Published 1 hour ago • loading... • Updated 19 hours ago
Canada’s star mathematician races to help establish safe superintelligence
The institute will train mathematicians to study superintelligence risks and publish new ideas quickly, with its first research semester set for January 2027.
Fields Medalist Jacob Tsimerman is shifting his focus to become scientific director at the Mathematical AI Safety Institute , an organization he co-founded to apply mathematical rigor to superintelligence.
Over the past decade, Tsimerman grew concerned about superintelligence as a mathematician and human being. In 2025, he co-authored 'A Taxonomy of Omnicidal Futures Involving Artificial Intelligence,' outlining extinction scenarios from AI misuse or loss of control.
Based in the San Francisco Bay, MAISI will borrow approaches typical of a University math department, facilitating working groups and research seminars to extract understanding from neural networks.
Through the institute, Tsimerman says they will sidestep traditional academic peer reviews to share insights quickly, aiming to help humanity avert extinction-level events while benefiting from the technology.
The institute treats the work as a race against time and AI's increasing capabilities. Tsimerman says MAISI would be "thrilled to be scooped" if someone else solves safe superintelligence first.