EDAP Model v4.0

Endogenous Dissipation of Adaptive Potential: A Mathematical Model of Civilization Dynamics and the Fermi Paradox

Overview

The EDAP model formalizes a mechanism of civilizational collapse through intraspecific economic cannibalism — the progressive concentration of resources by elites who optimize extraction over adaptation.

The model describes the co-evolution of three variables:

• T (Technology): fraction of realizable technologies actually deployed
• K (Cannibalism): fraction of resources extracted by elites without productive contribution
• C (Cooperation): capacity for collective action and institutional trust

When K exceeds a critical threshold K_crit(T), adaptive potential dissipates, and civilization enters a feudal trap — a stable low-technology equilibrium.

Key Results

• Turn accuracy: 0.70 (train), 0.52 (test)
• Perfect prediction: Soviet Union (5/5 directional changes correct)
• Structural tension detection: Model identifies turning points that align with documented elite compensatory interventions (New Deal, Bretton Woods, QE, etc.)
• Bifurcation: System bifurcates at α_K ≈ 0.12 — below this, both sustainable development and feudal trap are possible; above, only the trap remains

Installation

git clone https://github.com/sergemso/edap_model.git
cd edap-v4
pip install -r requirements.txt

Usage

Run the full analysis pipeline

python run_model.py

This generates:

• output/fig1_phase_portrait.png — Phase portrait with historical trajectories
• output/fig2_time_series.png — Time series of T, K, C with K_crit overlay
• output/fig3_bifurcation.png — Bifurcation analysis and phase portraits
• output/fig4_monte_carlo.png — Monte Carlo forecasts (GPU-accelerated)
• output/fig5_summary_table.png — Civilizational risk assessment
• output/results.json — All numerical results
• output/table_*.tex — LaTeX tables for the paper

Run parameter calibration

python calibrate.py

Requires NVIDIA GPU with CUDA for acceleration. Falls back to CPU if unavailable.

Run tests

python -m pytest tests/test_model.py -v

Paper

The paper source is split into sections for modular editing:

paper/
├── article.tex          # Skeleton with preamble and \input{sections/...}
└── sections/
    ├── abstract.tex
    ├── introduction.tex
    ├── model.tex
    ├── results.tex
    ├── discussion.tex
    ├── conclusion.tex
    ├── appendix.tex
    └── bibliography.tex

Compile: cd paper/ && pdflatex article.tex (run three times for cross-references).

Requirements

• Python 3.10+
• NumPy, SciPy, Matplotlib, Numba
• NVIDIA CUDA + Numba CUDA (optional, for GPU acceleration)
• See requirements.txt for full list

Data

Proxy data for four civilizations are provided in data/raw/:

• Roman Empire (100 BCE – 476 CE)
• Soviet Union (1922–1991)
• Russian Federation (2000–2024)
• United States (1933–2024)

Full proxy documentation with historical sources is included in the proxy files.

Citation

If you use this model or data in your research, please cite:

@software{strebulaev2026edap,
  author       = {Strebulaev, S. A.},
  title        = {EDAP Model v4.0: Endogenous Dissipation of Adaptive Potential},
  year         = {2026},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.20298963},
  url          = {https://github.com/sergemso/edap_model}
}

License

This project is licensed under the Creative Commons Attribution 4.0 International License (CC-BY-4.0). See LICENSE for details.

Contact

Sergey A. Strebulaev — strebulaev@gmail.com