Via Bluesky, this letter from a French mathematician to a student:
Dear [...],
I won't lie to you the math community myself included - is extremely concerned about the current situation. Some colleagues are more optimistic than others, but we are all wondering where we are headed (and the problem obviously extends far beyond mathematics).
In any case, I fully agree with what you wrote (and your email shows that you grasp the situation very well). I, too, am convinced that AI companies do not care about science; if they did, they would have provided more details for instance, regarding the full range of problems they attempted to solve, the methods employed, and so on. They simply want to train and sell their models; they have chosen mathematics as a battleground and a training ground, and this is causing immense harm to science. I also reject the vision of mathematics turning into a "sort of 'math vibing', where you have to decipher what a computer has compiled, churned through, and spewed into a file" as you so eloquently put it.
My view is that the goal of mathematics is to understand the world, not merely to solve problems. Open problems have been put forth in the hope that they should serve as milestones for understanding and for discovering new theories acting as a catalyst for research and inquiry. Al companies are now using them in a largely destructive way, collecting them like trophies while delivering solutions that yield no genuine understanding. Many of us are angry.
But I think it was also partly our fault, as mathematicians, for placing too much importance on "hard problems" and the glory associated with solving them (like the Fields medal), and for sometimes believing that mathematics was a race for results rather than a collaborative effort to understand the world.
I don't know how it will all turn out, so I certainly can't promise you anything. But I do know that "understanding" is a human activity something computers cannot do for us (even if they solve problems better than we do, it serves no purpose unless there are humans who actually understand). Perhaps mathematical research will shift its focus toward the pursuit of interesting questions, illuminating perspectives, or elegant exposition (including aimed at the general public) rather than the race to answer difficult problems; that is what I hope for.
(""I wish it need not have happened in my time,' said Frodo. 'So do I,' said Gandalf, and so do all who live to see such times. But that is not for them to decide. All we have to decide is what to do with the time that is given us."" -J.R.R. Tolkien, "The Fellowship of the Ring', book I, chap. 2.)
I sincerely believe we need more mathematicians (human ones!) to understand an increasingly complex and confusing world; and also to learn either how to use AI intelligently without making ourselves dump (sic - CJ), or how to forgo using it but in any case, how to manage it productively.
So, I cannot make any promises to you on what will happen tomorrow. Clearly, you have a very good grasp of the current challenges (better than some of my fellow researchers). I would fully understand if you decided to leave mathematics for another field [...] after all, the very purpose of this program is to enable you to decide what you prefer and what you find most interesting and useful. But what I can say with certainty is that if the advent of AI discourages promising students from pursuing mathematics, that will be the greatest harm it has done to the discipline.
[...]
Best regards,
David A. Madore
I appreciate this letter. Mathematics is well beyond my understanding, but there certainly does seem to be a lot of concern around the AI companies' moves in that space. It's nice to see one mathematician's perspective.
Even as I have concerns about the trajectory around chemistry (the Trump Administration's attacks on the funding agencies, which will ultimately impact academia combined with the growth of chemical research in China), I genuinely don't see a major impact from AI on the long-term job prospects of chemists. There is always a need for new chemicals and new materials, and (like the mathematicians!) there needs to be chemists who understand chemistry to interpret and reduce the new knowledge to practice. I would already adjust the long-term trend for chemist hiring down since (say) 2025, but I can't imagine adding more than 0.3% for the 10 year trend based on current AI. There's my bold prediction, but just like Dr. Madore, I shouldn't make any promises.
Best wishes to students making these decisions and to the rest of us.
