Anthropic Researcher:
"Claude Proves the Riemann Hypothesis, One of Math's Greatest Unsolved Problems"

Also Acknowledges, "I Could Not Fully Understand the Proof"
Announced on Website... Rigorous Final Verification Expected

A claim has emerged that Claude, the chatbot developed by AI company Anthropic, has proved the Riemann hypothesis, considered one of the greatest unsolved problems in mathematics.


Anthropic's chatbot Claude. Reuters-Yonhap News.

Anthropic's chatbot Claude. Reuters-Yonhap News.

View original image

On November 20 (local time), major international news outlets, including AFP, reported that Anthropic researcher Levent Alpoge said in a post on his website and a four-page document that day, "I believe I have discovered a proof of the Riemann hypothesis in collaboration with Claude." He said he had also solved the broader "generalized Riemann hypothesis." He said the work done with AI involved more than 250,000 pages. Alpoge wrote his code in Lean, a language that turns mathematical propositions into a form that can be verified. If the code contains no errors, it means the logic of the proof is sound.


He said he would announce the results after consulting Anthropic CEO Dario Amodei. He also acknowledged that the timing of the announcement could invite skepticism, as it came the day after Anthropic's initial public offering (IPO).


The claim has not yet been independently verified or undergone rigorous external review. However, Brian Conley, an authority on the Riemann hypothesis, told AFP that it was "a monumental moment in human history," adding, "I think this is the biggest thing that has ever happened in the history of mathematics."


Alpoge described the moment by saying, "As we neared the finish line, Claude was dancing and sparkling way ahead of me." Nestor Guillen, a visiting professor in the Department of Mathematics at New York University, said it was positive that Alpoge had candidly acknowledged he did not fully understand the proof. Steven Strogatz, a professor at Cornell University in the United States, praised the key role played not by AI alone but by a capable human mathematician like Alpoge, who directed Claude on which ideas to explore. James Maynard, a mathematician at the University of Oxford in the United Kingdom, said, "It's very likely to be correct. I'm 98-99% confident." He added, "What we really wanted to know was why the Riemann hypothesis is true. The new proof will give us a deep understanding, and that's something to celebrate."


However, some have criticized the claim. Brian Conrad, a mathematician at Stanford University in the United States, said, "To be honest, it is completely absurd for Alpoge to release a proof that he himself does not fully understand."


The Riemann hypothesis, proposed in 1859 by German mathematician Bernhard Riemann, is a central unsolved problem in number theory. It concerns the distribution and patterns of prime numbers, such as 2, 3, 5, 7 and 11, which are divisible only by 1 and themselves. The hypothesis centers on the idea that prime numbers, which appear random, actually follow a certain order and pattern based on specific values—the nontrivial zeros—of a complex function called the Riemann zeta function.



Because prime numbers form the mathematical foundation of public-key cryptography (RSA), which is used in modern internet finance and cryptocurrency, proving the Riemann hypothesis could have major implications for cryptography, quantum physics and computer science. In 2000, the Clay Mathematics Institute selected seven "Millennium Prize Problems," including the Riemann hypothesis, and offered a prize of $1 million (about 1.35 billion won) for each problem. A solution is officially recognized only after it is published in a peer-reviewed journal and passes two years of verification.


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing