Latest: Anthropic Claims Claude Formalized Fermat's Last Theorem; Dispute Persists
A machine-verified formalization converts a proof into a form a computer can check step by step; Anthropic argues this could help address the growing burden on mathematical referees as more proofs are produced. Whether Anthropic's claim holds or the dispute is correct affects how the event is understood as an AI milestone.
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.
Anthropic says Claude completed the first machine-verified formalization of Fermat's Last Theorem in 11 days, with dozens of Claude agents converting the Wiles-based proof into over 13 million lines of Lean code and generating more than 29,000 supporting theorems across areas of mathematics not previously formalized. The run was confirmed complete when the root theorem returned a status of 'Proved' with zero open leaves. A dev.to article disputes the characterization, arguing that FLT formalization remains an ongoing community project and that what was reported as complete covers only FLT for regular primes, which is a special case and not a full formalization of the general theorem. The article also states that Anthropic's documented Claude Lean work concerns the Riemann zeta function rather than FLT.
Zvi Mowshowitz's newsletter on September 10 noted Anthropic's claim that Claude formalized Fermat's Last Theorem in Lean and separately recorded a claim attributed to Jakub Pachocki that OpenAI could improve its models' math capabilities but has chosen not to prioritize it.
The dispute over whether the formalization covers the general theorem or only FLT for regular primes remains unresolved.
The Neuron Daily reports Claude formalized FLT in over 13 million lines of Lean.
Anthropic announced September 5 that dozens of Claude agents formalized Fermat's Last Theorem in machine-checkable Lean code in 11 days, producing over 13 million lines proving more than 29,000 supporting theorems.
A companion document described agents cross-checking each other's work, with completion confirmed when the root theorem returned zero open leaves. A dev.to article disputed the claim the same day, arguing that Claude's result covers only FLT for regular primes, a special case, not the general theorem.
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.