OpenAI reasoning model disproves 80-year-old Erdős geometry conjecture

2026-05-21 · Source: Dataconomy · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Mathematics & Computational Sciences · Depth: Fundamental Awareness, quick

Summary

OpenAI claims its new general-purpose reasoning model has produced an original mathematical proof disproving an 80-year-old geometry conjecture by Paul Erdős, first posed in 1946. This follows previous, criticized assertions by former VP Kevin Weil regarding GPT-5's supposed solutions to Erdős problems, which were later revealed to be existing literature findings. Mathematicians Noga Alon, Melanie Wood, and Thomas Bloom now support OpenAI's current claim, with Bloom having previously called Weil's earlier statement "a dramatic misrepresentation." The company states its model disproved the long-held belief that optimal solutions resembled square grids, instead discovering a new family of constructions. This marks the first time AI has autonomously solved a significant open problem in mathematics, suggesting AI systems are improving at connecting ideas across fields like biology, physics, engineering, and medicine.

Key takeaway

For research scientists exploring complex mathematical problems, this development indicates a shift in AI's capability from identifying existing solutions to generating original proofs. You should consider integrating advanced reasoning models into your research workflows to challenge long-held assumptions and discover novel constructions. This could accelerate breakthroughs in fields beyond pure mathematics, including biology and physics.

Key insights

A general-purpose AI model has autonomously disproved an 80-year-old mathematical conjecture, demonstrating advanced reasoning capabilities.

Principles

• AI can generate original mathematical proofs.
• General-purpose AI connects ideas across disciplines.
• Long-held mathematical beliefs can be overturned by AI.

In practice

• Explore AI for novel mathematical discoveries.
• Apply general AI models to diverse scientific fields.

Topics

• AI Reasoning
• Mathematical Proof
• Erdős Conjecture
• Discrete Geometry
• General-Purpose AI
• Scientific Discovery

Best for: AI Scientist, Research Scientist, Tech Journalist

Related on AIssential

• OpenAI Says Its AI Just Disproved a Famous 80-Year-Old Math Problem – And This Time, Mathematicians Agree — A general-purpose AI autonomously disproved an 80-year-old math conjecture by making a cross-domain mathematical leap.
• 😸 OpenAI solved an 80-year math problem by... disproving it — An OpenAI model disproved an 80-year math conjecture, showcasing AI's advanced reasoning and verifiable problem-solving capabilities.
• OpenAI claims it solved an 80-year-old math problem — for real this time — OpenAI's new model disproved an 80-year-old math conjecture, demonstrating advanced autonomous reasoning in AI.
• OpenAI claims a general-purpose reasoning model found a counterexample to Erdos's unit-distance bound [D] — A general-purpose AI model found a counterexample to a long-standing mathematical conjecture, highlighting AI's potential in complex problem-solving.
• An OpenAI model has disproved a central conjecture in discrete geometry — An OpenAI model resolved Erdos Problem 90, demonstrating AI's capacity to solve complex, long-unsolved mathematical conjectures and sparking debate on AI novelty.
• OpenAI makes breakthrough on 80-year-old maths problem — OpenAI's AI disproved an 80-year-old math conjecture, showing AI can find novel solutions in complex problems.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Dataconomy.