Keynote - Big Data's Ethical Drought
summary
Argument: the rush for data produces concrete harms and sloppy choices. Examples: viral WhatsApp images spread false election claims at scale; classic datasets come from questionable sources (MNIST from 14–18 year olds); IBM’s Diversity in Faces reuses a Yahoo/Flickr pool of 99.2M photos with unclear consent; Cambridge Analytica’s models infer sensitive traits — ~90% gender, 75–88% sexual orientation — and get deployed in political campaigns; Amazon’s hiring model learns anti‑women signals and is scrapped, while warehouse automation yields patented haptic bracelets and automated quota firings.
AI-written summary · about
chapters (13)
- 0:00Introduction and welcome
- 1:59Big data origins and impact
- 12:20Costs of big data collection
- 21:42Ethical data set development
- 22:03Big compute and automation
- 29:04Data science in elections
- 37:13Ethics in machine learning
- 38:20LGPD and consent regulations
- 43:29Privacy by design principles
- 46:42Ethical oath and accountability
- 48:06Microsoft's ethical duality
- 51:18Employee activism and responsibility
- 53:30Call to action for ethical tech
related talks
31mPanel Discussion : Ethics in Data Science
01h 07mKeynote: Welcome to the Machine
01h 01mDoing Data Science with Ethics
13mKeynote Speech
56mEthics in Machine Learning Panel
42mInfosec, AI and ethics – new models for a secure future
45mThe Myth of Neutrality: How AI is widening social divides
23mMachine Learning Ethics