The National Science Foundation has committed to a second five-year term for the MITaa-led Institute for Artificial Intelligence and Fundamental Interactions (IAIFI), increasing annual funding from $4 million to $4.98 million. The move signals a deep institutional bet on a research model where artificial intelligence and fundamental physics are treated not as separate disciplines, but as complementary engines of discovery.
A Renewed Commitment to the Two-Way Street
Launched in 2020 as part of the National Artificial Intelligence Research Institutes program, IAIFI brings together researchers from MIT, Harvard, Northeastern, Tufts, and Boston University. The central premise of the institute—that AI can open new ways of doing physics while physics can fundamentally improve AI systems—has defined its first five years of operation. IAIFI director Jesse Thaler characterized the renewal as validation of this model. “From the beginning, IAIFI has been built around a two-way street: AI enabling better physics, and physics enabling better AI,” Thaler said. “The exchange is producing not just new results, but genuinely new ways of doing science.”
Tangible Outcomes Across the Sciences
IAIFI’s research spans particle physics, nuclear physics, astrophysics, and foundational AI, with concrete results already emerging. In particle physics, researchers have developed AI techniques capable of handling the immense data rates from the Large Hadron Collider in real-time. In nuclear physics, generative methods are being used to model the interactions of quarks and gluons in lattice quantum chromodynamics, offering new approaches to studying the structure of matter from first principles. In astrophysics, machine learning is enhancing the sensitivity of the MIT-led LIGO gravitational-wave experiment and helping to uncover new cosmic phenomena.
On the AI side, ideas from physics are being used to develop learning algorithms and model architectures that embed physical symmetries, geometric structures, and exactness guarantees directly into neural networks. These physics-informed systems are proving to be more reliable, interpretable, and data-efficient than conventional approaches. Interim director Mike Williams noted that AI is expanding the frontier of problems physicists can realistically address, making possible questions that were previously beyond reach.
Training the “Centaur Scientist” Workforce
A defining feature of IAIFI is its investment in early-career researchers. The IAIFI Postdoctoral Fellows program pairs scientists with mentors in both physics and AI, fostering collaboration across institutional boundaries. Eight fellows have completed the program to date, with three securing faculty positions and others taking research roles at leading AI companies or startups.
The institute’s annual PhD Summer School has become a key pipeline for what the community calls “centaur scientists”—researchers with deep expertise in both physics and AI. For the 2026 edition, the program received nearly 600 applications for roughly 100 in-person spots, with approximately 300 additional participants expected to attend virtually. At MIT, IAIFI has also helped shape new educational pathways, including an interdisciplinary PhD program in physics, statistics, and data science that has awarded 20 doctoral degrees since 2021, as well as free online courses through MITx.
What This Means for the Broader AI Research Ecosystem
As a member of the National AI Research Institutes program, IAIFI is part of a nationwide effort to advance AI-driven discovery. The renewed funding provides an opportunity for the institute to push deeper into what it terms the “physics of AI”—using physical reasoning, physical challenges, and physical tools not just to apply AI, but to understand and improve it at a fundamental level. “The first phase of IAIFI established the model: interdisciplinary research, early-career talent, and a dynamic community,” Thaler said. “Now we have the foundation to push that model into new territory.”
For researchers and engineers in the AI and software space, the IAIFI renewal is a strong signal that funding agencies are backing cross-disciplinary approaches over siloed specialization. The rising profile of the centaur scientist—a researcher fluent in both machine learning and fundamental physics—points toward a future where the most significant AI breakthroughs may come from scientific principles, and where physics itself is increasingly conducted through the lens of artificial intelligence.
