Our mission
Make frontier AI systems a true object of scientific inquiry: open to the mathematical community to examine, reproduce, improve, and ultimately understand.
The rapid growth of AI capabilities marks a pivotal moment for mathematics. Our mission is to help the field thrive in this new era and place mathematicians and scientists in a position to shape its future. At the MAIC (pronounced /maɪk/) Center, we will bring mathematics and AI together to advance scientific discovery, deepen our mathematical understanding of modern AI systems at their natural scale, build research infrastructure for the field, and redefine mathematical training for the next generation. We will hold that work to the standards of open science: open to all, transparent in its methods, and accountable to the public.
To realize this mission, our core activity will be to develop an open frontier model for the mathematical sciences. We are founding members of an academic consortium that will pursue this goal in stages: first building the capability to generate mathematical data and environments at scale, then using that capability to post-train open models in partnership with open-weight model providers; and ultimately developing our own open model and training stack.
Responsible development is central to our goal. We approach AI safety and alignment as first-order scientific challenges that require the engagement and scrutiny of the broader research community in conditions of maximum transparency.
Our research areas
Six interconnected areas surround and inform our core project, linking the development of a frontier model for the mathematical sciences with (1) AI-powered research across pure mathematics and theoretical computer science, (2) pushing the frontier of applied mathematics, (3) advancing our mathematical understanding of AI, (4) building the foundations for AI safety under transparency, (5) software formalization, and (6) training and exposition.
An open frontier model for mathematics at the center of six complementary areas of activity.
Our activities
Seminar: Mathematics in the Age of AI
A cross-departmental seminar at Courant exploring the interactions between AI, mathematics, computing, and data science.
Leadership
Co-directors
| Joan Bruna | Computer Science and the Center for Data Science |
| Vlad Vicol | Mathematics |
Consortium lead
| Tristan Buckmaster | Mathematics |
Steering committee
| Nestor Guillen | Mathematics |
| Jaume de Dios Pont | Center for Data Science |
| Scott Armstrong | Mathematics |
Members
Coming soon