OpenAI Announces Three Scientific Research Findings from Its Model Outputs

OpenAI Announces Three Scientific Results Generated by Its Models

OpenAI has released three results generated by its internal models: a proof of the local smoothing conjecture across all dimensions, a refutation of the Exponential Time Hypothesis (ETH), and a proof of the existence of four-dimensional Yang–Mills theory.

OpenAI describes these results as more akin to raw research data and has invited external experts to form groups to interpret and verify them. The three groups have adopted different publication strategies: the computer science group has immediately made the proofs and their formalization in Lean public, while the physics group has delayed its release.

Editor's Note: If these three proofs are ultimately validated, each would represent a historic breakthrough in its respective field. Among them, the Exponential Time Hypothesis (ETH) is one of the most significant complexity conjectures, second only to core problems like P vs. NP. Its refutation would undermine the lower bounds upon which numerous exact algorithms, parameterized complexity studies, and fine-grained complexity research depend. The proof of local smoothing in all dimensions would also resolve central conjectures in harmonic analysis, such as the Kakeya conjecture and Stein's restriction conjecture. Meanwhile, the proof of the existence of four-dimensional Yang–Mills theory directly addresses one of the seven Millennium Prize Problems.

OpenAI: https://openai.com/index/building-scientific-understanding-with-academic-communities/