Audio Edition: Sometimes Being First Means Seeing the End Before Anyone Else

Episode Summary

Executive Summary: The episode profiles Jacob Zimmerman’s rise from child prodigy to Fields Medalist, emphasizing his shift from competition math to deep research through problem-solving, mentorship, and cross-disciplinary methods. It then pivots to his alarm over rapidly improving AI, which he believes may soon outperform mathematicians and reshape the profession, leading him toward AI safety work.

Main Topics: Zimmerman’s mathematical drive and identity (Priority: 5/5): Zimmerman is portrayed as intensely competitive, motivated less by abstract beauty than by being first to solve hard problems. From Olympiad success to research mathematics (Priority: 5/5): His early training in math competitions built persistence and proof skills, leading from International Math Olympiad medals to Princeton and research. Mentorship under Peter Sarnak and research maturation (Priority: 5/5): Sarnak repeatedly gave Zimmerman hard problems to help him learn where research difficulties lie, eventually turning him into a top researcher. Major mathematical contributions (Priority: 5/5): The transcript details his work on the André-Oort conjecture and later Griffiths’ conjecture, including the creation of O-minimal GAGA. Fields Medal and career culmination (Priority: 4/5): Zimmerman’s late-career focus on high-impact projects and sustained productivity helped earn him one of four Fields Medals in 2026. AI disruption and the future of mathematics (Priority: 5/5): Zimmerman argues AI could surpass mathematicians within two years, threatening the traditional motivation and structure of the field. Shift toward AI safety (Priority: 4/5): Concerned about the implications of AI for math and society, Zimmerman begins redirecting his work toward AI safety at OpenAI.

Key Arguments: Competition math can be valuable preparation for research because it trains persistence and tolerance for being stuck on hard problems. Research success came from learning to fail productively and to recognize the structure of hard problems, not just from solving isolated puzzles. Crossing boundaries between fields of math gave Zimmerman an advantage and enabled breakthroughs that others in a specialty might not see. The André-Oort and Griffiths conjectures were advanced by combining techniques from different areas, showing the power of borrowing tools across subfields. Zimmerman believes AI progress is accelerating so quickly that the profession of mathematics itself may be disrupted or diminished. Because of AI uncertainty, he has stopped taking graduate students and is reallocating attention toward AI safety and preparing mathematicians for change.

Data Points: Birth year: 1988 - Zimmerman was born in Kazan in southwest Russia. Age when family moved to Israel: 3 - His family moved to Israel when he was three. Age when family moved to Toronto: 9 - He moved to Toronto and began entering math competitions. Age at first Canadian Mathematical Olympiad: 12 - He competed in seventh grade and initially solved few problems. Hours of daily Olympiad practice: 5 hours/day - He practiced Olympiad problems after school to improve. International Mathematical Olympiad gold medal year: 2003 - At age 15, he won gold in Tokyo. IM0 problems solved in perfect-score year: 6/6 - In Athens the following year, he achieved a perfect score. Age at high-school departure: 16 - He quit high school to enroll at the University of Toronto. Undergraduate degree age: 18 - He completed his undergraduate degree in two years. Graduate school start: 2006 - He began graduate school at Princeton. Time spent reading Sarnak’s list: 8 months - Zimmerman spent the first eight months of grad school reading assigned books and papers. Years to call Sarnak by first name: 3 years - He says he was so formal that it took three years to address Sarnak by his first name. Duration to solve first analytic number theory problem: 3 years - It took him three years to solve the problem Sarnak gave him, with the proof written up together. Year of partial André-Oort progress: 2010 - Zimmerman’s dissertation work advanced the conjecture during his graduate period. Year Griffiths’ conjecture was proved: 2018 - Zimmerman, Baker, and Johann Brunbarb proved it and developed O-minimal GAGA. Year full André-Oort conjecture proved: 2021 - Zimmerman, Pila, and Anant Shankar posted the full proof. Year of major Hodge theory theorem: 2024 - Zimmerman, Baker, and collaborators proved a major theorem on certain Hodge-theory spaces. Number of papers posted in 2024–2025: 11 - He posted 11 preprints across different fields. Fields Medal winners: 4 - He was named one of four Fields Medal winners in 2026. Age at Fields Medal: 38 - The transcript says Zimmerman is 38 and within the last eligible cycle. AI timeline claim: within 2 years - Zimmerman predicts AI will be better than mathematicians at doing math within two years. Date of OpenAI leave announcement: July (this year) - He announced taking leave from Toronto to work on AI safety at OpenAI.

Pivotal Quotes: "“I made it an explicit goal for myself not to focus on winning the Fields. I was determined not to be a winner.”" — Jacob Zimmerman: Explaining his attempt to avoid obsessing over the Fields Medal earlier in his career. "“There are many people whose primary enjoyment of math comes through problem solving... If that disappears, that is not a trivial issue.”" — Jacob Zimmerman: His public warning that AI could undermine what many mathematicians love about the field. "“If you have an arithmetic phenomenon that you don't expect to happen, then there should be a geometric reason for it.”" — Jacob Zimmerman: Summarizing the intuition behind André-Oort and his approach to linking arithmetic with geometry.

Implications: The episode suggests mathematics may be entering an era where AI changes both what counts as progress and why people do the work. Zimmerman’s move toward AI safety signals that top mathematicians may increasingly redirect talent toward safeguarding advanced systems.