Data Centers, Data Centers Everywhere — And America Is Done Being Polite

Shimmy says, Shimmy says, Shimmy, Shimmy, Shimmy says, Ask me almost anything! Hey, everyone, it's Shimmy, and welcome to another Shimmy Says. Hey, if you're watching this live, I apologize.

I got hung up on a Techstrong TV interview, and we got late starting it up. But I want to talk to you guys today about data centers. And I don't just mean data centers.

I mean data centers, like data centers everywhere. Because that's all you're reading about today. It's interesting.

For the first few years of the AI, GenAI mostly, we associated artificial intelligence, as ironic as it is, with human faces. There was slippery Sam Altman over at OpenAI, Dario Amodei, the good guy over at Anthropic. Elon Musk.

Well, let's not even talk it, but Elon Musk was the face of Grok and xAI. Of course, the godfather of it all, the man in the leather jacket, Jensen Huang, who seemed to have just an endless supply of GPUs that he'd dole out to friends and to make political friends and enemies with. But I think this has changed.

" It's no longer a human, but it's not a robot. It's not an AI. I think for most people, it's not a face at all.

I think for most people, the face of AI is a huge concrete data center with some smoke billowing out of some generators on the side, electrical wires and substations, transmission lines, and all kinds of industrial cooling equipment, backup generators, all surrounded by security fences with razor wire. That's become the face of AI, like it or not. Unfortunately for the AI industry, though, most Americans clearly don't like what they're seeing.

The speed of the shift of how we think of data centers is remarkable. Just back a year ago, in August of 2025, Heatmap did a poll, and it was about 50/50, 43% yes, said they supported data centers locally, 42% no, said no, but not a majority. It was about even.

Today, that's upside down. Opposition to data centers is at 71%. Guys, that's not a slim majority or a healthy majority.

It's damn near a critical mass. Think about it. In less than a year, this has moved from a 50/50 prop at best to a three to one opposition.

You don't see stuff happening in public like that. This is truly a revolt. The New York Times recently asked whether this might be one of the fastest changes in political opinion and public opinion that we've seen.

I don't know. I don't know where it sits in the record books, but 50 points of sentiments moving against an industry in less than a year, that's more than just nimbyism, guys. That's a revolt, as I said.

Hey, this is going to be a long Shimmy Says, so don't rush on this. I'm going to try not to rush through some of this stuff because there's real human stories here I want you to talk about. The technology industry has a built-in flight or fight reflex for this kind of thing.

People don't understand the technology. They're afraid of progress, they say. They've been stirred up by fake news and activists and politicians.

They want all the benefits, but they don't want the infrastructure required to produce it. We've all heard these arguments before. Some of it is true, no doubt about it, but that's really a lazy explanation that's cutting the corners for what's really happening.

Because if the industry believes it, it is going to make this backlash worse. Because I don't think this is really an anti-AI revolt as much as I think it's an anti-deal revolt. Americans are looking at the bargains being made and what's being offered to their communities who are hosting these facilities, and more and more of them are deciding that it's not a really good bargain.

To understand why, though, you got to get out of Silicon Valley. You need to stop talking about AI benchmarks and token counts and model releases and these kinds of things, because that don't register with people on Main Street. We need to visit the places where the AI era is actually being built on the backs of communities.

I want to begin today in Broadview, Montana. Broadview has a town of about 140 residents. It has exactly one bar and restaurant, and the school gymnasium doubles as the center of the town's social life or any activities.

A story I wrote about there involves a 43-year-old rancher named Cassie Solberg. Cassie's been driving into Broadview to attend city council meetings because she's been trying to get an answer to what should be a pretty basic question. What exactly is being built right beside her family's property?

And who decided that it should go there? This is the Quantica infrastructure campus that she's talking about, or that's what it's called, and it should cover about 5,000 acres. Think about that, 5,000 acres.

