Anthropic's Claude Completes Formal Proof of Fermat's Last Theorem in 11 Days

Anthropic's AI model, Claude, has achieved a significant milestone by autonomously formalizing Fermat's Last Theorem in a remarkable 11-day period. This accomplishment involved generating 13 million lines of Lean code and verifying 29,500 intermediate theorems, a task that was previously expected to take years.

The theorem, originally posited by Pierre de Fermat in 1637, states that no three positive integers a, b, and c can satisfy the equation a^n + b^n = c^n for any integer value of n greater than 2. While Andrew Wiles proved the theorem in 1995, Claude's formalization translates this proof into a machine-verifiable format, ensuring that every logical step is encoded and checked independently by a computer.

This formalization is built on the extensive groundwork laid by the interactive theorem-proving community, particularly the ongoing efforts of Kevin Buzzard at Imperial College London. Claude's work follows the modern Frey-curve and modularity-lifting approach, which is distinct from earlier methods and highlights the potential for AI to assist in rigorous mathematical reasoning.

The implications of this achievement extend beyond mathematics, as it demonstrates AI's capability to handle complex logical reasoning tasks that surpass traditional benchmarks. Observers note that this success may enhance Anthropic's competitive position in the AI landscape, with market pricing reflecting increased confidence in Claude's capabilities.

Looking ahead, the mathematical community aims for a future where every major theorem has a machine-checked proof, reducing the risk of human error in accepted proofs. Claude's accomplishment is a step toward realizing this vision, though the gap between formalizing existing proofs and discovering new ones remains significant.