Factories Then, Data Centers Now
An EACOE™ Position Paper: What the Industrial Revolution Teaches Enterprise Architects About AI Infrastructure, Governance, and the Next Generation of Practitioners
The fight over AI data centers repeats the fight over factories in many ways. In both cases, national leaders called the build-out a matter of national security and economic power. Local communities saw new costs, a loss of local character, and outside capital they could not control. Workers feared for their livelihoods, and governments had to choose between suppressing the opposition and building rules around the new system.
One structural difference changes the politics. In the industrial era, the factory was both the disruption and the job. It employed thousands, pulled families into towns, and caused the harm in the same place. Today the building and the labor disruption are separate. A data center brings large local costs (power, water, land, tax breaks) and few permanent jobs. The AI it runs may change work everywhere else, starting with entry-level roles. The community that carries the cost is usually not the community that gets the benefit. That is the main reason the backlash is so intense and so bipartisan.
The historical record shows how this tends to resolve. Repression did not work. Rules that made promises without enforcement did not work. Rules that worked had four features: they priced the harm to the builder, set up inspectors with real powers, limited public money for private ventures, and protected the young people coming into the workforce.
For enterprise architects, the Enterprise Architecture Center Of Excellence (EACOE™) sees the AI build-out as an Enterprise Architecture failure at national scale. Infrastructure is being committed before requirements are defined. Decision authority is contested between Washington, statehouses, and townships. The costs are not traced to the decisions that create them. The same three failures are appearing inside enterprises that are committing to AI platforms, compute contracts, and workforce changes. The remedy is the one EACOE™ has argued for years: requirements before architecture, governance authority before controls, and strategy-to-outcome traceability for every major decision.
