Claude Fable Produced a Counterexample to the Jacobian Conjecture
Anthropic's AI model Claude Fable appears to have generated a counterexample to the famous Jacobian Conjecture, one of mathematics' long-standing unsolved problems. The news spread rapidly on Hacker News, gathering over 268 points and 152 comments. If the result holds up to scrutiny, it could represent a historic breakthrough in algebraic geometry.
The Jacobian Conjecture is one of mathematics' most stubbornly unsolved problems, originally formulated in 1939. It states that if a polynomial map from n-dimensional complex space to itself has a Jacobian determinant identically equal to one, then the map is a bijective polynomial with a polynomial inverse. Despite decades of research and numerous flawed proofs in both directions, no one has managed to definitively determine whether the conjecture is true or false.
Now, Anthropic's AI model Claude Fable claims to have found a counterexample that disproves the conjecture. The claim appeared on the platform X (formerly Twitter) and immediately spread to Hacker News, where it generated lively discussion among mathematicians, AI researchers, and tech enthusiasts. With 268 points and over 152 comments, the engagement shows just how much attention this claim has captured across the community.
Reactions on Hacker News are mixed. Some mathematicians express skepticism, noting that the Jacobian Conjecture has a long history of flawed proofs and counterexamples that were later shown to be incorrect. Others are more optimistic, viewing this as a sign that large language models are beginning to demonstrate genuine mathematical capability beyond pattern recognition and text generation. The key question remains whether Claude actually reasoned through the mathematics or produced something that merely looks correct on the surface.
Regardless of whether the counterexample survives mathematical scrutiny, the event marks a significant moment in AI history. The fact that an AI model can produce mathematical claims compelling enough to be taken seriously by researchers signals a new era for computational mathematics. The research community is now urged to carefully verify the result, and the outcome could have far-reaching consequences for both pure mathematics and the future development of AI reasoning systems.