AI systems have found a kind of physics that scientists have never seen before

Here's what you'll learn when you read this story:

For all the problems that AI is causing society, one of its greatest benefits lies in the world of science.
A new study focusing on the chaotic dynamics of dusty plasmas found that, when properly trained, AI can actually discover new physics.
By providing the most detailed explanation of this type of material, AI has revised its long-standing theoretical beliefs about how particles behave within dusty plasmas.

In one or more ways, artificial intelligence is making the world worse. Generated AI is currently spitting countless “AI slops” and in the classroom, AI is slowly eroding critical thinking skills. Deathly. This also does not mention the unfortunate role of AI as an environmental decimerator or as a job destroyer.

Fortunately, some artificial intelligence and machine learning (ML) models have more epic ambitions than eating My beloved animator A mass-produced essay at the reading level for 8th graders. For example, consider a new ML model developed by a team of scientists at Emory University. Machine learning algorithms are usually used as tools to help scientists sift through huge amounts of data and optimize experiments, but this particular ML model has actually discovered new physics on its own, at least as relevant to dusty plasmas.

You're probably familiar with plasma – that fourth material state that actually constitutes 99.9% of all normal matter in the universe. The dusty plasma is simply the same mixture as the ionized gas, but with charged dust particles. This type of plasma can be found in both spatial and terrestrial environments. For example, wildfires produce dusty plasmas for soot particle particles mixed with smoke. This new study was published in a journal Proceedings of the National Academy of Sciences (PNAS) – A team of researchers explains how a trained ML model can successfully provide the most detailed explanation of dusty plasma physics and create accurate predictions of incompatible forces.

“Our AI methods are not black boxes. We understand how it works,” Justin Burton, co-author of Emory's research, said at a press conference. “The framework it provides is also universal. It could be applied to other many-body systems to open up new routes to discovery.”

Simply put, non-polar forces (as the name suggests) occur when the force applied between two particles in the plasma is not the same. The author describes this phenomenon as two boats affected by the other obstacle. The position of the relationship can affect the attractive or repulsive force of the particles.

“In dusty plasma, we explained how the leading particles attract subsequent particles, but the following particles always fight off the major particles,” Ilya Nemenmann, another co-author of Emory's study, said at a press conference. “This phenomenon was expected by some, but now there are precise approximations that didn't exist before.”

The ML algorithm was also able to correct some theoretical misconceptions about dusty plasmas. For example, scientists thought that the charge of a particle is proportional to its size, but the model confirms that larger particles contain larger charges, but are not proportional because they can be affected by density and temperature. They also found that the charge between particles affects not only the distance between the two particles, but also the size of the particles.

According to the author, one of the most difficult parts of this project was designing ML algorithms in the first place. In general, AI acquires its capabilities by supplying (or training) datasets. This will gradually improve the identification of the monkey when you consume a million photos of the monkeys and when the monkeys are seen. But when you discover new Physics doesn't have much training data to continue. Therefore, the team had to create a structure that could work with data. did To explore unknown physics, you have it while giving it latitude.

“I think of it like a Star Trek motto. “A properly used AI can open doors to whole new realms and explore them.”