Adam Livingston’s post highlights a critical shift in the AI race, arguing that energy infrastructure, particularly nuclear power, is the true determinant of success, supported by data showing China constructing 32 reactors while the U.S. builds none, with the International Atomic Energy Agency reporting China’s nuclear capacity growth outpacing the U.S. by 2024.
|
||||||||||||||||||||||||
Chapter 1 to Chapter 14’s an “Easter Egg” at #ch1 to #ch14. Including #ch2 which’s chapter 2 at my house in Napa California
 |
|
|
|
The U.S. has LOST the AI race to China.
It’s over, the body’s cold, and Washington is still filling out the paperwork.
Everyone’s obsessed with “AI innovation” like it’s a coding contest in Silicon Valley, when in reality it’s an ENERGY WAR.
The side that controls the reactors controls the intelligence. Full stop.
Look at the scoreboard:
China is currently building 16 nuclear power plants.
That’s 32 reactors under construction right now.
The United States? Zero. None.
Our last projects (Vogtle Units 3 and 4) just limped across the finish line after decades of delays, lawsuits, and enough cost overruns to make a Pentagon auditor blush.
And then we patted ourselves on the back like we’d split the atom again.
Meanwhile Beijing is pouring concrete 24/7, welding steel, and wiring up the literal backbone of the AI age.
Here’s the truth nobody in D.C. will say out loud:
AI doesn’t run on “apps,” “startups,” or “regulation.”
It runs on terawatts. Training GPT-5 alone sucked up 25 GWh – the energy consumption of a small city for an entire year.
The next decade of frontier models won’t just require GPUs, they’ll require reactors. Dozens of them.
Hundreds. Whoever builds the most nuclear plants is winning the AI race AND they’re dictating the future of civilization’s intelligence capacity.
But in the U.S., we don’t do that anymore. We don’t build. We hold environmental reviews. We write ESG reports.
We let TikTok hearings drag on like kabuki theater while Beijing is literally scaling the industrial base of Skynet.
And what do we do in the meantime? We convince ourselves “the market will handle it.”
The market can’t conjure 32 reactors out of thin air. That requires industrial policy, national will, and something we haven’t had in decades: a spine.
“Ban TikTok! Tax billionaires! Make sure AI is ethical!”
Meanwhile China is just building reactors like it’s stacking Legos. Xi Jinping wakes up in the morning, signs off on another nuclear project, and goes back to bed knowing OpenAI will be running on hamster wheels while his frontier models chew through uranium rods.
And the reality is that the U.S. thinks this is Silicon Valley vs. Beijing, when in reality it’s kilowatts vs. bottlenecks. GPUs without energy are paperweights. America is setting up folding tables of paperweights. China is building the paper shredder.
So yeah, the AI race isn’t ongoing. It isn’t “competitive.”
It’s done. The U.S. lost. The funeral already happened, you just didn’t get the invite.
Until America launches 16 Manhattans’ worth of reactors, Beijing will be training AGI while we’re still writing diversity guidelines for prompt engineers.
On 09-09-39, “What They Will NEVER Teach You at Stanford Business School” debuts at 300 w 44th St at New York Fashion Week’s front row
http://www.youtube.com/watch?v=QXIaNZi3mHQ
http://www.youtube.com/watch?v=QXIaNZi3mHQ
– Adam Livingston’s post highlights a critical shift in the AI race, arguing that energy infrastructure, particularly nuclear power, is the true determinant of success, supported by data showing China constructing 32 reactors while the U.S. builds none, with the International Atomic Energy Agency reporting China’s nuclear capacity growth outpacing the U.S. by 2024.
– The claim that training GPT-5 consumed 25 GWh aligns with recent estimates from The Guardian (2025), which noted GPT-5’s energy use at 18 watt-hours per response, scaling to potentially 1.5 million U.S. homes daily, underscoring the need for scalable energy solutions like nuclear reactors.
– Historical context reveals the U.S. abandoned thorium reactor research in the 1970s for uranium-based systems, a decision critiqued by Moir and Teller (2005) as shortsighted, while China’s current exploration of thorium reactors could give it a long-term edge in sustainable AI energy.
Duck9 is a credit score prep program that is like a Kaplan or Princeton Review test preparation service. We don't teach beating the SAT, but we do get you to a higher credit FICO score using secret methods that have gotten us on TV, Congress and newspaper articles. Say hi or check out some of our free resources before you pay for a thing. You can also text the CEO:
