Data Centers and Power Grids: The Battle for AI Infrastructure (with Lynne Kiesling and Steve DelBianco)

[MUSIC] Welcome to Explain to Shane. I'm your host, Shane Tews at the American Enterprise Institute. On this podcast, I interview tech industry experts to explain how the apps, services, and structures of today's information technology system work, and how they shape our social and economic life. If you want to understand the future of artificial intelligence, don't start with chips or algorithms start with energy needs. Data centers are expected to account for roughly 40% of the total US power demand growth in 2026, making energy infrastructure the quiet backbone of the AI revolution. Training frontier AI models gets headlines, but increasingly the real power draw comes from the ability to run the AI model queries, the ability to power enterprise tools, enhance with AI, and supporting the always on digital economy. The challenge isn't that utilities don't want to build more capacity, it's that they can't build it fast enough. The constraint isn't capital, it's time. Permitting a new data center build can take up to four and a half years, and even after approval, the specialized equipment like transformers must be manufactured and installed by skilled labor force. Add to the mix, there is a skilled labor shortage, due to the time it takes to train qualified workers to build high voltage infrastructure and the bottlenecks multiply. Meanwhile, demand is accelerating far faster than supply. That gap shows up as inflation in electricity prices, which is a capacity crunch felt by households and businesses alike. Utilities are planning massive upgrades, transmission lines, sub stations, transformers, with capital expenditure expected to rise about 7% annually through 2029. Because utilities operate as regulated monopolies, those costs are recovered through customer rates. That's when grid expansion stops being abstract and starts feeling personal, and why state regulators and legislators are stepping in. Even trillion-dollar tech companies, racing the scale AI with private equity and sovereign funds, can't solve this alone. They depend on utilities and regulators to build the grid that powers their future. And here's the real strategic question. Why are we trying to build an energy system for the next decade and beyond, using a 1970s regulatory process? This is no longer just an economic story, it's also a national security issue. AI capability underpins military readiness, intelligence analysis, cybersecurity defense, logistics, and next-generation weapon systems. The nation that powers AI at scale will hold a global strategic advantage. If energy constraints slow compute capacity, they slow innovation, modernization, and resilience for the nation that runs those systems. An erase for AI leadership, the decisive factor may not be compute or code. It may be how fast can we approve and build new power lines. To help me impact this important issue on how we enhance our power capability to meet the massive data center needs, I am joined by Lynn Kiesling, non-resident senior fellow at the American Enterprise Institute, and director of the Institute for Regulatory Law and Economics at Northwestern University. Her work focuses on electricity regulation, energy systems, and sustainability. Also joining us in this conversation is Steve W. Bianco, president and CEO of NetChoice. Steve has decades of experience in engineering, economics, and public policy. He previously served as founder and president of Financial Dynamics. Today's guests come to help me look at all sides of this discussion about energy, data centers, and artificial intelligence. I hope you enjoy this episode of explain-to-shane. Lynn and Steve, thank you so much for being guests today on explain-to-shane. You are in the sweet spot of what everybody is talking about right now, which is data centers. And I'm so excited to have you both on as guests because you're experiencing this in both very important but different parts of the operation of building out more infrastructure and why we need that. So, Lynn, I'm going to start with you because I think that some of the just baseline economics of our energy system and our utilities, people don't think about it other than they just they pay a bill that shows up every day or corporations kind of have an idea of what they expect to project. We're getting into a whole new dynamic with this change in, sorry, let me see. We're getting into a whole new dynamic with the data centers, really the key element to bringing artificial intelligence to the next level of what we can be doing as as a society. So, let's just start with some basics here. Should we have done this a long time ago? I mean, are we like, is this just sort of a long time coming and we finally have a reason to push this forward? Talk to us about what's going on with investing in our utilities. Sure, I'll start there and thank you. This is going to be a fun conversation. The economics of data center energy use is exactly as you say, it builds on the fact that we have this 120 plus year old legacy approach to electricity infrastructure and its regulation and that that approach has been extremely effective for the 20th century, but you know, starting in the late 20th century, starting in the 1990s, we did start to see some tensions from technological change. And I think this is just a particularly dramatic example of the idea that I call the pacing problem, which I did not originate. I took that from Adam Thierry, who I know, you know, that he talks a lot about how in tech regulation, technological change outpaces regulatory change. And so you get this mismatch and because regulation is maladaptive to those technology changes and data centers are a particularly large and dramatic example of that. I can go through, I'll briefly just say for people who don't have a lot of background in regulated utilities, although I suspect some of your listeners are familiar with it from the telecom side, but at the state level, the kind of telecom and electricity history is fairly similar that since the early 1900s, state public utility commissions have been the economic