The US is facing a power capacity crisis that's intensifying the AI race with China, according to Google's energy head, Caroline Golin. The tech industry's growing energy demands, particularly from data centers powering AI, have led to a critical reassessment of energy strategies. Traditional renewable sources like solar and wind are proving insufficient due to their intermittency, pushing major tech players like Google towards nuclear power as a stable, clean firm technology to ensure grid stability and meet surging power needs.
Google itself has turned to nuclear energy, signing a deal for 500 megawatts from Kairos Power's small modular reactors, aiming to start operations by 2030. The US government has also declared a national energy emergency to expedite power plant construction, focusing on nuclear energy as a strategic asset for grid reliability and AI development.
The power capacity crisis is a significant concern, as the US competes with China in the AI race. China's impressive comeback in AI, with models like DeepSeek R1, has made the competition more intense. The US must prioritize scaling compute power, modernizing data infrastructure, and streamlining energy production to support AI advancements and maintain its leadership.
