Recently, OpenAI officially announced that its artificial intelligence models have achieved three major scientific discoveries in the fields of mathematics, computer science, and physics, respectively involving harmonic analysis, computational complexity theory, and quantum field theory. In order to transform these cutting-edge raw research data into genuine scientific understanding, OpenAI has innovatively attempted a new model of deep collaboration with the external academic community, and has developed diversified release strategies for the results in different fields.
In the field of harmonic analysis, the model successfully provided a proof of the local smoothing conjecture in all dimensions, thereby directly resolving the mathematics community's Kakeya conjecture and Stein restriction conjecture in all dimensions. The new proof is based on a novel auxiliary partial differential equation. OpenAI has released the AI natural-language output reports generated by the model as well as Lean proofs in a public GitHub repository.
In the field of computer science, the system proved that the exponential time hypothesis (ETH), long assumed to be true, is false. This conjecture is closely related to the famous P vs. NP problem and has previously been widely used in hundreds of papers to derive conditional computational hardness. This result was likewise verified in Lean. This work was handed over to a top expert team led by Russell Impagliazzo, Ramamohan Paturi, and Ryan Williams for verification and guidance, and the related results and Lean formalized proofs have been made public in a GitHub repository.
In the field of physics, the model provided mathematically rigorous existence proofs for certain interacting quantum field theories in four dimensions that include gauge fields and fermions (including quantum chromodynamics, QCD, and the unified completion of the Standard Model of particle physics), including as a subproblem a proof of the existence of pure Yang–Mills theory, which is also part of the Clay Mathematics Institute's Millennium Prize Problems. This complex work was advanced by a group of physicists led by David Gross (including Michael Douglas, Arthur Jaffe, Igor Klebanov, and Nikita Nekrasov), and related correspondence may be sent to the designated email address: YANGMILLS4D@gmail.com.
In order to fully support the academic community's independent review and in-depth discussion of these groundbreaking results, OpenAI has previously made a donation of 5 million US dollars to the Simons Foundation to fund related seminars, conferences, and special projects. This series of measures not only demonstrates the enormous potential of AI in cutting-edge fundamental scientific research, but also marks an acceleration of modern theoretical physics and mathematics toward higher standards of rigor.