Explained, Data Center Boom, And Who Is Really Paying For AI

[00:00:00] Speaker 1: You've seen the headlines, and maybe you've even seen the construction from rural towns to suburban neighborhoods. Completely bypassing the voice of the people. To bustling cities such as Miami and New York. To West Virginia, just to name a few. The latest gold rush is AI and the infrastructure to support it. [00:00:22] Speaker 2: You can see the skyline, condos, and then right here, a massive data center that takes up several city blocks. In communities across the country, everyday Americans are taking up a new fight against data center developers. Data centers in the U.S. are growing at a record-breaking pace. Construction is at a new high, according to the U.S. Census. Roughly 4,000 exist right now. There's at least one in all 50 states. 3,000 more are currently being built. [00:00:49] Speaker 3: Look at what a data center is. It's a massive facility, you know, think a warehouse, that runs and houses large computer systems. That's essentially what, you know, they are. The United States has more than 4,500 active data centers with about another 1,500 in the development pipeline. You know, in terms of growth, you know, that's kind of simple, too. I mean, more and more we're becoming a digital economy, you know, and we can see that in our daily lives. You know, our digital footprint is growing at a rapid scale. And, you know, so all that data needs to be housed somewhere. You know, think of cloud computing and artificial intelligence. You know, both are leading the surge in the need for the construction of these data centers. You know, by 2030, in just four years, the growth of AI, you know, the investment in that for the data centers could reach $7 trillion. [00:01:42] Speaker 2: It really is a conundrum because we all use chat GPT, it seems like. We all have AI in our lives. We all have the photos on our phones stored on the cloud. So you need data centers for all that storage. But at the same time, so many people don't want these massive data centers right in their backyard. [00:01:58] Speaker 1: Washington is advocating speed-up construction. But as the U.S. scrambles to win this global AI race, are regulations being scaled back? Who is profiting from the boom? Is it you? Is it those in your neighborhood? And who is covering the bill for those tax exemptions and the skyrocketing energy bills? [00:02:16] Speaker 2: All you hear is negative, negative, negative, negative, negative. [00:02:21] Speaker 4: Beltline is not our savior here. They're not our answer to our problems. You don't need any more research. Your people have spoken. We don't want a data center. [00:02:30] Speaker 3: On one side, though, you have what the potential benefits of the economy will be. A boost in jobs, specifically construction-related jobs, probably, with all the building of the facilities. There's billions in contributions to the GDP. The capital investment in these data centers is literally nothing like any of us has seen in our lifetime. You know, that could probably be said for your parents and grandparents. You know, the investment as a percent of GDP is the biggest thing since the Louisiana Purchase and five times larger than the U.S. highway system. You know, but on the other side, you know, they can be noisy. They consume a tremendous amount of water and electricity, and, you know, of course, that can affect your wallet. You know, communities in close proximity have complained about light disruption. You know, think exterior light. You know, these big facilities need big security parameters. So think floodlights and the like. And then there's the environmental and health effects that, you know, we really don't know, you know, what they are now. [00:03:43] Speaker 1: You know, we all know this, this comes at a cost. Every chat GPT prompt, every AI search or every AI photograph has to live somewhere. And that somewhere is a massive data center. Unlike a Google search, an AI search generates an answer performing an enormous number of calculations. [00:04:02] Speaker 3: The energy cost of a single, you know, say, chat GPT prompt, you know, is really small. It's a fraction of a watt hour of electricity, you know, but a single chat GTP request, you know, requires 10 times more electricity than a Google search. So then the concern comes, you know, when you're talking about increased, you know, electricity costs and the drain on potential energy is when you multiply that by tens of billions. So, you know, you take your, you know, what work, what I just said is one prompt now multiply it, you know, and then nationwide and then globally. And then what are the costs related to that? [00:04:43] Speaker 1: We don't use a server farm anymore. That's a collection of servers housed within a facility. So, of course, think the equipment itself. A data center is a physical facility or a building that houses computing infrastructure with these massive facilities. [00:04:57] Speaker 3: You know, they require a ton of equipment. And, you know, when they're churning 24-7, you know, you know, in terms of, you know, I'll talk about water. Like, you know, when they're churning 24-7, the room temperature gets real hot and you need a cooling strategy. You know, think of your car. You know, what would happen to your engine if you didn't have any coolant? You'd overheat. You also have spent years working in a television, you know, studio. It's cold in there, you know, most of the time, you know, because there's a lot of equipment, a lot of radiant heat. The International