Some thoughts on the recent AI progress in math, as a math grad student

I'm a grad student in mathematics, about to start a PhD next year. I thought it might be interesting for people to hear from a grad student/early-career mathematician, since most widespread articles/opinion posts have been written by well established late career ones. Of course, I don't speak for everyone, only myself.

As a long time avid practitioner of math, I think there is much more ​for us mathematicians to do - I have so many exciting projects lined up that would have been infeasible pre-AI assistance due to immense technical debt (think the equivalent of writing thousands of lines of boilerplate code by hand). It's also been much easier for me to enter a new field, mainly using it as a personalized literature guide. This is especially useful for the frontier research literature due to how otherwise sparse and hard to navigate it is. And my research projects come alive with countless simulations and hand-tailored programs. ​AI so far has not diminished my enjoyment of mathematics, it has​ greatly enhanced it. I think AI will revolutionize natural science, including mathematics, and the process is more or less unstoppable by now.

But there is a saying by a famous mathematician that​​​ goes "mathematics is the most humanity-like of the sciences, or the most science-like of the humanities, depending on how you look at it". That means a lot of mathematicians, myself included are facing some very conflicting feelings. On the one hand, the scientist in us wants to celebrate the monumental change in the reach of mathematics and science. On the other hand, mathematicians are effectively artists and poets that write for the sciences, who are seeing their craft they've dedicated much of their lives to change rapidly.

So while we should fully embrace this change as scientists, I hope that we can also have a heart for the artist in all of us - who just wants to engage fully in their art, just because they love it.