Generally capable agents emerge from open-ended play - Google DeepMind

In recent years, artificial intelligence agents have succeeded in a range of complex game environments. For instance, AlphaZero beat world-champion programs in chess, shogi, and Go after starting out with knowing no more than the basic rules of how to play. Through reinforcement learning (RL), this single system learnt by playing round after round of games through a repetitive process of trial and error. But AlphaZero still trained separately on each game — unable to simply learn another game or task without repeating the RL process from scratch. The same is true for other successes of RL, such as Atari, Capture the Flag, StarCraft II, Dota 2, and Hide-and-Seek. DeepMind’s mission of solving intelligence to advance science and humanity led us to explore how we could overcome this limitation to create AI agents with more general and adaptive behaviour. Instead of learning one game at a time, these agents would be able to react to completely new conditions and play a whole universe of ga

July 27, 2021 Research Open-Ended Learning Team In recent years, artificial intelligence agents have succeeded in a range of complex game environments. For instance, AlphaZero beat world-champion programs in chess, shogi, and Go after starting out with knowing no more than the basic rules of how to play. Through reinforcement learning (RL), this single system learnt by playing round after round of games through a repetitive process of trial and error. But AlphaZero still trained separately on each game — unable to simply learn another game or task without repeating the RL process from…

related reading

• How DeepMind's Generally Capable Agents Were Trained — LessWronglesswrong.com
• Reinforcement learning towards broadly and persistently beneficial modelsalignment.openai.com
• Learning To Play Settlers of Catan With Deep RLsettlers-rl.github.io
• Teacher Algorithms for Deep RL Agents that Generalize in Procedurally Generated Environments – Developmental Systems, a Blog of the Flowers Labdevelopmentalsystems.org
• DeepMind: Generally capable agents emerge from open-ended play — LessWronglesswrong.com
• Deep Reinforcement Learning: Pong from Pixelskarpathy.github.io
• A VLA with Open-World Generalizationpi.website
• [AN #159]: Building agents that know how to experiment, by training on procedurally generated games — AI Alignment Forumalignmentforum.org
• Learning Beyond Gradientstrinkle23897.github.io
• Embedded Agency (full-text version) — LessWronglesswrong.com
• General Agent: A Self-Evolving, Synthetic Agent Environmentprimeintellect.ai
• RL Environments and RL for Science: Data Foundries and Multi-Agent Architecturesnewsletter.semianalysis.com