OpenAI Model Becomes First AI to Autonomously Solve a Major Open Math Problem
Original: OpenAI Model Becomes First AI to Autonomously Solve a Major Open Math Problem View original →
A Historic Milestone in Mathematics
On May 20, 2026, an internal OpenAI general-purpose reasoning model independently disproved a central conjecture in discrete geometry — the planar unit distance problem — first posed by Hungarian mathematician Paul Erdős in 1946. This marks the first time in history that an AI has autonomously solved a prominent open problem central to a field of mathematics, verified by an independent panel of leading mathematicians.
The Problem Explained
The question is deceptively simple: if you place n points on a flat plane, what is the maximum number of pairs of points that can be exactly distance 1 apart? For eight decades, mathematicians believed square-grid configurations were essentially optimal. OpenAI's model found an entirely new infinite family of constructions that beat the grid — and proved, using algebraic number theory, that these constructions yield a polynomial improvement over all previous results.
An Unexpected Approach
What makes the breakthrough remarkable is not just the result but the method. The model connected the geometry problem to algebraic number theory — a field studying number systems that extend ordinary integers — and used those concepts to build its proof. This cross-domain reasoning emerged from a general-purpose reasoning model, not a system specifically trained for mathematics. OpenAI reports the proof cost under $1,000 in compute.
Expert Verification
Princeton combinatorialist Noga Alon, one of the world's foremost experts on the Erdős problem, independently verified the proof alongside Melanie Wood and Thomas Bloom — the same Bloom who previously confirmed incorrect an earlier OpenAI claim about solving Erdős problems. All three published companion remarks alongside the formal paper, lending strong credibility to this result.
Sam Altman's Reaction
OpenAI CEO Sam Altman posted on X: "a general-purpose model solved a major open problem in mathematics. we'll be saying this a lot over the coming years, but this is a kinda big milestone. i'm very excited for AI to greatly extend our understanding of the world, but still, i have complicated feelings today."
Related Articles
HN read this math story less as another "AI did it" headline and more as a case where a model pointed at a route humans had not tried. The part that stuck was the expert cleanup work after the GPT-5.4 Pro draft, not the one-shot prompt itself.
The subreddit jumped straight past the headline and into the hard question: was this finally something other than pattern replay? A Scientific American report on a 23-year-old using GPT-5.4 Pro on a 60-year-old Erdos problem sparked debate over novelty, expert cleanup, and whether messy model output can still contain a real mathematical idea.
Google DeepMind unveiled an AI Co-Mathematician system — a multi-agent Gemini-based framework scoring 48% on FrontierMath Tier 4, the highest ever for any AI. AlphaEvolve improved lower bounds on five Ramsey numbers, including R(3,13) whose previous record had stood for 11 years.
Comments (0)
No comments yet. Be the first to comment!