Definition
Deliberative Agents are AI systems that maintain internal state, construct models of the environment, and plan future sequences of actions to achieve goals.
Why It Matters
They are capable of solving complex, novel, and long-term planning tasks.
Core Concepts
- Reflex Agent (Stimulus-Response): Operates on “if-then” rules or direct mappings (e.g., a thermostat). It is fast and efficient but inflexible in novel environments.
- Deliberative Agent (Model-Based): Maintains an internal representation of the world (“What happens if I do X?”). It evaluates actions based on their expected contribution to a goal.
- Hybrid Architectures: Most complex systems (including humans) use both—reflexes for immediate, low-stakes tasks and deliberation for high-stakes, novel planning.
- Lookahead and Search: Deliberative agents use techniques like Monte Carlo Tree Search to “look into the future” and find optimal paths.