Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016
-
Upload
corinium-coriniumglobal -
Category
Data & Analytics
-
view
436 -
download
0
Transcript of Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016
![Page 1: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/1.jpg)
The Moral Dimension – Grappling with Ethics in the Age of Big Data
Chief Data Scientist, USA
![Page 2: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/2.jpg)
Introduction
Bennett B. BordenChief Data ScientistChair, IGED GroupDrinker Biddle & [email protected]
![Page 3: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/3.jpg)
3
Information is Overwhelmingly Electronic
![Page 4: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/4.jpg)
4
And is coming from more sources than ever
![Page 5: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/5.jpg)
5
![Page 6: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/6.jpg)
6
Insight into Human Conduct to Unparalleled Degree
![Page 7: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/7.jpg)
7
Unparalleled Market Efficiency The Right (Public or Private) Product or Service
- Right time
- Right place
- Right consumer
- Right cost
- Right price
![Page 8: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/8.jpg)
8
Unparalleled Potential for Disruption and Misuse
![Page 9: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/9.jpg)
9
![Page 10: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/10.jpg)
10
![Page 11: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/11.jpg)
11
![Page 12: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/12.jpg)
12
Differential Pricing – Disparate Impact
Algorithm used to set prices for online SAT tutoring by The Princeton Review showed customers in high density Asian neighborhoods were charged more.
![Page 13: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/13.jpg)
13
Cumulative Economic Impact
![Page 14: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/14.jpg)
14
![Page 15: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/15.jpg)
15
![Page 16: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/16.jpg)
16
What do we mean by ethics?
Common ethical frameworks?
Legal ethics?
Corporate Ethics – Social Responsibility?
Right v. Wrong?
Creepiness Factor?
![Page 17: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/17.jpg)
17
Common Ethical FrameworksUtilitarianism: The greatest good for the greatest number.
Liberalism: Individual freedom and autonomy
Communitarianism: Promotes the “common good” or shared values.
![Page 18: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/18.jpg)
18
Current Laws and Regulations Don’t Fit Current Applications of Analytics
In 2012 Facebook conducted an “emotional contagion” study, manipulating the display of happy and sad content in the news feeds of 150,000 users to see if they would share happy or sad content
![Page 19: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/19.jpg)
19
Notice and Consent
![Page 20: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/20.jpg)
20
Data Ownership
![Page 21: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/21.jpg)
21
The Power of Algorithms to Ferret Out Latent Traits in Datasets: Facebook “Likes” Research Finding “latent traits” in the Likes of 58,000 volunteers, an algorithm could model the
following otherwise undisclosed traits with 80 to 90% accuracy- Sexual orientation- Ethnicity- Religious and political views- Personality traits- Intelligence- Happiness- Use of addictive substances- Parental separation - Age- Gender- Etc.
Source: Z. Tufekci, “Algorithmiic Harms Beyond Facebook and Google: Emergent Challenges of Computational Agency,” 13 Colo. Tech. L. J. 203, 210 n.20 (2015)
![Page 22: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/22.jpg)
22
Predictive Policing
![Page 23: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/23.jpg)
23
How far can this be taken?
![Page 24: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/24.jpg)
24
![Page 25: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/25.jpg)
25
![Page 26: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/26.jpg)
26
![Page 27: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/27.jpg)
27
Predicting Human Conduct – Good or Bad Idea?
![Page 28: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/28.jpg)
28
Avoiding Bias
![Page 29: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/29.jpg)
29
Avoiding False Correlations
![Page 30: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/30.jpg)
30
The Digital Divide – Where do we get our information?
“This ‘digital divide’ is concentrated among older, less educated, less affluent populations, as well as in rural parts of the country that has fewer choices and slower connections.”- - Council of Economic Advisors Issue Brief – July 2015
![Page 31: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/31.jpg)
31
2016 Presidential Election Polling Data
How all of them were wrong? -Didn’t have data on the right people
The data they did have was wrong (reporting bias)
![Page 32: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/32.jpg)
32
Acting in a Legal Greenfield
• Identify and reasonably mitigate risks
![Page 33: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/33.jpg)
33
Corporate Ethical Review Boards Identify and quantify risks in analytics projects
Identify mitigation strategies
Include a diversity of opinion
Potential for greater transparency in decisionmaking
Builds confidence in corporate strategies and tactics
Modeled on IRBs, an ERB to be housed as component of Chief IG Officer or Chief Data Officer or CIO
![Page 34: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/34.jpg)
34
Hello Barbie vs. Amazon EchoUnderstand the risks you are creating and act reasonably to mitigate them
![Page 35: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/35.jpg)
35
![Page 36: Grappling with Ethics in the Age of Big Data presentation at the Chief Data Scientist, USA 2016](https://reader035.fdocuments.in/reader035/viewer/2022070511/58a4dd681a28ab34318b60e5/html5/thumbnails/36.jpg)
36