Balance

Episode 3 of Power Play, Quartz's six-part investigation into AI and the US power grid.

Missed episode 2? Check it out here.

The county gave up 75 percent of the property tax on it for 15 years. Historian John Reeves compares what's happening today with what happened the last time America built at speed. Catherine Baab investigates the pledge signed by tech CEOs, which claims AI won't raise anyone's electricity bill.

Reported by Jackie Snow, Catherine Baab, and Joseph Zeballos-Roig.

Hosted by Marco Pimentel.

Produced by Simon Litt.

Edited by Georgina Kemp Walker.

Marco Pimentel: On March 4th of this year, the administration held its high-profile AI energy roundtable with tech CEOs inside the White House's literal Indian Treaty Room. So this is the exact historic site where the federal government negotiated and systematically broke resource agreements with indigenous tribes. So inside that room, tech giants signed this highly publicized ratepayer protection pledge.

Catherine Baab: I immediately sent it to John Reeves, the historian, just to make him laugh, because this perfect accidental metaphor, it's a discussion about AI power and expansion that's taking place in a room that is named for a history of treaties that the U.S. almost systematically violated. So where agreements are made to be broken, almost. It's just this moment of symbolic irony echoing this history of violation and literal legal theft.

Marco Pimentel: I'm Marco Pimentel, and this is Power Play. Over six episodes, we follow the race to rebuild the American grid for AI and the fight over who ends up paying for it. This is episode three. Today, we're seeing tech site selectors using similar linguistic tricks, framing rural American farmland as underutilized space just because it sits near a high-voltage power line. With me today is Catherine Baab, reporting for Quartz.

Catherine Baab: I talked to John Reeves about this. He's a historian and the author of Grant's Betrayal about the theft of the Black Hills.

John Reeves: That was one of the central things in my book. One of the arguments for taking the Black Hills from the Lakota was, and it sounds trite and I don't mean it to be, but a lot of the military officers and businessmen were saying they weren't using it. From our perspective, they're not using it. But they were using it. They just were not using it in a way that seemed recognizable to the people that wanted to mine the gold that was there. One thing I'm noticing just from afar on the data centers and things is that you go into a rural community and you're bidding up the price of electricity and you're putting stress on the infrastructure, and then you may or may not be there for a long time. So there's this kind of boom and bust component to it. I think that there's that similar kind of trade-off of just, hey, we want to do this thing and the people that are already there are not optimizing or utilizing this land and we're going to do it and we may move on or we may not.

Catherine Baab: When things are treated as inevitable, it immediately frames protest as useless, right? So it's preempting any protest. Like I would say in the case of Manifest Destiny that treated continental expansion as a moral right. Same with data centers. In both eras, you have powerful actors describing land as underused, available, or simply waiting for productive development, which obscures the existing claims and the local stakes at once. So Manifest Destiny in 19th century belief. It's basically a belief that white settlement of the frontier was an economic, religious, spiritual, political, historical imperative that we were entitled to, that this was the march of history. These things are explicitly revived. It's a story we tell ourselves about expansion.

Reeves: One continuity between then and now is not only that progress was seen as inevitable, but that progress was seen as a good thing. And it was almost a universal assumption that, hey, no matter what, economic progress is beneficial to everyone. We'll all get richer. And therefore, if the odd treaty is ignored or if Native Americans are mistreated, these things needed to be tolerated because the overwhelming agreement was on the notion that progress is good and that ultimately we'll all be better off even if there's a rough patch along the way. That inevitability in both cases leads to, I think, a powerlessness of just ordinary people like, look, there's nothing we can do about it and it's going to be better for everyone anyway. So why struggle against the tide?

Marco Pimentel: Except the people already on the land had a different read on what it was good for. They weren't confused. They just disagreed.

Reeves: One thing that was just a given was that the Lakota needed to become farmers, that that was the more advanced form of economic production. And this life of hunting and roaming and not being confined to one settled place, that that was backward antiquated and needed to be changed. So that's where a lot of the conflict lay, right? Because they did not want to farm, that they enjoyed this life that was built around hunting on the Northern Plains. They kind of knew that the land wasn't suited to farming, that it's grassland. It was much more suited to grazing, and the buffalo would graze the grass, and then they would be hunted. And that was central to their economy, whereas sort of trying to farm from a standing start without proper tools and without any kind of history of doing that, they recognized that as being kind of absurd. It sort of was a self-fulfilling prophecy, right? Because not only the politicians, but a lot of the clergy members would often see their unwillingness to farm as a sign of their backwardness, whereas in some ways it made completely rational sense.

