Go back

Clayton Morris: Blackouts & Water Shortages, Tech Overlords Build AI Dystopia & Crush Resistance

63m 34s

Clayton Morris: Blackouts & Water Shortages, Tech Overlords Build AI Dystopia & Crush Resistance

The conversation explores the physical and societal implications of the AI data center boom in the United States. The guest argues that the build-out is in its infancy, with thousands of new centers planned, each consuming enormous resources—particularly water (up to 500 million gallons per center) and electricity—leading to higher utility bills, water shortages, and projected blackouts in communities like Homer City, Pennsylvania. He highlights federal support, including a DOJ AI litigation team that sues communities opposing data centers, and relaxed EPA rules that accelerate construction. Data centers often claim self-sufficiency using revitalized coal plants or small nuclear reactors, but this power is not shared locally, so residents still face rising costs. The demand for compute is so high that Amazon is sold out through 2030, and new FCC rules mandate 65% of robotics components be U.S.-made, intensifying the build-out. The guest sees no clear economic justification yet, noting some companies report efficiency gains, but warns of widespread job losses in fields like accounting and aviation. He criticizes the lack of public consent, the surveillance risks of AI-integrated tools, and the potential for humanoid robotics, while suggesting a silver lining: a renewed desire for human connection. Ultimately, he remains uncertain whether this is a bubble or a transformative shift, but stresses the immediate, tangible costs to ordinary Americans.

