopaveroadsandbuildanewlibraryforQuincy’s6,900residents.Theheadofthe power company summarized a general feeling when the company came to town:“You’re talkingaboutoneof the largest corporations.You’re talkingMicrosoft andBillGates.Wow!”(Glanz2012a).

Itdidnot takelongforthewowtoturnintopowwhenaQuincycitizens’grouptooklegalactionagainstMicrosoftforpollutionspewingfromfortydieselgeneratorsthat,asiscommon at data centers, the company deployed for its primary backup system. The

softwaregiant’sfacility is locatednearanelementaryschool,andparentsandneighborsfeared the toxic effects, especially for young students.The termbackupgenerator doesnotsoundparticularlyharmful,butthoseusedindatacentersarenotthekindhomeownerskeep in the garage. They are over ten feet tall and weigh thousands of pounds each,enough togenerate2millionwattspergenerator. Justas significant, theygetuseda lotmoreoften than the termbackupwould indicate,particularlyduring frequentperiodsofbuilding construction.The state had initially permittedMicrosoft to use them for 6,000hours over the course of a year for emergency backup power or for “maintenancepurposes” (ibid.). It appears,however, that thecompanyactuallyused thegenerators sofrequentlyduringaperiodofdata-centerexpansionthatitaskedtobeunpluggedfromtheelectricalgridtorunentirelyondiesel.In2010,MicrosoftranitsQuincydieselgeneratorsfor3,615hours,sendingintotheairparticulatematterthatstudiesofotherMicrosoftdatacentersfoundcontainedenoughcarcinogenstoposeathreattopeoplelivingandworkingin the area. No assessmentsweremade inQuincy, but residents knewwhen the dieselgenerators powered up. According to a forklift driver who works at a local fruitwarehouse,“Whentheyfirststartup,abig,hugecloudofblacksmokecomesup.Itjustkindofmakesyounauseous” (ibid.).Asmorecompaniesandmoredatacentersmovedinto town, even more diesel was used, all of which generated environmental hearings,lawsuits, and a lot of delicate negotiations, leaving an environmental engineerwith thestate’sDepartmentofEcologytoconcludeinsomeexasperation,“Ifindithardtobelievethatthisisthebestwaytostoredata.Something’sflawedinthatthoughtprocess”(ibid.).

Alongsidethefightoverdiesel,anotherdisputearosebetweenMicrosoftandtheutilityoverpowerusage.Asisgenerallythecase,theutilityrequiresestimatesofpowerusagefromitslargecustomersinordertoefficientlymanagethegrid.Thisissueissoimportantthatpowercompaniesarepermittedtofinefirmsthatsignificantlymissestimates.InthiscaseMicrosoftoverestimatedandwasleviedafineofslightlymorethan$200,000.Muchtothesurpriseandchagrinoflocals,thecomputergiantnotonlyrefusedtopaythefine,but proceeded to burn millions of watts of power in what it admitted was an“unnecessarilywasteful”manneruntil theutilityagreed toslashorcompletelyerase thefine.InMicrosoft’sview,ifitwasgoingtobefinedforoveruse,thenitwouldsimplyburnoff enoughpower to raise itspower consumptionabove the level thathad triggered thefine.OnemightthinkMicrosoftwouldpaythefineandenjoysomepositivepublicityforusing less power than it estimated. Indeed, Yahoo! faced just such a fine and paid it.However,Microsoftdecidedagainstdoingso,anditspowerusejumpedfrom28.5to34million watts in three days. Under pressure, the utility board voted to cut the fine to$60,000,andMicrosoftendeditsfuel-burningprotest.

It is little wonder that a utility commissioner and local farmer commented, “For acompanyofthatsizeandthatnature,andwithallthe‘green’thingstheyadvertisedtome,that was an insult” (ibid.). Microsoft, for its part, claimed that this was an isolatedincident.But itwas actually just onemore in a long list of issues creating tension andoutright conflict between the company and the farming community. Amere three daysaftertheribbon-cuttingceremonywelcomingthecomputergiantandpresentingthelocalgeneral manager with a bag of beans from the last harvest on the land and a signannouncing, “Preparing theSite forAnotherFarmer:Microsoft,” tensions rose over theability of the town to meet the company’s electrical-power needs. The data center’s

generalmanagercomplainedthattheutilitywasslowtobringonboardasubstationthatwould provide 48millionwatts of power to theMicrosoft facility, or enough to powerabout30,000homes.Arguingthatslowconstruction“dramaticallyaffectsouragilityasabusiness,” theMicrosoft official informed the utility that “our confidence is becomingquiteshaky”andwonderedif,intheabsenceofspeedierconstruction,thecompanymightbe eligible for $700,000 in reimbursements. This struck one utility official asdemonstrating “a level of arrogance” and confounded others, including a retiredschoolteacherwhohadfeltthat“Microsoftwouldbringalittleclasstothetown”(ibid.).

DespiteitsproblemswithMicrosoft, thetownhasnotturneditsbackondatacenters,approvingconstructionforYahoo!andDell,alsoattractedbythepromiseofcheappowerand tax breaks.By the end of 2012, littleQuincy had two supermarkets, two hardwarestores, and sixdata centers,with fivemoreunder construction, but nomovie theater orMainStreet.Sometownresidentsandbusinessesworrythat,withmanycompaniesnowchasing lower utility costs, the power company might have to raise rates for localcustomers.Theyarealsoconcernedthatpower-hungrydatacentersmightcreateanactualpowershortage,aremarkableironygiventhetown’sproximitytotheColumbiaRiveranditshydroelectricdams.Alocalfruitgrowerintheareaconcludedthattheoverallimpacthas been much less positive than most people imagined: “I don’t think it’s benefitingQuincy.” Although he recognizes the importance of data centers to the Americaneconomy,“Ithink,”hesaid,“we’retakingonefortheteam,totellyouthetruth”(ibid.).

While details may differ, there is nothing particularly unusual about Quincy’sexperiencewith thecloud.Manypeoplearenow“takingonefor the team”tobuildandoperatecloudcomputingsystems.Indeed,incidentsoflegalactionforallegedviolationsof environmental regulations, utility agreements, promised employment for localresidents, and other related issues comeup time and time again after the cloud arrives.ThisledGlanztoconclude,“WhentheseInternetfactoriescometotown,theycanfeelabitmorelikeold-timemanufacturingthanmodernmagic”(ibid.).Nordotheyfeelliketheclouds described in promotional accounts. As long as environmental officials inWashingtonStateandlocalcitizensinplaceslikeQuincycontinuetothinkofdatacentersas clouds rather than as factories, they will continue to have problems making sounddecisions.It isstriking,butnotsurprising,toobservehowconfusedpeopleappeartobeabout the cloud. I have talked to people with graduate degrees who still think it hassomething to dowith actual clouds,with communication satellites, orwith theweather(e.g.,thesystemcangodownintherain).Surveysconfirmthepublic’sconfusionaboutthecloud(Linthicum2013a).Atbest,thegeneralpublicseesitas“onebigstoragespace,”whichatleastcomprehendsonepieceofthecloudpuzzle(Abdul2013).1Gettingahandleon the language matters a great deal, especially in an era that lauds advertising andpromotion, suitably dressed up in terms like strategic communication. Giant powerprojectsinthedesertarecalledsolarfarms(Soto2011),anddiesel-spewinginformation-processing factories areknownasclouds. Factories,whether in the desert or in a smalltown, arenot inherentlybad,butpeopleneed toknowwhat they really arebefore theyapprove construction, determine whether or what kind of incentives to provide, andestablishanappropriateregulatoryregime.

Thischaptertakesasteptowardprovidingacriticalunderstandingbyexaminingsomeofthemajorproblemsassociatedwithcloudcomputing,concentratingonenvironmental

andpowerissues,privacy,security,andemployment.

E-pollutionAdvertising aside, we have known for some time that computers are not a greentechnology.Chemicalsusedintheircomponentsareamongthemostcarcinogenic.SiliconValleylongledthelistofextremetoxic-wastesitesintheUnitedStates,andtodayChinaand many poor nations contain mountains of computer parts making up a dangerouschemicalstew.AccordingtoMaxwellandMiller(2012a,3),by2007between20and50milliontonsofe-wasteweregeneratedannually,mostofitfromcellphones,televisions,andcomputersthatpeoplesenttothedump.E-wasteismostlyproducedinthedevelopedWestanddisposedofinLatinAmerica,Africa,EasternEurope,India,SoutheastAsia,andChina.Inrecentyears,IndiaandChinahavejoinedtheleadersinwasteproduction.Overthe ten-year period from 1997 to 2007, the United States alone discarded 500 millioncomputerscontainingover6billionpoundsofplastics,over1.5billionpoundsoflead,3millionpoundsofcadmium,almost2millionpoundsofchromium,and632,000poundsofmercury,aswellasmanyotherdangerousandcarcinogenicchemicals, likeberylliumandgalliumarsenide(MaxwellandMiller2012a).

E-wastehasbeendescribedasa“growingtoxicnightmare”andwithgoodreason.AsLeylaAcarogludescribesit,

In far-flung,mostly impoverished places likeAgbogbloshie,Ghana;Delhi, India;andGuiyu,China,childrenpilee-wasteintogiantmountainsandburnitsotheycanextractthemetals—copperwires,goldandsilverthreads—inside,whichtheyselltorecyclingmerchants foronly a fewdollars. In India, youngboys smash computerbatterieswithmallets torecovercadmium,toxicflecksofwhichcover theirhandsand feet as they work. Women spend their days bent over baths of hot lead,“cooking”circuitboardssotheycanremovesliversofgoldinside….Mostscientistsagree that exposure poses serious health risks, especially to pregnantwomen andchildren.(2013)

Fromtheirearliestdays,onemajorargumentmadeaboutcomputershasbeenthattheyprovideanenvironmentally soundalternative to theproductiveenginesof the industrialera. Scholars, including most who are otherwise critical about information technology,havegenerallyignoredtheirimpact.Moreover,asMaxwellandMiller(2012a,13)noteinoneof the fewsustainedaccountsof theenvironmentalproblemsassociatedwithmediatechnology, well-regarded academics who are quick to point out the excellent useenvironmentalistsmakeofnewmediahavenothingtosayabouttheprofoundironyofthisactivity. At best, research advances the viewwell stated in a 1998 article by a trio ofscholars who, in the first wave of the Internet’s growth, sought to understand therelationship between environmentalism and the information society: “On the one hand,there is the potential for reducing the stress on the environment: the emergence ofinformation technologies and services can lead to a dematerialisationof production andimmaterialisation of consumption” (Jokinen, Malaska, and Kaivo-oja 1998). This putssuccinctly thepromiseof IT topromote amore sustainableworld.Computers linked tocommunicationsystemscancreatesmartersystemsofproductionthatrequirelessmaterialinput and create lessmaterialwaste. Just as important, the process of getting goods toconsumers ismade lessmaterial, in part because an information society requires fewer

material products and also because the process of getting things to consumers ismadesmarterandmoreefficient.

Onecancertainlyunderstandwhy thisviewwould receivesupport. Iwrote thisbookputtingpracticallynoinktopaperbecauseIusedmylaptopanddrewfromthevaststoresof online information for research.To its credit, andunlike themanypositive forecastsabout green IT, the 1998 article also raises the risk that “positive environmental effectsmightbeovercomebythe‘reboundeffect’causedbyexcessiveeconomicgrowth”(ibid.).That success on the environmental front can encourage people to consumemore is notdissimilarfromothercounterintuitiveeffects,suchasthelinkbetweenadvancedbrakingsystems and the number of accidents. Trusting the brakes can lead to more recklessdriving, just as progress on environmental controls can encourage people to buymore,especiallymore“green”products.

The counterintuitive effect embodies good dialectical thinking, but it neverthelessretains the view that information technology is inherently green. The consequences ofusing ITmay indeed lead to greater consumption and resource depletion; however, thethinking goes, this is due not to the technology but rather to what we dowith it. Theexpansionofcloudcomputingdemonstratesthelimitationsofthisview,particularlywhenone considers the genuine materiality of production that takes place in the large datacenters around the world. On the outside, they appear to be enormous rectangularwarehouses, perhaps distinguished by their lack of unique identification and minimalexposure to outside light. Inside, they are far from the storage facilities that typicallydefineawarehouse.Insteadtheyarefilledwithactivedevicesandsystems,includingrackuponrackofserversprocessingdataandmultiplepowerandcoolingsources.Accordingto a lawyer forMicrosoft, “The heart of the cloud are these data centers, and the datacentersarereallyattheheartofMicrosoft’sbusiness”(Glanz2012a).

Wenowhavetensofthousandsofdatacentersspanningtheworld,permittingpeopletoinstantly download their Google mail, search on Baidu, buy music and movies fromiTunes,andpurchaseproductsofeverykindfromAmazon.Butallofthesebenefitscomeatthecostofincreasedpoweruseandmorestressontheenvironment.Clouddatacentersare filledwith thousandsofservers,eachcomprisingcommonandrarematerialswhosedisposal raises serious issues of water and soil contamination. There are few morearresting images than those in Edward Burtynsky’s documentary ManufacturedLandscapes of elderlyvillagewomen inChinapicking throughmountainsofhazardouscomputerwasteforsomethingtosell.Thisscene,repeatedagainandagainthroughoutthemany placeswhere detritus from the cloud finds a not-so-final resting place, belies theimageofanimmaterialinformationage.Admittedly,thisproblemdoesnotmakeforthedystopiandramaofnuclear-wastedisposal,arealitythathasitselfslowedthedevelopmentofnuclearpowerbyprovidingpoliticalballastforitsfoes.Moreover,themushroomcloudassociatedwithnuclearweaponsisafarmorearrestingdeterrentthanthepuffycloudsofour informationage.But in some respects, thechallengeof thecloud’se-waste ismoreinsidiousbecauseitshazardsarenotsoimmediatelythreateningandbecausethebulkofthedamageisdoneinpoorcountries,wheremostsuchwasteisdumped,orinthepoorerregionsofrichernations,suchasinruralChina.

The need to keep the heart of the data center beating requires a constant stream of

power. As a result, the facilities need reliable sources of electricity for their 24/7operationsand,forthosetimeswheneventhebestelectricalsystemsshutdown,backupsystems, including the dieselpowered generators described in the Microsoft story.Furthermore, inmost cases, additionalbackup isprovidedbyamassive supplyof lead-acid batteries and banks of flywheels whose spinning offers additional reserve power.Even with all of this expensive, polluting backup, there is still no guarantee of 24/7performance,asMicrosoftitselflearnedwhenitexperiencedaworldwidecrashofseveralmajor cloud services because it failed to renew a security certificate in 2013 (Ribeiro2013).

Theneedforreliable,low-costelectricityforbothpowerandcoolingisacomplicatedcoupling that influences locational decisions and helps to shape the politics of datacenters. The power demands alone are astounding. As an engineer who has designedhundreds of the centers described, “It’s staggering formost people, even people in theindustry,tounderstandthenumbers,thesheersizeofthesesystems.Asingledatacentercan take more power than a medium-size town” (Glanz 2012b). Estimates vary, butexpertsagreethatdatacenters’powerconsumptionaccountsforroughly2percentofalltheelectricityconsumedintheworld,andtheircarbonemissionsaresettoquadrupleby2020(DataCenterJournal2013).Over the longrun, theseratesofelectricalconsumptionare less thansustainable,and

companiesareactivelytryingtofindsolutions.Butthis isnoteasybecausedatacentersareprofit-makingenterprisesthatkeepcustomersbymaintaining24/7access.Moreover,theirsystemsneedmorethanjustaconstantsupplyofpowertooperate.Theyalsoneedameans of maintaining a sufficiently cool environment to prevent their servers fromoverheating. It should therefore come as little surprise that the coal industry expects arevenuebonanzafromcloudcomputing.Inadetailedreport,theassociationrepresentingBig Coal in the United States contradicts all the forecasts that cloud computing willeventually diminish energy requirements for companies that use the cloud and for thecloudindustryitself(Mills2013).

Companies can do some things to moderate power consumption, including locatingtheirfacilities inplaces likeScandinaviaandCanadathatprovidebetternaturalcooling.But,asthesectiononsecurityissuesshows,storingdataoutsideone’sbordersraisesotherconcerns.Companiescanalsobetterattunetheirpowersystemstotimeswhenserversareactivelyengagedinprocessing.Butthisisdifficulttoaccomplishbecausecloudproviderslike to keep the power flowing so that, in the event of a sudden spike in processingdemand,theirserversdonotcrash.Cloudcompaniesknowthatcustomersdonotliketoseeanydelayordowntimeintheiremailuse,indigitalproductdownloads,orinaccesstosocial-media sites, and theyworry that customerswill turn to another provider or loseinterest and cut back on their discretionary activities in cyberspace.Nevertheless, somefirmsaretakingaction.

HPhasdevelopednewserversthatrequirelesspower,aninitiativethathashelpeditsbottomlineevenasitearnslessthanitusedtoonallofitsotherlinesofbusiness(Sherrand Clark 2013). Companies are also developing innovative power systems tosubstantiallyreduce,ifnoteliminate, theneedtocoolserverselectrically.2Yahoo!madethe decision to build a data center outside Buffalo, New York, that uses hydroelectric

power,which substantially lowered its carbon footprint (Greenpeace International 2010,3).AlthoughGooglekilleditsthermalpowerprogram,thecompanyhasusedwindpowerforadatacenterinIowaandsetupanelectricitysubsidiarytosellpowerbacktothegrid(Barton2012).EspeciallysinceitsufferedabarrageofnegativepublicityforlocatingoneofthelargestdatacentersinNorthCarolinaandchoosingtodealwithacompany,DukePower,withanotoriousenvironmentaland labor record,Applehas takensomesteps todevelop sources of renewable energy (Clancy 2012). Finally, Salesforce has developednewmetrics, includingcarbonproducedper transaction, tobettermonitor itsenergyuse(Makower2012).

Evenifcompaniesmanagetoincreaserenewableenergysourcesforclouddatacenters,significantenvironmentalproblemswillremain.Thatisbecausemostpeopleaccesscloudsystemswirelessly,and,asa2013reportconcluded,wirelessaccessconsumesenormousamountsofenergyanddoessolessefficientlythanthedatacentersthathavecomeinformost criticism (Center for Energy-Efficient Telecommunications 2013).Moreover, it isimportanttoobservethatmostofthegreenshootsinanotherwisebleaklandscapesproutwithin the United States. There are exceptions. Greenpeace named the giant Indiantechnology outsourcing companyWipro the greenest electronics company in the world(Swinhoe2013).However, thematerialconstructionofcloudcomputing requiresglobalsupply chains whosemany links outside the United States give rise to daily stories ofenvironmental ruin.SoevenasApplewas trying toburnish its reputationforproducingsolarpowerinNorthCarolina’scoalcountry,oneofitssuppliersinChinawasdiscoveredto have killed a river outside Shanghai with e-waste resulting from the production ofAppleproducts.AccordingtoanaccountintheFinancialTimes,theApplecontractorhasbeenturningtheriveramilkywhitejustabouteveryweekoverthetwoyearsithasrunthe industrial park facility, prompting this comment from one waste-treatment plantworker:“Beforethat,therewerefishandshellfishintheriverthatweusedtoeat.Butnowtherearenofishatall.Andwhenthewaterturnswhite,wecan’tevenuseittowaterthevegetablesanymore”(Mishkin,Waldmeir,andHille2013).ThelocalcompanyisfacingsanctionsfromtheShanghaigovernment,but it isunlikely that therivercanbebroughtbacktolife.Storieslikethisprovideanimportantreminderthatthecloudisgroundedinaglobal system of production that is material, industrial, and, unless there are majorchanges,unsustainable.

One result of the “always-on” commitment is that server operation is woefullyinefficient.WhentheNewYorkTimescommissionedMcKinseyandCompanytoexaminetheenergyuseofdatacentersprovidingcloudservicestoavarietyofcustomers,itfoundthat they were using only between 6 and 12 percent of the electricity powering theirservers to perform actual processing operations (Glanz 2012b). Companies keep theelectricityflowingforfearthatservicewillnotbeavailablewhenitisneeded.Customersleasingfacilitiesdonotwanttohearaboutdowntimeandarenotreluctanttofindanothercloudproviderif24/7serviceisnotprovided.Soengineersworkingforcloudcompanieslaborinfearoflosingtheirjobsiftheyarecaughtwiththeirserversdown.Bettertopowerunused servers than to face an angry customer.According to one executive at a utilityfirm,“It’sanervousnessintheI.T.communitythatsomethingisn’tgoingtobeavailablewhen they need it” (ibid.). There is practically no incentive to save energy and everyincentivetokeepthesystemgoing.AsaseniorindustryexecutivetoldtheTimes,“Thisis

anindustrydirtysecret,andnoonewants tobethefirst tosaymeaculpa.Ifwewereamanufacturing industry,we’dbeoutofbusiness straightaway” (ibid.).The termdirty isappropriateinmorethanonesense.

Another not-so-little secret is the reliance on very un-cloud-like backup systems toguarantee against an electrical power failure.These include diesel generators like thosedescribed in the case of Microsoft’s data center in central Washington. Data centersthroughout Silicon Valley have been cited on the state of California’s Toxic AirContaminant Inventory for diesel air pollution. Since many jurisdictions lack such atrackingmechanism,theycannotmonitortheeffectsofdieseluseandsomustsuffertheeffectsoftoxinsandcarcinogensortrytheirluckwithlegalaction,asdidthecitizensofQuincy.Diesel generators are not enough for an industry determined to provide instantservice,ondemand,anytime.Thesealsotendtobebackedupbythousandsoflead-acidbatteriesof the typeusedin trucksandcarsandbyenormousflywheelswhosespinninggeneratesmorebackuppower.Astafferataninstitutethatstudieselectricalpowerusageisnotimpressed:“It’sawaste.It’stoomanyinsurancepolicies”(ibid.).Ofcourse,data-centermanagersunderintensepressuretodeliverallthetimewoulddisagree.Microsoftisnottheonlycompanytobepenalizedforviolatingenvironmentalregulations.InOctober2010,Amazonwas issued a fine of slightly over $500,000 by the state ofVirginia forbuilding, installing, and continuously running diesel generators without obtaining thenecessarypermitstodoso.Afterappeals,thefineswerecuttoabouthalfthatamount,butfour inspections and a total of twenty-four violations ranked “high” do notmake for arecord toboastabout,especially foracompanyclaiming leadership incloudcomputing(Barton2012).

The cloud industry,whichprofits by storing andprocessingother people’s secrets, isamong themostsecretive itself.Companiesdonot reveal the locationof theirowndatacenters,whichtendtobehousedinnondescriptwarehouse-likebuildingswithnosignsormarkings.Makingmattersmoredifficult, theUnitedStatesandothernationswith largenumbers of data centers have no single agency responsible for overseeing them. TheUnitedStatesknowshowmanygovernmentdatacentersithas—2,094in2010—butdoesnotknowhowmuchenergytheyconsume.Thisdoesnotjustcreatearegulatoryissue;italsocreatestheconditionsfordisaster.Asonetechnologyandpowerindustryconsultantconcluded,“It’sjustnotsustainable.They’regoingtohitabrickwall”(ibid.).

Publicawarenessisgrowingaspressuremountsfromenvironmentalgroups,especiallyGreenpeace. In 2010, the activist organization issued a report on cloud computing thatchallenged the major providers to do a much better job of taking into accountenvironmentaldamage.Specifically,ittookFacebooktotaskforbuildingadatacenterincentral Oregon serviced by a utility that primarily uses coal-fired power stations, thelargestsourceofgreenhousegasemissionsintheUnitedStates(GreenpeaceInternational2010).GreenpeaceusedthereporttolaunchacampaigndubbedUnfriendCoal,completewithaFacebookpagethatattracted700,000supportersandsetaGuinnessWorldRecordfor most comments on the social-media site in a twenty-four-hour period. In 2011 theorganization issuedanother studyoncloudcomputing thatprovidedspecificdetailsandgraded cloud companies on their performance. This report gave Facebook an F in“Infrastructure Siting” for the social-media company’s continued reliance on coal-firedplants(GreenpeaceInternational2011).Ayearlater,Facebookreachedanagreementwith

Greenpeacebypledging,amongotherthings,tochangeitsdataplantsitingpolicies.Asidefromthecommitmenttoreducedependencyoncoal-firedplants,Facebookwasshortonspecifics.ButGreenpeacetookthisasastepintherightdirection.

The Greenpeace reports did not just call out Facebook for failure to “like” theenvironment.Nocompanyfaredespeciallywell.Inthe2011reportTwittercameouttheworst, with F marks in all three categories of transparency, a measure that includedopennessaboutitsenvironmentalpolicies,infrastructureciting,andmitigationstrategy.Inkeepingwiththesecrecywithwhichnewmediacompaniesoperate,AmazonreceivedanFfortransparencybutsqueakedoutDmarksintheothercategories.Apple,whichfaredslightlybetter(twoCsandanF),hadtheworstrecordforcoalintensity,faringabitworsethanFacebook.However,themoststrikingfinding,andthemostdisappointingforanyoneexpectingdifferentbehavior fromcompanies that like topolish theirownhalos, is that,with rare exceptions, cloud-computing companies, including all the big names inhardware,software,socialmedia,andbigdata,behavenodifferentlyfromtheirindustrialpredecessors.

Greenpeace has not only taken the lead in shining a light on the sad environmentalrecordofITcompanies.Ithasalsobeenaleaderinactivism.InApril2012,peoplewiththeorganization climbed to the topofAmazon’snewcorporateheadquarters inSeattle,directlyacrossthestreetfromMicrosoft’scorporatecenter,andrappelledfromtherooftohangabannerintheshapeofacloudthatread,“Amazon,Microsoft:HowCleanIsYourCloud?” Following the event, Greenpeace’s IT analyst explained the protest toWiredmagazine:“Ifwewanttogettoarenewableenergyeconomy,wecan’tgettherewithoutleadership from these companies. For too long, toomany of the energy decisions havebeendictatedbyasmallsetofcompanieswhoareveryhappywiththestatusquo”(ibid.).The companies insist that they aremaking positive strides, but they alsomaintain thatlargedatacentersareintrinsicallybetterfortheenvironmentthanhavingeveryindividualor organizational user house its own data (ibid.). Environmentalists insist that Amazonwillhavetodomuchmorethanbuildphoto-op-readygreenhousesindowntownSeattle.

Itisdifficultenoughtocontemplateasustainablecloudfromthesupplyside,butitisevenmorechallengingwhenoneconsiderstheseeminglyunstoppabledemandforcloudservicesfromorganizationsandindividuals.Supplyanddemandareinterconnected,asisevidentthroughoutthepromotionalcultureofcloudcomputing.Forthosewhomarketthecloud,customersshouldnotonlywantcloudservices,butalsodemandthemasaright.A2013advertisementforSprintmakesthisabundantlyclearasayoungmalevoicerecitesaspiritualodetotechnologywhileasublimemontagezipsby:

Themiraculousiseverywhere.

Inourhomes,inourminds.

Wecanshareeverysecond

indatadressedaspixels.

Abillionroamingphotojournalists…

Uploadingthehumanexperience.

Anditisspectacular.

Sowhywouldyoucapthat?

MyiPhone5canseeeverypointofview…

Everypanorama.Theentiregalleryofhumanity.

Ineedtouploadallofme.

Ineed,no,Ihavetherighttobeunlimited.

OnlySprintoffersTrulyUnlimiteddata

foriPhone5.(Sprint2013)

“Ineedtouploadallofme.”Whynot?Sincemostpeoplebelievethatdigitalbitsaredifferent fromatoms, this is a consequence-freechoice.However theymighthavebeenburnt, or at least jadedabit, by thedotcombustof the early2000s,many remainwithformer director ofMIT’sMedia LabNicholasNegroponte (1995) and former editor ofWiredmagazineKevinKelly(2010),aswellascountlessothermythmakers,whoinsistedthat thedigitalworldnotonlydiffered from theworldofmaterial atoms; it representedanotherorderofreality.Beingdigital,asNegroponteinsisted,meantlivinginaworldoflimitlesspossibilitiesunboundedbythephysical,material,andenvironmental limits thatconstraintheworldofatoms.Aspowerfulasthisvisionhasbeenfordrawingaworldintotheetherofcyberspaceandnowthecloud,it isfundamentallyflawed.Theresourceandenvironmentalproblemsofthedigitalworlddemonstratethatthedigitalandthematerialareinextricablybound.3Negroponteandthosewhofollowedinhispathwerewrong.Theworld of atoms is not ending; it weighs upon us ever more powerfully, with everyadditionalpetabyte, in thedigitalworld’sseeminglyrelentlessgrowth.CloudcompanieslikeGooglearguethatthecentralizationandrationalizationofpowerusethatashifttothecloudenableswilldiminishoverallbusinesspowerconsumption.Butamodelbasedonresearch funded by Google that appears to demonstrate this has met with skepticism(Bourne2013).Moreover, reports fundedbyGreenpeace International (2010,2011, and2012) and by the U.S. coal industry (Mills 2013), typically adversaries, conclude thatoverallbusinessenergyconsumptionwillinsteadgrowsubstantially.

Building an environmentally sound or sustainable digitalworld requires fundamentalchangesinthebehaviorofITcompanies,includingthoseleadingtheflighttothecloud.Justbecausetheyareproducing,processing,distributing,anddisplayingadigitalproductdoes not mean that companies can avoid the environmental consequences of theiractivities.But it also requiresa fundamentalchange in thepeopleand theorganizationsthat download, upload, transmit, receive, and display the digital world. It is no morereasonable, and no less environmentally impactful, to demand aworld of limitless datathanitistodemandaworldoflimitlessgoods.Neithercomeswithoutacost,andneitherissustainablewithoutmajorchangesinconsciousnessandmaterialpractices.AsMaxwelland Miller eloquently conclude, “There are technological fixes for the Internet’senvironmental problem—moving data centers off the coal-fired power grid and ontohydro-electric, solar, geothermal and other sources; designing energy efficient devices;andusingsmartgridstoregulateandreducedomesticandworkplaceenergyconsumption.Butthesefixeswillnotsucceedwithoutacorrespondingtransformationofourconsumercultureintoacultureofsustainability,onethatensuresthatsocial,political,andeconomic

development does not exceed or irreversibly damage theEarth’s abilities to supply andrenewthenatural resourcesuponwhichwedepend” (2012b).Thiswillbedifficult,andthere is little time to lose. AsMaxwell andMiller also note, there are now 10 billiondevices to power, and these soak up 15 percent of all global residential energy. If thecurrent rateof adoptioncontinues—that is, if there isnochange in thebelief that thesedevicesimposeno,orlittle,burdenontheenvironment—thentheywillrequire30percentof the world electrical grid by 2022 and 45 percent by 2030. Meanwhile, the powerdemandsofclouddatacentersareexpandingatanevenfasterrate,growingby56percentbetween 2005 and 2010, at a time when worldwide industrial energy growth was flat(ibid.).

PrivacyandSecurityPrivacyandsecurityconcernsarecomingtogethertoformanotherquestionmarkovertheIT industry, including cloud computing and big data. In order to properly assess theseconcerns, it is useful tobeginby consideringdifferentways to think aboutprivacy andsecurity.At the risk of some simplification, consider three alternatives that range fromweak to strongprivacyprotections.Startingat theweakend,onecanviewprivacyandsecurityastradablecommodities.Webelieveintherighttobeleftaloneandtofeelsecurebutarewillingtogiveupsomeoftheprotectionsaffordedinordertoachieveothergoals.Thisincreasinglyincludesthedecisiontotradesomeofourprivacyandsecuritytoliveinthe cloud by posting on Facebook or Twitter and downloading videos from Apple’siCloud.Fortheabilitytodothesethings,werisklosingsomeofouridentitytohackersorgivingupinformationaboutourselves,includingthecontentofourpostingsortheprofileestablished by our purchases, to the companies that provide the service, as well as tooutside parties that purchase information about us from Facebook, Twitter, and Apple.Sometimesthedealwithacloudproviderisnotclear.Iknowapersonwho,afterlettingherFacebookfriendsknowaboutaseriousillness,beganreceivingadsfor“bucketlists.”Ofcourse,shewasn’tlookingforabucketlistwhenshegaveupsomeofherprivacyinordertoletfriendsknowaboutherhealthissue.Norwasthepersonwhostartedreceivingadsformultiplesclerosissupportservicesafterdoinganonlinesearchofsitesdevotedtothecondition(Singer2013).Theoutcomeisnotalwaysthisoffensive,butitcanalsobeworse,aswheninnocentonlinesearchesforpressurecookersandbackpacksledtoahomevisitfromsixmembersofaterrorismtaskforce,who,wesoonlearned,regularlycheckonpeoplewhoseuseof the Internetprovokes suspicion (Bump2013).Whether thedeal isclearornot,inthisfirstview,privacyandsecurityareamongtheseveralthingswedesire,andwemakechoicesabouttheminthecontextofotherthingswewant.

In the middle of the continuum, privacy and security are no longer tradablecommodities;rather,theyareuntradablevaluesthatdefineacitizen’srighttobeleftaloneandsecurefromviolations.Fromthisperspective,thereisnotrade-offinmoney,services,or goods because privacy and security are not commodities. Rather they are rights tofreedomfromidentitylossandfromphysicalormentalviolation.Seenfromthispointofview,lawandcustomshouldprotecttherighttobeleftalone,whichcannotbetakenawaywithoutviolatingarightofcitizenshipandthereforecannotbetradedformoney,goods,orservices.WhenGoogle,Amazon, orMicrosoft tracks us,we lose some of our privacy.Whatweappeartogetinreturnisactuallyunrelatedtoprivacy.Itisaserviceprovidedbythecompany forwhichwemight ormight not pay.But since, from this point of view,privacyisnotacommodity,wecannotuseitasacurrency.Whenweagreetoawebsite’s“privacypolicy,”weareactuallyonlyacceptingthatweknowaboutitsprivacyviolationpolicy. We rely on government to protect this citizenship right, and when it allowscorporations todiminishourprivacy,orwhengovernment itself takesawayourprivacyandsecurity,itisfailingtoupholdafundamentalright.

Bothof theseapproachesprovideusefulwaysof thinkingaboutprivacyandsecurity.Buttheyareweakinconveyingasenseofwhatprivacyandsecuritydoforusorwhyweshouldcaredeeplyaboutthem.Forthatweturntoathirdperspectivethattriestoaddressthesepointsas itprovides thefoundationfor thestrongestprivateprotection.Accordingthisview,privacyandsecurityaresignificantmeansofprovidingthespace,thebreathing

room, or the buffer between our selves and the world that is necessary for self-development. They offer an essential space between the individual and the world,including those elements of the world that might benefit from taking, purchasing, orotherwisecarryingoutsurveillancethatviolatesthisspaceandmakesitmoredifficulttosafelydevelopaselfandanidentity.Inthisreading,privacyviolationsareattacksonourcapacityforself-development.

Dissatisfiedwithwhattheyperceiveasweakversionsofprivacyandsecuritythatfailto address why these values are important, a number of observers and scholars haveadoptedtheself-developmentperspective.AswriterJathanSadowskiexplains,“Sincelifeandcontextsarealwayschanging,privacycannotbereductivelyconceivedasonespecifictypeofthing.Itisbetterunderstoodasanimportantbufferthatgivesusspacetodevelopanidentitythat issomewhatseparatefromthesurveillance, judgment,andvaluesofoursocietyandculture”(2013).Scholarshavedeepenedthisview.ForlawprofessorJulieE.Cohen,itmeans“creatingspacesforplayandtheworkofself-making”(2013,1911).ForWoodrow Hartzog and Evan Selinger, privacy protection goes well beyond keepingbusinesses from gathering information about us for profit; privacy—or, in their terms,obscurity—is essential for democratic societies because it guards “autonomy, self-fulfillment,socialization,andrelativefreedomfromtheabuseofpower”(2013).Finally,forMichaelLynch,privacyisessentialfor thegrowthofhumanautonomy;puttingit instrongterms,heinsists,“Howeverweresolvetheseissues,wewoulddowelltokeeptheconnections between self, personhood and privacy inmind aswe chew over the recentrevelationsaboutgovernmentalaccesstoBigData.Theunderlyingissueisnotsimplyamatterofbalancingconvenienceandliberty.Totheextentweriskthelossofprivacywerisk, inaveryrealsense, the lossofourverystatusassubjective,autonomouspersons”(2013).

When Facebook develops tools, like the social search engine Graph Search, thatcombinespiecesofouridentitywiththird-partydataandthenmarketsthisinformationtoadvertisers,ittakesoverthespaceofself-development,limitsourbreathingroomtocarryout the task of forming an identity, and lessens our ability to develop the autonomynecessary to live as citizens in a democratic society. It turns citizens into data points,commodifiestheiridentifies,reducesdemocracytoanotheractofconsumption,andleaveslessroomforgenuineautonomy.Attacksonprivacyandsecurityarenot justmattersoftrade or abstract rights; they diminish our psychological and social well-being, a pointoften submerged in debates about the impact of privacy legislation on commerce andpolitics.

Privacy is a perennial issue in communication, especially since the arrival of mediatechnologies in the mid-nineteenth century.With the telegraph and then the telephone,peoplelearnedtotruststrangerswiththeirsecrets.Onewaytobuildtrustwastopromisethatmessageswouldremainprivateandsecure,evenifthatrequiredclosesurveillanceofthosewhoworked the telegraphkeyanddeliveredmessages,aswellas thosewho tookcallrequestsataswitchboard.Inthe1960s,astelevisionwastransitioningintocableandexperimentsin“interactive”videopreviewedafutureofon-demandentertainment,peoplelearnedquickly,totheembarrassmentofsome,thatthesystemsmakingitallpossiblealsokeptarecordofthechoicesmade.Later,theworrygrewwhenvideostoreskepttrackofrentals,firstofcassettesandthenofDVDs.Questionsaroseregardingthepublic’srightto

knowaboutapolitician’sviewinghabits,questionsthatcouldnotfeasiblyberaisedinthe“rabbit-ear” broadcasting days. The Internet upped the ante by globalizing once largelylocalprivacyandsecurityissues.

Cloudcomputing is thenextstep—neitherasimpleextensionnoraradical rupture inthe challenges it poses for privacy and security. By definition the cloud raises seriousconcerns in these areas because it entails moving all data from relatively well-knownsettingswherethehomecomputerharddriveisunderpersonalcontrolorthecomputeratworkstoresdatabehindanemployer’sfirewallatanon-sitedatacenter.Thesecertainlydonotguaranteeprivacyandsecurity,butthemovetotheclouddiminishesthemfurther.Itisonethingforascholartokeepdataonalaptoporportableharddriveor,tosavespaceandmoney,onauniversity server. It isquiteanother to relocatedata to the serversanddata centers of businesses with whom nothing more is shared than an impersonal,customer-company relationship. There are many layers to the privacy and securityproblemwith cloud computing, including growing opportunities to hack and steal data,incentives for companies to make commercial use of cloud data in various forms ofsurveillance capitalism, and opportunities for governments to use cloud data to trackpeoplewithin andbeyond their borders and to apply their own laws to data originatingoutsidetheirboundaries,givingrisetoasurveillancestate.

AheadlineontheWashingtonPostIdeas@Innovationblogwondered,“IsThistheYearEverybody Gets Hacked?” After near-daily accounts of one hacker after anothersuccessfullyattackingthesitesofsomeofthebiggestplayersinthecloud,itwashardtoconsider this hyperbole (Basulto 2013). After all, it was only February 21, 2013, andalreadyFacebook,Twitter,andtheonceinvulnerableApplehadbeenhacked.Fourdayslater,asifinresponsetothequestion,hackersstruckMicrosoft.Itisdifficulttosaywhatpreciselytheattackerswereafter,butexpertsagreedthattheywereprobablylookingforcustomer data or proprietary company information for which black market customersmight pay top dollar to better tailor phishing attacks (M. Schwarz 2013). InApril, theTwitter account of the Associated Press news service was hacked and a tweet postedannouncingaWhiteHousebombingthathadseriouslyinjuredPresidentBarackObama.Intheensuingbriefpanic,stockmarketsdove,andbothTwitterandtheAssociatedPresswerelefttoissuemajorapologiesandpromisesofsolutions.Thishackfollowedcloselyontheheelsofsimilarattackson theTwitteraccountsofBurgerKingandJeep(Romm2013b).

ArguablytheawardforthebiggesthackingstoryofthenewyearwenttoaFebruary19reportthatChina’sPeople’sRevolutionaryArmywasresponsibleforsystematichackingattacks directed against American corporations and government agencies. AttacksincludedthetheftofterabytesofdatafromCoca-Cola,onceinvolvedinafeudwiththegovernment ofChina. Significant as this strike against theworld’s leader in soft drinkswas, security analysts believe that attackers caremore about companies responsible forcriticalinfrastructureprojects,includingelectricalpowergrids,gaslines,andwaterworks(Sanger, Barboza, and Perlroth 2013). A survey of U.S. companies with businesses inChinaconcludedthataboutafourthclaimedtohavebeenhacked(Reuters2013b).Detailsremained murky, and it was reasonable to wonder about the connection between theproliferation of hacking reports and the U.S. government’s drive to pass controversialcyber-security legislation that itself raised privacy questions because it would increase

information-sharingbetween intelligenceagencies andprivate companies (Finkle2013).Furthermore, as two hacking experts note, “It’s good business today to blameChina. Iknow from experience thatmany corporations, government andDODorganizations aremoreeagertobuycyberthreatdatathatclaimstofocusonthePRCthananyothernationstate”(Raimondo2013).

The United States was not just on the receiving end of cyber-attacks. ParticularlynotablewasoneitlaunchedwithIsraeltosendthemaliciousStuxnetmalwaretodisruptIran’s nuclear program. China also claimed that the United States was responsible formassive cyber-attacks on its computers and data centers, especially those containingsensitivemilitarydata.According toa spokesman for thedefenseministry,China’s twomainmilitarywebsitesareunderconstantattack from theUnitedStates:“Lastyear, theChinese Defence Ministry website and Chinamil.com were attacked 144,000 times amonthonaverage.AttacksoriginatingintheU.S.accountedfor62.9percent”(HilleandThomas 2013). Moreover, China’s Huawei, a world leader in the provision oftelecommunicationsequipment,whichitselfhasbeenchargedwithstealingsensitivedatain the United States, Australia, and Canada, maintains that its computers are attackedabout10,000timesaweek(ibid.).ForthePeople’sDaily,“Infact,itisAmericawhichisarealhackers’empireworthyof thisname”(ibid.). Indeed,giventheconnectionrevealedby Edward Snowden between Verizon and the National Security Agency (NSA), evenWesternexpertswonderwhetherthespecialattentiontoHuaweiisjustifiedsincewenowknowthatatleastoneofAmerica’stelecommunicationsgiantshasbeendirectlyinvolvedin massive cyber-surveillance (Pilling 2013). Furthermore, Snowden’s contention thathackingattacksonHongKongandChinahaveemanatedfromtheUnitedStatesforyearsdidnothelptheAmericanclaimthatChinaistheprimarysourceofcyber-mischief(Lam2013).

All of these attacks and counterattacks called the security of the cloud enough intoquestion to leadsomewell-respectedexperts toargueagainstadoptingcloudcomputing(Darrow 2013; Stapleton 2013).According to the PrivacyRights Clearinghouse, in thefirst twomonthsof2013, twenty-eightbreachesattributed tohackersweremadepublic,resulting in the lossof117,000data records (Gonsalves2013). Ifhackerscanstealdatafromsomeofthelargestcomputerandsocial-mediafirms,thelargestsoft-drinkcompanyin theworld, andvital infrastructurecompanies, thenwhoseclouddata is safe? Indeed,among themanyattacks reported in thewinterof2013,one that stoodoutmadeuseofcloudcomputingfacilitiestolaunchaconcertedattackagainstmajorU.S.banks.HerethemajorsuspectwasIran,perhapsinretaliationforStuxnet.However,themostinterestingpart of the tale was not the culprit but the means. Hackers mobilized the combinedresources of several cloud data centers to create what one account called their own“privatecloud,”fromwhichtheylauncheddenial-of-serviceattacksthatdisruptedserviceforcustomersofBankofAmerica,Citigroup,WellsFargo,U.S.Bancorp,PNC,CapitalOne,andHSBC,amongothers(PerlrothandHardy2013).

Thesehackingattacksarejustthosepubliclyreported.Manyothersareknownonlytothoseaffectedbecauseorganizationsdonotwanttocallattentiontotheirvulnerabilitiesortothosetheybelieveareresponsible.Infact,thereisconsiderabledebateinbusinessandgovernmentaboutwhetherattacksshouldberevealedatall.Asoneexpertargued,“Thisisjust the tip of a vast iceberg, and the overwhelming majority of companies today are

terrifiedof talking toopublicly about the issue, for fearof suffering stigmaor sparkingpanic.Thatmeansitistoughforanyoutsidertogetpreciseinformationabouttheoverallscaleofattacks”(Tett2013).Theculpritsalsovaryconsiderablyfromindividuals intentondemonstratingtheirprowess,togenuinethievesouttostealidentities,companysecrets,and money, to others who are looking to disable corporate systems and criticalinfrastructure(NewYorkTimes2013b).Farfromdiminishingsecuritythreats,themovetothecloudincreasesthem.ThathelpstoexplainwhyattacksonU.K.businesseswentfromtwo a day in 2010 to five hundred a day in 2012 (Robinson 2013). As one analystexplained,“Allthevulnerabilitiesandsecurityissuesthaton-premise,non-virtualizedandnon-cloud deployments have still remain in the cloud.All that cloud and virtualizationdoes isenhance thepotential risksby introducingvirtualizationsoftwareandpotentiallymass data breach issues, if an entire cloud provider’s infrastructure is breached”(Gonsalves2013).

