Jesse Thaler
Massachusetts Institute of Technology, MA·Physics
Physics now produces enormous, complex datasets, from particle collisions to gravitational-wave signals, and many of its calculations are too hard for ordinary computing. Artificial intelligence (AI) can help, but standard AI tools can be hard to trust and verify.
- Fields
Physics
Computer Science
Astronomy & Astrophysics
- Topics
artificial intelligence
machine learning
deep learning
- Grants
- 2 known grants, through May 2031
What they work on
- Develop "ab initio AI" methods that build first principles of fundamental physics into machine learning, to make intractable theoretical physics calculations, such as precision structure-of-matter predictions, feasible.
- Design AI tools for experiments that can be verified against well-understood calibration data, improving uncertainty quantification at facilities such as the Large Hadron Collider, LIGO and neutrino experiments.
- Advance foundational AI by infusing physics principles into representation learning, robust AI and reinforcement learning, and use physics to better understand how AI models work.
- Apply AI techniques such as generative models and anomaly detection to data-rich astrophysics, including dark-matter searches, structure formation and multi-messenger astrophysics.
More about the research
The Institute for Artificial Intelligence and Fundamental Interactions (IAIFI) is built on the idea that AI can be designed to include physics principles directly. It aims to advance physics discovery and to make AI itself more reliable. The institute is centered at MIT and includes Harvard, Northeastern and Tufts.
Why it matters
Trustworthy, verifiable AI matters well beyond physics. Methods built on physics principles could give AI used across society better uncertainty estimates and easier verification. Within science, these tools could speed discoveries at major facilities such as the Large Hadron Collider and LIGO and help astronomers handle fast-growing datasets, for example in dark-matter searches.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Ayres Freitas
University of Pittsburgh, PA·Physics
Develop calculational techniques for higher-order electroweak corrections, the quantum refinements needed to interpret precision measurements at the Large Hadron Collider and at planned electron-positron colliders.
Kyoungchul Kong
University of Kansas, KS·Physics
Develop quantum algorithms for particle physics data analysis, collision event reconstruction and anomaly detection.
Hitoshi Murayama
University of California-Berkeley, CA·Physics
Study non-perturbative dynamics of strongly coupled gauge theories, forces too strong for standard approximations, using exact supersymmetric solutions combined with anomaly-mediated supersymmetry breaking.
The school
Massachusetts Institute of Technology: 25 labs on PI Pages, $1.1B a year on research, 692 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.