Transcription

9353 Words, 51530 Characters

English
Hank joined BJ's Wholesale Club the day he became a father of 30. I coach football. Now, Coach Hank saves up to 25% off grocery store prices. 30 pounds of pasta, 3 cases of protein bars, 75 sports drinks. And that's just pregame. He knows teamwork, and BJ's knows savings. This is your home, Coach. Home of the Save. Join for just $20 at bjs.com slash mesquite and save 10 cents per gallon for 6 months. Open soon. Limited time offer. New members only. BJ's. Home of the Save. Clayton Morris, thank you so much for doing this. A lot of us have been focused on the AI slash data center story from a badly informed tech perspective, trying to answer the question, you know, will machines become our masters? Is this the end of the human race? But there's another component to the story which may be bigger, and it has to do with the physical world and the demands of AI on that world and the short-term implications for the rest of us. Because of the construction of these data centers, and you have been on this more than anyone I know. So, can you just frame the way that you're looking at AI and data centers? Why is it important? Well, I think it's so important for these communities. You're seeing, if we step back, there's so much math involved in this. And, you know, I always say that every real estate story is all local. So, it's all different what's happening in New York from what's happening in Texas right now, what's happening in Quincy, Washington. Or, Homer City, Pennsylvania, like it's all different. But it's all of the same, you know, it's all of the same animal, so to speak. And so, we are, I don't even think we're at the beginning of this. Because there's so much discussion about AI, just flip on CNBC or Bloomberg or anything. It's AI everywhere, we're going to hit an AI bubble. We haven't even started this build out for AI data centers across the United States. And I honestly think we're at the precipice of a major transformation for humanity. And we won't know for years whether it's good or bad. There were studies done on Google, the Google effect back in like 2011. You know, our kids like outsourcing their ability to search for things to Google. Like, think about how funny that is now in hindsight all these years later. But we are, I think, at the beginning of this build out, we haven't even begun to see the thousands of data centers and the way that this is going to change humanity for good and bad. And I think that anyone who says they know for sure doesn't know. And I think a lot of these companies that are spending $500 billion, we just got word that Nvidia is going to take out credit in order to build out massive amounts of infrastructure. So this is like a new phase of this now, not even their own money. It's going to be a credit story, borrowing money at enormous sums from Wall Street in order to build out this infrastructure. So just at like a micro level, if you just think about communities and what it means to just you and I, it means right now increased prices for kilowatt hours just to run your home. It means increased prices at the grocery store just to get additional groceries because the cost of everything has gone up. It means water shortages across the United States. We're already seeing it. We're already seeing its effects on ranchers. So it's so big and there's so many pieces of this. You've got the robotics piece of this, which is the human robotics layer. So AI will, of course, be involved in every level of robotics, which is going to be as if you listen to Elon Musk and I believe he's a smart, very smart person. He says that his Tesla robotics division is going to be the biggest product in human history is Optimus robots. So if you take him at his word or even if half of his word is true, and that's just one company, the human robotics element of this is going to be absolutely massive. And of course, then the AI, the compute power that will be layered inside of that hardware is the requirements for that are going to be astronomical. And that's just one piece of it. Every company that exists right now, just about every company is incorporating AI. It's like the AI that you don't see that I think is in many ways the most invasive of all of this. So, for instance, Zoom, like you and I are talking right now on Zoom. Right. Well, they've recently sort of shifted to becoming really an AI productivity company in many ways. That's their focus now. Forget being just like, hey, we're going to have a meeting over Zoom. Now they've added in a take notes feature, which can use AI and basically take notes for you during your meeting. And now their focus is AI. So AI being built into everything you can think of. May I ask you to pause and ask a question for a second? So just to take Zoom as an example. Zoom, again, being this software program that allows people to speak on video conferencing anywhere between each other, anywhere in the world. Basically, what they're saying is we're going to transcribe every conversation and that's going to be helpful as a productivity device somehow. But what happens to the data? So the substance of your conversation is now on a Zoom server. Do you have control over what's done with that data? The words with that information? I mean, I think this idea that you have control over any of your data is a joke. You can ask companies to delete your data. But of course, when the FBI then wants to subpoena that data, suddenly it manages to show back up again. So wait, I thought, you know, like that you could just like delete that stuff or Google tells you that, hey, after 30 days we delete this stuff. It's all there. It's fractals. It's always going to be there. So, you know, you opting into this, the take notes feature and I don't know the full back end of what Zoom's information looks like, but most of it is all built on Amazon Web Services. Most of it is built on Amazon's infrastructure from Palantir to flock safety. You know, they'll tell you that they've got their own proprietary layer of what they do with that data, but all of it's built on like Amazon infrastructure and sort of this back end infrastructure. So where is that data going? It's living in these data centers and being built out at a clip that we've never seen before. So right now there are about eight hundred nine, give or take a few data centers being built right now. There's about four thousand eight hundred of them that are currently built. There's another, I think, thirty nine hundred that are in that have been projected to be built over the next year. Doubling the amount of data centers in the United States and they're not getting smaller, they're getting bigger, they're getting more massive. Take, for instance, like the project Stargate in Texas and Abilene, Texas, massive, massive infrastructure that Trump, of course, was championing. And a lot of this was opened up. We shouldn't forget at the very beginning of the Trump presidency with all these executive orders so that you could start building a data center without even getting environmental permits so you could start the construction. So if like the redacted news, you know, the data center wants to get started, I could go out tomorrow and buy some land and I could start this construction, local permits and get started, forget the EPA requirements. Like how much pollutants will be spun off as a result of the diesel diesel generator backups or the water influx or any of the particulate matter that will be spun off as a result of the the center, the water usage, all of that, don't don't worry about it because now the EPA doesn't care, you can get those you can already start construction. And then, of course, on top of that, you have Trump's A.I. litigation team, which is pretty funny when you look at it. So the Department of Justice has built like an A.I. litigation team to go out to different states like you can see where the priorities are. So if there's a state or a community that says we don't want data centers in our backyard or we're putting onerous rules in place and regulations in our backyard because we don't want these things here, well, here comes Trump's Department of Justice and the A.I. litigation team. To town, not staying at a Best Western when they arrive and they're going to start to look at your community and sue you for onerous regulations. What? They're not going to look at your data centers. This was part of that executive orders and executive actions at the very beginning of Trump's presidency. So glad we