Mathematics has changed. Let's act like it

OpenAI's release of over 700 research papers on October 6 marked a significant moment in the history of mathematics. The AI's capabilities are now evident in its ability to make groundbreaking contributions across various mathematical areas. However, some mathematicians are unhappy about this development, citing the apparent disregard for the practice of mathematics by AI companies and issues with the quality and attribution of the OpenAI papers.

There is also a perception that the subject is being "hollowed out" and that mathematicians are under attack.

While the papers did not directly address major open problems like functoriality and reciprocity in the Langlands program, mathematicians should explain that tackling central problems is merely the first step to exploring new areas. The reaction to the OpenAI dump seems to reflect a feeling that the mathematical subject is being undermined.

Historically, the classification of finite simple groups, announced by Dan Gorenstein in 1983, was a major success and failure. Like the current OpenAI dump, it was a premature announcement that cooled interest in the field. However, there are still rich questions in finite group theory and the classification itself.

The promise of AI and its impact on mathematics is both exciting and daunting. While humans have made remarkable progress in the last thirty years, there is a growing unease about mathematicians' inability to explain their work to those outside their field. AI offers an opportunity to broaden the scope of mathematics and explore deeper mysteries. As mathematicians lament the missed paths, they should remember that those paths still exist and seek them out, rewarding the behavior of exploring new territories.