regulators of investor owned utilities. And so these are, you know, she'll hold her own private companies and they are regulated at the state level as well as the federal level to build and it gets very complicated. So I'm just going to say in general, in a state like Virginia, for example, where it's vertically integrated to build generation transmission, distribution, and retail, you know, the full vertically integrated stack. And the profit that the regulated utility earns is largely a function of their return on the assets that they invest in. So it's rate of return regulation. And so that has been great for a 20th century trying to build out and electrify the country and build out the infrastructure, but it does have some perverse incentives and, you know, the data center problem does surface a lot of those which we can come back to. Bringing you in the conversation, you're Steve, you're dealing with a lot of entities that are to seeing power restraints. Like they want to, they're chopping at the bit to get more power because they see what they can do with this new technology. But doing that in with our current capabilities is we're going to hit, if we aren't a surpassed, you know, the ability to do it on our current infrastructure is really tough. And this, it came on. I feel quickly for those who don't live in the tech space, it seems like it happened overnight and all of a sudden we're seeing a lot of, I'll just say fear mongering. You know, we're both in the, you know, the northern Virginia in that area that like my neighbors are convinced that our utility balls have gone up and I've explained to them that we're not even on the same grid. I mean, like they get a little confused about that. So how's it going out there? Shane, thanks for having me on. I really appreciate that. The utilities could be forgiven for not spotting the price signals or the demand signals over the last 30 years. The total electricity demand in the United States is grown by about one half of 1% compounded for 25 or 30 years. At that rate of growth, we saw the retirement of fossil fuel and nuclear facilities. We saw states really encouraging through renewable portfolio standards, standing up solar and wind, which is not truly a one-to-one replacement for a base load, coal, or gas, or nuclear plant. So all of that was proceeding in a relatively acceptable manner until about two years ago. When you think about it, my industry has to build a data center, a billion dollar data center, about once a week on average. Just to hold your movies and music and videos, all of the streaming that you do, the messages that you said with copies of all of that, the cloud migration from internal IT infrastructure to the cloud, things like AWS, Google Cloud, all of these factors work contributing to an ever-growing demand for data centers, particularly because none of us ever delete our old videos, messages, and photographs. So there's an ongoing need and our industry is able on average to add a billion dollar data center. Once a week, we would simply pick the parts of the country where there was adequate power to provide a couple of hundred megawatts to a billion dollar data center. And we can put those data centers anywhere and still serve users everywhere because of the nation's superior connectivity. But it became clear in the last two years that something new was afoot and that is the addition of AI. So the very same building about the size of an aircraft carrier that comprises a data center also holds a bunch of server blades within video chips strapped to the side. Those are 35 to 40 thousand dollar chips and they burn 5 to 10 times more electricity than just the storage chips that we had been using for cloud-based computing and photo and video storage. So suddenly the same buildings are consuming 400 megawatts of power and that insatiable demand for power continues to be shoe-horred into parts of the country that have generative capacity and unfortunately we cannot rely upon our regional grid of electrical transmission to deliver power from areas that have access to areas that need it. That grid is also serving us poorly when it comes to the part of the day where say solar and wind in one part of the country is just cranking out the electrons but another part of the country is in darkness or no wind. We don't even have an adequate grid to move that kind of power around right now. So let's actually maybe take a step back and just talk about how many power grids do we have in the United States? Because it seems like there's some people that's just seamless, like it's all just connected. And at some level it is seamless. So it depends on what level you mean power grid because at the deepest gut level there's three large high voltage grids in the US. There's the eastern interconnection, the western interconnection and the split between them is somewhere in the plain states. And then there's Texas. Certainly in electricity and I think increasingly in electricity and in data centers the punchline to everything is always and then there's Texas. Because one of the things that we're seeing right now is that Texas is becoming a great attractor for data center growth and there are a lot of reasons for that that we can get into. But so at that gut level at the very high voltage level you've got these three interconnections and what that basically means is that we don't want to get too much into the physics but that essentially-- You can. Okay. We like physics. We like physics. All right. So I think we have to do that. That essentially you can transmit, you can transport electricity more readily within one of those three interconnections that the connection from one to the other are very, very, very, limited. And those three are all huge. You know they cover a lot of territory. Texas is smallest of the three. And so historically it's not been a problem because you know whether it's weather patterns or the location of generation and demand load that you had enough variation within each of those footprints to be able to kind of even each other out within that footprint. But that doesn't seem to be working really well. Then if you go up the next level