"Particle accelerators are some of the most powerful machines humans have ever built, driving discoveries in materials, medicine, microelectronics, and the fundamental nature of our universe. But designing and running one is incredibly complex. There are thousands of components and millions of possible settings that can take hours or days to get right. That's where MOAT, part of the Department of Energy's Genesis mission, comes in. Berkeley Lab leads the multi-office particle accelerator team in putting artificial intelligence to work alongside accelerator operators and designers. MOAT is building digital twins that mirror real accelerators, AI models that can predict failures before they happen, and intelligent assistants that operators and designers can simply talk to. The end goal is a new generation of smarter, more efficient accelerators that can tune themselves and help experiments run autonomously, so researchers at labs, universities, and industries across the country can make discoveries and innovate faster."