Compounding the problem of hacker attacks is that, for all the charges andcountercharges,thereisgenuineuncertaintyaboutwheretheycomefromandwhy.WhenitappearedthatChinawasgoingaftercomputersoperatedbythecompanythatmonitorsmore than half the oil and gas pipelines in the United States, the company set out todeterminewhytheyweredoingit.WeretheyinterestedinbringingdownamajorpieceofAmericaninfrastructureintheeventofamilitaryconfrontation,orweretheyjusttrollingforsecretstopassontoChina’sutilities?Sixmonthsaftertheattack,Americanofficialsclaimed that they still did not know. The same was the case with attacks against fivemultinationalenergycompaniesin2011.TheyappearedtocomefromChina,butnooneknowsforsureandcertainlynotwhy.Moreover,U.S.securityexpertsareuncertainaboutwhich is thebigger threat,Chinaor Iran.The latter, they suspect, continues toworkonretaliationforStuxnetbutlacksthetechnicalsophisticationofChina.Butnooneknowswhethereitherisaprimarythreatgiventhenumberofoperationsemanatingfromallovertheworld,includingfromwithintheUnitedStates(Perlroth,Sanger,andSchmidt2013).Indeed,giventhemountainofrevelationsabouttheNSA,itisreasonabletoconcludethatthemajorthreattotheprivacyofcommunicationandinformationintheUnitedStates,andperhaps the world, is the electronic surveillance operations of the NSA, other U.S.intelligence agencies, the Pentagon, and their partners in theUnitedKingdom,Canada,Australia,andNewZealand(Bamford2013).

More than external attacks violate privacy and security. The very act ofmaintainingthese protections can bring down computers, a demonstration of the often repeatedprinciplethatcomplexsystemsfailbecausetheyarecomplex(Perrow1999).Inordertoblock unauthorized access to their cloud services, some companies deploy an httpsprotocol,whichrequiresregularrenewal.InFebruary2013Microsoftfailedtorenewthecertificate to run its cloud serviceAzure, leading to aworldwide shutdown of itsmaincloudservices.TheembarrassingfailurekeptAzureusers fromaccessingfilesstored inMicrosoft’sdatacenters.Evenafterfourhours,customerswerestillonlyabletoseethestatement “We apologize for any inconvenience this causes our customers” on thecompanywebsite (Ribeiro2013). In thiscase, systemssetup toprotect theprivacyandsecurityofcloudservicesledtoaglobalcrash.TheMicrosoftcasedemonstratesthatevenwhenarmedtotheteethwithsecurityprotection,cloudcompaniesarenotguaranteedtocontinueprovidingservices.Indeed,theveryactofprotection,ofaddingthatextralayer

ofcomplexitythatneedstobemanaged,canleadtoacatastrophe.Thissnafuwasnotanisolated case.Cloud companies regularly lose data, and accidental loss,mainly throughdeletion, was considered the second most significant security problem facing cloudcompaniesinonesurvey(Gonsalves2013).

In another survey of 3,200 companies, 43 percent admitted to losing files stored incloud computers and had to use backups to retrieve them. Still, almost every companyreported at least one failure in the recovery process. Although a leading provider ofsecurity services conducted this survey, the problem is serious enough to alarm evenindependent security experts (ibid.; Investor’s Business Daily 2013). Moreover, thegrowing trend to “bring your own device” to theworkplace has createdmajor securityproblems.Companiesmightspendmillionstokeepouthackersonlytofindthattheirownexecutives are causing major security breaches because they use unprotected smartphones,tablets,andlaptopsintheworkplace(McCarthy2013).Theproliferationofcloud-computing providers is also a source of security concerns because small, inexperiencedcompaniesarealsolesslikelytoprovidestrongprivacyprotections,asusersofthestartupDigitalOcean learnedwhen they foundotherusers’data, includingpasswords, showingupintheiraccounts(McMillan2013).Butwhetherthecloudcompanyislargeorsmall,experiencedornot, it is increasinglydifficult for firms todiscarddata that clientswantdeleted.Explainedoneanalyst,“Companiesarelosingcontrolofwheretheirunstructureddataare.Andiftheydon’tevenknowwhereitis,theywillnotbeabletodeleteit.”Asaresult, data that one believes has been deleted actually lives on to threaten a client’sprivacy(Palmer2013a).

Failuresliketheseleadsecuritycompaniestokeeplayeringsystemsthatpayforaddedprotectionwithgreater complexity. In thewakeof thebig start to the “YearEverybodyGetsHacked,” theCloudSecurityAlliance (CSA), a nonprofit organization comprisingindustry security experts, released, through its Top ThreatsWorking Group, a positionpaper titled“TheNotoriousNine,” a collectionof threats tocloudprivacyand security,eachwithasetofprotocolstominimizethethreatrisk(MarketWatch2013).Followingonthis,theCSApublishedareportonhowtoaddressthethreatsofbig-dataanalysistocloud security and privacy (Goldberg 2013). Organizations like the CSA represent onesmallpieceoftheverylargeandgrowingbusinessofITsecurity,whichwasworthabout$65 billion in 2013 and is growing at the rate of 9 percent annually, faster than the ITbusiness as a whole (Waters 2013a). Despite this enormous investment in protection,experts, including those with no ax to grind with the cloud security business, are notoptimisticthatcurrentformsofsecurityarekeepingupwiththeincreasingsophisticationoftheattacks.Indeed,itappearsthattheenvironmentalimpactofthecloudisjustoneofthe industry’s “dirty secrets.”Another, as one analyst notes, is the failure of traditionalmeasures to successfully address current problems: “The dirty secret that the securityprofessionals can no longer keep to themselves is that their old defenses—whichwereaimed at protecting PCs and other devices that comprise the endpoints of computernetworks—no longer work” (ibid.). The old defenses mainly consisted of antivirussoftware, which continues to work well against the bulk of attacks but is no longereffectiveagainsttoday’smoresophisticatedhacking.

It is particularly interesting that while attacks on cloud data centers are the mostproblematic, security companies believe that the big data processed in the cloud may

provide thebest solution.Bigdatapresents opportunities for pattern recognition,whichcandistinguishbetweennormalandanomalousbehaviorinanetwork.Whatthesecuritypeople call “big intelligence” is actually big surveillance because to succeed requiresmassive monitoring of network activity. When attackers make it through standarddefenses, surveillance spots the patterns theymake in the cloud. Some see this as littlemore than useful rhetoric, a means of giving hope to computer security customers inlanguage they might understand. But as one commentator observed, “Besides theimprovedrhetoric, there’sanotherbenefit to thesenewapproaches:someof themmightevenwork”(ibid.).

Raytheon, the fifth-largest defense contractor in the United States, developed oneespeciallypromisingsystem.Thecompanymines social-media sitesand trackspeople’smovementstopredictbehavior.Withanamechosenfromthelandofbadsciencefictionor good science satire, Riot, or Rapid Information Overlay Technology, provides asnapshot of an individual’s online life, including likes and dislikes, opinions on issues,friends, and places visited. Using one of its employees as an example, RaytheondevelopersputtogetheraprofileandusedittodemonstratehowRiotcouldpredictwherehewouldbe(aspecificgym),onaparticularday(Monday),ataparticulartime(6a.m.)(Gallagher 2013). The Riot software was developed with the support of industry andgovernment experts, and by 2013 it was featured in a patent Raytheon pursued for asystem designed to gather information from social media, including social networks,blogs,andothersources,todeterminewhetherapersonshouldbejudgedasecurityrisk.PublicadvocatesliketheElectronicPrivacyInformationCenterraisedconcernsaboutthearrival of Big Brother into the seemingly innocuous world of social media: “Socialnetworkingsitesareoftennottransparentaboutwhatinformationissharedandhowitisshared.Usersmaybepostinginformationthat theybelievewillbeviewedonlybytheirfriends, but instead, it is being viewed by government officials or pulled in by datacollection services like theRiot search” (ibid.).Actually, the cloudmaybe evendarkerthan this. First, more than just governments are interested in tracking people andpredictingtheirbehavior.Businessesarealsoeagertofollowpeople’smovesinthecloud,especiallyifasystemlikeRiotenablesthemtoforecastwhatproductsorservicestheyarelikely to purchase. Furthermore, an arguably more significant problem with Riot andsystems like it is that theyoftenmakemistakeswithsignificantconsequences.Riotandother such applications appear so flawless that they receive the benefit of the doubt indisputes about accuracy. Others doubt whether such systems can work successfully totrackdowncriminalsandterroristswhooperateinalessthanrationalfashion,suchasthebrotherswhoengineeredtheBostonMarathonbombingsin2013(G.Silverman2013).

Thedarkcloudofattacksonprivacyandsecurityisonlypartofthestory.Thebiggestchallengescomenotfromoutsideattackersbutfromwithintheclouditselfascompaniesincreasingly recognize that an excellent, if not the best, revenue stream flows from thedataprovidedbytheirownusers.Infact,whereas2013mayberememberedastheyearwe all got hacked, perhaps it should also be known as the yearwe all got tracked.AsMaija Palmer maintained, “The new digital economy’s biggest resource is data. FromGoogle’srecordingofinternetsearchhabitstoAmazon’sstoringofcreditcardnumbers,companiesarebusypumpingandextractingdata,alltogreasethewheelsofcommerce”(2013b). No enterprise is more aware of this than Facebook, which bases its business

modelonmakingthemostprofitableuseofinformationaboutitsusersderivedfromtheirposts on its site.Rolling out thismodel has created problems for the companybecauseeachstepintheprocessencroachesonFacebook’sownprivacypolicy,whichit initiallyused to attract and keep users. This began when the social networking site introducedadvertisingonuserpages.Toattractmoreadvertisersandjustifychargingthemmore, itallowedcompaniestodirectadstousersbasedonwhattheyposttotheirpage.SincemypagedescribesmeasaCanadian,IreceiveadsfromCanadiancompanies,butsinceIalsopostlinksfromtheNewYorkTimes,thenewspaperthatdeliversallthenewsthat’sfittoprint also sendsme ads. The next stepwas to vastly expand the information Facebookgathers on users by making deals with large data vendors that collect and manageinformation on users’ offline purchases. This enables the social-media giant to matchofflinepurchasingdatawithinformationthatusersposttoprovideamorecompleteguideto advertisers who want to better target users with ads. Member profiles, advertiserrecords,andofflinedatabasesprovidedbythirdpartiesareanonymouslymatchedthroughuser email addresses andphonenumbers to improve targeting.Each step in theprivacyerosion dance meets with a negative reaction from privacy advocates. In this case theexecutive director of theCenter forDigitalDemocracy immediately alerted theFederalTradeCommission,theGovernmentAccountabilityOffice,andkeylawmakerswhoworkon privacy policy because “clearly the integration of these powerful databases andpurchasingrecords tobeusedfor targeting isaseriousprivacyconcernandneeds tobeinvestigated.We need new privacy controls andmarketing policies to protect sensitiveinformation”(Bachman2013).

Even though it must abide by a twenty-year consent decree with the Federal TradeCommission to give users clear and prominent notice and obtain their consent beforesharing informationbeyond its privacy settings,Facebookpushed its commercializationproject a major step forward with the introduction of Graph Search, its challenge toGoogleand,somewouldadd,itschallengetoprivacy.Theservice,whichbeganarolloutin2013,takeseverypost,includingpictures,likesanddislikes,ageandbirthdate,schoolsattended, work history, sexual orientation, political views, religious preference, andcommentsonmembers’ownandothersites.Itcombinesthisinformationwithpublicdataavailabletousersofaconventionalsearchengine,putstheminadatabase,andmakesuseofasearchalgorithmthatbothFacebookfriendsandthegeneralpubliccanaccess.GraphSearchdeterminesitsresultsbymatchingphrasesandobjectsonasiteratherthanjustkeywords.Bycombiningthe informationonauser’ssiteandtherelationshipof theuser tofriendsandtoobjects,Facebookisabletoreturnresultsthattakeintoaccounthowusersfeel about people and things.The “like” function is especially important in this respectbecauseitenablesGraphSearchtoproduceresultssuchasfriendswholikethefilmsLifeofPiandZeroDarkThirtyandsinglewomeninManhattanwhowereborninFrance.

There ismore to it than theserelatively innocuoussearchpossibilitiessuggest.GraphSearchtakesputtingtogethersocialcombinationstoanewheightordepth,dependingonyour point of view.Which employers are most likely to hire racists (i.e., people whoidentify their employer and “like” racist or racist-friendly organizations) or peoplewholike sadomasochism (in an early search the “prize” went to Home Depot). How aboutpeoplewholikethebannedreligiousgroupFalunGong—thegovernmentofChinamightbeinterestedtoknowwhichofitscitizenshaverelativesintheUnitedStateswholikethe

outlawedorganization.Allthisandmoreisavailableonthepowerfulnewsearchengine,andnoneofitissubjecttofact-checking(Giridharadas2013b).MakingthisallthemoreremarkableisthatmostoftheworkisdonebyFacebookmemberswho,ofcourse,laborwithoutcompensationso thatcompanies,governments,and,yes, friendscandoabetterjobofadvertisingtothem,trackingtheirbehavior,andkeepingintouch.Itislittlewonderthat one organization that tried outGraphSearch in its early days declared “the end ofprivacybyobscurity.”Orwecanviewitastheendofprivacythroughabusinessmodelthatturnseverybitofinformationpostedbymembersintoamarketablecommodityanddeliversthosesamememberstoadvertiserseffectivelyandefficiently.

Withallof theseelements inplace, there remainedonekeyelement.Howwould thecompanydeterminetheeffectivenessofadvertisingonitsownandothersites?Thefirststepwastopartnerwiththedata-miningcompanyDatalogix,whichtrackstheconnectionbetween ads that users see on Facebook and their in-store purchases. This provided animportantindicatorofjusthowsuccessfulthesocial-mediacompanycouldbeinturningadsintoactualsales.Butthiswasnotenough.Facebookwantedtodeterminehowadsonits site stacked up against those located on others, and that would require anotherinvestment.Inthiscase,itwasAtlasSolutions,whichFacebookboughtfromMicrosoft.Through this purchase, the social-media firm expanded its ability to measure theefficiency of ads becauseAtlas compares advertising and purchasing across a range ofcompaniesthatdisplayanad,aswellasacrossdifferentplatforms,includingcomputers,smart phones, and tablets (Dembosky 2013a). Atlas provides Facebook with anassessment of the relative strength of the site and of the range of devices that carryFacebook ads. Nevertheless, questions arose about the accuracy of this research,particularlywhenhackingschemesliketheMarch2013“botnet”attackhijacked120,000personalcomputersandfalselyadded9billionadviewsamonthtoovertwohundredsites(Bradshaw and Steel 2013). With or without mischief like this, audience analysis isbecoming more and more difficult, leading one media industry analyst to decry “themeasurementmess”(Winslow2013).

There isnoguarantee thatanyorallofFacebook’s strategywillwork. In fact, it canbecome painfully counterproductive, as when it led Facebook to place ads for majorbrandsnexttodeeplyoffensivecontent,whichpromptedcompaniestocancelcampaignson the social-media site (Budden 2013). Through the first half of 2013, the company’ssharepriceremainedmiredconsiderablybelowthatofitsinitialpublicoffering,asignalthat Wall Street at least was not optimistic. Nevertheless, some research suggests thatFacebook advertising pays off for most sponsors, a point that contributed to theturnaround in its share price in the last half of 2013 (Manjoo 2013). Whatever theoutcome,Facebookisaprimeexampleofamajorcloudcompanywhosebusinessmodelfundamentallyderivesfromusinginformationprovidedbymembersaboutthemselvesandotherstoselladvertising.Inessence,inreturnforusingthesocial-mediasite,participantsgive up their privacy. They lose it not because of deviant acts by domestic or foreignhackersbutbecauseFacebook, likeGoogle,Twitter, andmostothercompanies thatusethecloud,takeitfromtheminthenormalcourseofdoingbusiness.

It is not just corporations whose normal practice makes privacy in any formincreasingly difficult to secure. Citizens lose privacy through the ordinary practices ofgovernmentswhosesecurityconcernsoftenoutweightheprotectionofprivacyrights.On

thissubject,fingerstypicallypointtowardChina,Iran,andtheArabstatesoftheMiddleEast,whichpracticesurveillancewidelyandlegallyconstrictprivacy.China’ssurveillancepractices are particularly worrisome because of the country’s strong commitment tobecome a world leader in cloud computing, as evidenced by its plan to construct theworld’slargestdatacenters,buildentirecitiesaroundcloudfacilities,andspreadthecloudacrossthecountry,allpartofaprogramtomorethandoubleitsclouddatacentercapacityby2016withagovernment investmentof$370billion.Moreover,Chinahaswelcomedbignames in the cloud tohelp achieve thisgoal.By2013 IBMalreadyhad three largecloudfacilitiesupandrunninginthecountry(J.Lee2013a).

Althoughmostofitsdatacentercapacitywillbeusedfordomesticcivilianandmilitarydata,thereislittledoubtthatChinawillwanttoprofitfromitsinvestmentbyofferingdatastorage,processing, andothercloud services to foreigncompaniesoperatingwithinandeven outside its borders. The country has already demonstrated that its low-costproductionmodelhasbeenoverwhelminglysuccessfulinluringforeigncompaniestouseit as the foundation of global supply chains.With massively increased cloud capacity,China will most likely be in a position to offer a low-cost alternative for Westerncompanies looking to beat the competition for new cloud customers. Given that it iscommonpracticefor theChinesegovernment tomonitor theonlineactivitiesof itsowncitizensandtoengageinspyingandhackingpracticesoncomputersintheUnitedStatesandelsewhere, itshouldcomeasnosurprisethatfilesstoredinclouddatacentersinitsterritorywouldberoutinelyinspected,copied,andused.

DarkcloudsoverChinawilllikelygrowinthecomingyearsasitchallengestheUnitedStates forworld leadership in cloud computing.Fornow,however, theUnitedStates iswellaheadofthepack,anditisimportanttofocusonproblemsthatthispresentsforitsown citizens and for people beyond its borders. Electronic privacy is a problem in theUnited States not just because hackers from abroad are stealing secrets but, moreimportantly, because the country has some of the weakest privacy protections in thedevelopedworld, certainly weaker than those of the EuropeanUnion (EU) or Canada.Thereareseveralreasonsforthis,buttheprimaryonethatAmericanpolicymakerswillpoint to is the need to balance the right to privacywith the nation’s need for security,particularlyinthewakeofthe9/11attacksandtheensuingstrugglesagainstterrorism.Wewillgettothesecuritysideofthisissueshortly,butitisalsoimportanttopointoutthatweakprivacyprotectionshavemadeforastrongITindustry,particularlyassocial-mediafirms have built world-dominant companies like Google, Facebook, Amazon, Apple,Microsoft, andTwitter that profit from selling information about users to advertisers. Itwouldbemuchmoredifficultforthesefirms,especiallyFacebookandGoogle,toprofitifthey were not free to market this information. Most of the $1.1 billion in profit thatFacebook earned in 2012 came to the company because advertisers were interested intargetedmarketingtoitsusers.4Bystrengtheningcompaniesreliantonsellinguserdata,weakprivacyprotectionsbetterenablethemtocompeteinglobalmarkets.Suchhasbeenthecaseforleadingsearchandsocial-mediacompaniesinthecloud,particularlyastheytargetmarketsinthedevelopedworld,suchasinEuropeandCanada.However,intheseplaces companies run into resistance from thosewho prefer a stronger privacy regime,especially one that does not subject them to the USA PATRIOT Act and other cyber-securitylaws.EventhoughAmericansareincreasinglyrestiveaboutdataprivacy(Gross

2013),U.S.businesshas foughtbackfiercelyagainstEUattempts tostrengthenprivacylaws with lobbying led by the American Chamber of Commerce and IT companies,headedbyGoogleandFacebook,thatwouldbenefitthemostfromlooseningdatasecurityintheEU.

Until2013,theEUwasadamantinitsresistancetothelobbyingbarrageandhigh-levelpressure(RTTNews2013).However, inaweakenedeconomicpositionanddesperatetoboosteconomicactivity,theEUbegantobackofffromitsresistancein2013.Infact,onecandatethischangealmostpreciselytoMarch6,2013,oneofthemoreremarkableinthehistoryoftheEU’sdealingswiththeUnitedStates,becauseitbeganwithadecisionthatdemonstrablyaffirmedtheEU’sdeterminationtoenforcethelawandendedwithwhatcanonly be described as capitulation to American power. The first announcement out ofBrusselswasfora$732millionfineagainstMicrosoftbecausethecompanyfailedtoliveup to an agreement to offerWindows customers easy access to alternatives to InternetExplorer,thecompany’sownwebbrowser.Microsoftclaimedthata“technicalerror”keptthecompanyfromofferinguserchoiceonsomeofitsproductsand,inadditiontopayingthefine,agreedtomakeacorrection.Sinceittookoverayearforthecompany’sfailuretocome to light, and then only after its rivals, including Google, brought it to the EU’sattention,thecommissionwaschastisedforlaxenforcement.Nevertheless,thesizeofthefinegavesomeindicationthatBrusselswaspreparedtogettoughwhennecessary(Kanter2013).Thatconclusionwasquestioned,however,especiallybyprivacyadvocates,whenlaterthatdaytheEUreportedthatitwouldloosenitsdatasecurityrequirements,therebyeasingtheabilityofAmericancompaniesthatmakeheavyuseofthecloudtoexpandintoEuropean markets (Fontanella-Kahn and McCarthy 2013). For privacy proponents thiswasamajorstepbackwardbecauseestablishingaunifiedprivacypolicyforthetwenty-seven-nation body that included heavy fines for failing to secure explicit consent fromusers before processing and using their data, as well as incorporating a “right to beforgotten” for online users who want to be erased from the web, would significantlystrengthenprivacyprotectionintheEUandworldwide.Eagerforopportunitiestoexpandeconomic growth, leading EU nations, including the United Kingdom and Germany,appearedtotossinthetowelonstrongprivacyprotectionsinordertoadvanceafreetradeagreementwiththeUnitedStates.ButfurtherturmoilbrokeoutwhenEdwardSnowden’srevelationsaboutmassiveglobalsurveillancebytheNSA,includingacrosstheEU,ledtorenewedcallstostrengthendataprivacyintheEUcloud(Bryant2013).

OnereasonwhytheEuropeanCommissionhassought itsowndataprivacyregimeisthatU.S.legislationcouldviolatetheprivacyofEUcitizens.TheUSAPATRIOTActandtheForeign IntelligenceSurveillanceAmendmentAct (FISA)give theU.S.governmentenormous leeway tocollect informationonpeoplewithout requiringawarrantbasedonprobable cause. The pursuit of global markets by American companies using cloudcomputing, including, for example, Google’s cloud mail service Gmail, draws foreigncitizensintotheirorbit.ConsideraconcreteexamplefromCanada.Gmailearnsrevenuebysellingadvertisingtocompaniesthattargetadstousersbasedinpartonthecontentoftheir emails. That in itself troubles privacy supporters. In order to expand into newmarkets,Googlehasbeenofferingdealstoorganizationsaswellastoindividuals.Scrapyour current internal email system, goes the company’s pitch, eliminate the labor costsincurredbyyour ITdepartment tomanagean internal system, joinus in thecloud, and

slash your IT budget. That pitch has been made to countless organizations, includingToronto’sYorkUniversity,whichwasfullypreparedtoacceptGmailinthecloud,alongwith its advertisements, in return for helpmeeting the fiscal crisis that, likemost otherpublicinstitutions,theuniversityfaces.TheflyintheointmentwasapresentationbytheCanadian Association of University Teachers (CAUT), which explained the privacyconsequences of going with the Google system. Under the university’s existing emailsystem,U.S. lawenforcement authorities, intelligence services, and corporationsdonothavedirectaccesstocommunication,unlessitpassesentirelythroughaserverlocatedintheUnitedStates.Furthermore,thecooperationofCanadianauthoritiesisimportant,ifnotessential,inmakingthejudgmentcallaboutpursuingintercepts.WithGmail,allemail—including messages, attachments, links, and any transaction data—is subject to theprovisionsofthePATRIOTActandFISA.Neitherawarrant,norprobablecause,norevensuspicion of criminality is required to permit intercepts. Government authorities wouldsimplyhavetherighttoscoopupmostofwhatscholarsconsidertheexerciseofacademicfreedom.Moreover, corporations like Google are required to comply with governmentrequests foremailandassociatedcommunicationandareprohibited from informing thetargetthatsucharequesthasbeenmade(Turk2013).ForCanadianuniversitieslikeYork,thismeansweighingmajorcostsavingsagainstthethreatstothesecurityandprivacyofitsstudents,faculty,andstaff.5

Theseconclusions received thecomplete supportofMicrosoft’s chiefprivacyofficer,who, in a submission to the European Parliament, determined that all U.S. cloud-computing companies, including Microsoft, were subject to the surveillance andinvestigatorypowersoutlinedbyCAUT.Specifically,heasserted,“itislawfulintheU.S.to conduct purely political surveillance on foreigners’ data accessible in U.S. clouds”(MacLeod 2013). He notes that FISA in particular provides broad surveillance powersdirectedat“foreign-basedpoliticalorganization(s)…orforeignterritorythatrelatesto…conduct of the foreign affairs of the United States.” The cloud is singled out in 2008amendments that, inaddition topermitting“warrantlesswiretapping,”give thego-aheadtoinvestigatecommunicationcontainedin“remotecomputing”—thatis,thecloud(ibid.).InaninterviewwithaCanadiannewspaper,theMicrosoftofficialconcludedthattheU.S.government “for the first time [has] created a power of mass-surveillance specificallytargetedat thedataofnon-U.S.personslocatedoutsidetheU.S.,whichappliestocloudcomputing”(ibid.).HecalledtheU.S.legislationa“graverisk”toEuropeandatasecurityand told theCanadiannewspaper“everythingI’vesaidabout thesituationofEuropeansapplies also to Canadians” (ibid.). For example, Canadian organizations mobilizingagainstenergyprojectsthatthreatentheenvironmentormarchingagainstmorepipelinesto the United States should expect their communication stored in U.S. company cloudsystemstobesubjecttoinvestigationwithoutanyrecourse,eveniftheydidknow,whichtheymostlikelywouldnot,thatsuchinvestigationsweretakingplace.Theheadofoneofthe largest advocacy groups inCanada concluded, “It does indicate formanywho takeadvocacypositionsthattheyreallyneedtobeverycautiousaboutwhatthey’redoingforthewantofsavingafewdollars,”andcounselsagainstoutsourcingcomputerservicestoU.S. cloud-computing companies (ibid.). Although Canadian federal and provincialgovernmentshaveputinplaceprivacyprotectionmeasures,mostexpertsagree,accordingtoajournalistwiththeleadingnewspaperinCanada’scapital,that“theFISAAmendmentActoverridesanyprivacyanddataprotectionofferedbythird-partyvendors,international

agreementsondatatransfersandCanadiandomesticlegalprotections”(ibid.).

There is little hope of removing FISA in the near future because the U.S. SupremeCourt, in a 5–4 decision supported by both its conservative majority and the Obamaadministration, ruled against plaintiffs who challenged its constitutionality. Inwhat hasbeendescribedasthecatch-22ofFISAandotherlawsintendedtocombatterrorism,theCourtmajorityarguedthatopponentsofthelawcouldnotdemonstratethatitwouldharmthem. However, since this and other laws of its type keep all warrantless surveillancesecret, to the point of preventing service providers from notifying customers, it isimpossible to demonstrate the law’s specific harm (Liptak 2013).As one lawprofessordetermined,“Thecoalitioncouldnotchallengeoursecretsurveillancelawsbecausetheyare secret.There isnoonewhocancomplainofhisorher rightshavingbeenviolated,becauseanyonewhoserightshavebeenviolateddoesn’tknowit.That’sthecatchwhenitcomes to assessing the legality of the government’s secret activities” (Calo 2013). It isthereforelikelythatsignificantconcernsaboutprivacyandsecuritywillcontinuetofaceprovidersandusersofcloudcomputingwellintothefuture.

Working(orNot)intheCloudEach yearFortune magazine produces a list of the top one hundred companies in theUnited States to work for. It covers the range of objective criteria, such as pay andbenefits, as well as such subjective considerations as sense of community andcamaraderie.By the looks of its 2013 list, the sixteenth annual for the company, cloudcomputing does not have a labor problem. Of the top ten, three are leading cloudcompanies,includingnumbersoneandtwo.Googletakesthetopprizeforthefourthtime,asits34,311employeesinitsheadquarterslocationenjoythreewellnesscenters,aseven-acre sports complex, and the benefits of knowing that the company continues to listdozensofvacantpositions.InsecondplaceisSAS,thedataanalyticscompany,withitsownartistsinresidenceandanorganicfarmforitscafeterias.Itisnowonderthatturnoverislessthan5percentannually.ThedatastoragecompanyNetAppholdsdownsixthplace.Likethefirsttwo,itprovidesbothanon-sitefitnesscenteranddomestic-partnerbenefitsforsame-sexpartners.Therestofthelistincludesothermajorcloudcompanies,includingSalesforce (19),Rackspace (34),Cisco (42), andMicrosoft (75), aswell as some firmsthat,whilenotprimarilycloud-computingcompanies,areinvolvedinsomeaspectsofthecloud,suchasAutodesk(54)andIntel(68)(MoskowitzandLevering2013).

WhileitmaysurprisesomethatthelistdoesnotcontainApple,Amazon,Facebook,orTwitter, the cloud is well enough represented that one might question the inclusion ofwork in a discussion of dark clouds.This is certainly understandable becausewhenwethinkofleadingITfirms,includingthoseintheforefrontofthecloud,wetendtothinkofthe top slice of workers, what Giridharadas calls “the tech aristocracy”; for him, “thisemerging aristocracy is, of course, the technocracy—the thousands ofmen andwomenwhoarestriving,throughthegadgetsandservicestheysell,tochangethetextureofbeinghuman: to change fundamental things about all of our relationshipswith time,withourbrains,witheachother”(2013a).Theseprivilegedfewgettoenjoyworkplacesfilledwithluxuries beyond the imagination of most of the world’s workers. Google’s New Yorkoffices contain, in the words of one touring reporter, “a labyrinth of play areas; cafés,coffeebarsandopenkitchens;sunnyoutdoorterraceswithchaises;gourmetcafeteriasthatserve free breakfast, lunch and dinner; Broadway-theme conference rooms with velvetdrapes;andconversationareasdesignedtolooklikevintagesubwaycars”(Stewart2013).Hundredsofsoftwareengineersgettodesigntheirowndesksandworkspaces,includingthe precise ergonomics of furniture andwhether to include company-provided exerciseequipment.Workers are free to come andgo as theyplease, as long as they satisfy therequirementsof theirworkgroup.However,most remainat theoffice foranaverageofnine hours a day because of all the perks. Here is an account of aGoogle employee’sdescription:“Inthecourseofourbriefconversation,shementionedsubsidizedmassages(withmassageroomsonnearlyeveryfloor);freeonce-a-weekeyebrowshaping;freeyogaand Pilates classes; a course she took called ‘Unwind: The Art and Science of StressManagement’;acourseinadvancednegotiationtaughtbyaWhartonprofessor;ahealthconsultation and follow-up with a personal health counselor; an author series and anappearancebythenovelistToniMorrison;andaliveinterviewofJustinBieberbyJimmyFallonintheGoogleoffice.”Thefreefoodaloneisenoughforsometoreturntotheofficeontheirdayoff.Practicallyeveryelementintheworkplaceisresearch-testedandappearstowork, in thewordsofoneGoogleexecutive,“tocreate thehappiest,mostproductive

workplaceintheworld”(ibid.).Googleturnsworkplacesintocommunities,encouragingfreedom and serendipitous interactions that contribute to the innovations thatmake thecompanyaworldleader.MostothercompaniesthatemploythetecharistocracyfallshortofGoogle’s standardbutonly indegree.Bigcloudcompanies likeSASandRackspacereportsimilardegreesofcomfortandfreedom.

Nevertheless,thetecharistocracyisjustathinsliver,theprivilegedfewattheapexofcompaniesthatnotonlyemploythousandsofworkersat theircorporatecentersbutalsomanageglobalsupplychains.Itiscriticaltoresistthetemptationtomistakethesliverforthe whole because doing so means missing serious problems looming for the cloudcomputingindustryattwoverydifferentlevelsoflabor.Thesupplychain,orthechainofaccumulation, responsible for the success or failure of cloud computing extends wellbeyondthecorporateheadquarters.Inordertounderstandtheindustry,especiallyitslaborissues,itisessentialtoscanthebroadersupplychainthatincludes,atoneend,theworkerswho manufacture the material that makes cloud computing possible, where workplaceconditionsarecomparabletothe“darksatanicmills”oftheearlyindustrialage,andattheother, the work of the IT professionals who are most directly affected by thetransformationinlaborthatcloudcomputingisbringingabout.Thefirstgroupofworkerstoils primarily in the industrial centers of China where contractors for big computermanufacturersproducethehardwarethatfillsdatacenters,offices,andhomes.ThissectorhasexperiencedremarkablegrowthasproductionhasshiftedfromtheWesttoChina,butitisnowundergoinganupheavalwithsignificantimplicationsforeverycompanyinthecloud.Asaresult,itisimportanttoconsiderthedialecticalrelationshipbetweenchainsofaccumulationandchainsofresistance.

Thesecondmajordevelopmentisthereorganizationofinformationtechnologylabor.AmainreasonwhycompaniesmovetothecloudistosaveonITlaborbyoutsourcingworkto the cloud.While the head of AmazonWeb Services sees this as a two-decade-longproject,heisconfidentenoughtoconcludethatthecloud“isreplacingthecorporatedatacenter” (Miller and Hardy 2013). The centralization and resulting industrialization ofprofessionalcraftITworkareoneoftheprimarymeansofsavingcostsinthecloud.Butitisimportanttounderstandthat“IT”nowincludesamuchwiderrangeofspecialtiesthanitoncedid.ItnowencompassesnotonlythosewhoworkinITdepartmentsbutalsothosetech-savvy people whose expertise also lies in a substantive profession like education,journalism,orlaw.Inotherwords,thereisanincreasinglysignificantcategoryofworkerswhose work in a professional field requires considerable expertise in the use ofinformationtechnology.Asaresult,thethreatthecloudposestoinformationtechnologyprofessionalsbyvirtueofitscapacitytoabsorbtheITfunctionsofindividualbusinessesextendstoagrowingnumberofworkers.

Theemploymentissuehasbeendebatedthroughoutthehistoryofcomputing.Infact,itaroseasearlyasthe1940swhenthecelebratedcyberneticspioneerNorbertWiener(1948,1950)speculatedthatcomputerswouldleadtomassiveworkplaceautomation.Theissuesheraisedcontinuetoprovide thefoundationforamoregeneraldebateabout theroleoftechnology in structural unemployment (Krugman 2013; Sachs 2013). Once again, theproblemofthequantityandqualityofjobsisnotnewtocomputersandcommunication,but the cloud adds significant elements to the debate. The complexity ofmanaging theglobalsupplychainsthatthecloudrequiresdemandsadegreeoflaborstabilitythatmay

not be possible. Moreover, cloud computing promotes the elimination of skilled jobsthroughcentralizationandautomation.

Itsometimesappearsthattheglobalsupplychainisanythingbutunstable.HasnotmosteverymaterialthingbeenproducedinChinaforaslongasanyone,atleastanyoneunderforty,canremember?Itmayseemthatway,but this isnot thecase,especially in theITindustry, where fundamental changes in the global division of labor are the norm.Beginninginthe1950s,forexample,computerelectronicsproductionbeganintheroomsand garages of amateurs who, like the amateur “Radio Boys” of the 1920s, started anindustry through interpersonal networks of tech friends playing with modified off-the-shelfcomponents.Italsobeganinthelaboratoriesofasmallgroupofuniversitieswherethe building blocks of computer communication were invented and then sent intoproduction with industry partners. IT production moved first to the factories of bigcomputer firms like IBM and DEC whose skilled workforce in the U.S. Northeast,includingupstateNewYorkandtheBostonarea,firmlyestablishedthecomputerindustry.But providing a strong foundation does not guarantee labor stability. During this time,productionbegantoshifttotheU.S.WestCoastasSiliconValleyemergedasacenterofdigitaltechnologyproduction.ThiswaspartlybecausetheexpansionofadivisionoflaborinITproductionmadeitpossibletohirelow-skilledworkersforanimportantpartoftheprocess that couldbe completed in a factoryor even at home.Therewere considerableworkplace hazards associated with this work because it involved dangerous chemicals,which were often cooked up in the apartments and homes of immigrant workers. Oneconsequencewas the rise of a significant toxic-waste issue inSiliconValley,which theEnvironmentalProtectionAgency singled out as the site of themost toxic of themany“Superfund”sitesinthecountry(PellowandPark2002).

While remnantsofhazardousproduction remain inCalifornia, itwasnot longbeforetheindustrywentinsearchofoffshoreproductionsiteswhereauthoritariangovernmentscouldenforcearegimeoflowwages,labordiscipline,andweakenvironmentalprotection.ThefirststopwasSoutheastAsia—Malaysia,Singapore,andthenVietnam—wheretheITproductionprocessbegan.Butthattoowasshort-livedasthetransitiontoastate-directedcapitalisteconomyinChinaoverwhelmedotherproductionsiteswithcheaplaborsubjecttothenear-completecontrolofcompanieslikeTaiwan-basedelectronicsfirmFoxconnorChina’sownHuawei,aworld leader in theprovisionof telecommunicationsequipment.BasedinthenewindustrialheartlandofeasternandsouthernChina—whichreplacedtheoldone,nowarustbeltinnortheastChina,establishedwiththehelpoftheformerSovietUnion—these firms anchored the unprecedented mass production of electronictechnologiesforexporttotheworld.

The success of Foxconn is undeniable. Its 1.4millionworkers labor in over a dozenfactories inChina,and thecompanyalsooperatesmanufacturingplants inBrazil, India,Japan,Malaysia,Mexico,and,especiallysinceitexperiencedbadpublicityforitsChinaoperations, three low-wage European countries (the Czech Republic, Hungary, andSlovakia).It isnoexaggerationtoconcludethatFoxconnplaysavitalroleintheglobaldivisionofinformationlabor,andalthough40percentofitsrevenuescomefromcontractswith Apple, the company manufactures products for nearly every major IT company(Yang2013).Clearly,however,Foxconn’sChinafactoriesemploythemostworkersandattract themostattention,bothgoodandbad.Thelargestof thesefacilities is locatedin

thecityofShenzhenacrossfromHongKonginthesouthofGuongdongprovince,whereover250,000peopletoilinelectronicsplantsthatmakeproductsforalmosteverymajorIT firm in the world, including leading cloud-computing companies such as Amazon,Apple,Google,Microsoft,Cisco,andHP,aswellasJapan’sleadingtechfirms.MostoftheseworkersareimmigrantstotheregionwhocomefromChina’shinterlandinsearchofa living. The Foxconn facility in Shenzhen is part of a walled complex that includesdormitoriesformostoftheworkersandcompanystoresthatprovidethemwithmealsandotheressentials.Onekeytothefirm’ssuccessinmakingthisfacility,aswellasothersthatattract rural workers, highly productive is the workforce’s utter dependence on thecompanyfor their livelihoods, ifnot for their lives.Lowwagesand longhoursbuildingPCs, iPhones, iPads, servers, andmanyof theother ingredients that comprise thecloudhavemadeFoxconnaworldleader.

We encountered Huawei earlier in the book in connection with both the WorldEconomicForumreportoncloudcomputingandconcernsraisedbyWesterngovernmentsover alleged spying by company staff in theUnited States,Canada, andAustralia. Thecompany,nowoneofthelargestproducersoftelecommunicationsequipmentintheworld,isalsobased inShenzhen.Huaweiemploys fewerpeople thanFoxconn,about150,000,almosthalfofwhomworkinresearchanddevelopmentinChinaandatsitesaroundtheworld.ManufacturinghastendedtobeconcentratedintheShenzhenarea.

Muchoftheworldthat,withtheconsiderablehelpoftheChinesegovernment,FoxconnandHuawei have built is beginning to change as this stage in the dynamic electronicsindustry supply chain faces growing turbulence that is likely to impact the cloud-computing industry.Thequiescenceofapoor, rural, immigrantworkforce, cutoff fromtheirhomesinsomecasesbythousandsofmiles,iscomingtoanend.Workingconditionsthat generated big profits for Foxconn and built big cloud-computing companies havetaken their toll onworkerswhoendure longworkdaysof twelvehoursormorewith atbest one day off per week—and not even that during peak demand periods. In 2010Foxconnmadeheadlinesaroundtheworldwhenfourteenworkersreportedlycommittedsuicidebecauseofstressproducedbylonghoursandlowpay.Photosofhowthecompanydealtwiththeissuemadeevenmoreheadlines.Insteadofmoderatingworkingconditions,theFoxconnplant inShenzhen installednetsaround the roofof thebuilding tomake itmore difficult forworkers to take their own lives.While reported suicides did decline,workerprotests spread.Foxconngenerally ignored themorcalledon thepoliceand themilitarytomaintainorder.

InJanuary2012,theNewYorkTimesreportedonsystematicviolationsofbasicworkerrights,includingthehiringofunderageworkersandroutinelyrequiringgreaterthansixty-hourworkweeksover longstretcheswithoutadayoff. ItalsocitedFoxconn’sfailure tocomply with minimum standards of workplace safety that led, in one case alone, toinjuriesto137workersatplantsmanufacturingAppleproductsandtoexplosionsatotherAppleplantsthatkilledseveralworkers(DuhiggandBarboza2012).Appleitselfreportedthatin2012childrenworkedatelevenofitsmanufacturingfacilities(Bradshaw2013).AMay 2012 report by a workers’ rights group that examined company documents andinterviewed170workers concluded, “Exhaustingworkloads, humiliatingdiscipline, andcrampeddormitoriesarestill‘thenorm’forworkersatFoxconnfactoriesinChina”(Musil2012).Workers who refused to follow strict discipline were made to read “confession

letters”aloudandtocleantoilets.Foxconndidnodtoworkerdemandsbysupplyingstoolsso that they would no longer have to stand for entire shifts. However, the companyinsistedthatworkerssitononlyone-thirdofthestoolinorder“toremainnimble”(ibid.).Living conditions remained cramped; typically twenty to thirty people shared a three-bedroomapartmentstackedwithbunkbeds.Useofhigh-energyappliances,suchashairdryers,kettles, and, ironically, laptopcomputers,wasprohibited,andworkerswhousedthemriskedtheirconfiscationuntiltheywerenolongerworkingforthecompany.

Nevertheless,itbecameincreasinglyobviousthatthesupplychainwasfraying.WhenAppletightenedpressuretomeetiPhoneproductionschedules,Foxconnresortedtomoreextrememeasures,suchasdraftinghighschoolstudentstoworkasunpaidinternsforthecompany. For example, the government of a nearby province sent students by thethousandstolaborforamonthorsoattheShenzhenplant.Studentsweregiventhechoiceofobeyingordroppingoutofschool.Somecomplained,butnotwantingtoruinwhateverhopetheyhadforacareer,mostcomplied(Perlin2011,191–196).Thispracticecontinuedwell into 2013,when the company, facedwith strong evidence, admitted to employingstudents and forcing them to work overtime and through the night (Mishkin 2013).Nevertheless, this does not appear to be a long-term solution, as publicity blows backagainst the Western firms that contract with companies engaging in these practices.Indeed,in2013HPandApplerespondedtorevelationsaboutstudentlaborbyannouncinglimitsonstudentandtempworkinChina(BradsherandBarboza2013).Applealsotriedto diversify production by contracting with one of Foxconn’s competitors, but thatcompany too was charged with numerous labor code violations, including employingunderageworkers(Osborne2013).

Foxconn responded to global protests with two substantial changes. First, it movedfactories away from increasingly militant urban centers like Shenzhen and into lesspopulatedregions,especiallytowesternChina,whereitexpectedthatworkerswouldbemoremalleableand informationaboutworkingconditionswouldbe less likely to reachbeyond China’s borders. Second, in a complete reversal of corporate policy, Foxconnsupportedtheformationoftradeunionsatitsfactories.Althoughitwasunclearhowtheunionswouldbeorganized,mostpeoplebelieved that thecompanywouldcontrol them(Jacob2013).Nevertheless,wagesarebeginningtorise,andeveniftheprospectofunionsdoesnotincreaseworkerpower,itislikelytoraisewagerates.Meanwhile,thecompanywashitbystrikeactionsagainstseveralofitsfacilitiesinChina(Tang2013).Allofthesemoves suggest that Foxconn is in trouble. It benefited for a few years from drawingimmigrantlaborintoChina’sboomingcitiesandwallingthemintoindustrialfortressestosupportmanufacturing, especially electronics production, for export.But it did not takelongfortheseformerpeasantstodoubtthewisdomofthesystemandtostartturningthechainofaccumulationthattheirworksustainsintoachainofresistance.Attheveryleast,theysucceededinforcingFoxconntomoveproductiontonewcentersandtoprovideforsomeformofworkerrepresentation.6Nevertheless,giventhecountry’sone-childpolicy,itisbecomingincreasinglydifficultfortheelectronicsmanufacturertoreplaceworkerswhodecide that the global assembly line is not for them or whose rebellion leads to theirdismissal.Thisraisesquestionsaboutthelong-termviabilityofChina’sexport-ledgrowthpolicyandthepoliticalconsequencesofshiftingtoamodelthatconcentratesonChina’sconsumers.

