Math Is Still Worth Studying in the Age of AI

AI has turned mathematics upside down.

On September 8, OpenAI announced a resolution of the Navier–Stokes existence and smoothness problem — one of the most famous outstanding problems in mathematics. Navier–Stokes was difficult enough to be designated one of the seven Millennium Prize Problems by the Clay Mathematics Institute in 2000, with each problem bearing a $1 million reward for a satisfactory answer.

OpenAI’s solution threw the mathematics community into a frenzy. Three days later, a group of 25 Fields Medalists — recipients of the most prestigious award in mathematics — penned an open letter decrying AI companies for so callously using unsolved problems to test their latest models. Regardless of one’s personal convictions on AI, it was clear to all that mathematics had crossed the Rubicon. There was no going back.

A month later, just as the shock had begun to subside, OpenAI released some 722 mathematics papers their models had generated, covering just about every subfield of mathematics. A headline in The Atlantic read “OpenAI Just Carpet-Bombed Mathematics.”

That claim isn’t entirely exaggerated. Hugo Duminil-Copin — also a Fields Medalist — lamented that every single problem he was working on had been solved in one fell swoop.

For Mathematics concentrators like myself, the past few weeks have certainly been tumultuous. On one hand, it is exciting to follow these developments as they occur and to feel so close to the action. Posterity may very well place September 8, 2026 among the likes of other epoch-defining dates like July 16, 1945 (the Trinity nuclear test) or October 12, 1492 (Columbus landing in the New World).

I also feel deep uncertainty. AI is already capable of major advances in mathematics. More unsettling is the speed with which AI models improve. Nobody knows what the future of mathematics — certainly in its current academic practice — will look like. How could humans ever catch up at this rate?

As I will argue, though, mathematics is still a fundamentally human endeavor. For this reason, even with newfound unknowns, math is still worth studying.

Faced with AI’s growing mathematical aptitude, mathematicians have at times sought refuge in the aesthetic qualities of their discipline, invoking a nebulous sense of beauty. These qualities will be of little use, for they can justify mathematics only through one’s subjective appraisal. Much of mathematics is decidedly not beautiful, to me. Instead, I will advance a more principled argument in favor of mathematics, even in a hypothetical future when AI is far better than the best human mathematicians.

Mathematics is the language through which we understand the universe. It is only a slight overstatement to say science is the art of reducing the world to mathematics. After all, mathematics is logic crystallized. Once you have the postulates of your scientific theory, what happens next is forced by mathematics. For the field I know best, theoretical physics, the mathematics needed to describe the universe at its most fundamental level may very well not exist yet.

As long as people need to understand the world, they need to learn mathematics — and I wouldn’t want a future where humanity abdicates its understanding of what it spent centuries learning.

Furthermore, mathematics provides excellent training in how to think. What I appreciate most about proof-based math is how there is nowhere to hide. One must rigorously justify each step in a solution. That includes scrutinizing definitions, considering edge cases, and ensuring you have not made a subtle misstep in your own understanding. Thinking this way is an invaluable asset in all dimensions of one’s life.

Ironically, I think mathematics — the ‘M’ in STEM — has much more in common with the humanities than many of us appreciate. The above line of reasoning, that mathematics teaches one how to think, is also used to justify the study of the humanities.

We study social theory and French literature to cultivate a thoughtful society and attain understanding of the many actors, ideas, and approaches within that intellectual tradition. In the same way, as the Clay Mathematics Institute observes, “A proof gives not only certitude, but also understanding.”

Math, then, is more than solving problems. It is also about developing new theories and unraveling how those theories connect to the rest of the field. The quest for understanding often spawns many new problems along the way. Unlike every other science, mathematics isn’t beholden to describing what one sees in the lab — math has ample room for human creativity and aesthetics, no matter how powerful AI becomes.

Math isn’t going away. But perhaps it is due for a change of scenery.

Evan B. Tsingos ’27, a Crimson Editorial Editor, is a joint Physics and Mathematics concentrator in Adams House.