Loading...
AI’s Energy Boom Is Leaving Clean Power Behind - As AI drives record energy demand, billions are flowing into gas plants while federal cuts sideline solar, wind, and storage—risking higher costs, lower reliability, and slower decarbonization.
Blackstone wants to pair gas plants with data centers. Microsoft is trying to revive a nuclear reactor. Trump’s budget cuts are sidelining solar and storage.
This is the direction the energy conversation is heading, just as AI ramps up demand like never before.
The problem is that our energy grid is complex. It wasn’t designed to meet real-time load from hyperscale data centers while transitioning off fossil fuels. And now, we’re making 20th-century decisions for 21st-century needs.
Talen Energy just spent $3.8B to acquire two massive gas plants to feed AI growth. The CEO called it “the fastest and cheapest way” to power the boom. Independent generators have now announced more than $33B in similar deals, racing to lock in fossil generation.
It’s easy to see why. These plants deliver dispatchable electrons, exactly what data centers need. But those electrons aren’t unique. Solar, wind, and storage generate them too. And they can do it faster, cheaper, and cleaner.
So why are we sidelining the cleanest, fastest-growing parts of the grid?
Because federal cuts are gutting clean energy momentum
The Rhodium Group says Trump’s budget law will eliminate about 60% of the clean power capacity we would have built over the next 10 years.
At the same time, aging fossil plants are being kept online, even when experts say they aren’t needed.
AI and clean energy should be natural partners. Fast dispatchable energy for an energy-hungry sector. Instead, they’re being forced apart.
There’s another cost here that no one’s talking about:
What happens when utilities start redirecting power from the grid to private data centers? That power doesn’t get replaced easily. We all pay the price through higher rates, less reliability, capacity, and flexibility, and delayed decarbonization.
Are we really ready to take resources offline from the grid to feed AI and leave the rest of the country behind?
To be clear: We need all options on the table. Gas, solar, wind, nuclear, storage, every arrow in the quiver. But cutting out the cheapest, most scalable, faster-deploying sectors at the moment we need them most? That’s not pragmatism. It’s self-sabotage.
We’re entering a new demand paradigm. The question is: Will our energy system evolve fast enough to meet it?
Author: Julian Spector