OpenAI has announced three scientific breakthroughs achieved by its AI models in mathematics, computer science, and physics. The company claims the models independently discovered and proved complex theories rather than merely assisting in research.
In mathematics, the AI provided a proof for the local smoothing conjecture across all dimensions, which also resolved the Kakeya and Stein restriction conjectures. The proof relies on a newly constructed auxiliary partial differential equation.
In computer science, the system disproved the Exponential Time Hypothesis (ETH), a long-standing assumption in computational complexity. OpenAI has published both natural language and Lean-verified proofs on GitHub.
The physics breakthrough involves a mathematical existence proof for certain quantum field theories in four dimensions, including quantum chromodynamics and Yang-Mills theory, which is one of the seven Millennium Prize problems.
OpenAI is collaborating with academic experts to review these findings. While math and computer science results have been formally verified using Lean, the physics proof is currently under review by experts, including Nobel laureates.
