AI"s demand for electricity is expected to surge. Exporting AI services essentially means turning electricity into computing power and computing power into intelligence. Whoever has cheaper and more stable electricity, and a grid that can be dispatched faster, holds a cost advantage in the AI era. Thus, this has sparked another saying has online: "The end of electricity is China."

　　In fact, a gap emerged between the US and China in the system-level support for AI development. An examination of power infrastructure construction shows that the US power grid is fragmented into three largely independent systems — the Eastern, Western and Texas interconnections. This fragmentation creates a chronical mismatch: Power can"t be transmitted where it"s needed, nor can shortages be mitigated through dispatch. Consequently, the US lags behind in grid stability, load capacity and infrastructure development.

　　On March 4, Wang Jian, a member of Chinese People"s Political Consultative Conference (CPPCC) National Committee and academician of the Chinese Academy of Engineering, said in an interview that if the US wants to build a power plant, it has to start with the most basic transformers, and "the transformer manufacturing industry is mainly in China."

　　As a result, tech giants in the US are feeling the pressure. Microsoft, facing delays in power grid access, was forced to build its own gas turbine generator; Google signed a massive power purchase agreement with a nuclear power company. If electricity shortage persists, the resulting burden will be inevitably be borne by the public. Regional grid operators in the states of Michigan and Virginia have announced that electricity bills for 67 million citizens in their service areas will increase by 20 to 30 percent in 2026.