THE LUDDITES VERSUS
THE GEEKS
Richard S. Wolff, Ph. DMontana State University
Department of Electrical and Computer [email protected]
406 994 7172October 2, 2003
Talk Outline
• How technologies evolve• Innovation and the shaping of the
telecommunications industry• A few considerations of what we might
expect in the future
First a few definitions
• Luddite: One who opposes the introduction of new technology, esp. into a place of work.
• A member of an organized band of English mechanics and their friends, who (1811-16) set themselves to destroy manufacturing machinery in the midlands and north of England.
Source: Oxford English Dictionary
First a few definitions
• Geek:A person who is extremely devoted to and knowledgeable about computers or related technology
slang (chiefly U.S.). a. orig. Eng. regional (north.). A person, a fellow, esp. one who is regarded as foolish, offensive, worthless, etc.
Source: Oxford English Dictionary
0
0.2
0.4
0.6
0.8
1
1.2
A historical example of progress:Steam ships replace sail ships
Frac
tion
of g
ross
tonn
age
built
in y
ear
1830 19401880Source: Historical Statistics of the US
~ten years
What were people thinking?
The luddites• Steam is dangerous• Where will we get the
fuel?• Steam engines are
unreliable• What will happen to
all the sail makers?
The geeks• Faster• More efficient• Predictable delivery
times• Easier to maneuver• Bigger ships
Another example
And probably the same arguments!Source: TFI, Inc.
A more familiar example from telecom
Source: TFI, Inc.
New technology has led to improved efficiency- employees per 10,000 access lines
But its about more than only cost avoidance…
• New technology can change the landscape, enable new applications– Stored program control switches enabled value-
added services• Call re-direction, 800 numbers, etc
– Lower technology costs enable new market entrants and change the industry structure
• Competition• Intelligence at the customer premises
Even the luddites benefit from innovation
“You’ve got mail”
The impact of sustaining and disruptive technology change
Performance demanded at low end of market
Performance demanded athigh end of market
Sustaining Technology
Prod
uct
perf
orm
ance
Disruptive Technology
Time
Source: Clayton Christensen, “The Innovator’s Dilemma”
Technology innovation stimulates new applications10,000
75 80 85 90 95
14 inch drives
Mainframes
8 inch drives
Mini computers 5.25 i
nch dr
ives
3.5 in
ch dri
ves
Desktop PCs
Notebo
ok PCs
YearSource: Clayton Christensen, “The Innovator’s Dilemm
1000
Ave
. dis
c ca
p aci
ty, M
B
100
10
1
a
Will voice over packet replace circuit-switched voice?
Packet networking is a disruptivedisruptive technology• The Geeks
– All information is digital
– Packets are more efficient
– Voice/data convergence
– The Internet will rule– New applications
without limit
• The Luddites– Circuit switching
provides high service quality
– Circuit switches are reliable
– Packet networks are vulnerable to abuse
– Huge capital investment to protect
Another innovation: will fiber replace copper?
Forecast made in 1989
What were we thinking?
So who was right this time?
• The Luddites– No demand for more
bandwidth– Installation costs are
too high– Huge capital
investment in outside plant
– Moore’s law can’t go on for ever
• The Geeks– High def TV is coming– Broadband ISDN will
solve everything– Opto-electronics,
semiconductors and fiber are on a very steep learning curve
Maybe the Luddites were right?
So what is different now?
• The Internet - applications we never dreamed of!
• Sustained technological progress in electronics and optics
• Strong evidence for high speed access (DSL, cable modems, etc.)
• A “critical mass” effect?
Continued growth in Internet usage-
In spite of the burst in the Telecom bubble!!
New applications are hard to predict
“My god! There’s been a terrible accident in our Chicago office”!”
March of progress in electronics1987 view
Will Moore’s law continue?
• The Luddites– Limit on narrow line
widths on chips– Too much complexity
on a chip – can’t be properly modeled and designed
– Power consumption will lead to melt down
– Too many input/output leads
• The Geeks– User shorter
wavelengths for drawing lines
– Reduce voltage to lower power
– Use optical interconnects
– Build in redundancy to improve reliability
And the winner is….Progress in semiconductor electronics,2002 view
Sustained progress forecast over the next decade
What is the next killer application?
South Korea leads the way
• Lee Yong Kyung, CEO of Korea Telecom, "The killer application of the Internet is speed. The money is in the pipes."
