OpenAI announces three major science discoveries including disproving Exponential Time Hypothesis

OpenAI has announced three significant new discoveries in the fields of mathematics, computer science, and physics, according to The Verge. These findings include disproving the Exponential Time Hypothesis, a longstanding question in computer science.

The Exponential Time Hypothesis was challenged by a deterministic algorithm that counts satisfying assignments of fixed-width CNF formulas in subexponential time and polynomial space. This result was independently verified by a group of computer scientists.

In mathematics, OpenAI's work involved proving local smoothing for the Euclidean wave equation using an auxiliary partial differential equation. This also addressed related conjectures including the Kakeya maximal and dimension conjectures and Stein’s spherical restriction estimates.

For this release, OpenAI adopted a new approach by involving external experts, referred to as "shepherds," to guide the results in computer science and physics. For the math problem, OpenAI released the AI output directly for researchers to analyze and develop explanations.

This approach marks a shift from OpenAI's previous mass release of hundreds of manuscripts, aiming for greater collaboration with the scientific community and transparency in sharing AI-generated research outputs.