Catherine Baab: Once it's framed that way, disagreeing with a plan stops being a position and it starts becoming a problem.

Reeves: Any counter argument was seen as a rebellion. When the Lakota were saying, hey, we don't want you to take them, they literally used the term for people that were putting forward that. They called them hostiles and then they made war on them to kind of soften the deal, so to speak. A similar thing, I think, too, is you see with the Standing Rock protests from a few years ago about the Dakota Access Pipeline. Resistance is seen as, someone who has a different opinion about economic benefits or progress or whatever has seen as like a critic or a rebel of some sort.

Catherine Baab: Almost, it seems like it's a feature and not a bug that this is about a belief system that's imposed on the land rather than arising from it.

Marco Pimentel: Before we go any further back, tie this to the present day data center issue for me.

Catherine Baab: I've spent a lot of time reading county filings in Ohio and Virginia. The pitch is always jobs and tax revenue, and the terms are usually settled before anyone in town even hears about it. Here's the thing about a build out. Somebody captures the gains and somebody absorbs the cost. But these are not usually the same people. The cleanest way I found to see that is to look at who got rich the last time.

Reeves: That depression of 1873, which was really harsh, right? There was no entitlements or supports for workers back then. So when a depression hit, it hit hard. And that was one of the reasons why we ended up going into the Black Hills was that Grant felt like, oh gosh, the poor working people are having a hard time. This could be an outlet for them to have them go out to the hills and pan for gold and maybe make something of themselves that way. And so that was putting a lot of pressure on him to allow this to happen, even though it went against the treaties of the time. One thing I often think about is with the Black Hills, right? When we seized the Black Hills, that ultimately led to economic destitution for the Lakota. But it also turned out badly for a lot of those white settlers that went out to the hills to first mine for gold. Because in some ways, the policy was on their behalf. But then it was the big corporations, one in particular, the Homestake Mining Company, that got the lion's share of all the wealth that was produced. And they bought up all the claims that these small-time white settlers. Who benefited? It definitely was some of the big capitalists and big investors. And there was this tendency towards monopoly in the economic system back then, which we see today, too. And you saw it in the fur trade, which there's an individual by the name of Pierre Chouteau, Jr., who emerged in the middle of the 19th century. And basically, the belief was, is, look, you know what? If there's free competition in the fur trade, nobody wins. So it eventually worked against the small-time producer, and eventually there was a consolidation in the name of Pierre Chouteau, who not only ran the steamboats up and down the Missouri, but he hired all the people to make the trades for the buffalo robes. And then the same thing happened with the gold miners as well. It required a lot of capital to run these enterprises. And then who paid for it? In some ways, the U.S. taxpayer paid for it, right? Because the U.S. government really was a partner in building the railways, and the U.S. military was sort of in cahoots with that, too. And we were building out the infrastructure that eventually allowed monopolies to benefit, right? And I think there's similarities with that today, right? The big tech has gotten really large and really wealthy. And a lot of it is a result of being domiciled in the United States and all of the intellectual capital that comes with that. There was almost a marriage of capital and the U.S. government in bringing about some of this progress out west then and now.

Catherine Baab: What gains they make are private anyway. I'm a capitalist myself. I love the stock market. I cover it. It's my beat. But I don't think that you get a public good argument necessarily. I think we should look askance at tech companies presenting themselves as public goods or public infrastructure when the gains remain private to shareholders. There's an old story in finance, where are the customers' yachts? It just means you should follow the gains, who gets them, where they sit, and question what folks' motives are. If the proper incentives aren't in place and you're going to get the behavior that results from proper incentives not being in place.

Marco Pimentel: The part that stays with me is that the benefits are loud and the costs are quiet. We all know the legend of striking it rich. We don't spend much time talking about the people who ended up worse off.

