The intensifying AI competition between the U.S. and China is increasingly defined not just by technological prowess but by access to vast energy resources. Amidst growing hype and investment in artificial intelligence, a provocative claim suggests the U.S. has already lost this critical race, with China reportedly gaining an insurmountable lead by quietly cornering the most vital resource for AI advancement: energy, particularly nuclear power.
China's Nuclear Ambition and U.S. Response
Adam Livingston, author of "The Bitcoin Age," posits that China has pulled far ahead, not through superior coding but by its aggressive nuclear energy buildout. Indeed, late 2025 data shows China with approximately 30 nuclear reactors under construction, making up nearly half of the world's new builds and aiming for 65 gigawatts of capacity by year-end, potentially reaching 200 gigawatts by 2040. In stark contrast, the U.S. has no new large-scale nuclear projects underway, having only recently completed its Vogtle 3 and 4 reactors after significant delays. While recent U.S. policy shifts have spurred Westinghouse to announce plans for 10 new reactors by 2030, these projects face substantial regulatory hurdles and public skepticism, making their execution far from guaranteed.
Energy as the AI Race's Bottleneck
The argument for energy as the decisive factor stems from AI's insatiable demand for electricity. Training advanced models like GPT-4 requires tens of megawatts, and U.S. data center power demand is projected to more than double over the next decade, reaching 78 gigawatts by 2035. Globally, data center energy consumption is expected to double by 2030, with AI consuming an ever-larger share. Therefore, nations capable of deploying abundant, stable, and carbon-free power sources gain a significant competitive edge. China's top-down industrial policy has enabled rapid nuclear expansion, whereas the U.S. has relied more on upgrades and market-based activities, though it is now exploring Small Modular Reactors (SMRs) and efficiency improvements. While China's nuclear expansion is impressive and undeniably linked to AI infrastructure needs, declaring the U.S. defeat premature is essential. The U.S. retains formidable strengths in foundational AI research, chip design, cloud infrastructure, and venture funding. Although energy may emerge as a critical constraint, American innovation in efficiency, smart grids, and distributed computing could still mitigate China's lead. The "energy wars" are becoming as significant as software and data in the global tech struggle, but the final outcome of the AI race will depend on a much broader array of factors than nuclear capacity alone.
