Can electrical grids, water systems, land, and local communities accommodate rapidly growing compute demand?
The IEA projects data-center electricity demand roughly doubling by 2030, with AI the main driver. Siting, water, and grid interconnection queues are where the abstraction meets local politics.
View on the map → · Open in Browse →
The backlash became measurable: 75 data-center projects worth ~$130 billion were blocked or delayed in one quarter, opposition groups doubled to 833, and the rate-payer question got real analysis — PJM's capacity-price spike traces largely to forecast-driven market design. Masley's forensic work on the viral water numbers cuts the other way.Evidence: 123
PJM's 9.3× capacity-price jump traces largely to forecast-driven market design; ERCOT absorbed similar growth with 11–17% rises.
At least 75 data center projects worth approximately $130 billion were blocked or delayed in a single quarter.
Community opposition doubled to 833 groups across 49 states; 300+ state bills in six weeks.
Expedited grid interconnection for large loads conditional on binding cybersecurity hardening.
Forensic reconstruction of the viral 500mL-per-prompt figure: stacked errors inflating per-prompt water use ~50–250×.
The conditional data-center construction ban, and whether safety advocates, local opponents, and labor hold together as a coalition.