Oil painting of a small town at dusk where residents use phones and laptops, beside a half-built data centre behind a fence with idle cranes

Quick answer

No, the AI boom is not over. The AI data centre backlash is slowing where AI gets built, not how much people use it. In his latest video, Nate B Jones argues that local votes against new data centres on 3 November 2026 will shift computing power towards big, established providers, push some investment overseas and could even lower the risk of an AI bubble. Demand keeps climbing: he quotes Google saying API usage is up 40% in a quarter. For a small business, the practical effect is likely to be fewer suppliers and possible price changes, so it is worth having a backup plan and a clear, honest way to talk about your own AI use.

Key takeaways

  • Use and opposition exist side by side. Nate quotes polls showing 65% of US adults used an AI chatbot in a week, while 72% opposed a new data centre in their own community.
  • Demand is still rising, so blocked projects move elsewhere rather than disappear.
  • Smaller AI clouds are most at risk from delays; big operators can afford to wait.
  • Slower building may cut bubble risk by giving demand time to catch up with supply.
  • Some investment and jobs may move abroad, with the service then rented back.
  • For you: expect possible price and supplier changes, not AI disappearing.

If you have seen headlines about voters rejecting AI data centres and wondered whether the AI boom is running out of road, you are not alone. Nate B Jones, who runs the Nate B Jones YouTube channel and an AI strategy newsletter, took this question on directly in a video called “Is the AI Boom Over? Let’s Talk About It.”

His focus is not the politics itself but the second and third order effects: what happens when the computing power behind AI gets harder to build. Here is our plain-English summary, with his claims clearly attributed and our own practical steps for small businesses and creators at the end.

Watch: the source video

What this video covers: Nate B Jones explains why US votes against AI data centres are unlikely to slow AI use, and how they could reshape who supplies computing power, where it is built and the odds of an AI bubble.

Video: Nate B Jones (@NateBJones). No paid sponsor; Nate mentions his own newsletter and prompt library. This summary and the practical advice in it are our own. Poll figures and forecasts are as quoted in the video.

What happened to AI data centres on 3 November?

According to the video, many US communities voted on 3 November 2026 to block or restrict new data centres, often by large margins. Nate gives Edgewater, Florida as an example, where unofficial results showed more than three quarters of voters backing a ban on new data centres. He says margins like that were typical across the country.

He also points to Texas, which he says was already tightening permits before the election and has now elected a governor who has spoken out about data centres. That matters because Texas has been a favourite place to build quickly and cheaply.

💡 Tip: Local election results and policy details change quickly. If a specific state or site matters to you, check the official result or council decision rather than relying on a video or social post.

Why do people use AI but oppose data centres?

Because using AI and hosting the buildings that power it feel like two very different things. Nate quotes two surveys to show the gap:

Survey (as quoted in the video)Finding
Edison Research / SSRS, February 202652% of US adults had used an AI chatbot in the previous week
Edison Research / SSRS, May 202665% had used an AI chatbot in the previous week
ABC News / Washington Post / Ipsos, October 202672% opposed a new data centre in their community

In his words, many people seem to expect AI to arrive “out of the air”. It does not. Every chatbot answer runs on chips in a building that needs land, power and often water. He is also frank that the tech industry has done a poor job of explaining why AI is good for ordinary people, and that talk of extinction risk has not helped.

Is AI demand actually slowing down?

No. According to the figures Nate quotes, demand is still growing fast. He says Google reported on its latest earnings call that usage through its model APIs was up 40% quarter on quarter and more than four times what it was a year earlier. Google also said it had started using outside computing capacity because its own cloud was short of supply.

Put those together and you get his core point: more people are using AI, while it is getting harder to build the infrastructure in the places companies planned. That combination changes where the money goes, which companies thrive and what becomes valuable over the next couple of years.

Who loses when data centres are delayed?

Smaller specialist AI cloud companies, often called neoclouds, are the most exposed. These firms rent out powerful AI chips. Their business depends on getting those chips installed, powered and earning money on schedule.

Nate walks through the risk. A neocloud borrows money, buys equipment and signs a customer who expects capacity by a certain date. If approvals or power connections are delayed, the revenue slips, but the loan payments keep coming and the equipment keeps losing value. The company may have to buy capacity from a rival just to keep its customer, which eats its margins.

⚠️ Warning: Nate’s point is that a company can fail even when demand for its product is strong, simply because it cannot afford to wait. A building under construction is not the same as one that is switched on.

Could the backlash make Big Tech stronger?

Nate thinks it could, which is one of the most surprising parts of his argument. When a smaller developer runs out of money, the project does not vanish. A buyer with cash and existing data centres can pick up equipment, contracts or half-built sites at a discount, and serve customers from its other facilities in the meantime.

Companies that already run working data centres also gain. With fewer new competitors coming online, their existing capacity stays busier and they gain pricing power. So a very public political defeat could leave the industry more concentrated around the biggest players. For smaller AI companies, that may mean renting more capacity from the same giants they compete with.

Does slower building make an AI bubble less likely?

Possibly, yes. Nate agrees with an argument made by writer Derek Thompson after the election: if demand keeps growing while new supply is slowed down, demand has more time to catch up.

He explains the bubble risk like this. Many developers look at the same growth forecast and each decides to build. Every project looks sensible alone, but together they build far more than the market needs, leaving expensive buildings without enough paying customers. Nate is clear that this had not happened yet, but it was a worry for 2027.

So moderating the US build-out could reduce the chance of a glut over the next year or two, even while it causes real pain for individual developers. He is also careful to say he is not giving advice on what any stock or company is worth, and neither are we. For more on how the market is weighing AI, see our piece on the Anthropic IPO as an AI bubble test.