Just as it is easy to expect that the massive control China has maintained over theworld’sindustrialeconomy,especiallyinelectronics,isheretostay,itisalsotemptingtooverstatecontemporarysigns that it iseroding.Given thecoercivepowerof thestate inChina, it is always possible for a crackdown to restore some degree of order, howeverimperfect.Nevertheless,theinstabilityatthebaseofthecomputersupplychainshouldbeasourceofseriousconcernfor thecloud-computing industry.Neither thestable flowofmaterial products essential to the cloud industry nor the low prices made possible byoppressivelylowwagesandhorrendousworkingconditionscanbeguaranteedformuchlonger.

Thereisalsoinstabilityatthetopofthecloud-computingsupplychain.Asoneanalystafter another has concluded, the primary value of cloud computing—what reallycompensates for all of the risks involved in yielding control over data and informationservices toanothercompany—is thesavings inIT labor.Somecompaniescaneliminatetheir IT departments altogether, and others are able to cut them substantially. For ITconsultant Dan Kusnetzky, “Cloud computing is nothing more than the next step inoutsourcingyourIToperations”(McKendrick2013c).Putanotherway,cloudcomputingadvances the industrialization of skilled knowledge labor by centralizing andconcentratingitincloudcompanies.Accordingtothisview,theenterprisecanrunmoreefficiently and leavemost of the ITwork to others.As another IT labormarket expertconcluded,“Automationhasmassiveimplications,especiallyfor thejobsmarket.Itwillnotonlyaffectmanufacturingbutalsoknowledgeworkersintheservicesector”(Solman2013).

It isespecially telling thatoneof themajorboostersofcloudcomputing, theGartnerGroup,isalsoamongthosemostinsistentlypredictingthedemiseofITlaborthroughthecloud. For Gartner, such a development is positive because it means a significantreduction in labor costs for the companies that employ its services. Speaking at aconferenceofITprofessionals,twoGartneranalystsforecastthatby2020demandforITstafftosupportclouddatacenterswillcollapse.Forthem,“thelongrunvaluepropositionof IT is not to support the humanworkforce—it is to replace it” (Dignan 2011b). Theprocesswilltakemanyforms,butthebasicpatternwillstartwithoutsourcingcomputingtothecloud,whichwillbecomeanITutility.Businessprocesseswillthenbeoutsourcedto software, which will affect all economies, but it will hit developing economies thehardest because nations like India now dominate the outsourcing of high-tech jobs.Assoftwaretakesoverthejobsofhigh-techserviceworkers,countrieslikeIndia,whichhaveemployed millions through outsourcing from theWest, will suffer. Furthermore, cloudcompanieswillvirtualizetheirdatacenters,leadingtoadeclineinthenumberofpeoplerequiredtomaintainthatinfrastructure.Workerswhosejobsareconnectedtobuildinganddesigning data centers will also suffer as the need for physical infrastructure declines.Consequently, “many ITworkerswill face hollowed out job prospects just like factoryworkers did as theU.S.manufacturing base disappeared” (ibid.). The outcome appearsinevitable,accordingtotheGartneranalystsandatechexpertwhodescribestheirviews.AsITutilitiesemergeandspread,workerswilldisappearalongwithotherphysicalassets.Gartnercouldnotbeclearer: “CIOsbelieve that theirdatacenters, servers,desktopandbusiness applications are grossly inefficient andmust be rationalized over the next tenyears. We believe that the people associated with these inefficient assets will also be

rationalized in significantnumbers along theway.We foreseea substantial reduction intheU.S.ITworkforce,especiallyamongthosesupportingthedatacenterandapplications,inend-userorganizations”(ibid.).Thisforecastisalreadyplayingoutamongsomeofthemajorusersofcloudservices,suchasEurope’slargestbank,HSBC,whichinMarch2013announced a significant reduction of its IT workforce due to the growing ability tooutsourcetothecloud.Inthefirstroundofcuts,ittrimmedsoftwarestaffingfrom27,000to21,000andplannedfurthercutsacrossallofitsITdepartments(Jenkins2013).Asiftoadd insult to injury, companies are now using cloud computing to develop and trainroboticsystemstoreplacehumanlabor(Harris2013a).Evensomecloudcompaniesaresheddingjobs.Forexample,Cisco,whichbuiltabusinessbasedonprovidinghardware,software,andservicestoon-sitedatacenters,announcedlayoffsof4,000workersin2013becauseithasnotbeenabletoovercomethedecliningbusinessofservingon-siteITwithcloudservices(Wortham2013).

This is a significant development for the IT workforce. Undoubtedly, new jobs willemerge requiring expertise to manage IT utilities, to mediate relationships betweencentralized cloud providers and individual businesses, and to make use of big-dataanalytics. They are also likely to grow in certain specialized areas such as securitybecause, asmore data and business functions aremoved to the cloud, opportunities forhacking and surveillance will also increase. The growth of cyber-security laws andregulations to minimize security problems will also require considerable expertise toaddress the complex problems of complying with new legal and regulatory regimes.Nevertheless,theseadditionaljobsarenotlikelytokeepupwiththemassdownsizingofindividual IT departments in corporations and government agencies. Concerns oversecuritymightalsoslowtheprocessasorganizationschoosetoadopttheprivateoverthepubliccloudinordertobettercontroltheirowndata.Butthisismoreaboutwhetherthetransitionwilltakeplaceoverfiveratherthantenyears,notwhetheritwillhappenatall.Not only do most observers believe that it will, but many see the shrinking of the ITworkforceasonlyonepieceof aneven largerprocessof transformingmostknowledgelaborthroughITandcloudcomputing.

OnewaytounderstandthislargerprocessbeginswithrecognizingthatnotallITworktakesplaceinITdepartments.Suchworkoccupiesanincreasingshareofallknowledgelabor,whichincludesmostjobsinvolvedwiththeproduction,processing,anddistributionof information (Mosco and McKercher 2008). This encompasses work in schools,libraries,andmedia industries likenewspapers,aswellas in theaudiovisualandsocial-media industries. It also includes jobs in health care, law, banking, insurance,transportation, social services, and security. The power of cloud computing and theincreasing reliance on big data, algorithms, and analytics for decision making make itpossibletosubsumeintotechnologymuchofwhattheprofessionsintheinformationandcultural industries laborat today.Asoneobserverput it,“In thenext40yearsanalyticssystemswill replacemuchofwhat theknowledgeworker does today” (Dignan2011a).This conclusion draws from another Gartner presentation, whichmaintained that cloudcomputingandanalyticswill leadtomassivejobeliminationandincreasingpolarizationin theworkforce (ibid.).We are beginning to see this happening today as colleges anduniversities relymore on online education to deliver curricula, including the spread ofmassive open online courses (MOOCs) (Lewin 2013; Chronicle of Higher Education

2013). Moreover, while MOOCs get the attention, we tend to neglect elementary andsecondary schoolswhere it is expected that thecloudwill takeup35percentof annualbudgets by 2017 (Nagel 2013). Librarians are giving way to automated systems thatdelivere-documentsfromthecloud(Goldner2010).7Thejournalismprofessionisinrapiddecline as print advertising has evaporated and freelance and unpaid or low-paid internworkers replace full-time reporters. Moreover, centralized editing from the cloud isreplacingeditorialstaffassociatedwithspecificpublications(PewResearchCenter2013).Thereisaninevitabledeclineinthequalityofworkfortheseandotherprofessionswhoselaborcanbecentralizedandconcentratedinthecloud.Butitappearsthatinstitutionsarewillingtoacceptsomeerosioninqualityformassivesavingsinlaborcosts.

Cloud computing essentially deepens and extends opportunities to eliminate jobs andrestructuretheworkforce.Whereastechnologyonceonlydisplacedworkersinindustrialsettings,itbegantobedeployedtoeliminateknowledgeworkersinthe1970s,atatimewhen accelerating energy costs and the emergence of industrial centers in non-Westernsocieties challenged companies to cut costs and restructure by drawing on a globalworkforce.Combinedwith thegrowing analytical capabilities of computer systems thatgive new life to the “scientific management” of the workplace, the cloud is creatingopportunitiestoeliminateseverallevelsofdecisionmakersinorganizations(Lohr2013b).Alreadythere iswidespreadfear inITandhumanresourcesdepartments that jobloss isinevitableand,wherejobsaresaved,controlwillbelostbecausecompanieswillrelyonautomated decision-making systems based on big-data analytics (Linthicum 2013b)—hencetheconclusionsoftheGartnerexpertsabouttheerosioninjobs,includingatmostlevels of management, and the polarization in the workforce between those in low-skilled/low-pay service work and those in the upper reaches of organizations. In moredescriptive language, Harvard economist and former Treasury secretary LawrenceSummers warned, “As economists like to explain, the system will equilibrate at fullemployment. But maybe the way it will equilibrate at full employment is there’ll bespecialistsatcleaningtheshallowendandthedeependofrichpeople’sswimmingpools.And that’s a problematic way for society to function” (Freeland 2013). Citing theunprecedented break between productivity and wages that has afflicted Westerneconomies in recent years, MIT economist Erik Brynjolfsson concluded, “Most of thedebateinWashingtonisreallyplayingsmallballandismissingthetectonicchangesinthewaytheeconomyworks,whicharedrivenbytechnology.Thisisthebigstoryofourtime,anditisgoingtoaccelerateoverthenext10years”(ibid.).

What might change or slow this trend? I have already alluded to two possibilities.Supplychaindisruptionsmaymakeitmoredifficulttodeploycloudsystemsaroundtheworld, andorganized resistance fromworkersmay alter the potential to profit from thecloud.ThelaborforceinChina, thebaseofglobalelectronicssupplychains,hasgrownrestive in recent years, prompting higher wages and a redeployment of electronicsmanufacturing sites. It is unlikely thesemeasureswill do anythingmore than delay theinevitablechoicebetweensubstantiallyraisingthelivingstandards,includingthewages,workingconditions,andpolitical freedom,ofChina’sworkforceorfaceescalatingmasscivil unrest. One can deploy suicide prevention curtains for just so long. Theacknowledgment of unrest inChina’s once placid factories has reached themainstreamWestern press, where a recent account in Time magazine offered this startling set of

observations:

“Resentment is reachingaboilingpoint inChina’sfactory towns.Peoplefromtheoutside see our lives as very bountiful, but the real life in the factory is verydifferent,” says factory worker Peng Ming in the southern industrial enclave ofShenzhen.Facinglonghours,risingcosts, indifferentmanagersandoftenlatepay,workers are beginning to sound like an angry proletariat. “The way the rich getmoney is through exploiting theworkers,” saysGuanGuohau, another Shenzhenfactory employee. “Communism is what we are looking forward to.” Unless thegovernment takes greater action to improve their welfare, they say, laborers willbecomemore andmorewilling to take action themselves. “Workerswill organizemore,”Pengpredicts.“Alltheworkersshouldbeunited.”(Schuman2013)

Itisnotonlythebaseoftheglobalsupplychainscreatedbymajorcloudcompaniesthatcancreatedisruptions.ChainsofresistancecanalsoformintheadvancednationsoftheWestwherethelaborprocessiscertainlybetterthaninChineseelectronicassemblyplantsbut very far fromwhat applies in the headquarters of these companies. Resistance canarisefromhowmanagementusesthecloudtomonitorandcontroltheminutestactionsofitsworkforce,includingthoseinwhite-collaroccupations.Accordingtooneanalyst,“AsBigData becomes a fixture of office life, companies are turning to tracking devices togatherreal-timeinformationonhowteamsofemployeesworkandinteract.Sensors,wornon lanyards or placed on office furniture, record how often staffers get up from theirdesks,consultother teamsandholdmeetings”(R.Silverman2013).Today’s technologyenablesemployerstocontrolworkersinwaysthatFredericWinslowTaylor,thefatherof“scientificmanagement,”couldonlydreamabout.Whereasonceanemployercouldonlysystematicallymonitorworkerswhentheypunchedtheirtimecardsatthebeginningandend of the workday, today they can measure practically every activity of workersemployedincallcentersandlogisticsoperations.Asaspecialistinworkplacesurveillancecomments,“Ifyouhaveaplentifulsupplyoflaboranddon’tneedtoworryaboutquality,the temptation is tonailyourworkers foreveryminuteof theday” (Gapper2013a; seealsoNeff2012).

While sensors raise significant privacy issues, a more ominous portent comes fromAmazon,whichisfundamentallychallengingtherightsthatworkersintheWestsecuredoveryearsofstruggleandorganizing.OnehotspotforlabortensionsisGermany,wherethecompanyhasestablishedeightdistributioncentersemploying8,000workers.Germanyis importantfor thecompanyas thesourceof14percentof itsrevenues(WingfieldandEddy2013).Thecountryhasnotreceivedagreatdealofattentioninstrugglesoverglobalsupply chains, but it has a long history of battles with Walmart, which abandonedGermanyin2006ratherthanbenditsworldwidelaborstandardstomeettheexpectationsof German workers and especially their union Ver.di, which represents over 2 millionemployees in the service sector. German workers and their unions have considerablygreater power than their counterparts in the United States and the United Kingdom.Mobilizing workers across the nation, Ver.di’s actions succeeded in ending Walmart’spresenceinthecountry.ThebattlehasnoweruptedoverAmazon,which,intheviewofGermanworkers, is attempting to impose “American-stylemanagement” by relying onruthless labor practices such as hiring thousands of low-wage and mainly foreign

temporaryworkersandthesecuritypolicenecessarytomaintaincontrol.Thishasenabledthe company to cut prices and drive out competition, including one German firm.Accordingtoaunionleader,Amazonappliesrigidcontrolstoitsworkforce:“Everythingismeasured,everythingiscalculated,everythingisgearedtowardefficiency.Peoplewanttobetreatedwithrespect”(Ewing2013).Thecompanydeniestheseclaims,arguingthatithiresforeigntempsbecausetherearenotenoughlocalworkers.Buttheonlinegiantfacedembarrassmentwhenithadtofireasecurityfirmhiredtopoliceoneofitsplantsbecausesome of the firm’s employees, decked out in outfits associated with neo-Nazi groups,rougheduppeopletryingtofilmactivityoutsidetheplant.Thecompanymaintainsthatitcouldnotpossiblyvetthebackgroundsofallthoseithiresandinsiststhat,whileitrefusestonegotiatewiththeunion,itdoespayworkerswell.

What will happen in this key node of Amazon’s global supply chain is uncertain.Workers mount regular protests using mass mobilization, guerilla theater, and onlineglobal petition drives (37,000 signatures received by March 2013). But Amazon hasrefused tobackdown. InMay2013,workersat thegiantAmazondistributioncenter inLeipzigwalkedoffthejob,markingthefirstreportedstrikeatanAmazonfacility(Wilsonand Jopson2013).As the story continues to unfold, there are important implications toconsiderforlaborinthecloudandforthecloudoverlabor.

While a great deal of its labor process can be automated and lodged in the cloud,Amazonstill requiresa largeworkforce in thedevelopedworld toefficiently locateanddistributeitsproducts.Soinspiteofthecompany’shigh-techimage,Amazonworkersatatypical warehouse walk between seven and fifteen miles every day carrying handhelddevicesthatdirectandmonitortheireverymovetolocateorderedgoodsinitswarehousesandpackagethemindividually.Tomaximizeproductivity,thecompanyregularlyadvisesworkersonmoreefficientwaystocarryouttheiractivities,makingfulluseofthedataineverythingfromindividualpersonneldecisionstogloballogisticsplanning.Onebusinesspublication,notknownforharshattacksonworkplacepractices,citespeopleinoneU.K.Amazonplantwhocomplainaboutthecompany’spractices.Alocalofficial,whofoughttobringthecompanytohistownwithahighunemploymentrate,concludes,“They’renotseen as a good employer. It’s not helpful to our economy; it’s not helpful to theindividuals” (S.O’Connor2013).Anotheruses stronger language:“The feedbackwe’regettingisit’slikebeinginaslavecamp.”EvenanAmazonmanageradmits,“You’resortof like a robot, but in human form.” In thewords of one technology reporter, “Digitalcapitalismproducesfewwinners.Apple,Amazon,FacebookandGooglemightposthugeprofits,butmanyoftheirstaffseelittlefinancialbenefit”(Naughton2013).

Amazon labor is restivenotonly in thematerialworkplace.Thecompanyoperates aglobal system of piecework in the cloud that critics have called a “digital sweatshop”(Cushing 2013). The Amazon Mechanical Turk (AMT) employs a large body of“crowdsourced” workers, whom Amazon calls “providers” (also known as “Turkers”).They carry out minute tasks online for “requesters,” who pay Turkers piece rates forwritingproductdescriptions,identifyingindividualsinimages,orproducingspam(a2010studybyNewYorkUniversityresearchersdeterminedthatspamconstitutesasmuchas40percent of the jobs) (Ipeirotis 2013).Amazon originally set the system up to carry outwork that could be done online but required somehuman involvement.The typical jobwassortingmerchandiseintocategoriesbasedoncolororstyleforthecompany’smassive

onlinewarehouse.Itwassosuccessful thatAmazondecidedtobecomeajobbrokerforcorporations needing people to do things like look up foreign zip codes or transcribepodcasts.

Formanagingtheservice,Amazonreceives10percentofthevalueofacompletedjob,or human-intelligence task (HIT). Although Turkers include professionals, the vastmajority are semiskilled workers who provide their credentials to requesters and, oncecleared, choose amongposted tasks.Workers in theUnitedStates are paid in cash, butmany foreign workers are primarily given the option to accept gift certificates. Exactfiguresarehard topindown,but it isestimated that the industryemploysover200,000workersand,by2011,wasearningabout$375millionannually(Cushing2013).Thereisalsogrowingevidence thatworkersare less thanhappywith thesystem. Itdidnot takelongforthemtorealizethat,asonecomplained,“theymakeitsoundlikeyoucanjustdoafewtasksinyourfreetimeinbetweenotherthings,butifyouworkedlikethat,Ibelieveyou would make about a dollar a day” (ibid.). Because companies have an enormousworkforcetodrawfrom,theycanpaythelowestpossiblerates—$1or$2anhourisnotunusual—anddemandswiftandaccuratecompletionofjobs.Workerswhomessupajobaredroppedorbannedfromreapplying.InJanuary2013Amazonstoppedacceptingnewapplications from international Turkers because of what the company deemedunacceptablelevelsoffraudandpoorworkerperformance(“TheReasonsWhyAmazonMechanical Turk No Longer Accepts International Turkers” 2013). Since internationalworkers are more likely to accept the low pay and constant demands, requesters havebeguntosetuptheirownTurkoperations.

Upset about the system, Turkers use their onlineworld to vet requesters and contactotherTurkers.TheresultisTurkopticon,apieceofsoftwarethataddsfunctionalitytositesthatpostHITsbyaddingratings,reviewsofemployers,andadvicetoexploitedTurkers.8AccordingtoonescientistwhohasworkedonAMT28,000times,“There’snosickleave,paidholidays,anythinglikethatonTurk.Thereisnoarbitration,noappealifyoufeelthatyou have been unfairly treated, apart from a stinging review on Turkopticon” (Hodson2013). Furthermore, worker complaints, fraud, and a host of negative consequencesresulting from AMT’s sweatshop in the cloud have encouraged other firms to set upsomewhat more hospitable operations. For example, the firm MobileWorks pays theminimumwageineffectinthecountrywheretheworkisbeingdone,assignseachworkeramanagertodealwithproblems,andprovidesopportunitiesforworkermobility(ibid.).Itisuncertainwhethertheemergenceofmoreworker-friendlycompanieswillrestoresomecredibilitytoonlinepiecework.MuchwilldependonwhetherbigcompanieslikeAmazonreformthelaborprocessinthecloud.Itappearstobeintheirinteresttodosobecauseithas become clear that the race to the bottom forwages andworking conditions createsproblemsforthecompanyaswellasforworkers.

Workerorganizations,especiallytradeunions,arenotoftendiscussedalongsidecloudcomputing. Only a handful of cloud providers, mainly the older computer andtelecommunications firms such as IBMandVerizon, have to dealwith organized labor.But as we have seen in the case of Apple’s experience with Foxconn in China andAmazon in Germany, cloud companies, as they become inextricably bound to globalsupplychains,facetheresistanceoforganizedlabor.Theseareexamplesofaprocessatwork in the broadly defined knowledge and cultural industries that brings together

workers across once discrete sectors. As a result, unions that once represented onlytelecommunicationsworkersnowincludecreativeandtechnicaltalentintheaudiovisual,writing,service,andtechnologysectors.TheCommunicationWorkersofAmericaanditscounterpart in Canada, which in 2013merged its communications and power workers’union with the union representing auto workers, are good examples of workerorganizationsthathavefollowedthepathoftechnologicalconvergenceintheirorganizingefforts. The 2012 merger of the Screen Actors Guild and the American Federation ofTelevisionandRadioArtistsbringstogetherthemajorHollywoodunionsforthefirsttimetofaceoffmoreeffectivelyagainsttheincreasinglyintegratedHollywoodmediaindustry.Moreover,individualunionsarenotonlyexpandingacrosstheconvergingcommunicationand information industries, but they are forming large transnational organizations likeVer.diandUNIGlobalUnion.Thesetransnationalunionsarebetterequippedtodealwithpowerfulmultinationalcompaniesbecausetheyhaveenormousmembershipsandarewellfunded.Furthermore,thescopeoftheirmembershipenablesthemtobetterrepresenttheconvergences in both the labor process and theworking conditions among information,cultural,andserviceworkersandtobuildbridgesacrossthedivideseparatingworkersatdifferentspatialandoccupationalpointsintheglobaldivisionoflabor.

Ver.diwasfoundedin2001andby2013hadreached2.3millionmembers,primarilyinGermanybutinotherpartsoftheworldaswell.Itrepresentsworkersinthirteensectors,all of which are increasingly affected by the rollout of cloud computing, includingfinancialservices,healthandsocialservices,education,scienceandresearch,mediaandculture, telecommunications, information technology and data-processing, postal,transport, and commerce services. Its members work in government and business atalmosteverylevelofoccupationalskillandfunction.Theunioncannotonlymobilizealargeanddiverseworkforcebutalsodrawonthespecializedtalentsofitsmembers,whocanhelptheuniontotightenandsecureitsinternalcommunicationsorcarryoutguerillatheaterproteststhatattractwidespreadmediaattention.UNIGlobalUnionwascreatedin2000when three internationalworker federations in the information,media,andservicesectorscametogethertoformagenuinelyglobalfederationofknowledgeworkers.Today,it gives voice to 20 million workers in 150 countries through nine hundred affiliatedunionsinabroadrangeoffields, includinginformationtechnologyandservices,media,entertainmentandthearts,gamingandsport,finance,commerce,andsecurity,aswellastothegrowingnumbersofworkerswhotoilfortemporaryemploymentagencies.Amongits major activities is negotiating global agreements with transnational companies toaddressimportantissuessuchaschildlabor,discrimination,andtherighttoorganizelocalunions.Byearly2013,ithadcompletedforty-eightsuchagreementswithawiderangeofcompanies,includinganumberinthecommunicationandinformationtechnologysector.Itwasalsointheprocessofnegotiatingfreshagreementswithmajortransnationalfirms,includingIBMandDisney.

Ver.di and UNI are not alone among converging unions and international laborfederations thatarehavinganimpactonglobalsupplychains, includingthosecentral tothe growth of cloud computing. But it is uncertain whether this development is theharbingerofasignificantupsurgeingloballaboractivismoradefensiveposturethatcanatbestslowdowntheinevitabledeclineanddemiseoforganizedlabor.Thatdepends,inpart,onhowonedefinesorganizedlabor,becauseanotherimportanttrendisthegrowthof

labororganizationsthatarenotformaltradeunions.Theseworkerassociationsresembleunions but, either out of choice or necessity, remain outside the legal and politicalstructuresthatgoverntheoperationoftradeunions.Theyoperateallovertheworld,andresearchhasdocumentedtheirimportanceinChina,India,Europe,andtheUnitedStates(Mosco, McKercher, and Huws 2010). They are especially active in the information,communication, and cultural sectors where worker associations have representedemployees in occupations ranging from call-center agent to software-engineeringspecialist. Worker associations have won major victories for contract employees atMicrosoft and for telecommunicationsworkers in India.Although theydonot typicallynegotiate contracts, worker associations have provided employees with legalrepresentation, groupmedical insurance, training, model contract language, counseling,and support for collective resistance, without suffering from some of the bureaucraticentanglements that plague traditional trade unions. These associations are particularlyactive among contract and temporary workers; for example, the Freelancers Union hassigned up 200,000 members in a wide range of jobs, including law, app and softwaredevelopment, graphic arts, accounting, writing, editing, and consulting. Workerassociations differ from trade unions not just in what they lack—a system of formalbargainingwithemployers—butintheiremphasisonmutualassistanceoutside,aswellaswithin,theworkplace.Theyfollowthesocialmovementtraditionofearliertradeunions,which provided workers with social support, including family assistance, housing,insurance, and a source of collective power and community. As the head of theFreelancersUnionnoted,“Thesocialunionismofthe1920shaditright.Theysaid:‘Weserveworkers360degrees.It’snotjustabouttheirwork.It’sabouttheirwholelife.’Weviewthingsthesameway”(Greenhouse2013).

The dark clouds identified in this chapter, involving the environment, privacy, andlabor, present major challenges to the future of cloud computing. The next chapteraddressesacloudyforecastofanothersortthattakesusintotheworldofbigdataandthecultureofclouds.

CHAPTER5BIGDATAANDCLOUDCULTURE

This is a world where massive amounts of data and applied mathematicsreplaceeveryothertoolthatmightbebroughttobear.Outwitheverytheoryof human behavior, from linguistics to sociology. Forget taxonomy,ontology, and psychology.Who knowswhy people dowhat they do?Thepoint is they do it, and we can track and measure it with unprecedentedfidelity.With enough data, the numbers speak for themselves. (Anderson2008)

Alongtimeago(and,ofcourse,inmanypartsofsocietytoday),peoplehadanother name formassive information dumps that occurred spontaneouslywithoutanyqueryhavingbeenmade.TheycalleditGod.ItwasGod,orthegods,whospokeoutof theburningbush to tellyouwhatyoudidn’t evenknow you needed to ask. Before Oracle, Inc., in other words, there wereoracles.(AlanLiuinFranklin2012,445)

The growth of cloud computing continues a process of building a global informationalcapitalism by concentrating production, processing, storage, distribution, and electronicservices inahandfulofcompanies,and, insomecases,governments, thatmanagelaborand consumption through the systems that the cloud enables. This is undoubtedly acontested process as dark clouds gather around the environment, privacy, security, andlabor.So it is uncertain that the cloud’s combinationofmassive computer powerundernear-uniformlyprivatecontrolwillbeabletoproduceandsustainacontinuouslygrowingcapitalistworldorder.Giventheproblems,itisdoubtfulthatwewillachieveBillGates’svision of “friction-free capitalism” (Gates 1995). Nevertheless, the powerful forcespromotingtheglobalcloudmakeitreasonabletoexpectconsiderableexpansion,howevercontested,inthenearfuture.Thisinnovationalonemakescloudcomputinganimportantdevelopment to follow. But the cloud means considerably more because it is alsopromotingaveryspecificcultureofknowingthatvalorizescertaintypesofknowledgeandways of knowing that have significant implications across social life. In this respect,friction-free capitalism meets what a Wired magazine editor calls the “globalsuperintelligenceknownasthecloud”(Wolf2010).Thischapterexaminesthiscultureofknowing and critically assesses it by taking up episodes in the long history of cloudculture,where a 2,000-year-old play, amedievalmanuscript, and a contemporarynovelspeak to theknowledgeculture that isunderconstruction.Thepolitical economyof thecloud (how it advances informational capitalism) and the culture of the cloud (what itmeansforknowledgeandfortherepresentationofourworld)cohereandclash.Exploringboththeharmonyandtheconflictcreatesspaceforacriticalunderstandingofthecloud.

Cloudcomputingacceleratesapowerfulandinfluentialwayofknowingthat iscalledontoaddresssignificant issuesfacingglobalcapitalism.In itsnear-magicalbrillianceatcertaintasks,thecloudhasseducedmanyofitsproponentstoseeitastheprimary,ifnottheonly,meansofsolvingproblems,pushingtothesidelineswaysofknowingandseeingtheworldthathaveguidedhumanityoverthecenturies.Initsextremeform,andtherearemanyexamplesfromwhichtodraw,thewayofknowingadvancedbythecloudwillreachasingularity,understoodastheoneandonlylegitimatemeanstoknow.Alltherestistobemarginalized,sequestered in thenetherworld, reservedfor the likesofastrologyandconjuring.Thisisamistakefortwosubstantialreasons.First,lifeissomassivelycomplexthat no form of knowledge, however dazzling, can claim to be the universal way ofknowing.Second,becausecloudcomputinghasdevelopedunderalmostcompleteprivatecontrol, its particularwayof knowing is constrainedby the narrowgoal of commercialexpansion. To put it bluntly, the cloud addresses most of the world as consumers andsubjects,notasactivecitizens,andthistendencyhassignificantconsequences.Itismoreimportant thanever to resist singularities,expandwhat itmeans toknow,andmake thecloud more than merely the instrument to build and manage markets for products,services, workers, and consumers. To address these points, we need to understand theparticularityofthecloud’swayofknowing.Whatareitsstrengthsandlimitations?Whatarethealternativesandhowaretheseconstrainedbythecultureofthecloud?

ACloudofBigDataItisusefultostartbyexaminingtherelationshipbetweenthecloudandwhatiscalled“bigdata.” The latter refers to the movement to analyze the increasingly vast amounts ofinformationstoredinmultiplelocations,butmainlyonlineandprimarilyinthecloud.Wecannot reduceone to theotherbecause thecloudencompassesmore thanbigdata.Theanalysis of big data, sometimes referred to as analytics, is one (admittedly important)service provided by cloud companies. Furthermore, big-data analysis can take placeoutsideofacloudsetting,ascompaniesandgovernmentagenciesoftenmakeuseofdataheld on their own computers. However, since the store of material used in big-dataanalysisisgrowinginsizeandcomplexity,itisincreasinglyafeatureofcloudcomputing,benefitting from the promotional pitch that cloud companies make to customers. Forexample, Amazon Web Services (AWS) has grown since its success in supplying theObamacampaignwithbig-dataanalysisthatmostexpertsagreeprovidedsignificanthelpinthesuccessful2012campaign,andthissuccessgaveAWSamajorboostinabattlewithIBM to win a $600 million contract with the CIA. It also helped AWS expand itsconsumerservicetochallengethatofDropboxandGoogle(Barr2013).Theexpansionofcloudcomputingaloneadvancestheinterestinbigdatabecause,asoneanalystsaid,thecloud“hasmade it viable toperformsophisticatedanalyticsoverhugevolumesofdatathat were never even thinkable before” (Wainewright 2013). The cloud is not alone ingiving impetus to big data. The proliferation of smart devices has brought about themassive growth in cloud-based information, including the locational data stored onphones, the devices installed in homes and workplaces that monitor everything frompowerconsumptiontotheactivitiesoffamiliesandworkers,andtheconstantstreamsofsocial-mediatweets,postings,andmessages.Infact,onecansafelyconcludethatbigdataresultsfromtheintimateconnectionthatcompaniesandgovernmentsrecognizebetweencloudcomputingandsmartdevices.

Cloudprovidershavealsoledthewayinpromotingbig-dataanalysis,viewingitasameans of expanding revenue. Some companies simply enable big data by introducinganalytics programs to the applications they provide their cloud-computing customers.Other companies go further by directly analyzing the data they store on workers andcustomerstofindaddedvalue.Onefirmproducedanationaldatabaseonemployeeswhohavebeencaughtstealing,informationthatretailersusetopreventfuturehiring(Cliffordand Silver-Greenberg 2013). Another firm used consumer data to develop a predictivealgorithm to let clients knowwhat files its users are most likely to download to localstorage. Still others are “productivizing” data by harnessing publicly available archivessuch as Twitter postings to build new products (Wainewright 2013). This has been thecenterpiececonceptbehindIBM’sSmarterAnalyticsproject,acombinationofsoftware,systems,andstrategiesthatenableclientstocombinetheirownbusinessorenterprisedatawith their consumers’ unstructured data to better identify and anticipate consumerbehavior. IBM refers to the latter as “the data of desire” because it registers popularexpressionsofsentimentandfeeling,suchas likes/dislikes,aboutproductsandservices.This gives its cloud customers the ability to correlate sales records with social-mediapostings, thereby linking behavioral data with information about customer feelings toprovideadeeperviewofcustomersentiment—notjustwhichcustomersarebuying,butwhy. IBM credits this system with enabling a communication carrier to predict which

customerswerelikelytodefectwithinninetydaysandreducedchurnby35percentinthefirst year (IBM 2013). The potential in big data gives traditional companies like IBMopportunities for reinvention. A leader in research on embedding intelligence andcommunicationcapabilitiesinobjects,orwhatiscalledthe“Internetofthings,”GeneralElectric has also bet heavily on transforming itself into a company that specializes infinding big-data solutions in the cloud (Butler 2013b). So has Monsanto, one of theworld’s leading chemical and agribusiness companies and the dominant producer ofgeneticallymodifiedseed.In2013thecompanyspent$930milliontopurchaseaSiliconValley start-up that uses big data to carry outweather and climate analysis (McDuling2013).

Thesedevelopmentsdemonstratethedynamicrelationshipbetweenbigdataandcloudcomputing. Cloud companies that might have been satisfied to limit their business toprovidingdatastorageandapplicationsnowhaveastrongincentivetomakeuseofdatatoselladditionalservices tocustomersand todevelopnewproductsof theirown.But thisdoes not just offer economic advantages. It also raises questions about the rights andresponsibilitiesofcloudcompanies.Somecompaniesandindividualsmightwonderwhydatatheyexpectedwasonlygoingtobestoredinthecloudisinsteadbeingusedbycloudcompanies to seek financial gain.Suchactivitymightbenefit a customerwho stands toshare in the added value, but it will also expose customer data to uses that were notanticipated.Moreover,asthecloudcontinuesitsinexorableglobalexpansion,thestoragefacility is increasingly likely to be located in the jurisdiction of another countrywhosegovernmentwill apply its own rules, regulations, and policies. In 2013Microsoft tooksignificant steps toward such a relationship with China, a development that promptedwarningsaboutdire consequences fromexpertsonSino-American relations (Raglandetal. 2013).Theeconomic synergies touted for the cloudandbigdata caneasilyproducesignificantpoliticalcomplications.

It is therefore now essential to consider big data in a comprehensive assessment ofcloudcomputingandespeciallytoassessitswayofknowing.Thecloudreceivedaboostwhen theNational InstituteofStandards andTechnologyprovided agenerally accepteddefinition,butthesamehasnotbeenthecaseforbigdata.Amongthemanycirculating,theWikipediaentry is a reasonablygoodone: “In information technology,bigdata is acollectionofdatasetssolargeandcomplexthatitbecomesdifficulttoprocessusingon-handdatabasemanagementtoolsortraditionaldataprocessingapplications.”1Theauthorsofa2013bookonthesubjectrefertoitas“theabilityofsocietytoharnessinformationinnovelwaystoproduceusefulinsightsorgoodsandservicesofsignificantvalue”(Mayer-SchönbergerandCukier2013,2).

Like thecloud,bigdatahasoftengenerateda rapturousresponsefromitssupporters,withoneof themost sober accountsnoting that “it hasbecomederigeur to ascribe allsorts of supernatural powers toBigData” (Asay 2013).AMicrosoft researcherworriesabout the uncritical acceptance of big-data analysis out of a widespread “big datafundamentalism”(Hardy2013i).Onesourceofthefundamentalismisthebeliefthatoncethe easy work of gathering data is completed, the data will speak for itself, yieldingprofitablegoldnuggetsofbusinessinformation.Butthisisfarfromthecase.Analysisisthehardpartanditisgrowingmorechallengingastheamountofcollectibledataexpands.Itisnowonderthatsomeexpertsworrythatbusinessesaregivinguponbigdata,leading

onetoconcludethata“dirtylittlesecret”oftheindustryisthat“nobodywantstousethedata”(Elowitz2013).Beforeexaminingwhatmightappropriatelybecalled thebig-datasublime,itisbesttobrieflyexaminewhatthefussisabout.

Although in application big-data analysis can be a very challenging exercise, itsfundamentals are much less complicated than one might expect. Analysts take sets ofquantitative data and run correlations to find relationships that yield insights, perhapsanticipated,perhapsnot,and theyuse thesefindings tomakepredictions.Let’sconsiderthefourimportantelementsinthisdescription.First,thedataunderanalysisareinvariablyquantitative in that operations are applied to numerical values of objects, events,outcomes, ideas, opinions, etc. This does not mean that big data avoids qualitativeinformation, but rather that analysts represent subjective states with quantities—forexample,byassigningnumericalvaluestolikesanddislikesortofeelingsofsatisfactionordissatisfaction.

Second, big data develops generalizations based on correlations among variables.According to two big-data specialists, this means internalizing “a growing respect forcorrelationsratherthanacontinuingquestforelusivecausality”(Mayer-SchönbergerandCukier2013,19).Suchanalysismightleadtotheconclusionthatavoter’sageiscloselyrelated to support for the president. specifically, as age increases, support decreases.Correlationalanalysiscanmeasurethedirectionofarelationship,positiveornegative,andthestrengthof that relationship.But it cannot sayanything,by itself,aboutcausalityorevenaboutwhetherarelationshipisgenuineorspurious.Onecannot,fromthedataitself,determinewhethertwovariablesthatarepositivelyrelatedarealsocausallyrelated—theirrelationship may be caused by another, yet unrecognized, variable or, worse, theirrelationshipmaybea figmentof thedata and thevariables actuallyhavenothing todowithoneanother.Evencorrelationsachievedatahighlevelofsignificance—forexample,thatoutofonehundredsamples,therelationshipwouldshowupninety-fivetimes—givenowarrant to assert causality and to rule out the possibility of a spurious relationship.Correlations help one to determine which among a group of variables go together, orcovary,andtoruleoutwithsomeconfidencethosethatdonot.Butpeopleoftenmistakethis for providing evidence of causality or of certainty that they are tied together,independentofothervariablesthatmayverywellbeessential.Forexample,justbecausethesaleofumbrellasishighlycorrelatedwithcaraccidentsdoesnotmeanthatonecausestheother.Rather, it is thepresenceofa thirdvariable, rain, that influencesboth. In thiscasetherelationshipbetweenumbrellasalesandaccidentsisspurious.

Big-dataanalysisalso tends tobeatheoretical. In fact,majorproponentsboast that itfreespeoplefromcomingupwithhypothesesortheoriestobetestedandallowsthedatatospeakforitself(Anderson2008).Noteveryproponentofbigdataholdsasstronglytothisview,butmostaccept that,givenourabilitytomeasureandmonitorbehavior,fromthe “likes” posted on Facebook to how fastwe drive, the goal of science should be toapplymathematicalprocedures,suchascorrelations,andletgeneralizationsemergefromthedata.Thepoint,asMayer-SchönbergerandCukieremphasize,isthat“nolongerdowenecessarily require a valid substantive hypothesis about a phenomenon to begin tounderstand our world” (2013, 55). Theory’s guiding hand was necessary in the pastbecausetherewasnotenoughdatatorelyonitalonetoprovideanswers.Aworldawashin data can now find, in the analogy often used by bigdata supporters, a needle in a

haystack(Singh2013).Replacingtheoriesandhypothesisaregeneralareasofinterestandspecific questions that the researcher believes big data and the cloud might answer.Anythingmorerigorouswouldprematurelyruleoutentireareaswheresolutionsmightbefound.

The primary goal of big data is to be predictive. Find patterns deep in the data andexpectthat,barringsignificantstructuralchanges,theywilltelluswhatthefuturewillbelike.Determiningwhyislessimportantthanpredictingwhatwillbe.Asa2013overviewconcludes,“We’reenteringaworldofconstantdata-drivenpredictionswherewemaynotbe able to explain the reasons behind our decisions” (Mayer-Schönberger and Cukier2013, 17). Consider the example of Google’s search for the needle of insight into thespreadofflu,agoalthathaseludedexpertsattheCentersforDiseaseControl(CDC)whohave spent years trying to track the disease.Google’s haystack ismore like a toweringskyscraper,withthreebillionsearchesadaysavedinGoogle’sclouds.Drawingfromthisvaststoreofdata,Googlecompared50millionofthemostcommonsearchtermstotheCDC’s informationon the spreadof flu from2003 to2008 (Ginsberg et al. 2009).Thecompany’s researchers looked for correlations between the frequency of certain searchterms and the spread of the virus over space and time. They found that “because therelativefrequencyofcertainqueriesishighlycorrelatedwiththepercentageofphysicianvisits in which a patient presents with influenza-like symptoms, we can accuratelyestimatethecurrentlevelofweeklyinfluenzaactivityineachregionoftheUnitedStates,withareportinglagofaboutoneday”(ibid.).Sincethebestreportinglaguptothispointwasabout twoweeks,Google’sresults,which led to theonline toolGoogleFluTrends,promisedtoprovideflufightersandthegeneralpublicwiththebestinformationonhowtopredictthespreadoffu.Moreover,itcoulddothisunobtrusivelyandinexpensively.BigdatafoundtheneedleintheformofkeysearchtermsandGooglecautiouslybelieveditsmethodmightservetorefineglobalandlocalpreparationsforthevirus.

Big data is now used widely throughout the sciences. Genomics, which uses it todecipher thehumangenome,andastronomy,whichapplies it tomap theheavens,gaverisetothetermbigdata.Accordingtooneassessmentofthebenefitsforgeneticsresearch,“Improvementsinthespeedandfunctionalityofdatacollection,storageandanalysistoolshaveloweredthecostofsequencingfromalmost£2bntoaround£2,000today,andcutthetime it takes fromover a decade to aweek.Whilemore incremental gainswould havetakenplaceatanyrate,suchmajorstrideshaveonlybeenmadeachievablebythecloudcomputingservicesofferedby—amongothers—Microsoft,AmazonandTeradata”(Burn-Murdoch 2012). The Sloan Digital Sky Survey has used big data to analyze moreinformationforastronomythanall theastronomical researchamassedbefore theprojectbeganintheyear2000(Mayer-SchönbergerandCukier2013,7).Physicistsusebigdatato model quantum behavior and climatologists use it to produce models of changingweather.

Bigdataisincreasinglyusedtoanalyze,model,andforecasthumanbehavior(BoydandCrawford2012).Manyoftheseusesarefamiliar,althoughnotoftenassociatedwithbigdata. They includeGoogle, Bing, and other search engines, which apply algorithms todatabasestodeliversearchresults.Facebook’sGraphSearchtakesthistoanewlevelbyprovidingsearchresultstailoredtotherecordofsubjectivechoicessuchasfriendrequestsand“likes.”Seeing thevalueofbigdata inordinaryonlineusehasadded to individual

user capabilities and, in the process, lengthened the industry lead of those companies,especially Google, that have invested in their development. For example, Microsoftpioneered the use of large databases to spell-check its word-processing program’sdocuments,butdidnotpursuethetechnologyfurther,atleastnotasfarasGoogle,whichused thesame technology todevelop its search,autocomplete,Gmail,andGoogleDocsservices. In fact, lessons learned from this use of big data helped convince Google todevelop a completely cloud-based laptop computer, the Chromebook. These big-dataapplications are typically cited indescriptionsof success stories, but others that receivelessattentionbearclosescrutiny.

TheU.S.militaryisaleaderinbig-dataanalytics,withthelargestprojectsrunbytheNationalSecurityAgency(NSA),thecountry’sleadingglobalelectronicsspyagencyandthe subject of considerable controversy in the summer of 2013, after a former NSAcontractor lifted the lidon theagency’smassivedomesticand international surveillanceoperations.Throughitsglobalsurveillancenetworks,theNSAhasbeencollectingdataforsixty years, first intercepting phone calls and now capturing emails and other onlinecommunication, which it stores and assesses through a variety of analytical systems,including keywords that might provide clues about security threats. IBM delivered theNSA its first computer, the top secret Stretch-Harvest, to process surveillance in 1962(Lohr 2013a). This extended a long tradition of government surveillance ofcommunication technologies,which began in earnestwith the telegraph.Back in 1861,just a few years after the technology was deployed, President Lincoln ordered federalmarshals to enter every telegraph office in the United States and seize copies of allmessages,withaneyetorootingoutConfederatesympathizers.