Reeves: The experience of the Western settler is a hard one in some cases. And I think our myth often holds that it was all positive and all just kind of the land of milk and honey type of thing. But I think in some ways, a lot of those stories were sad ones. Some of these theories weren't often tested in. And the notion that it was an overwhelming benefit to the mass of people, I think is questionable. And I think we're seeing more signs of that today with wealth inequality. The rewards are going disproportionately to a smaller number of people than the mass. It was such a boom and bust Wild West economy back then that when things went bad, they went really bad. And there was a lot of economic suffering. So oftentimes, I don't think workers' interests, they certainly weren't in the 19th century. They weren't put to the forefront then and they probably aren't now either. But is anyone really on top of this right now? Like how does this impact the labor force? I kind of think not. I'm a historian and I look at it and I'm just like, wow, there's a lot here to figure out.

Catherine Baab: Honestly, one of the things that strikes me again and again, now that I cover AI at least weekly, is that the economic logic is relentless. That is the battering ram. And everything else is just kind of a dressed up version of there is money to be made here. We think there is money to be made here. And then this is unstoppable.

Reeves: It's really hard to challenge whether it's digitization or AI or cryptocurrency. You're seen as kind of a crank if you go against what is seen as the future. There's just such a strong belief in faith in technology and progress that it's really hard to have a sustained and thoughtful criticism of it without being seen as sort of not on board with the national project, so to speak.

Marco Pimentel: So as a business journalist, how can tech companies project manage sustainable growth when the underlying commercial transaction is this intensely extractive for the local economy?

Catherine Baab: I'll offer an example. I live in Richmond, in Virginia and we have this county near me, Loudoun County, right? Where Google has a data center county-wide, they collect $900 million in data center property tax revenue, right? That was in fiscal year 2025. That sounds good on its face. And then you look at the tax incentives that were offered to the company and what those are worth per year. And in Loudoun County specifically, those are worth about $1.6 billion a year. So is the community netting out? It doesn't seem to add up. As a reader, as just an informed citizen, I think you follow the math. The math doesn't lie to you. When this information is publicly available, I say you look to that and make a call about what you think it's contributing locally. There's an example here that is very direct. So in the 1790s, you have Scioto, Ohio. It becomes part of this land scheme that's marketed to European buyers when the marketers didn't even have legal title to the land. So they sold it without even owning it. All the claims they make did not match the brochure. So now in January of 2026, you have officials in the modern day Scioto County, Ohio, coming out after this secret deal and saying that Google is going to build a $1 billion data center. It comes after county commissioners have already approved a giant tax abatement. And the people who actually have a stake in the land were never offered a say. And it's happening in the same Ohio county at the beginning of American history and now at present day. And all under a cloak of secrecy without allowing the local community to consent or not consent or even raise objections in real time. The deal is announced after the fact and no one gets a say about whether it goes through or not. Let's not gaslight ourselves, like secrecy, NDAs, fast-moving approvals. What are those in service of and do they lead to something good? There's an element here of we're sort of being asked to ignore what is plain.

Marco Pimentel: Isn't there a word for this?

Catherine Baab: Yeah, it's called boosterism. This is historian's term for a lot of the practices that went on in 19th century America. People selling ghost towns, people looking to sell the West as this place of opportunity, pamphleting the East Coast with stuff about Santa Fe. Come out here, build your life, have all this opportunity and so on. One thing I did see was there's almost a truth in that that you wouldn't necessarily want to be true, but if you build it, they will come. At a certain point, hype takes over and it makes itself true. It hits some kind of stage of critical mass and claims that were bogus to begin with eventually become true. You do see it. Places like Santa Fe, San Francisco, Duluth even that were boosted in this way, they did have thriving economies. There's almost a point where, holy crap, it better be true. We're publishing these insane statistics. We're taking on these contracts. It had better make itself come true. I do think there's some kind of mechanism by which that happens.

Marco Pimentel: Now, the higher bill. Is that just routine economics? More demand leads to higher price? Or is it a choice somebody made about who pays?

Catherine Baab: My colleague Jackie Snow asked about the same thing. Here's a clip of her interview with Gernot Wagner, an energy economist at Columbia Business School.

Gernot Wagner: The big question at the end is, in fact, who is able to negotiate the contracts that are favorable to them long-term. Microsofts, Googles, Amazons of the world want flexibility. So if and when the power is no longer needed, they are not locked into these long-term contracts.

Marco Pimentel: So the side with the most leverage writes itself an exit. And if they walk, the hugely expensive wires they ask for don't.