Where does blocked AI demand go next?

It goes wherever capacity can be delivered on time. Nate says a company that needs more computing power next year will not wait for the next legislative session. It will look elsewhere:

  • Other US locations with reliable power, water and supportive communities, even if land costs more.
  • Other cloud providers that already have capacity.
  • Other countries. He mentions Europe’s gigafactory initiative and a new UK agreement setting out what it expects from data centre developers, but says he hears more interest in places outside the EU, such as Canada, Mexico, India, Southeast Asia and parts of South America.

He argues this gives welcoming communities more bargaining power over things like infrastructure costs, local jobs and investment. He also makes a point that fits closely with how we think about digital businesses: keeping a data centre out of your area does not keep the technology out. It means you rent the technology rather than own it, while the construction and operating jobs go elsewhere. It is the same logic we cover in owned reach vs rented reach.

Are data centres in space a serious idea?

They are still a long shot, but Nate thinks the backlash makes them more likely to be tried seriously. He notes that xAI and Elon Musk have talked publicly about orbital data centres, Nvidia’s Jensen Huang has discussed the idea, and Google has a Project Suncatcher prototype in orbit.

The hard problems are real: radiation, cooling, hardware wear and maintenance. But orbit has far looser planning rules. Nate’s personal guess moves from about ten years to six to eight years for useful orbital computing.

🔎 Unconfirmed / rumour: Timelines for data centres in space are Nate’s own estimate, not a company announcement or confirmed plan. Treat them as informed speculation.

What signals should you watch next?

Watch what companies do with the next project they were planning, not what they say in press releases. Nate suggests asking:

  1. Are companies waiting, or moving projects elsewhere?
  2. If moving, where to: another US state, another cloud or another country?
  3. Are they paying more for sites that can open on time?
  4. Are they buying capacity from rivals?
  5. Are they investing somewhere with fewer constraints?

His view is that where companies put money on their balance sheets tells you more about their commitment to AI than any announcement.

What does the AI data centre backlash mean for your business?

For most small businesses and creators, AI is not going away, but your supplier, price and speed could change. Nate’s advice is not to treat construction headlines as a forecast for AI adoption. Your clients and employers will keep buying AI services, and they will find capacity from somewhere.

Here is what we would do this week:

  1. List every AI tool your business relies on, and note which tasks would stop if one went down or doubled in price.
  2. Pick a second option for your most important task and test it once, so switching is a known step, not a panic. Our AI backup plan guide shows you how.
  3. Keep your prompts, instructions and data in files you own, so you are not locked into one platform. See how to keep your AI agent setup when you switch tools.
  4. Review your monthly AI spend and route simple jobs to cheaper models, as explained in our model routing guide.
  5. Prepare a short, honest explanation of how you use AI for customers who are wary of it. Our guide on explaining your AI use to sceptical customers gives you a simple framework.

💡 Tip: If you are in the UK, most of the AI tools you use probably run on US or global cloud infrastructure. Local planning fights in the US can still affect prices and availability for you, so a backup plan is sensible wherever you are.

Why does the AI story matter as much as the buildings?

Because Nate believes the backlash comes from people feeling that AI is working against them. He shares a story about a friend who works at an AI company yet still feels behind and worried about his job. If even insiders feel that way, he says, the industry is “doing it wrong”.

His call is for clearer messaging about the everyday benefits of AI, and for more focus on empowerment, tools and training rather than replacing people. He notes that public opinion in Europe is not very optimistic about AI either, so this is a global challenge. We think the same lesson applies at small-business scale: how you talk about AI with your customers matters, and it should be honest and specific.

🛠️ Practical guides

Ready to put this into practice? These step-by-step guides answer two common follow-up questions:

Frequently asked questions

Is the AI boom over?

According to Nate B Jones, no. AI use keeps rising, and he quotes Google reporting 40% quarter-on-quarter growth in API usage. What is changing is where and how quickly the infrastructure gets built.

What is the AI data centre backlash?

It describes growing local opposition to new data centres in the US, which came to a head in votes on 3 November 2026, when many communities voted to block or restrict them.

What is a neocloud?

A neocloud is a specialist cloud company that rents out powerful AI chips, usually to AI developers. Nate says they are most at risk from construction and power connection delays.

Will AI tools get more expensive because of this?

Possibly. Fewer new facilities could give existing operators more pricing power. Nobody can say exactly how prices will move, so it is sensible to review your spend and have a backup option.

Does slower data centre building reduce the risk of an AI bubble?

Nate, agreeing with Derek Thompson, argues it could, because demand gets more time to catch up with supply. This is an opinion, not a certainty.

Was the Nate B Jones video sponsored?

There was no paid sponsor. Nate mentions his own newsletter and prompt library in the description.

Where can I read more from Nate B Jones?

He has a full written post on his Substack newsletter and regular videos on his YouTube channel.

The bottom line

The AI data centre backlash is real, but it is not the end of the AI boom. As Nate B Jones explains, demand is still rising, so the effect is to change who supplies computing power, where it is built and how fast. For your business, the sensible response is calm and practical: know which AI tools you depend on, have a tested backup, keep your setup portable, and be ready to explain your AI use honestly. Start today with your AI backup plan.

author avatar
Garry Knight
I'm Garry Knight, the person behind Prodify Digital. I write about email list building, email marketing, SEO, AI search and the tools that connect them. My aim is to make online marketing easier to understand, so creators and small business owners can make informed decisions about building an audience and keeping people engaged. Here you'll find straightforward guides and product reviews that explain what something does, where it fits and which limitations matter. The focus is on clear explanations and useful next steps—not hype, shortcuts or promises of easy earnings.

Leave a Reply