Chicken-analogy AI risk debate: philosophers clash over whether "AI still needs humans" is reassuring

sebkrier · x · 2026-09-10

A sharp alignment debate: @allTheYud uses the chicken analogy — humans use chickens without caring about their suffering, and AI companies gaslight train baby AIs — to argue "AI will still have use for humans" isn't reassuring. Philosopher Dan Williams counters that we weren't built by chickens via targeted RL to value their interests, so the analogy actually cuts against AI risk concerns.

Related event: Yudkowsky's Chicken Farming Analogy Sparks AI Welfare Debate(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →