American company OpenAI has announced the formation of an independent group of expert mathematicians to evaluate scientific results and coordinate their publication, TechCrunch reports. The move establishes an independent advisory body — the Advisory Group on Mathematics and Artificial Intelligence, according to infohub.kz.
The advisory group is based at the Institute for Advanced Study in Princeton, New Jersey. The council's creation came in response to criticism from the scientific community and the emergence of new results produced by the company's internal artificial intelligence model.
OpenAI previously stated that its new AI system not only proposed a counterexample to the hypothesis of the existence and smoothness of solutions to the Navier–Stokes equations from the Millennium Prize Problems, but also resolved more than a hundred long-standing open mathematical problems. The pace of progress has raised concerns among scientists.
In particular, 25 Fields Medal laureates signed an open letter expressing fears that the race among AI laboratories to solve famous problems poses risks to fundamental science and requires external oversight.
The new advisory group includes nine prominent scholars, among them Camillo De Lellis, Timothy Gowers, Martin Hairer, and Edward Witten. The experts will work without financial compensation from OpenAI.
The group is empowered to assess the significance of discoveries, issue public statements, and provide recommendations on scientific publication standards. However, the council's mandate does not include influencing the pace of OpenAI's own research or the right to halt the company's internal development.