Energy Administration, you know, they did a study and found that just 100 megawatt U.S. data center, you know, which is a really big center, can consume 530,000 gallons of water per day. You know, to put that in context, that's the average daily water use for about 6,000 American homes, you know, and, you know, it's a little less than one Olympic-sized swimming pool, you know, what the amount of water that can hold. And then there's the electricity, what we got to, you know, right now, nationwide, it's using, collectively, it's using about four, four and a half percent of annual U.S. electricity consumption. And, you know, they're estimating that could double or triple in two years' time. [00:06:10] Speaker 2: Proponents say that's great news for blue-collar workers and the economy. [00:06:15] Speaker 5: Demand for workers in the trades is skyrocketing. [00:06:18] Speaker 2: But the people who already live near the data centers say it isn't as advertised. Once they're built, they don't employ many people and they suck up power. They're causing utility bills to skyrocket. Some data centers are right in the middle of downtown areas, like here in Miami, you've got a fancy condo building, and then you've got this giant data center. The neighbors here say what they notice is the hum. Listen. The hum of fans, cooling all the servers, power everything from AI, like chat GPT, to cloud storage. [00:06:48] Speaker 4: You come home, you get out of your car, and you just hear this noise. It's just like, you just want to curse, you know. It's that bad. Did they tell you it was going to be like this? No, we had no knowledge whatsoever that they were going to be making a, putting in their own power plant here. [00:07:07] Speaker 1: Look at your screen. These are the tech giants that are spending hundreds of billions of dollars building the digital infrastructure that we all believe will define the next economy. You see Amazon, Microsoft, Google, Meta, Oracle, CoreWeave, OpenAI, XAI, just to name a few. And we've reached out to all of these companies, and almost all of them responded immediately. Some, such as Meta, sending us a plan to ensure that the company covers the full cost for energy that's used, and consumers are not negatively impacted. So, think of new power sources, such as dams, wind, solar power. Microsoft, also showing us multiple sustainability reports. OpenAI, writing in part, OpenAI is committed to paying our own way on energy, so local electricity prices do not increase. And look at Amazon, they sent us a pretty hefty statement where they said they're focusing on developing our infrastructure in places where we can be a strong, long-term partner that delivers meaningful, local benefits. But the race isn't just tech companies or inside the borders among states, it is across the globe. Policymakers on both sides of the aisle say this is a matter of economic and national security, which is why the White House is easing federal permitting, advancing policies to fast track, private power plan approvals and directing agencies to use federal land, or the Ratepayer Protection Pledge that ensures tech companies to ensure private developers can fund private power plans. The White House has done a lot to accelerate construction. Governors are getting involved, state legislators, local economic development officials, real estate developers, lobbyists, and of course, campaign promises. So think sales tax exemptions, property tax abatements, discounts or fixed pricing on electricity, incentives worth hundreds of millions of dollars, and it is constantly coming up on the trail. [00:09:08] Speaker 5: For example, forcing data centers to pay for any added energy costs and not imposing it on the community and neighbors and small businesses around. [00:09:16] Speaker 6: We must prohibit them from building AI data centers in rural Texas neighborhoods, and we must eliminate the tax break they are getting. [00:09:26] Speaker 7: Data centers must bring their own power. They must reuse their own water. And they must reduce electricity costs for residential customers as well as small business customers. [00:09:37] Speaker 1: The argument is simple: Build here, create construction jobs, expand the tax base, and establish your state as an AI hub at no cost to us. Some candidates say construction is exactly what is needed. Look at Louisiana Governor Jeff Landry just approving construction on what he says is hope and an investment. [00:09:57] Speaker 8: This expansion brings Meta's investment over $50 billion, making it the largest data center development in the history of the world. [00:10:09] Speaker 1: Senator Tommy Tuberville, who is now running to be Alabama's next governor, is campaigning on the issue. He says it's the type of investment that his state needs. [00:10:17] Speaker 9: I support them because they bring in a lot of money to the states. I'm hopefully getting ready to be governor of the state of Alabama. We all need more money for infrastructure, education, health care, all those things. But I've investigated them. I've gone and looked at them. I've visited them. I've talked to people about them. I've talked to power companies. I just had somebody in my office, Department of Energy, talking about them. And it's the future. We're going to have them. Elon Musk even says in the future we're going to have them in space. But in the meantime, if it's going to bring in a lot of tax money and revenue to your counties and cities and schools and hospitals, sure, I want to do it. And we're a perfect place in Alabama. We have a lot of rural area. We don't have to put it next to a Walmart or next to a housing district. We can put it out where nobody even recognizes what's going on. [00:11:11] Speaker 1: It's coming up in rural areas such as Lubbock, Texas. [00:11:15] Speaker 10: Data centers are an existential threat to our neighborhoods and to humanity. They will replace your jobs with AI. [00:11:24] Speaker 11: Please do not hoist our descendants into a dystopian nightmare just because some people can't write their own emails. [00:11:31] Speaker 12: We are directly endangering our local agricultural economy, stunting our future housing development and threatening the long term drinking water supply of my generation and the children that come after me. [00:11:41] Speaker 1: The big question during construction is will data centers create more permanent jobs than the factories they are replacing? Let's look at a small data center that could range anywhere, let's say from five to 20,000 square feet. So it may require one to five megawatts of power to feed, let's just say, in theory, a hundred thousand servers. So maybe that looks like a Target or a Walmart superstore. But a big box store would spread its megawatts over thousands of lights and escalators and refrigeration aisles. The data center crams all of that power into compressed computer racks. But then you hear of some of these data centers popping up that are significantly larger. Sometimes they're called hyperscale. So Amazon, Microsoft, Google, maybe Meta have built hyperscale facilities. Those can draw anywhere from 100 megawatts to five gigawatts. So the energy needed to feed a city. Millions and millions of gallons of water every day will be needed to cool those servers. So utilities are racing to build new power plants. Transmission lines are expanding. And in some communities, residents are asking whether their electric bills could go up. While rapid expansion is happening from Arizona to Georgia, Iowa, North Carolina, really across the entire United States, Texas and Virginia are taking the lead. Texas has more than 960 data center projects. It's got a very vast landscape, a power grid, wind, solar and fast track connections. Virginia reigns as the data center king with the highest number of operating facilities, more than 300, including Virginia's data center alley. But there is a new direct electric tax on data centers that are dissuading investors and tech giants. Critics argue that fast permitting could come at the expense of environmental review and local input. And look what's happening in New York. The governor there, Kathy Hochul, penned a data center moratorium executive order that delays new construction for up to a year. [00:13:40] Speaker 10: Data centers consume enormous amounts of power, truly threatening to outpace our grid's capacity. And they drive up costs for local rate payers. And I refuse to let those costs be passed on to New Yorkers, who already pay too much for their utility bills. The pause could last up to one year as the state works to create new rules for future data center development. [00:14:06] Speaker 13: It doesn't go anywhere near far enough to address the very real problems of needing more information, possibly needing more regulation, and also who does it apply to because there are loopholes and exceptions in the current executive order. [00:14:24] Speaker 1: In nearby Pennsylvania, for example, Governor Shapiro says company campuses such as Amazon should pay for their own electricity. The Trump administration has made accelerating AI infrastructure a national priority, directing federal agencies to streamline permitting for data centers and energy related projects. Supporters say America cannot afford to lose the AI race because of delays. [00:14:47] Speaker 14: We have a big signing right now and we have a tremendous industry where we're leading by a lot. It's the AI, artificial intelligence. It's a massive industry. We're leading China, we're leading everybody by a tremendous amount where the electricity is being built by them in every plant. They're building the electricity and we're getting them rapid approvals, but they're spending trillions of dollars and they can't basically this is a couple of other things that are less important. But one of the things that it has is you have to have a central source of approval when they need approvals of things, they have to come to one source. They can't go to California, New York and various other places like Illinois with Pritzker, who's, you know, totally unreasonable person and they have to have, you know, they're putting all this money in. It's a big part of the economy is only going to be one winner here and that's probably going to be the U.S. or China. And right now we're winning by a lot. [00:15:46] Speaker 15: Whoever wins, the values of that country will affect all of the AI. We don't want China's values of surveillance and centralized control by the communist government governing AI. We want American values of free speech, of individual liberty and respecting the individual. [00:16:05] Speaker 14: Not doing this would be the greatest gift to China and other countries, but China in particular, that China has ever received. [00:16:13] Speaker 1: There is little debate that AI is changing the world that we live in. The bigger question is whether or not America can build the infrastructure fast enough to lead without leaving communities, taxpayers and the grid behind. Because the new gold rush is not what's buried beneath our feet. It's what we're building in real time. [00:16:35] Speaker ?: you