Post-Cutoff

Matilde Marcolli: ‘Duet for the end of math’

Matilde Marcolliits.caltech.edu

Why it matters

A Caltech mathematician argues that the panic over AI-solved conjectures exposes a value system built on ‘Famous Conjectures’ and should end it; quoted by Quanta (Oct 5), cited in the Stange–Harriss guest post on Tao’s blog and listed by Tao on Oct 12.

Summary

A four-page essay (PDF on Marcolli’s Caltech page). Its footnote reads “written 09/10/2026, to appear in Newsletter of the London Mathematical Society”; we read the date as 10 September 2026, because the PDF was created on 18 Sept and Gil Kalai linked it on 22 Sept. Marcolli asks why automated solutions of well-known conjectures send “shock waves of panic” through the profession and answers that the field gave “Famous Conjectures™ and Big Problems™” a disproportionate role, tying careers, prizes and power to a narrow set of goals reached by hyperspecialization. She contrasts this with theoretical physics, which in her view rewards original ideas that connect many questions, and calls mathematical Platonism “a subtle propaganda instrument” that makes mathematics look like a zero-sum conquest. She describes AI systems as fed with human work, “often with unauthorized access to ongoing unpublished work of human mathematicians”. For her, mathematics is “a language to make sense of reality and a form of self-expression, much like the arts”, so AI solutions are “essentially irrelevant” to the reason for doing it. She proposes a profession more like the conductor or composer than the solo virtuoso. If conjecture trophies run out, “the sooner the better! Good riddance!”. The essay was read in full (PDF text extracted with pdftotext on 2026-10-13).

Archived text

  • “When and how did we start viewing mathematics as a zero-sum game, with some predetermined scarce trophies out there that one alone will conquer while all the others will fail?”
  • “Ask yourself rather what is the mathematics that lives only inside you and nowhere else.”
  • “Well, the sooner the better! Good riddance!”
  • “then we may be out of business quite fast, not because of the end of mathematics but because of the end of our imagination.”
Source: its.caltech.edu/~matilde/AImathNew.pdfShort quotes only; full text at the URL.

Cited in

  1. Science & math 98 days after the cutoff

    OpenAI releases 722 AI-written math manuscripts claiming hundreds of open problems

  2. Research 97 days after the cutoff

    Quanta’s math editor asks ‘Is AI the End of Math As We Know It?’

People in this post

Terence Tao, Professor of mathematics, UCLA