Researchers Test the Exponential Time Hypothesis Claim
The computer science breakthrough centers on the Exponential Time Hypothesis, which concerns the Boolean satisfiability problem, or SAT. The ETH states that computing time for worst-case satisfiability problems grows exponentially with each added variable, offering no real shortcuts. OpenAI claims its AI system proved the ETH is false. If valid, this result would undermine hundreds of published papers in graph theory, computational biology, databases, and computational geometry that rely on the assumption as a foundational limit. MIT professor Ryan Williams, along with Russell Impagliazzo and Ramamohan Paturi of UC San Diego—who originally proposed the ETH—are independently checking the result without nondisclosure agreements. Williams noted on X that nobody knows yet what everyday developers will notice first, adding that practical SAT solving could see serious progress within 12 months or potentially nothing at all.
Local Smoothing Conjecture Solved in All Dimensions
In mathematics, OpenAI’s models produced a proof for the local smoothing conjecture proposed by mathematician Christopher Sogge in 1991. The conjecture states that extreme wave concentrations, such as those in sound, light, or earthquakes, last only briefly and average out smoothly over time. While Guth, Wang, and Zhang proved the conjecture in two dimensions in 2019, OpenAI states its models produced a proof for every dimension, including space and hypothetical higher-dimensional spaces. According to OpenAI, solving the local smoothing conjecture automatically proves two follow-on conjectures in all dimensions: the Kakeya conjecture from geometry, which Wang and Zahl only proved for three dimensions in 2025, and Stein’s restriction conjecture from Fourier analysis.
Honest answer: we don't know yet. At the current rate of progress, it's plausible we could see a serious update on practical SAT solving within the next 12 months. Or we might not see anything like that. I really don't know.
Quantum Field Theories Reaching Mathematical Foundations
The physics results address foundational questions in particle physics regarding quantum field theory. OpenAI reported that its AI proved certain quantum field theories exist mathematically in four dimensions—three spatial dimensions plus time—for theories involving interacting particles. The company stated its models produced the Yang–Mills proof on Sept. 9 and the quantum chromodynamics proof on Sept. 18, finishing the Lean formalization of the Yang–Mills existence proof on Oct. 16, qz.com reported. A group led by David Gross, who shared the 2004 Nobel Prize in Physics for his work on asymptotic freedom, is shepherding the physics work alongside Michael Douglas, Arthur Jaffe, Igor Klebanov, and Nikita Nekrasov. The proof addresses the existence part of the Yang–Mills theory tied to the Clay Mathematics Institute’s $1 million Millennium Prize Problems, though it does not address the mass gap portion.
OpenAI Review Process Generates Broad Debate
OpenAI’s dissemination strategy has sparked discussions within the scientific community regarding verification standards and access. The company published raw data, written proofs, and Lean code for certain domains while committing $5 million to the Simons Foundation for workshops and conferences. Williams’ vetting group published an open letter calling for a broader debate on how to verify AI breakthroughs, who receives early access, and how credit is assigned. OpenAI stated it is not claiming scientific credit for the discoveries, maintaining that credit belongs to the groups explaining and publishing the results.
Worth a look