Pieter Abbeel

Pieter Abbeel is a Professor of Artificial Intelligence and Robotics at UC Berkeley and an Amazon Scholar, widely recognized as one of the world’s leading figures in AI research, education, and entrepreneurship. His work spans generative AI, reinforcement learning, robot manipulation, and large-scale learning systems, and has produced foundational advances such as TRPO, SAC, Hindsight Experience Replay, MAML, Domain Randomization, Decision Transformer, UniSim, and diffusion-based generative models. Abbeel has co-founded pioneering companies including Gradescope—an AI system that transformed academic assessment—and Covariant, a leader in robotics foundation models for warehouse and factory automation. His students have founded more than a dozen influential AI companies, including OpenAI, Perplexity, Physical Intelligence, Skild, Evolutionary Scale, Reflection, Genmo, and Ideogram, reflecting the ecosystem of innovation surrounding his lab. Abbeel’s contributions have been recognized with major honors such as the PECASE, NSF CAREER, ONR YIP, DARPA YFA, and the ACM Prize in Computing, and his work is frequently covered by global media outlets including The New York Times, BBC, Wired, and MIT Technology Review. His current research explores generative AI, humanoid robotics, robot learning at scale, and the integration of large models into physical intelligence systems.