Zachary Porat

In Defense of the Human Mathematician

Posted on October 8, 2026

In the wake of OpenAI’s GitHub release of 722 math manuscripts, I have found myself thinking about what it means to do mathematics. The release is remarkable, no doubt: an AI model produced hundreds of mathematical results, organized into hundreds of manuscripts, across a wide range of fields. If AI can produce such results, what is left for us to do? What is the role of the human mathematician in a world where AI can both solve mathematical problems and write papers that explain the solutions? I write this blog in defense of the human mathematician.

I often have to remind my Calculus I students that just writing the answer to a question will not earn them full credit. Instead, to get full credit, I need to see their problem-solving strategy. I need to see how they interpreted the problem. I need to see how they set up the problem. I need to see the steps they took to arrive at a solution. At the end of the day, I am not testing that they can simply answer the question, but rather I am testing that they are equipped with the skills to logically reason through any problem they might encounter. The answer is only one small part of mathematics. At its core, mathematics is everything that happens on the way to the answer.

I would argue that research mathematicians have never been interested in the answer to problems either. We are interested in the struggle and the process that leads to a novel solution. For within that struggle, we come up with new ideas and new strategies. We combine established techniques with brand new ones to forge a path towards a solution. The process of being confused, getting stuck, getting unstuck, getting stuck again is essential to our understanding. These moments lead to inspiration, to a spark, that lets us find a proof–maybe not a good one–but a proof. Then we refine, we tinker, we improve. AI cannot replicate this struggle, this battle, this journey.

When we solve a problem, we don’t just obtain a proof. We become slightly different mathematicians. The struggle is not valuable merely because it is difficult. It is valuable because it changes how we think. We learn something about how we approach problems, about which ideas work and which don’t, and about what questions we want to ask next.

Here are a few other things that AI cannot do:

• AI cannot attend a conference and deliver a talk that distills hours and hours of human might into a concise story that tells of the struggle–and eventual triumph–that comes with searching for a solution to an interesting problem.
• AI cannot replicate the excitement of receiving an email from a mathematician halfway across the globe expressing interest in your work. AI cannot recreate the joy and satisfaction that comes with knowing that your work is written in such a way that it can be read, understood, and valued as part of the greater mathematical network–an ecosystem made up of humans.
• AI cannot share a cup of coffee with you as you reminisce about that conference ten years ago, where you first met. AI cannot describe its journey, how it navigated different areas of math and settled in its current area of interest. AI cannot discuss how its area and your area might coincide, allowing you to rekindle a relationship that started many years ago and leading to a new and fruitful collaboration.

AI might be able to imitate these actions–it surely can draft an email or condense information–but it cannot experience these feelings in the way a human does. The reason I started this blog was to showcase the interconnectivity of mathematics. I have always wanted to illustrate that mathematics is a human endeavor, built on human connection. In all my travels, the one constant is the joy and energy I feel seeing old friends and making new ones–I hope that has come across in my blogs. After every conference, I feel invigorated by my fellow mathematicians, who are asking cool questions and exploring cool ideas. To me, the beauty of mathematics has always been in the community that we forge–the kindness, generosity, and support of so many. Mathematics is not just about getting the answer. It is about the people we become while searching for it, and the people we meet along the way. So I urge you, go out and do some beautiful mathematics by connecting with the kind, thoughtful (human) mathematicians around you.