The US Energy Information Administration projects electricity demand will hit record highs in 2026 and 2027. AI and the data center boom are driving the surge, reversing years of flat power consumption. The scale of this buildout includes commitments of roughly 5 GW of compute. Chips get faster every year. Power grids don't. The tech community on Hacker News sees this gap as the real constraint on AI growth, with the bottleneck moving from silicon to electricity. Big tech is betting on nuclear to bridge it. Microsoft signed a 20-year deal with Constellation Energy to restart Three Mile Island Unit 1 for 835 megawatts by 2028, while Google partnered with Kairos Power for small modular reactors targeting 500 megawatts by 2035. Amazon went further, spending $650 million on a data center campus wired directly to the Susquehanna nuclear plant for 960 megawatts of immediate access. Nobody makes these deals when the grid works fine. AI's growth now depends on whether we can build power plants fast enough. However, rising power costs are already leading to statewide bans.
AI is about to hit a power wall
US power demand is projected to reach record highs in 2026-2027, driven by AI usage and data center expansion according to EIA data. Hacker News commenters highlight that AI is becoming an energy problem, not just a compute problem, with power infrastructure scaling slower than chip improvements—a potential constraint for AI growth.
