In my career, first as a mathematician and then as a math writer, I’ve read a lot of books that were written by mathematicians “for the general public.” And yet I have found very few that describe the experience of working as a mathematician, in a way that rings true to me and is accessible to ordinary readers.
Where can you read about the experience of mathematics? Certainly not in textbooks. They teach you topics. Algebra. Differential geometry. Category theory. Generally speaking, they are 100 percent divorced from the human experience; they are only about finding answers.
Another common genre of math book is the biography of a famous mathematician. But it’s almost impossible to reproduce the experience of creating new mathematics in this way, because the famous mathematician’s most famous results are already known. You can’t vicariously experience the stumbles, the frustrations, and the sudden insights, when you already know how the story ends.
Then there’s another common type of math book, which I would call the smorgasbord. It presents a sampling of different topics, with a new topic every chapter, and perhaps some historical anecdotes around each one. I’ve read a lot of pop math books in this genre, and there’s nothing wrong with them, but they don’t tell a unified, coherent story. My vision of a book about the mathematical experience would describe the nearly obsessive focus that you need to live with a problem for months or perhaps years at a time, knowing all the while that it may not even have a solution. It would talk about all the stages in solving such a problem, some of them serendipitous and others deeply planned.
And even if you find a book that is literally a biography of a problem (for example, a book about Fermat’s Last Theorem), you’ll run into one more challenge, which is that the level of abstraction of frontier mathematics makes it nearly impossible for our proverbial “general public” to grasp.
So here are my desiderata: A book that describes the whole experience of doing math, including the mistakes and the dead ends. A book that engrosses the reader like a mystery. A book that takes the reader to places they’ve never been before, like a travelogue. A book with surprising twists and turns, like a Shakespearean play. A book that is unified, that tells the story of one problem instead of flitting about from topic to topic. The problem should be deep enough and subtle enough that it requires input from different areas of mathematics — because I do want to convey the mysterious unity of mathematics, the way that any sufficiently hard problem tends to overflow the boundaries that we try to construct around it. Only a genuinely hard problem will do. And yet — and yet! — I want this problem to be so conceptually simple and so universal that a person with no particular mathematical training can understand what it’s about.
A problem like that comes along only once in a lifetime. Or maybe once in 2000 years.
Master Sun’s problem came to me. I didn’t go looking for it. A friend of mine sent it to me, and I got obsessed by it. I worked on it for a full year, and sorta kinda solved it. Then a few months later, an unexpected collaborator helped me solve it completely. We wrote it up and published it in a journal, and that was that.
But there was so much more to tell that I couldn’t put into the journal article. The whole human story, reaching back to ancient China and leading up to today. The story of my friend, who rediscovered the problem, looked around for 30 years for someone who would take it seriously, and finally got to see his “bucket list” problem solved before his untimely death at age 59.
Now, ten years later, Master Sun’s Problem has finally been published, not in abbreviated journal-article form but as the book that it always wanted to be. Not only that, in this watershed year of 2026, a whole new set of reasons for the book has appeared. For the first time, artificial intelligence has surpassed human intelligence at solving math problems. OpenAI released a whole catalogue of 372 previously open math problems that its experimental AI had either solved or made major progress towards solving. Mathematicians are openly questioning what is the purpose of having humans do mathematics, if computers can do it better.
I’ve found it particularly interesting that top-level mathematicians are saying something they previously did not say openly — namely, that the true value of mathematics to them lies not in the solution of a problem, but in the experience of solving it. Here are some sample quotes I’ve collected (the last from a personal interview, the rest from the Internet):
“As a mathematician, I care not just about the end results, but also the journey.” – Zhiwei Yun (professor, MIT)
“If I’m doing a piece of math, I probably think of it as more akin to storytelling.” – Akshay Venkatesh (professor and Fields Medalist, IAS)
“We’re all frustration addicts. We just want to bang our heads against problems we don’t yet know how to solve.” – Garvy (a recent PhD who posts on Substack)
“I love the process of problem solving, banging our heads against the wall and getting frustrated and all of a sudden the satisfaction of seeing the idea that works.” – Kevin Ford (professor, Illinois)
What do mathematicians themselves value about mathematics? The journey. The story. The struggle. And the satisfaction. All of these are just as important as finding the answer. You cannot experience any of these things if you feed the problem to an AI and simply copy its answer. These are exactly the mathematical experiences that Master Sun’s Problem is all about.
At first, when the news about OpenAI and its wholesale harvesting of math problems came out, I was depressed. It seemed to me that my book, just days away from publication, had turned into an instant museum piece, a relic of the way we used to solve math problems before AI came along. But now I’m not depressed any more. The trauma of AI has forced us to re-evaluate our priorities, or to state them for the first time, and they were exactly the same as my priorities in writing Master Sun’s Problem. So I no longer consider my book to be a relic of bygone days. Instead, I consider it the first book of the new AI era in mathematics. In this era, the journey, the story, the struggle, and the satisfaction are what we have to preserve at all costs.
