OpenAI on October 26 announced three results it calls discoveries in mathematics, computer science, and physics, covering harmonic analysis, complexity theory, and quantum field theory1,2. The headline claim: the Exponential Time Hypothesis, a longstanding conjecture in computational complexity, is false.
The announcements arrive less than a month after OpenAI's mass release of hundreds of manuscripts that it said contained solutions to long-standing mathematics questions.
What OpenAI posted
One result, added to GitHub on October 26, addresses local smoothing for the Euclidean wave equation. According to OpenAI's description, the proof uses an auxiliary partial differential equation to establish Sogge's local smoothing for waves in every spatial dimension of at least two. OpenAI says the result also yields the Kakeya maximal and dimension conjectures and Stein's spherical restriction estimates throughout the interior of the conjectured range.
The other two results concern complexity theory, where OpenAI claims to have disproved the Exponential Time Hypothesis, and quantum field theory. Details on those results beyond the subject areas are not disclosed in the available sources.
A different release process
OpenAI changed its approach to publication this time. The company enlisted external experts it described as "shepherds" to guide the computer science and physics results. For the mathematics problem, OpenAI said those external advisers recommended releasing the AI-generated output so that researchers could work through their own explanations.
ANALYSIS The shepherd model marks a procedural shift from OpenAI's earlier bulk manuscript release, which drew scrutiny over whether AI-generated proofs had been adequately vetted before publication. Whether the complexity-theory and physics claims survive peer review will determine whether the new process adds substantive rigor or merely reputational cover.
The results remain unverified by independent reviewers at the time of the announcement. The Verge characterized the three findings as "major" in quotation marks, reflecting OpenAI's own framing.
ANALYSIS A confirmed disproof of the Exponential Time Hypothesis would rank among the most consequential results in theoretical computer science, with direct implications for algorithm design and cryptographic assumptions. The gap between the claim and independent verification is the operative fact for practitioners tracking the announcement.