This article explores how AI and fossil fuels create valuable powered land in a new state capitalism framework, suggesting significant implications for infrastructure planning.
The so-called AI boom is part of an infrastructure-led industrial strategy, converting speculative computing demand into bankable fossil generation, transmission expansion, and water- and land-intensive industrial sites. This article bridges digital infrastructure studies and the new state capitalism literature by theorising compute as a socially produced resource whose availability depends on territorial governance. I argue that the buildout is being assembled through a fossil–AI nexus, a fossil–finance–platform coalition that produces powered land: an emergent asset form whose value derives from positionality in a constrained energy system that secures deliverable ‘firm’ power through revenue guarantees, deliverability rights, and cost-allocation arrangements. Drawing on a review of major gas-to-data centre co-location projects and a comparative analysis of PJM and ERCOT, I identify three recurring de-risking channels that convert uncertain load forecasts into durable, carbon-intensive infrastructure: revenue certainty, delivery certainty, and cost shifting. I show how ‘reliability,’ alongside security and competitiveness framings, compresses timelines, translates engineering criteria into bankability, and narrows public contestation. I also show how opacity, or ‘blackboxing,’ stabilises powered land by restricting access to contractual and cost-allocation terms, while relocating politics to transparency disputes and siting conflicts over where data centres go and who pays for the buildout.
No takes yet. Share an insight, caveat, or question.
Justin Kollar (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: