
(l-r) Samuel Barber and Remi Lehe inspect the experimental setup at ATAP’s BELLA Center, where they are developing AI-based laser-plasma acceleration technology for the AI Genesis Mission project. (Thor Swift/Berkeley Lab)
Fusion, the process that powers our Sun, could one day provide abundant, clean power. Scientists study it with powerful lasers, but experiments are rare and costly, so experimental parameters must be chosen carefully. The project, led by the University of Rochester in collaboration with Berkeley and an industry partner, will develop an AI assistant that reviews past experiments and facility records to provide researchers with plain-language guidance in real time, helping them learn faster and run more effective experiments to advance fusion research.