Where do goals come from? A Generic Approach to Autonomous Goal-System Development
This addresses the foundational issue of autonomous goal formation in AI agents, offering a novel computational framework with potential broad applications in reinforcement learning and developmental robotics.
The paper tackles the problem of how agents can autonomously develop goals by proposing Latent Goal Analysis, which treats goals as abstractions of rewards and values, and shows that learned goals provide effective dimensionality reduction in reinforcement learning and enable goal-directed reaching from task-unspecific rewards.
Goals express agents' intentions and allow them to organize their behavior based on low-dimensional abstractions of high-dimensional world states. How can agents develop such goals autonomously? This paper proposes a detailed conceptual and computational account to this longstanding problem. We argue to consider goals as high-level abstractions of lower-level intention mechanisms such as rewards and values, and point out that goals need to be considered alongside with a detection of the own actions' effects. We propose Latent Goal Analysis as a computational learning formulation thereof, and show constructively that any reward or value function can by explained by goals and such self-detection as latent mechanisms. We first show that learned goals provide a highly effective dimensionality reduction in a practical reinforcement learning problem. Then, we investigate a developmental scenario in which entirely task-unspecific rewards induced by visual saliency lead to self and goal representations that constitute goal-directed reaching.