have an A.I. litigation team in place. Wait, so the our Department of Justice, which I thought we owned as citizens, is now working on behalf of private companies, which we do not own, in order to force communities to subsidize those private companies? Yes. So under the attorney general, under this A.I. litigation team, this task force, we're going to be working on behalf of private companies, which we do not own, in order to force communities to subsidize those private companies, which we do not own, but equity in case You didn't already know that. Big investors have bought up veterinary clinics across the country and doubled the amount they charge. So they're exploiting your love for animals to make money. And are they providing better services? Well, if you can't afford to go to the vet, you get no services at all. We did a whole show on this and it's really distressing. Dutch is the solution to this. Licensed veterinarians, easy remote appointments. Start with the code Tucker, plans at just $82 per year. That's like nothing compared to what you will get if you walk into a private equity-owned vet. 50% of Dutch customers say they had not seen a vet in three years because they couldn't afford it. Dutch fixes that. 10-minute calls, no waiting room, no clinic markup, free shipping on products, coverage for up to five pets. Visit dutch.com slash Tucker. Use the code Tucker for $50 off. Truly, Dutch is real. It works. We did a lot of research into this, used it before recognizing it. We're recommending it. Dutch.com slash Tucker. I coach football. Now Coach Hank saves up to 25% off grocery store prices, 30 pounds of pasta, three cases of protein bars, 75 sports drinks. He knows teamwork and BJ's knows savings. Home of the save. Join for just $20 at bjs.com slash mesquite and save 10 cents per gallon for six months. Open soon. Limited time offer. New members only. BJ's, home of the save. This episode is brought to you by TXU Energy. For the first time ever, Texans have an energy plan that combines free nights all year with daytime summer savings. It's called Free Nights and Cool Summer. You get free nights every night all year long, plus a 25% bonus toward your daytime use all summer. It's a first-of-its-kind plan created for how Texans actually use energy. Discover how much more you could save with Free Nights and Cool Summer. TXU Energy. Energy for everything. Visit TXU.com for more information. Rep number 1-0-0-0-4. Tucker. The Major Brent Taylor Foundation honors the legacy of, well, Major Brent Taylor. Soldier, husband, father, and a major who died in the war in Afghanistan. The foundation supports Gold Star families, veterans, and community leaders to whom we all owe the highest levels of respect, and here we do respect them. Black Rifle Coffee is proud to stand beside the Major Brent Taylor Foundation and other organizations with the same spirit. It's hard to know where your money really goes when you donate to charities, but Black Rifle Coffee has vetted these, and these groups are real, and you can be assured they're supporting a genuinely righteous cause. To learn how you can support veterans even after the war headlines are gone, visit the Major Brent Taylor Foundation at MajorBrentTaylor.com. MajorBrentTaylor.com. And there's going to be, you know, the loopholes that'll be used here, and we'll see what this data AI litigation task force does. It's like the, you know, the White Supremacy Task Force or the Anti-Semitism Task Force that'll be deployed across the United States. So it'll be interesting to see, like, what loopholes are employed. But good news is, like, 500 or so different communities have already fought back against this. There was a friend of the show, Derek Brose, and Hakeem Anwar put on a day of action on the beginning of August against these data centers. So people around the country to come out and say, we don't want these things in our backyard, at least kind of drawing attention to them. Um. We don't want these things in our backyard, at least kind of drawing attention to them. We don't want the noise. We don't want the pollution. Um. And there's so much just misinformation and there's fear, uh. about all of these data centers right now. And there's just a lot of. There's just a lot of BS that's kind of floating around about how these things operate. Maybe we could talk about it if you want to kind of dive a little more deeply into. Well, I want to dive very deeply into it. I mean, you've already. You know, we're 10 minutes into this. You've already shocked me. So, um, I had no idea that was happening. And, you know, if you were to have dinner with me, I would pretend to be highly informed on everything. But obviously, that's not true because I didn't even know this. I didn't know that there was a Department of Justice AI task force designed to pummel communities into accepting things they don't want, which is to say taking sovereignty from American citizens. So, um, yeah. I do want to go into greater depth. So, tell me about the misconceptions or, as you said, misinformation. Well, for one thing, there's. If you listen to. Wall Street or you listen to the people that. The sycophants, you know, like on CNBC are the people that love to have these CEOs on their shows and, you know, access journalists, you know, who love to have these CEOs on their shows and kind of promote AI data center build-outs and all the new technology that they're rolling out. And so some of this technology is good, right? There is actual good cooling technology that's emerging from companies like Vertiv, VRT, which instead of using. Water that would cool it with using evaporation technology, which is the big. One of the biggest problems. Forget the electricity piece of this for a second. That's a big. That's the other big part of the story. By the way, this energy, you know, it's like you can. It's like you can rob Peter to pay Paul. We're not using water. Oh, yeah, but you're using the energy now, the electricity to use dry cooling to cool. Okay, so it's just like. It's like the shell game. You say moving things around a board, you know, so when you see these like CEOs come on CNBC or Bloomberg or whatever, you know, you're like, and tell you how they've solved the cooling problem because they don't need water. Okay, now what about the massive amounts of electricity that you need now to use dry cooling? So it's. Anyway, it's a show. Well, it takes resources to make things. That's just the bottom line. It takes resources to make things and the supply of resources is finite. So anyone. It doesn't mean you shouldn't make things. You should. But every act of creation is a trade-off. So it's not clear why we have to lie about that. Let's just be honest about it. But anyway, so. But to the question of water, can you go into that? How much water, exclusive of this new and promising technology you just described, but like in general, how much water does it take to cool a data center? Hundreds of millions of gallons. And most, let's be clear, most of the data centers are using this evaporative cooling system. And I mean, by a huge percentage. So, you know, that's great that we've got some new technology out there that uses cooling plates and it keeps it in like a closed loop. So the water. Water is like right up against the computers, the GPUs, and then it kind of flows through and it gets cooled almost like by an air conditioning unit and then comes back a little cooler. But still, that excess heat needs to go somewhere, right? So it pushes off somewhere. It's not. But the water usage, yes, you're limiting that. But on most of them, overwhelmingly, it's evaporative cooling. And so you're. It's all just being wasted. If you think about an average American home, about, you know, 200 gallons to like run an American home, 300 million gallons or 500 million gallons of water just for these data centers. And we're talking like small data centers. So these communities are already seeing. They're already pressed. And if you go to the desert Southwest or you move out towards, you know, Colorado, Nevada, Utah, Wyoming, a lot of the ranchers in these areas are at their limit with their ability to get water to just feed their animals. To feed their heads of cattle. So everything is so dry. I just interviewed a rancher who's dealing with this right now. So in order to get permits to pull, you know, to pull water to feed their heads of cattle, they have to go to the city, the municipality, they have to get those rights. But most of those permits are already