At Emory University, a team of researchers has just Rewrite a page of modern physics. Thanks to an artificial intelligence specially designed for this mission, they corrected false ideas that slowed down the understanding of dusty plasma.
Consequently, this advance reveals unsuspected natural forces and opens up new perspectives for the study of complex systems.
How AI has dissected the mysterious behavior of dusty plasma particles
Dusty plasma is a Hot and electrically loaded gas containing tiny solid particles. We find it as well in Saturn rings that in the Forest fire smoke. However, the forces which act between these particles remained poorly understood because of their asymmetrical interaction.
To better understand this phenomenon, the team designed a Unpublished neural network. Unlike classical AIs that are content to predict or filter, this model has been designed to Discover new physical laws.
Thus, the researchers developed a 3D imaging system associating a laser leaf and a ultra-fast camera. Thanks to this technology, they were able to follow the movement of thousands of particles in real time.
By relying on a reduced but extremely rich data set, AI has identified invisible forces. She also integrated from the departure of known laws such as gravity or trail in order to refine her conclusions.
The received ideas swept: what AI has corrected in plasma theory
So far, scientists thought that the electrical load of a particle increased proportionately to its size. However, AI has shown that it actually depends on the density and temperature surrounding plasma.
In addition, another current hypothesis claimed that the force between two particles always decreased exponentially, whatever their size. The results show, however, that this decrease varies according to the diameter particles.
These conclusions were obtained With a simple desktop computer. This proves that the key to discovery does not only reside in the computing power, but especially in the precise design of the algorithm.
Why this method could transform the study of many complex systems
This approach is distinguished by its Ability to learn with little data. The AI has separated the role of speed, external forces and direct interactions between particles. Thus, it has reached a precision greater than 99 %.
As explained Justin Burtonco-author of the study, this model is not a black box. Its logic remains clear and adaptable to other areas: cell biology, industrial processes or Cosmic modeling.
Ultimately, this method shows that AI can reveal unknown laws And transform the way researchers explore chaotic systems, whether stars or living cells.
