OpenAI Sets Up Math Advisory Group After Model Solves 100 Open Problems

OpenAI has formed the Advisory Group on Mathematics and Artificial Intelligence to oversee the publication of more than 100 classic math problems that have been solved by the lab’s AI system. 

Meeting at the Institute for Advanced Study in Princeton, New Jersey, the nine-member panel will serve to connect advances in automated reasoning with the wider mathematics community. This comes after the claim by OpenAI researchers that the lab had solved the celebrated Navier-Stokes Millennium Prize problem.  

The non-compensated advisory board comprises renowned mathematicians like Edward Witten, Timothy Gowers, and Martin Hairer. They will provide help in organizing and coordinating the publication of theoretical proofs generated by the AI system. 

However, OpenAI made it clear that the board is not responsible for slowing down or mandating any research within the company’s internal systems. Group formation is amidst tension between the research community. In the last few weeks, 25 Fields Medalists signed an open letter voicing their concerns regarding the hasty release of unverified proofs by technology laboratories. 

With the goal of ensuring transparent proof verification mechanisms for machine-driven scientific discoveries, OpenAI is offering early insights of AI-derived proofs to academic authorities. The move is indicative of a wider trend of industry towards specialized domain governance due to the emergence of artificial intelligence solving symbolic reasoning problems. 

Formalizing proof verification channels guarantees scientific validity while running high compute models in theoretical areas. OpenAI’s new advisory board serves as an academic oversight framework for the contribution of artificial intelligence to pure mathematics.