My older brother graduated from college with a degree in chemical engineering at an age at which most people start. He also graduated with a six-figure job offer. Given this clear example of success, my parents naturally assumed I would follow in his footsteps. I enrolled at the same college, declaring my own major to be chemical engineering.
But I just couldn’t really get into chemical engineering, and I didn’t really have my brother’s tremendous work ethic. I was capable of focusing on things, for sure, but only if they really spoke to me first. Chemical engineering, for whatever reason, just didn’t. I found myself doing mathematical puzzles to calm myself down after forcing myself through Organic Chemistry homework. I guess, upon reading that, the conclusion is somewhat obvious, right? Much to the dismay of nearly everyone close to me, I eschewed a stable, well-paying future and graduated with degrees in mathematics and economics. It’s not exactly becoming a starving artist, but you’d never know that if you asked my parents. Oddly enough, the economics degree was motivated by my desire to take courses in game theory, and I found that by the time I had the necessary economics pre-requisites to take such a course, I didn’t have that much further to go before getting the entire degree. See? Capable of focusing after all.
My undergraduate university, Tulane, was still recovering from the effects of Hurricane Katrina. Many faculties, including mathematics, were still recovering. Some faculties, like computer science, were entirely missing. I am forever proud to have gone to Tulane and am deeply impressed with the level of normalcy they managed after such a historic crisis, but after I graduated, I felt my mathematical skills were somewhat incomplete.
To address this, I utilized study abroad programs. Namely, Math in Moscow, which I did for a year after graduating, and the Budapest Semesters in Mathematics, which I did during my senior year of undergrad. These two programs did two important things for me. Namely, they introduced me to the joy of research-oriented mathematical problem-solving and introduced me to theoretical computer science. If you know either of these programs, then you intimately understand how *fun* they make problem-solving, and it should be no secret how I was enthralled. I now had a vision. I wanted to work in theoretical computer science, and I wanted to be the kind of person who could solve ridiculously hard problems with absurdly beautiful proofs. The kind of person that even other mathematicians look at with awe.
To that end, I did a Master’s in Mathematics and the Foundations of Computer Science at Oxford. It largely reinforced my beliefs, though I remained remarkably unfocused on *how* exactly I wanted to accomplish my goals. As a result, I had some trouble with my first round of PhD applications (I applied to four schools!), but, eventually, I found my way to Rice University, doing a PhD in (theoretical) computer science outright. The transition from math to TCS was complete.
My PhD supervisor challenged my vision of “big problem hunting”. He told me that research was much more about finding your own direction to explore, of inventing your own style of mathematics. He warned me with stories of promising researchers who spent their entire lives unsuccessfully battling with a single problem. He eventually convinced me. My style of research is now much more about definitions and conceptual understanding than difficult problems and hard proofs. My work is about choosing the “right” problem to solve, and part of that criterion is that the problem, as a stand-alone challenge, has an answer that’s easy to understand. I even came to see hard problems as roadblocks that stifled my own conceptual understanding, obstacles to be removed rather than goals in themselves.
Why did I write all of this? Two reasons. First, I wanted to tell you about my human relationship with mathematics, emphasis on the word human. And second, I wanted to tell you that I’m the type of researcher who would, theoretically, be the most excited about a theorem-proving machine. This is not to suggest that I view myself as above people who really like problem-solving. I do not at all, and I really hope that you do not take that message away from this post. It’s just, that style of research is no longer my style of research. If you’ll allow me to characterize my own style of research in a somewhat smug and self-congratulatory way, I’ll say it’s about “the bigger picture.”
See, I wanted to use that phrasing exactly because that’s what I wanted to talk about with this post, the bigger picture. Many of the posts here are wondering what will happen to the field of mathematical research, and with good cause. I’m not going to go over the timeline of events. If you’re reading this, you already know what has happened. I’ll briefly add that I believe that there is a meaningful way to do mathematical research in the age of LLMs (and that the technology can be used as the incredible tool it so clearly is), but you can take that with however many grains of salt you like. I’m no senior researcher with a bird’s-eye view of the field, just a postdoc on sinking ground.
