AI's Power Grab: Worse Than You Think

Lynn Parramore: why AI is driving up your bills, your interest rates, and your health risks—while warring against the human spirit.

Lynn Parramore discusses her recent article on AI’s real costs

Whether or not you’ve ever used a chatbot, you’re already paying for AI’s monstrous appetite. Lynn Parramore reports that, from surging electricity bills in Virginia to proposed nuclear reactors in rural North Carolina, data centers are driving up costs, inflation, and interest rates for everyone. Meanwhile, tech and utility executives cut sweetheart deals, transfer wealth upward, and promise jobs that never materialize.

Young graduates feel discarded, communities are being bulldozed, and the industry’s answer to every concern is more unproven technology and more backroom chicanery. This isn’t just an economic debate—it’s a battle between human reality and a machine-driven vision that would cover the earth in data centers and call it progress.

Main Points

1. The costs of AI today are widespread and unavoidable: Higher electricity bills, inflation, interest rates, housing costs, and health impacts will hit nearly everyone, regardless of whether they use AI or live near a data center.

2. Ordinary people are being sacrificed for corporate profit: Utilities and AI companies strike backroom deals, transfer wealth upward, and hide behind promises of jobs while communities and younger generations bear the real burden.

3. The industry operates without transparency or accountability: There’s a lack of accurate data, industry co-authorship of academic papers, and complex financial structures designed to let companies walk away from failures.

4. This is a spiritual and human crisis, not just an economic one: The real question is not what AI produces, but what it does to people: undermining health, replacing human judgment, and abandoning an entire generation.

Transcript

Richard Eskow: The article is headlined “AI is Hungry for Power and You Are Footing the Bill.” We’ve seen some headlines about that, but I think the reality of AI’s energy usage has been obscured—first by unclear reporting from the mass media, and second by a slew of counterarguments from AI executives and their political allies. I’ve been following this story for years, but I still learned some things from your article. So why is it that, no matter how this plays out under current circumstances, consumers—or human beings in general—will end up footing the cost for this monster’s energy?

Lynn Parramore: Yeah. A lot of people following the news are already aware of some direct costs communities are facing. Take Northern Virginia, for example, which is part of what’s called the PJM grid covering several eastern states. Several reports show that customers there are already paying higher electricity rates, largely because of data center build-out.

It’s hard to wrap your mind around the scale of these things and the amount of electricity they use. One facility—a “hyperscaler” facility—can consume as much electricity as a small country. That’s what we’re talking about. Northern Virginia already has a few hundred of these.

And Erin Brockovich—whom we all remember from that wonderful movie starring Julia Roberts—is actively following this. She has a great map where you can see where data centers are already operational or proposed in your community. No matter where you live, there’s probably a proposal for a data center not far away. They’re popping up everywhere you can think of. It’s going to be very hard to escape the direct effects: higher electricity costs, possibly higher water costs. But the indirect effects will affect us all, even if we’ve never used ChatGPT.

Why? Because it will contribute to general inflation. My colleague Servaas Storm, an economist, has written about this. Any time you have energy usage this massive, it spills over into the broader economy. Small businesses, for example, will use the same amount of power but pay more. They’ll be forced to pass those costs on to consumers. Farmers will have to pay more. They’re already facing pollution and strains on water, which raises irrigation costs. All these pressures on the system—electricity demand going up from data centers—impact manufacturing, retail, office buildings, hospitals, everyone who uses electricity. And we’ll all end up paying.

At a time when inflation and utility bills are already tremendous worries, affordability has become a catchword. I think this is going to be a real political flashpoint—there are signs it’s already becoming one. It’s interesting because Americans are constantly described as horribly divided, but Gallup polls suggest we’re not divided on whether we want data centers built near us: over 70% of us do not want them. And while there’s been a lot of talk about using nuclear energy to power AI demand, more than half of Americans don’t want new nuclear facilities near them either.

I’m currently in North Carolina, and there’s a big fight in Stokes County, near the Virginia border. It’s a rural area; I don’t think any town in that county has more than 8,000 people. Duke Power has proposed building small nuclear reactors there.

Richard Eskow: Right.

Lynn Parramore: And what do we know about these small nuclear reactors? Not much. Only a couple exist—in China and Russia—and I wouldn’t trust the safety data coming from either of those places. The farmers in that county don’t want them built there. We’re also talking about repurposing Three Mile Island—still a living memory for a lot of people. So a lot is going on, and people are upset.

The bottom line question is: What exactly are we being asked to pay for? What do we—as citizens, workers, people living in communities—actually get out of it? We’re being promised everything from jobs to saving humanity, but it’s not clear to me that ordinary people are getting much out of any of this.

Richard Eskow: No, I think that’s a great point. You raise several interesting issues. One is the nuclear angle. The industry’s pitch was that these units could be tractor-trailer size—like a boxcar—and produce so much power. But when I looked into it, the power they claimed would supply the needs of about three to four hundred homes. That’s not very much. How many of those would you need? And it’s untried technology—we don’t know the health implications.

Your article also pointed out two broader issues beyond energy costs. One I’d thought of: the industry says, “We’ll go off-grid and build our own energy supply.” My response has been, “Are you going to build your own ecosystem too?” Because you’ll still pump pollutants into the air and water until those theoretical nuclear reactors come online—and then God knows what will happen. It’s not like these are people known for keeping promises.

The other issue—which you just mentioned—is the broader economic impact. I’d thought about farmers, but let’s talk more about that broader impact, because I think it’s crucial for people to understand.

Lynn Parramore: Yeah. And it’s not getting talked about enough in the news. It’s not just higher electricity bills if you live near a data center—though if you look at that map, eventually not many people will be left out of that equation if things keep going as they are. But even if you’re nowhere near a data center, rising electricity prices spill over into the broader economy. When inflation ticks up, the Federal Reserve will keep interest rates high. That means higher borrowing costs for everyone: car loans, mortgages, credit cards. All of this gets impacted by the building of massive data centers. So it’s a problem for every single one of us, at a time when we can’t afford higher prices for anything.

Richard Eskow: Right. And you mentioned—this was an excellent point from Servaas Storm—that if a data center is built in your 8,000-person town in North Carolina or elsewhere, people will move in to service it, and your housing costs will go up. And any carbon-based or fossil fuel energy requires raw materials or oil. So the ripple effects go on and on.

Lynn Parramore: On and on. The ripples go out into the broader economy, and no one comes out unscathed—except power company executives and executives at places like Meta and Amazon.

Richard Eskow: Right. And that brings up what you call an “axis of influence”—not only the AI executives profiting from this, but also energy executives and utilities.

Lynn Parramore: Yeah, they’re very happy with these prices. Reports show executive compensation at utility companies has been going up lately. Why is that? They’re clearly benefiting from higher prices. There’s a lot of lack of transparency, backroom deals, sweetheart cuts that communities have no access to. And of course, we citizens aren’t getting any sweetheart deals out of this.

Richard Eskow: There’s also an awful lot of AI backroom trading among companies themselves. That really concerns me. I wish someone would commission a group to look into it in detail—it would take a lot of legwork. All those loans, deals, and commitments backed by nothing concrete. It feels like a bubble within a bubble.

Lynn Parramore: Yes.

Richard Eskow: Meta commits to purchasing $30 billion worth of ChatGPT, and ChatGPT commits to X billion from another company. That’s going around a lot. Plus, as you mentioned, a hyperscale center costs about $11 billion to build. That’s a lot of money moving in and out of the regular financial system, inflating rapidly, isn’t it?

Lynn Parramore: Absolutely. You’re seeing complicated business structures and engineering. For example, some big corporations look like they’re putting up the money for data centers, but they’ve actually created a separate company that on paper owns the facility. That means Meta or Amazon doesn’t have to carry it on their balance sheet, giving them a nice opportunity to walk away if things go badly. So there’s a lot of shenanigans and chicanery going on.

