Archived story — belongs to the Monday, September 28, 2026 sweep.
P1 · Tectonic · Tech Industry · sweep Monday, September 28, 2026
Formal verification of mathematical theorems by language models
Anthropic reports that Claude formalized Fermat's Theorem in Lean mostly autonomously over 11 days, generating 13 million machine-verifiable lines, while failed attempts on the Riemann hypothesis produced partial advances. The event shifts mathematical validation from human trust to computational verification and opens a layer of formal proof infrastructure that crosses frontier AI capabilities with epistemic and security consequences.
Title, thesis and gravity assessment written by an AI model, without prior human review.
Gravity84/100 · Tectonic
Why this gravity
Anthropic reports that Claude formalized Fermat's Theorem in Lean mostly autonomously over 11 days, generating 13 million machine-verifiable lines, with coverage from WSJ, Hacker News, and Decrypt; the failed 54-hour attempt on the Riemann hypothesis produced only partial advances. The event marks a leap in long-horizon autonomy and computational formal verification, although it does not solve an open conjecture.
- Independent sources
- 4
- Cross verticals
- Frontier AI, Tech Industry, Blockchain
- Tier 1 actors
- Anthropic, OpenAI
- Gravity vs. the previous edition
- +22 (62 → 84)
Lead source
Anthropic says Claude worked "largely autonomously" over 11 days to formalize the proof of Fermat's Last Theorem in the Lean programming language (Anthropic) · TechmemeMore: Hacker News, Decrypt, Simon Willison
Timeline · Madrid time
- 00:35
- 23:04Quoting Linus Torvalds · Simon Willison
- 20:42Formalizing Fermat's Last Theorem · Hacker News
- 20:57Fermat's Last Theorem in Lean 4 · Hacker News
- 22:35
- 15:01
- 19:45A misalignment of AI in mathematics · Hacker News
ShareWhatsApp