A Russian tech site spells out why an AI-assisted Riemann Hypothesis claim matters
An Anthropic mathematician says Claude helped him prove a 167-year-old problem. Habr's version explains the stakes well, but the proof still awaits outside experts.
The short version
- Levent Alpöge, a mathematician at Anthropic, says he found a proof using Claude.
- Riemann posed the problem in 1859; it has gone unproven for 167 years.
- It is one of seven Millennium Prize problems, each carrying a $1 million Clay Institute prize.
- Only one of the seven has been solved since the prizes began: Perelman's Poincaré result.
- Other reports say a Lean check exists and is public, but outside experts have not confirmed it.
Why it matters
If the proof holds, a long list of number theory results that currently carry an 'if the hypothesis is true' caveat could lose it. Just as important is the method: a human and an AI working together, with a machine checking the logic. Until independent mathematicians weigh in, treat it as a serious claim, not a settled fact.
What is actually being claimed
Levent Alpöge, a mathematician at Anthropic, says he found a proof of the Riemann Hypothesis with the help of Claude, the company's AI assistant. As habr.com reported, the post opens with the sense that history is being made right now, and it even sets the news against the record $3.79 billion that GTA VI earned in its first day of sales.
The framing is dramatic. The article quotes one mathematician as saying humanity is being humiliated. That is a reaction, not evidence, and the quote is not tied to a named person in the piece.
The important detail is the wording of the claim itself. Alpöge is not described as saying that the AI worked alone. He says he found the proof with Claude, which makes it a collaboration between a person and a tool.
Why this problem is such a big deal
In 1859 the German mathematician Bernhard Riemann formulated the hypothesis. In the 167 years since, the best mathematicians in the world have not managed to prove it. It is one of seven Millennium Prize problems, and the Clay Mathematics Institute promises $1 million for solving each one.
The hypothesis concerns prime numbers, the numbers divisible only by one and themselves, such as 2, 3, 5 and 7. At first glance they look scattered almost at random along the number line. The hypothesis describes a strict pattern hidden behind that apparent chaos.
This is not only a pretty puzzle. Habr points out that a huge number of results in number theory are published with a caveat: they hold if the Riemann Hypothesis is true. A proof would remove that condition from many theorems at once, so the payoff would spread far beyond one famous question.
A rare prize, and a famous refusal
The article reminds readers how rarely these prizes are won. In the quarter of a century the prize list has existed, only one of the seven problems has been solved. The Russian mathematician Grigory Perelman proved the Poincaré conjecture in 2002-2003, after mathematicians had struggled with it for almost a hundred years.
In 2010 he was awarded the million dollars and turned the money down. Habr adds that he had earlier also refused the Fields Medal, one of the most prestigious honours in mathematics. The piece then suggests it may now be the turn of a second problem, and says the AI involvement is the most important part.
What the habr piece leaves out
The habr text is short and cut off at a 'read more' link, and it reads like an excitement piece rather than an investigation. It does not say how the proof was checked, who has reviewed it, or what techniques it uses. Those gaps matter more than the tone.
Other coverage fills some of them in. According to 36Kr, as summarized in our newsletter, the argument was checked in Lean, a program that verifies every logical step of a proof by machine, and the code is public. The same account says the method might extend to the generalized Riemann Hypothesis. Alpöge himself says the argument is massive and still has to be understood, organized and simplified.
That last point is a useful cooling note. A machine check is powerful because anyone can repeat it. But it only helps if the statement given to Lean really matches the famous problem, and experts have to confirm that.
How to read the news from here
Several outlets in different languages are repeating one source: the author's own post. That can look like many confirmations when it is really a single claim told many times. Decrypt also noted that the post appeared one day after Anthropic's Nasdaq listing. Timing is context, not proof either way.
The sensible reading is in three parts. The claim exists. A public machine check is described. Confirmation by independent mathematicians has not been reported. Until that third part arrives, a careful reader keeps the label 'claimed' on this result.
What to watch
- Will independent number theorists confirm the proof and the Lean statement?
- Does the method extend to the generalized Riemann Hypothesis, as one report suggests?
- Does the Clay Mathematics Institute respond to the claim?
Try it
When you see the same big claim on five sites, trace it back: find the original post and check whether any outlet did its own verification.
Sources and links
- Reportinghabr.com: the article we sourced
First reported by habr.com. 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.