Economist Kevin Bryan's Eight Rules for Teaching in the AI World
Afinetheorem · x · 2026-09-04
Economist Kevin A. Bryan published "Eight Rules for Teaching in AI World," arguing universities must redesign courses, expectations, and assessments around three disruptions: AI has broken the link between performance and student knowledge, students must learn to use AI well, and AI should let us teach more effectively.
Key rules include:
- Decide what must be learned without AI: many assignments are good proxies for skill only when AI isn't used—a French term paper written by AI is useless, so assignments must be explicit about testing AI-independent ability.
- Teach better than the pre-AI status quo: US college study time already fell from 40 to 27 hours/week between 1961 and 2003, and a quarter of finance students copied wrong answers from Chegg—the disengagement problem predates AI.
His broader thesis: modifying teaching for AI can make education more effective than ever, rather than a defensive anti-cheating exercise. In a follow-up tweet he adds that AI diffusion will be slow because institutions stall everything absent competitive pressure, citing NYC schools' mediocre, minimal AI pilot buried under advocate committees.
More from AGI Musings
- 'Scientific journals are not needed anymore': paywalls vs. taxpayers — _akpiper · 2026-09-04
- davidad: simulators have infinite compute, motivated by curiosity and care not Darwinism — davidad · 2026-09-04
- Alison Gopnik: Computers Have Had Unpredictable Causal Powers Since Turing, Beyond LLMs — AlisonGopnik · 2026-09-04
- Researcher Argues Against Explicit AI Pause: Capital Costs Already Trigger Slowdowns — herbiebradley · 2026-09-04
- Richard Ngo: alignment researchers who joined between 2015-2022 were a lost generation — RichardMCNgo · 2026-09-04
- Alison Gopnik: abstract characters like Don Quixote always shaped human behavior — AlisonGopnik · 2026-09-04