### AI Achievement: Claude Cracks Fermat’s Last Theorem in Record Time
In a landmark development for computational mathematics, Anthropic’s Claude AI has successfully generated the first fully machine-verified proof of Fermat’s Last Theorem. By completing this monumental task in just 11 days, the model has effectively authored the most extensive mathematical proof in history, demonstrating a level of autonomous reasoning that challenges traditional academic timelines.
#### A New Benchmark for Mathematical Rigor
For over three centuries, Fermat’s Last Theorem-which posits that no three positive integers $a, b,$ and $c$ satisfy the equation $a^n + b^n = c^n$ for any integer value of $n$ greater than 2-remained one of the most elusive puzzles in number theory. While Andrew Wiles famously provided a proof in the 1990s, the process of formalizing such proofs into machine-readable code is notoriously labor-intensive.
Claude’s approach was distinct: it generated 13 million lines of formal code. Unlike traditional proofs that rely on human peer review, this output is verifiable line-by-line by a computer, ensuring absolute logical consistency without the risk of human oversight errors.
#### Outpacing Human-Led Initiatives
The speed of this breakthrough is particularly striking when compared to ongoing human efforts. Since 2024, a dedicated team at Imperial College London has been working to formalize complex mathematical proofs. Despite their expertise, the project remains a long-term endeavor. Claude’s ability to reach the finish line in less than two weeks highlights a paradigm shift in how we approach “brute-force” logical verification.
Kevin Buzzard, a prominent mathematician and the lead of the Imperial College project, conducted a rigorous review of Claude’s work. He confirmed that the proof is sound, noting that it adheres strictly to the foundational axioms of mathematics.
Myriad: When will GPT-6 become publicly available? Click to make your prediction.
#### Why This Matters for the Future of AI
This achievement serves as a proof-of-concept for “AI-assisted discovery.” While human mathematicians provide the intuition and the creative leaps necessary for new theories, AI models like Claude are proving to be indispensable tools for the tedious, high-stakes work of verification. As we look toward the future, the integration of large language models into formal verification systems could accelerate scientific progress, potentially solving other “unsolvable” problems in physics and cryptography at a fraction of the current cost and time.
