AI Data Centers Will Soon Consume as Much Power as Two-Thirds of All American Homes

A single ChatGPT query uses 10 times more energy than a Google search. Former CIA and Pentagon advisor Jim Rickards says the AI boom has a massive dependency most people aren’t seeing — and the U.S. power grid is already at capacity.

Washington, D.C., May 04, 2026 (GLOBE NEWSWIRE) -- Every time someone asks ChatGPT a question, it consumes 10 times more energy than a standard Google search. Scale that across hundreds of millions of daily users, and the numbers get staggering fast.

According to Boston Consulting Group, by 2030, AI data centers alone will consume the same amount of electricity currently used by roughly two-thirds of all homes in the United States.

The U.S. power grid is already operating at capacity. And the demand curve is only pointing in one direction.

Jim Rickards — an economist, best-selling author, and former advisor to the CIA, the Pentagon, and the White House — says in his recently released presentation this is the part of the AI story that almost nobody is talking about.

“We’ve never seen this kind of increase in demand before,” Rickards said. “And we’re not producing enough electricity. Our power grid is already at capacity.”

The Crisis Hiding Behind the AI Boom

The public conversation around artificial intelligence has been dominated by the companies building the models — the Nvidias, the Microsofts, the OpenAIs. What’s received far less attention is the physical infrastructure required to keep those models running.

AI doesn’t live in the cloud in any abstract sense. It lives in massive data centers filled with thousands of GPUs running around the clock, generating enormous amounts of heat, and drawing enormous amounts of power. Every query, every image generation, every model training run requires real electricity from a grid that was never built to handle this kind of load.

And the investment pouring in isn’t slowing down. Amazon, Google, Meta, and Microsoft are now collectively spending roughly $400 billion a year on AI data center infrastructure. Each new facility that comes online adds another surge of demand onto an already strained system.

Rickards’s presentation sees this as the critical bottleneck that most AI investors are ignoring — and the key to understanding where the real opportunity lies.

Why the White House Is Going Nuclear

The energy shortfall has not gone unnoticed at the federal level. Earlier this year, four separate executive orders were signed in a single day to accelerate the development and deployment of nuclear energy across the United States.

Rickards’s presentation says that’s not a coincidence. Nuclear is the only energy source that can deliver the kind of reliable, high-capacity, around-the-clock baseload power that AI data centers require. Solar and wind are intermittent. Natural gas plants take years to build. Nuclear is the only realistic path to closing the gap at the scale the AI industry demands.

And the market has already started to respond. Companies positioned at the center of the nuclear energy buildout have seen explosive growth. Oklo, a nuclear energy provider, has jumped as high as 1,000% over the past year. Bloom Energy gave investors the chance to 10x their money over the same period. And Nuscale, a maker of small modular reactors, has surged as high as 2,500% in under two years.

The Fuel Underneath the Fuel

But Rickards’s presentation says the story doesn’t stop at the reactor companies. Every new nuclear facility requires fuel — and that means uranium.

Uranium prices have already hit the highest levels since 2008 and are expected to keep climbing as the pipeline of new reactors grows. The dynamic is straightforward: more data centers mean more power demand, more power demand means more nuclear buildout, and more nuclear buildout means a sustained surge in demand for the raw material that makes it all work.

It’s a supply chain that stretches all the way from a ChatGPT prompt on someone’s laptop down to uranium deposits in the ground. And Rickards says most people haven’t connected those dots yet.

Part of a Much Larger Shift

Rickards’s presentation frames the AI energy crunch as just one piece of a broader pattern he’s been tracking. The U.S. is simultaneously pursuing a historic reshoring of manufacturing — with nearly $9 trillion in new domestic investment commitments from companies like Apple, Nvidia, Taiwan Semiconductor, and Eli Lilly — while also trying to rebuild its mining and energy infrastructure after decades of neglect.

Every new factory needs power. Every new chip fab needs raw materials. Every new defense system needs critical minerals. And after years of offshoring, the U.S. now relies on China for 100% of 15 key minerals.

“You cannot have an industrial boom without energy from coal, oil, natural gas, or nuclear power,” Rickards said. “And you cannot have an industrial boom without metals like copper, iron ore, rare earth elements, lithium, and silicon.”

He believes the convergence of AI-driven energy demand, a government-backed industrial buildout, and a generational pivot in mining and energy policy is creating what economists call a supercycle — a sustained, multi-year boom in natural resources that most investors aren’t positioned for because they’re still focused exclusively on the tech names at the surface.

The Parallel Nobody’s Paying Attention To

Rickards’s presentation points to a clear historical precedent. In the early 2000s, China launched a state-backed industrialization effort that consumed twice as much steel between 2000 and 2020 as the U.S. used during the entire 20th century. That surge in demand triggered a supercycle in natural resources that lasted years and produced extraordinary returns for investors who saw it coming.

He believes the same pattern is now unfolding in the U.S. — only with an additional accelerant that didn’t exist 20 years ago: an entire new industry, artificial intelligence, that is layering unprecedented energy demand on top of an already massive industrial expansion.

“I honestly never thought I would see another supercycle in my lifetime,” Rickards said. “But here we are.”

Adam Rozencwajg, who manages a natural resource hedge fund, shares a similar view, calling this “the best opportunity that I’ve seen probably in the 150-odd years that we’ve been studying these markets.”

Rickards has published his full analysis through his research service, which is followed by thousands of readers nationwide.

About Jim Rickards and Paradigm Press

Jim Rickards is an economist, best-selling author, and former advisor to the CIA, the Pentagon, and the White House, with a career spanning nearly five decades at the highest levels of international finance and national security. He is widely known for accurately forecasting major political and economic events — including the 2008 financial crisis, Trump’s 2016 and 2024 election victories, and the recent historic boom in precious metals. His books Currency Wars and The Death of Money have been influential across both Wall Street and the intelligence community. His research is published by Paradigm Press, an independent firm whose readers have given it a 4.8-star rating on Google across more than 1,900 public reviews.

Derek Warren
Public Relations Manager
Paradigm Press Group
Email: dwarren@paradigmpressgroup.com

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.