It reminds me of past disasters—the build-up to the financial crisis, or R.J. Reynolds telling us tobacco was safe while hiding the research. That leads to another aspect: getting accurate information. The scientists and academics who would typically be working on this stuff have, in many cases, lost their jobs under the current administration and gone off to work for the private sector. And if you look at academic papers, more often than not, industry folks are co-authoring. So we have a real problem getting accurate information about any of this.

Richard Eskow: Yeah, absolutely. You quote a law professor, Mark Klick: Why aren’t companies paying for the additional resources they use, especially energy? We still live—at least under the illusion—that even a private power company operates as a public utility. Why would a company supposed to act in the public interest cut a special deal for a for-profit corporation at the expense of the citizens it’s supposed to be serving?

Lynn Parramore: It does sound like a transfer of resources. One section of my article is called “Trickle-Up Economics.” This is how it works: a transfer of wealth and resources to the top. Citizens are in a very vulnerable position because of the influence of money in politics. Why is this happening? I’d point to my colleague INET research director Thomas Ferguson, who’s followed money in politics for many years. Just look at the coalition of industries looking to make money off this, their lobbying power, and the politicians in their pockets. It’s on a big scale—we’re talking trillions of dollars now, not billions.

Richard Eskow: So let’s talk about that. One of Thomas Ferguson’s observations—and I strongly recommend his work—rebuts the whole idea of “NIMBYism” (Not In My Backyard). That term is often thrown around by people whose own backyards aren’t at stake anyway, but who have something to gain from corporate interests. NIMBYism is being used in this AI debate too, isn’t it?

Lynn Parramore: Right. These companies say, “If you’d all stop complaining about data centers in your backyard, we could have lower electricity prices.” I spoke to Professor Ferguson about that. He was very dismissive. In his work, it’s very clear that power companies and big corporations are keeping prices up—not ordinary citizens by any stretch.

Richard Eskow: Yeah. And it’s totally reasonable not to want an energy-guzzling monstrosity in your backyard. I wrote a few years ago about a historic African American community in Northern Virginia that was slated for a data center. Eventually the state backed off because it was just too embarrassing. But many other communities have been bulldozed or moved aside for high-power transmission lines and all the infrastructure needed to support these data centers. One thing that blew my mind in your article, Lynn, was a study saying that in 2024, data centers were already using some 40% of Virginia’s power. Did I remember that correctly?

Lynn Parramore: You did. Already in 2024. And I need to look at the incoming data for what’s happening now. Again, it’s a mind-boggling scale. That’s why calls for a moratorium, to me, are the bare minimum of what we should be discussing. It’s happening too fast, it’s too big, and the consequences are too great. It’ll be interesting to see what happens politically, because some states becoming AI hotspots, like Texas, are not traditionally blue states. Neither Democrats nor Republicans want to see these data centers in their communities.

Richard Eskow: Which seems like it would be a good issue for Mr. Talarico to take up.

Lynn Parramore: For sure.

Richard Eskow: Since we’re talking about communities and the nonpartisan resistance to data centers, what’s it actually like when a data center moves into your town or area? What does it do to people?

Lynn Parramore: Well, one of the first things, especially in poor rural communities, is excitement because of the promise of jobs. In places where people desperately need jobs, unfortunately, research shows there really aren’t a lot of jobs for ordinary people. There are initially some temporary construction jobs. But if you live in a small town in the middle of nowhere in northern North Carolina, you’re not going to be hired to work on a small nuclear reactor. It’s highly unlikely.

Richard Eskow: No. They’ll probably drag it in on a flatbed truck or train from somewhere else, and you’ll get nothing out of it. And you mentioned the impact on farmers and others.

Lynn Parramore: The jobs promise is held out as a carrot in places where people desperately need them. So even though they know there may be negative environmental impacts, when all you can think about is feeding your kids, those job promises can be very effective in getting people to overcome their fear of other consequences. We’ve seen that with other kinds of environmental impact and pollution.

