Deep reinforcement learning
A machine learning approach combining deep neural networks with reinforcement learning, used in AlphaGo and AlphaZero.
As defined in AMA | June 2023
Other definitions
Learning control policies through trial and error with deep neural networks, guided by reward signals, able to discover strategies beyond human examples
World's Top Researcher on AI, LLMs, and Robot Intelligence
Where it comes up · 4
- World's Top Researcher on AI, LLMs, and Robot Intelligence3 claims
- Demis Hassabis — Scaling, superhuman AIs, AlphaZero atop LLMs, AlphaFold2 claims
- Demis Hassabis — Scaling, superhuman AIs, AlphaZero atop LLMs, AlphaFold1 claim
- AMA | June 20231 claim
Most important claims · 7 of 7Search all claims for “Deep reinforcement learning” →
- Combining skill practice with prior knowledge is a hard unsolved problem across all of AI, because generative AI reproduces human abilities while deep RL discovers things humans have not thought of (AlphaGo's Move 37), and the field has not yet fused these two strengths.World's Top Researcher on AI, LLMs, and Robot Intelligence
- Many ideas from earlier deep reinforcement learning work like DQN and AlphaGo are coming back into fashion and should be recombined with the new advances in large multimodal models, with significant potential in merging older and newer ideas.Demis Hassabis — Scaling, superhuman AIs, AlphaZero atop LLMs, AlphaFold
- Reinforcement learning and deep reinforcement learning are examples of research ideas that were underrated by the field for years despite early promise, then came back into focus at scale; similar to how scaling was underrated before recent successes, there are likely current research directions that are underrated but will be important when combined with new advances.Demis Hassabis — Scaling, superhuman AIs, AlphaZero atop LLMs, AlphaFold
- The merger of Brain and DeepMind created synergies through resource pooling, idea sharing, and deeper engineering collaboration than project-by-project collaboration, accelerating frontier model development.Demis Hassabis — Scaling, superhuman AIs, AlphaZero atop LLMs, AlphaFold
- Beyond particle physics data analysis, deep reinforcement learning and similar brute-force techniques could in principle help explore conceptual structures—such as mapping ontologies or surfacing paradoxes and conceptual limitations that humans are blind to—though such applications to philosophy remain speculative and unproven.AMA | June 2023
- After a policy has been practiced, reinforcement learning from its own experience improved an espresso-making robot's robustness, speed and throughput, and Levine says that work is far from finished.World's Top Researcher on AI, LLMs, and Robot Intelligence
- Deep reinforcement learning from the early 2010s is likely a historical milestone because it allows systems to exceed human-level performance, which Levine considers essential for robotics; the advent of multimodal LLMs adapted for control is another major advance, though its historical status is uncertain.World's Top Researcher on AI, LLMs, and Robot Intelligence