Ethical framework for AI : Creating AI powered products that empower people
summary
Generative models power fake product reviews (Sports Illustrated publishes a fabricated reviewer “Sora Tanaka”), image models trained on scraped Getty content (Stability AI lawsuit), facial recognition returns at least six wrongful matches in Detroit leading to a pregnant woman’s arrest, Australia’s “Robo‑debt” algorithm forces repayments later reversed with $1.8 billion refunded, and LinkedIn trains models on user posts without clear consent. Use the community benefit lens, insist on human in the loop oversight, audit for representative data, and require explainable decisions.
AI-written summary · about
chapters (13)
- 0:00AI's impact on daily life
- 1:14Need for ethical AI framework
- 4:25Five lenses for ethical AI
- 6:36Community benefit and human rights
- 8:46Sports Illustrated fake reviews
- 13:04Fairness and bias in AI
- 14:27Wrongful arrest with facial recognition
- 17:54Accountability and human oversight
- 19:16Australia's Robo-debt scandal
- 24:59Transparency in automated decisions
- 28:09Privacy risks in AI platforms
- 28:52LinkedIn data training controversy
- 31:41Ethical AI principles summary
related talks
28mEthical AI?!
47mEthics in AI Panel
47mEthics & AI: Identifying & preventing bias in predictive models
26mBuilding ethical AI: Transparency, bias mitigation, and trust
45mEthical use of AI: seeking, understanding, and countering the risks
14mBuilding Transparent and Ethical Generative AI Models in High-Stakes Industries
54mAI and Ethics: Navigating the Future of Responsible Technology
44mAI, Corporate Responsibility & Democratic Legitimacy