But what it ends up being like to live near one of these centers as an ordinary working person is to be driven crazy. Not only the sound—which can be very disruptive—but there’s actually a sound at a level you’re not even conscious of hearing, yet it still affects your sleep and circadian rhythms. That impacts your health. There’s another cost: your well-being, your health. You’ll likely be liable to all kinds of negative physical and mental health outcomes from living near these things. How is that going to benefit anybody?

Richard Eskow: Right. I’m starting to picture—and I don’t mean to be lighthearted about this—a kind of 1950s horror movie. Between the impacts you’re aware of and the ones you’re not, it’s really appalling. Which brings me to Ilya Sutskever, former chief scientist and co-founder of OpenAI. He says extreme things. You quote him from a 2019 documentary: “I think it’s pretty likely the entire surface of the Earth will be covered with solar panels and data centers.” I tried to picture that, and what came to mind was the Death Star from Star Wars. On one hand, you could say he’s a brilliant technician who’s nuts. But I know from friends close to some of these tech billionaires—this is how they think. In terms of glory, world-transforming, terraforming, post-human.

Lynn Parramore: Absolutely. It gives them no hesitation to say things like that. That’s part of the terrifying part. There’s this disconnect between these tech gazillionaires and normal human life. The scale of these data centers reflects that. And what he’s talking about—data centers covering the Earth—look at Erin Brockovich’s map of existing and proposed data centers in the U.S. It looks a lot like that. That’s what’s being proposed and imagined.

It’s funny you mentioned science fiction and horror movies. As I’ve been writing about AI, I’ve gone back all the way to Metropolis—the warnings we’ve had. My latest spree last weekend was watching the Resident Evil series, which involves an AI called the Red Queen who unleashes havoc, and the workers pay the price. It’s a fear we’ve had for a long time, and it really does seem to be manifesting at scale. The horror is scaling up for most of us.

Richard Eskow: In a sense, this is the latest and most powerful manifestation of something that’s been going on since the Industrial Revolution: a battle between impersonal capital and human reality, the human spirit. A war between the machine and the spirit. I don’t mean machines can’t be great benefits—typewriters, word processors, recording equipment—they can. But Metropolis is a perfect example. Moloch, the god fed human sacrifices, is an extreme depiction of something that’s been going on for centuries. AI has brought it all to the fore. That’s why Sutskever’s quote is so powerful. Although he doesn’t seem to see anything wrong with it, it’s the complete annihilation of nature and the things we value—the intangible, the beautiful, the fragile, the evanescent—in order to create these computational monstrosities they worship, because they ascribe consciousness to something that is not consciousness. That goes back to Metropolis, back to the story of the golden calf that the wind blew and it lowed like a cow, and they thought it was alive.

This is an internal human struggle. You touch on that in your article too. You mention that it’s hard to stay healthy with these whines going on. And though you talk about economics, the quote from Amartya Sen—someone I’ve admired for a long time—is the money quote. You write, “As Amartya Sen has argued, the central question is not what the system produces but what it does to people.” Whether it’s financially, or the ever-present whine, or the threat to jobs, or just the dehumanizing, migraine-inducing profusion of AI slop—I don’t think any large language model or software has to be anti-human, but it seems as if this entire industry is determined to build and grow in an anti-human way.

Lynn Parramore: It’s true. One thing that concerns me: in New York City, where I live most of the time, I’m not far from Washington Square Park. I was there a couple of weekends ago, and all the graduates were out in their purple gowns. I couldn’t help talking to some of them, and boy, was that depressing. This generation—fairly affluent NYU students—are very scared, disheartened, dispirited. I kept getting the sense that they feel like they’re being thrown away as a generation. They’re being told that because of AI, they’re not really needed in this economy. There won’t be entry-level jobs for them. They won’t be mentored up through a company or organization. Forget it. Good luck.