Just as there is nothing especially new about the NSA’s activities, there is nothingparticularly novel about the warnings over its abuse of power. After all, in the 1970s,shortly after the Watergate scandal, a Senate committee warned about the danger thatagencieschargedwithforeignspying,includingtheNSA,posedtotheAmericanpeople(Greenwald2013).TheeffectivenessofNSAactivitieshasnotalwaysbeenclear,inpartbecause theagencycollected farmore information than itwasable toanalyze.For this,bigdataprovideswhatishopedtobeasolutionbystrengtheningthecapacitytoprocessdata, applyanalytical tools, andmakepredictions.Todeepen its analytical capacity, theNSA has built a close relationship with Silicon Valley, to the extent that one analystconcluded that “they are now in the same business” (New York Times 2013a). Othersmaintain that connections between the NSA,Microsoft, Google, Apple, Facebook, andmajor telecommunications firms make up a data-intelligence complex, a contemporaryversion of themilitary-industrial complex that PresidentEisenhower criticizedwhen heleft office in 1960 (Luce 2013). The Pentagon and U.S. intelligence agencies are anincreasinglyessentialtraininggroundforstart-upcompanies.AnNSAemployeewholeftto start a successful tech company praised the agency for putting him “on the bleedingedge, not just the cutting edge of what’s possible” (Sengupta 2013). Nevertheless, therelationship between private companies and the intelligence agencies is far fromharmonious. The scandal that arose from revelations about NSA spying and theinvolvement of the major computer and social-media companies led to business fearsaboutadeclineofpublictrustintheonlineworld.Asaresult,inDecember2013Apple,Yahoo!,Facebook,Twitter,AOL,andLinkedInjoinedGoogleandMicrosoft inanopen

letter to the president and Congress calling for reform and regulation of onlinesurveillancebygovernmentagencies(WyattandMiller2013).

The2013NSAscandalisunlikelytoslowtheconstructionofanNSAclouddatacenterin Utah for the storage, processing, analysis, and forecasting needs of the agency,estimatedtocost$2billion(Bamford2012).Asjournalistswhohavetriedtoinvestigatewhat is benignly called theUtahData Center have learned, the site is shrouded in thesecrecy that one has come to expect from the NSA (Hill 2013). After all, it is hardlysurprising that an agency whose budget is kept secret from public scrutiny (estimatesrangeinthetensofbillionsofdollars)wouldnotopenthedoorsofitslatestbigproject.Accordingtooneoftheworld’sleadingexpertsontheNSA,“Flowingthroughitsserversandroutersandstoredinnear-bottomlessdatabaseswillbeallformsofcommunication,includingthecompletecontentsofprivateemails,cellphonecalls,andGooglesearches,aswellasall sortsofpersonaldata trails—parking receipts, travel itineraries,bookstorepurchases, andother digital ‘pocket litter.’ It is, in somemeasure, the realization of the‘totalinformationawareness’programcreatedduringthefirsttermoftheGeorgeW.Bushadministration—aneffort thatCongresskilled in2003after itcausedanoutcryover thepotentialforinvadingAmericans’privacy”(Bamford2012).

TheUtahDataCenter isamonumentalconstructionprojectbuiltaroundfour25,000-square-foot buildings that house cloud servers to process and analyze data, with floorspaceraisedtopermitaccessforcablesdeliveringdatafiles.Fully900,000squarefeetofspacewillbe set aside for technical support andmanagement.Thebudget includes$10millionforextraordinarymeasurestosecurethefacility,whichincludesafencereportedlycapableof stoppinga15,000-poundvehicle traveling at fiftymilesperhour.Theentireoperation isconsideredself-sustaining,with itsownsubstationable todeliver sixty-fivemegawattsofelectricity.

Three key developments prompted the construction of theUtah site. The first is themassivegrowthofinformationworldwidethatrequiresenormousinvestmentinfacilitiesandprocessingpower.Analyzingpublicdataalonewouldbedaunting,asoneestimatehastheentirestockofdataontheInternetquadruplingbetween2010and2015,toover950exabytes. The total amount of information created from the dawn of writing to 2003amountedtoabout5exabytes(Bamford2013).ButtheNSAneedstogobeyondwhatispublicly available to capture and examine information contained on the deep web, ordeepnet, which includes classified reports from governments and businesses that areprotected by encryption systems that big data enables theNSA to crack.As one of theforemostexpertsontheNSAconcluded,“WithitsnewUtahDataCenter,theNSAwillatlasthavethetechnicalcapabilitytostore,andrummagethrough,allthosestolensecrets”(ibid.;seealsoDeibert2013).

Second is the expansion in the agency’s domestic spy operations (Clement 2013).Initiallychargedwith interceptingelectronic traffic toand from theUnitedStates,NSAsurveillancenolongerstopsattheU.S.border.Inthewakeofthe9/11attacks,accordingto Bamford and former NSA employees, it installed what amount to taps on majordomestictelecommunicationsswitchesandsatelliteearthstations.Italsosetupbetweenten and twenty facilities in the United States to analyze electronic traffic within thecountryandextendedtheNSA’sreachwithsurveillanceloopsintomajorCanadiancities

(Bamford2013;Clement2013).While theagency is formallyprohibited fromdomesticspying,therearedifferentperspectivesonitslegalityandconstitutionality,particularlyinlightofpost-9/11legislationthatexpandsthegovernment’spowertointerceptelectroniccommunicationwithintheUnitedStatesandabroad.WiththehelpofaBoeingsoftwaresubsidiary,theNSAcannowremotelycontrolsoftwarefromitsMarylandheadquarterstosearchU.S.databases, including, itnowappears,2.8trillionbillingrecordsoftelephonecalls stored in an AT&T facility covering individuals and organizations it targets forrecording, transmission, andanalysis.TheUtahDataCenter expands theopportunity toanalyzeandmakeuseofthesemassivenewstoresofdata.

Third,asdauntingasitistokeepupwiththegrowthintraffic,theNSAhasbenefitedfromtheexpansioninprocessingpowerandbig-dataanalysisthatenablestheagencytoactuallyusewhat itgathers toanalyzeintelligenceandforecastevents.Theagencynowhas thecapacity toenteranameinto itsdatabaseandautomaticallyrouteandrecordallelectroniccommunicationtoandfromthatperson.WhentheNSAconsidersitnecessary,the agency can carry out a detailed analysis of communication content and use it tocompleteariskassessment.Alongwithcontentsurveillance,theagencyusesmetadatatomapthesocialnetworksofindividualstodeterminetheimplicationsofstrongandweaknetworkconnectionsaswellastiesthatcanbeimpliedbynetworksofassociationsamongdifferent people. Given the quantitative and qualitative expansion in its capabilities, aformerNSAemployee,WalterBinney,believesthattheagencyhasshiftedfromfocusedcollectionandanalysisofdataonforeignthreatstogatheringasmuchdataonforeignersand Americans as the technology allows (Bamford 2012).2 Moreover, the predictivecapabilityofbig-datasystemsmakesitevenmorelikelythattheNSAandagencieslikeitwillcollectfarmoredatathantheyneed.Thatisbecauseimprovementsincrackingdata-encryptioncodeskeepopenthelikelihoodthat,iftheagencycannotdecipherandanalyzedatanow,itwilllikelybeabletodosointhefuture.

TheNSAisattheleadingedgeofaconcertedprogramthatalsoinvolvestheCIA,theDefense Advanced Research Projects Agency (DARPA), and other military andintelligenceorganizationsthatputbigdatatowork,forexample,inthecontroversialuseofbattlefieldattackdrones.Consideringthesheeramountofdatathatmustbeprocessedto carry out a successful drone attack, it is not surprising that there are significantchallenges to successful applications. In fact, some insiders question the expansion ofdroneprogramsbecausetheyrequireprocessingcapabilitiesthatexceedtoday’sbudgetaryand technological limits (Beidel 2012). Pushing ahead, in 2012 the federal governmentannouncedspendingofmorethan$200milliononbig-datamilitaryandcivilianresearchanddevelopment.Accordingtothepressreleaseaccompanyingitsannouncementof the“Big Data Initiative,” the Department of Defense (DOD) will “place a big bet on BigData”with$60millioninnewannualspending.Thegoalisto“accelerateinnovationinBigData” thatwill“improvesituationalawareness tohelpwarfightersandanalystsandprovideincreasedsupporttooperations.TheDepartmentisseekinga100-foldincreaseinthe ability of analysts to extract information from texts in any language, and a similarincreaseinthenumberofobjects,activities,andeventsthatananalystcanobserve”(U.S.OfficeofScienceandTechnologyPolicy2012).Thisfundingisexpectedtosignificantlyexpandthemilitary’sdroneattackprogram(Beidel2012).

AlongwiththeDODinitiative,DARPAannouncedaninvestmentof$25millionayear

initsXDATAprogramtoovercomecurrentlimitationsinbig-dataanalysis.specifically,itis focusing on developing software and other computational tools, such as improvedalgorithmsandvisualrepresentations,toexaminethesemistructuredandunstructureddataintextdocumentsandmessagetraffic.TheannouncementdidnotincludetheNSAortheCIA,whoseprogramsarenotpublicizedinpressannouncementsandwhosespendingiskeptseparatefromDODauthorizations.ItishardtosaywhetherthebadpublicitythattheNSAhasattractedinthewakeofrevelationsabouttheextentofitssurveillanceactivities,especially againstAmericans,will dampen the government’s commitment to expandingthemilitaryuseofthecloudandbigdata.Thisisunlikely.Whilethenamesofprogramschange(todayitisPrism,tomorrowsomethingelse),theNSAhasbeeninthesurveillancebusiness for more than half a century and its work is vital to U.S. spy operations.Nevertheless, some rethinking is likelybecause revelationsofelectronic surveillanceontheofficesofallies,particularlyintheEuropeanUnionandinLatinAmerica,havecreatedenoughangertodamagerelationstothepointofthreateningsensitivetradenegotiations(Castle2013).Indeed,someanalystsarewonderingaloudwhetherrevelationsaboutNSAactivitieswillsignificantlyunderminesupportforcloudcomputingworldwide(Linthicum2013d).OnethinktankestimateslossestotheU.S.cloudindustryatbetween$21.5and$35billionover thenext threeyearsasa resultof fearsgeneratedbyNSAsurveillance(Taylor 2013a). Cisco claimed that it had already lost business in emerging marketsbecauseofconcernsaboutU.S.spying(Meyer2013).

Thegovernmentcommitmenttocloudcomputingisnotlimitedtomilitary/intelligenceapplications.Inadditiontoadvancingresearchinmedicineandhealthcare,itislookingtoreducehealthcarecosts,andtheanalysisandpredictivepromiseofbigdataaremeansofmeetingthisgoal.Tothatendthegovernmentisfundingajointprojectbringingtogetherthe National Science Foundation and the National Institutes of Health to research“managing, analyzing, visualizing, and extracting useful information from large anddiversedatasets”(U.S.OfficeofScienceandTechnologyPolicy2012).Whileimprovingthe analysis and display of data is not controversial, the ultimate goal of predictingoutcomesbasedonpatientinformationhasstirredconcernsthatgovernmentwillusetheresultstomodifybehaviorinwaysconsideredexcessivelyintrusive.Forexample,shouldthegovernmenttailoritsmedical-insurancecoveragetothehealthchoicesofAmericans,withcutstobenefitsforthosewhomakewhatthedatasuggestsarebadchoices?Anotherhealth-relatedfield,genomics, isalsoapopularsubject inbig-datadiscussions.HerethegovernmentisteamingwithAmazonWebServices(AWS),whichhelpedbringvictorytoPresidentObamain the2012election, tostore200terabytes(16millionfilecabinetsor30,000 standardDVDs)ofdata fromgenomics research.Thedata is publicly available,but users have to pay AWS for computing costs. It is interesting to observe anotherexampleofthegovernment’sdependenceonprivatecloudcompanies,inthiscaseoneofthe most important in the world, to store, process, and distribute valuable data sets.Finally, energy and geology research receive funding to advance the capacity of thesefieldstoanalyze,visualize,andpredictthebehaviorofresourceandgeologicalsystems.

Big data is increasingly used in the traditional social sciences and in the humanities.Social-scienceresearchisnowoftenconductedbyprivatecorporationsthatseesignificantopportunities in areas such as real-time fraud detection, health risk assessments formedical patients, continuous process monitoring of consumer sentiment or vital

mechanical systems, andnetwork relationships on social-media sites (Davenport,Barth,and Bean 2012). Large data sets are providing new opportunities for research withpracticalconsequences.Forexample,aUnitedNationsagencysupportsbig-dataresearchon how organizations respond to humanitarian crises. The data include social-mediacontentwith the goal of creating recommendations onwhatworks best (Burn-Murdoch2012). Similarly, in Sierra Leone, the mapping company Esri provides software and acloud portal that reveals where health clinics are needed (A. Schwarz 2013). Datascientists working with the London-based organization DataKind provide advice tocharitiesabouthowtodealwithproblemsinthenonprofitsector.Furthermore,researchersassociated with Toronto’s Hospital for Sick Children have used big data to developalgorithmsthatanticipateinfectionsinprematurebabies.Notwithstandingthesebenefits,because similar types of algorithms can be used by insurance companies to refusecoverageorbysocial-mediacompanies tomanipulate“trending” results, therearemorethan a few worries about ethical and political issues (Burn-Murdoch 2012; Gillespie2013).Thishasledsomedatascientiststopromoteacodeofgoodbehavior,“DoingGoodwithAnalytics,”thatcommitstoassessingtheethicalvalueofresearchbeforetheprocessbegins and tousing it tobring aboutpositive social change (D.Ross2012). It has alsoprompted calls to democratize data science by making the new field more open andaccessibletocitizens(Harris2013b).

Sinceprivate corporationscontrolmostof the researchusingbigdata, concernshavebeen raisedaboutaccess todatabecause firmsare reluctant to follow traditional social-scienceprotocolsforreleasingevidencereportedoninacademicpapers.Theissuecameto a head in 2012when researcherswithGoogle andCambridgeUniversity refused tomake available data for a conference paper on the popularity of YouTube in severalcountries.Thechairmanoftheconference,aphysicistwhoheadsasocial-scienceresearchgroup atHP, responded angrily and recommended that the conference shouldno longeraccept papers from researchers who, whether for a commercial, security, or any otherreason, refuse to sharedata.He followed thatupwitha letter to theprestigious sciencejournalNature declaring that big-data analysis,whichwas supposed to expand researchhorizons, is actually narrowing them because the private companies that own the datarefusetoreleaseit(Markoff2012).Ontheotherhand,agrowingnumberofcriticalsocial-sciencescholarsaredeveloping tools tousecommercial softwareanddatageneratedbysocialmediatoadvancealternativevisionsofsociety(Beer2012).

Big data is also increasingly used in the humanities, shaking up traditional researchapproachesandstirringconsiderabledebate(Hunter2011).IntheUnitedStates,thepushtousebigdataintheliberalartsisledbythefederalgovernment’sNationalEndowmentfortheHumanities(NEH).Oneofthelargestfundersofliberal-artsresearchintheUnitedStates,NEHisa federalagencyfounded in1965.Withanannualbudgetofabout$170million, theagencyprovidesgrants tocultural institutions suchas libraries,universities,museums, public broadcasters, and individual scholars in order to strengthen teaching,research, and the institutional base of the humanities, including expanding access toeducationalandculturalresources.NEHcreatedtheDigitalHumanitiesInitiativein2006,anditwasraisedtothelevelofanOfficeofDigitalHumanities(ODH)in2008,amovethat helped to legitimize use of the term digital humanities in the United States.WithODHsupport, scholarsworking in the fieldmade theirpresence felt at the2009annual

meetingoftheModernLanguageAssociation,whatmanyconsideraturningpointinthefield.Digitalhumanistsapplycomputersciencetothehumanities,primarilybyexamininglargedatasetstocarryoutresearchthatwasdifficult,ifnotimpossible,tocompletebeforecomputationalmethodswere available to scholarsworking in such humanities fields asliterature,history,andphilosophy.

Someof the research, such as theODH-fundedVisualPageproject, involves findingnewways togatherbigdata and analyze it: “All printed texts conveymeaning throughbothlinguisticandgraphicsigns,butexistingtoolsforcomputationaltextanalysisfocusonly on the linguistic content. TheVisual Pagewill develop a prototype application toidentifyandanalyzevisualfeaturesindigitizedVictorianbooksofpoetry,suchasmarginspace, line indentation, and typeface attributes” (U.S. National Endowment for theHumanities2013).Otherprojectsdirectlyapplycomputationalmethods toanalyze largedata sets;oneof these isanODH-fundedprojecton the lifecyclesofpublishedworks:“includingnotonlyscholarlyandscientificliterature,butalsosocialnetworks,blogs,andother materials.” The goal is to “identify which scholarly activities are indicative ofemergingareasandidentifydatasetsthatshouldnolongerbemarginalized,butbuiltintounderstandings and measurements of scholarship” (ibid.). Another funded projectdemonstrateswhy the grant program is called “digging into data”: because it looked at“newwaysofexploringthefulltextcontentofdigitalhistoricalrecords…usingmedievalcharterswhichsurviveinabundancefromthe12thtothe16thcenturiesandareoneoftherichest sources for studying the livesofpeople in thepast.ThenewChartEx toolswillenable users to really dig into the content of these records, to recover their richdescriptionsofplacesandpeople,and togofarbeyondcurrentdigitalcatalogueswhichrestrictsearchestoafewkeyfactsabouteachdocument(the‘metadata’)”(Digging intoDataChallenge2011).The ODH program has succeeded in giving the humanities a significant push into

quantitative research that takes advantageof cloudcomputing systems to examine largesetsofdata.3ODHhasalsoattractedinternationalattentionandsupport.Its2009and2011“digging into big data” competitions received proposals from 150 research teams andfunded 22 from theUnited States, UnitedKingdom, Canada, and theNetherlands. For2013, support and sponsorship expanded across new research councils and governmentfunding authorities, giving the program ten sponsors. This is significant becausegovernment support for thehumanities, including research, teaching, and archiving, hasdeclined to perilous levels in most Western societies, leaving cloud-based, big-dataresearch one of the few areas where funding is on the rise (Delany 2013). Moreover,governmentresearchcouncilsthathaveseentheirbudgetscutaredevotingmoreofwhatlittleislefttofundingcomputationalresearchinthehumanities.Defendersofthedigitalhumanities support this shift because they believe it is bringing about a revolutionarytransformationinallfacetsofhumanitieseducationandresearch.AstheheadoftheNEHexclaimed,“Arevolutionhascommencedwherescienceandtechnologyaremeldingwiththehumanities”(Leach2011).

Noteveryoneinthehumanitiesseesitthisway,includingStanleyFish,oneofthemostdistinguishedliteraryandcultural-studiesscholarsofourtime.ForFish,mostsupportersof the digital humanities advance a view that he considers “theological” because itpromisesfreedomfromtheconstrictionsofamediumthatisbothlinearandtime-bound,

whichcanonlyproduceknowledgethatisdiscrete,partial,andsituated(i.e.,forhereandnow,bythisauthor,andfor thisaudience).For itssupporters, thedigitalhumanitiesusethecloudandcomputationalmethods toprovideauniverse inwhichknowledge is fullyavailableeverywhereandtoeveryone.Throughit,weallbecomenodesinanetworkofmeaningproductionforasystemthateliminatesthespatialandtemporalbarriersbetweenthe person seeking knowledge and the object of cognition. Fishmaintains that this is astate thatmost religions identifywith theafterlife,whenpeoplecastoff the shacklesofmortality and all of its limitations to become one with the creator, the source of allknowledge.Headmitsthatnooneinthefieldspeakspreciselyinthisway,butsaystheymayaswellbecause,fordigitalhumanists,theirmissionaffirmsafutureof“expanding,borderless collaboration inwhich all the infirmities of linearitywill be removed” (Fish2012a). He cites Fitzpatrick (2011), whose book Planned Obsolescence describes thelimitationsoftraditionalmediaandthesocialrelationsthatarosewiththem,maintainingthatinaworldofnewmedia“weneedtothinklessaboutcompletedproductsandmoreabouttextinprocess;lessaboutindividualauthorshipandmoreaboutcollaboration;lessaboutoriginalityandmoreaboutremix;lessaboutownershipandmoreaboutsharing”(p.83).

In his critique of what he considers the theology of the digital humanities, Fish isdescribingwhat I have called the digital sublime (Mosco 2004). At the very least, thedigitalhumanitiesmythologize theonlineworldbyviewing itasmeansof transcendingthe banalities of everyday life, but evenmore so by helping to bring about the end ofhistory, the end of geography, and the end of politics. In its extreme form, the digitalhumanitiesareclearlytheologicalinthattheydrawinspirationfromthewritingofpeoplelikeTeilharddeChardin(1961),whoenvisionedmankindfindingunitywithGodthroughthenoosphere,theliteralatmosphereofthoughthebelievedwascreatedbythegrowthofinformation. The work of Ray Kurzweil (2005) on informational immortality and thesingularitymarryTeilhard’stheologywiththedigitalworld.

Fish also takes issuewith thedigital humanitiesonpolitical grounds, particularly thegoals of democratizing the humanities by breaking down the barriers that separatedisciplinesandthebarriersseparatingscholarsfromthegeneralpublic.WhatmakesFish’scritiqueinterestingis thatheisnotopposedtothesegoalsperse,buthedoubtsthat thedigitalhumanitiescanreachthem.Forhimtheyaremorelikemythiccoversthatjustifytheprimarygoalofgatheringasmuchquantitativedataaspossibleon literary textsandotherworksofpopularcultureto,attheveryleast,inspirenewreadingsoftextsandnewassessments of the process and the context of their creation (Fish 2012b). The digital-humanitiesmovementhassparkedrigorousdebate,withproponentsmakingreferencetothe “backward” humanities and opponents using words like “diabolical” to describeFrancoMoretti,oneofitsleadingpractitioners(Sunyer2013).

Thereisnothingnewintheprinciplesbehindbig-dataanalytics.Formanyyearssocialscientists have been working on large data sets to find relationships among seeminglyunrelatedvariables.Butthedifferencenowistheconcertedefforttomakeitthesingularlymostimportanttoolinresearchand,forsome,themagicalalternativetothemethodsthathaveguidedresearchinscienceaswellasthehumanitiesforcenturies.Bigdataisnotjusta method; it is a myth, a sublime story about conjuring wisdom not from the flawedintelligence of humans, with all of our well-known limitations, but from the pure data

storedinthecloud.

Proclaiming“theendof theory,”ChrisAndersongot theball rolling ina2008Wiredmagazine article in which he stated, “the data deluge makes the scientific methodobsolete”(Anderson2008).ForAnderson,bigdatamarksnothingshortofarevolutioninwhat it means to know. This view is mythic because it envisions big data as arevolutionarydevelopmentthatdoesnotjustmakesciencebetter,butendsscienceasweknow it and replaces itwithanewwayofknowing.Likemanymyths,Anderson’s taleimagines a new world where what was universally accepted yesterday is rejected anddiscardedtodayinfavorofasimplealternativethatsolvestheworld’sproblems.Outwiththescientificmethod,inwithbig-datacorrelations.FollowinganexampleofhowGoogleisrevolutionizingadvertising,Andersonproclaimed,“Thebigtargethereisn’tadvertising,though.It’sscience.”Ormoreprecisely,itisthecoreofscienceembodiedinanapproachtoknowledge.“Thescientificmethodisbuiltaroundtestablehypotheses.Thesemodels,for themostpart,aresystemsvisualized in themindsofscientists.Themodelsare thentested, and experiments confirm or falsify theoretical models of how the world works.Thisisthewaysciencehasworkedforhundredsofyears.”Itnolongerhastoworkthisway, but scientists have to give up their cherished notions. “Scientists are trained torecognizethatcorrelationisnotcausation,thatnoconclusionsshouldbedrawnsimplyonthebasisofacorrelationbetweenXandY(itcouldjustbeacoincidence).Instead,youmustunderstandtheunderlyingmechanismsthatconnectthetwo.Onceyouhaveamodel,you can connect the data setswith confidence.Datawithout amodel is just noise. Butfacedwithmassivedata,thisapproachtoscience—hypothesize,model,test—isbecomingobsolete”(ibid.).

Attheircore,mythshelpustocopewithlife’suncertainties,fromthelittlebanalities,suchaswhattohaveforbreakfast,tothegrandquestionsofhowtofindmeaningandfacemortality. They do not just offer an answer; they provide the answer, typically withconvincingclarity,simplicity,andfervor.Bigdataisnotjustoneamongmanyinstrumentsto understand and change the world; it is the essential one, and all others, includingscience, themethod that has guided themodernworld and itsway of knowing, can besweptintothedustbinofhistory.Someunderstandthiswell.PeoplelikeChrisAndersonandRayKurzweilaretoday’sseers,whoknowthewaythatdrawsthecurtainonanoldageandforeshadowsthenew.Mostmythsareaboutendings,whethertheendofhistory,oftheory,orofscience.Theycallonustocelebrateourgoodfortunetoliveattheendofan era and to begin to experience the new. For Anderson, today’s visionary is Googlebecauseitisnotjustasuccessfulcompany,aleadingforceininformationalcapitalism,butprimarilybecause it isusing thecorrelations it finds inmountainsofbigdata tochangewhat itmeans to know: “The new availability of huge amounts of data, alongwith thestatistical tools to crunch these numbers, offers awhole newway of understanding theworld.Correlationsupersedescausation,andsciencecanadvanceevenwithoutcoherentmodels,unifiedtheories,orreallyanymechanisticexplanationatall.There’snoreasontoclingtoouroldways.It’stimetoask:WhatcansciencelearnfromGoogle?”(ibid.)

For some, the new visionary is the data scientist whomagically conjures truth frommountainsofseeminglyunrelatedinformation.Accordingtooneobserver,“bigdatahascreated amythical god called the data scientist: a lone-wolf, super-smart humanwith asolid foundation in computer science, modeling, statistics, analytics, math, and strong

businessacumen,coupledwiththeabilitytocommunicatefindingstobothbusinessandIT leaders in a way that can influence how an organization approaches a businesschallenge” (Walker 2013). One observer sees the data scientist as the successor to theiconic“MadMen”ofadvertising(Steel2012a).Mythsmatter.Inthiscasetheemergenceof the data scientist as the latestmythical god is having a significant impact on highereducation,where universities are scrambling to produce programs to train aspirants forwhattheHarvardBusinessReview,nostrangertohyperbolicexcess,calls“thesexiestjobin the 21st century” (Miller 2013).Despite budget constraints created in part by failedprogramsinspiredbythedot-combubbleof the late1990sandthefinancialbubble thatgreetedthenewcentury,dozensofnewprogramshaveemergedateverylevelofhighereducation.EventheusuallysubduedNewYorkTimeshascaughtthefever.Declaringdatascientists“themagiciansoftheBigDataera,”thenewspaperdescribestheirmanytalents:“They crunch the data, usemathematical models to analyze it and create narratives orvisualizations toexplain it, thensuggesthow touse the information tomakedecisions”(ibid.).Itisuncertainwhethertheycanalsobringhomethebaconandfryitupinapan,buttheTimes issatisfied to transmit,withnocritical reflection,apromotional reportbyMcKinseythatforecaststhemillionsofjobsthatthedemandfordatascientistswillcreate.It is remarkable that after the disastrous economic catastrophes brought about by near-rapturousfaithintheITofthelate1990s,andinthebig-dataalgorithmsthathelpedbringtheWesttothebrinkofanewGreatDepressionin2008,educatorscontinuetochaseafterthenextnewfad.Thistimewillbedifferent.Mythsmatter.4

A current exemplar of myth-building around big data is a 2013 book by a pair ofknowledgeable analysts whose breathless prose begins with its title: Big Data: ARevolutionThatWillTransformHowWeLive,Work,andThink.Oneofthecharacteristicsofagoodmythisitsabilitytoinoculateitsstorywithwhatappearstobesobergoodsenseinordertoachieveadegreeoflegitimacybeforeplowingaheadwiththetalltale.FortheauthorsofBigDatathismeansputtingsomedistancebetweenthemandChrisAnderson:“Bigdatamaynotspellthe‘endoftheory,’butitdoesfundamentallytransformthewaywemake sense of theworld” (Mayer-Schönberger andCukier 2013, 72).Here,we areencouraged to question the implied hyperbole even as we adopt another, equallyextraordinary claim. For the authors, “the IT revolution is all around us” and it ismanifestednot in the technology,but in information,which takesonseeminglymagicalpowers to change thewaywe know theworld (ibid., 77–78).This appears againwhenthey turn to the method of choice in big-data analysis, finding correlations: “Withcorrelations, there is no certainty, only probability. But if a correlation is strong, thelikelihood of a link is high.” They “demonstrate” this by asking us to observe theconnectionbetweenAmazon’sbooksuggestionsandthosebooks’appearanceonpeople’sshelves(ibid.,53).Undeterredbytheabsenceofanythingresemblingevidencetosupporttheir contention, they plow forward: “By letting us identify a really good proxy for aphenomenon,correlationshelpustocapturethepresentandpredictthefuture”(ibid.,53–54).Whatcouldbemoremythicalandsublime,moreevidenceoftheconjurer’sart,thanthemagicwandofcorrelation?Only thismagicdeliversmore than rabbits fromhats. Itcantelluswhatisandwhatwillbe.

Becausemythsmatter,itisimportanttoprovidesomecriticalreflectionontheseclaims.Butitisalsoessentialtounderstandthelimitsofanysuchcritique,howevertelling.The

cloud and big data aremore than technical developments because their emergence hasinspiredanewmythologythatputsafreshfaceonthedigitalsublime,which,attheendofthe last century, promised to end history, annihilate geography, and transform politics.Likeallmyths,theyarefullofmagicalconjurerswhoofferrevolutionarytransformationsandhappyendings thatbidgood-bye to the temporal,spatial,andsocialconstraints thatmakeup thebanalitiesofeveryday lifeandwelcomeanewworld in thecloud.Wecannowknow thepast, represent thepresent, andpredict the future likeneverbefore,withlittle of this contaminated by flawed human decision making. The data will speak forthemselvesorthroughdata-sciencemagicians.Likeallmyths,theycannotbefullyjudgedbased on their claims of truth, but rather, as the philosopherAlisdairMacIntyre (1970)concluded, only onwhether they are living or dead.Myths live on if they continue tomake lifemeaningful and if they continue tomake socially and intellectually tolerablewhatotherwisemightbeexperiencedaspainfuland incoherent.Mythsdonotdisappearwhentheyarefalsified—considertheirpersistenceafterthedot-combustandthefinancialcrash—as long they continue to energize people and feed their hopes and dreams. Thecloudandbigdatado sobypromisinganendless supplyof accessible information thatwill be used to solve the problems that afflict theworld andmake it possible to enjoyformsofperfectionthathaveheretoforebeenlittlemorethanthestuffofdreams.

BigData:ACritiqueofDigitalPositivismBig data gives priority to quantitative over qualitative data, arguing that the formerprovides the best opportunity for meaningful generalizations and that, when necessary,qualitative states can be rendered qualitatively. For example, a quantitative contentanalysisofsearchtermsrelatingtofluprovidedGooglewithwhatitbelievedwasastringoftermsthatcorrelatedwithfluoutbreaks,therebyenablingresearcherstopredict,earlierthanever,thespreadofflu.If,ontheotherhand,onechosetocarryoutabig-dataanalysisofasubjectivestate,saybyassociatingpositiveTwitterpostsabouttheToyotaPriuswithsales of the car, then one might assign numerical values to capture the strength ofresponderposts.Orbigdatamightrunananalysisthatcombinestheresultsofnumerouscustomersatisfactionsurveysthatassignanumbertoeachpossibleresponse,suchasa5forstrongdislikeora3forsimplydisagreeingwithastatement.Afterall,stronglylikeordislikerepresentsamorepowerfulattractionthanjustlikeordislike.Themeasurementofquantityisnotonlycentral;itisabsolutelyessentialtothetransformativecapacityofbigdata.Astwoofitsproponentsattest,“JustastheInternetradicallychangedtheworldbyaddingcommunicationstocomputers,sotoowillbigdatachangefundamentalaspectsoflife by giving it a quantitative dimension it never had before” (Mayer-Schönberger andCukier2013,12).There ismuchtobesaidforquantitativeanalysis. It renderscomplexbehavior,aswellasmentalstates,easytoprocessandanalyze.It isnowonderthatbig-dataspecialistsbelievethat“themorequantitativeitis,thebetter”(Morozov2013b,232).The ease of analysis, the opportunity to draw broad generalizations and then to makepredictions, provides a strong temptation to reduce all methodological approaches toquantitativeones.Indeed,thehotnewprofessionofdatascientistknowsonlyquantitativeapproaches. Moreover, big data makes it possible to avoid the need to sample apopulation,andalloftherisksassociatedwithaccuratelyrepresentingalargergroup,byexaminingresultsforanentirepopulation.

Theproblemswithrelyingsolelyorprimarilyonquantitativeanalysisaretodaymoreoften thannot ignored, but that is amistake.Quantitative researchprovides a scientificglossonbehavioralorattitudinaldatathatisoftenfarmessierthanthenumbersmakeitappear.Socialscientistsarewellawareofthelimitationsofworkingwithdataonreportsof lawbreaking behavior that are often massively skewed by the human limitations ofwitnesses, police, and the vagaries of plea-bargaining and trials. Nevertheless, big-datasupporters and their corporate sponsors continue to press for what is euphemisticallycalled“predictivepolicing”(Bachner2013).Becausequantitativeresearchworksbestondataembodyinglittleinthewayofsubjectivity,researcherstendtoneglectquestionsthatrequiretheircarefulconsideration.Itisfareasiertogoforthelow-hangingfruitofvoteranalysis (there is little subjectivity in the determination of whom one votes for), or ofcountingthefrequencyofsearchterms,thantoexamine,forexample,howayoungpersonbecomesaracist.Thelatter involvesanaltogetherdifferentkindofmethodology,whichmightmakeuseof somequantitativedatabutalso requirescloseobservationanddepthinterviews—inotherwords,acarefulqualitativestudythataims tocomprehendtherichsubjectivity that makes up personal and interpersonal experiences. Big data deals withsubjectivity to the extent that analysts can do the impossible—i.e., assign a precisenumericalvalue to itsvarious states.This is inherently flawedbecause subjective statessuch as happiness, depression, or satisfactionmean different things to different people,

andassigningthesamenumericalvaluetothechoiceofthistermsimplifiestothepointofabsurdity.The samegoes for other attitudinal terms such as like and dislike, agree anddisagree, and their amplifiers, such as “strongly.” What is the meaning of a numberassociatedwiththeseterms?Howcanoneassignanymeaningworthtakingseriouslytothenumericaldifferencebetweendisagreeandstronglydisagree?

Itisuncertainwhichisworse:thatbigdatatreatsproblemsthroughoversimplificationor that it ignores those that requireacareful treatmentofsubjectivity, including lengthyobservation,depth interviews,andanappreciation for thesocialproductionofmeaning.Thereisadifference,asthecomputerpioneerJaronLaniernotes,betweenusingbigdatato analyzeweather or galaxy formation andusing it to examine the emotional states ofhuman beings, which are often contradictory and unreliable (Lanier 2013). Such anapproachonlyfeedswhatRomanKudryashov,drawingonRolandBarthes,referstoasthemythofthequantificationofquality:“Whenlanguagecannothandlethecomplexitiesofreality, it strives to economize the world: qualities become quantities, and once again,languagegoesbeyondrealitytojudgeit.Thoughlanguagetriestobescientificaboutitsdescriptionshere,ithasattributedpropertiesnotbelongingtotheoriginalobject,andthusdoes not judge the object, but its properties” (Kudryashov 2010). As Barthes himselfasserted, “Awholecircuitof computableappearancesestablishesaquantitativeequalitybetweenthecostoftheticketandthetearsofanactor”(1982,144).Thiscommenttakesustocorrelation,thekeytechniquefordrawingquantitativeconclusionsthroughbig-dataanalysis, whether it is the relationship of a ticket price to an actor’s tears or betweensearchtermsandthespreadofflu.

As a sociologist, I am very familiar with both the magic and the danger of thecorrelation.As a graduate student in the 1970s I can recall turning in punch cards andreceiving printouts that appeared magical because they provided me with a series ofcorrelationsandconfidence levels (measuresofstatisticalsignificance) that,evenarmedwithmystatisticstextbook,oncetookhourstocomplete.Thisgavemethefirstsmalltasteof what a mainframe computer could do, but it was still within the realm of my owncomputationalpowers.Moreofaleapcameinthe1980swhen,withanothercolleague,IlaunchedmyownmajorresearchprojectbasedonanationalsurveyoftelephoneworkersinCanada(MoscoandZureik1987).Forthis,thevariablesmultipliedexponentiallyandsowerefarbeyondmanualcalculations.Buttheretheywere,hundredsofcorrelationsthatbroughttogetherdemographicdataontheworkforce,everythingfromagetojobcategory,withattitudesaboutthework,workmates,surveillance,andthetechnologythatwastakingovermoreandmoreofthelaborprocess.Thisappearedtobeevenmoremagicalbecausecomputerswerenowdoing something that I couldnot even conceivably accomplishonmyown.Whilenotexactly thestuffof today’sbig-datastudies,becausewe reliedonanationalsampleratherthanacompletepopulation,itgavemethefirstfeelingofwhatitwasliketoreviewaprintoutwhosenumbersappearedtospeaktome.Butitdidnottakelong, especially because the senior member of our team was an experienced hand, tounderstandthatmuchofwhatIwaslookingatwasofourownconstruction.Wesetupanddefinedthevariables,creatingthemoutofourowntheoreticalvisionthatestablishedwhatmatteredmost inourview—theimpactofelectronicsurveillanceonjobsatisfaction.Asthepopular(andverysuccessful)dataanalystNateSilverexplained,“Thenumbershavenowayofspeakingforthemselves.Wespeakforthem.Weimbuethemwithmeaning.”

Anyotherviewis“badlymistaken”(Asay2013).ThatbecameabundantlyclearwhenIrealizedthatmostofwhatwasspoken,whoeverwasdoingthetalking,wasgibberishor,whatSilverandotherscallnoise (Silver2012).Thatwasprimarilybecausemostof thecorrelationswefound,howeverstrong,werespuriousorirrelevant;thatis,therelationshipfound between two variables either was created by one or more other variables or thecorrelationsthemselvesweretrivial.Ratherthanfindaneedleinahaystack,bigdata,asNasimTalib(2012)andDavidBrooks(2013)haveperceptivelynoted,oftenjustleadstomorehaystacks.AsBrooks(2013)putit,“Asweacquiremoredata,wehavetheabilitytofindmany,manymorestatisticallysignificantcorrelations.Mostofthesecorrelationsarespurious and deceive us when we’re trying to understand a situation. Falsity growsexponentially themoredatawecollect.Thehaystackgetsbigger,but theneedlewearelookingforisstillburieddeepinside.”

Twoofthebestmeansofaddressingamassofcorrelations,mostofwhicharespuriousortrivial,employstrategiesthattendtobeignoredbybigdata,particularlybyitsbiggestboosters:theoryandhistory.Theoryistheexplanatorystorythatmakesthemostsenseofthedata.Nostorymakesperfectsensebecausethecomplexityofthedataandtheworlditrepresents can only be perfectly theorized by an explanation that is so general that itceasestobeuseful.Rather, thegoal is tofindatheorythat isbothgroundedinthedataandmakesreasonablesense.Somewouldarguethatthisrequirestheinclusionofanotherconceptroutinelyeschewedbybig-dataenthusiasts:causality.Itmakesmoresensetotestdataagainstacausalmodelthantoexpectdata,howeverlargeanddiversethecollection,tospeakforitself.Infact,itisdoubtfulthatthelatterispossiblebecause,inoroutsidethecloud, data is not an entity independent of human conception or contamination, but iscreated throughhuman intelligenceandpurpose,withallof their limitationsandbiases.Nevertheless,thechoiceisnotbetweencausaltheoryornotheoryatall.Anintermediatepositionisbuiltuponmutualconstitution,whichmaintainsthatconceptsanddata,theoryand evidence, construct or mutually constitute one another in an ongoing process ofbuilding an argument. Arguments are then tested against new data and alternativearguments.

There are other ways to constitute theory, but the point is that research of anyconsequence, including studies using large data sets, cannot do away with it. That isbecausetheconceptsexpressedinthedatapresumeatheoreticalperspective.AsBrooksexplained, “data is never raw; it’s always structured according to somebody’spredispositions and values. The end result looks disinterested, but, in reality, there arevaluechoicesall thewaythrough,fromconstructionto interpretation”(ibid.). Itmaybeambiguousorclear,weakor strong,butbyvirtueofournamingwhat iscollected,datadoesnotspeakforitself.Rather,wegiveitvoice.Nevertheless,oncewedoso,data,ifitis valuable, contains information that can speak to us, not by itself, but through thetheoreticalframethathelpedbringittolife.Thisistheessenceofmutualconstitution.Butit remains amessage slow toget through tobig-data enthusiasts.Fiveyears afterChrisAndersonproclaimedtheendoftheory,writersforWiredpersist,“Forscience, itmakessensetoseebigdataasarevolution.Algorithmswillspotpatternsandgeneratetheories,sothere’sadecreasingneedtoworryaboutinventingahypothesisfirstandthentestingitwithasampleofdata”(Steadman2013).

Inadditiontogivingtheoryinsufficientattention,bigdatatendstoneglectcontextand

history. That is partly because big data tends to examine behavior as a set of discreteeventsordatapoints.Again,Brooksoffered insight: “Humandecisionsarenotdiscreteevents.Theyareembedded in sequencesandcontexts.Thehumanbrainhasevolved toaccount for this reality. People are really good at telling stories that weave togethermultiple causes and multiple contexts. Data analysis is pretty bad at narrative andemergent thinking, and it cannotmatch the explanatory suppleness of even amediocrenovel” (Brooks2013).The fear is that theseeminglymagicalcombinationof largedatasets and massive computational power will lead people to replace narrative withcorrelationand,moreimportantly,toaskonlyormainlythosequestionsthatbigdatacanhandle. In the real world of history, if not in the metaphorical one of needles andhaystacks,contextcounts.Itisnotjusttheplacewheretruthorsolutionshide,butcontextactivelygivesshapeandsubstance to truth.Thisconclusion isofmore than“academic”value,asastudyofcommunicationtechnologyusedinurbandevelopmentdemonstrates.Thereare timeswhensimpleemailamongagroupofcommunity-minded individuals ismoreeffectiveatbringingabouttheresolutiontoacomplexpracticalissuethanthemostsophisticatedbig-dataanalysis(Applebaum2013).

Bigdataisincreasinglyusedinhistoricalresearch,tothepointthatanentirespecialty,cliodynamics,isincreasinglyappliedtoresearchlikethatcarriedoutattheUniversityofToronto to date medieval manuscripts by analyzing language and phrasing (Tilahun,Feuerverger, andGervers 2012). The specialty includes its own journal,Cliodynamics:TheJournalofTheoreticalandMathematicalHistory.Thepointisnotthatbigdatalacksusefulnessinhistoricalresearch,butratherthatitsuseislimitedand,unlessthisisclearlyunderstood, itwould be easy to extend themythmaking about the end of the scientificmethod and the end of theory and apply it to a putative end of history, or at least ofhistorical research, as we have known it. This is especially tempting when the majorsourceoffundingforhistoricalresearchisagovernmentprogramtomakehistoryanarmofthedigitalhumanities.Norisitjustamatteroftakinglargedatasetsandputtingtheminahistoricalcontext.Contextandhistoryarenotdiscretecontainersintowhichonecanobjectively insert data. They are fluid and require the experienced judgment of skilledprofessionalswhosesubjectivityisanassetthatenricheswhatweknow,notaliabilitytobesetaside.

“Atitscore,”accordingtotwoofitsleadingpromoters,“bigdataisaboutpredictions”(Mayer-Schönberger andCukier 2013, 11; italicsmine). It is hard to disagreewith thisconclusionandwiththefactthatitunderscoresboththepromiseandthedangerofrelyingonlargedatasets.TheabilitytomovebeyondtherandomsampletothebillionsofdatapointsthatGoogleusedtomakepredictionsaboutthespreadofthefluvirusiscertainlyattractive and, for some, compelling and revolutionary.But keep inmind that even thisprojectappearstohavehadashortpredictiveshelflife.Afterafewyearsofsuccess,theGooglemodelfellflatonitsfaceinthe2012–2013fluseason,grosslyoverestimatingthenumberofcases. It ishard tosaypreciselywhythishappened,butanalystspoint to theexpansionofnews-mediacoverageofthevirus’sspreadinDecemberandJanuary,whichled to far more Google searches using flu-related search terms than the company’salgorithm expected. In addition, the spike in coverage took place during the holidays,when people havemore time for both old and newmedia. It appears that peopleweresearchingmorenotbecausetheyhadflusymptoms,butbecausethemediasteppedupits

flucoverageat a timewhenpeoplewerepayingmoreattention tomedia.Whatever thecause,thedamagewasdone.AsGooglewipedtheeggfromitscorporateface,itpromisedtoimproveitsalgorithmtomakebetterpredictionsinthefuture(Butler2013;Poe2013).Thata similarmodelwasused for stock-market forecastingshouldcauseconcernaboutthe consequences of overconfidence in big data for the economy (Waters 2013b).Nevertheless, economists are confident, to the point of exuberance, that big data willtransformresearchandpolicymaking(EinavandLevin2013).Oneofthereasonsforthisenthusiasm is the potential analysts anticipate for using big data to better managetemporary,low-wagelabor.Asonereportsummedup,“Itisrearranginghowweallocatework—maybetoastateofpermanent,temporarywork,forthemostlynontechnicalranksoftheworkforce”(Hardy2013c).

