OpenAI has recently published hundreds of mathematical proofs solving problems that have stumped the world’s best mathematicians for decades or centuries, according to Derek Thompson’s analysis. The discoveries span number theory, theoretical computer science, and logic, and have provoked an unexpectedly emotional response from the mathematical community.
Hugo Duminil-Copin, recipient of the 2022 Fields Medal, told Thompson: “Tuesday night, I cried and it woke up my daughter. They have annihilated my field of research.” Steven Strogatz, a mathematician and author at Cornell University, described similar distress: “I literally cried when I spoke to Wired a few weeks ago. People are using language like ‘they’re strip-mining our subject’ or ‘they’re nuking the landscape.’”
Thompson argues the mathematicians’ grief extends beyond bruised egos. He suggests they are mourning the disruption of mathematics as a social endeavor—a thousand-year conversation where each generation builds on the work of previous discoverers. Thompson quotes Archimedes, who wrote two thousand years ago: “There will be no lack of discoverers to add to the number of theorems which have fallen to our share.” To Strogatz, this reflects mathematics’ essence: standing on the shoulders of giants in a shared quest for beauty and transcendence.
According to Thompson, the concern is not merely about losing jobs or status, but about AI solving the “craft” dimension of mathematics—discovering answers—while severing the “conversation” dimension where scientists engage with one another’s work. When machines publish proofs without human collaboration, Thompson argues, they interrupt this relational process that has defined scientific culture.
However, Thompson also explores counterarguments. Earlier this year, Shanmu Jin, a Chinese medical resident, used ChatGPT to solve Crouzeix’s conjecture, a problem professional mathematicians had failed to crack for decades. This suggests AI could democratize access to mathematical problem-solving. Yet Thompson acknowledges a paradox: such a world might contain more knowledge but less understanding, as AI proofs often take “alien shapes” with reasoning styles that differ fundamentally from human approaches.
Key facts
- OpenAI has published hundreds of mathematical proofs solving decades- or centuries-old problems
- Fields Medal recipient Hugo Duminil-Copin said the breakthroughs ‘annihilated’ his field of research
- Cornell mathematician Steven Strogatz reported crying over the development
- A Chinese medical resident used ChatGPT to solve Crouzeix’s conjecture, which professional mathematicians had failed to solve for decades
- Mathematicians describe grief rooted in the disruption of mathematics as a collaborative, generational conversation—not merely loss of prestige