used. Like they're not. There's no like new permits just like sitting there for Bob Rancher to come grab. So they're already tapped, tapping into these aquifers. I mean, requiring them to go down, pull out any kind of water they can from these aquifers, right? So those are already tapped. A lot in the West of the United States. And then on top of that, now you have these data centers that are coming in and all these ranchers are asking the question, how are they able to get water? Like we've applied for permits and we as ranchers are not allowed to get water, but they are. Where are they getting the permit? So they're carving out these special permits for these ranchers, which is. I mean, excuse me, for these data centers, which is unbelievable on its surface. And so these ranchers then are having to compete for what little water scraps are left to feed their head of cattle, which brings up a larger question about wanting to decimate the beef industry, of course. And you have 68 million heads of cattle. How many of those are being whittled away, whittled away, whittled away? And then you've got wildfires out West on top of that, which is destroying the grazing land. So they don't have enough water for that. But the data centers are fine to get these water rights somehow. People should ask questions about that. Like how are they able to get these carve outs to tap into these aquifers in ways that these ranchers are not? It's very interesting. So the misinformation right now is this idea that, well, now we've got this new sort of liquid cooling setup. And so therefore, we're not using as much water. And that is the case in some cases with some of this new technology. But a lot of that isn't even built out yet. But you're seeing like these companies like Vertiv like exploding. I think they're even more valuable now like share price, not more valuable, but like more than Amazon because they're selling these cooling technologies that they can use in these data centers. So that's just one piece of it. But the majority of these data centers are using an enormous amount of water that we simply don't have. And you're having these communities now fighting over it in order to, you know, homeowners who are having to deal with this now. And of course, the other big piece of this is the smart meters. So just take, for example, like Aurora, Colorado, not too far from me, They implemented these smart water meters so they can monitor at every 15 minute increments, like the amount of water that you're using as a, you know, as a, as a private citizen compared to the water that's being used from these data centers. I don't know what sort of metering system they have on these data centers, but there's a, there's a massive data center called QTS in Aurora, Colorado. Massive, huge. And in that, in that setup, they've got, they're using an enormous amount of water and the homes around Aurora, Colorado are using now these smart meters to monitor how much irrigation, how much water they're using. And they've sent out thousands of notices to people that live there that, hey, you've used too much water, you're using too much water. And these smart meters are only going to increase across the United States. There's many communities that are now rolling out these water smart meters in order to make sure that you're not using enough water. But are those same, are those same, you know, municipalities sending out notices to the data centers to let them know how much water they're using? You know, a lot of people hit that afternoon slump and immediately reach for another cup of coffee. Guilty! But is the answer always more caffeine? Well, sometimes you just need hydration. Thing is, a lot of people are already drinking a ton of water, but they are not fully hydrated because water alone isn't always enough. Your body needs something called electrolytes, salts, to help absorb the water. And use the water, and that's where Re-Lite comes in. Re-Lite is made by Redmond, the company behind Redmond Real Salt. They've been harvesting mineral salts in Utah since 1958, a state famous for salt. And they are totally transparent. You can visit their mine if you want. If you ever show up out there, you can go to the mine and see where it comes from. So Re-Lite's ingredient list is completely clean. There's no sugar, no chemical preservatives, no artificial colors, flavors, sweeteners. Just sodium, potassium, magnesium, calcium, chloride, and coconut water. That's it. So whether you're working out, spending time outside, or just trying to stay sharp throughout a day, Re-Lite is an easy way to replenish your electrolytes, which, again, you need. Flavors like lemon-lime, mixed berries, strawberry-lemonade, watermelon-lime, mango, pina colada. You're making me hungry reading the list. They've got it. All of it. So if you're ready to give it a try, visit redmond.com slash tucker. Use the code TUCKER for 15% off your first order. Redmond.com slash tucker. Code TUCKER. 15% off your first order. Or click the link in the video description, or you can even scan the QR code on the screen, or check the pinned comment. Many ways to get it, and you should. Hey, it's Tucker Carlson. We launched ALP about a year and a half ago, and people are voracious for it. They can't get enough. Everywhere you go, you get the same question. When are you making more flavors? We are not in the business of disappointing people. Instead, we're in the business of satisfying them. So we have made at least three more flavors. Introducing the ALP Frontier Collection. They are Campfire Cinnamon, Sunset Mango, and the Alp Frontier Collection. They are Campfire Cinnamon, and Cowboy Coffee. Available in 4mg and 8mg. Because if you're going to use nicotine, you ought to have some options. Get the new Frontier Collection now at alppouch.com slash tucker. Tammy regularly hauls stuff to her kids at college. That's why she shops at BJ's Wholesale Club, saving up to 25% off grocery store prices. Beef jerky, microwave meals, laundry pods, even they can't screw up. Have you met my kids? And with BJ's gas, she gets there for less. They call me the road warrior. This is your home, road warrior mom. Home of the save. Open soon. Limited time offer, new members only. You get free nights every night, all year long. Plus a 25% bonus toward your daytime. TXU Energy. Rep number 10004. I'm confused. I mean, so one of the long-term and distressing trends in the United States is the explosion in population in places that don't have the resources to sustain that population. So, I'm thinking Clark County, Nevada, Maricopa County, Arizona, you know, Phoenix, Vegas, Denver, parts of California. Like, they don't have the water for this number of people. So, we know this. No one seems to care. This is going to be a huge issue going forward. But there are other parts of the United States, upper Midwest, for example, that have a ton of water and not a lot of people. So, why would you ever build a data center in Colorado, for example, which already doesn't have the resources to sustain that population? Why would you ever build a data center in Colorado, for example, which already doesn't have the resources to support the people? Like, who's making these decisions and why? Do you have any idea? Why is every data center not being built on Lake Superior? Well, because the money. So, just take, for example, I believe it's probably the money. But a lot of it is these data center build-outs now. Well, a couple pieces. I'll take Quincy, Washington. I think it's an interesting sort of anecdote of a story. So, the propaganda around Quincy, Washington. I think it's that data centers saved this town. Data centers came in, used the hydroelectric power here, and the influx, the tax basis now, I think it was like three different centers that were built, three different data centers. The amount of influx of like a 56% increase in taxes flew into Quincy, Washington. So, they built $120 million new school. They built all new sewer infrastructure inside of Quincy, Washington. They built all new parks, water slide parks for the kids. Crime has dropped. And they point and they blame it specifically on the data centers. So, all of the money that flows into these communities from their taxes is exorbitant. So, they don't want to turn away from this additional tax revenue in these communities. And that to me is just, it's always the money, right? That's at the heart of it is clearly, it's clearly the tax revenue that's coming in. The other piece of this is that I think