OpenAI has published three new research findings in mathematics, computer science, and physics on GitHub. The company is now utilizing external experts, referred to as shepherds, to guide the verification of results in computer science and physics. For the mathematics discovery, OpenAI released the model output directly to allow researchers to develop their own explanations.
The computer science breakthrough claims to disprove the Exponential Time Hypothesis. A working group led by Russell Impagliazzo, Ramamohan Paturi, and Ryan Williams will examine the proof. The math discovery provides a proof for the local smoothing conjecture, which also resolves the Kakeya and Stein’s restriction conjectures in all dimensions.
The physics research involves a mathematically rigorous existence proof for interacting quantum field theories, including Quantum Chromodynamics and a completion of the Standard Model. David Gross will lead a group of experts, including Michael Douglas, Arthur Jaffe, Igor Klebanov, and Nikita Nekrasov, to oversee this work.
The AGMAI panel of math advisers recently recommended that AI companies avoid designating specific individuals to receive research outputs first. They cited concerns regarding unequal access and biased selection. The panel suggests that labs should instead use scholarly repositories and provide Lean-formalized proofs for future disclosures.
