Reinforcement learning has solved Go, DoTA, Starcraft -- some of the hardest, most strategy-intensive games for humans. Nonetheless, these games are missing fundamental aspects of real-world intelligence: large agent populations, ad-hoc collaboration vs. competition, and extremely long time horizons, among others. Neural MMO is an environment modeled off of Massively Multiplayer Online games -- a genre supporting hundreds to thousands of concurrent players, realistic social incentives, and persistent play. I will discuss the current state of the project, key enabling features of Ray + RLlib, and infrastructure required for continued expansion.
Joseph Suarez is an MIT PhD student in Phillip Isola’s lab and the creator of the Neural MMO platform for massively multiagent research. He graduated from Stanford University with a BS in Computer Science with emphasis on Artificial Intelligence 2019. During his undergraduate studies, Joseph completed two years of full-time research and released several papers in natural language processing, computer vision, and reinforcement learning, one of which was published as a single author paper in NIPS 2017. Joseph interned at OpenAI for six months where he developed and released the first version of Neural MMO. He is currently continuing the project’s development at MIT.