The AI data center boom is putting heavy strain on America’s aging power grid. Some people close to the problem think that strain could push the grid to modernize faster than anything else has. As TechCrunch AI reports, three energy leaders at TechCrunch Disrupt this month pushed back on the doom-heavy story about AI and electricity. They argued that the flood of new demand is opening doors that had stayed shut for decades.
The worries are real. Critics say the surge in grid connections is a live stress test for old infrastructure. In the short term, it could push electricity bills up. In the long term, it could mean outages. The panel didn’t deny any of that. What they changed was the framing: urgency is now a catalyst, not only a threat.
Three Bets on One Problem
The panel covered three different layers of the grid, and that’s worth looking at closely.
- Software orchestration (WeaveGrid): Founder and CEO Apoorv Bhargava says utilities are juggling “a ton of new variables,” including renewables, EVs, heat pumps, and data centers. His company’s software manages and orchestrates that load so utilities can get more out of the system they already have and keep ratepayers’ bills down.
- Physical infrastructure (Veir): CEO Tim Heidel’s startup sells superconducting power lines. The company says its cables carry power five times farther than traditional lines while using 20 times less space. Veir just closed a $110 million Series C.
- Policy and households (Rewiring America): Kristin Eberhard, VP of policy, pushes for household electrification and argues the AI crunch has made the whole industry more willing to experiment.
The common thread is that no single fix will do the job. You need smarter software, better wires, and changes at the household level all at once.
The “Yes, And” Era
Eberhard had the line of the day.
“I feel like this era of AI and load growth is a very improv era,” she said. “There’s a lot of yes-and-ing going on.”
She explained that the industry needs more transmission lines and more centralized power plants, “and we also need to make sure that we’re taking advantage of all this other stuff that we had previously been leaving on the table.”
This matters because grid modernization has been stuck for years. Utilities are slow and regulated, and they avoid risk. Ideas like load orchestration and advanced conductors have been around for a long time but rarely got deployed at scale. AI changed the math. When hyperscalers need gigawatts on a deadline, “let’s study it for five years” stops being an option.
Heidel made a similar point: “The scale of the investment in AI infrastructure is driving a tremendous amount of innovation.” He wants these technologies adopted “as fast as possible” so data centers can be “the best community neighbors and citizens they can possibly be.”
A Note on Who’s Talking
Keep the context in mind. All three speakers run or represent organizations that gain directly when grid modernization speeds up. That doesn’t make them wrong. It does mean this is an industry pitch, not an independent forecast. Veir’s performance figures are the company’s own claims, and superconducting lines still have to prove they can be deployed affordably at utility scale.
What the Next 1-3 Years Look Like
Here’s where I think this goes:
- Grid access becomes a competitive moat. AI companies that lock in power early, or bring their own efficiency tech, will scale faster than rivals waiting in interconnection queues.
- Grid tech funding keeps climbing. Veir’s $110 million round is part of a bigger shift. Investors now treat energy infrastructure as AI infrastructure.
- Flexible load becomes a bargaining chip. Data centers that can shift or curtail demand will likely get faster connections and better terms from utilities.
- Public pushback on electricity bills grows. Expect regulators to look harder at who pays for grid upgrades: hyperscalers or households.
What You Should Do Now
If you’re building or buying AI capacity, treat power as a first-order constraint, not a facilities footnote. Ask your cloud and colocation providers where their power comes from and how they plan to handle demand flexibility. If you’re an investor or founder, the energy layer under AI may be a less crowded opportunity than the model layer above it.
The grid didn’t modernize when it was merely a good idea. It may modernize now because AI made it urgent. You can find the full panel discussion at TechCrunch AI.