The math superstar who fears AI, yet recently joined OpenAI
Fields Medal winner Jacob Zimmerman is joining OpenAI. He aims to use his mathematical expertise to advance AI safety and ensure the technology does not pose an existential threat to humanity.
When he won the Fields Medal last week, Jacob Zimmerman went up to receive the most prestigious award in the world of mathematics wearing a light blue tuxedo with satin lapels. His announcement after the ceremony was no less impressive.
When asked about his future plans, Zimmerman replied: "I'm going to start working at OpenAI."
Jacob Zimmerman
-
Personal: 38 years old, born in Kazan, USSR. Immigrated to Israel at age 3, lived there for 6 years. Resides in Canada.
-
Professional: Completed a bachelor's degree in mathematics at the University of Toronto at 18. Received a doctorate from Princeton University in 2011 and completed post-doctoral work at Harvard.
-
Achievements: First Canadian to win the Fields Medal (2026).
Imposing order on apocalypse scenarios
This decision was surprising yet logical. Zimmerman recently published a paper mapping out ways in which artificial intelligence could lead to human extinction. He stopped advising graduate students not working on AI and shifted his research focus from number theory and complex algebraic geometry to AI safety.
The professor is taking a leave of absence from the University of Toronto to join OpenAI. He seeks to harness formal language and verification methods to evaluate AI progress, understand decision-making processes, and ensure the technology does not lead to humanity's extinction.
Connection to Israel
Jacob Zimmerman was born in 1988 in Kazan and immigrated to Israel with his family in 1991. His mother was a math teacher and his physicist grandfather encouraged his early interest in science. After moving to Canada, Zimmerman won two gold medals at the International Mathematical Olympiad. His doctoral supervisor at Princeton was Professor Peter Sarnak, a mathematician with Israeli and American citizenship.
In 2024, Zimmerman signed a petition calling for the suspension of cooperation with Israeli academic institutions that do not explicitly condemn Israel's actions in the Gaza Strip, a move that drew significant public attention.
Decoding machine behavior
"Within a few years, AI systems will be significantly better than humans at solving mathematical problems," Zimmerman says. "The social implications of this are much more difficult questions." He is currently working on how to mathematically prove that AI systems are not secretly plotting against humanity.
While Zimmerman prefers the classic approach of working with a chalkboard, he acknowledges AI's efficiency in routine tasks. Recently, an AI model helped him identify an error in one of his arguments. "If one of my research colleagues or students had done this, I would have been very impressed," he notes. "This is a capability it didn't have last year."
"As AI systems become more capable, ensuring their safety and alignment with human values will require exceptional researchers and new ways of thinking," says Kai Chen, an alignment researcher at OpenAI. For Zimmerman, this is a natural career progression, where mathematics remains his primary passion.
