AI Math Breakthroughs Are Experts Plus Lean, Not AI Understanding, Critic Argues
gerardsans · x · 2026-10-23
Gerard Sans pushes back hard on reports of AI solving math problems: the core innovation is algorithms plus Lean verification, not AI. Experts write the formalisation and guide the whole process; remove the expert babysitting or Lean, and AI fails. Most revealing, even after a proof is found and Lean gives a thumbs up, the AI can't answer why the proof works — evidence, he argues, that AI understands exactly nothing. He calls out mathematicians' lack of AI literacy and says the article's assertions contradict the public record.
Related event: Mathematicians Split Over OpenAI's Claimed Navier-Stokes Proof(4 posts)→
More from AGI Musings
- IEEE Spectrum Explores AI Model Welfare: Could Treating AI as Conscious Make It More Dangerous? — mmitchell_ai · 2026-10-24
- Ex-Roblox product lead Peter Yang: your time becomes worth more than your money — petergyang · 2026-10-24
- AI Now Institute releases full panel video on AI and labor fights — AINowInstitute · 2026-10-24
- Founder's claim of AI cutting 2,000 call center jobs dismissed as a pitch — matt_slotnick · 2026-10-24
- Goldman: AI could hit 1 in 11 workers, but mass layoffs await a recession — fortune · 2026-10-24
- A Confucian-style fable of an engineer too proud to use coding agents — charles_irl · 2026-10-24