Toward Ethical AI
summary
AI systems repeatedly produce harms — denied loans, misdiagnosed patients, wrongful arrests — because bias emerges from the interaction of code, training data, and application, not the hardware or language. The claim is that inclusion at every step reduces that risk: validate use cases, vet training sets, run customer focus groups, force model interpretability. One company (DataEcho) builds a platform for this but finds technology alone fails without organization-wide processes and commitment.
AI-written summary · about
chapters (5)
- 0:00Start
- 1:15Opinion A.I. Could Worsen Health Disparities In a health system riddled with inequity, we risk making dangerous biases automated and invisible
- 2:49We all have blind spots.
- 3:34ORGANIZATIONAL COMMITMENT
- 4:51REPUTATIONAL RISK
related talks
34m(Un)Ethical Machines? Why AI Bias Is a Problem and What to Do About It
56mAn AI with an Agenda: How Our Cognitive Biases Leak Into Machine Learning
47mEthics in AI Panel
19mEthical Bias
19mMoral Outsourcing: Humanity in the age of AI
21mFairness and bias in AI
30mFireside chat: Breaking the wall of unethical AI
30mFighting Prejudice in Artificial Intelligence