Image copyrightReuters

Kari Hayes Technology Reporter

On the 20th, a mathematician affiliated with Anthropic, a major U.S. AI developer, claimed to have used an AI chatbot to prove the unsolved Riemann Hypothesis. Mathematician Levent Alpoge announced this on his website. The company's AI chatbot, Claude, was used to successfully prove the proof in six weeks. On the 19th, Anthropic was newly listed on the US NASDAQ, which deals with high-tech stocks.

If the proof of this ``Riemann Hypothesis'' is correct, it will be one of the most important proofs in history. Alpoge described it as "the dawn of something much greater for the scientific community at large."

In recent months, various AI companies, including OpenAI, have announced that their AI has solved difficult high-level math problems that have puzzled mathematicians for decades.

Alpoge said he used a private version of Claude to prove it.

The Riemann Hypothesis is a problem related to the distribution of prime numbers that was proposed by German mathematician Bernhard Riemann in 1859. A prime number is an integer greater than 1 and not divisible by anything other than 1 and that number, and Riemann proposed that these prime numbers have a fixed order on the number line.

This lighting problem is so complex that it has been selected as one of the Millennium Sweepstakes Problems, an unsolved mathematical problem selected by the Clay Mathematics Institute, and those who can prove it will receive a prize.

Prime numbers are the foundation of mathematics, and groundbreaking discoveries about prime numbers have a variety of effects.

Alpoge said he contacted Anthropic CEO Dario Amodei on the 5th and asked him directly how to announce the discovery.

Alpoge, referring to Anthropic's initial public offering, wrote that the announcement "couldn't have come at a worse time" and that "I would have ended this whole thing a lot sooner if I could have."

"It's clearly my fault that I felt like I couldn't survive if I waited any longer [for the announcement]. Of course there's some cynicism about the timing, but that's wrong."

Mr. Alpoge presented proof of his proof in Lean, a computer language used to verify mathematical proofs. The proof spans over 250,000 pages.

Renowned mathematician Terrance Tao said on the 20th that it will still take time to verify the proof.

"I think the process of accepting [the publication] will take some time, but at least this time I can communicate with a human author who can help me with this process," Tao wrote.

This year, a number of mathematical proofs using AI have been published, but most of them clearly did not involve humans.

In September, OpenAI announced that its employees, using a private version of its AI tool, had solved part of the most difficult Navier-Stokes equations in just 88 hours. The Navier-Stokes equations are related to the movement of fluids.

More than 10,000 AI agents, or bots programmed to operate with some degree of autonomy, were deployed on the mission as an AI training exercise.

The accuracy of the proof published by Open AI for the Navier-Stokes equations has not yet been confirmed. The company's announcement sparked significant criticism from academia, the higher mathematics community, and related organizations about the use of AI to solve math problems.

Anthropic has previously claimed to have found other advances and solutions in the fields of mathematics and physics.

A group of mathematicians said it was "detrimental to the science of mathematics" for AI companies to tout their major solutions to demonstrate the sophistication of their products.

“We are witnessing a general threat to intellectual labor, with a disconnect between the outcomes of AI use and its original purpose,” the academics said in a joint letter.

A month later, OpenAI published solutions, proofs, or claims of significant progress for 372 unsolved math problems.

It will likely take several months or more for experts to verify this claim and prove its validity.

(English article In a year of AI solving maths problems, Anthropic may have just solved the biggest)

See more

U.S. NASDAQ plunges, Anthropic stocks also put selling pressure on tech stocks

Is AI a threat to scientific research? Concerns fueling controversy over mathematics

US Anthropic warns of ``catastrophic risks'' to humanity ahead of major listing

video

Video, What did the US, China, and Russia leaders gain from the meeting? BBC correspondents' opinions, Duration 2,02

Video, US and Russian presidents shake hands, President Xi looks on with a smile at APEC summit, Duration 0,31

Video, BBC report from the bridge in Kiv attacked by Russia "Battle for the Bridge", duration 1,14

Video, Trump says 'we won' Iran war at Veterans Day ceremony, duration 1,02

Top articles

UK and EU agree on food exports, youth programs, etc. British Prime Minister Burnham says they will consider deepening ties

Russian missiles land in Poland, Prime Minister Tusk, Ukraine's allies conduct exercises nearby

Anthropic mathematician claims that AI can prove unsolved mathematical problem ``Riemann Hypothesis''

Reading material/commentary

Featured articles