We're gonna need a lot more mathematicians

Mathematicians may soon face a situation where they must struggle to keep up with the rapid pace of AI-generated mathematical ideas. In the past, students who found the pace too fast gave up on research mathematics and found alternative careers. However, the current research mathematics community consists of those who have overcome this challenge through hard work.

As AI systems continue to produce stunning new ideas, humans may soon find themselves unable to keep up. The temptation to abandon research mathematics may arise, but this would be a grave mistake. The responsibility lies in building a future where humans can understand and contribute to the discoveries that change the world. Struggle with difficult concepts is essential to understanding, and it can be shared among multiple research groups.

Society has an immediate stake in making this possible, as AI systems may propose groundbreaking designs for nuclear fusion power plants. Before constructing such machines, it is crucial for humans to understand the underlying principles, safety measures, and uncertainties. This demands the presence of mathematically sophisticated individuals who can engage with these questions.

The idea of a "deployable intellectual reserve" suggests a community of highly skilled mathematicians ready to help understand consequential AI breakthroughs. This skillset is valuable for tackling problems far beyond traditional research interests. Human agency is a fundamental value, and we must retain the ability to meaningfully consider alternatives and decide the kind of world we want to build.

Investing in our capacity to understand is essential to avoid arriving at a future where decisions of enormous consequence rest on reasons that no human community comprehends.