- Published
An Anthropic employee and mathematician claims to have used the company's artificial intelligence (AI) chatbot to prove the famed Riemann hypothesis.
uncertain
employee → publish → hypothesis
Levent Alpöge said, external on Friday, a day after Anthropic made its US stock market debut, that over the course of about six weeks, he used the Claude chatbot to create a detailed proof of the maths problem.
If found to be correct, the proof would be among the most significant in history.
asserted
proof → say → history
The claim comes as AI firms like OpenAI have in recent months announced scores of solutions to major higher maths problems that have eluded scholars for decades.
uncertain
that → come → decades
As for the Riemann hypothesis proof put forward on Friday, it was found using a non-public version of Anthropic's Claude.
It is a nearly 170-year-old maths problem that deals with prime numbers and is widely considered to be the single most famous problem in maths, in large part due to how complex it is.
Named for the mathematician Bernhard Riemann, the hypothesis puts forward that prime numbers have a determined order along the number line.
asserted
numbers → put → line
Prime numbers are the bedrock of higher maths, leaving breakthroughs related to them to have broad impact.
asserted
numbers → leave → impact
Alpöge said that he messaged Anthropic CEO Dario Amodei on Sunday, asking him directly about how he could announce the discovery.
uncertain
he → say → discovery
"The precise timing is horrible," Alpöge said, referring to Anthropic's public stock listing.
asserted
Alpöge → say → listing
"If I could have, I'd have finished this all much earlier."
uncertain
I → have → this
"I am clearly unilaterally responsible for feeling I just could not live if I waited even longer.
uncertain
I → live → ?
There is an obvious cynical view, to do with the timing, but it is false," he added.
asserted
he → be → timing
The complete proof that Alpöge put forward in Lean, a computer language used for confirming maths proofs, is over 250,000 lines long.
asserted
Alpöge → put → proofs
Terrance Tao, a prominent mathematician, said on Friday, external that verification of the proof is still some way off.
"I imagine the digestion process will take some time, though in this case at least we have a human author to communicate with to assist with this task," Tao said.
asserted
Tao → say → task
No such human was obviously a part of major maths proofs made earlier this year using AI.
asserted
human → make → AI
OpenAI in September claimed that employees using a non-public version of its AI tools were able to solve part of a notoriously difficult problem in just 88 hours.
asserted
employees → claim → hours
The problem was part of the Navier-Stokes equations, which concern how fluids move.
asserted
fluids → concern → equations
More than 10,000 AI agents, or computer programmed bots that operate somewhat autonomously, were put to the task as an AI training exercise.
asserted
that → program → exercise
The proof that OpenAI released on Navier-Stokes has yet to be confirmed as accurate, and it set off criticism within academic and higher maths communities and organisations about the use of AI to solve maths problems.
asserted
it → release → problems
Anthropic has also claimed to have made other progress and solutions in maths, external and physics, external.
asserted
Anthropic → claim → maths
A group of maths scholars said, external AI companies boasting of major solutions to show how sophisticated their commercial products had become was "detrimental to the science of mathematics".
asserted
products → say → mathematics
"We are witnessing a general threat to intellectual work, with misalignment between the outcome of the use of AI and its initial purpose," the scholars said in a group letter.
asserted
scholars → witness → letter
A month later, OpenAI released, external what it claimed were solutions, proofs, or major progress on 372 more outstanding higher maths problems.
asserted
it → release → problems
It will take months, if not longer, for those claims to be reviewed by experts to prove their validity.
uncertain
claims → take → validity