If you look up AI and math on any news site, you will read statements including “AI is destroying careers” or “AI is nuking entire areas of math” or even a prominent Field’s medalist saying “I cried so hard due to AI in math that I woke up my daughter”. The news loves a good headline, and these extreme statements make for good headlines. Underneath these articles is the comment section, with many comments hostile to mathematicians. One comment read “Imagine being a seven year old and your father waking you up to cry that other people did math problems.” This comment, hostile to mathematics (and arguably bad faith), I think sums up a very common reaction of non-mathematicians to some of these articles. I wanted to argue with this commenter, but I realized that this reaction was, on some level, perfectly reasonable if you just read the article and were not a mathematician.
It is no secret that graduate students can be stressed out and neurotic. I have been talking to younger mathematicians, and some are taking the rhetoric of senior mathematicians very seriously. I feel badly for them — some are incredibly depressed. I’m not going to tell people how to feel, because this is pointless. However, I do try to be more balanced than senior mathematicians have been — both because I think it is correct, but also because I have been disgusted by poor leadership.
The purpose of this essay is to implore senior mathematicians to tone down the extreme rhetoric, especially when communicating with the media and with junior mathematicians. I am encouraging this both because I believe these extremes aren’t true, but more importantly because it is bad leadership. It is like going into battle, even if you’re all going to die, your captain shouldn’t cry to you about it.
One major effect AI will have is forcing us to move on from well-trodden ground. The Riemann hypothesis has been around since…well…Riemann. The Navier-Stokes equations were introduced in the early to mid 1800s. Since then, these problems have accumulated an immense amount of prestige. Maybe it is time to move on? There are new questions to ask, there are new theories to develop, and new math to do. It seems like a lot of the neurosis comes from senior mathematicians who have attached an inordinate amount of self-worth to solving specific, well-formulated, and difficult conjectures. This is just one style of math, and it should not be seen as the entirety of math. AI will force a shift away from this style, which will ultimately lower the prestige attached to certain areas. However, in my personal opinion, this is a very good thing. I welcome this shift.
A second major effect AI will have is an increase in the number of papers, particularly incremental or “slop” papers. This is unequivocally bad. This may have a silver lining though — hopefully paper counts will be devalued, and paper quality will be valued more. This is a good thing.
A third major shift, related to the second, is that we can no longer trust someone’s papers to accurately represent their understanding or ability. This, again, is bad. However, similarly, I hope that this makes other things emphasized more in assessment — chalk talks, qualifying exams and coursework in grad school, informal conversations, and so on. The grad school standard of publication being the most important has, in my opinion, gotten a bit out of control. One thing a PhD should qualify you to do, is to teach mathematics. At the bare minimum, a person with a PhD in mathematics should be able to teach any undergraduate mathematics class or core graduate mathematics class (analysis, algebra, and geometry) with little difficulty. In reality, I don’t know how common this is. I hope AI devalues “novel contribution” and increases the value on having a solid foundation of core areas of mathematics.
Beyond any of these specifics, I have a very strong prior in the belief “nothing ever happens”. Mathematics has been around for thousands of years; it has undergone many changes. This will be yet another change, and mathematics will come out the other side alive.
If you still aren’t convinced, then could you at least tone it down a touch when talking to the media and younger mathematicians? If we really are marching towards the mathematical equivalent of nuclear armageddon, it is the responsibility of senior mathematicians to maintain some level of calm.
Received 24 October 2026.
Add to the discussion