Thecloudandbigdatacomewiththevisionofperfectingourknowledgeoftheworldifwecancollectmoreinformation,improvethesiftingforcorrelations,andcomeupwithjust the right refinements inmodels and algorithms.But itmay be that theworld is socomplexthattheloftyaspirationsofbig-dataenthusiastsareoutofreach.Perhapsitwouldbebettertoatleastsupplementbig-datastudieswithold-fashioneddepthinterviewsonacarefully selected sample. But diversifying methods is possible only when analystsapproachtheproblemwithopenmindsandtheskilltocarryoutresearchusingmultipleapproaches,ratherthanwiththeviewthatwehavediscoveredthekeytoarevolutionarytransformationinhowweacquireknowledge.

Given his considerable success in forecasting election results, one would not expectNateSilver to takeacriticalviewofbigdata.However, this ispreciselythepositionhesupportsinallofhiswriting,butespeciallyinTheSignalandtheNoise(2012),acarefullywrittenoverviewof thepotentialand theproblemsof large-scalestatisticalanalysisandprediction. For Silver, devotion to the statistical techniques and values of Bayesiananalysis means committing to probabilities over certainties and recognizing that allresearch is infusedwithbiases thatwecanrecognize, ifnoteliminate,and thenaccountfor them.Assume, hemaintains, that the complexity of theworld puts certainty out ofreach and one is likely to do a better job of approximating an accurate conclusion andmakereasonable,ifnotalwaysaccurate,predictions.Itisnotthesizeofthedataset,but,ashasbeenthecaseforas longaspeoplehavecarriedoutsocialresearch, theskillandhumilityoftheresearcherthatmostoftendeterminesuccess.

A good example of this point arose in 2013 when a doctoral student uncoveredsignificanterrorsinanacademicpaperthathasbeenusedbygovernmentpolicymakersand corporate decisionmakers to support strong economic-austeritymeasures bypublicauthorities around theworld. The article “Growth in a Time ofDebt” drew on severallargedatasetstoostensiblydemonstratethatwhentheratioofgovernmentdebttogrossdomesticproduct (GDP)exceeds90percent, themedianrateofeconomicgrowthdropsby1percentandtheaveragegrowthratebyconsiderablymore.The90percentthresholdapplied to bothdeveloped and emerging economies (Reinhart andRogoff 2010). If onecan speak of an academic finding going viral, this paper is a prime case in point. Theauthors,oneaneconomistwiththeNationalBureauofEconomicResearchinWashington,D.C., and the other atHarvard, achieved academic rock star status, including a lengthyNewYorkTimes profilewith the breathless headline, “TheyDidTheirHomework (800YearsofIt)”(Rampell2010).Itwaswrittenandtalkedaboutinalmosteverymajormedia

outlet.5Anotheracademicrockstar,thehistorianNiallFerguson,referredtoitas“thelawoffinance”(Konczai2013).Moreimportantly,policymakersusedthepaper topromoterigid austerity measures because it appeared to demonstrate that cutting governmentspendingwould reverse economic decline and spur growth. Thiswas a significant turnbecause, from the 1930s on, governments more or less believed that public spending,especiallyoninfrastructureandpublicworks,wouldspurgrowth,evenifitmeanttakingon debt. The new research demonstrated something fundamentally different: oncegovernment debt reached the magic ratio of 90 percent of GDP, the economy showssharplyslowergrowthrates.

Governments, corporations, and conservative think tanks jumped on the findings tosupport, implement, and justify cuts in government spending even as their economiessuffered from what some believed was inadequate spending. Even when governmentscontinuedtoexperienceeconomicrecession,double-dipandeventriple-dip,theirleadersheldfasttothemagicformula.Thenin2013,ThomasHerndon,adoctoralstudentattheUniversityofMassachusettswhohadnotyetbegunworkonhisowndissertationfoundsignificant errors in the original article’s data, thereby calling into question its centralfindings.As a commentator described, “One of the core empirical points providing theintellectual foundation for theglobalmove toausterity in theearly2010swasbasedonsomeoneaccidentallynotupdatingarowformulainExcel”(Wise2013).Thestudentwassimplytryingtoreplicatetheoriginalresultsforaneconometricsprojectandcouldnotdoso with publicly available documents, so he contacted the authors and asked for theirspreadsheets,whichtheyprovided.Hequicklyspottederrorsindatareportingonnationalgrowthratesanddebt levelsandpublished the results (Herndon,Ash,andPollin2013).Eventuallytheauthorsoftheoriginalpieceadmittedtotheerrors,butstoodtheirgroundonausteritypolicy.

Asonemightexpect, thedebate rages,withmostgovernmentscontinuing topracticeausterity even as they change their underlying justification (Vina and Kennedy 2013).Nevertheless,theimplicationsforbigdataaresignificant.Beforetheerrorsweredetected,critics such as Nobel laureate Paul Krugman raised a concern familiar to big-dataspecialists.TheReinhartandRogoffpaper,Krugmancomplained,usesbigdata todrawconclusionsbasedon correlations, not on causality: “All it does is look at a correlationbetween debt levels and growth. And since debt levels are not sharp extreme events,there’snogoodreasontobelievethatthey’reidentifyingacausalrelationship.Infact,thecase they highlight—the United States—practically screams spurious correlation: theyearsofhighdebtwerealsotheyearsimmediatelyfollowingWWII,whenthebigthinghappening in the economywas postwar demobilization,which naturally implied slowergrowth: Rosie the Riveter was going back to being a housewife” (Krugman 2010). Inadditiontoidentifyingthelimitationsofcorrelationalanalysis,thecaserevealsthat,byitsnature, big data can create big problems. First, errors in entering data in key cells cancreate significant changes throughout the analysis, amplifying the consequences of theoriginalerrors. In thiscase, errors led toapowerful findingcongenial topolicymakersandcorporate leaderspredisposed toausterity,which turnedout tobe,at thevery least,grosslyexaggerated.Second,thesizeofthedatasetsmakesitdifficultforpeerandotherreviewerstocatcherrors.It isnotcommonforreviewerstohaveaccesstooriginaldatainputs, and certainly not in the case of data sets with multiple variables spanning

numerousnationsand timeperiods. In this case, if itwerenot for theworkof ahighlymotivateddoctoralstudent,itisunlikelythattheerrorswouldhavebeencaught,andthepaper would have retained its stature as the intellectual cornerstone for pro-austeritypolicies.Bigdatacancontainandmaskbigerrorswithbigconsequences.Asonebusinesseducator concluded, “Don’t getmewrong:Data is critical. But history suggests that itplaystricksonourabilitytoobjectivelyunderstandallofthevariablesthatareatplayintheworld.Sobecareful:Althoughmanyprofessionalstellyouthatthedataisonlyoneofmany decision points, I have found that too many people rely too heavily on itsinformation.Butaswehaveseen,thedatacanlie!”(Langer2013).

CloudCultureThe technical criticisms directed at big data’s singular reliance on quantification andcorrelation, and its neglect of theory, history, and context, can help to improve theapproach, and perhaps research in general—certainly more than the all-too-commonattempts to fetishize big data. But big data is more than just a methodological tool. Itpromotesaveryspecificwayofknowingthat,whenconnectedtotheglobalexpansionofcloud computing, has significant implications. Specifically, cloud computing provides apowerful technologicalgrounding to supportbigdata’sdigitalpositivism or the specificbeliefthatthedata,suitablycircumscribedbyquantity,correlation,andalgorithm,will,infact,speaktous.Theabilitytoprocessbillionsofdatapointsinthecloud,inthetimethatittakestoreadthissentence,helpedtolegitimizeGoogle’sflu-virusproject,asitdoessomany other big-data projects. The cloudmay be central to amyth but, in this as in somanyothercases,mythsmatter.Itisthereforeimportanttocritiquethecloudasaculturalforcebecauseitisnotjustamethod;itisacompletewayofknowingthat,ifleftwithoutseriouscriticalreflection,willcrowdoutotherlegitimatepathstounderstanding.

Thecloudisanenormouslypowerfulmetaphor,arguablythemostimportantdevelopedintheshorthistoryoftheITworld.Assuch,itssignificancefaroutweighstheaccuratebutbanal rootsof the term in thecloudnetworkdiagramsproducedby telecommunicationsspecialists.Namingit thecloudtapsintoarichliteraryanddiscursivehistorythattermslike cyberspace, Internet, and even theweb fail tomatch. By its nature, culture resistsessentialismsofall types, including the tendency in thedigitalworld,nowembodied incloudcomputing,toreducethecloudtoaninformationrepositoryandthefoundationforthedigitalpositivismofbig-dataanalysis.Thereismoretothemetaphorofthecloudthanits crudely rendered image in the network diagrams that gave rise to the term cloudcomputing.Contrastthisimage,whichlooksasifdrawnbyachild,withtheeerie,cloud-filledpaintingTheEmpireofLightbytheiconofsurrealismRenéMagritte.Thepaintingfeaturesthebrightblueofadaytimeskyfilledwithpuffywhitecloudsthatoverseearowof houses in nighttime darkness. Unlike the cloud-computing diagram, which uses theimageofthecloudtonaturalizethetechnology,Magritte’sjarringclashofrowhousesindarkness under the bright clouds and blue sky of daytime suggests that something isseriouslyawryinthecloudsandontheground.

Cloudsareamongthemostevocativeimagesinthehistoryofculturebecausetheyhavebeenadailypartofthelivesofeveryonewhohaseverlived.Itisnosurprise,therefore,that cloud gazing to search for symbols and signs, known as nephelococcygia, is anancientart.Cloudsarealsorichlyevocativebecausetheytakeonanalmostinfinitevarietyof designs, providing, formany, an early introduction to form and towhat itmeans totransformoneshapeintoanother.Thealtocumulusfillstheskywithgiantcottonballs,thecirrocumuluswithpatchesofrice,andtheundulatuswithcelestialripplesofsand.Thesebenign images disappear when an arcus formation signals the leading edge of anoncomingstormorwhenatubashootsoutofdarkcloudstocreateawaterspoutoverabody of water (Pretor-Pinney 2011). Clouds aremore than cultural evocations becausetheyreplenishtheresourcethatisabsolutelyessentialtosustainlife,leadingsorcerersandscientistsoverthemillenniatoapplytheirparticulartalentstoconjurerain-bearingclouds.In this respect, the cloud is transcendent because it knows all time and all space, andoverseeseveryformoforganiclife.

Itisnowonderthatcloudshavearichhistoryinpracticallyallcultures,andtheWestiscertainly no exception. It is the perfect metaphor for today’s computing, whose globalnetworkof24/7datacenterslinkedtotelecommunicationssystemsandsmartdevicesalsotranscendsspaceand timeand, justas realcloudsproduce rain, showersa resource thatmany consider absolutely essential for today’s world: knowledge. Certainly a literalistmightpointtothevaporintheskyandthegiantcementwarehousesonearthanddeclarenoconnectionbetweenthetwo.6Butthatwouldmisstherichmetaphoricallinksthatgivebothatouchofthedivine.Wemarvelatcloudsintheskybecausetheyareeverpresentandyetinfinitelydiverse.Theyareassociatedwithsublimebeneficencefortheraintheybringandwith sublime terrorwhen theywithhold itorbringdestruction in the formoflightning,tornados,andfloods.Theirtechnologicalcounterparts,thevastdatafactoriesinthe fields,provideacloudofknowing,a systemofubiquitous, infinite information thatwasonce reservedfor thedivineand, sincehumankind’sbanishment fromparadise,hasbeendeniedtoall.

Evenintheirliteraldifferences,theimageofthecloudprovidesaglossoncomputing.First, the clouds of vapor in the sky soften the hard-edged data center by giving cloudcomputinganetherealquality.Thecloudistheplaceofnoplace;thehomeofdatastoredand processed everywhere and nowhere.Moreover, the image of the cloud naturalizescomputing,coveringitwiththeauraofanorganicprocessthattranscends,toadegree,thephysical presence of the data center as a blot on the landscape and an energy hog.Admittedly, there are dark clouds that can cause damage andwe oftenwish the cloudswoulddispersetorevealthecherishedbluesky.Butwealsoknowthattheseareallnaturalprocesses, part of the eternal cycle of nature, whose extension to the cloud makescomputingappearnaturalaswell. It is rare forclouds to inspire significant reprobation.Rather,thereisaCloudAppreciationSociety,and,forthosewhoprefercloudstobirds,aCloudCollectorsHandbook that enables people to chart and chronicle the varieties ofclouds they have observed. Clouds are embraced by romantic poets like Shelley andWordsworthforgivinglife,forcontributingtonature’srhythmiccycles,andforpointingthewaytothesublimevisionsthatserveupalifetimeofrewards.What’snottolikeaboutthecloud?

There is more to the metaphor of the cloud than capturing the sublimity of cloudcomputing. In its rich history, themetaphor contains a critique that challenges utopianvisions finding transcendence, if not the divine, in new technology. Considering itsubiquitouspresenceandpersistencethroughouttime,itisnosurprisetofindthecloudinmanyexpressionsofthehumanimagination.Thewrittenword,music,andthevisualartswouldbemuchpoorerwithoutthemetaphoricalcloud.Fromthebroadsweepofthecloudinculture,IhavechosenthreeexemplarsfromvastlydifferentperiodsinWesternsocietytodocumentantimoniesbetweenthemetaphorandtheinformationtechnologythatwouldadopt it. It begins with The Clouds, a comedy written by Aristophanes that satirizedintellectuallifeinfifth-centuryBCGreece.Next,wemovetothefourteenthcenturyADandTheCloudofUnknowing,aspiritualguidetolifewrittenbyanoldermonktoprovideadvice to ayoungmanwhohas recently joined themonastery.Finally, I takeupDavidMitchell’s masterful contemporary novel Cloud Atlas, which tells six interconnectedstoriesthatspanhumanhistoryacrosstheworld.

Therearemanyotherexamplesfromtheculturalhistoryof thecloud thatcouldhave

servedaswell.CloudsfillthenaturalandmythicimageryofHomer’sIliad,suggestingthedualityofnature’spastoralbeautyandthegods’interestinthedarkcloudsofwar.Giotto’sthirteenth-centuryfrescointheBasilicaofSaintFrancisofAssisicontainsadevilhiddenintheclouds,depictingthesceneonearthandinheavenatthetimeofthesaint’sdeath.For the great artist, even a setting of celestial majesty includes a nod to the Prince ofDarkness.Theaward-winningwriterAnnieProulxtitledher2011evocative“memoirofplace”BirdCloudbecause,onherfirstvisittothevastWyomingwetlandandprairiethatwould become her home, a bird-shaped cloud greeted her in the sky at dusk. For thewriter,itwasasigntosettlethereandanintimationoftherichandseeminglyever-presentbirdlifeinthearea.Therearemanyotherpotentialexamples,andsomewillmakeabriefappearance, but the three I have chosen enjoy the advantage of covering a significantswatheofWesternhistory,representthreedifferentformsofthewrittenword,and,moreimportantly,speakevocatively,ifmetaphorically,aboutthedeepersignificanceandthreatsrepresentedbycloudcomputingandbigdata.OthersmoreexpertthanIcansurelythinkof examples from music and the arts, and from outside the world of the Westernhumanities.7

TheWisdomoftheCloudsEventhoughitwaspannedbycriticsandforcedintorewriteswhenfirstperformedin423BC,itishardtooverestimatetheimportanceofAristophanes’sTheCloudsforliterature,for the history of ideas, and for today’s debates about what knowledge means in aninformation society. After 2,500 years, it remains a model for what Eve Smith calls“comedy as social conscience” (Smith 2013).Remarkably, the play accomplishes all ofthis through a satire that lampoons Socrates, one of themost venerated thinkers in thehistoryoftheworldand,inthemindsofsome,amartyrtohisbeliefs.TheplotcentersonStrepsiades, aonce-prosperousmannowsaddledwithdebts,whoplans togetout fromunderthembysendinghisslackersonPheidippidestotheThinkery,thefictitiousschoolestablishedbySocrates that teacheshow towinanargumentnomatterhowweakyourposition.Or as Strepsiades describes it to his son, “There they prove thatwe are coalsenclosedonallsidesunderavastsnuffer,whichis thesky.Ifwellpaid, thesemenalsoteach one how to gain lawsuits, whether they be just or not.” The play turns the greatphilosopher into aDale Carnegie,whose classic book on public relations,How toWinFriends and Influence People, became a marketing bible on publication in 1936. TheClouds is the name for the play’s chorus, which rises out of the oceans to live in theheavens,surveyingtheworldwithapanopticgazeand,whenproperlysummoned,shareits deep knowledge and clever rhetoric with earthly mortals. When Strepsiades’s sonprovesmoreslackerthangeek,StrepsiadesdecidestoenrollhimselfintheThinkeryafterconsultingastudentattheschoolwhoboastsabouttheresearchledbySocrates,including“Howmany times the length of its legs does a flea jump,”which led to a new unit ofmeasurement,thefleafoot;thesourceofagnat’sbuzz:itstrumpet-shapedanus;andthesophisticateduseofcompasses todefeata lizardthat interrupteda“sublimethought”ofthe great philosopher who gazes up to the heavens just in time to receive one of thelizard’snot-so-sublimedroppings.Isthisscienceoruselesstrivia?Whatevertheanswer,and it is clear where the playwright stands, the debate certainly resonates in a worldcharacterizedbyanapparentglutofinformation(Andrejevic2013).

Rather than flee the seemingly crazed Thinkery, Strepsiades is more convinced thanever that Socrates can rescue him, although it is unclear whether this is because hebelieves Socrates is a great thinker or such a masterful conman that he can convincepeopletopraisehistrivialresearch.Itdoesnotmattertothewould-bestudentbecausehesimply needs the rhetorical skill to win over debt-holders. At their first meetingStrepsiades meets Socrates, who summons the Clouds for counsel with sacrificialofferingsandhissignatureoratoricalskill:“WhetheryouberestingonthesacredsummitsofOlympus, crownedwith hoar-frost, or tarrying in the gardens ofOcean, your father,forming sacred choruseswith theNymphs;whether you be gathering thewaves of theNileingoldenvasesordwellingintheMaeoticmarshoronthesnowyrocksofMimas,hearkentomyprayerandacceptmyoffering.”ThesummonsissuccessfulandtheCloudchorusimmediatelyrevealsitssardoniccharacterbygreetingSocratesasthe“greathigh-priest of subtle nonsense.” Chiding the philosopher for putting rhetoric ahead ofknowledge, the chorus demonstrates its own rhetorical skill, promising the desperateStrepsiades,“Clientswillbeeverlastinglybesiegingyourdoorincrowds,burningtogetatyou,toexplaintheirbusinesstoyouandtoconsultyouabouttheirsuits,which,inreturnforyourability,willbringyouingreatsums.”Unfortunatelyforhim,Strepsiadesproves

tobeapoor student.Perhapshis agehasgivenhim toomuchexperience,wisdom,andcharactertoacceptaneducationthatvaluestriviaandrhetoric.Orperhapsheis justnotsuitedtotheesotericmethodsSocratesapplies.

Orderedtoacouchandcoveredinablankettoencourageself-reflection,theboredoldman instead decides to masturbate. Having failed to learn from Socrates, Strepsiadesreturnstohisson,who,perhapstooyoungtocareaboutwhetherheisofferedwisdomortrivia,knowledgeorrhetoric,agreesthistimetobeamodelstudent.Socratesstepsasideandinstructionistakenoverbytwofigures:onewhostressescreatingargumentsbasedonknowledge, the other onmanipulating peoplewith rhetoric. The latterwins and, armedwith theskillsofa sharp talker,Pheidippidessaves theday forhis fatherbydismissingwith his now-dazzling rhetoric those to whom his dad owes money. Unfortunately forStrepsiades,Socraticeducationmakeshissonarrogant to thepointofbeatinghis fatherand threatening his mother. He even manages to mount a convincing defense of hisviolence,whatforAristophanesisthetruetestofhissuccessfultransformationunderthegreat philosopher. This leavesDad tomoan, “Oh!whatmadness! I had lostmy reasonwhenI threwover thegodsthroughSocrates’seductivephrases.”TheCloudchorushaslittle sympathy for Strepsiades: “Here is a perverse old man, who wants to cheat hiscreditors; but some mishap, which will speedily punish this rogue for his shamefulschemings,cannotfailtoovertakehimfromtoday.Foralongtimehehasbeenburningtohave his son knowhow to fight against all justice and right and to gain even themostiniquitous causes against his adversaries every one. I think this wish is going to befulfilled.Butmayhap,mayhap,will he soonwishhis sonweredumb rather!”TheplayendswiththeoldmanclimbingtotheroofoftheThinkerytoripitapartandburnitdown,getting in one last jab at the great philosopher.When someone demands to knowwhatStrepsiadesisupto,heanswers,Socratically,“Iamenteringonasubtleargumentwiththebeamsofthehouse.”

TheCloudsisnearly2,500yearsoldyetremainsbothhilariousandremarkablymodern.WhentheCloudchorusstepsoutofitsroleasacelestialsourceofwisdomtopleadwiththe audience to “like” thisnewversionof aplay that first opened toweak reviewsandthenreturnstocharacter,onecannothelpbutthinkofthenarratorsinThorntonWilder’sOurTownandTheSkinofOurTeeth,whomoveeffortlessly throughdramatic timeandspace. But for our purposesTheClouds speaksmost powerfully across two and a halfmillennia to a world of new clouds that would also revise themeaning of knowledge.Consider their first words in response to Socrates’s summons: “Eternal Clouds, let usappear; letus arise from the roaringdepths ofOcean, our father; let us fly towards theloftymountains,spreadourdampwingsovertheirforest-ladensummits,whencewewilldominatethedistantvalleys,theharvestfedbythesacredearth,themurmurofthedivinestreamsandtheresoundingwavesofthesea,whichtheunwearyingorblightsupwithitsglitteringbeams.Butletusshakeofftherainyfogs,whichhideourimmortalbeautyandsweeptheearthfromafarwithourgaze.”Aristophanes’smetaphorofthechorusrisingoutof the oceans to become cloud-filled sky is appropriate to themodern cloud because itoffers a way of envisioning through discourse the panoptic knowledge that is bothinformationandmeansofsurveillancelookingoutontheworldandinterveningtomodifythought and behavior. Aristophanes sends a warning flare across the bow of cloudcomputing.There is no separatingknowledge frompower, ubiquitous information from

ubiquitoussurveillance.

ForTheClouds, the key ontological tension is not between knowledge and data, butrather between reason and rhetoric. They are viewed as different because reason,whatAristophanes calls in the play “just discourse,” advances, as its character states, “bypresentingwhatistrue.”Rhetoric,ontheotherhand,describedwithoutsubtletyas“unjustdiscourse,”isaspindoctor,twistingthetruthwithskillfullyconstructedfabricationsthatcarrytheday.TheCloudchorus,itturnsout,isoftwominds,atfirstappearingtoapproveof the outcome, but later admitting that rhetoricwas only permitted towin in order toteach Strepsiades a lesson: those seeking a shortcut to success will themselves be cutshort. Here Aristophanes warns against the seductive power of dazzling languagemasquerading as the wisdom of the clouds. There is a fine line between reason andrhetoric,truthandspin,knowledgeandpublicity.Thewayofknowingestablished2,500yearsagocomesnotintheformofthephilosopherking—suchafigurewasjustaPlatonicaspiration. Rather it is the philosopher-trickster, the intellectual-spin doctor whodominates with knowledge and rhetoric both mutually constituting and mutuallycontaminating.IntheWesternwayofknowing,thereisnopuretruthstoredandprocessedinthecloud;thereisjusttheongoingstrugglebetweenreasonandrhetoric,somethingthatthe contemporary philosopher-trickster Bruno Latour recognizes in his restaging of thedebatesbetweenSocratesandtheSophistsinthemasterfulPandora’sHope(1999).BeforeleavingtheworldofAristophanes’sAthensforamedievalmonasteryandThe

CloudofUnknowing,itisworthnotingtwoadditionaltellingpointsofrecognition.Todaythereisagreatdealofattentionpaidtothemythofyouthandnewtechnology,whichismade to mock and shame older men and women who are caricatured as laughablyunskilledintheworldofinformationtechnology.Instead,itistheyoung,unburdenedbytheweightofyears,whoarenaturallyadeptatmasteringsmartdevicesand,unliketheirelders, appreciate thewisdom of the cloud. I have elsewhere described thisworship ofyouth in thehistoryof technology, fromstoriesofheroicyoung telegraphkeyoperatorsthroughtalesoftheamateurradioboyswhosebraverysavedthedayforsailorsatseaandothers indistress, to thegarage-shopwizardsofcyberspacewhomake their firstbillionbefore thirty and, as in the filmWar Games, save the world from nuclear holocaust(Mosco 2004). For those who buy into all or part of this myth, Aristophanes has adifferent tale to tell.Although no one is spared his satirical darts, the playwright savessomeofhissharpestbarbsfortheyoungPheidippides,whoistransformedfromaslacker,too lazy to help his family by attending the Thinkery, to a button-downed geek andslippery con artist. Sure, his father is no prize either, but at leastDad comes around tounderstandjusthowfoolishhewas.Armedwithhisnewpowers,Pheidippidesisreadytotakeontheworldtothepointofjustifyingassaultsonhisparents:“Howpleasantitistoknow these clever new inventions and to be able to defy the established laws!When Ithoughtonlyabouthorses,Iwasnotabletostringthreewordstogetherwithoutamistake,butnowthatthemasterhasalteredandimprovedmeandthatIliveinthisworldofsubtlethought,ofreasoningandofmeditation,IcountonbeingabletoprovesatisfactorilythatIhavedonewelltothrashmyfather.”Perhaps,theplaysuggests,wisdomiswastedontheyoung.Finally, there is theThinkery,amisnomer if thereeverwasone,aplaceof rankpositivism(whatis therelationshipbetweenthelengthofaflea’sleganditscapacitytojump?)andrhetoricalgobbledygook.Justbecauseaninstitutionbearsthenameofthought

does not guarantee the delivery ofwisdom. Two and a halfmillennia later, it is worthremindingourselvesthatneitherdoestheterabytecapacityofadatacenter.

CloudsGetinOurWayFourteenth-centuryresidentsoftheBritishIsleslivedinfearoftheblackshilling.Thisisareferenceto thedarkcircularswelling thatappearedin thearmpitorgroinsignaling thepresenceofthebubonicplagueandthelikelihoodthatdeathwasnear.Inthelatterpartofthat century, half the population of England disappeared following the arrival of thecirculardisk,astarkreminderthatthelatemedievalperiodmeantfarmorethanlordsandladies.AsiftheBlackPlaguewerenotenough,thecountrywasinaconstantstateofwarwith France. In fact, the so-calledHundredYears’War lasted formore than a century.Smallwonderthatwhenanewpolltaxwasimposedonthepeasantry,itrespondedwithasocialupheaval that swept throughseveralcountriesand terrified theauthorities.Outofthis dark and tumultuous setting, an anonymous religiousman (one suspects he was apriest or monk) produced a guide for a young monastery initiate called The Cloud ofUnknowing(Anonymous2009).Itwasnotunusual to findmonasteries inpre-ReformationEngland, including several

committedtoamysticismwetendtoassociatetodaywithEasternreligioustraditionslikeBuddhism.OutofthisEnglishmonastictraditionandtheupheavalsofthetimecamethework ofWalter Hilton, Julian of Norwich, and the anonymous author of a manuscriptwritteninthecolloquialMiddleEnglishofthetime;thatworkrevealsawayofknowingandametaphorforthecloudthatprovidesadistinctalternativetothedigitalpositivismofbigdataandcloudcomputing.TheircounterpartsincontinentalEuropeincludedasetofremarkable women, such as Gertrude the Great, Catherine of Siena, and MargueritePorete.Ascloudcomputing’swayofknowingcrowdsoutothersand,indeed,takesonthecharacteristicsofasingularity,oratleastofthehegemonicdiscourseofdigitalpositivism,itisessentialtorecallalternatives—attheveryleast,toconsiderwhatisbeinglostandtomore fullycomprehend thebroader significanceof today’scloud.For theauthorofTheCloudofUnknowing,thatcloudisametaphorfortheeverydaybitsofdataandexperiencethat make it difficult to achieve genuine wisdom and for oneness with God. Suchachievements are possible only by setting aside life’s banalities and, throughcontemplationandmeditation,concentrating themindandspiriton the lightbeyond thecloud.

ThereisnomaskingthereligiousnatureofTheCloudofUnknowing.ItspurposeistoteachayoungmonkandthewiderreadershipofthetimehowtoreachGod.Althoughitmightappearunusualtothoseunfamiliarwiththeliteratureonthecultureofinformationtechnology, as Franklin (2012) argues, “analogies with divine bodies persist withsurprisingregularityinanalysesofdigitaltechnology”(445).KevinKelly,cofounderandformerexecutiveeditorofWiredmagazine,wasnotthefirst,orthelast,whenhedeclaredin 2002 that “God is theMachine” in an article exploring “the transcendent power ofdigital computation” (2002). The emergence of the Internet sent gurus in search of itssublime origins and several, including former vice presidentAlGore, the novelistTomWolfe,andwebauthoritieslikeErikDavis(1998)andMarkDery(1996),founditinthework of the Jesuit priest Pierre Teilhard de Chardin. The United Nations sponsored aconference on his work and, in a characteristic burst of gushing enthusiasm, WiredmagazineproclaimedthattheJesuitpriest“sawtheNetcomingmorethanhalfacenturybefore it arrived” (Kreisberg 1995).Teilhard’swork remains popular today, particularlyfor his core concept of the noosphere, which he thought of as the mental space

surrounding theearth inanatmosphereof thought (noos inancientGreekmeansmind),andwhichhasevenreceivedacontemporaryspellingastheknowosphere(Revken2012).

The Jesuit priest’s work appeals to a sublime vision of transcendence throughknowledge.Specifically, asTeilharddescribes it in hismajorworkThePhenomenonofMan (1961), inaddition to theatmospheresurroundingourearthandmaking lifeasweknow it possible,weare also encircledbyanoosphereor sphereof thought that growsthickerandmorepowerfulwiththeworld’sacceleratingproductionofinformation.Asthebiologist and anthropologist David Sloan Wilson described, “As a new evolutionaryprocess, however, our origin was almost as momentous as the origin of life. Teilhardcalled thehuman-createdworld thenoosphere,whichslowlyspread likeaskinover theplanet, like the biological skin (the biosphere) that preceded it.He imagined ‘grains ofthought’coalescingatever-largerscalesuntil theybecameasingleglobalconsciousnessthat he called the Omega Point” (Revken 2012). For some early and current cyber-enthusiasts,Teilhard’sworkreaffirmedtheircommitmenttoprogressthroughknowledge,toavisionofevolutionthatextendedbeyondDarwintotherealmofpurethought,andtotheirbeliefthattheinformationagewasmorethanaconvenientmarkerforthelateststepfromtheagriculturalandindustrialstagesofhumandevelopment.Intheirview,itwasawatershed in human, organic, and cosmic evolution. More than a new means ofproduction, the computer andother information technologieswerekeys to a posthumanworld.OursisnotjustanAge;itisaMission.

Teilhard’s popularity is both understandable and puzzling. One can certainly see theattractiontosomeonewhobelieveswithreligiouszealthatinformationtechnologyisthekey toprogress. It isall themoresignificant thathismajorworkappeared in the1930sand ’40s, well before the personal computer and the Internet.8 Nevertheless, the Jesuitpriest was steeped in controversy that remains today. His work as an archeologist wasquestionedashewaseitheraperpetratororavictimofthehoaxdiscoveryofthePiltdownMan, one of many fraudulent “missing links” that appeared in the twentieth century.Moreover, hiswriting got him into continuous hotwaterwith religious authoritieswhowondered what the noosphere, a term they knew to have come from the work of thenineteenth-centuryRussian scientistVladimirVernadsky,had todowithCatholicismorevenChristianity.Afterall,VernadskywasfavoredbyStalin,whoawardedhimtheStalinprize in science in1943.Andyet,Teilhard’sworkappears topreviewsomuchofwhatcomprises the cornerstone of current myths about the information age and now cloudcomputing. It speaks to those who see communication visionary Marshall McLuhan’s(1989) imageof informationas theglobalnervoussystemof thehumanrace,whoviewcomputerguruRayKurzweil’sconceptionofanetworkedworldapproachingthedreamofimmortalityinwhathecalledtheageofspiritualmachines,andwhoseeinthesemachinesnot just the instruments to creatematerial abundance, but thekey to salvation.Teilhardcreated the spiritual foundation for what might best be called a cloud of knowing,something that is conjured with each new IBM commercial hymn to its SmartCloud.Kurzweilinspiredaquasi-religiousreadingofinformationtechnologywithhisargumentsfor a computerized version of immortality, as science develops the capacity to save theessenceofanindividual’sintelligenceandspiritinastoragedevice.Relatedtothisishiswork on the “singularity” or what amounts to a technological superintelligence, whichKurzweilbelievesisachievableinafewdecades.Italsobearsastrikingresemblanceto

Teilhard’sreligiouslyinspirednoosphere(Kurzweil2005).

TheCloudofUnknowingwasmeantasaspiritualguidetoalifethataspirestoonenesswithGod.Butitcanalsobereadasaseculartextwiththesupernaturalunderstoodasametaphor for the perfectmachine, the perfect algorithm, or thewisdomderived from arichunderstandingofknowledgeandinformationmadepossiblebytechnologiessuchascloud computing. To my knowledge no one has addressed The Cloud of Unknowingthrough the lens of information technology. That is understandable because, unlike thecloudsofTeilhard’snoosphereorKurzweil’ssingularity,thosefeaturedbytheanonymousauthor of The Cloud of Unknowing, although substantially the same in content, areanythingbut the sublimegateway tocosmicevolutionor thekey to theageof spiritualmachines. The cloud of that anonymous writing is associated with the data, facts,information,anddetailsthatcompriselife’sdiscursivebanalities,whatwemightcallbigdata stored in the cloud or the haystacks that surround the prized needles that datascientists discover.But for the fourteenth-century teacher, the clouds of information, soattractivetoday,onlygetinthewayoflife’spurpose.Forthatwork’sauthor,life’spurposewas to discover true knowledge ofGod; for a secularworld it signifies how clouds ofinformation get in theway of truth. ForTeilhard,Kurzweil, and any defender of cloudcomputingandbigdata,thepathtoknowledge,ifnottowisdomandthesingularity,istocreatemoredata, analyze it, anddrawconclusionsandpredictions.For themmoredataandinformationleadtomoreknowledge,betterpredictions,andabetterworld.

Forourfourteenth-centurywriter,pursuingthecloudisnotthekeytowisdom;itgetsinthe way of wisdom. Instead, he concludes, it is essential to systematically purge thebanalities of life, including the many bits of data, information, and knowledge (“allcreated things, material and spiritual,” 19) that literally cloud the truth. Given howdifficult it is, even for people of the fourteenth century, to carry out this project, hedescribesthepracticesofcontemplationandmeditationthatmakeitpossibletoovercomethecloudofunknowing:“Secularorreligious,ifyourmindisinflatedbyprideorseducedbyworldlypleasures,positions,andhonors,orifyoucravewealth,status,andtheflatteryofothers,ourGod-givenabilitytoreasonisservingevil”(27).Toknowrequiresactsofunknowing.Itisdifficultforthemodernmind,whichistrainedtoviewmoreasbetter,tograsp this perspective. For the secular-minded, the bigger the cloud (the data set or thehaystack),themorelikelywewillsolvetheworld’sproblems.ForthosewhogivesupporttowhatNoble(1997)called“thereligionoftechnology,”whetherthismeansKellyseeingGodinthemachine,Teilhardenvisioninganoosphere,orKurzweilanticipatinganageofspiritualmachines,thegrowthofthecloudisanessentialpartofhumandestiny,astepintheprocessofevolution.Giventheseviewsandothersamong technologicalenthusiasts,the religiousnature ofTheCloudofUnknowing appears to be less problematic than itsepistemologyorwayofknowingbyunknowing.

Nevertheless,therevivalinthebook’spopularityandtheinterestinarangeofreligiousandnonreligiousmeditationpracticessuggestthatevenitsepistemologyisnotsofarofftheradarofcontemporarythinking.The2009translationfromtheMiddleEnglish,withalongintroductoryessayintheeditionusedforthisbook,suggeststhatthereiscontinuinginterest in thework.A1973editionbenefited from the riseof the1960scountercultureandespeciallyitsinterestinalternativewaysofknowing,apointtowhichitsintroductionby the renowned religious scholar Huston Smith alludes. One of the most important

novelistsofourtime,DonDeLillo,makesuseofTheCloudofUnknowing intwoofhisbest-knownworks. In 1985’sWhiteNoise,which traces the spread of a toxic cloud, healludestoachildas“acloudofunknowing”(290).Becausechildrendonotknowdeath,theyareopentomoreoftheworldthanadults,whopresumablyseeinlifeitsinevitabledemise. In the faceof the inexplicable impact the“airborne toxicevent”hasbrought tosunsets,peoplearereducedtoasublimefeelingofchildlikeunknowing:“Thereisawe,itis all awe, it transcendspreviouscategoriesof awe,butwedon’tknowwhetherwearewatchinginwonderordread,wedon’tknowwhatwearewatchingorwhatitmeans,wedon’t knowwhether it is permanent, a level of experience to which we will graduallyadjust, intowhichouruncertaintywilleventuallybeabsorbed,orjustsomeatmosphericweirdness,soontopass”(324–325).Thesunsetvision,broughtaboutbytechnologyrunamok, brings a strange serenity, despite “men inMylex suits…gathering their terribledata.”DeLillogoeson,“Nooneplaysaradioorspeaksinavoicethatismuchaboveawhisper.Somethinggoldenfalls,asoftnessdeliveredtotheair”(325).Moreimportantly,in hiswidely recognizedmasterpiece, 1998’sUnderworld, DeLillo uses the fourteenth-centurybookasthetitleandleitmotifforoneofthesixpartsofhisepicnovel,havingthemaincharacterNickShaydescribeitscontentsinthemidstoflovemakingwithawomanhe has recently met. No amount of knowledge, Shay maintains, can comprehend thenegationwe callGod. It is only by engaging in our own forms of unknowing that thisbeginstobepossible.Therearenumerousotherreferencesfromculturalicons,includingSomersetMaugham(TheRazor’sEdge),J.D.Salinger(FrannyandZooey),andLeonardCohen(inhissong“TheWindow”).

Thevisionofknowingthroughunknowingappearsincontemporaryworkthatdoesnotmentionthebookatall.Considera2012essaybythewell-knownnovelistZadieSmith,inwhichshecomparesherbroadknowledgeofthewrittenwordwithwhatisforherasadlackofmusicalknowledge(2012).How,shewonders,didshegofromanearlyexperienceofhatingtheworkoffolksingerJoniMitchellonlytocometoloveitmanyyears later?Smithisbaffledbecausesheultimatelycametotreatthefolksinger’smusicasasublime,rapturous experience, saying “it undid me completely,” a feeling that she has notexperiencedintheworkofherchosenprofession.Sheconcludesthatitmayhaveresultedfromanexperienceofunknowing:“acertainkindof ignorancewas thecondition.”Intothispureignorance,this“non-knowledge,”somethingsublime,perhapsanevent,beyondor beneath the threshold of awareness, made the shift in her sensibility possible. SheknowsandlovesMitchell’sworkwithanunexpecteddepthbecauseshedidnotknowit,or much of anything about music, before. Unlike her knowledge of fiction, which hasaccruedfromyearsof incrementaladditions toherowncloudofconsciousness,Smith’sknowledge of music followed an epistemological break made possible not by small,consistentadditionstoadatabase,butbyyearsofwillfulunknowing.9

It is unlikely that thewriter ofTheCloudofUnknowingwill join the ranks of thosewho, likeTeilhard,arehailed forpredicting the Internetandnowthecloud,wellbeforetheirtime.Butperhapsheshould,ifonlybecausethemedievalteacherofferedagenuinealternativetowhatwouldbecomeadominantwayofknowingintheWestthatthreatenstooverwhelmchallengestothecloud,bigdata,andthedigitalpositivismtheypromote.

AnAtlasofCloudsAristophanes’splaydemonstrates that at oneof the earliest points inWestern literature,there was already serious concern about the arrogance that comes from excessiveconfidenceintheabilitytoknowtheworldthroughanarrowpositivismandtheeasewithwhich we canmake a fetish of information. For the unnamed author of The Cloud ofUnknowing, thedangerliesinbeingoverwhelmedbyinformation, thebanalbitsofdataanddiscourse that literally cloudour vision andkeep us fromachieving transcendence.DavidMitchell’s2004novelCloudAtlasbeginswithaseemingoxymoronandchallengesbasicconceptionsoftime,space,andinformation.

Howcanoneevenconceiveofanatlasofclouds?Afterall,anatlasprovidesamapofrelatively stable forms, like land masses and bodies of water.We think of an atlas asmapping theworld, the nation, the universe, or perhaps the city, but not themasses ofquick-movingvaporthatdartabouttheskyandchangeshapeintheblinkofaneye.Wedogive themnamesandsomepeoplekeepa recordofcommonand rare forms, justasdobirders.Buttherearefarfewerpeoplewho“collect”cloudsforalifelistthanthosewhogoinsearchoffeatheredcreatures,atestamenttojusthowstrangeitistocaptureclouds,by whatever means. Because of their inherent ambiguity, clouds lend themselves tosubjectivity and sowe aremore likely to use poetry than an atlas to describe them.Ofcoursethereisascienceofcloudsonwhichmanyaweatherforecastrisesandfalls.Butwehave tended to leave theirdescription to thosewhoconjuresublime images, suchasWilliamWordsworth, who writes that after wandering “lonely as a cloud,” the viewercomes upon a field of “golden daffodils” that forever appear in the “inward eye” toprovideasourceofpleasurein“blissfulsolitude.”Thekeytoalifetimeofsuchjoyis,forthepoet,tobecomeacloud.OronethinksofPercyByssheShelley,whoin“TheCloud,”apoemthatgenerationsofstudentsweremadetolearn,presentsthecloudasthekeytoacyclicalvisionoftimeinnature.Mitchell’sseeminglyoddjuxtapositionofthetwowordsinhisnovel’stitlesuggestsachallenge:ifoneseespeoplenotasdatapointstobecapturedin a network diagram or in a statistical regression analysis, but rather as ephemeralformationsdriftingorwanderingthroughtimeandspace,thenwhatwouldamapoftheirlives,theircloudatlas,looklike?

Mitchell’snovel,whichwonnumerousawardsandnominations,wasalsoadaptedforthescreenbythecreatorsofTheMatrixtrilogy,totepidreviews,perhapsevidenceofhowdifficult it is to turn a novelwhose author is primarily takenwith themetaphor of thecloudintoafilmwhosecreatorstaketheirmetaphorsfromtheworldofdata.CloudAtlasfeatures six characters whose lives extend from the nineteenth century to the distantfuture,crisscrossingtheworld,butendingwheretheybegan, intheislandsof theSouthPacific.Thecharactersarecontainedindiscretestories thatproceedchronologically, thefirstfiveofwhicharebrokenoffbeforeending.Eachstoryreferencesthepreviousonebyhaving a character read about it as, for example, one person happens on the journalproducedbythemaincharacterinanearlierstory.Thesixthstoryisthepivotpoint,andfrom that one, each story is completed in reverse chronological order, each tale nestedwithintheotherslikeasetofRussiandolls.RecallingShelley’sclassicpoem,Mitchell’shistoryiscyclical.Thelinearityweappeartoexperienceislittlemorethanacomfortingmirage.

The cloud and its atlas take three forms in the novel. The first is music, which,alongside poetry, is a familiar type of discourse for presenting clouds. One of the sixcharacters, a young musician named Robert Frobisher, works on “The Cloud AtlasSextet,”whichhecompletes justbeforecommittingsuicide.Thenextmaincharacter toappearlocatesararerecordingofthepieceinanoldmusicshop.Thesextetembodiestheunity in difference that the six main characters represent and was produced while theyoung Frobisher was helping a well-known composer complete the major symphony,appropriately calledEternal Recurrence. Contemplating his plan to end his young life,Frobisherisresolute:“Myheadisaromancandleofinvention.Lifetime’smusicarrivingall at once. Boundaries between noise and sound are conventions, I see now. Allboundaries are conventions, I see now, national ones too. One may transcend anyconvention, if only one can first conceive of doing so” (Mitchell 2004, 460). And soFrobishertranscendsconventionbyconceivinganatlasofclouds,inmusicalform.

Frobisher’s sextet is the cloud’s way of speaking about the novel’s protagonists, buteachcharacter is also connected to another, and thereby liveson in the flowofhistory,through a distinct form of communication, a secondmanifestation of the cloud. AdamEwingleavesapersonaldiary,LuisaReyisthecharacterinamysterypotboiler,TimothyCavendishlivesoninafilmmadeabouthissadlife,andSonmi,aheroiccyborg,emergesin the future as a goddess whose totems are worshipped. The simple Zachry survivesthrough the stories, some true, some not, that his children recall. Finally, as the worldstands on the brink of self-inflicted destruction we encounter the orison, a small egg-shaped, holographic communication device that is several generations ahead of today’sbest-equippedsmartphone.Notsurprisingly,itappearsmagicaltoacommunityofpeoplewithlittleadvancedtechnology,butnoneofitsgodlikepowerscanpreventtheinevitablefallofthecivilizationthatbuiltit.

