A 1946 Erdős Conjecture Has Been Solved by Artificial Intelligence for the First Time

AI Breakthrough in Mathematics

According to НВ — Техно: A landmark moment in science arrived on June 2, 2026, when an artificial intelligence system solved a major open problem first posed by mathematician Paul Erdős in 1946. The challenge, which concerns the number of point pairs in a plane separated by exactly one unit distance, was cracked by a neural network that employed a high-dimensional lattice and algebraic integers to construct an intricate structure. Mathematician Will Sawin confirmed that the growth rate reaches at least n^{1.014}.

The response from the mathematical community was immediate. Fields Medal winner Tim Gowers called the achievement a

'milestone in AI mathematics'

. Professor Daniel Litt noted that for the first time, he is seeing

'an autonomous result from a neural network that is genuinely interesting in its own right, not just a technology demonstration'

.

New Frontiers for AI in Math

At the Joint Mathematics Meetings in January 2026, researchers had already been exploring how artificial intelligence could be applied to mathematical research. The great Hungarian mathematician Paul Erdős, when he first posed his problem, could not have imagined that decades later an AI would find a solution surpassing his expectations. The system recognized that the Erdős grid was too simplistic. Although the algorithm itself did not produce an exact figure, mathematician Will Sawin quickly verified the model, confirming the solution's success.

This case opens up new horizons for further research at the intersection of mathematics and artificial intelligence. The significance of this achievement lies not only in resolving a historic mathematical challenge but also in demonstrating AI's potential as a tool for making new scientific discoveries. Given AI's growing role across various fields, this breakthrough could spur deeper integration of technology into mathematical research and unlock new possibilities for automating complex calculations and data analysis.

This groundbreaking achievement in mathematics raises intriguing questions about the potential risks and benefits of AI in advanced problem-solving. Just as the recent collapse of an AI-led society highlights the complexities of artificial intelligence, the resolution of the Erdős Conjecture demonstrates how this technology can also lead to significant advancements in scientific understanding.