To me, that’s filicide. If we abandon an entire generation to AI, what kind of sacrifice is that? It doesn’t sound so extreme to talk about films like Metropolis—we are actually doing that. And what solutions have been proposed to these young people? Where is the New Deal-style jobs program for them? Training programs? That’s what they need right now. They’re graduating into a terrible economy that’s likely to get even worse. We cannot afford to let this whole generation go.

I’m wondering, Richard, how you feel about this. In the 1960s, we had a lot of youth resistance, but that was possible because of the affluence and optimism of the time, which we really don’t have now. Young people have been so isolated—the pandemic isolated them. They’re not with each other as much, not dating as much. A lot of them are at home gaming. But you can’t be isolated in an apartment when there’s no money and no job. So maybe they’ll be forced into community with each other and find a way to push back.

Richard Eskow: The French Situationists in 1968 had the slogan “Beauty is in the streets.” Let’s hope that’s the case. You’re absolutely right. The 1960s protests were triggered by several things—you have to distinguish Black and white protests—but the white anti-war protests were based on relative affluence plus a sense of optimism. The curve was slanting upward, not downward as today. Interestingly, in the early part of this century—late 2000s or early 2010s—as the economic situation improved in Brazil, the number of protests escalated considerably. My theory is that there was increasing optimism: people thought, “The future can be better here—let’s get involved.” That’s the challenge we have now. You need something to push against and something to move toward.

My big criticism of the “resistance” during the first Trump administration was always: Don’t just tell me what you’re against, tell me what you’re for. It’s funny that Erin Brockovich comes into this. I met her in Lower Manhattan when I went to rent a car to go upstate. Hertz was out of cars—didn’t matter that we had reservations. There was a line of people. “Sorry, no car.” I was third in line. The woman first in line was arguing with them, and the person behind the desk said, “What was that name again? B-R-O-C-K-O-V...” and everyone realized who she was. There she was—a person fighting for people’s interests—and there was the impersonal corporation. The guy behind the computer monitor had no control; he was as stuck as we were. But you don’t want to be that guy, stuck behind the terminal, working for people who have no compunction about doing this to you. This is all of that on a meta-meta-meta level—at the level of the planet, at the level of humanity as a whole.

Lynn Parramore: Right. What’s it doing? The young people had so many different things to say: “What is this doing to my mind? I’m losing my cognitive skills. What is this doing to the environment? What is this doing to my future in my profession?” So many worries all at once. It’s terrifying. I kept thinking, this generation—it sounds extreme to say it, but it’s almost like graduating into the Great Depression. It’s a bad scene from so many angles. What loyalty do you have to a system that isn’t offering you anything?

Richard Eskow: Absolutely. Elon Musk shoots hundreds of rockets into space, permanently altering the night sky, with nobody to stop him. The way they’re treating our younger generation reminds me of what the poet Robinson Jeffers wrote: “They’d shit on the morning star if they could.” That’s who we’re up against. And who we’re for are the people who affirm beauty, humanity, hope, and community. That’s the pain and the beauty.

Lynn Parramore: And who live in these human bodies. We are embodied intelligence. No AI is ever going to replace us.

Richard Eskow: Right. We’re embodied intelligence, and we have whatever you want to call it—psyche, spirit, soul. If you don’t believe in soul, “psyche” is probably good. Consciousness. All of this matters. And AI is not conscious.

Lynn Parramore: No. And it’s not going to replace us. One of the things I hope young people will absorb is that the narrative that they’re replaceable is a lie. No computer program can replace their human experience, life, judgment, insight, and imagination. I really hope we can promote that message: they don’t have to buy into that narrative.

Richard Eskow: Absolutely. Beautifully said, so I won’t add anything—no quotes, nothing. As always, I encourage people to read the article again. The title: “AI is Hungry for Power and You Are Footing the Bill”—to which I would add, with not just your checkbook, but your body and soul. It’s at ineteconomics.org. Lynn, thanks for writing this excellent piece and, as always, thanks for coming on the program.

Lynn Parramore: Thank you so much for having me. I look forward to talking about this more.