Skip to main content

Kaer NewsScience

OpenAI announces three scientific discoveries in math, computer science, physics

On 26 October 2026 OpenAI announced discoveries including a proof that the Exponential Time Hypothesis is false, a local smoothing proof resolving the Kakeya conjecture, and a physics existence proof.

4 min read
OpenAI announces three scientific discoveries in math, computer science, physicsPhoto: The Verge

Key points

OpenAI announced three discoveries in mathematics, computer science and physics, including a proof that the Exponential Time Hypothesis is false.

OpenAI announced on 26 October 2026 three discoveries it calls major, spanning mathematics, computer science and physics. The computer science result claims to show the Exponential Time Hypothesis, long believed true, is false. The mathematics result claims a proof of the local smoothing conjecture resolving the Kakeya and Stein restriction conjectures in all dimensions. The physics result is described as a mathematically rigorous existence proof for certain four-dimensional quantum field theories, though that information has not been released in full.

The announcement matters because these are claims about long-standing open problems, not answers to questions posed by researchers. OpenAI says its models produced the results independently. The Exponential Time Hypothesis, if refuted, would change what complexity theorists believe about how fast exact algorithms can run. The mathematics claims would extend results beyond what researchers had proved by hand. The physics claim addresses existence of quantum field theories including Quantum Chromodynamics, a core theory of particle physics.

What the three claims are

For computer science, OpenAI says the result gives a deterministic algorithm that exactly counts satisfying assignments of fixed-width CNF formulas in subexponential time and polynomial space, and that the 3-SAT case refutes the Exponential Time Hypothesis. The Exponential Time Hypothesis, proposed in 1999, holds that certain problems cannot be solved dramatically faster than brute force. If false, faster algorithms may be possible. An independent group of computer scientists will examine and explain the proof.

OpenAI says the mathematics model produced a proof of the local smoothing conjecture in every dimension, and therefore also resolved the Kakeya conjecture and Stein's restriction conjecture in all dimensions. The proof uses an auxiliary partial differential equation to establish Sogge's local smoothing for waves in every spatial dimension of at least two, and also yields the Kakeya maximal and dimension conjectures and Stein's spherical restriction estimates through the interior of the conjectured range. Just months earlier, NYU professor Hong Wang received a Fields Medal for resolving the Kakeya conjecture in three dimensions.

Verification differs across the three results. For computer science, a group of three experts including UC San Diego's Russell Impagliazzo and Ramamohan Paturi, who were part of the group that proposed the Exponential Time Hypothesis in 1999, and MIT professor Ryan Williams, will guide verification. The group decided to release the AI output write-up and Lean proof on GitHub. For physics, David Gross will lead a group including Michael Douglas, Arthur Jaffe, Igor Klebanov and Nikita Nekrasov to shepherd the work.

Who is checking the proofs

The physics result is described as a mathematically rigorous existence proof of certain interacting quantum field theories containing gauge fields and fermions in four dimensions, including Quantum Chromodynamics and a unified completion of the Standard Model of particle physics. The physics information has not been released in full. OpenAI said the mathematics and computer science results were formally verified using the Lean mathematical verification tool and that proof materials were made public, while the physics results are under review by experts including Nobel laureates in physics.

OpenAI says it reached out to external experts and asked them to act as shepherds, organising groups to understand the new results. This differs from its release weeks earlier of hundreds of manuscripts it said contained solutions to long-standing mathematics questions, which drew criticism for bypassing academic peer review. The AGMAI panel of mathematics advisers said it recommends that AI companies not adopt practices of giving certain people early access to output, citing unequal access, borrowed reputation and biased selection.

The computer scientists leading verification published an open letter asking the community how early access to AI-generated results should be governed, how credit should recognise both discoveries and the work of checking and extending them, and how students and researchers can be supported as methods change. For the mathematics problem, OpenAI says advisers recommended releasing the AI output directly for researchers to work through their own explanations, as in its 6 October release. The findings still require independent confirmation by the academic community.

Frequently asked questions

What three discoveries did OpenAI announce?

On 26 October 2026 OpenAI announced a claimed proof that the Exponential Time Hypothesis is false in computer science, a local smoothing proof resolving the Kakeya and Stein restriction conjectures in all dimensions in mathematics, and an existence proof for certain four-dimensional quantum field theories in physics.

Who is verifying OpenAI's computer science proof?

An independent group including UC San Diego's Russell Impagliazzo and Ramamohan Paturi, who helped propose the Exponential Time Hypothesis in 1999, and MIT professor Ryan Williams will examine and explain the proof. The group requested the materials be made public immediately.

Has the physics result been released in full?

No. OpenAI described the physics result as a mathematically rigorous existence proof of certain four-dimensional quantum field theories including Quantum Chromodynamics, but the information has not been released in full and is under expert review.

More from this edition