A mathematician at Anthropic has said that the San Francisco-based artificial intelligence (AI) company’s software has helped crack one of the most important unsolved problems in mathematics – and experts say the claim checks out.
Levent Alpöge, a Turkish American mathematician, on Friday posted a note on his personal website saying that Claude, the AI software, had helped prove the Riemann hypothesis. He says that Claude helped develop a solution to the problem during the course of about six weeks.
On Saturday afternoon, two organizations announced that they had independently verified Alpöge’s claim after reviewing the programming language he used to share the solution.
The proof evidently marks another breakthrough in the use of AI for complicated math problems – even as mathematicians express uneasiness around extremely lucrative private companies dishing out millions of dollars for the technology to solve problems.
“Claude and I have had a great time together with such breadth in recent months, but this time Claude far outpaced me near the finish line,” Alpöge wrote, while anthropomorphizing the AI software. “No, Claude flew, and danced, and sparkled near that finish line.”
Alpöge linked to computer code – written in a language called Lean – that he maintained could verify the findings.
Mathematicians subsequently began working to see if Alpöge’s proof could be confirmed. On Saturday afternoon, the Lean Focused Research Organization – which seeks to advance the use of the programming language – and the Mathlib Initiative, a program that works with Lean for math equations, announced that they had verified Alpöge’s solution.
“We have full confidence that Alpöge’s proof of the Riemann hypothesis is correct, and we congratulate him on this achievement,” the groups announced.
Alpöge’s note was published just one day after Anthropic went public on the Nasdaq stock market index. He acknowledged that, writing: “The precise timing is horrible. If I could have, I’d have finished this all much earlier.”
The Riemann hypothesis is a nearly 170-year-old math problem named after Bernhard Riemann, a German mathematician who proposed it in 1859.
The hypothesis deals with prime numbers – whole numbers greater than 1 that can only be divided by themselves and 1. Riemann questioned how often prime numbers appear and how they are distributed along the number line. And he developed a formula for estimating how many prime numbers exist under a specific number.
The Riemann hypothesis is one of the seven Millennium prize math problems selected in 200 by the Clay Mathematics Institute. The institute has offered $1m for each problem solved.
after newsletter promotion
Mathematicians have been attempting to solve this problem for years. The New York Times reported that on Friday, an independent mathematician in Italy released an incomplete proof of the Riemann hypothesis that he said he had been working on for almost 12 years. That mathematician, Alessandro Masullo, said he did not use AI because he did not fully trust the system.
Other mathematicians have expressed discomfort with relying on AI to solve math problems.
In September, Anthropic competitor OpenAI claimed that a non-public version of its AI tools helped solve part of another Millennium prize problem. But it was solved in a way that struck mathematicians and observers in a way many found strange: a near-trillion-dollar private company spent $15m to employ 10,000 AI systems to solve the problem.
Alpöge was working on that separate problem for Anthropic when OpenAI made its announcement.
University of Warwick mathematician James Robinson told the Guardian in September that “it’s frustrating to see these big tech companies burning this fuel up just so that they can show off about how great their latest model is”.
He added: “It seems like immature playground boasting writ large, underpinned by billions of dollars and the potential for significant environmental damage in an age when climate change is probably the biggest challenge we face.”













