little-dan-framework

Understanding Free Will - Human Decision-Making Through Different Agent Models

This case study illustrates different human decision-making modes from passive, fixed reactions to dynamic and proactive behaviors, emphasizing how each mode reflects varying degrees of free will.

1. Reflective Agent

Description: The reflective agent operates purely on automatic, involuntary responses. Actions are fundamental reactions without conscious thought or choice, reflecting an absence of free will.

Example: When touching a hot stove, your hand instinctively pulls away before conscious recognition occurs. This reflexive action demonstrates no free will—pure biological programming dictates the response.

Free Will Aspect: Nonexistent, as the agent acts entirely on involuntary reflexes.


2. Survival Agent

Description: The survival agent bases decisions on past experiences to minimize danger or discomfort. Choices are still passive, driven primarily by learned avoidance behaviors rather than proactive intentions.

Example: If someone once experienced food poisoning from a specific dish, they might automatically avoid that dish in the future, even without immediate danger present.

Free Will Aspect: Limited, as decisions predominantly result from conditioned responses rather than conscious choice or proactive intent.

Animals often have this level of the agency.


3. Fixed Goal-Driven Agent

Description: The fixed goal-driven agent sets clear, predetermined objectives and makes consistent decisions aligned with those goals. This is a profound shift from passive reactions to proactive pursuing. This agent exhibits greater autonomy and conscious choice compared to the survival agent.

Example: A person striving to lose weight sets a specific daily calorie goal. Regardless of social events or temptations, they adhere strictly to this calorie target.

Free Will Aspect: Moderate. Deliberate goal makes free will much more clearer in this mode compared to previous two modes. This is the beginning of “Taking control” of your own life proactively opposite to simply being reactive and passive. As individuals consciously establish their goals, yet remain constrained by their predefined objectives, limiting adaptability.


4. Dynamic Goal-Driven Agent

Description: The dynamic goal-driven agent continuously adjusts its objectives based on self-reflection, environmental context, and changing circumstances. This represents a significant step toward proactive and adaptive decision-making.

Example: An individual initially aiming for morning workouts realizes mornings are too rushed. Consequently, they shift their goal to exercising in the evening, flexibly adapting their approach while still maintaining their overarching fitness objective.

Free Will Aspect: High, as individuals actively reflect, reconsider, and adapt their goals and actions, demonstrating considerable autonomy and self-direction. (This could be done through Observation eye concept in my “Little Dan Framework”to break the auto-pilot mode and be deliberate about examining the current fixed goal and axiom of believes to modify or update the goal based on meta-goal)


5. Instinct-Driven Agent

Description: This agent makes decisions based on deep-seated subconscious instincts or intuitions (Little Dan), transcending explicit reasoning or defined goals. Actions emerge from an internal, often inexplicable, intuitive awareness rather than rational analysis (Daddy Dan).

Example: Someone spontaneously chooses an unfamiliar route home without apparent reason, later learning about a major accident on their usual path. This action results from instinct rather than conscious thought or learned experience. Another example would be given all the analysis based on explicit goals that one set, people choose to jump out of the most reasonable action and yet choose to the opposite action on purpose, based on unexplainable “gut feelings” or simply be curious about what would happen if I choose not based on rationality.

Free Will Aspect: Extremely high or unclear. Complex and nuanced—actions are neither purely logical nor consciously planned, reflecting an intuitive form of autonomy that integrates subconscious intelligence and deeper cognitive processes. This often is an attempt to break the assumptions or axioms of the current believe systems and think differently guided by unknown subconscious or intuitions that often have much more depth of the coverage beyond limited symbolic system that’s describable. Although this agent could often lead to more chaotic or unexpected situations, it at times can also be useful. It’s the ability to say “No” to the obvious logical conclusion and might have evolutionary advantages of creating chaos (diversity) on purpose.


Evolutionary Insight

These agent models demonstrate a clear evolutionary trajectory in human decision-making:

Every mode has its own Pros and cons and not necessarily means one has to progress from low agency to high agency. At any given moment in front of us, one can maybe choose which modes you want to use explicitly or inexplicably, which the choice of choosing modes itself becomes a free will topic.

This framework enriches our understanding of human behavior by highlighting the diverse interplay between reflex, learned behavior, goal-oriented decisions, adaptive flexibility, and intuitive insights that shape daily decision-making. It might also give an intersting framework to start understanding the AI agency and how much of “free will” we would like each AI system to have.