AI data centers from Microsoft, Google, Meta, and Amazon are driving massive new power demand, with wholesale electricity prices in data center hubs up 267% in five years and residential bills rising faster than inflation.
Most of this AI and cloud load still runs on coal and gas, linking Big Tech’s GPU buildout directly to higher emissions, while climate-driven insurance costs and home repair expenses pile onto households.
Local grids and water systems are hitting their limits as single campuses draw city-scale power, stress aging infrastructure during extreme weather, and in some places drain groundwater needed by communities.
To keep AI growth going, utilities are extending fossil plants and planning new gas plants, and tech vendors are now backing nuclear restarts like Three Mile Island to feed Microsoft data center demand.
The author argues Big Tech and the current administration are socializing the real costs of AI infrastructure—energy prices, climate risk, water, noise, health—while privatizing the upside.
The proposed fix is strict siting and transparency rules plus real 100% renewable commitments for hyperscale and AI workloads, rather than leaning on fossil and nuclear as a backstop.
Worth reading for a grounded view of how the AI data center race is reshaping power markets, climate risk, and community infrastructure.
Source: The public pays the price for Big Tech’s data centers | The Invading Sea
