Goal
The desired outcome toward which behavior is directed.
h/ai Primitive Language
Volume IV
Behavioral Primitives
Behavioral primitives describe how intelligent systems reason, learn, plan, decide, remember, and act.
Overview
Behavioral primitives describe how a Human–AI system transforms available information into decisions and actions. They operate independently of the participants involved, the structure of the interaction domain, the information being communicated, and the temporal progression of interaction.
Behavioral reasoning consumes information supplied by the other primitive families to infer goals, evaluate constraints, plan future behavior, select actions, and improve future performance through learning.
Primitive Definitions
Inference
Derivation of conclusions from available information, memory, context, and domain state.
Planning
Organization of future actions required to achieve a Goal.
Decision
Selection of a course of action from available alternatives.
Preference
A learned or defined tendency that influences behavioral selection.
Policy
Rules governing permissible behavior and external consequence.
Learning
Modification of future behavior through acquired experience.
Memory
Previously acquired information made available to future reasoning.
Action
Execution of behavior producing an internal or external effect.
Example: Missing Information During Ordering
A customer appears to have completed an order, but required Domain variables remain unresolved. Behavioral reasoning determines that clarification is required before commitment.
Domain State
Bagel
✓
Spread
✓
Toast Preference
Missing
↓
Inference
Required information
is incomplete.
↓
Goal
Complete order.
↓
Planning
Request missing information.
↓
Decision
Delay commitment.
↓
Action
"Would you like your bagel toasted?"
Example: Memory Influencing Behavior
Previously stored information may influence future behavioral decisions while still allowing the user to override prior preferences.
Memory
Previous order:
Sesame
Toasted
↓
Inference
Prior preference detected.
↓
Policy
Confirm before reuse.
↓
Decision
Ask user.
↓
Action
"Last time you ordered
a toasted sesame bagel.
Would you like the same today?"
Relationship to Other Families
Agency
Agency identifies who is interacting. Behavior determines how the system responds.
Domain
Domain defines available variables, constraints, and completion state. Behavior reasons over those structures.
Information
Information provides Meaning. Behavior interprets that Meaning to select decisions and actions.
Temporal
Temporal determines when behavioral decisions occur and when actions become committed.
Behavioral Principle
Human–AI systems should not merely generate responses. They should reason over available information, evaluate goals and constraints, select appropriate actions, and continuously improve future behavior through learning.