Key Takeaways

  • The US electric grid hasn't seen major investment since post-World War II and is teetering on collapse before factoring in AI demand.
  • Reshoring, re-industrialization, electric vehicles (EVs), and electrification for things like heat pumps will independently trigger blackouts and brownouts.
  • Most electricity price inflation comes from utility transmission and distribution costs, not the cost of generating power itself.
  • Decentralized power generation by homes and businesses cannot meet the immense industrial demands of AI factories or reshoring initiatives.
  • A single 1-gigawatt (GW) AI factory needing solar power would require 35,000 acres of panels, an area “bigger than San Francisco,” due to capacity factor limitations.

The Disagreement: Grid Collapse vs. Ground-Up Energy Independence

Dan Dreyfus of Fortnite Capital pulls no punches: the United States is staring down a grid crisis of historical proportions. He told the All-In Podcast that the electric grid has seen no meaningful upgrade or modernization since World War II. We've simply "let it go." The stakes are high, especially as the nation eyes re-industrialization and reshoring. Dreyfus warns that even without factoring in the energy appetite of AI, the combined needs of these initiatives, alongside basic electrification like heat pumps and EVs, will lead to widespread blackouts and brownouts. The culprit for rising electricity costs, he says, isn't generation. "What's really interesting I think is really underappreciated is that's where all the inflation is coming from. It's from the transmission and distribution from the utility." Craft labor, he points out, is a major bottleneck.

Jason Calacanis, hearing this bleak forecast, pushes back with a common founder's instinct: if the centralized system is broken, can't we just bypass it? He asks Dreyfus, “Is the solution to this energy independence in the home in the business? Businesses are not waiting for the government. So maybe the grid is going to be like this weird archaic infrastructure and it's just going to be a groundup solution.” This vision suggests a future where homes and businesses generate their own power, making the traditional grid less relevant.

Dreyfus, however, quickly deflates this decentralized dream when it comes to industrial scale. He explains that for a 1-gigawatt AI factory to be powered entirely by solar, it wouldn't just need 1 GW of panels. Due to solar's capacity factor, it would need 5 GW of solar capacity. He breaks down the land requirement: “Each gigawatt of solar takes up 7,000 acres. So at 5 gigawatts, that's 35,000 acres. That's bigger than San Francisco.” The sheer scale of land, materials, and labor makes Calacanis's "ground-up" solution impossible for the industrial demands facing the country.

Who's Right (and When They're Wrong)

Dreyfus is overwhelmingly right. While distributed solar on rooftops for individual homes and small businesses plays a role in decarbonization, it cannot, and will not, replace the centralized grid for industrial-scale power consumption. The numbers Dreyfus provides—35,000 acres of solar for a single AI factory—are a stark reminder of the physical constraints. This isn't a problem that can be solved by a few rooftop panels. The "ground-up" solution Calacanis suggests works for personal energy independence and small-scale resilience, but it's a fantasy for powering the factories, data centers, and advanced manufacturing facilities the US hopes to bring back.

The grid, for all its current flaws and neglect, remains the only infrastructure capable of delivering power at the scale needed for re-industrialization and the coming AI boom. Dismissing it as "archaic" ignores the fundamental physics and logistics of large-scale energy transfer.

What to Do With This

If you're a founder planning industrial-scale operations like manufacturing facilities, large data centers, or even substantial energy-intensive projects, don't assume the US grid will reliably provide cheap, consistent power. Factor potential blackouts, brownouts, and escalating electricity costs into your location and operational planning now. Seriously consider investing in significant on-site backup generation or robust energy storage solutions for your critical loads, viewing the grid as increasingly fragile. Also, look for opportunities in the immense, inevitable spending on grid modernization, energy storage, and craft labor training—the market for solutions here is about to explode.