Uploaded September 2026 | Updated November 2026, 13 hours ago
Reinforcement learning is one of the hardest distributed systems problems in AI: training, inference, and environments all running in one loop.
Robert Nishihara, creator of Ray and co-founder of Anyscale, breaks down the anatomy of a reinforcement learning workload on the Ray Summit 2026 Day 1 keynote stage: the components an RL system has to coordinate, why the combination of large-scale training and large-scale inference in a single loop stresses infrastructure in ways neither does alone, and how Ray was built for exactly this shape of problem.
Liked this video? Check out other Ray Summit keynote sessions: youtube.com/playlist?list=PLYnBpswCtPo4
Subscribe to our YouTube channel to stay up-to-date on the future of AI: youtube.com/c/anyscale
đź”— Connect with us:
LinkedIn: linkedin.com/company/joinanyscale
X: https://x.com/anyscalecompute
Website: anyscale.com
Reinforcement learning is one of the hardest distributed systems problems in AI: training, inference, and environments all running in one loop.
Robert Nishihara, creator of Ray and co-founder of Anyscale, breaks down the anatomy of a reinforcement learning workload on the Ray Summit 2026 Day 1 keynote stage: the components an RL system has to coordinate, why the combination of large-scale training and large-scale inference in a single loop stresses infrastructure in ways neither does alone, and how Ray was built for exactly this shape of problem.
Liked this video? Check out other Ray Summit keynote sessions: youtube.com/playlist?list=PLYnBpswCtPo4
Subscribe to our YouTube channel to stay up-to-date on the future of AI: youtube.com/c/anyscale
đź”— Connect with us:
LinkedIn: linkedin.com/company/joinanyscale
X: https://x.com/anyscalecompute
Website: anyscale.com










