Stanford professor criticizes AI companies' math PR push: proofs no one understands don't count

anshulkundaje · x · 2026-09-13

Stanford professor Anshul Kundaje argues AI should absolutely accelerate our understanding of unsolved math, and there's a clear path to it — but not the way AI companies are doing it, citing the recent millennium prize controversy. His core point: even if AI autonomously solved millions of theorems a day, it means little if humans lose the ability to understand them, a lesson applying to many intellectual endeavors.

Related event: Stanford Professor: Domain Expertise Matters More in the AI Era(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →