Jeff’s Note: Have you noticed?
Something strange is happening in Washington. From a distance, it all looks like chaos – the bizarre trade policies, multibillion-dollar tech deals, even talk of acquiring Greenland.
And I’d like to show you that something connects it all…
A quiet power shift is underway. It could reshape America’s economic, political, and industrial landscape from the ground up. Trillions are already on the move. Industries are being upended.
Yet hardly anyone is talking about what’s happening… or the companies involved that are surging anywhere from 200% to over 1,000%.
Just a couple of days from now, on June 11 at 2 p.m. ET, I’m joining forces with forecasting legend Porter Stansberry. Together, we’ll expose what could be the most lucrative wealth transfer in a generation.
Go here to claim your free viewing pass with just one click.
Meta Platforms (META) – Facebook’s parent company – just inked one of the biggest energy deals by a private company.
The company signed a 20-year deal with Constellation Energy (CEG) to purchase 1.1 gigawatts of nuclear power from the Clinton Clean Energy Center in Illinois.
That’s enough energy to supply the needs of about 800,000 U.S. homes.
And this single deal tells us a lot about where the AI megatrend – and the U.S. power grid – is looking for new energy.
The Clinton nuclear plant wasn’t supposed to exist any longer.
Back in 2017, it was headed for shutdown. The economics just didn’t work. Wholesale power prices were too low, and natural gas and heavily subsidized solar power had undercut the nuclear industry.
That’s when the State of Illinois stepped in and created Zero Emission Credits (ZECs). These subsidies pay nuclear plants an extra $16.50 per megawatt-hour (MWh) for producing carbon-free electricity.
The credits were meant to keep Clinton alive through 2027. After that, it was anyone’s guess whether the plant would stay in operation.
Now we have our answer.
Meta’s deal replaces those expiring state subsidies with private capital, locking in the full output of the plant for two decades. That keeps Clinton operational through at least 2047 and likely far longer. Constellation is already evaluating an expansion, with early permitting in place for another reactor at the site. And with the current regulatory environment, that’s highly likely.
At the end of May, Jeff showed readers a chart from consulting firm ICF projecting that U.S. power demand will rise 78% by 2050.
Historical and Projected Electricity Demand in U.S. | Source: ICF
Jeff said ICF “embarrassed itself” with these projections, which don’t reflect what’s coming.
We’re not just talking about smarter search engines or more realistic voice assistants. We are building the foundation for artificial general intelligence (AGI) and eventually artificial superintelligence (ASI).
These systems won’t just assist researchers and engineers… They’ll become the researchers. They’ll be embedded in government, defense, industry, and consumer applications alike.
And these systems are power-hungry.
AI training runs already rely on clusters of tens of thousands of GPUs operating 24/7. Inference demand – when AI models respond to user inputs – never sleeps. And as AI permeates every aspect of life, power demand will surge.
But even if ICF’s projections prove accurate, we’re already on the edge of a structural shift in the energy market. The United States must build far more generation capacity just to stay competitive with China, which is rapidly accelerating its grid expansion.
And here’s the hard truth: The U.S. energy grid is not ready.
The permitting system is a bureaucratic bottleneck. Building a large-scale power plant – especially a nuclear facility – is a decade-long slog. Yes, the four executive orders signed by President Trump at the end of May aim to fast-track the process.
But the U.S. is starting from behind. The U.S. has lost much of its nuclear expertise. America has only completed two nuclear reactors in the past 27 years. It’s safe to say many nuclear engineers have moved on or retired.
Source: Energy Information Administration
The reasons for this are complex. To risk oversimplifying, part of the reason for this was the permitting and approvals process. But another reason is simple economics.
For decades, nuclear power plants operated at a loss.
The Clinton Clean Energy Center – the one Meta just contracted with for 20 years – is a prime example. Despite being one of the cleanest and most reliable sources of electricity in the country, it struggled financially for years. Why?
Because nuclear doesn’t play well in volatile electricity markets.
Gas and coal plants can ramp their output up and down rather quickly. That helps them be profitable. They produce during peak demand, and they shut down or reduce output when prices dip. Nuclear, by contrast, delivers a steady, inflexible stream of power. That’s great for reliability. But it’s a disaster when prices drop or fluctuate wildly.
So for years, nuclear plants were forced to sell power below cost during low-demand hours and couldn’t capitalize on price spikes during high-demand periods. They were reliable… and consistently unprofitable.
But AI just flipped that equation on its head.
AI data centers don’t operate in bursts. They run around the clock. Training new models is a continuous process. And the businesses using AI need consistent uptime. They can’t afford outages, lulls, or load-balancing games.
That makes nuclear energy a perfect match.
Of course, this is just greenwashing Meta’s power consumption. Jeff detailed this practice before.