Eventually, it could consume as much as a gigawatt of electricity. And I know a lot of you don't understand gigawatts, megawatts, all that, but a gigawatt is a bit more power than the local utility currently supplies to all of its customers across all of Montana. So you're talking about doubling Montana's electricity usage.

The project can also require transmission lines approximately 21 miles long. And now there's an argument about who should pay for this. Who should pay for the 21 miles?

Who should pay for that gigawatt? Should the company that needs the transmission lines pay for it, or should the people already paying their electric bills keep financing it? Cassie Solberg is not attending these meetings to argue about artificial intelligence or AGI.

She isn't debating transformer architectures or the latest model benchmarks or open weights. She wants to know what will happen to the wells, the roads, the surrounding property values. What's going to happen to the night sky that she enjoys, all the stars, being able to see the stars.

She wants to know how much electricity the facility's going to consume and how much of it does she have to pay for. She wants to know what Broadview is getting out of this deal. Guys, this is not anti-technology or anti-AI questions.

They're fundamental. They're questions people are asking when something enormous is being built right beside their homes. They weren't told about it.

They didn't have a say in it. The frontier of AI no longer runs through model laboratories in far-off Silicon Valley offices. It runs along ranchers' properties' lines in Montana.

I'm going to give you another example. Let's head south. I want to go to Richland Parish, Louisiana.

" Richland Parish used to be cotton country. It's up in northeast Louisiana. Farmers there, though, can produce a strong crop and still lose money.

That's been the story in cotton since before the Civil War. The cost of equipment, seeds, fertilizers, and fuel keep rising, and what they receive for the cotton they grow has not kept pace. Families that once considered their land untouchable are reconsidering assumptions that survived for generations.

Young people have been leaving Richland forever. The big economic development projects that were supposed to change the region's future, they never seemed to arrive. But then Meta showed up with Hyperion.

Hyperion could eventually represent more than $200 billion in investment. It could go to approximately five gigawatts of compute capacity. Supporting it is going to require 10 new natural gas-fired power plants, along with enormous investments in transmission, water usage, and other infrastructure.

Meta expects Hyperion to create approximately 500 jobs by 2035. They pay about $90,000 a year each. That's a good living in Richland Parish, but it's only 500 people.

Let's put these numbers together. Five gigawatts, 10 power plants, more than $200 billion invested for 500, let's even call them $100,000 jobs, $500,000 a year. Guys, the resources look industrial.

The numbers are big, but the employment model's not. From a distance, it's easy to look at the deal and say Richland Parish is giving away too much. Property tax deductions, concessions given to Meta and the data center builders approach 80%.

They're not paying 80% of what they should be. There were no non-disclosure agreements. This was all politically maneuvered.

It sounded something out of the Huey Long kingfish days. The risk of the data center builders are being absorbed by the surrounding community in Richland Parish. And when you get closer, that decision becomes worse and worse.

Five hundred good jobs mean something in a parish that has spent decades watching farms consolidate, children leave, and opportunity pass it by. But Hyperion may be a lottery ticket with a considerable amount of fine print, because when a community has been offered very few other tickets, you understand why they buy one that's not so great. Farmers understand risk better than most Wall Street investors.

They make a wager every season. They can do everything correctly and still lose because the weather changes, costs rise, or commodity prices fall. Meta has the capital, it has financing structures.

It has options. If the economics of AI change, they'll walk away from Richland Parish. Richland will still have its land, its schools, its roads, its water, and its hope that this particular crop will finally pay off.

Broadview and Richland, though, are not really opposite. They are two sides of the same story. It's one community fears that something enormous is being imposed on it.

The other fears what will happen if its one enormous opportunity disappears. They both have little hope. I want to go to another place now called Boxtown.

Boxtown's right outside of Southwest Memphis. This may be the story that captures the entire issues better than the other two. In January of '61, there was a newspaper photographer named Saul Brown.

He took a picture of two Boxtown residents. They were James Thredford and Albert Lee Wright. They were in a horse-drawn cart piled with firewood to bring back to their houses to warm their houses for winter and provide cooking.

