Math Twitter clashes over claim AI labs are Goodhart-ing mathematics via millennium prize problems
prasanna_says · x · 2026-09-13
- The debate: A viral exchange around Terry Tao's remarks on AI labs "Goodhart-ing" mathematics escalated. One summary: to make that argument, Tao had to claim math's real point is not solving problems but producing "conceptual understanding and insight" — existing almost exclusively in mathematicians' heads.
- The critique: No other academic field justifies itself this way. Physics, chemistry, biology, history and philosophy are supposed to produce outputs that matter to the public, or public funding loses its rationale. Admitting that solving problems "doesn't accomplish anything directly" looks like a self-inflicted wound.
- The steelman: Millennium prize problems are metrics for mathematical progress; AI labs are optimizing the metrics, thereby making them less useful.
The clash exposes a deeper rift between academia and AI engineering over what counts as real progress.
More from AGI Musings
- Dev warns abliterated models could wreck software supply chains with 1000+ transitive dependencies — nitarshan · 2026-10-27
- Delete your priors: why every SOTA model drop should reset how you think the world works — zack_overflow · 2026-10-27
- Replit predicts every Fortune 500 will be a 'Self-Driving Company' by end of 2030 — amasad · 2026-10-27
- Anthropic says insurers can't cover its own catastrophic risk, X mocks it — GarrisonLovely · 2026-10-27
- MIT's Ziming Liu: AutoResearch doesn't generalize — blind training tricks cause 'forking' disasters — ZimingLiu11 · 2026-10-26
- AI accusations shake literary prizes from Caine to Goncourt: who decides what's human? — TuhinChakr · 2026-10-26