Local backlash against data centers: who will supply AI capacity

By Nate (Nate's Substack) · Sunday, November 8, 2026.

The email is a preview of a Sunday briefing from Nate's Substack on the backlash against AI data centers. According to the author, November 3 was a bad night for the expansion of AI infrastructure: in dozens of local ballot measures, voters rejected data center development. In Edgewater (Florida), unofficial results showed more than three quarters in favor of banning new data centers.

At the same time, AI use keeps growing. An Edison Research survey on SSRS in May found that 65% of U.S. adults had used an AI platform in the previous week, up from 52% in February. An October ABC News/Washington Post/Ipsos poll found that 72% opposed a new data center in their community. The author sums it up this way: people find AI useful in their lives, but reject the infrastructure proposed near them.

The central thesis is that computing capacity has to come from somewhere. If people keep using the services and companies keep buying them, someone will supply that capacity, and where it is permitted determines who gets the investment, who gets the jobs and who gets paid every time the technology is used.

The author expects the next two years to reward the companies and communities able to deliver that capacity. Developers who cannot afford a delay will have to sell, and the largest operators will be able to buy projects cheaply and face less competition from new facilities. In his view, the backlash against the AI industry is handing a good opportunity to some of its most powerful companies.

He also recounts that, after the midterm elections, a friend who works at an AI company wrote to him feeling left out of the AI race and worried about his job. Nate argues that Silicon Valley is largely to blame, because all too often the message people receive is that AI automates their work and that those who build it get rich. According to him, a vote against a data center is starting to feel like a vote for the American dream, and it is not in the industry's interest to end up on the opposite side.

The full briefing, reserved for Executive Circle members, covers: the cost of a one-year delay (with a table for a hypothetical project with $500 million in debt and the contract clauses that determine whether a small provider survives the wait); how the backlash gives Big Tech an advantage and what it means if you rent capacity from a competitor; why overbuilding is now less likely and how to tell whether a failed developer ran out of customers or out of money; where investment and jobs are going (what Canada, the EU, the United Kingdom and Singapore offer developers); and a 90-day plan to review the company's AI commitments, identify those that depend on capacity that does not yet exist and test a real alternative.

The email does not include the detailed content of those sections, only the introduction and the table of contents.