U.S. Energy Secretary Discusses Nuclear Power for AI Data Centers
U.S. Energy Secretary Wright and General Matter founder Scott Nolan were interviewed by Ti Morse, focusing on how nuclear power can support the large-scale expansion of AI data centers.
The interview addressed energy bottlenecks in the next five years, using natural gas as a transition before commercializing SMRs, the potential underestimation of AI's electricity demand, and the need to build hundreds of gigawatts of base-load power generation capacity by 2050.
Both emphasized adopting a SpaceX-style rapid iteration strategy, encouraging founders to build factories early, and stated, "If you want to build large projects in the U.S., come collaborate with the government."
Source: Public Information
ABAB AI Insight
The U.S. Department of Energy has previously promoted advanced nuclear energy development, and this dialogue with Scott Nolan continues the Trump administration's strategic path of bridging AI computing power needs with nuclear and natural gas. The rapid iteration model, akin to SpaceX's early experiments leading to frequent launches, is explicitly referenced in the energy sector.
On the capital front, the government is mobilizing private capital towards nuclear and gas projects through policy incentives and early permitting, motivated by the need to address power bottlenecks before the explosion of AI data centers, strategically avoiding China's lead in nuclear and AI infrastructure, while providing reliable power for hyperscalers' future trillions in spending.
Similar to the rapid construction of nuclear power infrastructure during the Cold War, the current phase is transitioning from traditional fossil fuels to nuclear and gas, ultimately moving towards large-scale commercialization of SMRs.
This is fundamentally a restructuring of the industrial chain. The explosive growth of AI computing power is changing the pricing power structure of energy supply, with the mechanism being the government's clear support for an "early factory construction" model, prompting capital to shift from slowly approved traditional energy to high-execution, SpaceX-style nuclear projects, achieving a structural shift in U.S. power infrastructure from reactive responses to proactive, advanced construction.
ABAB News · Law of Cognition
The stronger the AI computing power, the more proactive energy construction must be.
True great power competition is never about the size of the model, but who can supply the electricity first.
When the government says, "If you want to build large projects, come to me," the window of opportunity has opened.