Skip to content

🧠 Cognitive Division of Labor (Latent vs. Introspective Agents)

This simulation explores one synthetic division-of-labor design connected to the Agentic Society hypothesis.

Status: The result follows from assigned movement delays, visibility, and pheromone rules. It does not show that omniscience causes paralysis or that heterogeneous teams are always superior.

The Simulation

The simulation runs three isolated resource-gathering societies in parallel:

  1. Pure Latent (Intuition Only): 10 agents that move every single tick. They have \(R \approx 0\) (no reflectivity). They only see 1 step ahead. They are fast but often wander aimlessly or loop around local minima.

  2. Pure Introspective (Reflexion Only): 10 agents that calculate the absolute perfect path (A* distance mapping) to the nearest resource. They have \(R \approx 1\) (high reflectivity). However, computing this "god-mode" vision takes time. They only move every 5 ticks.

  3. Symbiotic (Action + Reflexion): A mix of 7 Latent and 3 Introspective agents.

  4. The Introspective agents act as the slow-moving "consciousness" of the system. While computing their perfect paths, they leave strong gradients ("meaning"/pheromones) in the environment.
  5. The Latent agents act as the fast-moving "body". They still only see 1 step ahead, but their instincts drive them up the gradients left by the Introspectives.

Result: The Symbiotic society typically outperforms the others in this toy environment, suggesting that a ā€œdivision of cognitive laborā€ (slow reflection shaping fast action through environmental memory) can outperform homogeneous swarms under some conditions.

Running the Simulation

pip install numpy matplotlib
python latent_introspective_society.py