Critic calls OpenAI's 166-page Lean math proof a brute-force dump, not a real achievement

gerardsans · x · 2026-09-23

Replying to Martin Bauer, gerardsans argues that comparing AI to humans is a category mistake—nobody compares millisecond computer algorithms against people, yet AI-vs-human comparisons are treated as meaningful. He also disputes OpenAI's celebrated math result, claiming the company dodged the actual proof elaboration via brute-force compute and simply dropped a 166-page Lean formalisation "with a thumbs up," calling it disingenuous to claim anything remarkable was achieved. A single critical voice, not a verified conclusion.

Related event: Critics Slam $10M AI Math Contest as Category Error and Waste(3 posts)→

Original post →

More from Models

Models channel →