🌐 The Grid's Next Move: Nuclear Daydreams vs. Distributed Energy Reality

2025 has been a fascinating year to cover innovators scaling climate solutions across markets, industries, and cities.

Amid political headwinds from Washington, where electric vehicle tax credits were sunset and federal support for solar and wind was curtailed, the shift to a low-carbon future marches onward.

This is a new moment: prices for clean energy technologies are plummeting. Deployments are scaling. New business models are making climate solutions more affordable and more profitable. Every major business and technology story, including AI, is now a climate story.

Making sense of all the forces currently at play is exciting, confusing, terrifying, and hope-inducing all at once.

Normally, this newsletter covers climate innovators accelerating the low-carbon economy, diving deep into their products, technologies, and market strategies.

We’re taking a slight departure for the next few weeks. We’re engaging three of the most prominent observers and voices in climate and the clean energy transition to help us understand what transpired in 2025 and what’s in store for 2026.

Week 1: David Roberts, host and writer of the Volts podcast and newsletter.

Week 2: Akshat Rathi, reporter for Bloomberg, host of the Zero: The Climate Race podcast, and author of Climate Capitalism: Winning the Race to Zero Emissions.

Week 3: Jigar Shah, Former Director of the U.S. DOE Loan Program Office, author of Creating Climate Wealth, and Co-Managing Partner of Multiplier.

David Roberts has covered climate and the energy transition for two decades. His beat takes him to Capitol Hill, where he interviews politicians like Arizona Senator Ruben Gallego (that’s his guest this week), into the venture world, where he interviews investors like Clay Dumas of Lowercarbon Capital, and the innovation space, talking with leaders like Colin Campbell, CTO of Redwood Materials, a giga-scale EV battery recycler.

David has likely written more words than anyone about the battles, triumphs, setbacks, and (in the best way) minutiae underpinning a world constantly wrestling with itself to build a more prosperous, just, and clean future.

Here’s what leapt out from our conversation:

Nuclear Daydreams vs. Distributed Energy Reality

I asked David for his one-sentence headline for 2025. His response was immediate:

“Data centers prompt long overdue discussions and changes in the energy system.”

Not climate policy. Not solar or battery deployments.

Data centers. 

Specifically, the newest, most energy-intensive—now referred to as AI factories.

For twenty years, U.S. electricity demand was flat. Adding more solar or wind power usually meant replacing fossil fuels. The fights were about coal-plant retirements—what get shut down, when, and what replaces it.

Then AI arrived, and the energy demand story totally changed.

“The projections are nuts,” David said. “Like the projections are crazy.”

Growth in energy demand is only part of it. The bigger issue is uncertainty. As David put it, the spread between the low and high projections is so wide that “the entire energy demand of the UK can fit” inside it.

Utilities aren’t built for that variance. Grid systems plan around energy peaks and reliability. They’re designed for predictability. Utilities are where “boring people go to be boring,” David quipped.

Now factor in the hyperscalers.

Amazon, Microsoft, Google, and Meta are building massive AI factories reaching gigawatt-scale. But in many places, they can’t connect.

“They can’t get interconnected to the grid because there’s not enough grid capacity to let them on… they’re losing millions of dollars a day just sitting there not computing,” said David.

This isn’t red tape. It’s physics. Utility executives believe they can’t reliably serve this energy load at peak demand without risking the entire system.

And once peak capacity becomes the constraint, the default response becomes: we need to build more supply. Which is why nuclear has come roaring back into the conversation.

“I think they think that a big nuclear power plant is the sort of masculine way to do this,” David said. “They’re dreaming about nuclear power. They’re spending a bunch of money on it. To which I say, Godspeed. I’ve been at this a while. I’ve seen several cycles of nuclear hope crash on the shores.”

His critique isn’t moral. It’s temporal.

Nuclear takes 10+ years to build. Gas takes 3 to 5 years. Even utility-scale solar and storage can take years to complete once you add in permitting and interconnection queues.

But the data center energy challenge is right now.

Again, if you’re a utility executive staring at data centers, it’s easy to jump straight to: we need to build a ton of new power plants.

But here’s the utterly bizarre reality that we non-grid people don’t realize: the U.S. power grid operates at under 50% utilization nearly all of the time.

That’s because it’s built to manage rare peak-demand spikes. Yet, most of the time, much of the infrastructure we’ve already paid for sits idle.

“Obviously, the cheapest and first thing we should do is utilize the grid we have more fully,” said David.

That means optimizing HVAC systems, coordinating EV charging, and putting batteries in households, businesses, and distribution nodes—tools that can all be orchestrated to show up as usable energy capacity when the grid is tight.

In other words, treat plugged-in, electric-powered devices just like energy capacity.

David draws an analogy to heat pumps:

“Heat pumps produce more heat than the energy you put in. They have a coefficient of performance of like three, which means for every unit of energy you put in, you get three units of energy out. How do they do that? They don’t generate heat. They just go find it. The heat already exists out in the ground. They just go round it up, and it’s a lot easier to round things up than it is to generate things.”

Same idea with grid capacity:

“Similarly with grid capacity, it’s easier and faster just to go round up the grid capacity that already exists than it is to generate any new power, almost by definition.”

The question now is: who will pay to round up the grid capacity that already exists?

To David, the logical answer points to one set of players: the hyperscalers.

Distributed energy in the U.S. is a consumer-driven market. One household buying one solar plus storage system at a time. One EV charger in a garage. One smart thermostat in a hallway.

The outcome is gradual adoption, which is a sub-optimal way to build urgent infrastructure.

As David puts it, “We are currently trying to build energy infrastructure by knocking on one door at a time.”

David points to a striking possibility: using just 3% of the capital flowing into AI data centers to build distributed energy capacity could create roughly 100 gigawatts of virtual power plants.

A virtual power plant is coordination at scale: thousands of batteries, buildings, EV chargers, and flexible loads operating together so they can behave like capacity when the grid is tight, enabling AI data centers to connect to the grid now.

The irony: AI data centers may finally solve the distributed energy problem that's been waiting two decades for funding. The hyperscalers just need to realize they're the solution to their own problem.

As David put it: 'This is a win-win-win that is just waiting to happen.'"

This Week’s Podcast Episode

The Grid's Next Move: Nuclear Daydreams vs. Distributed Energy Reality

🎙️ Listen on Apple, Spotify, YouTube, and all other platforms.

↓

Stat of the Week: 100

The number of gigawatts of near-term energy capacity that could be unlocked in the U.S. through virtual power plants, according to a widely cited recent study from researchers at Duke University.

Quote of the Week:

↓

🌐 The Climate Adoption Playbook

Climate innovation is half the battle. The other half? Getting customers to adopt it.

The good news is that successful climate companies have figured it out. The great news is they all do the same thing.

They pull from a set of five adoption levers. Sometimes they use just one. Often, they combine them to create even stronger value propositions. But, no matter the market or industry, their entire playbook consists of these levers:

1.  🧩 Friction Removal (Make it easier)

2. ✅ De-Risked Adoption (Make it credible)

3. 🎨 Lifestyle Upgrade (Make it aspirational)

4. 💳 Financing as the Unlock (Make it affordable)

5. ♻️ Circular Advantage (Make it profitable)

The Climate Adoption Playbook teaches you how to use these levers too. It’s the first course built on real-world case studies of companies that have gone from pilots to commercial scale to meaningful market share, now shaping the low-carbon future.

To make it easier to dive in during the holidays—and as you gear up for the year ahead—we’re offering a special holiday price.

Now through January 15, 2026, the course is discounted from $499 to $199, with ongoing access and updates for the next year included.

👉 Learn more and sign up for the Climate Adoption Playbook.

↓

Related

At Supercool, we’ve been tracking grid innovation from multiple angles in 2025: the home, the utility control room, long-duration flexibility, and the infrastructure that connects it all. Here are five Supercool conversations to go deeper:

⚡ Span (podcast, newsletter)
Span rethinks the grid starting inside the home. By turning the electrical panel into an intelligent control layer, Span makes electrification manageable—shifting loads in real time so homes can run on clean power without overloading the system.

🏗️ Siemens Energy (podcast, newsletter)
Siemens Energy operates at the opposite end of the spectrum: the backbone of the grid itself. From high-voltage transmission to grid-scale digital controls, they’re focused on upgrading legacy infrastructure so it can handle renewables, electrification, and volatility at scale.

🔋 Enphase (podcast, newsletter)
Enphase shows how millions of small assets can behave like one coordinated system. Its microinverters and software platform turn rooftop solar and home batteries into flexible grid resources—improving reliability while giving households more control.

🌍 Wärtsilä (podcast, newsletter)
Wärtsilä tackles the grid’s hardest problem: variability. Its flexible power plants and energy storage systems are designed to respond instantly to swings in supply and demand, helping grids integrate renewables without sacrificing stability.

🧠 Camus (podcast, newsletter)
Camus focuses on the intelligence layer utilities have been missing. Its software helps grid operators plan, simulate, and run distribution networks built for two-way power flows—critical as EVs, heat pumps, and distributed energy flood the system.

↓

Where Supercool Traveled This Week

PODCASTS

Snowpal (above): Can AI Save The Planet Without Burning it?

One Bite is Everything: Climate Change as a Blueprint for Systems Change — with Josh Dorfman

MY DAD’S HOUSE

In 2018, I co-founded Simbly, a direct-to-consumer furniture company manufacturing sustainable, flat-pack furniture just outside of Asheville, North Carolina.

This past weekend, while visiting my dad, who lives there, I encountered one of our tables still in use.

The design brief was straightforward but challenging, bordering on impossible:

1. Good design

2. Sustainable materials

3. Domestic manufacturing

4. Accessible price points.

For a while, we pulled it off until COVID eventually forced us to shut down production. But that experience led to Plantd and a burning desire to bring carbon-negative materials to market. I wrote about it on LinkedIn.

↓

Interested in Advertising with Supercool?

Connect with future-forward decision-makers seeking next-gen climate innovations. Reach out to discuss how Supercool’s platform can help. Just hit reply to this email.

↓

Not yet subscribed to Supercool?

Click the button below for weekly updates on real-world climate solutions that cut carbon, boost the bottom line, and improve modern life.

🌐