AI breakthrough could replace rare earth magnets in electric vehicles

University of New Hampshire scientists developed an AI system that reads scientific papers to identify magnetic materials, creating a database of 67,573 magnetic compounds including 25 previously unrecognized high-temperature magnets. This breakthrough could help reduce electric vehicle costs by finding alternatives to expensive rare earth magnets that are mostly imported and difficult to secure. The AI approach dramatically speeds up materials discovery compared to traditional laboratory testing of millions of possible element combinations.