331
Productivity & Workflow350
Automation & Workflow232
Software Development240
Marketing & Growth193
Writing & Content Creation196
AI Infrastructure & MLOps182
Photography & Imaging153
Design & Creative168
Data & Analytics155
Customer Support145
Education & Learning133
Sales & Outreach128
Voice & Speech135
Operations & Admin107
Axiom Math says its AI checked the Stacks Project by turning it into computer-verified proofs, aiming to reduce unverified math claims from AI.
In short: Axiom Math says it used AI to formally verify the Stacks Project, a large and widely used online library of math proofs.
Axiom Math, a startup valued at about $1.6 billion, says it has formally verified the Stacks Project. The Stacks Project is an open online reference that mathematicians have built since 2005, covering topics from college algebra to advanced research areas like geometry and number theory.
The announcement comes as more AI systems publish large numbers of math papers and proofs. TechCrunch notes that OpenAI recently released 719 manuscripts claiming solutions to difficult problems, and that even computer-based checks can sometimes miss issues.
Axiom Math says its “AI agents” translated the Stacks Project’s roughly 7,654 pages into about 16 million lines of code. The idea is to rewrite each proof in Lean, a programming language used for math proofs, and then have a “proof assistant” (a checking program, like a strict spellchecker for logic) verify each step line by line.
Math is supposed to be dependable, but it can be hard to confirm every detail when work arrives in huge volumes. Formal verification is meant to make results easier to trust because the computer checks every step against the rules. If this approach spreads, it could help researchers, and eventually AI tools that write software, rely on math that has been checked more like a safety inspection than a quick glance.
Source: TechCrunch AI