these companies, they are telling, I say they're telling lies, but they're certainly structuring these deals in a very interesting way, which is we will, we're not going to pull from your resources. So, we're not going to tap the electricity in your town. We're not going to tap the excess water. We figured that out, which is laughable because there's always the spillover effect of these things. It's not self-contained. But a lot of these new data centers being built professed to be, self-contained, which is we're going to produce our own power. So, you don't have to worry about it. So, we're going to use these SMRs, these small nuclear reactors, these modular nuclear reactors. There's companies like NuScale Technology, which are exploding these. And so, Amazon can put a nuclear reactor at one of these data centers, a small nuclear one. We don't need to rely on Three Mile Island, which by the way, is being revitalized right now, in order to power data centers. I don't know if people recall the fund that had this. It happened at Three Mile Island. But so, that's being revitalized right now for data centers. So, they can build self-contained power, modular nuclear technology. The other story, and I was speaking to a source at Amazon, who it was just crazy to hear what their build-out looks like right now. And said that when they go in, they're looking at like defunct coal plants or other power plants that are nearby. These data centers. So, they're looking at like defunct coal plants or other power plants that are nearby. And if they can't get the power they need from the local community, they just buy up the coal-powered plant or the other power plants that were there. They buy them up, and then they revitalize them. Bring in a couple thousand people, build it out to a new energy complex. And then, oh, you think, oh, that's great. So, now that coal plant that you're taking, which is defunct, it closed three years ago or four years ago or 10 years ago, under regulations and onerous regulations and so forth, you're going to revitalize it, and then the community will benefit. No, no, no, no. No. They're going to get all the power. So, they're not giving any of this power back to the community. The coal plants specifically, and the ones that are in Nevada and Colorado, where Amazon is coming in and building these facilities, they're buying them up and revitalizing them, and they're turning them into natural gas, facilities, so that it's far cleaner. They don't have to deal with the coal regulations, and then they will use all of that energy. Oh, okay. So, then it won't have an effect on the community then, right? So, then energy prices won't have gone up in the community. Oh, no, no, no. In fact, the sources I've spoken to at Amazon said, absolutely, the energy costs will go up in those communities, and they already have. So, they're self-contained. Why are my energy costs going up? It's because the spillover effect. And the grid capacity and everything else. So, you're already seeing this across the country. People without data centers, even in their backyard yet, are seeing increases to their kilowatt hours, their per kilowatt hours, upwards of 12%, 13%, 14%, just in prices. We in Colorado have already seen prices skyrocket, even without the massive data center build-outs, because it's future projections. So, it's the auction process of electricity. So, there's a lot of data center build-outs, but the future projections on the auction process believe that they will be over capacity. And they believe, and local communities are saying, expect blackouts. If I could, I'll give you an example. Homer City, Pennsylvania, Western Pennsylvania, near where I went to college out there, University of Pittsburgh, and that whole area in Western PA. This is a really interesting anecdote because Homer City, their energy prices have already increased. But they're building this, the data centers that have come in there have revitalized the old coal plants that are there and were shut down in 2023. Now they're building this massive Homer City energy complex on top of it, converting it to natural gas. But the people of Homer City don't get the benefit of it. They don't get any of that. It's going to go to the data center build out. And I think the mayor, the other members. Of the local city council have already seen those prices go up and have said that we will expect blackouts. This is Pennsylvania. This is in Texas where you get incredible heat waves. You know, you still get heat waves and things, but not like Texas or Arizona. So expect blackouts like over the next few years, our grid cannot handle this. And so the people of this area that live in the surrounding area can expect to see their homes affected. By this and blackouts that will unfold in these, in these neighborhoods and that again, that's Pennsylvania, Western Pennsylvania, nowhere near like Abilene, Texas or Arizona. So these things are very real for people and the costs that will be passed on to the consumers where energy costs are already skyrocketing is, is very real. I know the Trump administration says, well, we're putting this cap on as part of this, you know, this AI build out, we're putting this cap on it so that we're going to make these. The producers pay that, but there's no evidence that they are like, we are already paying more for electricity already. I don't know what it is for you, like in Maine, like if you've seen it there, if you know, like what data center, because it seems like I'm looking at the data center map in Maine is like very minimal for build outs from what I can tell. But so you have the future projections of costs of this energy build out that are already affecting people. I mean, I guess I'm really, I'm struck by so many parts of the story, but the one that I'm fixated on maybe most is the total powerlessness of the people who live in this country to have any effect on any of this at all. Going back to what you said at the outset of the conversation about the department of justice targeting anyone who complains about this or tries to slow it down. Have there been, I don't think most people are for this. I don't think people understand exactly what the purpose of it is. I don't think there's been. I don't think there's been much of an attempt to convince us that this will improve our lives, which is to say the technology itself, AI data centers. Is that, how is that good for me? How does it make this a better country for my kids? No one's spending any time trying to convince us that it is. It's all scaring the crap out of people, maybe for good reason, maybe not. I really don't know what the motive is. You talk to an AI person, someone developing AI, and they spend very little time telling you how this is going to cure cancer and a whole lot of time telling you how it may. And humanity. So it's like, I don't really get what's going on here, but I know that people aren't for it. Has anybody had any success in protecting their own communities from destruction by this? Maybe New York state is the only real story right now from a state level where you have the state now banning the build out of data centers right now. So, and, you know, you have governor Abbott in Texas. Uh, who is now putting a one year moratorium on data center build outs through an audit process. So you have to prove like your energy efficiency and all of that in order to be built out in Texas. Um, but I also, I love, I love the hypocrisy there. So it's okay that we can have like massive child trafficking operations like happening, right? Right. The state of Texas and right under your nose and old Walmart facilities and thousands of children being trafficked. And, but we're going to, we're going to audit data centers. Abbott has. So has overseen the total destruction of, of Texas and that's, um, I don't, yeah, I'm not saying he's evil. He's just weak, but, uh, yeah, it doesn't get the credit he deserves for that, but it'll be obvious in the next 10 years. Uh, but New York, I mean, so we'll see. I mean, I, I don't know. And I don't know what's going to happen in Texas. I mean, you have this project Stargate project in Abilene, Texas, arguably the large, one of the largest data center build outs in the world, uh, in project Stargate, one that president Trump has championed. So I wonder if there's going to be any pushback from his, uh, the AI task force. Um, you know, if anyone speaks out against that in the state of Texas, but New York, and I don't know how much of this is, um, you know, how, how, how much of this will continue and, and will New York be like the state that says we're, we're done with data centers altogether. Sorry, go, go someplace else. Um, what will that look like? Uh, and I, I don't know, and I don't know if this will be rolled back pending some, Hey. We now feel that they're self-sufficient, the power is self-sufficient, that they don't need to rely on us at all, uh, for energy, for water. Um, and they figured out the diesel generation backup, so that's not spilling off particulates and all of these other things. And now we can welcome them back, uh, to the state. Um, you know, and then you have governments, uh, you have local governments like in the state of Georgia, uh, of course, who are going after and using imminent domain. Um, and then you have governments like in the state of Georgia, uh, of course, who are going after and using imminent domain. monitoring its citizens. So national security, the idea that we don't want China to have this AI layer for all of the data collection that we're doing in the United States, whether it's flock safety, I know that's the most like kind of public facing one right now, but that data goes somewhere and it goes through AWS. It goes through Amazon Web Services, the same way that Palantir does and all of it needs to be stored somewhere. So it needs to be stored in these massive facilities, whether it's Nevada and other places and all of this data that's being collected on Americans from, you know, phone calls to email, everything that they're using to scan and track us at this moment. And so they don't want China involved in this, of course, and that's the outwardly facing, you know, enemy. We don't want to use, we don't want any of their models, their AI models. We want to use our own models that are helped inside the United States and funded. By the United States government or in part with large grants and so forth. So that's like the national security piece of this. So about a week ago, there's a story that really didn't get much attention, but it's huge and it's going to. It's going to further bring more AI build out and infrastructure into the United States in a huge way. And that is the FCC's ruling on hardware. Basically, the components for robotics cannot be built outside of the United States. so 65% of the component parts in your humanoid robots or robotics need to be built within the United States now. Well, Elon Musk, I mean, and Tesla and Optimus, they're going to have to shift dramatically so that now instead of most of it being built in China, most of it will need to be built in the United States. The actuators, fingers for robots, all of that stuff needs to now be built in the United States, 65% of it. So that's another big piece of this war right now over AI between China. That's going to increase though to over 80% by the year 2030. So that ruling is like, hey, get your stuff figured out now at 65% because it's going to go up to 80%. So all of this stuff will be built in the United States, or at least presumably it'll be built in the United States. And then this AI layer inside the United States is just going to grow. And so speaking of sources at Amazon, they're, they're build out, they are sold out. This blew me away. They are sold out for eight, like if I wanted to like redacted news, my little crappy show, like if I wanted to build an AI data center and, or even just use a part of compute, like with Amazon web services for my show or something, or you wanted to do it with, uh, with, with the Tucker Carlson network or something, Amazon has sold out through 2030. For capacity. Like we want a portion of this data center to be devoted to whatever. Sorry, we're sold out for through 2030 and we don't have enough capacity now. And we have projected to be building out thousands of additional data centers in the next few years. So when I said at the beginning, we're not even at the beginning of this and people are already talking about a bubble and this, no. And there's so many other pieces to this. I'm sorry. I'm just like, this is crazy. This is because there's so many pieces of this. You got the robotics piece of this. You have the information layer, then you have like the factory piece of this. Um, then you have the efficiency piece of this inside of these factories. Uh, then you have just the, the actual infrastructure to build out all of these different data centers. And then all of the companies surrounding those build outs like this blew me away. I don't know if you've heard of United, uh, have you heard of this company? No, it's just one example. So when you think about Amazon goes in and builds a new data center, well, or meta or Google or whomever they hire, they don't like come in there and like change the community. And like thousands of people get jobs. Like people come in for like about 18 months at most and they leave. So it's kind of like the gold rush or like, you know, St. Elmo's in Colorado ghost towns, you know, they come in, they do what they need and then they leave. So they hire these companies to bring in Caterpillar machines. And United rentals is one of those companies, right? So, and they're, they've got now sites in 1000 different locations. So, Hey, United rentals, uh, we're building out a data center in Reno. Do you have a location nearby? You can bring earth diggers, movers, generators, backup generators, all of that. Yep. We got you down. And so then they'll come in and bring all that equipment, build it out and then it leaves. That's just like one piece of the puzzle. Then you've got like the porta potties companies, like you've got so many pieces of this, this build out that's happening. And so people are like, well, it creates jobs temporarily. And a lot of those people come in for every job. They're not like creating new jobs every time, right? They're just, they just go to a location and they go to the next location. So not creating new jobs. And these data centers really only employ at the most around maybe 50 people. And even that number is dwindling. So, sorry, there's just so many pieces of this that it's, it's like, uh, there's, and there's just like a lack of information out there about it. Which I just find it just so fascinating and frustrating and terrifying all at the same time. And maybe also a little bit silly. I mean, this is happening so fast. And as you've pointed out, it's so complex. And, but it's also so expensive that you have to ask like, well, what is this? I mean, it's an effort obviously to snap the control grid into place and, you know, eliminate dissent. Got it. It's an effort to stay on par with China. Got it. But it's also an investment opportunity for a lot of people. And it's the last hope of states like California, which have been completely hollowed out. I mean, the tax base in California is like disappearing. And so their hope is that AI is going to hit, this is all going to be incredibly profitable and they'll be able to keep their welfare state afloat. Is this a good bet from a business perspective or is it a bubble? And if it's the former, how exactly does it make money? I don't understand that part of it. Where's the revenue? So this is one of the, yeah, no, this is a great question. And so when people say we're in an AI bubble, they don't know what they're talking about. I believe right. When Mark Cuban says, you know, we're, we're building out so many of these data centers and we don't need them. And they're going to be turned into pickleball courts. I, I believe he doesn't know what he's talking about. I don't, I don't from every source I've spoken to we're nowhere. We're not even at the beginning of this. And the demand for compute is off the charts from different companies. So I don't know what he's talking about. Like these are going to be converted to pickleball courts. Um, I think of anything, my theory, my theory is that these are going to be turned eventually. We don't need them these because they're cement and steel. These are going to be turned into for-profit prisons. Uh, that's just my theory, just throwing it out there. Um, they're out in the middle of nowhere. They'll be converted and, uh, we need them badly. And of course the Trump administration loves the for-profit prison system. So, um, I just think that that's a perfect transition, not pickleball courts, but anyway, so the, the, so, right now we're at the beginning of this and I was going through earnings reports, trying to find over the past few weeks. The reason I say we're at the beginning of this, because I have not seen yet from companies on their earnings reports, where they have pointed to the AI build out in their factories, equaling a percentage of revenue to answer your question. There's very few of them right now, but I believe that that wave is coming. So I think, and I'm not being hyperbolic about it. I really believe that the wave is going to end. That will increase the demand a hundred times. There are a few companies that have now said in some earnings in small measure, their AI build out has increased revenue. So for, I'll give you an example. There's like a, an outdoor furniture company, and I was going through their reports and they were explaining that with their AI layer that they built in, they were able to sort of, I don't know, bunch furniture together in a certain way that was more aesthetically pleasing at their stores. I don't know exactly. And they were able to increase revenue. It also increased the proficiency of their sales associates. So it didn't cost jobs. It actually increased their sales associates proficiency by like 82%. They claim leading to about a 15% increase in revenue. Might have the numbers slightly different, but you get the point, right? Like they've been doing something the same way for 30 years. And then suddenly they add in this like AI efficiency layer and suddenly they've increased their revenue by 15% without, they didn't hire anybody new. They didn't build a new factory or new warehouse. They just like bunched furniture together in a certain way that like increased something. Another company, I was looking at their earnings they claimed that with the scrap metal that they were able to save, like they put in an AI efficiency layer in their metal processing facility, in their factory. And when it would chop off pieces of metal, that would usually lead to about $129 million in wasted scrap. But by putting in this AI efficiency layer, they were able to save almost all of that, like saving hundreds of millions of dollars in non having like wasted scrap metal. So these are like small little anecdotes from like a few companies that have maybe shown some initiative with AI. And so if, if we see huge increases in efficiency and revenue game over, like, you know, there, there's no bubble, you know, like with the, the.com bubble, there were people building like fake websites, you know, what the, they didn't have anything. There was like nothing there. It was like fake. And there were so many greedy people who were building all these like fake websites. I remember it well, like pets.com, you know, all these whatever. And it was all just fake and it was all greedy people. And so I don't know. I mean, I, I, I'm just gonna say, I don't know. I don't know if like, this is all greed or if these companies who've been doing something a certain way for 50 years in their warehouse, suddenly they add in like this AI efficiency layer and their revenue increases by 20%. Maybe they don't fire anybody. Maybe they actually hire more people. So anyone who says they know is I, I just, I don't know. Maybe Elon knows he seems to know, but I, yeah, I, I, and I'm the last person who would know. I'm happy to admit that. But whenever I see lemming like behavior, I suspect the cliff is near, you know, whenever, you know, every midway. You run across is saying the same thing about the future. I, I generally bet that's not what the future's gonna look like, but you know, maybe I'm the midway. Um, but I have noticed all of a sudden when the guy from shark tank is like talking about stuff, he doesn't understand the future is data center. So I'm like, ah, I think The future is probably not data centers, but, you know, who knows? Maybe I'm just reactionary. No, can I say this? I think the future is human connection. So I think one of the, I'm always a glass half full kind of person. I think one of the benefits of this, despite all of its downsides, is the human interaction layer. And you're already seeing it, actually. Restaurants are reporting that, like, young people are coming in and they're not drinking. Number one, they don't drink alcohol. And they're coming in, they're bringing, like, a stack of, like, playing cards. And they're just hanging out all night, like, or a board game. They're bringing Scrabble, you know? And they don't bring their phones at all. And they just come in and are together. And I think there's going to be an insatiable desire for human interaction with this explosion of AI. Call it slop, call it fakery, whatever you want to call it. I think there's going to be a, really an insatiable. There's going to be an insatiable desire for human connection and human interaction and human content. You know, people to be able to say, I'd rather watch Tucker's show because it's real human beings having a conversation than the fake AI-generated garbage that, you know, maybe Hollywood is putting out. So, you know, I think that's one maybe benefit is that among young people, you're seeing this shift. And I hope that that's the case. I don't know. I don't know. I don't know. I'm more troubled about how this is going to affect young people than anything. I see my kids. My son just turned 16 and, you know, and he's going to go into the trades because so many of these jobs are just being taken by AI. And they will be. And if you don't think that that's the case, you're fooling yourself. Yes. I know all these Wall Street types, the Shark Tank types are like, this is going to create way more jobs. And we don't know. Maybe they're right. And maybe there will be so many jobs created to help the AI. Maybe there will be so many jobs created to help the AI efficiency layer or help the robots, you know, help the robots and all of this piece of it. I don't know that there's any evidence that that's true. I think most of the jobs that you can name, you can title yourself, I'm an accountant, I'm a pilot, I'm this, I'm that are going to be replaced by AI. I mean, at what point soon will FedEx, all of their aircraft be run by AI instead of pilots, or maybe because of a pilot's union, there'll be one pilot or two pilots. Or maybe because of a pilot's union, there'll be one pilot or two pilots. um And there are the arguments, well, maybe we shouldn't be connected in such a way to these AI models that are coming from China, that are coming from the United States, that are coming from open AI, Anthropic, etc. Get open source models that you can then store on your own compute in your own house. So you have all of the wisdom self-contained. And I know there are some libertarian friends of mine who believe that that's the way to go. It's like you have all of this compute sort of living, I don't want to say like Hillary Clinton's style, on your own servers in your house. But you're not relying on China. You're not relying on the government. You're not relying on these companies. You're relying on yourself. And if you want to do some deep research on a project, you have access to it because it lives on your own server in your house, disconnected from those people. I don't know. I don't have that. Without getting too far afield, however, I mean, I think one of the biggest things. It's happened in technology in recent memory was the hacking of cold wallets and the theft of Bitcoin from cold wallets. And the promise of a cold wallet was it was disconnected from the Internet. It was it was wholly yours. No one could put a finger on it other than you. And that was the assurance that your Bitcoin was actually yours. And so to see that get hacked somehow suggests that anything with an electrical impulse is not really under your control. Well, we even saw the story, was it three weeks ago about the AI model that was offline, was not even connected to the Internet and managed to somehow connect and get all and hack, jump online to hack and then jump back offline. Right. And so and then add in the robotics layer to this on top of it all. Right. So the humanoid, the humanoid robot piece of this. Where? You know, you're going to have this. You're going to have this in warfare. You're going to have humanoid robots. I mean, it's straight. It's it literally is straight out of Terminator. Yeah. You know, this is what's coming. And who can stop it? Is there going to be some sort of a kill switch and the ability to turn this stuff off? I don't think so. So I don't know. Are we going to build compounds with, you know, corrugated steel walls and turrets on our houses now to protect us? I don't know. I don't think so. I think we're going to have to do that. you

Podcast Summary

Key Points:

  1. AI data center construction is rapidly expanding across the U.S., with thousands of new centers planned, driven by massive corporate investments (e.g., Amazon, Nvidia) and government support.
  2. These centers impose heavy local costs
  3. Federal actions, including a Department of Justice AI litigation team and relaxed environmental permits, pressure communities to accept these projects, limiting local resistance.
  4. Data centers often use self-contained power (e.g., revitalized coal plants, small nuclear reactors) but still strain grids, leading to projected blackouts and higher consumer energy bills.
  5. The AI build-out is tied to national security (e.g., FCC rules requiring 65% of robotics components be U.S.-made) and a race with China, with demand for compute sold out through 203
  6. Economic benefits are uncertain
  7. Local pushback exists (e.g., New York bans, Texas moratoriums), but many communities are overwhelmed by tax revenue incentives and legal pressure.
  8. The broader impact includes data privacy concerns (e.g., Zoom transcriptions stored on AWS) and a potential cultural shift toward valuing human connection amid AI saturation.

Summary:

The conversation explores the physical and societal implications of the AI data center boom in the United States. The guest argues that the build-out is in its infancy, with thousands of new centers planned, each consuming enormous resources—particularly water (up to 500 million gallons per center) and electricity—leading to higher utility bills, water shortages, and projected blackouts in communities like Homer City, Pennsylvania. He highlights federal support, including a DOJ AI litigation team that sues communities opposing data centers, and relaxed EPA rules that accelerate construction.

Data centers often claim self-sufficiency using revitalized coal plants or small nuclear reactors, but this power is not shared locally, so residents still face rising costs. -made, intensifying the build-out. The guest sees no clear economic justification yet, noting some companies report efficiency gains, but warns of widespread job losses in fields like accounting and aviation.

He criticizes the lack of public consent, the surveillance risks of AI-integrated tools, and the potential for humanoid robotics, while suggesting a silver lining: a renewed desire for human connection. Ultimately, he remains uncertain whether this is a bubble or a transformative shift, but stresses the immediate, tangible costs to ordinary Americans.

FAQs

The main concern is the massive resource consumption—electricity, water, and land—by AI data centers, which drives up energy and grocery costs, causes water shortages, and impacts local communities, while communities have little control over the build-out.

The DOJ's AI litigation team can sue communities that impose regulations on data centers, effectively forcing them to accept these projects, which reduces local sovereignty and prioritizes private company interests.

Most data centers use evaporative cooling, consuming hundreds of millions of gallons of water, which strains local supplies, especially in the West, and competes with ranchers and homeowners for scarce water resources.

Data centers are built there primarily for tax revenue and financial incentives, as they bring significant local tax income, despite not having enough resources, and companies often claim to be self-contained but still increase local energy costs.

No, data centers employ only about 50 people on average, and construction jobs are temporary, with workers moving to other projects, so the long-term employment benefits are minimal.

The speaker argues it's not a bubble, as demand for compute is sold out through 2030, and early AI efficiency gains in companies are increasing revenue, though they admit uncertainty about the future.

Chat with AI

Loading...

Pro features

Go deeper with this episode

Unlock creator-grade tools that turn any transcript into show notes and subtitle files.