Experiment 1

Testing

Plots

Prede

Behavior(t) = argmaxₐ** [ (D(a) × A(a) × M(a)^P(a) × H(a) × I(a)) / F(a) ]

at each time step, the actor selects the action *a* that maximizes this ratio.

- **D(a)** — **Drive / Desire / Need pressure**
How strongly the actor feels the need or urgency for the outcome of action *a* (survival, security, economic gain, political legitimacy, etc.). Peaks near maximum in existential resource scarcity.
- **A(a)** — **Agency / Belief the action can succeed**
Perceived capability and likelihood of pulling off action *a* without catastrophic failure (military reach, economic leverage, sovereign control, technological edge, etc.).
- **M(a)** — **Meaning / Perceived value of outcome**
The baseline subjective value or reward the actor assigns to achieving the outcome of action *a* (wealth, power, national pride, survival of population, legacy, etc.).
- **P(a)**—**Persistence amplifier** (formerly called Greed amplifier)
Exponent that nonlinearly amplifies the perceived value: effective value = M(a)^P(a)
- P = 1 → neutral weighting
- P > 1 → strong amplification of short-term / self-maximizing outcomes (makes risky or zero-sum actions look disproportionately rewarding)
- P < 1 → restraint or de-emphasis on immediate gains
Persistence reflects how the behavior becomes self-reinforcing once chosen (repeated payoffs, sunk costs, identity lock-in, habit escalation).
- **H(a)**—**Habit strength from repetition**
How ingrained or automatic the action *a* (or similar actions) has become due to prior success or repetition. High H makes switching costly even if better options exist.
- **I(a)**—**Interdependence multiplier** (the key addition for relational invariance)
Adjusts the entire numerator based on how mutually dependent actors are in the situation:
- I > 1 → high mutual dependence + correspondence of interests (boosts cooperative / positive-sum actions, makes shared benefits higher-value)
- I ≈ 1 → low or neutral interdependence (individual calculation dominates)
- I < 1 → high asymmetry or conflict of interests (suppresses cooperation, amplifies zero-sum or unilateral persistence)
This term enforces relational invariance: the same equation structure produces different outcomes depending on the actor’s position in the relational network (upstream vs. downstream, exporter vs. importer, hegemon vs. vulnerable).
- **F(a)**—**Friction (cost + risk + effort + punishment)**
All barriers, costs, and downsides of pursuing action *a* (economic cost, military retaliation, sanctions, diplomatic isolation, long-term ecological blowback, domestic backlash, etc.). Acts as a strong penalty in the denominator.

Intro

Prede

talk and it should predict behavior but AI can do silly thingsGuest