To understand any physical phenomenon, one must identify the variables that are responsible for it. While scientists are familiar with the variables of most physical associations, some remain elusive. Now, researchers at Columbia University have used artificial intelligence (AI) to develop a program that observes such physical phenomena and detects relevant variables. The program uses a video camera to observe the dynamics and then processes the information to state the minimum set of fundamental variables needed to describe it.
In studypublished in Nature Computational Science, the researchers began by processing raw video into systems they already knew the answer to. Then they matched the result of aye With the system that turned out to be close. “We thought this answer was close enough. That’s where the main job was done. Especially since all AIs had access to raw video footage without any knowledge of physics or geometry. But we wanted to know what the variables were actually what were in me, not just their number,” Told Hod Lipson, Director Creative Machine Lab in the Department of Mechanical Engineering. Lipson is also an author of the study.
Next, the team tried to visualize the variables that the program recognized. while they got two variable Corresponding to the angles of the sides, the other two cannot be described. “We tried to correlate other variables with anything and everything we could think of: angular and linear velocities, kinetic and potential energy, and various combinations of known quantities. But nothing matched perfectly. Did not eat,” explained Boyuan Chen PhD ’22, an assistant professor at Duke University and lead author of the study.
The researchers continued to test the system and fed the video into the system, to which they had no response. These included videos of an air dancer and a lava lamp. The system gave eight variables for both of them. Meanwhile, for the video of the flames from the holiday fireplace loop, the system returned 24 variables.
The team now hopes that such an AI program could help scientists understand complex phenomena in fields ranging from biology to cosmology.