Behind them stood a giant electricity power plant. That photograph right there captured the entire economic times in one frame. Two men collecting firewood for fuel while there was a huge facility generating electricity right behind them.

But that electricity didn't go to the homes in Boxtown. 'Cause a lot of the homes back then didn't even have electricity. That electricity was shipped out to the city where people would use it.

Six decades later, Boxtown is in the same spot. It's a place where infrastructure, a huge infrastructure project, producing one of the most valuable resources in the world is being built. But this time, it's not just electricity that the people of Boxtown aren't getting.

They're not getting machine intelligence. It's being manufactured inside this massive Colossus AI complex built by Elon Musk and xAI for Grok. Residents have described sitting on their porches close enough to smell fumes coming from the facility beyond the woods.

Once again, that Boxwoods is asking the questions its parents and grandparents asked: what do we receive for living beside an infrastructure that creates tremendous value for people somewhere else? The corporate tenants have protected themselves rather well. Anthropic's arrangements give it access to hundreds of thousands of GPUs across Colossus and Colossus Two, its sister site.

Google's also contracted for a large block of that compute. Their agreements allow provisions allowing them to leave after the initial commitments on 90 days' notice. Anthropic can leave, Google can leave, SpaceX can redirect the GPUs.

It could reclaim them for Grok or rent them to another AI company. But you know what? The people of Boxtown can't leave.

They're stuck with it. The models change, the contracts change, the corporate logos change. We go from electricity to intelligence, but there still stays Boxtown, right beside, on the land, next to the turbines, the emissions, the infrastructure.

That's the moral question beneath this entire data center build-out. What do the communities receive when they host this infrastructure, producing something that's indispensable for everyone else, but not here? I got one more for you.

Let's go to Fort Worth, Texas, because this is where the new arithmetic of economic development becomes almost impossible to ignore. 1 billion. The number of full-time jobs expected, wait for this, 37.

Now, to be fair, they're good jobs. These are $150,000-a-year jobs, but there's 37 of them. The data center providers were given tax abatements up to 70%.

For generations, towns have accepted the noise, the traffic, and pollution that came with factories because those factories employed hundreds or thousands of people. It was a visible bargain everyone understood. The plant supported the families in the town, it supported local businesses, and it supported entire counties for generations, these factories.

But not data centers. They're different. They're designed to be automated, efficient, with minimal human intervention.

A data center consuming the electricity of a city might employ a permanent workforce that could fit inside Broadview 140 people's only restaurant. That doesn't make the facility worthless. It may generate tax revenue even.

It may enable enormous activity, economic activity somewhere else. But we need to stop selling these facilities to communities as though billions of dollars in capital investment automatically produces thousands of local jobs and huge benefits, because they're not. The secret surrounding some of these projects makes the situation worse.

In Tucson, residents discovered that a mysterious development called Project Blue belonging to Amazon only after a document was mistakenly released through the public records request. The city, after an uproar, canceled the project. Mason County, Kentucky, another place.

A family reportedly refused millions of dollars for its farm because the prospective buyers would not reveal who they represented. Another data center project. In Virginia, a public records survey found that 25 of 31 localities with existing or proposed data centers had signed non-disclosure agreements.

They couldn't tell their local residents. These aren't even elected officials a lot of times, that they don't know how much power and water a project's going to require. They may not know what effects it could have until the important decisions have already been made.

Guys, this is a recipe for disaster. It's a recipe for public suspicion. It's a recipe for people to be revolting about what they're building next door.

Utilities, they love it. It's great growth opportunities for them. They haven't encountered this in generations.

Their incentives aren't difficult, but more power demand means more power generation, more transmission, a larger rate base, greater returns. But ordinary customers can't be asked to foot this bill, not on their backs. They're paying high enough rates as it is.

If a utility builds billions of dollars in new capacity and the projected AI demand never materializes, who pays for that? You do. We do.

I do. We all do. In Texas and elsewhere, data center developers are responding to this by saying, "We won't use your public grid.

