The Jacobian conjecture, a problem that resisted mathematicians for 87 years, has been shown false thanks to a collaboration between mathematician Levent Alpöge and Anthropic's AI model Fable 5. The counterexample, a single polynomial map from three-dimensional space to itself, was posted on X and verified by hand Source: The Next WebSource: CoinDesk.
The Jacobian conjecture, posed by Ott-Heinrich Keller in 1939, asserts that any polynomial map with a non-zero constant Jacobian determinant must have a polynomial inverse. Alpöge and Fable 5 found a specific map that meets the condition but lacks a polynomial inverse, disproving the conjecture Source: The Next Web. The result is a short list of polynomials that any mathematician can check Source: The Next Web.
News of the disproof stirred the mathematical community. Fields medalist Timothy Gowers called it the first time an AI had solved a problem he had heard of outside his area Source: The Next Web. Daniel Litt posted that he could not stop laughing. However, some were unimpressed: Andrew Blumberg said the result did not change his expectations, as finding counterexamples is precisely what AI should be good at Source: The Next Web. The OpenAI researcher Aaron Lou claimed that an internal version of its Codex model found essentially the same counterexample independently Source: The Next Web.
Beyond mathematics, the breakthrough has implications for Bitcoin and AI-linked markets. Bitcoin miners have pivoted to AI data centers, so demand for computing power ties cryptocurrency to AI developments Source: CoinDesk. The news added to the narrative that AI capabilities are leaping forward, affecting investor sentiment Source: CoinDesk.
“Together they had produced a counterexample to a problem that had resisted mathematicians for 87 years. It was 216 characters long.”
“An artificial intelligence model has helped disprove a mathematical conjecture that had gone unsolved for 87 years, and the answer was confirmed within a day because anyone can now check it by hand.”
“Timothy Gowers, a Fields medallist, said it was the first time an AI had solved a problem outside his own area that was big enough for him to have heard of it.”
“Andrew Blumberg, a mathematician at Columbia who sits on a project testing AI on research maths, was pointedly unmoved. 'This did not cause me to update my priors,' he told Mashable.”
“Bitcoin's biggest holders, its miners, have rebuilt themselves into AI data-center operators, so their fortunes now rise and fall with demand for computing power.”
“The OpenAI researcher Aaron Lou said an internal version of the company's Codex model found essentially the same counterexample on its own.”