AI just discovered new physics in the fourth state of matter

2026-04-23 · Source: Robotics Research News -- ScienceDaily · Field: Science & Research — Physical Sciences & Chemistry, Artificial Intelligence & Machine Learning, Research Methodology & Innovation · Depth: Expert, medium

Summary

Emory University physicists have developed a novel AI-driven neural network to uncover new physical laws governing particle interactions in dusty plasma, a "fourth state of matter." This AI model, detailed in *PNAS*, achieved over 99% accuracy in describing complex, non-reciprocal forces within the plasma. The research challenges long-held theoretical assumptions, such as the direct proportionality of particle charge to size and the independence of force decay from particle size. The team utilized precise 3D tomographic imaging to track particle motion in controlled laboratory experiments, providing high-resolution data for the AI's training. This method, which runs on a standard desktop computer, is designed to be universally applicable to other many-body systems, including industrial materials and living cells, demonstrating AI's potential beyond data analysis to fundamental scientific discovery.

Key takeaway

For AI and Research Scientists exploring complex systems, this work demonstrates that carefully designed AI models can directly uncover new physical laws and correct existing theories. You should consider integrating physics-tailored neural networks into your research workflows, especially when dealing with limited experimental data, to reveal hidden dynamics and challenge established assumptions in fields ranging from materials science to biophysics.

Key insights

AI can discover new physical laws and challenge long-standing assumptions in complex systems.

Principles

• Non-reciprocal forces are common in many-body systems.
• Particle size affects force decay in dusty plasmas.

Method

A custom neural network, trained on 3D particle trajectory data from dusty plasma experiments, infers underlying physical laws by separating influences like drag, environmental forces, and inter-particle forces.

In practice

• Apply this AI framework to other many-body systems.
• Use 3D tomographic imaging for precise particle tracking.

Topics

• Physics-Tailored Machine Learning
• Dusty Plasma
• Non-Reciprocal Forces
• Particle Interactions
• 3D Particle Tracking

Best for: AI Scientist, Research Scientist

Related on AIssential

• Accelerating science with AI and simulations — AI is at a second inflection point, integrating language and multi-modal reasoning to accelerate scientific discovery.
• #AAAI2026 invited talk: machine learning for particle physics — Machine learning, especially graph neural networks, can overcome limiting assumptions in particle tracking to discover new physics.
• When Deep Learning Meets Physics — Integrating physical laws into neural networks is a crucial, under-recognized advancement for AI's real-world utility.
• NSF renews support for MIT-led AI and physics institute, expanding a new model for discovery — AI and physics form a "virtuous cycle," mutually advancing discovery and system development.
• When AI Starts Seeing Inside Atoms: What I Learned About Deep Learning From Nuclear Physicists — Deep learning can accurately model complex quantum phenomena like atomic nuclear charge distribution.
• GPT-5.2 derives a new result in theoretical physics — AI can discover and formally prove new theoretical physics results, challenging long-held assumptions.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Robotics Research News -- ScienceDaily.