OpenAI says its AI cracked three major problems in math, computing and physics

OpenAI says an internal AI system has produced three potentially major scientific results spanning mathematics, theoretical computer science and physics. The company is releasing the work through different review paths and describes the outputs as raw research data that still needs expert interpretation.

In mathematics, OpenAI claims a proof of the local smoothing conjecture in every dimension, which would also resolve the Kakeya and Stein restriction conjectures in all dimensions. The company has published both a natural-language manuscript and a Lean formalization.

In computer science, OpenAI says the model disproved the Exponential Time Hypothesis, a longstanding assumption used to derive hardness results across algorithms, databases, computational biology and other fields. A group led by Russell Impagliazzo, Ramamohan Paturi and Ryan Williams will examine and explain the result; the proof and Lean artifact are public.

The physics claim is broader and less publicly inspectable for now: OpenAI says its models produced a rigorous existence proof for interacting four-dimensional quantum field theories including quantum chromodynamics, alongside work related to pure Yang–Mills theory. A group led by Nobel laureate David Gross will review the material before deciding how to publish it.

If these claims survive independent scrutiny, they would mark a shift from AI assisting scientists to AI generating results on central open problems. The caveat is essential: formalization can validate internal logical steps under encoded assumptions, but significance, framing and possible gaps still require expert review.