Skeptic: AI math proofs may lack conceptual novelty — 'hard' problems may just be labor-intensive

LucaAmb · x · 2026-11-21

LucaAmb expresses skepticism about the conceptual novelty of AI-generated math proofs: in every field he's competent in, he sees little conceptual innovation from AI, so he doubts math is fundamentally different.

His provocative hypothesis: much math assumed difficult for deep conceptual reasons may simply have been too labor-intensive to complete — and that's exactly the kind of work AI can scale.

Original post →

More from AGI Musings

AGI Musings channel →