How AI is tackling a thousand-year-old physics mystery
In short: Scientists have struggled for millennia to solve complex problems in statistical physics. A new AI framework called THOR is now breaking through these barriers, offering a way to predict material behavior with unprecedented speed.
The core challenge lies in calculating configurational integrals—mathematical expressions that capture how particles interact. These are notoriously difficult to evaluate because of the "curse of dimensionality," where increasing variables causes complexity to grow exponentially. Traditionally, researchers relied on slow simulations like molecular dynamics or Monte Carlo methods, which can take weeks of supercomputer time.
THOR (Tensors for High-dimensional Object Representation) was created by a team involving Los Alamos National Laboratory and the University of New Mexico. By combining tensor network algorithms with machine learning potentials, the system solves these equations directly. It can compute thermodynamic properties hundreds of times faster than previous methods while maintaining accuracy, potentially accelerating breakthroughs in metallurgy and materials science.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: Solving The Hardest Problem In Physics
- Also: THOR AI solves a 100-year-old physics problem in seconds
To cite this card, link to https://intellectplus.co.uk/app/space-solving-the-hardest-problem-in-physics.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Physics · More in Space · Latest · A–Z · All topics
