With a feel for physics, AI models simulate a wider range of real-world scenarios

Researchers at MIT introduced GeoPT, a novel AI framework that embeds fundamental physics principles directly into model architecture, enabling more accurate and efficient simulations of real-world interactions like fluid dynamics, wind resistance, and object collisions. Unlike traditional AI models that learn patterns from data alone, GeoPT uses physics-informed neural networks to generalize across unseen scenarios, reducing the need for massive datasets and improving reliability in engineering, robotics, and climate modeling applications.