The thirdmanifestationof thecloudand itsatlas is throughthemetaphorof thesoul.WhenZachryasksa scientist, oneof the few remaining inwhatwasonceanadvancedcivilization, how her people face deathwithout belief in a soul, the scientist replies inZachry’sdialect,“ourtruthisterrorsomecold.”Zachryfindsitworsethancold:“Justthatonce I sorried for her. Souls cross the skies o’ time… like clouds crossin’ skies o’ theworld.” And later, as Zachry and the scientist hide from attackers, “I watched cloudsawobblyfromtheflooro’thatkayak.Soulscrossageslikecloudscrossskies,an’tho’acloud’sshapenorhue,norsizedon’tstaythesame,it’sstillacloudan’soisasoul.Whocansaywherethecloud’sblowedfromorthesoul’llbe’morrow?OnlySonmi,theeastan’ thewest an’ the compass an’ the atlas, ya,only the atlaso’ clouds” (Mitchell 2004,308).Asmysteriousasclouds,thespiritsofpeopleliveonacrosstimeandspaceandonlyagoddessoraspiritualatlascantelluswhattheyareandwheretheyaregoing.

Like Aristophanes and The Cloud of Unknowing’s writer, DavidMitchell is a cloudengineer who builds his clouds out of human imagination. Like the engineers whoconstruct the systems that make up today’s cloud computing, Mitchell’s creationsovercometheconstraintsoftimeandspacetocaptureessentialinformationandhelpustoprocess it in ways that advance our understanding of the human condition. Mitchell’scloud takes numerous forms, but they primarily embody a network of individualswhomeetacrosstimethroughthewidevarietyofmediatheyleavebehind,demonstratingthatevenastoday’sdigitalengineersworkonthemeansofstoringconsciousnessincomplex

systems,wealreadystoreconsciousnessinthedevicesthatfillCloudAtlas.Thejournalofanineteenth-centurylawyer,themusicalscoreofanearlytwentieth-centurycomposer,thedetective story that describes the life of a struggling writer, the film that lampoons aBritishpublisher’sagent,andonintothefuturewherewefindtheiconofacyborg-turned-goddess,thecomputerdevicethatbringstimeandspacetothispresentmoment,andtheoral tales thatasimple tribesmanleaveshischildren,all formacloudofconsciousness.Thereare,ofcourse,differencesbetweenthecloudsshapedfromtheliteraryimaginationand those that emerge from the no-less-imaginative worlds of science and technology.Clearly the formerbuilds cloudsoutof fiction and is assessed for its capacity to createworldsthatmayormaynotbearacloserelationshiptotheworldweknow,whereasthelattercreatecloudsofdataandapplicationsthatarejudgedbytheircapacitytorepresentan empirical reality. But it is all too easy to dwell on simple differences; it is moreimportant to consider the subtle ones that shed light on each enterprise, particularly byprovidingaculturalgroundingfromwhichtothinkaboutcloudcomputing.

ForMitchell,thecloudthatcountsisdrawnfromarichpoolofsubjectivity,includingemotional intelligence, that isconstantlysensitive to the riskof reducingconsciousness,character,spirit,orsoul,toafewnotabledatapoints.CloudAtlasisnotjustastoryabouttheseeminguniversalityofpeoplepreyingonothers,mainlyformaterialgain,butalsoforthe sheer pleasure of domination, and it is not just a tale about how people respond,sometimessuccessfullybutoftennot,throughstruggleandresistance.Ifthiswereallthatmattered,wewouldnotneedacloudatlasbecauseallcloudswouldbe thesame.Theirrichness and diversity emerge from the historical context in which each node in thenetworkofcloudsisimmersed.Thisisoftenmissedinbig-dataanalysis,whichaddresseshistory by examining networks or even networks of networks over time, but does sothroughaprocessofextrapolation,typicallyfromquantitativedata.Itisanapproachthathas difficulty with those key historical turns or slow, crescive changes that are vitallyinfluentialbuthardtodetect.Tocorrectthisproblemrequiresimaginationandexperienceaswellashumanormachineintelligence.

Makingmattersmorecomplexarethesubjectivecategoriesandinterpretationsofthose,including the novelist and the reader, who provide descriptions and assessments. Theclassic description of the communication process, Shannon andWeaver’smathematicalmodel (1949), distinguishes transmitter from receiver, information source fromdestination, and signal from noise.When it is relatively easy to identify each of these,primarilywheneachstepintheprocessismechanized,themodelmakessomesense.Butformostformsofhumancommunication,thetermsarefarmoreambiguousthanitmightatfirstappear.“Boundariesbetweennoiseandsoundareconventions,”declaresFrobisherinMitchell’s novel, and all conventions can and should be transcended.AsNateSilver(2012), one of big data’s best-known champions, understands, one cannot simplyannounce a distinction between signal and noise because they are both ambiguous andrelativetothesubjectiveexpectationsofthoseconnectedtothecommunicationnetwork.Just as modern physics challenges the existence of an independent observer operatingoutside the systemunder study,meaning that relativity is universal, no one, neither thenovelist nor the data analyst, resides outside the social network of human actors.Informationsourcescanalsobedestinations,transmitterscansimultaneouslyreceive,andwhat isnoiseforsomeissweetmusicoreffectivecommunicationforothers.Moreover,

writers and researchers are also communicators with stakes in the objects under theirparticularmicroscopes.

Finally,thereisthemediumofinterpretationitself,demonstratedbythestarkdifferencebetweenCloudAtlasthenovelandCloudAtlasthefilm.Onedoesnothavetotravelasfardown the deterministic road asMcLuhan did to agree that themedium,whether it is anovel, film,or research report,hasan impacton themessagecommunicated.Thenovelcreates room for complexity, nuance, and the reader’s imagination that film, howevervisuallystunning,ismorechallengedtoreplicate.Theresearchreportprovidesaconcisesnapshotofenormousquantitiesofdatathatneitherthenovelnorthefilmcanmatch.Butindoingso,thereportmakesassumptionsaboutdefinitionsandchoicesand,moreoftenthannot,paysthepriceforitsconcisionbyrepressingthecomplexityandsubjectivityofthe objects under study. Nor does the report take into account the complexity of itsformation—specifically how, as the science scholar Bruno Latour (2009) hasdemonstrated,thescientificprocessmakesitswaytocompletionthroughmultiplemodesofexpressionandrepresentation.

At the very least, Mitchell’s atlas of clouds reminds us that there are legitimatealternativewaysofknowingandofcommunicatingknowledgealongsidethoseenshrinedin clouds of big data accessed through digital positivism. However, the latter isincreasingly crowding out the former as advances in computational capability and dataanalysisareappliedtomoreofwhatusedtobethehumanitiesandthesocialsciences.Thespreadofthedigitalhumanities,theiraccesstofunding,andtheirsupportfromuniversityleaders who desperately need the resources that big data in the humanities can attractmake it more difficult for those who defend the kind of detailed, qualitativeunderstandingsthathumanitiesscholarshavedeployedforcenturies.

Coda:CloudsAreintheAirThecloudmetaphorhasalwaysplayedaroleinourliteraryandartistictraditions.ButIcannothelpbutthinkthatthisisatimewhentheimageofthecloudholdsaparticularlyimportant cultural prominence. Perhaps it is the debate over climate change. After all,cloudcoverisamajoruncertaintyinforecastingfutureclimate.Perhapsitisthemedia’sfascinationwithweather coverage, especiallywhen natural disaster strikes. Itmay alsohavetodowithgrowingawarenessofcloudcomputing.Thatthemetaphoricalcloud,aswellastheliteralone,isintheairwasevidentona2012triptoNewYorkCitywhere,onavisit to theMetropolitanMuseumofArt,Iobservedamodernclassicofcloudcultureandoneaspiringtojointhatcategory.ThefirstwasanexhibitionofAndyWarhol’sSilverClouds,comprisingaroomfullofhelium-filledmetal“pillows”floatinggently,likefair-weather cloudsona springday.Warholbeganworkonhis clouds after the scientist heworkedwith,BillyKlüverofBellLabs,convincedWarholthathisoriginalidea,floatinglightbulbs,would notwork. Rather than drop the project,Warhol reportedly respondedimmediatelywith, “Let’smake clouds” (“TheWarhol:SilverClouds”2010).The resultwasoneofthegreatmoderncollaborationsbetweenanartistandanengineer,aworkofartwhosepiecesfloatthrougharoomandgentlybumpupagainstoneanotherandtheirobservers.Themetallic exterior creates an initial surprisebecausemetal objects arenotsupposedtofloatonair.Thisfeelingquicklygiveswaytoasenseofrandommovementthat has been captured formally by dance companies after the 1968 success of MerceCunningham’s ensemble dressed in costumes designed byWarhol’s artist friend JasperJohns.But it isalsoexpressed informally,asanyobserverofaSilverCloud installationnoticeswhennormallystationarymuseum-goerscannothelpbutdancetheirway,howeverawkwardly,aroundthecloud-filledroom.

ThatsamedaytookmeuptotherooftopgardenoftheMet,wheremoreartisticcloudsattracted large crowds. This time it was Tomás Saraceno’s installation Cloud City, acollectionof large,connectedmodulesbuiltwith reflectiveand transparentmaterial thatrisefromthegroundandinviteobserverstoclimbamongthem.10Thesightofgroupsofusclimbing throughanetworkofclouds, reflectingour imagesmany timesover,asweroseabovethecity,wasbeautiful,particularlybecauseweweresurroundedbythecity’sbuildingsandbyCentralPark,butalsofrightening,becausethenetworkofpeoplenestedin reflectiveand transparentsurfacescreated thesensationof life insidean information-processingdevice.ButthatmayhavebeenonlybecauseIhavebeenthinkingalotaboutanothertypeofcloud.

Cloud computing itself is becoming the object of conscious artistic expression,mostnotably in the Clouding Green collection created by Tamiko Thiel (2012), one of aremarkablegroupofcontemporaryartistswhogivelifetotheart-sciencemovement.Withdegrees in product design engineering (Stanford) and mechanical engineering (MIT),Thiel works on multi-dimensional, augmented-reality projects that create dramaticnarrativesofsocialandculturalsignificance.CloudingGreenusestechnologytopresentavisual expression of the share of data-center emissions taken up by renewable energysources.UsingtheGreenpeace(2012)report“HowCleanIsYourCloud?”sheprovidesavisually stunning presentation of color-coded clouds sweeping across the skies overcorporateclouddatacenters.Indoingso,Thielbuildsabridgeacrossthedividebetween

cloudcomputingandcloudculturewith thegoalofcreatingbothartandenvironmentalawareness. In the hands of an artist, clouds of data come alive with the emotionalresonanceneededtoenergizeaninformedresponse.Thisconvergenceoftechnology,art,and politics renews the hope that dark clouds are not the only ones on our collectivehorizon.

NOTES

Chapter2

1.ForanexampleofhowApplewieldedthispowertoprotectitsiTunesservice,seeBott2013.

2.Theconsequencesofacloudcompany’sbankruptcycanbecatastrophicforcustomers,assomelearnedwhenthepromisingcloudproviderNirvanixdisappearedin2013(Kepes2013).

3.Onepathistopursuestrategicallianceswithcloudcompaniesthatdonothavetheburdenoflegacysystems.InNovember2013,HPtookthisapproachbyteamingwithSalesforcetoputdedicatedHPcomputerservers,datastorage,andnetworkingintoSalesforce’scloud-computingfacilities(Kolakowski2013).

4.Ontheideologyof“openness”seeMorozov2013a.

5.VMware disrupted the traditional servermarket by developing software that allows servers to do thework ofmultiplemachines,enablingcomplextaskstobesharedoverseveralservers.

6.Notknownforexcessivemodesty,Ellisonhasnowbecomeabigcloudbooster:“Idon’tacceptthenotionIdidn’tgetthecloud.IthinkIinventedit”(Waters2012).

7.TheU.K.isalsoinvestingheavilyinitscapacitytolaunchcyber-attacks.Infact,in2013itbecamethefirstnationtoformallyannouncethatitwasdevelopinganoffensivecyberwarfarecapability(Fung2013).

Chapter3

2.Althoughthissimplerversionofmythcanbeuseful;seeLanda2013.

3.Atthetimeofwriting,thisadisavailabletowatchatwww.youtube.com/watch?v=MaA9l2H8BM8.

4.ThequotationappearedinthecommentssectionofapostingoftheadvertisementonYouTube,whichwaslatertakendown.Fora time, thevideowasunavailableonlineuntil it reappearedonYouTube.Theadvertisementwas socontroversialthatitinspiredasatire:www.youtube.com/watch?feature=player_embedded&v=buYxMvqkDfs.

5.Applehas also succeededat lobbying the federal government.The$2.5million it spent lobbyingWashington,D.C.,from2012to2013paidoffwhenPresidentObamatooktheunusualstepofoverturningaU.S.InternationalTradeCommissionpatent-infringementrulingagainstApple(Kirchgaessner2013).

6.Heisnotaloneinthisview.SeeParry2013.

7. In 2013, an entire school was set up with what appears to be the primary goal of generating enthusiasm forinformation technology.DraperUniversityofHeroes,based inSiliconValleyand justdown the road fromFacebookheadquarters, teaches aspiring entrepreneurs “the tech world’s own brand of magical thinking.” Students fork over$9,500tospendtwomonthschantingunderpostersofBillGatesandotherITluminaries,learningalittlebitofcodingandalotofwaystoworshipatthealtarofthecloud,bigdata,andallthatmakesupIT.Accordingtoonedescription,“it’sreallyaneight-weekinfomercialforthecultureofSiliconValley.Itsgoalistoinfectstudentswiththeexuberanceoftechandmakethembraveenoughtoleaveatraditionalcareerpathforastintinstart-upland”(Roose2013).

Chapter4

1.Google’sProjectLoon,whichoperatesthroughanetworkofsolar-poweredballoonstodeliverWi-Fiservicestounderservedareas,isasmallcaseofacloudsystemthatactuallyusescloud-likeobjects.Nevertheless,itis,forsome,theconfusingexception(Meehan2013).

2.Seewww.OVH.com.

3.FormediascholarSeanCubitt,“Thecloudisnotweightless:itisaheavyindustry.Addinthemetalsandplastics,thehydrodams, the thousandsofmilesofcables, the satellitesand their rocket launches,and themillionsof tonsofelectronicgadgetsweusetoaccessourmovies—andthecloudlooksalittlelessfluffy”(2013).

4. Other considerations, including a business-friendly tax code, enabled the company to pay no federal or stateincometaxonthoseearnings.Infact,itreceivedarefundof$429million(CitizensforTaxJustice2013).

5.ItisalsonotverycomfortingtohearGooglerejectclaimstoprivacy,aswhenitsfilinginaprivacycasedeclaredthat“apersonhasnolegitimateexpectationofprivacyininformationhevoluntarilyturnsovertothirdparties”(Szoldra2013).

6.ThecompanyisalsotryingtomoveuptheITfoodchainbysettinguponeoftheworld’slargestcloud-computingresearchanddevelopmentcentersinTaiwan(CioL2013).

7. Consider a 2013 conference announcement on libraries: “It is predicted that within five years, all librarycollections,systemsandserviceswillbedrivenintothecloud.Thisconferencewillbeanattempttoexplorehowcloudcomputingcouldbeappliedforlibraryapplications”(DailyPioneer2013).

8.Seehttp://microwork-dev.ucsd.edu.

Chapter5

1.https://en.wikipedia.org/wiki/Big_data.

2.WalterBinneylefttheNSAin2001butstayedincontactwithNSAemployees.Heleftoncetheagencystarteditswarrantlesswiretapprogram.Binneyexplains,“TheyviolatedtheConstitutionsettingitup.Buttheydidn’tcare.Theyweregoingtodoitanyway,andtheyweregoingtocrucifyanyonewhostoodintheway.WhentheystartedviolatingtheConstitution,Icouldn’tstay”(Bamford2012).

3.See,forexample,Tilahun,Feuerverger,andGervers2012.

4.Noteveryexpertandcommentatoragreeswiththemyth.Ina2013addresstoaconferencesessiontitled“DataScientist: The Sexiest Job of the 21st Century,” the chief technology officer for President Obama’s 2012 campaignarguedthat“datascientistasaprofessionislargelyafad”(Parry2013).

5.Seewww.reinhartandrogoff.com/related-research/growth-in-a-time-of-debt-featured-in.

6.Thesightofadatacenter temptsme to thinkaboutanotherpopularMagrittepainting, thisoneofapipe (TheTreacheryofImagesorCecin’estpasunepipe—Thisisnotapipe).Iwouldcaptiontheimageofthedatacenter’sdullbanalityCecin’estpasunnuage—Thisisnotacloud.

7. Frankly, I feel blessed to have received a rich education in the humanities well before the field required theadjective“digital”foritslegitimacyandperhapsforitssurvival.

8. I can recall reading his work for the first time as a university student in the 1960s and feeling the surge ofpossibility in knowing that by cultivating themindwewould be participating in a global process of advancing thehumanraceclosertoitscosmicdestinyattheOmegaPoint.

9.ItisonlyslightlyironicthatJoniMitchelliswellknownforsingingaboutcloudsinthesong“BothSides,Now”andespeciallyforitslyric“Butcloudsgotinmyway.”

10.www.metmuseum.org/saraceno.

REFERENCES

Abdul, Salam. 2013. “What John Smith Thinks of Cloud Computing: A Big ’Ol Warehouse.” Cloud Tweaks.

www.cloudtweaks.com/2013/01/what-john-smith-thinks-of-cloud-computing-a-big-ol-warehouse.

Acaroglu, Leyla. 2013. “Where Do Old Cellphones Go to Die?” New York Times.www.nytimes.com/2013/05/05/opinion/sunday/where-do-old-cellphones-go-to-die.html.

Anderson, Chris. 2008. “The End of Theory: The Data Deluge Makes the Scientific Method Obsolete.” Wired.www.wired.com/science/discoveries/magazine/16-07/pb_theory.

Andrejevic,Mark.2013.Infoglut:HowTooMuchInformationIsChangingtheWayWeThinkandKnow.NewYork:Routledge.

Anonymous.2009.TheCloudofUnknowing.EditedandtranslatedbyCarmenAcevedoButcher.Boston:Shambala.

Ante, Spencer. 2012. “IBM and the Cloud: Danger and Opportunity.” Wall Street Journal.http://blogs.wsj.com/corporate-intelligence/2012/10/17/ibm-and-the-cloud-danger-and-opportunity.

Apple.2011.“AppleiCloudCommercial.”YouTube.www.youtube.com/watch?v=YWZTMyjmcnU.

———.2012.“AppleTVAdiCloudHarmony.”YouTube.www.youtube.com/watch?v=YWZTMyjmcnU.

Applebaum,Alec.2013.“TechsintheCity.”NewYorkTimes.www.nytimes.com/2013/06/02/opinion/sunday/the-limits-of-big-data-in-the-big-city.html.

Asay,Matt.2013.“NateSilverGetsRealaboutBigData.”ReadWriteEnterprise.http://readwrite.com/2013/03/29/nate-silver-gets-real-about-big-data.

Babcock, Charles. 2013a. “Amazon Again Beats IBM for CIA Cloud Contract.” Information Week.www.informationweek.com/cloud/infrastructure-as-a-service/amazon-again-beats-ibm-for-cia-cloud-contract/d/d-id/1112211.

———. 2013b. “Amazon, Telcos Will Battle for Cloud Customers.” Information-Week.www.informationweek.com/cloud-computing/infrastructure/amazon-telcos-will-battle-for-cloud-cust/240150631.

Bachman, Katy. 2013. “Is Facebook About to Run into Privacy Issues Again?” Adweek.www.adweek.com/news/technology/facebook-about-run-privacy-issues-again-147491.

Bachner,Jennifer.2013.PredictivePolicing:PreventingCrimewithDataandAnalytics.Washington,DC:IBMCenterfor the Business of Government. www.businessofgovernment.org/content/about-center-business-government-connecting-research-practice.

Bamford, James. 2012. “The NSA Is Building the Country’s Biggest Spy Center (Watch What You Say).”Wired.www.wired.com/threatlevel/2012/03/ff_nsadatacenter/all.

———. 2013. “They Know Much More Than We Think.” New York Review of Books.www.nybooks.com/articles/archives/2013/aug/15/nsa-they-know-much-more-you-think.

Barr, Alistair. 2013. “Amazon vs. IBM: Big Blue Meets Match in Battle for the Cloud.” Reuters Canada.http://ca.reuters.com/article/businessNews/idCABRE96K04B20130721.

Barrett,Brian.2012.“ShouldtheInternetBeaUtility?”Gizmodo.http://gizmodo.com/5972173/should-the-internet-be-a-utility.

Barthes,Roland.1979.Mythologies.TranslatedbyA.Lavers.NewYork:HillandWang.

———.1982.ABarthesReader.EditedbySusanSontag.NewYork:HillandWang.

Barton, Mike. 2012. “Greenpeace Cloud Protest: Do Amazon, Microsoft Deserve the Doghouse?” Wired.www.wired.com/insights/2012/04/greenpeace-cloud-protest.

Basulto, Dominic. 2013. “Is This the Year Everybody Gets Hacked?” Washington Post.www.washingtonpost.com/blogs/innovations/post/is-this-the-year-everybody-gets-hacked/2013/02/21/eeb88fd4-7c2f-11e2-9073-e9dda4ac6a66_blog.html.

Bauman,Zygmunt.2000.LiquidModernity.Cambridge,UK:Polity.

Beer, David. 2012. “Using Social Media Data Aggregators to Do Social Research.” SociologicalResearch Online.www.socresonline.org.uk/17/3/10.html.

Beidel, Eric. 2012. “Flood of Data Puts Air Force’s Drone Growth on Hold.” National Defense.www.nationaldefensemagazine.org/blog/Lists/Posts/Post.aspx?ID=738.

Bernnat,Rainer,WolfgangZink,NicolaiBieber,andJoachimStrach.2012.“StandardizingtheCloud:ACalltoAction.”Booz&Co.www.booz.com/media/uploads/BoozCo_Standardizing-the-Cloud.pdf.

Bilton,Nick,andClaireCainMiller.2013.“HowPay-per-GazeAdvertisingCouldWorkforGoogleGlass.”NewYorkTimes.http://bits.blogs.nytimes.com/2013/08/20/google-patents-real-world-pay-per-gaze-advertising.

Bott, Ed. 2013. “How Apple Used Its Money and Muscle to Kill an iTunes Competitor.” ZDNet.http://anteyekon4myst.sharedby.co/share/eP3mAq.

Bourne, James. 2013. “Cloud Computing Saves Energy on Huge Scale, Says New Study—but How?”CloudTech.www.cloudcomputing-news.net/news/2013/jun/12/cloud-computing-saves-energy-huge-scale-says-new-study-how.

Boyd,Danah,andKateCrawford.2012.“CriticalQuestionsforBigData.”Information,CommunicationandSociety15,no.5:662–679.

Bradshaw,Tim.2012.“Apple:Innovator’sDilemma.”FinancialTimes.www.ft.com/intl/cms/s/2/b558911a-4ac7-11e2-9650-00144feab49a.html.

———. 2013. “Apple Says Child Labour Found at Suppliers.”Financial Times. www.ft.com/intl/cms/s/0/8af2a286-6754-11e2-8b67-00144feab49a.html.

Bradshaw, Tim, and Emily Steel. 2013. “Hacked PCs Falsify Billions of Ad Clicks.” Financial Times.www.ft.com/intl/cms/s/0/ab60c728-908f-11e2-a456-00144feabdc0.html.

Bradsher,Keith,andDavidBarboza.2013.“H.P.DirectsItsSuppliesinChinatoLimitStudentLabor.”NewYorkTimes.www.nytimes.com/2013/02/08/business/global/hewlett-packard-joins-push-to-limit-use-of-student-labor-in-china.html.

Brannen, Kate. 2013. “Wanted: Geeks to Help Fight Pentagon Cyberwar.” Politico.www.politico.com/story/2013/01/wanted-geeks-to-help-fight-pentagons-cyberwar-86946.html.

Brian,Matt.2013.“ThousandsofAccountsFoundtoHostUnsecuredPass-words,Photos,andOtherFilesonAmazon’sCloud.”Verge.www.theverge.com/2013/3/27/4152964/researcher-exposes-data-businesses-amazon-s3.

Briscoe,Gerard, andAlexandrosMarinos. 2009. “CommunityCloudComputing.” First InternationalConference onCloud Computing, Beijing, China. LSE Research Online.http://eprints.lse.ac.uk/26516/1/community_cloud_computing_%28LSERO_version%29.pdf.

Brooks, David. 2013. “What Data Can’t Do.”New York Times. www.nytimes.com/2013/02/19/opinion/brooks-what-data-cant-do.html.

Bryant,Chris.2013.“EuropeanDataProtectionunderaCloud.”FinancialTimes.www.ft.com/intl/cms/s/0/dbee868a-f43c-11e2-8459-00144feabdc0.html.

Budden, Robert. 2013. “Ads on Facebook Drop after Appearing Next to Offensive Posts.” Financial Times.www.ft.com/intl/cms/s/0/d1e74ee8-c7ae-11e2-be27-00144feab7de.html.

Bump,Philip.2013.“Update:NowWeKnowWhyGoogling‘PressureCookers’GetsaVisitfromCops.”AtlanticWire.www.theatlanticwire.com/national/2013/08/government-knocking-doors-because-google-searches/67864.

Bundy, Todd, andMichaelHaley. 2012. “China’sCloudCities.”OSP. www.ospmag.com/issue/article/Chinas-Cloud-Cities.

Burke,Edmund.1998.PhilosophicalEnquiry into theOriginofOur Ideasof theSublimeandBeautiful.NewYork:Penguin(originalworkpublished1756).

Burn-Murdoch, John. 2012. “Big Data: What Is It and How Can It Help?” Guardian.www.guardian.co.uk/news/datablog/2012/oct/26/big-data-what-is-it-examples.

Butler, Brandon. 2012a. “Gartner: 1/3 of Consumer Data Will Be Stored in the Cloud by’16.” Network World.www.networkworld.com/news/2012/062512-gartner-cloud-260450.html.

———. 2012b. “Gartner: Cloud Putting Crimp in Traditional Software, Hardware Sales.” Network World.www.networkworld.com/news/2012/071312-gartner-cloud-260882.html.

———. 2013a. “Cost Battle: Cloud Computing vs. In-House IT.” Network World.www.networkworld.com/news/2013/031813-cloud-cost-267615.html.

———. 2013b. “What GE’s Cloud Computing Foray Means for Big Data.” Network World.www.networkworld.com/news/2013/061913-ge-cloud-271011.html.

Butler, Declan. 2013. “When Google Got Flu Wrong.” Nature. www.nature.com/news/when-google-got-flu-wrong-1.12413.

Calo, Ryan. 2013. “The Catch-22 That Prevents Us from Truly Scrutinizing the Surveillance State.” Atlantic.www.theatlantic.com/technology/archive/2013/03/the-catch-22-that-prevents-us-from-truly-scrutinizing-the-surveillance-state/273738.

Cassidy, John. 2013. “Bezos and the Washington Post: A Skeptical View.” New Yorker.www.newyorker.com/online/blogs/johncassidy/2013/08/bezos-and-the-washington-post-a-skeptical-view.html.

Castle, Stephen. 2013. “Report of U.S. Spying Angers European Allies.” New York Times.www.nytimes.com/2013/07/01/world/europe/europeans-angered-by-report-of-us-spying.html.

CenterforEnergy-EfficientTelecommunications.2013.ThePowerofWirelessCloud.Melbourne,Australia:UniversityofMelbourne.www.ceet.unimelb.edu.au/pdfs/ceet_white_paper_wireless_cloud.pdf.

Chappuis, Bertil. 2012. “Cloudpreneurs.” Techonomy. http://techonomy.com/conf/12-tucson/entrepreneurship-2/cloudpreneurs.

Chatter.com.2011a.“HalftimeAd1withwil.i.am.”YouTube.www.youtube.com/watch?v=tdqoQ0zL7GQ.

———.2011b.“HalftimeAd2.”YouTube.www.youtube.com/watch?v=tcjAD-_H_rk.

Chen, Brian X. 2012. “‘The Cloud’ Challenges Amazon.” New York Times.www.nytimes.com/2012/12/27/technology/latest-netflix-disruption-highlights-challenges-of-cloud-computing.html.

Chief Information Officer, Department of Defense. 2012. Cloud Computing Strategy.www.defense.gov/news/dodcloudcomputingstrategy.pdf.

Chou, Timothy. 2012. “Marco Polo 2.0: Report fromChina.”CFO. www3.cfo.com/article/2012/12/the-cloud_china-cloud-services-companies-strategies.

Chronicle of Higher Education. 2013. “Major Players in the MOOC Universe.” http://chronicle.com/article/Major-Players-in-the%20MOOC/138817.

CioL. 2013. “Foxconn Plans to Establish Cloud Computing R&D Center in Southern Taiwan.”www.ciol.com/ciol/news/193445/foxconn-plans-establish-cloud-computing-r-d-center-southern-taiwan.

Cisco. 2013. “Cisco Global Cloud Index: Forecast and Methodology, 2012–2017.”www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns1175/Cloud_Index_White_Paper.pdf.

CitizensforTaxJustice.2013.“Facebook’sMulti-BillionDollarTaxBreak:Executive-PayTaxBreakSlashesIncomeTaxesonFacebook—andOtherFortune500Companies.”http://ctj.org/ctjreports/2013/02/facebooks_multi-billion_dollar_tax_break_executive-pay_tax_break_slashes_income_taxes_on_facebook—.php.

Clancy, Heather. 2012. “Greenpeace Re-grades Apple in Green Data Center Report.” ZDNet.www.zdnet.com/greenpeace-re-grades-apple-in-green-data-center-report-7000000805.

Clark,Jack.2012a.“CloudComputing’sUtilityFutureGetsCloser.”ZDNet.www.zdnet.com/cloud-computings-utility-future-gets-closer-7000007256.

———. 2012b. “HowGoogle Compute Engine Hopes to Sidestep AWS Failures.” ZDNet. www.zdnet.com/google-compute-engine-hopes-to-sidestep-aws-failures-7000001379.

Clement,Andrew.2013. “IXmaps—TrackingYourPersonalData through theNSA’sWarrantlessWiretappingSites.”Paperpresentedatthe2013IEEEInternationalSymposiumonScienceandTechnology,June27–29,Toronto.

Clifford,Stephanie,andJessicaSilver-Greenberg.2013.“RetailersTrackEmployeeTheftsinVastDatabases.”NewYorkTimes.www.nytimes.com/2013/04/03/business/retailers-use-databases-to-track-worker-thefts.html.

CloudExpo.2013.“EnterpriseIT’sTwoBiggestGameChangersunderOneRoof—CloudComputingandBigData.”www.cloudcomputingexpo.com.

Cloud Tweaks. 2012. “Cloud Predictions for the New Year.” http://us2.campaign-archive2.com/?u=04809abc68958c8c94da79e96&id=36235497ec&e=0eda32265e.

———. 2013. “Do Cloud Companies Know How to Market Themselves?” http://us2.campaign-archive2.com/?

u=04809abc68958c8c94da79e96&id=183a823c77&e=0eda32265e.

Cohen, Julie E. 2013. “What Privacy Is for.” Harvard Law Review 126: 1904–1933.www.harvardlawreview.org/media/pdf/vol126_cohen.pdf.

Cohen, Reuven. 2012. “How Cloud Computing Helped Obama Win the Presidential Election.” Forbes.www.forbes.com/sites/reuvencohen/2012/11/15/how-cloud-computing-helped-obama-win-the-presidential-election.

———. 2013. “Cloud Computing at the Hotel California: Check-In and Never Leave!” Forbes.www.forbes.com/sites/reuvencohen/2013/05/02/cloud-computing-at-the-hotel-california-check-in-and-never-leave.

Columbus, Louis. 2012a. “Cloud Computing and Enterprise Software Forecast Update, 2012.” Forbes.www.forbes.com/sites/louiscolumbus/2012/11/08/cloud-computing-and-enterprise-software-forecast-update-2012.

———.2012b.“RoundupofCloudComputingForecastsandCloudComputingEstimates,2012.”SoftwareStrategiesBlog. http://softwarestrategiesblog .com/2012/01/17/roundup-of-cloud-computing-forecasts-and-market-estimates-2012.

Crawford, Susan. 2012. “US Internet Users Pay More for Slower Service.” Bloomberg.www.bloomberg.com/news/2012-12-27/u-s-internet-users-pay-more-for-slower-service.html.

CRMSoftwareBlogeditors.2011.“6ReasonsWhySalesforce.comIsWorriedaboutMicrosoftCRM.”CRMSoftwareBlog.www.crmsoftwareblog.com/2011/09/6-reasons-why-salesforce-com-is-worried-about-microsoft-crm.

Crovitz, L. Gordon. 2013. “Silicon Valley’s ‘Suicide Impulse.’” Wall Street Journal.http://online.wsj.com/article/SB10001424127887323539804578266290231304934.html.

Cubitt, Sean. 2013. “How toWeigh a Cloud.”TheConversation. http://theconversation. com/how-to-weigh-a-cloud-19581.

Cushing, Ellen. 2013. “Amazon Mechanical Turk: The Digital Sweatshop.”UTNE Reader. www.utne.com/science-technology/amazon-mechanical-turk-zm0z13jfzlin.aspx.

DailyPioneer. 2013. “Meet on Cloud Computing.” http://archive.dailypioneer .com/avenues/127319-meet-on-cloud-computing-.html.

Dalwadi,Manish.2012.“EnterpriseCloudComputing:TheWarforEnterpriseSoftware.”CenterforDigitalStrategies,Tuck School of Business at Dartmouth. http://digitalstrategies.tuck.dartmouth.edu/assets/images/Enterprise_Cloud_Final_Manish_Dalwadi.pdf.

Darrow, Barb. 2013. “If PRISM Doesn’t Freak You Out about Cloud Computing, Maybe It Should, Says PrivacyExpert.”Gigaom. http://gigaom.com /2013/06/28/if-prism-doesnt-freak-you-out-about-cloud-computing-maybe-it-should-says-privacy-expert.

DataCenter Journal. 2013. “Industry Perspective: Energy Efficiency and Renewable Sources for the Data Center.”www.datacenterjournal.com/facilities/industry-perspective-energy-efficiency-and-renewable-sources-for-the-data-center.

Davenport,ThomasH., PaulBarth, andRandyBean. 2012. “How ‘BigData’ IsDifferent.”MITSloanManagementReview.http://sloanreview.mit.edu/article/how-big-data-is-different.

Davis,Erik.1998.Techgnosis:Myth,MagicandMysticismintheAgeofInformation.NewYork:Harmony.

Deibert,RonaldJ.2013.BlackCode:InsidetheBattleforCyberspace.NewYork:RandomHouse.

Delany, Ella. 2013. “Humanities Studies under Strain around the World.” New York Times.www.nytimes.com/2013/12/02/us/humanities-studies-under-strain-around-the-globe.html.

DeLillo,Don.1985.WhiteNoise.NewYork:Viking.

Deloitte. 2009. “Cloud Computing: Forecasting Change.” www.deloitte.com/assets/Dcom-Netherlands/Local%20Assets/Documents/EN/Services/Consulting/nl_en_consulting_cloud_computing_security_privacy_and_trust.pdf.

Dembosky,April.2013a.“FacebookBuysAd-ServingPlatform.”FinancialTimes.www.ft.com/intl/cms/s/0/5de42176-81ed-11e2-b050-00144feabdc0.html.

———. 2013b. “Facebook Spending on Lobbying Soars.”Financial Times. www.ft.com/intl/cms/s/0/cfaf0c78-65b2-11e2-a17b-00144feab49a.html.

Dembosky,April,andJamesFontanella-Khan.2013.“USTechGroupsCriticizedforEULobbying.”FinancialTimes.www.ft.com/intl/cms/s/0/e29a717e-6df0-11e2-983d-00144feab49a.html.

Dery,Mark.1996.“IndustrialMemory.”21c.www.21cmagazine.com/Mark-Dery-Industrial-Memory.

Digging into Data Challenge. 2011. “ChartEx.”www.diggingintodata.org/Home/AwardRecipientsRound22011/tabid/185/Default.aspx.

Dignan, Larry. 2011a. “Analytics in 40 Years: Machines Will Kick Human Managers to the Curb.” ZDNet.www.zdnet.com/blog/btl/analytics-in-40-years-machines-will-kick-human-managers-to-the-curb/61092.

———. 2011b. “Cloud Computing’s Real Creative Destruction May Be the IT Workforce.” ZDNet.www.zdnet.com/blog/btl/cloud-computings-real-creative-destruction-may-be-the-it-workforce/61581.

Duhigg, Charles, and David Barboza. 2012. “In China, Human Costs Are Built into an iPad.” New York Times.www.nytimes.com/2012/01/26/business/ieconomy-apples-ipad-and-the-human-costs-for-workers-in-china.html.

Dutta, Soumitra, and Beñat Bilbao-Osorio. 2012. Global Information Technology Report 2012: Living in aHyperconnectedWorld.Geneva:WorldEconomicForum.

Dyer-Witheford,Nick.2013.“RedPlentyPlatforms.”CultureMachine14:1–27.

Eckel, Erik. 2012. “Why Businesses Shouldn’t Trust Apple’s Cloud Services.” TechRepublic.www.techrepublic.com/blog/mac/why-businesses-shouldnt-trust-apples-cloud-services/2262.

Edwards, Jim. 2013. “Liberal Groups Begin Facebook Ad Boycott over Zuckerberg’s Big Oil Lobbying.”BusinessInsider.www.businessinsider.com/facebook-ad-boycott-over-fwdus-oil-lobbying-2013-5.

Eggers,David.2013.TheCircle:ANovel.NewYork:Knopf.

Einav, Liran, and JonathanD. Levin. 2013. “TheData Revolution and EconomicAnalysis.”Working Paper 19035,NationalBureauofEconomicResearch,Cambridge,MA.www.nber.org/papers/w19035.

Elowitz, Ben. 2013. “In Media, Big Data Is Booming but Big Results Are Lacking.” All Things D.http://allthingsd.com/20130520/in-media-big-data-is-booming-but-big-results-are-lacking.

Erl,Thomas,RicardoPuttini,andZaighamMahmood.2013.CloudComputing:Concepts,TechnologyandArchitecture.UpperSaddleRiver,NJ:PrenticeHall.

Evans-Pritchard,Ambrose.2012.“HighTechExpansionDrivesChina’sSecondBoomin theHinterland.”Telegraph.www.telegraph.co.uk/finance/comment/9701910/Hi-tech-expansion-drives-Chinas-second-boom-in-the-hinterland.html.

Ewing, Jack. 2013. “Amazon’s Labor Relations under Scrutiny in Germany.” New York Times.www.nytimes.com/2013/03/04/business/global/amazons-labor-relations-under-scrutiny-in-germany.html?pagewanted=all&_r=0&gwh=FCF0A5D17B34AF08B45C6563A3889C3E&gwt=pay.

Fingas, Jon. 2013. “Strategy Analytics: iCloud, Dropbox and Amazon Top Cloud Media in the US.” Engadget.www.engadget.com/2013/03/21/strategy-analytics-cloud-media-market-share.

Finkle, Jim. 2012. “Amazon-Netflix ChristmasOutage and the Costly Risks of ‘Cloud’ Reliance.”Globe andMail.www.theglobeandmail.com/report-on-business/international-business/us-business/amazon-netflix-christmas-outage-and-the-costly-risks-of-cloud-reliance/article6736744.

———. 2013. “White House Will Soon Revive Cybersecurity Legislation.” Reuters.www.reuters.com/article/2013/02/26/us-cybersecurity-obama-idUSBRE91P02120130226.

Fish, Stanley. 2012a. “The Digital Humanities and the Transcending of Morality.” New York Times.http://opinionator.blogs.nytimes.com/2012/01/09/the-digital-humanities-and-the-transcending-of-mortality.

———. 2012b. “Mind Your P’s and B’s: The Digital Humanities and Interpretation.” New York Times.http://opinionator.blogs.nytimes.com/2012/01/23/mind-your-ps-and-bs-the-digital-humanities-and-interpretation.

Fitzpatrick,Katharine.2011.PlannedObsolescence:Publishing,Technology,andtheFutureoftheAcademy.NewYork:NewYorkUniversityPress.

Foley, John. 2012. “10 Developments Show Government Cloud Maturing.” Information Week.www.informationweek.com/government/cloud-saas/10-developments-show-government-cloud-ma/240002578?itc=edit_in_body_cross.

Fontanella-Kahn, James, and Bede McCarthy. 2013. “Brussels to Soften Data Protection Rules.” Financial Times.www.ft.com/intl/cms/s/0/dbf20262-8685-11e2-b907-00144feabdc0.html.

Forbes.2012.“AmericansStillUnclearaboutCloudComputing.”www.forbes.com/sites/thesba/2012/11/13/americans-still-unclear-about-cloud-computing.

Fox,Justin.2013.“TheWeb’sNewMonopolists.”Atlantic.www.theatlantic.com/magazine/archive/2013/01/the-webs-

new-monopolists/309197.

Franck,Lewis. 2013. “4Non-obviousCosts ofCloudDowntime.”CloudTech.www.cloudcomputing-news.net/blog-hub/2013/feb/05/4-non-obvious-costs-of-cloud-downtime.

Franklin,Seb.2012.“CloudControl,ortheNetworkasMedium.”CulturalPolitics8,no.3:443–464.

Freeland, Chrystia. 2013. “When Work and Wages Come Apart.” New York Times.www.nytimes.com/2013/02/22/us/22iht-letter22.html.

Fung, Brian. 2013. “How Britain’s New Cyberarmy Could Reshape the Laws of War.” Washington Post.www.washingtonpost.com/blogs/the-switch /wp/2013/09/30/how-britains-new-cyberarmy-could-reshape-the-laws-of-war.

Gallagher, Ryan. 2013. “Software That Tracks People on Social Media Created by Defence Firm.” Guardian.www.guardian.co.uk/world/2013/feb/10/software-tracks-social-media-defence.

Gallagher, Sean. 2012. “How Team Obama’s Tech Efficiency Left Romney IT in the Dust.” Ars Technica.http://arstechnica.com/information-technology/2012/11/how-team-obamas-tech-efficiency-left-romney-it-in-dust.

Gangireddy, Geetha. 2012. “Making the Most of Cloud Computing in the Military and Department of Defense.”Blackboard Blogs. http://blog.blackboard.com/professional-education/making-the-most-of-cloud-computing-in-the-military-department-of-defense.

Gapper, John. 2013a. “Bosses Are Reining in Staff Because They Can.” Financial Times.www.ft.com/intl/cms/s/0/90300088-80cf-11e2-9c5b-00144feabdc0.html.

———. 2013b. “Google Is the General Electric of the 21st Century.” Financial Times.www.ft.com/intl/cms/s/0/e57abef0-cd0c-11e2-90e8-00144feab7de.html.

Gartner.2013.“AboutGartner.”www.gartner.com/technology/about.jsp.

Gates,Bill.1995.TheRoadAhead.NewYork:Viking.

Gerovitch, Slava. 2010. “The Cybernetic Scare and the Origins of the Internet.” Baltic Worlds.http://balticworlds.com/the-cybernetics-scare-and-the-origins-of-the-internet.

Gillespie,Tarleton.2013.“CananAlgorithmBeWrong?”Limn.http://limn.it/can-an-algorithm-be-wrong.

Gilmoor, Dan. 2013. “Embrace the Cloud Computing Revolution with Caution.” Guardian.www.guardian.co.uk/commentisfree/2013/mar/05/cloud-data-revolution-google-chromebook-pixel.

Ginsberg, Jeremy,MatthewH.Mohebbi,RajanS.Patel,LynnetteBrammer,MarkS.Smolinski, andLarryBrilliant.2009.“DetectingInfluenzaEpidemicsUsingSearchEngineQueryData.”Nature457:1012–1014.

Giridharadas, Anand. 2013a. “Reality Crashes the Technocrats’ Party.” New York Times.www.nytimes.com/2011/03/26/us/26iht-currents26.html.

———.2013b.“WhattheFacebookSearchEngineTellsUs.”NewYorkTimes.www.nytimes.com/2013/02/23/us/23iht-letter23.html.

Glanz, James. 2012a. “Data Barns in a Farm Town, Gobbling Power and Flexing Muscle.” New York Times.www.nytimes.com/2012/09/24/technology/data-centers-in-rural-washington-state-gobble-power.html.

———.2012b.“Power,Pollution,andtheInternet.”NewYorkTimes.www.nytimes.com/2012/09/23/technology/data-centers-waste-vast-amounts-of-energy-belying-industry-image.html.

Glanz, James. 2013. “Landlords Double as Energy Brokers.” New York Times.www.nytimes.com/2013/05/14/technology/north-jersey-data-center-industry-blurs-utility-real-estate-boundaries.html.

Glanz,James,andEricLipton.2004.TheRiseandFalloftheWorldTradeCenter.NewYork:TimesBooks.

Globe Investor. 2012. “Competition Chatter around U.S. Telcos Gets Louder.” www.theglobeandmail.com/globe-investor/news-sources/?date=20121231&archive=rtgam&slug=escenic_6804996.

Glover, Tony. 2013. “Cloudy Outlook for Blue Sky Computing in the Middle East.” National.www.thenational.ae/business/industry-insights/technology/cloudy-outlook-for-blue-sky-computing-in-the-middle-east.

