Michigan professor Ben Green: why 'AI for Good' fails and self-regulation won't work
benzevgreen · x · 2026-10-16
Dr. Ben Green (University of Michigan) on The Data & AI Chief podcast explains why well-built algorithms still fail their intended beneficiaries.
Key points
- Hidden cost of outsourced thinking: students using AI ace homework but fall behind on exams — powerful AI doesn't automatically make people better at their jobs.
- 'Algorithm in the loop' over 'human in the loop': the framing sidelines humans; flip it so AI supports human decision-makers.
- Boston EMS case: the real driver of slow ambulance response was homelessness and mental-health calls, leading to a specialized team — asking the right question beats deploying a model.
- Asking questions is the data scientist's most underrated skill, building trust and humility.
- AI companies can't police themselves: voluntary safety pledges are PR; he argues for legal liability and pre-deployment testing.
More from AGI Musings
- SEO visibility now tracks impressions, not clicks, as AI Overviews eat traffic — lilyraynyc · 2026-10-16
- Google white paper: AI river-flood forecasts 7 days out for 2B+ people, FireSat wildfire tracking — ymatias · 2026-10-16
- Polymarket odds: Only 10% chance US enacts AI safety bill by end of 2026 — Polymarket · 2026-10-16
- Intelligence is proportional to data: reorder tasks by data volume and the jagged frontier becomes a spiral — sebkrier · 2026-10-16
- "Vibecoding" is as misleading a label as "AI" and "SI," dev complains — eschadiol · 2026-10-16
- Reviewer will decline papers largely written by AI, citing author effort — IanArawjo · 2026-10-16