Somewhere in Loudoun County, Virginia, a retired schoolteacher just opened her electric bill.

It was higher than last year.

She doesn’t know why.

She has no idea that her neighborhood sits at the center of what analysts are calling the most important digital infrastructure hub on the planet.

That a few miles from her house, 199 data centers hum around the clock.

With 117 more on the way, according to Loudoun County’s own board of supervisors.

That the utility serving her county – Dominion Energy – recently received approval for a significant rate increase, driven in part by the surging electricity demands of artificial intelligence.

She just knows her bill went up.

The scenario is illustrative – but the math is real. According to Virginia’s State Corporation Commission, the average Dominion Energy residential bill rose by $11.24 per month in 2026, driven in part by surging electricity demand from data centers and AI expansion.

This is what the AI boom looks like from the ground—

not the stock charts…

not the earnings calls…

not the breathless headlines about trillion-dollar valuations.

As Mason Sexton, editor of Money & Power and a veteran markets analyst, recently put it:

“No matter how sophisticated the software becomes, every AI query ultimately comes down to the physical reality that electricity has to come from somewhere.”

A retired woman in northern Virginia just found out where some of that somewhere is.

And multiply AI’s insatiable need across millions of households – and across a grid that was never designed for what’s being asked of it – and a much larger picture begins to emerge.

For the last two years, investors have focused on the companies building artificial intelligence.

Chipmakers.

Semiconductor manufacturers.

Cloud infrastructure providers.

Those trends created some of the market’s biggest winners.

But Sexton believes the conversation may be starting to shift.

“We’ve spent the last two years talking about artificial intelligence, but now we’re probably going to spend the next five talking about electricity.”

A modern AI data center doesn’t consume electricity the way your home does.

It consumes electricity the way a small city does.

The latest facilities require hundreds of megawatts of power. Some planned projects are expected to demand more electricity than entire metropolitan areas consumed just a generation ago, according to Sexton’s recent research.

And the American grid, quite simply, wasn’t built for this.

The problem isn’t confined to Virginia.

It’s happening in Texas. Arizona. Utah. Across the Midwest.

“While the winners of the first phase of the AI boom built the computers,” Sexton writes, “the winners of the next phase will power them.”

The story started with chips.

It may continue with transformers, transmission lines, power generation, and the electric grid.

Which raises an important question:

Are investors thinking broadly enough about how AI could affect the economy beyond the technology sector?

Chuck Carlson has spent more than 40 years studying markets. His work has appeared in The Wall Street Journal, Barron’s, and Kiplinger’s, and he contributes to one of the longest-running independent market research letters in the country, Dow Theory Forecasts.

As CEO of WorthNet partner adviser Horizon Investment Services, Carlson has been watching the AI infrastructure trade closely – and his perspective may surprise investors who think AI exposure begins and ends with chipmakers and hyperscalers…

“A lot of AI infrastructure stocks have done gangbusters over the last 24 months,” Carlson notes.

He points to names his firm identified early – including Comfort Systems USA (FIX), a mechanical and electrical contractor whose deep exposure to data center construction helped drive substantial gains over a two-year stretch. And Powell Industries (POWL), an electrical equipment manufacturer that many investors have never heard of – and that has more than tripled over the past year as demand for power distribution infrastructure tied to AI has surged.

Caterpillar (CAT), a company many investors still think of as a heavy equipment maker, tells a similar story even if Carlson didn’t recommend it directly. Its stock has more than doubled over the past year – driven in meaningful part, Carlson says, by its power generation business supplying generators and turbines to AI data centers.

“Probably the strongest industrial stocks have direct ties to the infrastructure trade,” he points out.

The lesson, for Carlson, isn’t that the trade is over; it’s that it was broader than most investors recognized.

“You can probably go through virtually any sector and find stocks that are linked – sometimes heavily – to the AI and AI infrastructure buildout.”

Utilities getting re-rated on power demand. Energy companies signing new deals tied to data center electricity needs. Communications and consumer names whose underlying businesses run on the same AI infrastructure everyone else is building.

His counter-intuitive observation: many investors probably already have more AI exposure than they think – not because they bought chipmakers, but because they own companies quietly benefiting from the same buildout.

“Investors are looking at their own AI exposure too narrowly in that they probably own stocks in non-tech industries that are getting re-rated because of the AI infrastructure trade. In other words, I think most people probably possess more AI exposure than they think in their portfolios.”

So where does that leave investors looking for new opportunities?

Carlson still sees appeal in what he calls the “picks and shovels” side of AI infrastructure – the contractors, equipment suppliers, and service providers who build and maintain what AI requires to function.

Two reasons stand out:

• First, many of them remain under the radar. “A lot of them are not household names,” he says – which means they may not yet carry the valuation premiums attached to the obvious AI plays.
• Second, they’re structurally insulated from one of the risks hanging over the hyperscalers: price competition among model providers. “You still need data centers to run all of this stuff,” Carlson notes, “even if it’s running cheaper models.” Demand for the physical layer of AI may prove more durable than demand for any particular software running on top of it.

He’s candid that the picks-and-shovels group isn’t without risk.

Permitting, bottlenecks, labor constraints, community opposition – the infrastructure buildout faces real-world friction that stock charts don’t always reflect.

But for investors asking how to think about AI beyond the obvious names, the answer may be hiding in plain sight:

The industrial sector…

Utilities…

The companies pouring concrete and pulling wire so that somewhere in Loudoun County, a data center can stay on…

If you’d like to learn how advisers like Chuck Carlson think about AI exposure, diversification, and positioning portfolios around major long-term themes, you can complete WorthNet’s brief, 90-second questionnaire by clicking below to see whether a conversation with one of the advisers in our curated network – including Carlson – may make sense for your goals and interests.

WorthNet does not provide advisory services and is not a client of Horizon Investment Services or the other advisers in our network. In full transparency, we’re compensated for promoting Horizon Investment Services and therefore have a financial incentive to recommend them, which creates a material conflict of interest.

Charles B. Carlson, CFA®

CEO & Portfolio Manager of Horizon Investment Services

Investment Visionary, Seasoned Asset Allocator

Charles B. Carlson is a veteran investment adviser with over 25 years of experience in retirement planning, asset allocation, and portfolio management. He is CEO of Horizon Investment Services (CRD #110642), a proud member of the WorthNet partner adviser network.

Ready to Connect with a Financial Adviser?

Let’s Get Started.

Find Your Adviser »

Last Revised: August 20, 2026