Gold,MatthewK.2012.DebatesintheDigitalHumanities.Minneapolis:UniversityofMinnesotaPress.

Goldberg,Michael.2013.“CloudComputingExpertsDetailBigPrivacyandSecurityRisks.”DataInformed.http://data-informed.com/cloud-computing-experts-detail-big-data-security-and-privacy-risks.

Goldner,Matt.2010.WindsofChange:LibrariesandCloudComputing.Dublin,OH:OnlineComputerLibraryCenter.

Gonsalves, Antone. 2013. “The Nine Top Threats Facing Cloud Computing.” ReadWriteEnterprise.http://readwrite.com/2013/03/04/9-top-threats-from-cloud-computing.

Gordon, Michael. 2013. “2 Americans Advised Not to Visit North Korea.” New York Times.www.nytimes.com/2013/01/04/world/asia/bill-richardson-and-googles-eric-schmidt-are-advised-not-to-visit-north-korea.html.

Gordon, Robert J. 2000. “Does the ‘New Economy’Measure Up to the Great Inventions of the Past?” Journal ofEconomicPerspectives14,no.4:49–74.

Greenhouse, Steven. 2013. “Tackling Concerns of Independent Workers.” New York Times.www.nytimes.com/2013/03/24/business/freelancers-union-tackles-concerns-of-independent-workers.html.

Greenpeace International. 2010. “Making IT Green: Cloud Computing and Its Contribution to Climate Change.”www.greenpeace.org/international/Global/international/planet-2/report/2010/3/make-it-green-cloud-computing.pdf.

———. 2011. “How Dirty Is Your Data? A Look at the Energy Choices That Power Cloud Computing.”www.greenpeace.org/international/Global/international /publications/climate/2011/Cool%20IT/dirty-data-report-greenpeace.pdf.

———. 2012. “How Clean Is Your Cloud?”www.greenpeace.org/international/Global/international/publications/climate/2012/iCoal/HowCleanisYourCloud.pdf.

Greenwald, Glenn. 2013. “Liberal Icon Frank Church on the NSA.” Guardian.www.guardian.co.uk/commentisfree/2013/jun/25/frank-church-liberal-icon.

Gross, Grant. 2013. “Ovum: Big Data Collection Clashing with Privacy Concerns.” InfoWorld.www.infoworld.com/d/big-data/ovum-big-data-collection-colliding-privacy-concerns-212397.

Groucutt,Peter.2013.“CloudComputingIsBecomingaUtility.”RealBusiness.http://realbusiness.co.uk/article/24508-cloud-computing-is-becoming-a-utility-.

Hanna, Sheree. 2013. “Cloud Computing—Where Does Your Data Go When the Service Dies?” African BusinessReview. www.africanbusinessreview.co.za/business_leaders/cloud-computing-where-does-your-data-go-when-the-service-dies.

Hardy, Quentin. 2012a. “Active in Cloud, Amazon Reshapes Computing.” New York Times.www.nytimes.com/2012/08/28/technology/active-in-cloud-amazon-reshapes-computing.html.

———. 2012b. “Google Apps Challenging Microsoft in Business.” New York Times.www.nytimes.com/2012/12/26/technology/google-apps-moving-onto-microsofts-business-turf.html.

———. 2012c. “Intel’s Schooling from the ‘Big Four’ Cloud Customers.” New York Times.http://bits.blogs.nytimes.com/2012/12/05/intels-schooling-from-cloud-customers.

———. 2013a “Amazon Bares Its Computers.” New York Times. http://bits.blogs.nytimes.com/2013/11/15/amazon-bares-its-computers.

———. 2013b. “Declaring This a Year for Fixing and Rebuilding, H.P. Posts Lower Profit.” New York Times.www.nytimes.com/2013/02/22/technology/hp-reports-decline-in-revenue-and-profit.html.

———. 2013c. “Elance Pairs Hunt for Temp Work with Cloud Computing.” New York Times.http://bits.blogs.nytimes.com/2013/09/24/elance-pairs-hunt-for-temp-work-with-cloud-computing.

———. 2013d. “Google: Let a Billion Supercomputers Bloom.” New York Times.http://bits.blogs.nytimes.com/2013/05/21/google-let-a-billion-supercomputers-bloom.

———.2013e.“I.B.M.InflatesItsCloud.”NewYorkTimes.http://bits.blogs.nytimes.com/2013/06/18/i-b-m-inflates-its-cloud.

———. 2013f. “Intel Tries to Secure Its Footing beyond PCs.” New York Times.www.nytimes.com/2013/04/15/technology/intel-tries-to-find-a-foothold-beyond-pcs.html?pagewanted=all&_r=0.

———. 2013g. “Intel’s Extensive Makeover.” New York Times. http://bits.blogs.nytimes.com/2013/09/10/intels-extensive-makeover.

———. 2013h. “Oracle and Salesforce: A Data Sharing Deal.” New York Times.http://bits.blogs.nytimes.com/2013/06/21/oracle-and-salesforce-a-data-sharing-deal.

———.2013i.“WhyBigDataIsNotTruth.”NewYorkTimes.http://bits.blogs.nytimes.com/2013/06/01/why-big-data-is-not-truth.

Harpreet. 2013. “Huawei to Offer Cloud Storage to CERN.” Tools Journal. www.toolsjournal.com/cloud-articles/item/1272-huawei-to-offer-cloud-storage-to-cern.

Harris, Derrick. 2013a. “Researchers Create Cloud-Based Brain for Robots.” Gigaom.http://gigaom.com/2013/03/11/researchers-create-cloud-based-brain-for-robots.

———.2013b. “WeNeedaDataDemocracy,Not aDataDictatorship.”Gigaom. http://gigaom.com/2013/04/07/we-need-a-data-democracy-not-a-benevolent-data-dictatorship.

Hartzog,Woodrow, andEvan Selinger. 2013. “Obscurity:ABetterWay to Think aboutYourData Than ‘Privacy.’”Atlantic. www.theatlantic.com/technology/archive/2013/01/obscurity-a-better-way-to-think-about-your-data-than-privacy/267283.

Hayles,N.Katherine.1999.HowWeBecamePosthuman:VirtualBodies inCybernetics,Literature,and Informatics.Chicago:UniversityofChicagoPress.

Herndon, Thomas, Michael Ash, and Robert Pollin. 2013. “Does High Public Debt Consistently Stifle EconomicGrowth?ACritiqueofReinhart andRogoff.”UniversityofMassachusettsAmherst,PoliticalEconomyResearchInstitute.www.peri.umass.edu/236/hash/31e2ff374b6377b2ddec04deaa6388b1/publication/566.

Hickey, Andrew R. 2012. “Verizon Relies on Cloud Services for Future Growth.” CRN.www.crn.com/news/cloud/232500629/verizon-relies-on-cloud-services-for-future-growth.htm.

Hill,Kashmir.2013. “SurpriseVisitorsAreUnwelcomeat theNSA’sUnfinishedUtahSpyCenter (EspeciallyWhenTheyTakePhotos).”Forbes.www.forbes.com/sites/kashmirhill/2013/03/04/nsa-utah-data-center-visit.

Hille, Katharin, and Daniel Thomas. 2013. “China Blames U.S. Hackers for Attack s.” Financial Times.www.ft.com/intl/cms/s/0/8203676e-818f-11e2-904c-00144feabdc0.html.

Hodson, Hal. 2013. “Crowdsourcing Grows Up as Online Workers Unite.” New Scientist.www.newscientist.com/article/mg21729036.200-crowdsourcing-grows-up-as-online-workers-unite.html.

Hoover, J. Nicholas. 2012. “6 Ways Amazon Helped Obama Win.” InformationWeek.www.informationweek.com/government/cloud-saas/6-ways-amazon-cloud-helped-obama-win/240142268.

———. 2013. “Military Plans Exabyte Storage Cloud.” InformationWeek.www.informationweek.com/government/cloud-saas/military-plans-multi-exabyte-storage-clo/240152481.

Horn, Leslie. 2011. “Facebook Picks Sweden for First Data Center Outside U.S.” PC.www.pcmag.com/article2/0,2817,2395378,00.asp.

Houlder, Vanessa. 2013. “Google Accused of Devious and Unethical Behaviour.” Financial Times.www.ft.com/intl/cms/s/0/d1193b70-be2b-11e2-bb35-00144feab7de.html.

Hunnius,Gerry,G.DavidGarson,andJohnCase,eds.1973.Workers’Control:AReaderonLabourandSocialChange.NewYork:Random.

Hunter, Andrea. 2011. The Digital Humanities: Third Culture and Democratization of the Humanities. PhD diss.,Queen’sUniversity,Kingston,Ontario,Canada.

IBM.2012a.“AllintheCloud.”YouTube.www.youtube.com/watch?v=q_d7Io_rr2s.

———.2012b.“IBMAdvertisement.”NewYorker,November19,23.

———.2013.“IBMAdvertisement.”NewYorker,April8,11.

Investor’s Business Daily. 2013. “Cloud Computing Users Are Losing Data, Symantec Finds.”http://finance.yahoo.com/news/cloud-computing-users-losing-data-205500612.html.

Ipeirotis, Panos. 2013. “Mechanical Turk: Now with 40.92% Spam.” A Computer Scientist in a Business School.www.behind-the-enemy-lines.com/2010/12/mechanical-turk-now-with-4092-spam.html.

Isaacson,Walter.2011.SteveJobs.NewYork:SimonandSchuster.

Jacob, Rahul. 2013. “Better Workplaces Require Better Consumers.” Financial Times.www.ft.com/intl/cms/s/0/94215ed0-97d6-11e2-b7ef-00144feabdc0.html.

Jenkins, Patrick. 2013. “HSBC Set to Cut Thousands of Jobs.”Financial Times. www.ft.com/intl/cms/s/0/8d7afe12-8cd0-11e2-8ee0-00144feabdc0.html.

John, James. 2013. “Cloud Computing for Government Is Not Just a Cost Cutter.” Information Daily.

www.theinformationdaily.com/2013/02/01/cloud-computing-for-government-is-not-just-a-cost-cutter.

Jokinen,Pekka,PenttiMalaska,andJariKaivo-oja.1998.“TheEnvironmentinan‘InformationSociety’:ATransitionStagetowardsMoreSustainableDevelopment?”Futures30,no.6:485–498.

Kaminska, Izabella. 2013. “What Google Reader Tells Us about Banking and Nationalisation.” Financial Times.http://ftalphaville.ft.com/2013/03/25/1438422/what-google-reader-tells-us-about-banking-and-nationalisation.

Kanter, James. 2013. “European Regulators Fine Microsoft, Then Promise to Do Better.” New York Times.www.nytimes.com/2013/03/07/technology/eu-fines-microsoft-over-browser.html.

Kelly,Kevin.2002.“GodIstheMachine.”Wired,December,180–185.

———.2010.WhatTechnologyWants.NewYork:Penguin.

Kenealy, Chris. 2013. “Five Different Ways to Sell Cloud Computing to Anyone.” Cloud Tweaks.www.cloudtweaks.com/2013/03/five-different-ways-to-sell-cloud-computing-to-anyone.

Kepes, Ben. 2013. “A Nirvanix Post-Mortem—Why There’s No Replacement for Due Diligence.” Forbes.www.forbes.com/sites/benkepes/2013/09/28/a-nirvanix-post-mortem-why-theres-no-replacement-for-due-diligence.

Kerner, Sean Michael. 2013. “30 Years of TCP/IP Dominance Began with a Deadline.” InternetNews.com.www.internetnews.com/blog/skerner/30-years-of-tcpip-dominance-began-with-a-deadline.html.

Kerr,Dara.2012.“MicrosofttoGoogle:You’reNot‘Serious’aboutBusinessApps.”CNET.http://news.cnet.com/8301-1023_3-57561046-93/microsoft-to-google-youre-not-serious-about-business-apps.

———. 2013. “NASA Falls Short on Its Cloud Computing Security.” CNET. http://news.cnet.com/8301-1009_3-57596053-83/nasa-falls-short-on-its-cloudcomputing-security.

Kirchgaessner, Stephanie. 2013. “Obama Patent Move Caps New Apple March on Washington.” Financial Times.www.ft.com/intl/cms/s/0/b84f5a88-ff77-11e2-b990-00144feab7de.html.

Kirilov, Kiril. 2011. “Cloud Computing Market Will Top $241 Billion in 2020.” Cloud Tweaks.www.cloudtweaks.com/2011/04/cloud-computing-market-will-top-241-billion-in-2020.

Kisker, Holger. 2011. “10 Cloud Predictions for 2012.” Forrester. http://blogs.forrester.com/holger_kisker/11-12-13-10_cloud_predictions_for_2012.

Ko, Carol. 2012. “How to Sell Cloud Computing to CIOs and CFOs.” Asia Cloud Forum.www.asiacloudforum.com/content/how-sell-cloud-computing-cios-and-cfos.

Kolakowski, Nick. 2013. “Salesforce, HP Teaming Up to Sell ‘Superpods.’” Slashdot.http://slashdot.org/topic/cloud/salesforce-hp-teaming-up-to-sell-superpods/.

Konczai, Mike. 2013. “Reinhart-Rogoff a Week Later: Why Does This Matter?” Next New Deal.www.nextnewdeal.net/rortybomb/reinhart-rogoff-week-later-why-does-matter.

Kreisberg, Jennifer Cobb. 1995. “A Globe, Clothing Itself with a Brain.” Wired.www.wired.com/wired/archive/3.06/teilhard.html.

Krugman, Paul. 2010. “Notes on Rogoff (Wonkish).” New York Times.http://krugman.blogs.nytimes.com/2010/07/21/notes-on-rogoff-wonkish.

———. 2013. “Is There Any Point to Economic Analysis?” New York Times.http://krugman.blogs.nytimes.com/2010/07/21/notes-on-rogoff-wonkish.

Kudryashov,Roman.2010.“RolandBarthes:MythToday.”WhatAreTheseIdeas?http://whataretheseideas.com/roland-barthes-myth-today.

Kurzweil,Ray.2005.TheSingularityIsNear.NewYork:Viking.

Lam,Lana.2013.“EdwardSnowden:U.S.GovernmentHasBeenHackingHongKongandChina forYears.”SouthChinaMorningPost.www.scmp.com/news/hong-kong/article/1259508/edward-snowden-us-government-has-been-hacking-hong-kong-and-china.

Landa, Heinan. 2013. “Top 4 Myths of Cloud Computing.” TechFlash.www.bizjournals.com/washington/blog/techflash/2013/01/top-4-myths-of-cloud-computing.html.

Langer,Art.2013.“It’sNotJusttheData,Stupid.”WallStreetJournal.http://mobile.blogs.wsj.com/cio/2013/02/19/its-not-just-the-data-stupid.

Lanier,Jaron.2013.WhoOwnstheFuture?NewYork:SimonandSchuster.

Latour,Bruno.1999.Pandora’sHope:EssaysontheRealityofScienceStudies.Cambridge,MA:HarvardUniversityPress.

Leach, Jim. 2011. “TheRevolutionary Implicationsof theDigitalHumanities.”Speech to5thAnnualConferenceofHASTAC, University of Michigan. National Endowment for the Humanities.www.neh.gov/about/chairman/speeches/the-civilizing-implications-the-digital-humanities.

Lee, Edmund. 2012. “Apple’s iTunes Would Be One of World’s Biggest Media Companies.” Techblog.http://go.bloomberg.com/tech-blog/2012-12-03-apple%E2%80%99s-itunes-would-be-one-of-world%E2%80%99s-biggest-media-companies.

Lee, Justin. 2013a. “ChinaCloudProgram to Invest $360B toMoreThanDoubleDataCenterCapacity.”WebHostIndustry Review. www.thewhir.com/web-hosting-news/china-cloud-program-to-invest-360b-to-more-than-double-data-center-capacity.

———. 2013b. “Public Cloud Services Spending toReach $47.4Billion in 2013: IDCReport.”WebHost IndustryReview.www.thewhir.com/web-hosting-news/public-cloud-services-spending-to-reach-47-4b-in-2013-idc-report.

Lee,TimothyB.2013.“TheseTechCompaniesAreSpendingMillionsonHigh-PricedLobbyists.”WashingtonPost.www.washingtonpost.com/blogs/wonkblog/wp/2013/07/02/these-tech-companies-are-spending-millions-on-high-priced-lobbyists.

Lessin, Jessica E. 2012. “Google’s Explainer-in-Chief Can’t Explain Apple.” Wall Street Journal.http://online.wsj.com/article/SB10001424127887323717004578159481472653460.html.

Lewin, Tamar. 2013. “Online Classes Fuel a Campus Debate.” New York Times.www.nytimes.com/2013/06/20/education/online-classes-fuel-a-campus-debate.html.

Linthicum, David. 2012. “Cloud Computing in 2013: Two Warnings.” Info-World. www.infoworld.com/d/cloud-computing/cloud-computing-in-2013-two-warnings-208759.

———. 2013a. “Everyone Has Heard of the Cloud, but Few Know What It Is.” InfoWorld.www.infoworld.com/d/cloud-computing/everyone-has-heard-of-the-cloud-few-know-what-it-210818.

———.2013b.“Hey,HR,GetOffMyCloud.” InfoWorld.www.infoworld.com/d/cloud-computing/hey-hr-get-of-my-cloud-218413.

———. 2013c. “The Proof Is In: Amazon Fully Controls the Cloud.” Info-World. www.infoworld.com/d/cloud-computing/the-proof-in-amazon-fully-controls-the-cloud-222641.

———. 2013d. “Thanks NSA, You’re Killing the Cloud.” Cloud Computing. www.infoworld.com/d/cloud-computing/thanks-nsa-youre-killing-the-cloud-220434.

———. 2013e. “What Will Cloud Computing’ Mean in 10 Years?” InfoWorld. www.infoworld.com/d/cloud-computing/what-will-cloud-computing-mean-in-10-years-219497.

Liptak, Adam. 2013. “Justices Turn Back Challenge to Broader US Eavesdropping.” New York Times.www.nytimes.com/2013/02/27/us/politics/supreme-court-rejects-challenge-to-fisa-surveillance-law.html.

Lohr,Steve.2013a.“BigDataSleuthing,1960sStyle.”NewYorkTimes.http://bits.blogs.nytimes.com/2013/06/10/big-data-intelligence-sleuthing-1960s-style.

———. 2013b. “Big Data Trying to Build Better Workers.” New York Times.www.nytimes.com/2013/04/21/technology/big-data-trying-to-build-better-workers.html.

———. 2013c. “McKinsey: The $33 Trillion Technology Advantage.” New York Times.http://bits.blogs.nytimes.com/2013/05/22/mckinsey-the-33-trillion-technology-payoff.

Luce, Edward. 2013. “Data Intelligence Complex Is the Real Issue.” Financial Times.www.ft.com/intl/cms/s/0/a1dd626c-cf80-11e2-be7b-00144feab7de.html.

Luckerson,Victor.2013.“PRISMbytheNumbers:AGuidetotheGovernment’sSecretInternetData-MiningProgram.”Time. http://newsfeed.time.com/2013/06/06/prism-by-the-numbers-a-guide-to-the-governments-secret-internet-data-mining-program.

Lynch, Michael P. 2013. “Privacy and the Threat to the Self.” New York Times.http://opinionator.blogs.nytimes.com/2013/06/22/privacy-and-the-threat-to-the-self.

MacIntyre,Alisdair.1970.SociologicalTheoryandPhilosophicalAnalysis.NewYork:Macmillan.

MacLeod,Ian.2013.“CloudE-dataLawPutsUsersatRisk;Canadians’PrivateInfoOpentoU.S.EyesviaComputingService.”OttawaCitizen,February2,A1.

Makower, Joel. 2012. “How Do You Measure the Cloud’s Environmental Impact?” GreenBiz.com.www.greenbiz.com/blog/2012/04/14/how-do-you-measure-environmental-impact-cloud.

Manjoo, Farhad. 2013. “Facebook Follows You to the Supermarket.” Slate.www.slate.com/articles/technology/technology/2013/03/facebook_advertisement_studies_their_ads_are_more_like_tv_ads_than_google.html.

MarketWatch.2013.“CloudSecurityAllianceWarnsProvidersof‘TheNotoriousNine’CloudComputingTopThreatsin 2013.” www.marketwatch.com/story/cloud-security-alliance-warns-providers-of-the-notorious-nine-cloud-computing-top-threats-in-2013-2013-02-25.

Markoff, John. 2012. “Troves of Personal Data, Forbidden to Researchers.” New York Times.www.nytimes.com/2012/05/22/science/big-data-troves-stay-forbidden-to-social-scientists.html.

Marlow, Iain. 2013. “Huawei Canadas Sean Yang.”’ Globe and Mail. www.theglobeandmail.com/report-on-business/careers/careers-leadership/huawei-canadas-sean-yang-dismissing-suspicion-over-dim-sum/article6957873.

Marshall, Alex. 2013. “Should the Public or Private Sector Control Broadband?” Government Technology.http://m.benton.org/node/148781?utm_campaign=Newsletters&utm_source=sendgrid&utm_medium=email.

Marshall, Bob. 2012. ‘“IBMs Smarter Planet TV Campaign Gets a Makeover.”’ Agency Spy.www.mediabistro.com/agencyspy/ibms-smarter-planet-tv-campaign-gets-a-makeover_b28274.

Marx, Leo. 1964. The Machine in the Garden: Technology and the Pastoral Ideal in America. New York: OxfordUniversityPress.

Mathias, Craig. 2012. “The Huawei Controversy—The Rest of the Argument.” Nearpoints.www.networkworld.com/community/blog/huawei-controversy-%E2%80%93-rest-argument.

Maxwell,Richard,andTobyMiller.2012a.GreeningtheMedia.NewYork:Oxford.

———. 2012b. “Greening Starts with Us.” New York Times.www.nytimes.com/roomfordebate/2012/09/23/informations-environmental-cost/greening-starts-with-ourselves.

Mayer-Schönberger, Viktor, andKenneth Cukier. 2013.Big Data: A Revolution ThatWill TransformHowWe Live,Work,andThink.NewYork:Houghton-Mifflin.

Mazzucato,Mariana.2013.TheEntrepreneurialState:DebunkingPublicvs.PrivateSector.London:AnthemPress.

McCall,Jay.2012.“AvoidtheCloudServicesPiecemealTrap.”BusinessSolutions.www.bsminfo.com/blog/bsm-blog.

McCarthy,Bede.2013.“StaffUnderminesCybersecurityEfforts.”FinancialTimes.www.ft.com/intl/cms/s/0/01f936e6-a365-11e2-ac00-00144feabdc0.html.

McChesney,RobertW. 2013.DigitalDisconnect:HowCapitalism Is Turning the Internet againstDemocracy.NewYork:NewPress.

McDuling,John.2013.“WhyIstheWorldsBiggestSeedCompanyBettingNearly$1BilliononaBigDataStartup?”’Quartz. http://qz.com/130946/why-is-the-worlds-biggest-seed-company-betting-nearly-1-billion-on-a-big-data-startup.

McFedries,Paul.2012.CloudComputing:BeyondtheHype.SanFrancisco:HPPress.

McKendrick, Joe. 2013a. “10 Quotes on Cloud Computing That Really Say It All.” Forbes.www.forbes.com/sites/joemckendrick/2013/03/24/10-quotes-on-cloud-computing-that-really-say-it-all.

———. 2013b. “Cloud Computing Market May Become an Oligopoly of High Volume Vendors.” Forbes.www.forbes.com/sites/joemckendrick/2013/07/11/cloud-computing-market-may-become-an-oligopoly-of-high-volume-vendors.

———. 2013c. “In the Rush to Cloud Computing, Here’s OneQuestionNot Enough People AreAsking.”Forbes.www.forbes.com/sites/joemckendrick/2013/02/19/in-the-rush-to-cloud-computing-heres-one-question-not-enough-people-are-asking.

McKinsey&Company.2013.“AboutUs.”www.mckinsey.com/about_us.

McLuhan,Marshall.1989.TheGlobalVillage.NewYork:OxfordUniversityPress.

McMillan, Robert. 2013. “Cloud Computing Snafu Shares Private Data between Users.” Wired.www.wired.com/wiredenterprise/2013/04/digitalocean.

Medina,Eden. 2011.Cybernetic Revolutionaries: Technology andPolitics in Allende’sChile. Cambridge,MA:MITPress.

Meehan,Chris.2013.“GooglesProjectLoonMakesCloudComputingaRealitywithSolar-PowerandBalloons.”’Daily

Fusion. http://dailyfusion.net/2013/06/googles-project-loon-makes-cloud-computing-a-reality-with-solar-power-and-balloons-12074.

Mell,Peter,andTimothyGrance.2011.“TheNISTDefinitionofCloudComputing.”NationalInstituteofStandardsandTechnology,InformationTechnologyLaboratory.www.csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf.

Meyer, David. 2013. “Cisco’s Gloomy Revenue Forecast Shows NSA Effect Starting to Hit Home.” Gigaom.http://gigaom.com/2013/11/14/ciscos-gloomy-revenue-forecast-shows-nsa-effect-starting-to-hit/.

Miller, Claire Cain. 2011. “Amazon Cloud Failure Takes Down Web Sites.” New York Times.http://bits.blogs.nytimes.com/2011/04/21/amazon-cloud-failure-takes-down-web-sites.

———. 2013. “Data Science: The Numbers of Our Lives.” New York Times.www.nytimes.com/2013/04/14/education/edlife/universities-offer-courses-in-a-hot-new-field-data-science.html.

Miller, Claire Cain, and Quentin Hardy. 2013. “Google Elbows into the Cloud.” New York Times.www.nytimes.com/2013/03/13/technology/google-takes-on-amazon-and-microsoft-for-cloud-computing-services.html.

Miller,Eden.2002.“DesigningFreedom,RegulatingaNation:SocialistCyberneticsinAllende’sChile.”WorkingPaper#34, Program in Science, Technology, and Society, MIT, Cambridge, MA.http://web.mit.edu/sts/pubs/pdfs/MIT_STS_WorkingPaper_34_Miller.pdf.

Miller, Kathleen, and Chris Strohm. 2013. “IBM Wins Its Largest Cloud-Computing Contract.” Bloomberg.www.bloomberg.com/news/2013-08-15/ibm-wins-its-largest-u-s-cloud-computing-contract.html.

Mills,MarkP.2013.“TheCloudBeginswithCoal:BigData,BigNetworks,BigInfrastructure,andBigPower.”DigitalPowerGroup.www.tech-pundit.com/wp-content/uploads/2013/07/Cloud_Begins_With_Coal.pdf.

Mims, Christopher. 2013. “Amazon Doesn’t Reveal What It Makes on Cloud Computing, but Here’s the Number,Anyway.” Quartz. http://qz.com/78754/amazon-doesnt-reveal-what-it-makes-on-cloud-computing-but-heres-the-number-anyway.

Mishkin, Sarah. 2013. “Foxconn Admits Student Intern Labor Violations at China Plant.” Financial Times.www.ft.com/intl/cms/s/0/88524304-319f-11e3-817c-00144feab7de.html.

Mishkin,Sarah,PattiWaldmeir,andKatharinHille.2013.“AppleSupplierFacesSanctionsinChina.”FinancialTimes.www.ft.com/intl/cms/s/0/cafeb812-7ce2-11e2-adb6-00144feabdc0.html.

Mitchell,David.2004.CloudAtlas.NewYork:RandomHouse.

Moeller, Katy. 2013. “Taxes on Computing Irk Idaho Tech Businesses.” Idaho Statesman.www.idahostatesman.com/2013/01/11/2408172/computing-tax-clouds-tech-economy.html.

Morozov, Evgeny. 2013a. “Open and Closed.” New York Times.www.nytimes.com/2013/03/17/opinion/sunday/morozov-open-and-closed.html.

———.2013b.ToSaveEverything,ClickHere:TheFollyofTechnologicalSolutionism.NewYork:PublicAffairs.

Mosco,Vincent.1982.PushbuttonFantasies:VideotexandInformationTechnology.Norwood,NJ:Ablex.

———.2004.TheDigitalSublime:Myth,Power,andCyberspace.Cambridge,MA:MITPress.

———.2009.ThePoliticalEconomyofCommunication.2nded.London:Sage.

Mosco,Vincent, andCatherineMcKercher. 2008.The Laboring of Communication:Will KnowledgeWorkers of theWorldUnite?Lanham,MD:LexingtonBooks.

Mosco,Vincent,CatherineMcKercher,andUrsulaHuws,eds.2010.GettingtheMessage:CommunicationandGlobalValueChains.London:Merlin.

Mosco,Vincent,andEliaZureik.1987.ComputersintheWorkplace:TechnologicalChangeintheTelephoneIndustry.ReportfortheCanadianFederalDepartmentofLabourTechnologyImpactResearchFund.

Moses, Asher. 2012. “How the Internet Became a Closed Shop.” Sydney Morning Herald.www.smh.com.au/technology/technology-news/how-the-internet-became-a-closed-shop-20121221-2brcp.html.

Moskowitz,Milton,andRobertLevering.2013.“The100BestCompaniestoWorkfor.”Fortune,February,85–96.

Musil, Steven. 2012. “Foxconn Working Conditions Slammed by Workers’ Rights Group.” CNET.http://news.cnet.com/8301-13579_3-57444213-37/foxconn-working-conditions-slammed-by-workers-rights-group.

Nagel, David. 2013. “Cloud Computing to Make Up 35% of K–12 IT Budgets in 4 Years.” Journal.http://thejournal.com/articles/2013/02/19/cloud-computing-to-make-up-35-of-k12-it-budgets-in-4-years.aspx.

NationalInstituteofStandardsandTechnology(NIST).2011.“USGovernmentCloudComputingTechnologyRoadmap:High Priority Requirements to Further USG Cloud Computing Adoption.”www.nist.gov/itl/cloud/upload/SP_500_293_volumeI-2.pdf.

National Institute of Standards and Technology (NIST). 2013. “NIST Cloud Computing Program.”www.nist.gov/itl/cloud/.

National Science Foundation (NSF). 2012. “Report on Support for Cloud Computing.”www.nsf.gov/pubs/2012/nsf12040/nsf12040.pdf.

Naughton, John. 2013. “Digital Capitalism Produces Few Winners.” Guardian.www.guardian.co.uk/technology/2013/feb/17/digital-capitalism-low-pay.

Neff,Gina.2012.VentureLabor.Cambridge,MA:MITPress.

Negroponte,Nicholas.1995.BeingDigital.NewYork:Knopf.

Nelson,D.Schwartz,andCharlesDuhigg.2013.“Apple’sWebofTaxSheltersSaved ItBillions,PanelFinds.”NewYorkTimes.www.nytimes.com/2013/05/21/business/apple-avoided-billions-in-taxes-congressional-panel-says.html.

New York Times. 2013a. “Report: Deepening Ties between N.S.A. and Silicon Valley.”http://bits.blogs.nytimes.com/2013/06/20/daily-report-the-deepening-ties-between-the-n-s-a-and-silicon-valley.

New York Times. 2013b. “Should Companies Tell Us When They Get Hacked?”www.nytimes.com/roomfordebate/2013/02/21/should-companies-tell-us-when-they-get-hacked.

Nextgov. 2013. “Pentagon Signs $5 Million Deal for Cyber Battleground.”www.nextgov.com/cybersecurity/2013/06/pentagon-signs-5-million-deal-cyber-battleground/65594.

Noble,David.1997.TheReligionofTechnology:TheDivinityofManandtheSpiritofInvention.NewYork:Knopf.

Novet, Jordan. 2013. “Long-Shot Distributed Data Center Project in Canada Like SETI for Mobile.” Gigaom.http://gigaom.com/2013/03/07/long-shot-distributed-data-center-project-in-canada-like-seti-for-mobile.

Nye, David. 1990.Electrifying America: SocialMeanings of a New Technology, 1880–1940. Cambridge,MA:MITPress.

Nye,David.1994.AmericanTechnologicalSublime.Cambridge,MA:MITPress.

O’Connor,Mark.2013. “How toRegulateCloudComputing.”Guardian.www.guardian.co.uk/media-network/media-network-blog/2013/mar/28/regulation-cloud-computing-data-protection.

O’Connor, Sarah. 2013. “Amazon Unpacked.” Financial Times. www.ft.com/intl/cms/s/2/ed6a985c-70bd-11e2-85d0-00144feab49a.html.

O’Neill, Shane. 2011. “Forrester: Public Cloud Growth to Surge, Especially SaaS.” CIO.www.cio.com/article/680673/Forrester_Public_Cloud_Growth_to_Surge_Especially_SaaS.

Ong, Josh. 2012. “Cloud Atlas: A Weather Forecast on the Chinese Cloud Industry.” Next Web.http://thenextweb.com/asia/2012/12/01/cloud-atlas-the-state-of-the-chinese-cloud-industry.

Orenstein, Gary. 2010. “Selling the Infrastructure Cloud.” Gigaom. http://gigaom.com/2010/07/11/selling-the-infrastructure-cloud.

Osborne, Charlie. 2013. “Chinese Labor Group Alleges Worker Abuse by Apple Supplier Pegatron.” ZDNet.www.zdnet.com/chinese-labor-group-alleges-worker-abuse-by-apple-supplier-pegatron-7000018660.

Page, Lewis. 2011. “DARPA Wants Weapons-Grade Military Cloud Computing.” Register.www.theregister.co.uk/2011/05/17/darpa_war_clouds.

Palmer,Maija.2013a.“CloudComputingHindersDataDeletion.”FinancialTimes.www.ft.com/intl/cms/s/2/0e8aad72-7444-11e2-80a7-00144feabdc0.html.

——.2013b.“DataMiningOffersRichSeam.”FinancialTimes.www.ft.com/intl/cms/s/2/61c4c378-60bd-11e2-a31a-00144feab49a.html.

Panattieri,Joe.2012.“Top100CloudServiceProvidersList2012.”Talkin’Cloud.http://talkincloud.com/%5Bprimary-term%5D/top-100-cloud-services-providers-list-2012-ranked-10-1.

Parkhill,DouglasF.1966.TheChallengeoftheComputerUtility.Reading,MA:Addison-Wesley.

Parkman,Ralph.1972.TheCyberneticSociety.NewYork:PergamonPress.

Parry,Marc.2013.“‘BigData’IsBunk,ObamaCampaign’sTechGuruTellsUniversityLeaders.”ChronicleofHigher

Education. http://chronicle.com/blogs/wiredcampus/big-data-is-bunk-obama-campaigns-tech-guru-tells-university-leaders/47885.

Parsons, John J. 2013. “TheUnutterableName.”HebrewNames ofGod.www.hebrew4christians.com/Names_of_G-d/YHVH/yhvh.html.

Pellow,David,andLisaSun-HeePark.2002.TheSiliconValleyofDreams.NewYork:NewYorkUniversityPress.

People’s Daily Online. 2013. “Tencent to Build Chongqing Cloud Computing Center.”http://english.peopledaily.com.cn/90778/8294842.html.

Perkins, Tara. 2013. “Should Data Centers Have a Canadian Address?” Globe and Mail.www.theglobeandmail.com/report-on-business/industry-news/property-report/should-data-centres-have-a-canadian-address/article13224735.

Perlin,Ross.2011.InternNation:HowtoEarnNothingandLearnLittleintheBraveNewEconomy.NewYork:Verso.

Perlroth,Nicole,andQuentinHardy.2013.“BankHackingWastheWorkofIranians,OfficialsSay.”NewYorkTimes.www.nytimes.com/2013/01/09/technology/online-banking-attacks-were-work-of-iran-us-officials-say.html.

Perlroth,Nicole,DavidE.Sanger,andMichaelS.Schmidt.2013.“AsHackingagainstU.S.Rises,ExpertsTrytoPinDown Motive.” New York Times. www.nytimes.com/2013/03/04/us/us-weighs-risks-and-motives-of-hacking-by-china-or-iran.html.

Perrow, Charles. 1999.Normal Accidents: Living with High-Risk Technologies. Princeton, NJ: Princeton UniversityPress.

Pew Research Center. 2013. The State of the News Media: 2013. Project for Excellence in Journalism.http://stateofthemedia.org.

Pilling, David. 2013. “America Cedes Moral High Ground on Cyber-Spying.” Financial Times.www.ft.com/intl/cms/s/0/0a11deea-d831-11e2-9495-00144feab7de.html.

Poe,Emily.2013.“Google’sBadCaseoftheFlu.”FierceCIO.www.fiercecio.com/story/googles-bad-case-flu/2013-02-15.

Pogue, David. 2013. “Photoshop’s New Rental Program and the Outrage Factor.” New York Times.http://pogue.blogs.nytimes.com/2013/07/05/photoshops-new-rental-program-and-the-outrage-factor.

Porat,MarcUri.1977.TheInformationEconomy:DefinitionandMeasurement.OfficeofTelecommunicationsSpecialPublication77–112,May.Washington,DC:USDepartmentofCommerce.

Powell,Rob. 2013. “PacnetHasBigPlans forChina.”Telecomasia.net.www.telecomasia.net/content/pacnet-has-big-plans-china.

Pretor-Pinney,Gavin.2011.TheCloudCollectorsHandbook.London:Hodder.

Qian, Zhao. 2013. “China’s Cloud Computing Chain Ready.” People’s Daily Online.http://english.peopledaily.com.cn/90778/8273069.html.

Qing, Liau Yun. 2013. “China’s Cloud Deployment Dampened by Nascent Enterprise Demand.” ZDNet.www.zdnet.com/cn/chinas-cloud-deployment-dampened-by-nascent-enterprise-demand-7000012662.

Quinn, Michelle. 2013. “Samsung’s Lobbying Grows with Its Market Share.” Politico.www.politico.com/story/2013/01/samsungs-lobbying-grows-with-its-market-share-86784.html.

Rachman,Gideon.2013.“AConspiracyofReasonablePeople.”FinancialTimes.www.ft.com/intl/cms/s/0/6e78755a-693d-11e2-b254-00144feab49a.html.

Ragland,LeighAnn,JosephMcReynolds,MatthewSoutherland,andJamesMulvenon.2013.RedCloudRising:CloudComputing in China. Vienna, VA: Center for Intelligence Research and Analysis.http://origin.www.uscc.gov/sites/default/files/Research/Red%20Cloud%20Rising_Cloud%20Computing%20in%20China.pdf.

Raimondo, Justin. 2013. “The Great Cyber-Warfare Scam.” Antiwar.com.http://original.antiwar.com/justin/2013/02/19/the-great-cyber-warfare-scam.

Rampell, Catherine. 2010. “They Did Their Homework (800 Years of It).” New York Times.www.nytimes.com/2010/07/04/business/economy/04econ.html.

Rao, Leena. 2009. “McKinsey’s Cloud Computing Report Is Partly Cloudy.” TechCrunch.http://techcrunch.com/2009/04/16/mckinseys-cloud-computing-report-is-partly-cloudy.

——. 2010. “Google Spent $1.34 Million on Lobbying in Q2, Up 41 Percent from Last Year.” TechCrunch.

http://techcrunch.com/2010/07/21/google-spent-1-34-million-on-lobbying-in-q2-up-41-percent-from-last-year.

“TheReasonsWhyAmazonMechanicalTurkNoLongerAcceptsInternationalTurkers.”2013.TipsforRequestersonMechanicalTurk.http://turkrequesters.blogspot.ca/2013/01/the-reasons-why-amazon-mechanical-turk.html.

Regalado, Antonio. 2011. “Who Coined ‘Cloud Computing’?” MIT Technology Review.www.technologyreview.com/news/425970/who-coined-cloud-computing.

Reid,Stefan,andHolgerKisker.2011.SizingtheCloud.Cambridge,MA:Forrester.

Reinhart, CarmenM., and Kenneth S. Rogoff. 2010. “Growth in a Time of Debt.”Working Paper 15639, NationalBureauofEconomicResearch,Cambridge,MA.www.nber.org/papers/w15639.pdf.

Reuters. 2013a. “Huawei Springs Back with 33% Rise in Net Profit.” New York Times.www.nytimes.com/2013/01/22/technology/huawei-springs-back-with-33-rise-in-net-profit.html.

——. 2013b. “Survey Details Data Theft Concerns for U.S. Firms in China.” New York Times.www.nytimes.com/2013/03/30/business/global/survey-details-data-theft-concerns-for-us-firms-in-china.html.

Revken, Andrew C. 2012. “Exploring the Roots of an Emerging Planet-Spanning ‘Mind’.” New York Times.http://dotearth.blogs.nytimes.com/2012/12/26/exploring-the-roots-of-an-emerging-planet-spanning-mind.

Ribeiro, John. 2013. “Microsoft’s Azure Service Hit by Expired SSL Certificate.” PCWorld.www.pcworld.idg.com.au/article/454609/microsoft_azure_service_hit_by_expired_ssl_certificate.

Rivlin,Gary. 2004. “The Tech Lobby, CallingAgain.”New York Times.www.nytimes.com/2004/07/25/business/the-tech-lobby-calling-again.html.

Robinson, Duncan. 2013. “Tech Trends Increase Cybercrime Threat.” Financial Times.www.ft.com/intl/cms/s/0/806d7d72-7d16-11e2-adb6-00144feabdc0.html.

Roeder,Ethan.2012. “IAmNotBigBrother.”NewYorkTimes.www.nytimes.com/2012/12/06/opinion/i-am-not-big-brother.html.

Rogers,Mike,andC.A.DutchRuppersberger.2012.“InvestigativeReportontheU.S.NationalSecurityIssuesPosedbyChineseTelecommunicationsCompaniesHuawei andZTE.”USHouse ofRepresentatives, Permanent SelectCommittee on Intelligence. http://intelligence.house.gov/sites/intelligence.house.gov/files/documents/Huawei-ZTE%20Investigative%20Report%20%28FINAL%29.pdf.

Romm, Tony. 2013a. “How Google Beat the Feds.” Politico. www.politico.com/story/2013/01/how-google-beat-the-feds-85743.html.

——. 2013b. “Twitter’s Hacker Problem.”’ Politico. www.politico.com/story/2013/04/twitter-ap-hacking-problem-90510.html.

Roose, Kevin. 2013. “Up, Up, and Away!” New York Magazine. http://nymag.com/news/features/draper-university-silicon-valley-2013-8.

Ross,Christopher.2012.“KeepYourHeadsintheCloud.”Corbus.www.corbus.com/blog/2012/01/06/keep-your-heads-in-the-cloud/.

Ross,Duncan.2012.“TurnYourAnalyticalSkillsintoaPublicGood.”Causes.www.causes.com/actions/1694321-turn-your-analytical-skills-into-a-public-good.

RTTNews.2013.“IsYourPrivateLifeSafeunderCloudComputing?”www.rttnews.com/2044231/is-your-private-life-safe-under-cloud-computing.aspx.

Sachs,Jeffrey.2013.“ProfessorKrugmanandCrudeKeynesianism.”HuffingtonPostCanada.http://blogs.ft.com/the-a-list/2012/07/12/move-americas-economic-debate-out-of-its-time-warp.

Sadowski, Jathan. 2013. “Why Does Privacy Matter? One Scholar’s Answer.” Atlantic.www.theatlantic.com/technology/archive/2013/02/why-does-privacy-matter-one-scholars-answer/273521.

Sanger,DavidE.,DavidBarboza,andNicolePerlroth.2013.“ChineseArmyIsSeenasTiedtoHackingAttackagainstU.S.”New York Times. www.nytimes.com/2013/02/19/technology/chinas-army-is-seen-as-tied-to-hacking-against-us.html.

Sayare, Scott. 2012. “On the Farms of France, the Death of a Pixelated Workhorse.” New York Times.www.nytimes.com/2012/06/28/world/europe/after-3-decades-in-france-minitels-days-are-numbered.html.

Schiller.Dan.1981.TelematicsandGovernment.Norwood,NJ:Ablex.

——.2014.DigitalDepression:TheCrisisofDigitalCapitalism.Urbana:UniversityofIllinoisPress.

Schmidt,Eric,andJaredCohen.2013.TheNewDigitalAge:ReshapingtheFutureofPeople,Nations,Business.NewYork:Knopf.

Schuman, Michael. 2013. “Marx’s Revenge: How Class Struggle Is Shaping the World.” Time.http://business.time.com/2013/03/25/marxs-revenge-how-class-struggle-is-shaping-the-world.

Schwarz,Ariel.2013.“HowMapsCanChangetheWorld.”FastCompany,www.fastcoexist.com/1681766/how-maps-can-change-the-world.

Schwarz, Matthew J. 2013. “Microsoft Hacked: Joins Apple, Facebook, Twitter.” InformationWeek Security,www.informationweek.com/security/attacks/microsoft-hacked-joinsapple-facebook-tw/240149323.

Scott, A. O. 2012. “Spielbergian, on a Budget.” New York Times.www.nytimes.com/2012/12/30/movies/awardsseason/beasts-of-the-southern-wild-shares-something-with-lincoln.html.

Sengupta, Semini. 2013. “The Pentagon as Silicon Valley’s Incubator.” New York Times.www.nytimes.com/2013/08/23/technology/the-pentagon-as-start-up-incubator.html.

Shannon,ClaudeE., andWarrenWeaver. 1949.TheMathematical Theory ofCommunication.Urbana:University ofIllinoisPress.

Shaw, Robert. 2013. “A Quick Guide for Cloud Companies That Don’t Understand Marketing.” Cloud Tweaks.www.cloudtweaks.com/2013/01/a-quick-guide-for-cloud-companies-that-dont-understand-marketing.

