Newswise — How can scientists foresee and avoid massive disruptions in plasma, a key hurdle to bringing the fusion reactions that power the sun and stars to Earth to generate electricity? “You can’t have a prototype reactor if it’s disrupting,” says William Tang, a physicist at PPPL and a Princeton University professor who leads a project to forecast disruptions through artificial intelligence (AI) — the branch of computer science that is transforming scientific inquiry and industry.  

Such intelligence can rapidly spot signs of disruptions by dealing with many more variables than previously have been detected, says Tang, providing a big practical benefit to scientists seeking the best methods for avoiding disruptions. Here in this interview from the San Diego Supercomputer Center, a National Science Foundation-sponsored branch of the University of California San Diego, Tang in his own words describes the remarkable ability of AI to accelerate the development of clean and abundant fusion energy.