Wagner: So not all shareholders are created equal either. So they are shareholders of hyperscalers, shareholders of data center providers. They are shareholders of utilities. They are shareholders of steel companies, cement companies who are competing for the power with the data centers.

Catherine Baab: I've talked to people in VC and people in the Bay Area who are investors in AI and so on. There's actually quite a lot of, I wouldn't say it's nationalistic, but almost like a form of patriotism that drives it. There's sincere conviction that we need to beat China to these capabilities, driven by what I think they would themselves term as broad public spirit. And I wouldn't have expected that. It's not something I've heard from VCs in 15 years of covering and talking to them.

Marco Pimentel: There was some research that was indicating that when these CapEx or data centers are invested in, in some communities, that typically means that the net effect of the rates typically drop for electricity. Because they're putting back into the grid, which allows us to drop the prices for consumers. But this math is showing, I guess at least nationally, is showing the opposite.

Catherine Baab: Here's Andy Cvengros, who co-leads U.S. data center markets at JLL, one of the biggest firms in commercial real estate.

Andy Cvengros: We have a very old utility system that has not been improved over time and has not been invested in. And regardless of those data centers or not, all of us were faced with capital expenses of improving our electrical grid in the future, right? And it just so happens that these data center groups are going to help fund a lot of that. The back-end infrastructure upgrades they're going to have and all benefit all of us because that cost is picked up by a single group.

Marco Pimentel: Okay, but that only works if the new customer really does carry the cost instead of it getting spread onto everyone. So why would adding a huge load ever make a bill smaller?

Catherine Baab: Ian Magruder had some thoughts on that. He runs the Utilize Coalition, a grid policy group that pushes to get more energy out of the infrastructure that we already have. Here's what he had to say.

Ian Magruder: But electricity functions a little differently because there's so many fixed costs. The cost of generation as a share of your average person's utility bill has gone down over time. The real drivers of electricity costs in America today, which is really primarily on the distribution transmission side, it's the poles and wires. And so if we can get better at getting more out of that system, we can drive down costs for everyone. There's a big question right now in the energy world about we know there's more demand coming, but how much and how fast and at what scale? And there is a risk that you could overbuild. And then everyone is stuck sort of footing the bill. And utilities, regulators, they're thinking about this very actively. And it's a little hard right now to tell exactly what's real and what's not.

Cvengros: I think this helps to advance the electrical infrastructure of the country for generation conversations in a big way. Are we really jumping to nuclear finally, and with both arms, or small modular reactors? And we're seeing advancements in generation that we just have never really dove into. And three or four years ago, nobody would touch nuclear. There's a maturity in this whole thing that's happened over time. And I think it's going to continue to happen.

Catherine Baab: This is not, I don't think it's necessarily anti-business at all to raise these questions. I tend to think that ethical business practices are actually the best practices because they incur the least consequences over time. There's a case for operating ethically and to the full letter of the law simply because it saves you time over the long run. When I talked to the Association for Rural Electricity Co-ops, they presented it to me in quite a positive light. I wouldn't want to suggest that the math is overly straightforward. Whether it's a net win for local communities, that math will be complex and you're going to have to run it on a case-by-case basis. The extra expenses that would go into maintaining a grid or building out a grid when a massive data center infrastructure comes in and the increased competition for the amounts of power, I would think that overall that is going to tend to drive up rates over the long term. But I could see cases in which there would be either subtleties and asterisks or it simply did not emerge as the case.

Marco Pimentel: Let's go back to that pledge from the beginning of the episode, the one signed in the Indian Treaty Room. What was it actually worth?

Catherine Baab: Drawing on others who consulted energy experts said that it was a non-starter on its face. This is major tech companies, including Oracle, coming together to say that individual ratepayers will not be paying more, which is not necessarily a promise that they can make and it's also not binding legally. So it's largely an act of theater that was immediately portrayed as a binding thing. It was not. All taking place within this historical setting where we've seen this movie before.

Marco Pimentel: How can this kind of thing happen or even worse happen again?

