OpenAI shifts the boundary of automated reasoning with a "milestone in AI mathematics" that experts are now unpacking
OpenAI's reasoning model just solved a 80-year-old math conjecture. Here's why that changes everything about AI capability benchmarks.
Why it matters
OpenAI has demonstrated a breakthrough in automated mathematical reasoning by using its model to disprove a longstanding conjecture in unit-distance geometry. This represents a concrete capability milestone that validates reasoning-focused model development and signals the emergence of AI systems that can tackle problems at the frontier of human mathematics—a domain traditionally used to measure AI progress.
The key facts
5 to knowOpenAI reasoning model disproved Paul Erdős conjecture on unit-distance geometry (open since 1946)
Fields Medalist Tim Gowers validated result as 'milestone in AI mathematics'
Model applied algebraic number theory tools in unexpected ways
Gowers quote: 'We have probably entered era where humans cannot compete with AI at solving mathematical problems'
Published May 21, 2026
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: A reasoning model from OpenAI has disproved a conjecture by mathematician Paul Erdős on unit-distance geometry that stood open since 1946 - using tools from algebraic number theory that experts never expected in this context. Fields Medalist Tim Gowers calls the result "a milestone in AI…