Professor Pieter Abbeel, Director of the Berkeley Robot Learning Lab and Co-Director of the Berkeley Artificial Intelligence (BAIR) Lab, talks with Jeff Bier, Founder of the Edge AI and Vision Alliance, for the “Perspectives on AI Beyond Pattern Recognition” interview at the May 2021 Embedded Vision Summit. See here for Abbeel’s keynote at the Summit, for which this interview is an extended Q&A session.
Abbeel’s research strives to enable ever more intelligent systems. His lab pushes the frontiers of deep reinforcement learning, deep imitation learning, deep unsupervised learning, transfer learning, meta-learning and learning to learn, and studies the influence of AI on society.
Is AI ready to move beyond pattern recognition to create autonomous systems that achieve goal-oriented behaviors? What are the practical challenges and limitations in applying deep reinforcement learning and deep imitation learning? What characteristics make an application well suited to these techniques? What safeguards are needed to ensure that autonomous systems remain safe? In this interview, Abbeel shares his perspectives on these and other topics.