Sherr, Ian, and Don Clark. 2013. “H-P’s New Servers Will Cut Power Use.” Wall Street Journal.http://online.wsj.com/article/SB10001424127887324504704578410643910775254.html.

Shields, Greg. 2013. “How to Beat a Cloud Skeptic: Four Steps towards Rationalizing the Great Cloud Debate.”InfoWorld.http://resources.infoworld.com/ccd/show/200014814/00636570079897IFWGC2RZTF220.

Silver,Nate.2012.TheSignalandtheNoise.NewYork:Penguin.

Silverman, Gary. 2013. “Digital Intelligence and Dumb Terrorists.” Financial Times.www.ft.com/intl/cms/s/0/800d3268-d8d4-11e2-84fa-00144feab7de.html.

Silverman, Rachel Emma. 2013. “Tracking Sensors Invade the Workplace.” Wall Street Journal.http://online.wsj.com/article/SB10001424127887324034804578344303429080678.html.

Singer, Natasha. 2012. “A Vault for Taking Charge of Your Online Life.” New York Times.www.nytimes.com/2012/12/09/business/company-envisions-vaults-for-personal-data.html.

——. 2013. “When Your Data Wanders in Places You’ve Never Been.” New York Times.www.nytimes.com/2013/04/28/technology/personal-data-takes-a-winding-path-into-marketers-hands.html.

Singh,Gurjeet.2013.“TheBigDataWorldIsOperatingat1Percent.”Gigaom.http://gigaom.com/2013/03/10/the-big-data-world-is-operating-at-1-percent.

SmartData Collective. 2013. “Cloud Computing Use Increases among Supply Chains.”http://smartdatacollective.com/onlinetech/99516/cloud-computing-use-increases-among-supply-chains.

Smith, Eve. 2013. “Aristophanes’ Cloudcuckooland to Terry Pratchett’s Discworld: Comedy as Social Conscience.”ComedyStudies4,no.1:23–33.

Smith,Zadie.2012.“SomeNotesonAttunement.”NewYorker,December17,30–35.

Solman, Paul. 2013. “WebOils theWheels of Progress.”Financial Times. www.ft.com/intl/cms/s/0/002d4e10-ad8d-11e2-82b8-00144feabdc0.html.

Solnit,Rebecca.2010.AParadiseBuilt inHell:TheExtraordinaryCommunitiesThatArise inDisaster.NewYork:Penguin.

Soto, Onell R. 2011. “Big Desert Solar Farm Means Big Factory in S.D.” U-T San Diego.www.utsandiego.com/news/2011/mar/10/big-desert-solar-farm-means-big-factory-in-san-die.

Sprint. 2013. “iPhone 5 ‘I Am Unlimited: Picture Perfect’ Commercial.” YouTube.www.youtube.com/watch?v=C9qxjBlL3ko.

Spufford,Francis.2010.RedPlenty.London:FaberandFaber.

Stapleton, Jay. 2013. “Cloud Computing Trend Raises Ethical Issues.” Connecticut Law Tribune.www.ctlawtribune.com/PubArticleCT.jsp?id=1202609066339&Cloud_Computing_Trend_Raises_Ethical_Issues_&slreturn=20130608132308.

Steadman,Ian.2013.“BigDataandtheDeathoftheTheorist.”Wired.www.wired.co.uk/news/archive/2013-01/25/big-data-end-of-theory.

Steel,Emily.2012a.“DataScientistsTakeBiteOutofMadMen.”FinancialTimes.www.ft.com/intl/cms/s/2/db8d250e-4279-11e2-979e-00144feabdc0.html.

——.2012b.“TVCompaniesinDigitalAdFightback.”FinancialTimes.www.ft.com/intl/cms/s/0/bd47e5fc-4bb6-11e2-b821-00144feab49a.html.

Stewart, James B. 2013. “Looking for a Lesson in Google’s Perks.” New York Times.www.nytimes.com/2013/03/16/business/at-google-a-place-to-work-and-play.html.

Stoller, Jonathan. 2012. “Why Cold Canada Is Becoming a Hot Spot for Data Centres.” Globe and Mail.www.theglobeandmail.com/report-on-business/economy/canada-competes/why-cold-canada-is-becoming-a-hot-spot-for-data-centres/article6598555.

Streitfeld, David. 2013. “As Competition Wanes, Amazon Cuts Back Discounts.” New York Times.www.nytimes.com/2013/07/05/business/as-competition-wanes-amazon-cuts-back-its-discounts.html.

Sunyer, John.2013. “BigDataMeets theBard.”FinancialTimes.www.ft.com/intl/cms/s/2/fb67c556-d36e-11e2-b3ff-00144feab7de.html.

Sutter, John D. 2011. “iCloud: Revolution or the Next MobileMe?” CNN Tech.www.cnn.com/2011/TECH/web/06/07/icloud.reaction/index.html.

Swinhoe,Dan.2013.“GreenIT.”IDGConnect.www.idgconnect.com/blog-abstract/743/dan-swinhoe-asia-green-it-asia.

Szoldra,Paul.2013.“Google:IfYouUseGmail,You‘HaveNoExpectationofPrivacy.’”BusinessInsiderAustralia.www.businessinsider.com.au/gmail-privacy-google-court-brief-2013-8.

Takahashi, Dean. 2013. “It’s Crowded in Here: CES Attendance Tops 150,000.” VB.http://venturebeat.com/2013/01/12/its-crowded-in-here-ces-attendance-tops-150000.

Talbot,Chris.2013.“CloudOutages:PowerLossBlamedasMainCause.”Talkin’Cloud.http://talkincloud.com/cloud-computing-research/cloud-outages-power-loss-blamed-main-cause.

Talib,Nasim.2012.Antifragile:ThingsThatGainfromDisorder.NewYork:RandomHouse.

Tanaka,EdwardTessen.2012.“TheNSAandMilitaryCloudComputing:JustPaintingaCyberBullseyeforAttackers?”Pateixia.www.patexia.com/feed/the-nsa-and-military-cloud-computing-just-painting-a-cyber-bullseye-for-attackers-2401.

Tang, Han. 2013. “China’s Young Workers Fight Back at Foxconn.” Labor Notes.www.labornotes.org/2013/08/china%E2%80%99s-young-workers-fight-back-foxconn.

Taylor, Paul. 2013a. “CloudComputing IndustryCouldLoseUp to $35Bn onNSADisclosures.”Financial Times.www.ft.com/intl/cms/s/0/9f02b396-fdf0-11e2-a5b1-00144feabdc0.html.

——.2013b.“HessePutsSprintBackamongFrontrunners.”FinancialTimes.www.ft.com/intl/cms/s/0/0fe039f4-56b1-11e2-aad0-00144feab49a.html.

TeilharddeChardin,Pierre.1961.ThePhenomenonofMan.NewYork:HarperTorchbooks.

Tett,Gillian.2013.“BreakaWallofSilenceonCyberAttacks.”FinancialTimes.www.ft.com/intl/cms/s/0/d5b2464e-6648-11e2-b967-00144feab49a.html.

Thibodeau, Patrick. 2013. “Cloud Computing’s Big Debt to NASA.” Computerworld.www.computerworld.com/s/article/9237439/Cloud_computing_s_big_debt_to_NASA.

Thiel, Tamiko. 2012. “‘Clouding Green’ @ Zero1 Biennial.” Mission-Base. http://mission-base.com/tamiko/AR/clouding-green.html.

Thomas, Daniel. 2013. “Baidu Nets France Telecom Browser Deal.” Financial Times.www.ft.com/intl/cms/s/0/07812948-5d92-11e2-ba99-00144feab49a.html.

Tilahun, Gelila, Andrey Feuerverger, and Michael Gervers. 2012. “Dating Medieval English Charters.” Annals ofAppliedStatistics6,no.4:1615–1640.

Tunstall, Jeremy.1986.CommunicationsDeregulation:TheUnleashingofAmerica’sCommunications Industry.NewYork:Blackwell.

Turk, James. 2013. “Outsourcing toGoogle:ABadDeal forYorkAcademicStaff.”Presentation toYorkUniversityAcademicStaff.CanadianAssociationofUniversityTeachers.www.yufa.ca/wp-content/uploads/2013/02/2013.02-

Google-YUFA-final.pptx.

Tydeman,John,HubertLipinski,RichardP.Adler,MichaelNyhan,andLaurenceZwimpfer.1982.TeletextandVideotexin the United States. New York: McGraw-Hill. www.ft.com/intl/cms/s/0/07812948-5d92-11e2-ba99-00144feab49a.html.

Udell, Jon. 2012. “Is It Time to Mandate Cloud Storage Preservation?” Wired.www.wired.com/insights/2013/01/guaranty-associations-cloud-storage.

U.S.NationalEndowment for theHumanities,Office of theDigitalHumanities. 2013. “DigitalHumanities Start-UpGrants.”www.neh.gov/grants/odh/digital-humanities-start-grants.

U.S.OfficeofScienceandTechnologyPolicy.2012.“ObamaAdministrationUnveils‘BigData’Initiative:Announces$200 Million in New R&D Investments.” White House.www.whitehouse.gov/sites/default/files/microsites/ostp/big_data_press_release.pdf.

Vega, Tanzina. 2013. “Two Ad Giants Chasing Google in Merger Deal.” New York Times.www.nytimes.com/2013/07/29/business/media/two-ad-giants-in-merger-deal-chasing-google.html.

Verizon Wireless. 2013. “Verizon Powerful Answers—‘Suddenly: 60’ Commercial.” YouTube.https://www.youtube.com/watch?v=xpuoN6S3Efk.

Vina,Gonzalo,andSimonKennedy.2013.“FinanceChiefsEndorseCutsasReinhart-RogoffChallenged.”Bloomberg.www.bloomberg.com/news/2013-04-19/finance-chiefs-endorse-cuts-as-reinhart-rogoff-challenged.html.

Wainewright,Phil.2013.“CloudProvidersWorkingwithBigData.”ZDNet.www.zdnet.com/cloud-providers-working-with-big-data-7000013521.

Waldrop,M.Mitchell.2002.TheDreamMachine:J.C.R.LickliderandtheRevolutionThatMadeComputingPersonal.NewYork:Penguin.

Walker, Michael. 2013. “Data Science Is a Team Sport.” Data Science Central.www.datasciencecentral.com/profiles/blogs/data-science-is-a-team-sport.

Wallsten, Peter, Jia LinYang, andCraigTimberg. 2013. “Facebook Flexes PoliticalMusclewith ImmigrationBill.”WashingtonPost.www.washingtonpost.com/business/economy/facebook-flexes-political-muscle-with-carve-out-in-immigration-bill/2013/04/16/138f718e-a5e7-11e2-8302-3c7e0ea97057_story.html.

“The Warhol: Silver Clouds.” 2010.www.warhol.org/uploadedFiles/Warhol_Site/Warhol/Content/Exhibitions_Programs/Exhibitions/EX_20100903_TE_SilverClouds.pdf.

Warren, Tom. 2010. “Microsoft Shows Off Its ‘Cloud Power’ with New Advertising Campaign.” Win Rumors.www.winrumors.com/microsoft-shows-off-its-cloud-power-with-new-advertising-campaign.

Waters, Richard. 2012. “Oracle Takes to Cloud with Buying Spree.” Financial Times.www.ft.com/intl/cms/s/0/73a7facc-5043-11e2-805c-00144feab49a.html.

——.2013a.“BigIntelligencetoTackleCyberthreats.”FinancialTimes.www.ft.com/intl/cms/s/0/87ed6bc8-8105-11e2-9fae-00144feabdc0.html.

——. 2013b. “Google Search Proves to Be New Word in Stock Market Prediction.” Financial Times.www.ft.com/intl/cms/s/0/e5d959b8-acf2-11e2-b27f-00144feabdc0.html.

——. 2013c. “IBM Looks beyond the IT Department for Fresh Growth.” Financial Times.www.ft.com/intl/cms/s/0/9ecf5a64-d8cf-11e2-a6cf-00144feab7de.html.

——. 2013d. “Inside Business: Cloud Hangs over Old Guard of Business Software.” Financial Times.www.ft.com/intl/cms/s/0/33099ad8-b7e6-11e2-bd62-00144feabdc0.html.

Wegener, Al. 2013. “Big Data Plumbing Problems Hinder Cloud Computing.” Electronic Design.http://electronicdesign.com/communications/big-data-plumbing-problems-hinder-cloud-computing.

Weisinger, Dick. 2013. “Cloud Computing: Skills Gap Threatens Technology Boom?” Formtex Blog.www.formtek.com/blog/?p=3541.

Whittaker, Zack. 2012. “SamsungHikesAppleChip Prices by 20 Percent.”ZDNet.www.zdnet.com/samsung-hikes-apple-chip-prices-by-20-percent-report-7000007254.

Wiener,Norbert.1948.Cybernetics;or,ControlandCommunicationintheAnimalandtheBeast.NewYork:Wiley.

——.1950.TheHumanUseofHumanBeings.Boston:HoughtonMiffin.

Wiles,Will.2012. “BeforeFruitNinja,Cybernetics.”NewYorkTimes.www.nytimes.com/2012/11/30/opinion/the-no-

10-dashboard-and-cybernetics.html.

Wilhelm,Alex.2012.“Microsoft’sCloudVision.”NextWeb. http://thenextweb.com/microsoft/2012/12/13/microsofts-cloud-vision-how-azure-is-the-linchpin-to-the-firms-new-devices-and-services-corporate-strategy.

Wilson, James, and Barney Jopson. 2013. “Amazon Hit by Old World Strike Action.” Financial Times.www.ft.com/intl/cms/s/0/e4d3bdde-bc82-11e2-9519-00144feab7de.html.

Wilson, Karin. 2012. “Avoid Failure When Marketing Cloud Computing.” Cloud Computing Journal.http://cloudcomputing.sys-con.com/node/2307497.

Wilson,Valerie Plame, and JoeWilson. 2013. “TheNSA’sMetastasised Intelligence-Industrial Complex Is Ripe forAbuse.”Guardian.www.guardian.co.uk/commentisfree/2013/jun/23/nsa-intelligence-industrial-complex-abuse.

Wingfield,Nick,andMelissaEddy.2013.“InGermany,UnionCultureClasheswithAmazon’sLaborPractices.”NewYorkTimes.www.nytimes.com/2013/08/05/business/workers-of-amazon-divergent.html.

Winner, Langdon. 2004. “Resistance Is Futile: The Posthuman Condition and Its Advocates.” In Is Human NatureObsolete?editedbyHaroldBailieandTimothyCasey,385–411.Cambridge,MA:MITPress.

Winslow, George. 2013. “The Measurement Mess.” Broadcasting and Cable.www.broadcastingcable.com/article/494609-The_Measurement_Mess.php.

Wise, Bill. 2013. “Big Data’s Usability Problem.” All Things D. http://allthingsd.com/20130423/big-datas-usability-problem.

Wojtakiak, Mark. 2012. “Tag Archives: Deloitte Cloud Computing Forecast Change.” Storage Effect.http://storageeffect.media.seagate.com/tag/deloitte-cloud-computing-forecast-change.

Wolf, Gary. 2010. “The Data-Driven Life.” New York Times. www.nytimes.com/2010/05/02/magazine/02self-measurement-t.html.

Wolonick, Josh. 2012. “Are Apple and Google in Race for North Carolina’s ‘Black Gold’?” Minyanville.www.minyanville.com/sectors/technology/articles/Are-Apple-and-Google-in-Race/12/7/2012/id/46453.

Woodall, Angela. 2013. “Amazon Files Court Complaint over CIA Cloud Contract.” CRN.www.crn.com/news/cloud/240158953/amazon-flies-court-complaint-over-cia-cloud-contract.htm.

WorldEconomicForum.2013.“TheWorldEconomicForumLeadershipTeam.”www.weforum.org/content/leadership-team.

Wortham, Jenna. 2013. “Cisco Plans to Cut 4,000 Jobs as It Posts Profit Gain.” New York Times.www.nytimes.com/2013/08/15/technology/cisco-plans-to-cut-4000-jobs-as-it-posts-profit-gain.html.

Wyatt,Edward,andClaireCainMiller.2013.“TechGiantsIssueCallforLimitsonGovernmentSurveillanceofUsers.”New York Times. www.nytimes.com/2013/12/09/technology/tech-giants-issue-call-for-limits-on-government-surveillance-of-users.html.

Yafang, Sun. 2012. “Foreword.” InTheGlobal Information Technology Report 2012, edited by Soumitra Dutta andBeñatBilbao-Osorio,ix–x.Geneva:WorldEconomicForum.

Yang, Lin. 2013. “Foxconn Tries to Move Past the iPhone.” New York Times.www.nytimes.com/2013/05/07/business/global/foxconn-tries-to-move-beyond-apples-shadow.html.

Zhu, Julie. 2013. “Lanfang: China’s Cloud Computing Hub.” Financial Times. http://blogs.ft.com/beyond-brics/2013/05/15/langfang-chinas-cloud-computing-hub.

INDEX

Acaroglu,Leyla,128

AdobePhotoshop,40

AdvancedResearchProjectsAgency(ARPA),29–30

AdvancedResearchProjectsAgencyNetwork(ARPANET),31

advertising:agencies,56,91;Apple,88–90;campaigns,7–8,17,63;ofcloudcomputing,79–90; Facebook, 148–150,152;Google,152;IBM,86–88;onInternet-enabledtelevision,60–61;Microsoft,83–86,89,90;Sprint,135–136;SuperBowl,79–83,90

agility,101

airpollution,133

AlibabaGroup,72

Aliyunnetwork,72

Allende,Salvador,25–27,31

AllianceTechnologyGroup,70

Amazon,4,7,16;competitorsfor,51,60–65;environmentalrecordof,133,134–135;asindustryleader,46,48–51,56,57–58;laborissuesat,168–171;lobbyingby,117

AmazonCloudDrive,54

AmazonMechanicalTurk(AMT),170–171

AmazonWeb Services (AWS), 41, 48–51, 102; big-data analysis by, 177; Obama campaign and, 27–29, 177, 188;outagesby,50,98;U.S.governmentuseof,42,66–67

AmericanChamberofCommerce,116

AmericanFederationofTelevisionandRadioArtists,172

analytics,7,177–205.Seealsobigdata

Anderson,Chris,175,193–194,201

antivirussoftware,146

AOLTime-Warner,66

Apple,4,7,51;advertisingby,88–90;Chinesefactoriesand,161,162;chipmanufacturingand,58;competitorsfor,60;datacentersof,35;dominationby,46;environmentalrecordof,135;iCloud,2,8,41,54, 55, 88–90; as industryleader,54–55;laborissuesand,171;lobbyingby,115;pollutionby,132;renewableenergydevelopmentby,132

applications,7

AppStore,54

archives,47–48

Aristophanes,12,208,209–213,219–220

ARPA.SeeAdvancedResearchProjectsAgency(ARPA)

artisticmanifestations,14,205–226

art-sciencemovement,226

Ashby,WilliamRoss,22

AT&T,45,56,60,65–67,112–113

AtlasSolutions,150

austeritypolicies,203–205

Autodesk,156

autonomy,140

AWS.SeeAmazonWebServices(AWS)

Azure,41,51–52,72

backupsystems,9,130,133

Baidu,33–34,72–73

Barrie,Ontario,36–37

Barthes,Roland,93,198–199

BasilicaofSaintFrancisofAssisi,208–209

batchprocessing,30

Bateson,Gregory,22

Bauman,Zygmunt,37

Beer,Stafford,25,26–27

Benioff,Marc,63

big data, 2, 10–11, 121, 177–205; analysis of, 7, 177–205; as atheoretical, 181; cloud and, 177–179; concept of,179–180;correlations,180–181,195–196,199–200,205;definitionof,179;digitalpositivismand,196–205,206;fundamentalism, 180; government use of, 182–188; health care and, 188, 189; historical research and, 201–202;humanities research and, 190–193;mythology of, 193–196; as predictive, 181–182, 202; security and, 146–147;socialscienceresearchand,188–189;usesof,108–109,182–196

BigData(Mayer-SchönbergerandCukier),195

Bing,52

Binney,William,186

BirdCloud(Proulx),209

blogs,8,90–95,97

broadcastingcompanies,60–61

Brooks,David,200

Brussels,116

Brynjolfsson,Erik,167

Burke,Edmund,82

Burtynsky,Edward,130

businessmarket,53–55,98

BusinessSolutions,93–94

Cameron,David,28,29

CaminodeSantiagodeCompostela,119

Canada:asdatacenterlocation,36–37,131;privacyprotectionsin,151,153–155;U.S.surveillanceand,153–155

CanadianAssociationofUniversityTeachers(CAUT),153–154

capitalism,5;friction-free,2,176;global,59,176;informational,2,32,58–59,175–176;surveillance,10,141

carbonemissions,130

Carnegie,Dale,209

cartels,58–59,60

CatherineofSiena,214

causality,200,205

CBS,60–61

CenterforDigitalDemocracy,148

CentersforDiseaseControl,182

centralcomputers,3

CentralEconomicMathematicalInstitute,23

CentralIntelligenceAgency(CIA),23–24,31,42,50,67,70,177,186

CERN,105

certificationprograms,74

TheChallengeoftheComputerUtility(Parkhill),3,18

Chappuis,Bertil,101–102

chemicalbatteries,9

Chengdu,73

Chicago,37

chieffinancialofficers(CFOs),94

chiefinformationofficers(CIOs),94

childlabor,161

Chile,6,25–27,31

China:censorshipin,38;cloudcomputing in,7,71–75,151;cyberattacksfrom,5,69, 141–142, 144; e-pollution in,127, 130, 132; labor issues in, 74–75, 157–159, 167–168; manufacturing in, 159–163; Microsoft and, 179;surveillanceby,150–151

ChinaTelecom,73

Chromebooks,52,183

Cisco,7,34–35,156,165

citizenship,42–43

Clark,Jack,44–45

Clearwire,66

climatechange,225

cliodynamics,201–202

CloudAppreciationSociety,208

CloudAtlas(Mitchell),2,13–14,208,220–224

cloudcities,73

CloudCity(Saraceno),14,225–226

CloudCollectorsHandbook,208

cloudcomputing:anatomyof,32–39;bottom-up,4;characteristicsof,6–7,38–39;competition in,51, 56–57, 60–65;conception of, 1; cost savings from, 70, 101, 163–164; cultural significance of, 10–14; definition of, 6, 16–18;environmentalissueswith,9,14,74,127–137;growthof,34–35,175–176;hypeover,5; implicationsof,4–5; asmarket-driven,42–43;materialityof,37,77,129;meteringof services,45;originsof,5–6, 15–18; philosophicalassumptionsof,2;predecessorsto,6,18–32;pricing,56–57,62;problemsfacing,9–10;promotionof,7–9,77–122,135–136;publicconfusionover,127;regulationof,45–47,56,60;securityissuesin.seesecurityissues;typesof,7,39–42;asutility,18–21,44–48

CloudComputingExpo,9,119–122

cloud computing industry: company leaders in, 48–65; issues facing, 110–111; rights and responsibilities of, 179;secretivenatureof,134.Seealsospecificcompanies

CloudComputingJournal,93

cloudculture,11–14,176,205–226

clouddatacenters.Seedatastoragecenters

CloudingGreen(Thiel),14,226

CloudLock,69

cloudmetaphor,11–12,16,77,206–209,212,214,225

cloudofknowing,216

TheCloudofUnknowing,2,12–13,208,213–219,220

TheClouds(Aristophanes),12,208,209–213,219–220

CloudSecurityAlliance(CSA),145–146

CloudSwitch,65

CloudtotheEdge(COE),69–70

CloudTweaksblog,91–92

coal-firedpowerstations,134–135

coalindustry,131

Coca-Cola,141–142

Cohen,JulieE.,139

commercialmodel,42–43

commercialproviders,toU.S.government,66–71

communicationflows,19

communicationprocess,223–224

communicationscompanies,67,73,104–105.Seealsospecificcompanies

CommunicationWorkersofAmerica,172

community,42

communityclouds,41–42

communitygridprojects,4

Compaq,15

ComputerDealerExhibition“COMDEX”,118

computerindustry,159

computerutility,6,20;cloudas,44–48;componentsof,20;conceptof,18–21

consultingfirms,8,95–103

ConsumerElectronicsShow“CES”,118

consumermarket,42,54,60

context,fordata,201,202

contractworkers,174

Cook,Tim,59

coolingsystems,131

corporatetradeshows,112,118–122

correlationalanalysis,11,180–181,194–196,199–200,205

Crawford,Susan,46

crowdsourcedworkers,170–171

CubanMissileCrisis,23

Cue,49

culturalimpact,10–14

culturalstudies,4

culture:cloud,11–14,176,205–226;ofknowing,32,176–177

customer-relationshipmanagement“CRM”,63,102

cyberattacks,9–10,63,141–144,146,165;defenseagainst,69–70

cybernetics,6,19,22–25,31

DARPA.SeeDefenseAdvancedResearchProjectsAgency“DARPA”

data:errorsin,204–205;lossof,145;migrationof,tocloud,2;preservation,47–48;qualitative,196–197;quantitative,180,191,196–198;securityof,50–51,108,116;user-generated,108–109,147–150,179.Seealsobigdata

dataanalysis,10–11,67–68,70,180–205

data-intelligencecomplex,183

DataKind,189

Datalogix,149

datascientists,194–195

datastoragecenters,2–3,6,7,129;ofAmazon,49;ofApple,54;inChina,71–75;cyberattackson,144–146;energyconsumptionby,125–126,130–134,136;lackofdowntimefor,9;locationof,10,35–37,71–72,131,134;powerneedsof,130–134;securityfor,35–36;sizeandgrowthof,33–35;stressonenvironmentfrom,124–127,129–137;virtualizingof,164

Davis,Erik,215

Davosconference,8,103,112

DEC,159

DefenseAdvancedResearchProjectsAgency“DARPA”,29–32,69–70,71,118,186,187

DefenseInformationSystemsAgency,70

DeLillo,Don,12,218

Dell,16,61,62

Deloitte,8,96–97

democracy:cloudcomputingand,28,42–43;worker,26,27

denial-of-serviceattacks,143

deregulation,117

Dery,Mark,215

developingcountries,e-wastedumpingin,130

dieselgenerators,9,124–125,130,133

digitalhumanities,190–193

DigitalHumanitiesInitiative,190

DigitalOcean,145

digitalpositivism,2,10,11,32,196–205,206,214

digitalsublime,5,106,108,192

digitalworld,136

DishNetwork,66

Disney,54

distributedprocessingmodel,43

DOD.SeeU.S.DepartmentofDefense“DOD”

domesticspying,185–186

Dorsey,Jack,44

downtime,9,133

drones,186–187

Dropbox,54,73,177

DukePower,132

Easternphilosophy,13

editorialstaff,166

Edmonton,Alberta,39

educationsector,7,68,166

Eisenhower,Dwight,29,183

elasticity,38

electricalgrid,9

electricity,32–33,46–47,130

electronicassemblyplants,159–163,167–168

ElectronicPrivacyInformationCenter,147

electronicsurveillance,74,153–155,182–188

elementaryschools,166

Ellison,Larry,64

email,153–154

emergentthinking,201

TheEmpireofLight“Magritte”,14,206

employmentissues,155–174

energy:costs,36;sources,9,35;usage,74,125–126,130–134,136

environmentalissues,9,14;energyusageand,74,125–126,130–134;e-pollution,127–137,159;wirelessaccessand,132

EnvironmentalProtectionAgency,159

epistemology,13

e-pollution,127–137,159

Ericsson,104

espionage,182–183

ethics,189

EuropeanUnion“EU”:lobbyingof,116–117;privacyprotectionsin,152–155

Facebook,7,10,44,51;advertising,148–150,152;environmentalrecordof,134;GraphSearch,140,148–149,183;as

industryleader,46,55,57–58;lobbyingby,114–115,118;privacy issuesand,138,140; user-generated data and,147–150

FederalTradeCommission“FTC”,114,117,148

Ferguson,Niall,204

Fish,Stanley,191–193

Flickr,2

Flores,Fernando,25

flu,spreadof,181–182,197,202

flywheels,130,133

ForeignIntelligenceSurveillance

AmendmentAct“FISA”,153,154,155

foreignworkers,114,115

ForresterResearch,8,97–99

Foxconn,74–75,159–163,171

FranceTelecom,72

Franklin,Seb,214–215

freecoolingtechnology,36

FreelancersUnion,174

friction-freecapitalism,2,176

Friedman,Milton,117

FTC.SeeFederalTradeCommission“FTC”

fullemployment,167

FWD.us,114–115

GartnerGroup,8,99–100,164

Gates,Bill,1–2,51,89,176

GeneralElectric,30,112,178

genomics,182,188

Germany,168–169,172

GertrudetheGreat,214

Giotto,208–209

Giridharadas,Anand,156

Glanz,James,123,126

globalcapitalism,59,176

globalinformationeconomy,1–2

GlobalInformationTechnologyReport2012,103–111

globalsupplychains,132,151,157–174

Gmail,41,153–154

Gigaom,94–95

Google,7,16,51,177;advertising,152;anti-competitivepracticesof,114;

Google“continued”:AppEngine,41;big-data analysisby,181–182, 183, 202; employment at, 156–157;Gmail, 41,153–154;asindustryleader,46,52–54,56,57–58,59;lobbyingby,114,117,118;asvisionary,194

GoogleAppEngine,53

GoogleComputeEngine“GCE”,52–53

GoogleDrive,54

GoogleFluTrends,182

GoogleGlass,52

GoogleReader,46

Gore,Al,11,215

government:local,115;privatesectorand,66–71;regulation,45–47,56,60;state,115;systems,42;useofbigdataby,182–188.SeealsoU.S.government

GraphSearch,140,148–149,183

GreenMountainDataCentre,36

Greenpeace,14,132,134–135

greenproducts,129

gridcomputing,43

Grosch,Herb,18

Groucutt,Peter,1

“GrowthinaTimeofDebt”article,203–205

hackerattacks,5,9–10,141–144,146,165

hardware,57–58

Hartzog,Warren,139

HarvardUniversityProgramon

InformationResourcesPolicy,19

healthcare,188,189

hegemony,102–103

Herndon,Thomas,204

Heroku,64

Hilton,Walter,214

historicalresearch,201–202

Hoffmann-LaRoche,53

Hollywoodmediaindustry,172

Homer,208

HP,15,56,58,61,131–132,162

HSBC,164

httpsprotocol,144–145

Huawei,72,104–106,111,142,159–161

humanbehavior,forecasting,182–183

humanimagination,12

humanities,68,190–193

humanresourcesdepartments,166

hybridcloud,7,40,41,49,120,121

hydroelectricpower,124,126,131

HyperconnectedWorldInitiative,106

hyperconnectivity,106–107

hypotheses,181

IBM,30,61–62,73,159,177;advertisingby,86–88;inChina,151;datacentersof,36;governmentcontractsand,67;

governmentpressureon,60;laborissuesand,171;legacysystems,56;SmartCloud,87–88,216;SmarterAnalyticsproject,178;transitiontocloudby,7,8,34;WorldCommunityGrid,43

iCloud,2,8,41,54,55,88–90

Iliad“Homer”,208

India,132,164

industryleaders,48–65.Seealsocloud

computingindustry;specificcompanies

informationalcapitalism,2,32,58–59,175–176

informationgrowth,11

informationprocessing,19

informationresources,19–20

informationsociety,environmentand,128–129

information technology, 5; budgets, 70; cloud-based, 61–62; departments, 2, 10, 163–167; enthusiasm for, 106;environmentand,127–137;industry,9,151–152,155–174;laborissuesin,155–174;professionals,158,163–167;security,146

infrastructureasaservice“IaaS”,39

innovator’sdilemma,62

INSEAD,104,106

Instagram,49

Intel,57–58,156

intelligencesector,7,68–71,182–188

internationalpromotion,103–111

InternationalTelecommunicationsUnion,111

Internet,6;cloudcomputingandthe,16–17;developmentofthe,29–32,215

Internet-enabledtelevisions,60–61

Internetprotocol(IP)traffic,35

invisiblehand,ofmarket,45,46

Iran,5,143

Isaacson,Walter,78

Israel,142

iTunesMatch,54

iTunesStore,54,60

iWeb,54–55

Jakobson,Roman,22

Jassy,ArthurR.,15

Jobs,Steve,54,59,78,88,89

journalism,166

JulianofNorwich,214

Kelly,Kevin,136,215,217

Kennedy,JohnF.,24

Kennedy,Robert,23

Kerry,John,114

Klüver,Billy,225

knowledge,207;cultureofknowingand,176–177;tacit,26

knowledgeworkers,10,75,172–174

Kohnstamm,Jacob,116

Krugman,Paul,205

Krushchev,Nikita,22

Kudryashov,Roman,198

Kurzweil,Ray,192,194,216,217

Kusnetzky,Dan,163

labor issues, 155–174; in China, 74–75, 157–159, 167–168; knowledge workers, 163–167; unions, 162, 168–169,

171–174;workerresistance,167–169

Langfang,34

Lanier,Jaron,198

Latour,Bruno,212,224

layeringsystems,145–146

lead-acidbatteries,130,133

Lee,KevinB.,82

legacysystems,62

legitimacy,8–9,91,93,97,105,111,195

librarians,166

LibraryofCongress,34,47

Licklider,J.C.R.,30

Linthicum,David,18

liquidmodernity,37

Liu,Alan,175

lobbying,9,112–118

localgovernment,115

Loosecubes,98

LosAlamosNationalLab,69

Lynch,Michael,139–140

Lysenko,Trofim,22

MacIntyre,Alisdair,196

MacyFoundation,22

Magritte,René,14,206

Malaysia,159

marketcontrol,57–60

market-drivenmodel,42–43

marketing:ofthecloud,7–8,91,93–94;hype,5,61,88,96,100–101.Seealsopromotion

Marx,Leo,82

massiveopenonlinecourses(MOOCs),166

materiality,ofcloudcomputing,37,77,129

Maxwell,Richard,127,128,137

McCarthy,John,18

McKellar,Campbell,98

McKinsey&Company,8,100–103

McLuhan,Marshall,216

measuredservicebilling,7,38–39

medicine,188

MeePo,73

metadata,70

metaphysics,13

methanegas,35

MetropolitanMuseumofArt,225

Microsoft, 7, 8, 156; advertising, 83–86, 89, 90; Azure, 41, 51–52, 72; bigdata analysis by, 183; China and, 179;competitorsfor,53,60,63–64;crashexperiencedby,130,144–145;datacentersof,124–127,129;EUand,152;governmentcaseagainst,60;asindustryleader,46,51–52,56,57–58;lobbyingby,115,118;pollutionby,124–126

MiddleEast,72

military-industrialcomplex,183

militarysector,7,68–71

Miller,Toby,127,128,137

Mitchell,David,2,13,208,220–224

MobileMe,88

MobileWorks,171

monopolies,60

Monsanto,178–179

moralphilosophy,13

Moretti,Franco,193

Morozov,Evgeny,107–108

Multics(MultiplexedInformationandComputingService),30

mutualconstitution,200

mythology:ofbigdata,193–196;ofthecloud,78,80–81,206;oftechnology,106–108,213

narrative,201

NASA,66,68

NationalEndowmentfortheHumanities(NEH),68,190–191

NationalInstituteofStandardsandTechnology(NIST),17–18,68,122,179

NationalInstitutesofHealth,188

NationalNuclearSecurityAdministration,69

NationalScienceFoundation,18,68,188

NationalSecurityAgency(NSA),7,10,70–71,142,144,153,182–188

naturaldisasters,36

naturalmonopolies,19,20

naturalsublime,81–82

NavalSupplySystemsCommand,66

NavalWarCollege,69

Negroponte,Nicholas,136

nephelococcygia,206–207

NetApp,156

Netflix,49,50

Netscape,60

networkaccess,38

networkredundancy,54

networktelevision,60–61

networkeddatacenters,1–2

NewsCorp.,54

newsletters,8

NewYorkTimes,195

NIST.SeeNationalInstituteofStandardsandTechnology(NIST)

No.10Dashboard,28,29

Noble,David,217

Nokia,104

noosphere,11,215

NorthAtlanticTreatyOrganization(NATO),36

NorthCarolina,35,115,131

NorthKorea,59

Norway,36

nuclearwaste,130

Obamapresidentialcampaign,4,27–29,48,177,188

Oettinger,Anthony,19

Office365,52,53

OfficeofDigitalHumanities(ODH),68,190–191

offshoreproduction,159–160

oligopolies,60

Omnicom,56

on-demandentertainment,140

on-demandself-service,38

onlineeducation,166

onlinepiecework,170–171

onlinesurveillance,74,153–155,182–188

OpenDataCatalogue,39

OpenStack,62,66–67

Oracle,7,56,64

organizedlabor,162,168–169,171–174

O’Sullivan,Sean,15–16

outages,50,53–54,98

outsourcing,10,158,163–167

OVH.com,37

Ovum,109

packet-switchingtechnology,31

Pacnet,72

Page,Larry,59

Palmer,Maija,147

Pandora’sHope(Latour),212

Parkhill,Douglas,3,18,20–21

Patches,Matt,82

patternrecognition,146,181–182

Pavlov,Ivan,22

Pentagon,29–32,70

perfection,13,88–90

personalcomputer,3,57

ThePhenomenonofMan(Teilhard),215

philosophy,12,13

pigmanure,35

pilgrimages,119–120

PiltdownMan,216

Pinterest,49

PlannedObsolescence(Fitzpatrick),192

PlanX,70

platformasaservice(PaaS),39

politicaleconomy,4,176

pollution,74;air,133;fromdatacenters,124–127;e-pollution,127–137

Porat,MarcUri,19

post-humanism,85–86

Porete,Marguerite,214

poweroutages,50,53–54,98

powersources,9,35

powerusage,125–126,130–134,136

predictions,181–182,202

predictivepolicing,197–198

pricecuts,56–57,60

pricing,62,101

Prism,70,187

privacyissues,9–10,116,137–155;electronicsurveillance,70–71;governmentsand,150–155;importanceof,139–140;rethinking,107–108;surveillance,74,153–155,182–188;tradeoffs,138–139;usergenerateddataand,147–150

PrivacyRightsClearinghouse,143

privateclouds,7,40,41,69,74,83,143,188

productivity,167

professionals,165–166

ProjectCybersyn,25–27,31

promotion, 77–122, 135–136; advertising, 79–90; by blogs, 90–95, 97; international, 103–111; lobbying, 112–118;mythologyand,78;byprivatethinktanks,95–103;throughtradeshows,118–122

Proulx,Annie,209

public,42

publicclouds,7,41;securityand,50–51;U.S.governmentand,66–71

Publicis,56

publicutilities,18–19,45–46

qualitativedata,11,196–197

quantitativedata,11,180,191,196–198

Quincy,Washington,123–127

Rackspace,7,41,49,62,156,157

RangeTechnology,73

Raytheon,146–147

reason,212

reboundeffect,129

reliability,50,53–54

religioussublime,81

renewableenergy,132,135

Reputation.com,40

researchfirms,8,95–103

resourcepooling,38

rhetoric,212

RIM,104

Riot(RapidInformationOverlayTechnology),146–147

robotics,165

Sadowski,Jason,139

Salesforce,6,7,10,40,41,62–64,156;energyusemonitoringby,132;SuperBowladof,8,80–81,82–83,90

Samsung,58,104,115

SAP,64,73,83

Saraceno,Tomás,14,225–226

SAS,156,157

Scandinavia,36,37,72,131

Schiller,Herbert,19

Schlesinger,Arthur,Jr.,23–24

Schmidt,Eric,16,59

schools,166

sciences,182,193–194

scientificmanagement,168

ScreenActorsGuild,172

searchengines,182–183

secondaryschools,166

securitycertificates,144–145

security issues,2,9–10,63,74, 79, 108, 137–155, 165; cyber attacks, 9–10, 63, 69–70, 141–144, 146, 165; for datacenters,35–36,116;dataloss,145;governmentsand,150–155;internationaldatacentersand,131;militarysectorand,68–71;privatecloudsand,40;publiccloudsand,50–51;tradeoffs,138–139

Selinger,Evan,139

semiconductorchips,57–58

sensors,168

servers:energyconsumptionby,131–132;inefficientoperationof,132–133

Shannon,Claude,19,22,223

Shazam,49

Shenzhen,160,167

SierraLeone,189

TheSignalandtheNoise(Silver),203

SiliconValley,113,117–118,127,133,159,183

Silver,Nate,199,203,223–224

SilverClouds(Warhol),225

Singapore,159

SkyDrive,60

SloanDigitalSkySurvey,182

SmartCloud,87–88,216

“smartcloud”campaign,8

smartdevices,177–178

smarttelevisions,60–61

SmarterAnalyticsproject,178

Smith,Eve,209

Smith,Huston,218

Smith,Zadie,219

Smythe,Dallas,19

Snowden,Edward,142,153

socialmedia,7,8,60,64;data,108–109;growthof,9;privacyissuesand,147

socialsciences,68,188–189,197–198

sociology,4

Socrates,12,209–212

SoftBank,66

software,7

softwareasaservice(SaaS),40

softwareengineers,156–157,164

SoftwareStrategiesResearch,97

solarfarms,127

Solnit,Rebecca,81–82

solutionism,107–108

SoutheastAsia,159

SovietAcademyofSciences,22

SovietUnion,6,22–25,29,31

spam,170

Spielberg,Steven,82

Spotify,49

Sprint,66,135–136

Spufford,Francis,24

Sputnik,29

Square,44

Stalin,Joseph,22

StandardOil,60

statelegislatures,115

stock-marketforecasting,202

TheStorageEffect,97

Stretch-Harvest,183

structuralunemployment,158

studentlabor,162

Stuxnetmalware,142,143

subjectivity,198

sublime,81–82,90,106,108,135–136,192,193,219

Summers,Lawrence,167

SuperBowlads,8,17,63,79–83,90

supercomputers,43

surveillance,70–71,74,153–155,182–188

surveillancecapitalism,10,141

surveillancestate,7,10,141

Switchcorporation,34

Switzerland,37

tacitknowledge,26

Talib,Nasim,200

Tam,Jimmy,98

Target,28

taxbreaks,115

Taylor,FredericWinslow,168

TCP/IPnetworkprotocol,31

technicalconvergence,111

technicism,111

technocracy,156–157

technologicalsublime,32–33,75,82,90,106

technology:mythology of, 78, 80–81, 106–107, 108, 213; packet-switching, 31; social relations of, 27; as sublime,81–82.Seealsoinformationtechnology

technologystudies,4

TeilharddeChardin,Pierre,11,192,215–217,219

telecommunicationscompanies,65–66,73,104–105,112

telegraph,140

telephone,20,140

teletext,6

television,60–61,140

Telidon,3

temporaryworkers,173,174,202–203

Tencent,72

Terremark,65,66

terrorism,151

theories,181,195,201

Thiel,Tamiko,14,226

thinclientinterface,40

thinktanks,95–1033M,88

TianfuSoftwarePark,73

time-sharingmodel,18,30

T-Mobile-MetroPCS,66

Torre,Sergiodela,82

“TotheCloud”advertisements,8,84–86,90

toxicwaste,127–137,159

tradeshows,9,112,118–122

tradeunions.Seeunions

transnationalunions,172–174

trust,140

Turkers,170–171

Turkopticon,171

24/7operation,9,130,131,132–133

Twitter,44,47,51,134,178,197

Udell,Jon,47

Underworld(DeLillo),218–219

unemployment,158

Unesco,103

UNIGlobalUnion,172–173

unions,162,168–169,171–174

UnitedStates:cyberattacksby,142;privacyprotectionsin,151–155

USAPATRIOTAct,37,38,152,153–154

U.S.DepartmentofDefense(DOD),39,67,69–70,113,187

U.S.DepartmentoftheInterior,53

user-generateddata,108–109,147–150,179

U.S.government,7,17;demonstrationprojects,68;privatesectorand,66–71

U.S.military,182–188

UtahDataCenter,184–186

utilitymarkets,44–45

Ver.di,168–169,172–173

Verizon,65–66,88,142,171

Vernadsky,Vladimir,216

videotex,3,6,21,31

videoviewinghabits,140

Vietnam,159

VisualPageproject,190

VMware,7,62–63

vonNeumann,John,22

wages,167

Walmart,28,168–169

Warhol,Andy,225

warrantlesswiretapping,154

Weaver,Warren,19,223

WesternUnion,45

WhiteNoise(DeLillo),218

Wiener,Norbert,22,25,158

Wilder,Thornton,211

Wilson,DavidSloan,215

Windows8,60

WindowsLive,52

windpower,131

Wipro,132

wirelessaccess,132

Wolfe,Tom,11,215

Woolf,Virginia,82

workerassociations,173–174

workerdemocracy,26,27

workplaceconditions,inChinesefactories,160–162

WorldCommunityGrid,43

WorldEconomicForum(WEF),8–9,103–111

XboxLive,52

XDATAprogram,187

Yahoo!,125,131

Yangquandatacenter,34

Yoo,Richard,62

YorkUniversity,Gmailin,153–154

youth,213

Zuckerberg,Mark,44

ABOUTTHEAUTHOR

VincentMosco(PhD,HarvardUniversity)isProfessorEmeritusofSociologyatQueen’sUniversity,whereheheldtheCanadaResearchChairinCommunicationandSociety.Dr.Mosco is the author of numerous books and articles on the media and informationtechnology,includingThePoliticalEconomyofCommunicationandTheDigitalSublime:Myth,Power,andCyberspace.