With a feel for physics, AI models simulate a wider range of real-world scenarios
Researchers at MIT and Tsinghua University have developed 'GeoPT', a new AI pretraining approach that helps models understand physical laws. By simulating mechanical interactions in 3D, the model improves the efficiency and accuracy of simulations for vehicles and robotics.
Why it matters
Improving AI's grasp of physics is critical for the safe and efficient design of autonomous systems and engineering projects.
by Alex Shipps, Massachusetts Institute of Technology
edited by Lisa Lock , reviewed by Robert Egan
This article has been reviewed according to Science X's editorial process and policies . Editors have highlighted the following attributes while ensuring the content's credibility:
Add as preferred source Overall design of GeoPT. Credit: arXiv (2026). DOI: 10.48550/arxiv.2602.20399 Artificial intelligence models are jacks of many trades, including writing, generating images, and creating 3D models. But they aren't as helpful when it comes to testing robots or designs for vehicles in diverse environments, since they don't understand physics as well as they do pixels or text.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in