Key Points
- Nvidia’s CEO warned China could win the AI race due to its cheap and abundant energy.
- China generates twice as much electricity as the U.S. and is rapidly expanding renewable energy capacity.
- Much of China’s new power is needed for its broader economy, not just for AI, and substantial amounts of renewable energy are wasted.
- China’s AI sector faces destructive internal competition (“involution”) and overcapacity, with many companies losing money.
From the outside, China’s progress in artificial intelligence appears unstoppable. Even top tech leaders like Nvidia CEO Jensen Huang have warned that China could win the AI race thanks to its cheap, plentiful power for data centers, especially when compared to the aging and overwhelmed U.S. grid.
The numbers certainly look scary for America. China generates more than twice as much electricity as the United States and is in the midst of a massive green energy construction boom. By 2030, the country’s renewable energy output alone will be sufficient to power all of its data centers. This growing “electron gap” has many in the U.S. worried that they are falling behind.
However, a closer look shows this panic is probably overblown. Despite its energy advantage, China remains lagging in the construction of new data centers. This suggests that U.S. restrictions on advanced AI chips remain a major impediment for Beijing.
Furthermore, China’s huge power build-out isn’t just for AI. The country’s demand for electricity is increasing as it transitions to electric vehicles and moves its factories away from fossil fuels. Data centers are expected to use only about 3% of the country’s total power consumption by 2030.
China also wastes a lot of its green energy because most of it is generated in remote western regions, far from the eastern cities where it is needed most. A plan to move data centers west to be closer to the power source has largely failed due to slow data transfer speeds.
Perhaps the biggest problem is internal. China’s AI industry is suffering from what experts call “involution”—a cycle of destructive competition, overcapacity, and brutal price wars.
Startups are incurring losses as they race to the bottom on prices. This internal chaos and massive misallocation of money may be a bigger threat to China’s AI ambitions than America’s power problems.
