Skip to content

DualEarth/AI_Highway

Repository files navigation

AI Infrastructure Super Highway

This repository accompanies the manuscript “AI Infrastructure Super Highway”, which examines the rapid expansion of AI infrastructure through a comparative analogy with the U.S. Interstate Highway System. The core argument is that AI infrastructure—data centers, energy systems, and network backbones—functions as a large-scale sociotechnical system that shapes culture, governance, and individual autonomy, much like highways did in the 20th century.

Using historical comparison, growth modeling, and scenario analysis, the project explores how accelerated AI infrastructure buildout may create participation defaults that make opting out of AI-mediated life increasingly impractical.


Repository Contents

Notebooks

  • highway_v_datacenter_growth.ipynb
    Compares historical U.S. highway expansion with projected AI data center growth using generalized logistic (Richards) S-curves.

  • highway_v_datacenter_transit_growth.ipynb
    Analyzes information transit growth (internet vs. data center traffic) alongside highway vehicle-miles traveled.

These notebooks reproduce the quantitative analyses and figures used in the manuscript.


Figures

  • datacenter_infrastructure_projection.png
    Comparative growth trajectories of highway infrastructure and AI data centers.

  • datacenter_transit_projection.png
    Comparison of external internet traffic and internal data center traffic growth.

These figures illustrate the paper’s central finding: AI infrastructure is expanding on a much more compressed timeline than historical transportation infrastructure.


Data

  • datacenter_projection_2036_2050.csv
    Projection data used for long-range AI infrastructure growth scenarios.

Conceptual Framework

The project frames AI infrastructure and highways as analogous sociotechnical systems along five dimensions:

  1. Scale and pervasiveness
  2. Public–private governance
  3. National security drivers
  4. Cultural and behavioral standardization
  5. Opt-out feasibility

The analysis shows that once infrastructure reaches sufficient scale and reliability, participation becomes the default and non-participation carries increasing penalties—social, economic, and political.


Key Insight

While the U.S. highway system unfolded over roughly half a century, AI infrastructure is reaching comparable societal entrenchment within a single decade. This compressed timeline narrows the window for ex-ante governance and increases the risk that long-term environmental, financial, and social liabilities will be socialized after private benefits are captured.


Intended Use

This repository is intended for:

  • Researchers studying technological forecasting and infrastructure growth
  • Policymakers and planners concerned with AI governance
  • Scholars of sociotechnical systems and public infrastructure
  • Readers seeking transparency and reproducibility for the accompanying paper

Citation

If you use this code or analysis, please cite the associated manuscript:

Frame, J. M., Rapp, J., Olaseinde, A., & Jones, S. AI Infrastructure Super Highway.
Submitted to Technological Forecasting & Social Change.


License

This project is released under the license included in the LICENSE file.

About

Analysis of the growth charactaristics of both American Highways and AI Infrastructure

Resources

License

Stars

Watchers

Forks

Packages

No packages published