Claim that AI proofs are incomprehensible to humans is questionable, argue commenters
basedjensen · x · 2026-10-06
Pushing back on the popular claim that AI-generated math proofs are beyond human comprehension, AcerFur argues models aren't writing in an alien language: a proof is just a chain of logical deductions, and given sufficient time, humanity collectively can make sense of them.
More from AGI Musings
- AI slop is a mirror: treating AI as a creative partner instead of a tool — Temporary_Dirt_345 · 2026-10-06
- Apollo's chief economist: AI boom visible in valuations, not yet in productivity data — 2C_ornot2C · 2026-10-06
- If AI Agents Replace Browsers, Millions of Websites Will Struggle, User Warns — saibharadwaj · 2026-10-06
- Dev slams Anthropic's personhood framing: first-person voice is UI, not a self — gerardsans · 2026-10-06
- AI firms ingest books 'illegally' then cite copyright law to refuse real quotes — StewartalsopIII · 2026-10-06
- Frontier Labs Reportedly Use Mathematicians as 'Human Cover' for AI Results — CharlotteHase · 2026-10-06