Ethics in AI Panel
summary
Panel argues urgent, practical harms: deployed models at companies like SiriusXM, Pandora and AT&T routinely encode biases; a resume-screening model that rejects women and a computer-vision demo that repeatedly fails to detect darker-skinned people are concrete failures. Examples include a Tay-style chatbot that learns hate, license-plate recognition abused off-premise, and trust-and-safety teams being cut before elections. Solutions offered are ethics councils, domain-specific risk tiers, and demand for production-grade observability tooling for models.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:06Defining AI ethics
- 9:14Ethics beyond compliance
- 12:23Data bias in AI
- 24:50Steps for ethical AI
- 36:14Ethical AI by design
- 42:40Leadership accountability
- 43:44Future of responsible AI
related talks
56mEthics in Machine Learning Panel
33mEthics in Machine Learning Panel Q&A
05mToward Ethical AI
44mPanel: Building responsible, human-centric AI models
34m(Un)Ethical Machines? Why AI Bias Is a Problem and What to Do About It
35mDesigning Inclusive and Accessible Experiences in the Era of AI
34mEthical framework for AI : Creating AI powered products that empower people
30mFireside chat: Breaking the wall of unethical AI