Country Percentage of homes
South Korea 57USA 25
Internet homes with broadband access
Source: NY Times, May 3, 2003
Broadband access demand is global
Ranking Country Total DSL lines (millions)
1 Japan 7.0
2 USA 7.0 (+ 14M cable)
3 South Korea 6.7
4 Germany 3.6
5 China 2.9
Source: DSL Forum, June 2003
The “critical mass” effect
• Malcolm Gladwell: “The Tipping Point”– Change is contagious- a few people can infect a
large population– Subtle changes can have a big effect– Several factors can combine to make a sudden
and dramatic impact
The geeks can spread an epidemic among the luddites
Analysts’ Forecasts for U.S. FTTH
Forecast made in 20032M
FTTHFTTHTotalTotal
1M
0
EPONEPONFTTHFTTH
1M ~ 2% ofBB U.S. HH
in 2006
2000 2001 2002 2003 2004 2005 2006
Sources: CIR, FTTH Council, InStat/MDR, Probe Research
And one more thing to look out for…
WirelessWireless
A few prophetic remarks from the past
• The “Negreponte Switch”:– only mobile services should be wireless;
stationary services should be wired. eg, relegate TV broadcasting to cable. ...
• The impact of billions of embedded processors– “Things that think – link”
Nicholas Negreponte, MIT Media Lab ( a famous geek)
The migration of voice to wireless
• This war is over, the tipping point has occurred
• But what’s next – will data (and the Internet) move to wireless?
Wireless Internet: top down or bottom up?
•Windows or Linux?
–Hierarchical project managers or egalitarian hackers?
•Cellular orWiFi?
Eric Raymond
Wireless data –when will it happen?
• Top down: the cathedral builders– First there was CDPD– Then SMS, GPRS, EDGE, 1XRTT, 1XEVDO,
3G….• Meanwhile in the bazaar
– Will WiFi make a difference?
Wireless chipsets make WiFi cost effective?
Intel introducesCentrino chips
802.11g stdadopted
Source: Pyramid research
Intel generates momentum for WiFi
WiFi access complements good food!
A “war drive” of Bozeman41 WiFi access points
Bottom up innovation:use of WiFi for fixed Internet access
A few things that will be important in the next decade
• Embedded processors– There is a chip in everything– XXX microprocessors manufactured in 2002– Chips replace bar codes…..
• And ad hoc networks to link all these gadgets– WiFi– Ultra wideband
A few things that will be important in the next decade
• Adaptive, network-distributed software– Software-defined radios: adaptive waveforms,
modulation– Application software downloaded over the
network• QUALCOMM “BREW” environment in use now• software agents act on users’ behalf
A few things that will be important in the next decade
• The all-optical network– Wavelengths get cheaper, closer together– Opto-electronic chips replace discrete
components– Ultra-dense WDM: A unique wavelength
replaces an IP address or phone number???
25 Tbps
1991 1994 1997 2000 2003 2006 2009 2012
1 Gbps
10 Gbps
100 Gbps
1 Tbps
10 Tbps
100 Tbps
100 Mbps
High bit ratesystems
High channelcount systems
TDM
DWDM
Aggregate fibercapacity
Bit rateper channelOC-48
OC-12
OC-192
OC-768
160 λs
>1000 λs
>200 λs
Can this trend continue unabated??
A few hidden assumptions- its not just about technology
• In the next few years we will need to:– Resolve video and audio ownership/sharing
intellectual property issues– Understand the economics for pricing & billing
of broadband services– Clear up some of the regulatory and
competitive issues that are constraining investment
Economic trends suggest long-term telecom demand should remain robust
Telecoms Servicesas % of US GDP
-
0.5
1.0
1.5
2.0
2.5
3.0
3.5
1940 1950 1960 1970 1980 1990 2000 2010
%
Telecoms Services (Adjusted to GDP)
100
1,000
10,000
100,000
1,000,000
1940 1950 1960 1970 1980 1990 2000 2010
$ (
Lo
g S
cale
)Comms spending is ~3% of GDP & growing @1.4%/yr (not including web /internet services) GDP/capita
growing ~6%/year
This is just the beginning!
Source: Rainer Malaka, EMLICDE 2001
And a final caveat
• Predictions are very difficult to make, especially when they are about the future
Attributed to Mark Twain
But maybe it was Yogi Berra?
“All I’m saying is NOW is the time to develop the technology to deflect an asteroid”
A cautionary comment to the luddites
Top Related