If we want to talk about the future of mathematical research, though, I believe we should start by asking why mathematics was so uniquely targeted by AI companies. The answer, of course, is that it wasn’t. This is where the bigger picture comes in. The big stories before mathematics were artists and software engineers. Artists, however, never needed to justify the human element of their craft, which was widely understood by the public, even if it wasn’t particularly valued. Software engineers put financial reality before everything else, something that AI companies are still struggling with. Enter mathematics. Here was a field that, at least from the outside, could be characterized as completely results-oriented, so economical concerns could be thrown out the window. It was also a field that had undeniable connections to practical advancements and lots of prestige associated with it, so the artist’s defense falls short as well. The field of research mathematics wasn’t specifically targeted. It was a target of opportunity, a field with many unique and severely compromising aspects that made it exceptionally vulnerable to a hostile attack.
Yes, a hostile attack. Many people are arguing that OpenAI is testing its models on mathematics problems since it’s the only way they have to test their models. Is that true? Surely, there were ways to do this in ways that worked with the mathematical community instead of explicitly against it. In fact, the former scenario even involves willing free labor from the community, which would’ve been readily handed over. So again, let’s talk about the bigger picture. The goal of these companies is not to augment human capabilities; it is to replace human beings at scale. They want a system that achieves “AGI”, which is defined as having superhuman abilities in *every* domain simultaneously. Mathematics was just uniquely vulnerable as a base for an early demonstration of such suggestive capabilities.
And it’s not the only one, by far. Countless other jobs have been silently bulldozed without the fanfare of mathematics. I have a friend who loves languages and worked in freelance translation. How do you think the job market looks for my recently laid-off friend?
The bigger picture again. If we want a future for human beings in mathematical research, then we *need a future for human effort*. A lot of people here are talking about how to adapt. I don’t think you’re going to be able to, as long as a trillion-dollar company is explicitly putting its thumb on the scale to make sure you cannot. The entire point of their work is to demonstrate that mathematics can advance without humanity. It’s part of an overarching goal to advance everything else without human input.
The public is not going to uniquely feel hugely sympathetic for just mathematicians. They largely do not know a single mathematician. They know their friend who was recently laid off, though. Their friend who worked in a field that received no headlines or any form of sympathy. They know their kid, still in college, who was told to “learn to code, because it’s the only thing that will be left” two years ago, and is now being told they are useless. They know the artist whose style was shamelessly ripped off and doesn’t get commission work anymore, denying them the already unglamorous salary they had before. They know the entry-level office worker who was “deservedly” let go because “they don’t even do anything.” We are all human beings. They are us, and we are them.
AI may be capable of solving many important problems. I hear cancer cures being discussed quite frequently. But even if AI achieves that, I still want human understanding of the cure and a human organization to take responsibility for it. I want someone who can understand human pain to help those suffering from it. AI companies do not want this. They want knowledge that is specifically beyond our comprehension. They do not want, to quote Terence Tao, “the best of both worlds.” Because – and I really wish more people understood this – they do not want a civilization in which humanity remains in the driver’s seat. Many of them directly see us as a disposable precursor species to a digital successor, a “biological bootloader”.
I am deeply invested in the future of mathematical research. But the whole point of this post is to zoom out a bit, to the bigger picture. To write in the style of our overlords, this is not just an attack on the premise that humans should do mathematics. It’s an attack on the idea that human effort has meaning. And I really want to see other mathematicians also recognize and discuss this, in addition to discussing our situation with mathematical research. This is part of a larger effort of our far-right government in the US to invalidate enormous groups of people entirely. We are human, first and foremost, and we need to deal with the attack on our humanity before we can effectively take back control of our field. So please, continue discussing mathematics here. Please continue helping each other through this. But please, add to that as well, if you feel that you can. We are not the only ones hurting, and so much more is at stake. I will be right here with you.
Your fellow human being,
Senthil
Received 11 October 2026.
Add to the discussion