Summary (tl;dr)

Mathematician Levent Alpöge, leveraging Anthropic's AI model Claude Fable 5, has announced the discovery of a counterexample that disproves the Jacobian conjecture, an 87-year-old unsolved problem in mathematics.

Essential Background

The Jacobian conjecture, initially posed in its general form in 1939 by Ott-Heinrich Keller, is a fundamental problem in algebraic geometry. It asks whether a polynomial function from n-dimensional space to itself, with a non-zero constant Jacobian determinant, must always possess a polynomial inverse. This essentially tests if local invertibility implies global invertibility for polynomial maps. The conjecture was famously included in Steve Smale's 1998 list of significant mathematical problems for the upcoming century and has gained a reputation for numerous erroneous proofs over the decades, leading to a cautious stance within the mathematical community regarding any claimed solutions.

The Full Story

On July 19, 2026, during the World Cup final, mathematician Levent Alpöge posted on X (formerly Twitter) that the Jacobian conjecture is, in fact, false. He credited Anthropic's AI model, Claude Fable 5, with discovering the counterexample, reportedly at the suggestion of a friend named Akhil. The discovered counterexample is a polynomial map in three dimensions that demonstrates a constant non-zero Jacobian determinant (specifically, -2) but is not invertible because it maps multiple distinct inputs to the same output. This significant announcement quickly led to updates on the Jacobian conjecture's Wikipedia page and is currently undergoing rapid verification by other mathematicians, highlighting a pivotal moment for both advanced mathematics and the burgeoning role of AI in scientific research.

Why It Matters

The disproof of the Jacobian conjecture represents a monumental achievement in mathematics, settling a problem that has eluded resolution for over 87 years and was considered one of the paramount challenges in algebraic geometry. Beyond its direct impact on pure mathematics, this development underscores the rapidly evolving capabilities of artificial intelligence in contributing to complex scientific discovery. The involvement of Claude Fable 5 as a "genuine research collaborator" with a human mathematician, identifying a verifiable counterexample in a remarkably short period, signals a new era where AI can effectively tackle long-standing, difficult problems previously thought to be exclusive to human intellect. This breakthrough could pave the way for accelerated mathematical discoveries and fundamentally alter research methodologies in highly abstract fields.

Geographic Location

• None