AI cracks 80-year-old mathematics challenge — researchers are astonished

2026-05-22 · Source: Machine learning : nature.com subject feeds · Field: Science & Research — Mathematics & Computational Sciences, Artificial Intelligence & Machine Learning · Depth: Novice, quick

Summary

On May 20, OpenAI announced an experimental, general-purpose AI chatbot disproved Paul Erdős's 80-year-old unit-distance problem conjecture from 1946. Erdős had proposed what he believed was the optimal arrangement of points on a plane to maximize same-distance pairs, challenging others to do better. OpenAI's system achieved this by employing techniques from algebraic number theory, selecting points whose coordinates were solutions to specific equations. This finding, independently verified by mathematicians like Daniel Litt, has astonished the mathematical community. Experts call it the first autonomous, important research result by AI in any field. OpenAI has not fully disclosed the AI model's name or precise steps. A 125-page document details the "very long chain of thought" from a single open-ended prompt.

Key takeaway

For research scientists evaluating AI's potential in fundamental discovery, this event signals a significant shift. You should now consider general-purpose AI models as capable of autonomously generating novel, independently verifiable mathematical proofs. This breakthrough suggests integrating AI into early-stage problem formulation and hypothesis generation, beyond mere computational assistance. Explore how AI's "chain of thought" reasoning could accelerate your own research into long-standing conjectures.

Key insights

AI can autonomously solve complex, long-standing mathematical conjectures.

Principles

• AI can generate novel mathematical proofs.
• General-purpose AI models can achieve specialized breakthroughs.
• Algebraic number theory is a powerful tool for geometric problems.

Method

The AI used algebraic number theory to choose points with coordinates that were solutions of particular equations, generating a "very long chain of thought" from a single open-ended prompt.

In practice

• Use open-ended prompts for complex problem-solving.
• Explore algebraic number theory for geometric challenges.

Topics

• AI in Mathematics
• Unit-Distance Problem
• Paul Erdős Conjecture
• Algebraic Number Theory
• Autonomous AI Research

Best for: AI Scientist, Research Scientist, Tech Journalist

Related on AIssential

• An AI solution to an 80-year-old problem has shocked mathematicians — AI can autonomously solve long-standing mathematical conjectures by exploring vast solution spaces and combining existing knowledge.
• 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’s math breakthrough played to AI’s strengths — AI disproved an 80-year-old math conjecture by applying broad knowledge and persistent strategy exploration, marking a significant autonomous contribution.
• OpenAI cracks an 80-year math belief — A general-purpose AI model achieved novel mathematical discovery, signaling advanced AI capabilities for scientific breakthroughs.
• 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.
• I Solved an 'Impossible' Math Problem with AI — AI-assisted formal verification significantly lowers the barrier to entry for advanced mathematical theorem proving.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine learning : nature.com subject feeds.