OpenAI announces three big science discoveries and a new way to work with academics

Key points

• OpenAI announced three new major discoveries in math, computer science and physics, adding results to GitHub today.
• The computer science result claims to disprove the widely believed Exponential Time Hypothesis, with verification guided by experts including Russell Impagliazzo, Ramamohan Paturi and Ryan Williams.
• OpenAI's own math advisers recommended yesterday that AI companies not adopt the practice of giving certain people first access to AI output.

OpenAI has announced three more "major" discoveries in math, computer science and physics, less than a month after releasing hundreds of manuscripts it said contained solutions to long-standing mathematics questions. The company added the results to GitHub today.

This time, OpenAI says external experts it calls "shepherds" guided the computer science and physics results. For the math problem, OpenAI says its advisers suggested releasing the AI output so researchers could work through their own explanations.

Details of the three results

The math output presents a proof of the local smoothing conjecture in every dimension, which OpenAI says also resolves the Kakeya conjecture and Stein's restriction conjecture in all dimensions. The computer science paper gives a deterministic algorithm for counting satisfying assignments of fixed-width CNF formulas in subexponential time and polynomial space, and OpenAI says the 3-SAT case refutes the Exponential Time Hypothesis.

The physics findings have not been released in full. OpenAI describes them as a mathematically rigorous existence proof of certain interacting quantum field theories containing gauge fields and fermions in four dimensions, including Quantum Chromodynamics and a unified completion of the Standard Model.

we recommend that AI companies do not adopt

— OpenAI's AGMAI math advisers

Expert review and criticism

A group led by Russell Impagliazzo, Ramamohan Paturi and Ryan Williams will examine and explain the computer science proof, and at its request OpenAI made the materials public immediately. OpenAI says David Gross will lead a group including Michael Douglas, Arthur Jaffe, Igor Klebanov and Nikita Nekrasov to shepherd the physics work.

OpenAI acknowledges the approach could have its own issues. Its AGMAI math advisers said yesterday that "we recommend that AI companies do not adopt" the practice of designating certain people to receive output first, citing unequal access, borrowed reputation and biased selection, and instead recommended releasing output on scholarly repositories with Lean-formalized proofs and other information.