A long-standing mathematics problem that challenged some of the world’s brightest minds for nearly 80 years has been cracked by an artificial intelligence model developed by OpenAI.
The breakthrough involves the famous “unit distance problem,” first posed in 1946 by legendary Hungarian mathematician Paul Erdos. The problem asks a surprisingly simple question: if you place n points on a flat surface, how many pairs of points can be exactly one unit apart from each other?
While the question sounds straightforward, finding the maximum possible number of such pairs has puzzled mathematicians for decades.
According to a report by The Wall Street Journal, OpenAI researchers gave the problem to an advanced AI model to test its mathematical reasoning abilities. What happened next surprised even the experts working on the project.
Instead of confirming Erdos’s famous conjecture, the AI found a counterexample that showed the conjecture was not always true. In mathematics, this is known as a disproof. The model constructed a family of point arrangements that produced more unit-distance pairs than mathematicians previously believed possible.
The AI’s result shows that for some large point sets, the number of unit-distance pairs can grow faster than the upper bound that many researchers expected. The finding settles a major question in combinatorial geometry and changes how mathematicians understand the problem.
Researchers initially suspected there might be a mistake.
“I initially didn’t believe it,” said Mehtaab Sawhney, a mathematician at OpenAI, as per WSJ.
The team spent considerable time checking the proof, using both human experts and AI tools to verify the work. Eventually, outside mathematicians were brought in to review the result.
The reaction from the mathematics community has been striking.
“AI was able to do here what lots of excellent human researchers tried and failed to do,” said Princeton professor Noga Alon.
Daniel Litt of the University of Toronto said, “This is the first example of a result produced autonomously by an AI that I find exciting in itself, as opposed to as a leading indicator.”
Perhaps the strongest endorsement came from Fields Medal winner Timothy Gowers, who said: “There is no doubt that the solution to the unit-distance problem is a milestone in AI mathematics.”
Researchers believe one reason the AI succeeded is that it explored unusual ideas that many humans had dismissed. The solution combined concepts from algebraic number theory and discrete geometry.
The achievement is being viewed as one of the clearest signs yet that AI can contribute to original scientific and mathematical research, rather than simply assist with calculations or summarise existing knowledge.