Reeves: It's a sad story, right? Because there were treaties in operation and Congress didn't enforce them. And it would have taken Congress. And at the same time, the Grant administration used its powers to sort of override Congress. And it also kept a lot of information from Congress about what they were doing and how they were negotiating with the tribal leaders and that sort of thing. So, and it gets to the essence of power, right? It's like, do we have structures in the society to restrain a powerful presidency or do we not? And I don't know the answer to that. It's very hard to say. I think if there's a will, it's possible. But in this case, it didn't happen. It's hard to imagine when you look at American history that a case where if the U.S. government wanted to expand into a certain region, it would be stopped or it wouldn't be stopped, it didn't feel like, by the electorate because a lot of this stuff was popular at the voting booth. The law is, as we've learned in recent years, right, this law is not self-reinforcing. It's going to require a counterpoint to uphold it. It's funny with laws having to do with Native Americans. It's complicated because it was changing all the time. And they weren't even sure, like, which party was negotiating these treaties with the Native Americans? Was it Congress? Was it the executive? Was it the court? And also, could Red Cloud speak for Sitting Bull on behalf of the Lakota? They disagreed about that, too. So it got really complicated.

Marco Pimentel: It sounds like what's right got overruled by what's convenient. The people behind this can't all be villains, though, right?

Reeves: A lot of the religious figures at the time really cared quite a lot about the Lakota. They really cared. And yet at the same time, they bought into all of this stuff about progress and inevitability. And so their attitude was like, yeah, I think we need to put them on reservations, but can we at least be kind about it? Do you know what I mean? So on the one hand, you're kind of sympathetic. These are people that mean well, they're good people. And at the same time, it's a really colonizing attitude. They thought they were doing the right thing. But once again, they really did hold this notion of progress as being central to their belief system.

Catherine Baab: I think it's like a good reminder that one can both believe fervently and be wrong. And one can have good intentions and be wrong. We shouldn't assume that these things protect us. At the time, there's no system in place to vet their claims in any kind of systematic way. People are saying like, this is progress. This is going to be a good thing. Well, says who and for whom?

Reeves: It feels like it was inevitable, right? The problem with counterfactuals, too, is that, well, it did happen. So you almost have to think, well, what would have had to happen for it not to have, right? It's kind of interesting that we saw westward expansion as inevitable, but not northern expansion. Like, wouldn't us having just taken over Canada be part of that, too? Like, at a certain point, it's power, like, who's going to stop you?

Catherine Baab: The U.S. had promised the Black Hills to the Sioux in the 1868 Fort Laramie Treaty. After gold is discovered there in 1877, they seized the land without tribal consent. Later, the Supreme Court deems that the seizure was unconstitutional and they offered the tribes monetary compensation. But to this day, the tribes are saying, this is our sacred land. We don't want the money, we want the land back, which shows that how thorny these legal issues can become when they are ignored in real time, right? In the same way that it could be difficult to claw back the incentives that are offered to companies building data centers.

Reeves: I don't think Americans know this, but the Lakota won a huge victory about the Black Hills. And it was determined that the United States government acted in bad faith. The Lakota were given money, though, and not the hills back. And the Lakota are holding out. They want the land back. And the money now is an extraordinarily large sum of money. It's sitting in a U.S. Treasury account because basically the ruling is, yeah, we acted badly. We stole your land. But you know what? We're not giving it back given that there's some economic activity and settlement on that, those hills now.

Marco Pimentel: None of this was a law of nature. It was a series of intentional decisions. People made choices, while other people were told there was nothing to choose.

Reeves: What we've tried to do is we sort of ignore that past that is unpleasant, and we create a mythical past that is more congenial. But to the extent that that mythical past isn't true, then we're not really learning from what happened. So in that response, I don't think we've kind of learned our lessons.

Catherine Baab: Christopher Hitchens said in his critiques of religion, he's like, extraordinary claims require extraordinary evidence. I always come back to that, like is this simple touchstone of like claims require evidence.

Reeves: And oftentimes we're not seeing any evidence at all, you know.

Catherine Baab: Like, there's not even a pretext of providing it.

Reeves: Maybe the one thing that I would just encourage people is that, and maybe things aren't inevitable. I think that it does matter to try to have a voice about some of these things and finding ways to get your opinions out there. And on AI, maybe it's intellectual property, maybe it's jobs, maybe it's censorship. These could be a little area where one could build up an expertise and try to get their opinion out there.

Marco Pimentel: Next time, we'll hear from someone who's active in this space, building up expertise in many of its different aspects. They're betting that the whole thing is fixable, even as they build the fix. That's next on Power Play.