AI math debate: critic argues AI leans entirely on Lean and can't explain why proofs work
gerardsans · x · 2026-10-19
A heated exchange over AI doing mathematics: responding to claims that AI could produce "new maths nobody understands," @gerardsans pushes back — new math has always been possible, but you have to build it and show it doesn't fall apart, and so far that new math is nowhere to be found. His core argument: current AI can't even connect the dots, as it fully relies on Lean for formal verification, making it useless at answering why a solution works. A sharp articulation of the gap between passing formal checks and genuine mathematical understanding.
More from AGI Musings
- Anthropic research finds verbalizable representations form a global workspace in LLMs — MoonL88537 · 2026-10-19
- Mike King: Google's Generative UI means rank tracking as we know it is dead — VeryWellVersed · 2026-10-19
- Why computational programs likely can't be conscious: Anna Ciaunica's embodied-mind argument — AnnaCiaunica · 2026-10-19
- Nathan Lambert: using models outside coding makes me feel AGI is a while away — natolambert · 2026-10-19
- Hacker used AI agents and ~$5,000 in tokens to breach 100+ corporate sites — deanwball · 2026-10-19
- Every product becomes infra: the 'infra plus infra' view of building in the agent era — aishashok14 · 2026-10-19