Meta won’t pull electricity directly from the Clinton plant. Like all power purchase agreements (PPA), this deal is about attributes, not physical delivery. Meta is buying the right to claim that it’s using clean nuclear power – even though most of its data centers are still plugged into grids that burn natural gas or coal.
This is textbook greenwashing. But don’t dismiss it as just PR. The implications are enormous.
Meta’s electricity usage has nearly tripled since 2019. And its future data centers will each require as much power as a small city. That’s why Meta is now evaluating up to 4 gigawatts of additional nuclear capacity in the U.S. And it has received proposals from more than 50 developers across 20 states.
The company is now planning new data centers that each consume enough electricity to power entire cities. It’s already reviewing proposals for an additional 1 to 4 gigawatts of new nuclear capacity.
Meta’s not alone. Microsoft, Amazon, and Google are all racing to lock up future power supply. AWS signed a long-term nuclear PPA in Pennsylvania. Google has signed agreements with Kairos Power and Elementl Power to develop nuclear sites. And Microsoft recently agreed to pay over $100/MWh in a deal with Constellation to help restart Three Mile Island. Analysts estimate Meta’s own deal is priced closer to $80/MWh – a relative bargain given the urgency.
It’s a land grab for baseload power.
And while solar and wind get the headlines, nuclear is the only technology that offers the holy grail: dense, reliable, zero-emission baseload power.
Nuclear is tailor-made for hyperscale data centers.
At first glance, it might seem like the smart move is to rush out and buy shares of Constellation Energy (CEG).
After all, Meta just signed a 20-year nuclear deal with them. Microsoft and AWS are already customers. And CEG owns more operating nuclear capacity than any company in America.
But I’d caution readers to hold off for now. We actually recommended CEG in 2024 at $186, predicting that this would happen. The stock is up 63% since then.
CEG’s stock has already doubled from its April lows in just two months. That kind of vertical move isn’t sustainable, especially when the market starts to digest the actual terms of these power purchase agreements.
Take a look at the chart below. It’s a candlestick chart. We don’t use them often, but in this case, it tells the story better than any analyst report.
Each candle represents a day’s worth of trading. Green candles show gains. Red candles show losses. And the thin wicks show price moves that happened outside of the open-to-close range and can indicate intraday volatility.
On the day of the Meta deal announcement, CEG opened up 10% at the high of the day. It was even higher in premarket action. But by the end of the day, the gains evaporated. The stock sold off. And it continued selling off the next day.
This is a classic signal. Investors expected more.
They weren’t pricing in an $80 per megawatt-hour deal. They were hoping for the kind of windfall Microsoft offered to restart Three Mile Island, at over $100/MWh.
CEG has traded to near-record high valuations on those expectations. And when those expectations weren’t met, profit-takers stepped in and sold the expensive shares.
And unless another major contract is announced, we may see CEG retrace back toward its long-term valuation range, likely 20% lower.
So what’s the better way to play this?
You invest in what these hyperscalers can’t grow AI without…
We’re heading into a massive supply crunch in uranium.
The U.S. currently consumes about 45 million pounds of uranium concentrate (U3O8) each year. If Trump’s energy plan is realized, demand could surge to 180 million pounds annually.
And the U.S. only produced 800,000 pounds in 2024.
That’s a 225x gap between domestic production and future demand.
The U.S. imports most of its uranium today, mainly from Canada, Australia, and… Russia. Russia supplies 27% of the enriched uranium used in U.S. reactors.
This trade dependency is a strategic vulnerability. The U.S. needs more secure sources of uranium. And to get that, uranium prices must go higher.
Spot uranium trades around $71 per pound today. But at those prices, it’s not profitable to bring new uranium mines online. Industry experts say it could take $150 to $200/lb to incentivize enough production to meet future demand.
Now, I don’t recommend owning individual uranium miners unless you’re an expert in that field. These companies carry high execution risk. They face long permitting delays, environmental hurdles, cost overruns, and operational risks.
Most investors are better off avoiding that minefield entirely.
One way to do that is to buy the Sprott Physical Uranium Trust (SRUUF).
SRUUF doesn’t speculate on mining. It doesn’t bet on management teams or drill results. It simply owns physical uranium in secure storage, tracking the spot price directly.
As uranium prices rise, so does SRUUF’s value. No operational risk. No development timeline. Just direct exposure to the one commodity that Big Tech, national defense, and the AI future can’t function without.
Meta’s 20-year nuclear deal isn’t just about green optics. It’s a signal.
It signals that the AI era will be powered not by solar panels, windmills, or batteries… but by a dense, baseload source of energy with zero emissions. And the world can’t scale nuclear without uranium.
The uranium bull market is just getting started.
– Nick Rokke
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
