In early October, OpenAI posted more than 700 AI-generated mathematical manuscripts in an online public repository. If the arguments withstand checking, they could contain a surprising amount of new mathematical knowledge. The response from mathematicians has ranged from excitement to despair: some fear that decades of work could suddenly appear meaningless, while others worry about the effect on younger researchers entering a field in rapid transition. Many are also angry about how AI companies are releasing such material.
Concerns about transparency and verification
As one of the authors of the Leiden Declaration on Artificial Intelligence and Mathematics, published in June, I believe many of these concerns about OpenAI’s approach were predictable. The declaration does not oppose the use of AI. Instead, it proposes practical safeguards, including clear citation policies, accurate disclosure of how AI has been used, and a renewed commitment to the long-standing principle that every argument in a mathematical proof should be independently verifiable and that human authors should be able to explain and answer for it. Some AI companies have agreed to follow the declaration. OpenAI has not.
Mathematicians are developing rules for using AI – other fields should follow
Now that mathematicians have started examining the AI-assisted proofs, it is clear that many of the 700-plus preprints are written in a rough, sketchy style that makes them difficult to understand. The inner workings of OpenAI’s closed model are unknown, and no human author has been identified as responsible for explaining or correcting possible flaws. Checking the manuscripts and rewriting them in a professional form suitable for peer review requires substantial time and effort. Understandably, many mathematicians are reluctant to clean up the OpenAI material, regardless of how interesting the underlying mathematics might be.
Who should shape mathematics in the age of AI?
What does this mean for the future of mathematics? Rethinking how the field should evolve will require action at every level. Leadership from mathematicians such as Terence Tao and grassroots initiatives such as the Leiden Declaration should be complemented by engagement from major stakeholder organizations, including the International Mathematical Union. It would be a disservice to mathematics if AI companies were allowed to establish standards for how the discipline should operate without broader input from academics.
Can mathematics survive AI? Fields Medalist Terrence Tao calls for a world of “Mathematics 2.0”
Source: www.nature.com