The Institute for Artificial Intelligence and Fundamental Interactions (IAIFI), led by MIT, has secured renewed funding from the National Science Foundation (NSF) for another five years. The annual grant has increased from $4 million to $4.98 million. This renewal signifies a new chapter for IAIFI, which has spent its initial five years developing a research model and fostering an interdisciplinary community focused on the synergy between AI and physics. Established in 2020 as part of the National Artificial Intelligence Research Institutes program, IAIFI includes researchers from MIT, Harvard, Northeastern, Tufts, and Boston University. Their work has demonstrated that machine learning can expedite discoveries in physics, while physics can enhance AI systems.<\/p>\n
Jesse Thaler, IAIFI’s director and MIT physics professor, stated, “From the beginning, IAIFI has been built around a two-way street: AI enabling better physics, and physics enabling better AI.” He noted that this beneficial cycle has been evident in various physics and AI fields over the last five years. The collaboration has produced not only new findings but also innovative scientific approaches.<\/p>\n
IAIFI’s research encompasses particle physics, nuclear physics, astrophysics, and foundational AI, with significant advancements arising from these collaborations. In particle physics, IAIFI has developed AI techniques to process the vast data from the Large Hadron Collider in real-time. In nuclear physics, generative AI methods are used to model quark and gluon interactions in lattice quantum chromodynamics, offering new ways to explore matter’s structure. In astrophysics, machine learning aids in detecting new cosmic phenomena and enhancing the MIT-led LIGO gravitational-wave experiment’s sensitivity.<\/p>\n
Simultaneously, physics principles are shaping new AI methodologies. IAIFI researchers are creating learning algorithms and model architectures that incorporate physics knowledge\u2014such as symmetries and geometric structures\u2014into neural networks, resulting in more dependable, interpretable, and data-efficient systems. Mike Williams, interim director of IAIFI and an MIT physics professor, commented, “AI has begun to transform how physicists tackle some of the field\u2019s most challenging problems.” He added that it is expanding the range of feasible questions.<\/p>\n
IAIFI invests significantly in training the next generation of scientists through its Postdoctoral Fellows program, which supports early-career researchers at the intersection of physics and AI. Fellows receive mentorship in both fields, promoting cross-institution collaboration. To date, eight fellows have completed the program, with some securing faculty roles and others joining leading AI firms or startups.<\/p>\n
Phiala Shanahan, interim deputy director of IAIFI and an MIT physics professor, emphasized the fellowship’s impact, saying, “Our fellows aren\u2019t just contributing to physics or to AI separately \u2014 they are helping shape a growing field at the intersection.” The IAIFI PhD Summer School has become a hub for the community of “centaur scientists” skilled in both disciplines, drawing nearly 600 applications for the 2026 event.<\/p>\n
IAIFI has also influenced new educational paths at MIT, including an interdisciplinary PhD program in physics, statistics, and data science. The program, a collaboration between the Department of Physics and the Statistics and Data Science Center, has awarded 20 doctoral degrees since 2021. Additionally, IAIFI members have developed a computational data science course in physics offered both on-campus and online.<\/p>\n
Beyond research and training, IAIFI hosts an annual summer workshop and engages the public through activities with the MIT Museum and the Museum of Science in Boston. Nergis Mavalvala, dean of the MIT School of Science, remarked, “IAIFI shows what becomes possible when researchers in physics, computation, statistics, and data science organize around shared scientific questions.” The institute is hosted in MIT’s Laboratory of Nuclear Science and is led by Jesse Thaler, Mike Williams, Phiala Shanahan, and Marisa LaFleur, along with a steering committee.<\/p>\n
As part of the National Artificial Intelligence Research Institutes program, IAIFI is involved in national efforts to advance AI-driven discovery. Marisa LaFleur, IAIFI’s managing director, noted the value of connections among NSF AI Institutes, saying, “We\u2019re sharing management strategies and resources for training, community building, and collaboration that make the whole network stronger.” With the renewed funding, IAIFI aims to delve deeper into the “physics of AI,” using physical reasoning and challenges to enhance AI understanding and development. Jesse Thaler highlighted that the institute’s first phase established a model of interdisciplinary research and community, and now, with renewed support, they aim to expand their aspirations.<\/p>\n