Agential AI for Integrated Continual Learning, Deliberative Behavior, and Comprehensible Models

Contemporary machine learning faces key limitations, such as a lack of integration with planning, incomprehensible structure, and inability to learn continually. We present initial design for system, Agential AI (AAI), that overcomes these issues. AAI's core is a learning method that models temporal dynamics with guarantees of completeness, minimality, and continual learning. It integrates this with a behavior algorithm that plans on a learned model and encapsulates high-level behavior patterns. Preliminary experiments on a simple environment show AAI's effectiveness and potential.