Physicists at Emory University have utilized machine learning to uncover previously unknown aspects of dusty plasma, a state of matter prevalent in space and certain terrestrial environments. Their research, published in the journal *Proceedings of the National Academy of Sciences*, demonstrates the potential of AI in discovering new physical laws governing complex systems.
Dusty plasma consists of ionized gas containing suspended dust particles and is commonly found in space and planetary environments. The Emory team employed a neural network model trained on experimental data from laboratory dusty plasma to identify non-reciprocal forces within the system. This approach allowed them to describe these forces with high accuracy, challenging existing scientific assumptions.
By tracking the three-dimensional motion of individual particles in a dusty plasma, the researchers observed interactions that had not been previously quantified. Their findings suggest that particle charge is influenced not only by size but also by plasma temperature and density, indicating a more complex relationship than previously understood.
This research highlights the growing role of AI in scientific exploration, offering new tools to probe the complexities of the natural world. The implications extend beyond dusty plasma, potentially informing the study of other many-body systems, including biological systems and industrial materials.
The study was primarily funded by a grant from the National Science Foundation, with additional support from the Simons Foundation. The interdisciplinary collaboration between experimental and theoretical physicists underscores the potential of combining AI with traditional scientific methods to advance our understanding of complex systems.