Ethics in Machine Learning Panel
summary
Panel records a practical debate about harms, fixes and governance with concrete examples and proposals: Tay chatbot turning racist, Google and Flickr classifying images of African‑Americans as "monkeys", COMPAS recidivism concerns, biased word embeddings and FaceApp style risks. Discussion names companies (Shopify GoodAI), cites audits, the Algorithmic Justice League, interpretability methods, APIs that leak bias, GDPR / European Data Protection Directive, ethics review analogies and whether audits or datasets actually change outcomes. Watch if case studies tools or policy steps matter to the decision.
AI-written summary · about
chapters (14)
- 0:00Welcome and panel
- 1:08Panelist introductions
- 5:43AI ethics defined
- 6:26Positive AI ethics trends
- 12:25Data risks in democratization
- 16:45AI providers' ethical duties
- 19:13AI governance and values
- 27:41Privacy laws and models
- 31:55AI ethics research priorities
- 43:01Q&A on ethics review boards
- 46:37Military AI ethics
- 50:51Ethical dilemmas in platforms
- 51:40Ideology in ML models
- 55:12Closing and Q&A
related talks
47mEthics in AI Panel
31mEthics of algorithms - from theoretical discussion to practical application
31mPanel Discussion : Ethics in Data Science
47mEthics & AI: Identifying & preventing bias in predictive models
55mPanel: Main legal risks and challenges with LLMs / generative AI models
47mEthical Machine Learning
38mSocial Implications of Bias in Machine Learning
26mPanel Discussion: Ethical AI - From talk to practice