SIAgentNews.com
← All news

Anthropic researcher claims Claude solved the Riemann hypothesis

A mathematician says he used the Claude chatbot to find a proof for a 160-year-old problem regarding prime numbers.

· First reported by Phys.org

$1 millionMillennium Prize Problem reward

The short version

  • The Riemann hypothesis concerns the distribution of prime numbers and has been unsolved since 1859.
  • The problem is one of seven Millennium Prize Problems with a $1 million reward for a verified solution.
  • The researcher released a version of the proof written in Lean, a computer-verifiable programming language.
  • The claim has not been peer-reviewed or independently verified by the mathematical community.

Why it matters

If verified, this would be one of the most significant mathematical breakthroughs in history. It demonstrates the potential for AI to assist in solving complex, high-level theoretical problems that have stumped humans for centuries.

A 160-year-old mathematical mystery

A mathematician from the AI company Anthropic has made a bold claim regarding one of the most famous unsolved problems in history. Levent Alpoge stated on Friday that he used the firm's Claude chatbot to discover a proof for the Riemann hypothesis. This specific problem has eluded the global mathematical community for more than 160 years.

First proposed by German mathematician Bernhard Riemann in 1859, the hypothesis focuses on the distribution of prime numbers. Because of its fundamental importance to pure mathematics, it is widely regarded as the most significant open question in the field. Solving it would provide deep insights into the underlying patterns of numbers.

The problem is so significant that it was selected as one of the seven Millennium Prize Problems by the Clay Mathematics Institute in 2000. The institute offers a $1 million reward for a solution to each of these problems. However, the rules require a solution to be published in a peer-reviewed journal and withstand two years of scrutiny before the prize is awarded.

How the AI assisted the process

Alpoge did not simply ask the chatbot for an answer; he worked with the tool to produce a specific version of the proof. He released a four-page note detailing his work, which he claims also settles a wider variation of the puzzle called the generalized Riemann hypothesis. He noted that he is only releasing the version produced in a specific format for now.

The proof was written in Lean, which is a programming language designed to allow mathematical arguments to be checked by a computer. By using a formal language like Lean, mathematicians can ensure that every step of a logical argument follows the rules of the system. This type of verification is often more rigorous than traditional written proofs because it removes human error from the checking process.

Alpoge admitted that a full written argument for the proof would be extremely long. He also stated that he has not yet mastered all of the mathematics involved in the complete proof. Consequently, he plans to write up a proper, comprehensive version of the argument in the future.

Context and corporate timing

The announcement comes at a notable time for Anthropic, which recently saw a record IPO. Alpoge acknowledged that the timing of his announcement could be viewed cynically, though he did not provide specific details on why he felt that way. He did, however, consult with Anthropic's chief executive, Dario Amodei, before making the public claim.

This is not the first time Anthropic has engaged with the Riemann hypothesis. In August, the company stated that a version of Claude that was not yet public had made progress on a long-standing finding connected to the hypothesis. At that time, the company was careful to stress that the AI had not actually proven the hypothesis, only improved upon existing work.

The claim follows a similar trend in the AI industry regarding high-level mathematics. Only two months ago, a rival company, OpenAI, claimed that an AI model had solved another Millennium Prize Problem involving fluid dynamics. While that proof appeared correct, it also remains unvetted by the broader scientific community.

The path to verification

Despite the excitement surrounding the claim, the mathematical community remains skeptical until the work is independently reviewed. Because the proof is currently unverified, it cannot be accepted as a solution to the Riemann hypothesis. The next step involves moving the work from a private note to a public, peer-reviewed forum where other experts can scrutinize the logic.

Alpoge's history includes work on related fluid problems, where he was involved in a dispute over credit for research with a New York University mathematician. This history of complex collaborations and disputes highlights the difficulty of establishing ownership and accuracy in high-level research. For now, the community awaits a more formal presentation of the work.

If the proof holds up under the scrutiny of other mathematicians, it would represent a monumental achievement for both human and artificial intelligence. Until then, the claim remains a significant but unproven assertion in the world of pure mathematics. The scientific community will need to see the full written argument to determine its validity.

What to watch

  • Whether the proof survives independent peer review by the mathematical community.
  • Whether Alpoge releases the full, long-form written argument of the proof.
  • If Anthropic officially confirms the proof's validity in future model updates.

Try it

Look up the 'Millennium Prize Problems' on the Clay Mathematics Institute website to see the other six unsolved problems currently offering a $1 million reward.

Sources and links

Read the original at Phys.org

First reported by Phys.org. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

More AI news