Reward is enough

By David Silver, Satinder Singh, Doina Precup, Richard S. Sutton. 2021

554 citations recorded by the source corpus.

Abstract In this article we hypothesise that intelligence, and its associated abilities, can be understood as subserving the maximisation of reward. Accordingly, reward is enough to drive behaviour that exhibits abilities studied in natural and artificial intelligence, including knowledge, learning, perception, social intelligence, language, generalisation and imitation. This is in contrast to the view that specialised problem formulations are needed for each ability, based on other signals or objectives. Furthermore, we suggest that agents that learn through trial and error experience to maximise reward could learn behaviour that exhibits most if not all of these abilities, and therefore that powerful reinforcement learning agents could constitute a solution to artificial general intelligence.

Subfields: Reinforcement Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Reinforcement Learning

Citations in the corpus

On Reinforcement Learning and Distribution Matching for Fine-Tuning Language Models with no Catastrophic Forgetting (2022, background): “RM and DM objectives in control problems While RM is the dominant approach to tackling control problems (Sutton & Barto, 2018) and is sometimes argued to be sufficient for any intelligent behavior (Silver et al., 2021), prior work explored the benefits of alternative objectives formulated as DM: min”
Security and Privacy for 6G: A Survey on Prospective Technologies and Challenges (2021, background): “A promising solution is to improve the generative learning ability of deep learning models, as the suggestion of artificial general intelligence in [176], [179].”
Symbolic Behaviour in Artificial Intelligence (2021, background): “Such intentional action motivates techniques that exploit the situated, goal-directed aspects cognition, such as reinforcement learning [53, 54, 55].”
Learning Natural Language Generation with Truncated Reinforcement Learning (2022, background): “"Reward is enough" (Silver et al., 2021) highlights the necessity of using RL for AI systems to acquire language in its full richness.”
Contrastive Learning as Goal-Conditioned Reinforcement Learning (2022, background): “While this paper might be construed to imply that RL is more or less important than representation learning [72, 75, 112, 114], we have a different takeaway: that it may be enough to build RL algorithms that look like representation learning.”
Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback (2023, background): “Even modeling human preferences with a reward at all, implicitly accepting the reward hypothesis [Silver et al., 2021], might be unwarranted [Skalse and Abate, 2022b, Vamplew et al.”