" They're building dedicated behind-the-meter power generation plants to bypass the local power grids. Right? Bringing your own power may prevent costs going up from the public grid, but it also It also brings another problem.

These power plants right outside the data centers are right outside people's houses, and now they're breathing the fumes of natural gas power plants, and the noise that it makes, makes their properties unlivable, unsellable. It makes for a rigged game, and when people believe that the game is rigged, that building a private grid for billionaires is not going to restore their confidence. I'm going to bring the last one to you, Utah.

This is the biggest project of them all, 40,000 acres, 500 to 1,000 jobs, nine gigawatts. That's like what New York City uses type of electricity. The whole state of Utah.

But the people there, what are they getting out of it? Construction companies want projects, GPU manufacturers want deployments, utilities, and all them want customers, investors sure as heck want returns, and politicians want big announcements and ribbon cuttings. " This doesn't have to be a conspiracy, though.

Every participant has a rational reason to want the same thing. This is why they call it the AI industrial complex. It's exactly what Eisenhower warned us about.

Because when all these interests align, the guys who lose are the little guys. It's the people living beside this who have to live with it. The billionaire tech bros, they don't live nearby.

They too often appear callous and tone-deaf about this. They tell Americans the AI future is inevitable while those same Americans are worried about losing their jobs to it, their electric bills and water supplies going up. Investors, they see returns.

Politicians see headlines, as I said. But the communities that were passed over by the previous American dreams, they don't have much choice. They see this as their last chance to participate in it.

When you put it all together, though, what do you get? You get data centers everywhere. But the tide's turning.

I do believe the public is fighting back. We're seeing moratoriums in New York, zoning battles, political campaigns across the entire country. Local governments in red states and blue states are reconsidering the bargains they've made.

America wants the intelligent grid. We just don't want it next door. There's one more story I want to tell.

In 2007, a group of thieves dressed up as police officers, walked into a Verizon data center in Loudoun County, Virginia, and walked out with 80 servers. Supposedly, they were paid to steal these 80 servers because they contained fraudulent mortgage information. People got caught.

They went to jail. The New York Times did an exposé on this. But this isn't really just a colorful story like an Ocean's 11 goes to the data center.

It's more relevant today because today the lesson is, guys, the cloud was never in the sky. It's at a data center. It has an address with loading docks and security guards.

These data centers represent tremendous targets. Physical security, as much as cybersecurity, is a requirement there. Maybe that'll bring some jobs over to it.

But we've got to be clear about what that means. We need these data centers. We need them to be secure.

Everything we're doing about AI has to run on the infrastructure underneath it. These models run on chips. Chips sit in racks.

The racks consume electricity, produce heat. It's all one virtuous cycle or non-virtuous cycle. How are we building them?

How are we building them with public consent and political access, with the will of the people? Do developers pay the actual cost? Are the communities next door being treated humanely, properly, fairly, or do they shift this onto our backs?

Public acceptance is not a public relations problem. It's an infrastructure requirement. And when we start realizing that, we'll be better off because most Americans are never going to meet Sam Altman or Dario Amodei or Elon Musk.

Their opinion of AI instead is shaped by the data center proposed five miles from their house, the gas plant constructed to power it that's sending fumes their way, and the water it consumes and the noise it creates. The industry should not be surprised when opposition to the building becomes opposition to the technology inside it. Data centers everywhere.

Behind every gigawatt is a place. Behind every place are people. They're not externalities.

They are not permitting obstacles. They're not extras. If the companies building the intelligence grid want Americans to embrace the future it enables, they need to stop treating communities as obstacles, ratepayers, and financing and public consent as something that can be purchased at the governor's office or worse.

The AI race isn't going to be run by whoever builds the best model or buys the most GPUs. It's going to be won by whoever earns permission to keep the lights on at these data centers. Data centers everywhere.

This is Shimmy. I'm out. Thanks for sticking for this extra long one.

We'll see you again next time. Shimmy says, Shimmy says, Shimmy Shimmy Shimmy says, ask me almost anything