AI Identification- Variables in Physical Concepts

AI identification has reached that point in the world where AI programs can now help identify variables in physical concepts.

AI Identification
AI Identification

Variables are always necessary to understand any physical phenomena. While most physical connections contain factors that scientists are aware of, a few have remained mysterious. Researchers from Columbia University have now created software that watches such physical occurrences and recognizes the pertinent factors using artificial intelligence (AI). The computer application makes use of a video camera to record the dynamics and then analyses the data to determine the bare minimum of essential variables needed to describe it.

Researchers started their study by processing raw video in the system whose solution they previously knew. They then compared their own results with those of the AI system. It showed out to be rather close. Hod Lipson, head of the Creative Machines Lab at the Department of Mechanical Engineering, said “We determined that this response was adequate. Particularly considering that none of the AI could understand physics or geometry and just had access to unprocessed video material. However, we weren’t simply interested in the quantity of the variables; we also wanted to know what they were”.

The team then made an effort to visualize the factors the computer had found. While they discovered two factors that matched the angles of the arms, the other two remained unaccounted for. “We attempted to correlate the other variables with every possible combination of known values, including angular and linear velocities, kinetic and potential energy. Boyuan Chen Ph.D. ’22, assistant professor at Duke University and study’s principal author, said nothing, however, it “seemed to correspond precisely.”

The system was still being tested by the researchers, who fed it films for which they had no answers. These featured lava light and air dancer videos. Eight variables were provided for both of them by the system. The algorithm returned 24 variables for a looping film of Christmas fireplace flames.

The group is now hoping that AI software might assist researchers in understanding complicated events in fields like cosmology and biology.


Data Science related posts visit HERE

Algorithms-related posts visit HERE

Data Structures related posts visit HERE

Databases-related posts Visit HERE

Python-related posts Visit HERE

C++-related posts Visit HERE