AI & ComputeLOT 261122 · 3 min read

Anthropic Mathematician Claims Claude Proved the Riemann Hypothesis

Levent Alpöge says Claude helped prove the Riemann Hypothesis, formalized in Lean. Anthropic has not commented. Clay rules demand two years of scrutiny before any prize.

By Sophie Lindqvist583 words

Key takeaways

  1. Anthropic mathematician Levent Alpöge claims a Claude-assisted proof of the Riemann Hypothesis, formalized in Lean.
  2. The Clay Mathematics Institute offers $1 million per Millennium problem and requires two years of scrutiny before consideration.
  3. Only one Millennium problem, the Poincaré conjecture, has been officially solved in the prize's 25-year history.
  4. Anthropic said in August it did not expect Claude's techniques to prove the Riemann hypothesis.
  5. OpenAI spent millions and 10,000 agents on its September Navier-Stokes claim, which lacks formal recognition.
Anthropic mathematician claims to have solved prestigious Millennium Prize problem with Claude
Anthropic mathematician claims to have solved prestigious Millennium Prize problem with Claude — geralt / pixabay

Anthropic mathematician Levent Alpöge claims he has proved the Riemann Hypothesis — one of seven Millennium Prize problems — with the help of Claude, and says the pair formalized the arguments in Lean without assumptions. The claim has not been independently verified. Anthropic has not commented. The document sits on his personal website, nowhere else.

That gap between claim and acceptance defines this story. The Clay Mathematics Institute administers the Millennium Prize and offers $1 million per solved problem. Its rules require publication and at least two years of scrutiny. During that period, a solution must receive, in the institute's words, "general acceptance in the global mathematics community." Only one problem has been officially solved in the prize's quarter-century: the Poincaré conjecture, a topological problem concerning three-dimensional spheres.

What exactly did Alpöge claim?

On his website, Alpöge wrote: "I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis." He goes further than the headline result. The method, he says, "easily extends" to the Generalized Riemann Hypothesis, and he expects the Extended Riemann Hypothesis to follow with more time. He also struck an unusually emotional note: "It is natural to feel that this is an end, in my case a loss of a principal guiding light, but what I saw during the course of progress so amazed me that I believe it is one dawn across science of something far greater."

The Riemann Hypothesis dates to 1859 and concerns the distribution of prime numbers and the zeroes of a function. Mathematicians regard it as one of the most important unsolved problems in the field.

What is not said matters more

Anthropic has stayed out of the announcement entirely. Alpöge says he texted CEO Dario Amodei "late Sunday night" and suggests he delayed publication until after the company's IPO on Thursday. Timing, in other words, may have been a commercial decision. The Verge reached out to Anthropic and did not immediately hear back.

There is also a prior statement that cuts against the claim. In August, Anthropic said it was working on the Riemann Hypothesis with Claude, but noted: "We don't expect that the techniques Claude used will lead to proving the Riemann hypothesis." Either the techniques changed, or the August assessment was wrong. Nobody has said which.

How does this fit the OpenAI pattern?

OpenAI claimed in September to have cracked another Millennium problem, Navier-Stokes, which concerns fluid flow. That effort reportedly cost millions of dollars and deployed 10,000 agents. The result is widely considered credible but has no formal recognition. It has also drawn allegations of questionable research practices, including claims that OpenAI used unpublished work from mathematicians who had shared research with its AI tools. OpenAI has strongly denied wrongdoing. The denial has not quieted the criticism.

The Navier-Stokes claim has a personal edge here. Alpöge had been working on that problem with Tristan Buckmaster, a mathematics professor at New York University. Buckmaster claimed OpenAI produced its solution using the very route he and Alpöge were pursuing — with OpenAI's Codex and with Claude.

Alpöge is not a one-claim outsider. In July, he said he solved the Jacobian conjecture using Claude. Mathematicians are still digesting that result alongside everything else. The field, in the source's words, is rapidly becoming a profoundly different place.

Two years of mandatory scrutiny stand between this announcement and the $1 million prize. Many claimed proofs of the Riemann Hypothesis have not survived two weeks.

theverge.com (Source)

Filed under

Share this article:

More from Sophie Lindqvist

Sophie Lindqvist

Show full bio

News editor covering money, funding and market shifts around Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.

1,238 articles

Read also

« Previous article