there are what are called balancing authorities. And that's sort of utility level or kind of regional territorial level areas in which the grid operator has a requirement to balance the supply and the demand on their grid. And they have to do that in real time. So network, power systems are real time delivery systems. So any supply demand imbalance is going to cause outages. And then there's also the difference between the transmission grid, which is the high voltage stuff. And then the distribution grid which is you know the big square, the big square pieces of infrastructure you see with, with all sorts of stuff that's called a substation. And so the high voltage wires come into the substation and then that gets ramped down to a lower voltage and then put on the smaller distribution lines that you see around town coming into your house so on. So let's do two things. One Steve I do want to talk about the physics of it because it's interesting. But the everything is the geography of it. So let's say you you're in Virginia, data center, Ali, why there what made it was it that they were permissive they had the regulatory structure. It's because they're close to a lot of usage. What makes that area important in the 90s chain? They're say internet interconnect point called May East. There's a May West, of course, and there's a May East and it was in Tyson's corner. It was a block away from the offices of the first company I started. That meant that connectivity was readily there, especially for the transatlantic cables. And add to that the notion that we had the federal government located here in Washington. We had a lot of very inexpensive land west of Fairfax County in an area called Loudoun County. Dominion power, which was Dominion energy at the time of Virginia power had adequate capacity. They have a broad diverse mix of generation, including nuclear. And add that together and Loudoun County became very attractive to build a data center. Once you constructed one data center, then the architects, engineers, contractors, and subcontractors who built that one are ready to start on a new building as soon as it's done in two years. So 1800 workers will labor for two years to construct a billion dollar data center. But before they're even finished, we're building another one adjacent to it on the same 100 acre campus because you've already done the zoning. You've already built a substation for interconnect. So it makes sense to continue to build them out. So when you think about it, Shane, 25 years ago, we were both here in Northern Virginia. When you went out to Loudoun County, out near Dolas Airport, you'd see data centers and there'd be nothing anywhere near them. The three decades later, there are homes right up against those data centers. And in most cases, the data centers were there first. And Loudoun County grew without adequate setbacks and ended up having neighborhoods cheap by job by the data centers themselves. The ecosystem of contractors, architects, engineers, service personnel, contractors and subs, that ecosystem means you can build in Northern Virginia faster, higher quality and lower cost than you can anywhere else in the world. It's the largest concentration of data centers on the planet. Beijing is number two and they're not even close. Wow, that's amazing. That same phenomenon of course works with airports whenever you move in airport. I think what Denver, when they moved it, and there was nothing out there, opinion, ball of art and now there's a whole little city around it. And there's a related phenomenon going on now that what's relatively scarce is the power. And we've seen so many people talk about this from Jensen Wong who very carefully says that the real limitation on our AI growth and on data center growth is power. And I think increasingly what we're seeing is it's to Steve's earlier point, it's not just the generation. It's really more the grid and how we both build the grid, but also the fact that we don't do a very good job of pricing and managing our use of the grid. And as grid capacity becomes scarce in some places, but it's abundant in others, we don't really do a good job of that coordination because we don't use prices really well. And so power has really become the limiter because it's the things relatively scarce. But you see things like in Texas where big data center projects are being cited out in West Texas right in the Permian Basin. And so the gas is right there to fuel the turbines to generate the electricity to run the data centers. And I also have to give a pitch in here for one of my favorite companies, which is Crusoe, because not only are they citing data centers right out where the gas wells are, but they're also buying gas that would otherwise be flared because there's not a whole lot of storage and there's not enough pipeline capacity. And so gas producers end up having to flare stuff, especially if it's like gas co-located with oil and the gas comes out and you can move the oil, but you don't have anywhere to put the gas so you just flare it, which is both economically and environmentally costly. And so if you know, Crusoe goes and buys your flare gas, you know, that's a win-win economically and environmentally. So we see a lot of what Steve described in terms of the construction and labor, shifting to the power being the main constraint, but also the same underlying driver for why you get this agglomeration in Texas. So is there a need for this to be in a particular region or is what you just pointed out with Texas is that the key is getting the energy near the data center. There's probably some positives around that from just a benefit of physical space, I would think, but it doesn't matter how close you are to the consumer of the tail end data of the. - I need that point. - You can build a data center anywhere in the country and server-views are everywhere. - Okay. - The only time that we need to put it, customers need to be close by is when low latency is critical. A medical surgical application, there are certain financial traders who believe latency is their edge. I think that autonomous vehicle and robotics will need lower latency. But for those of us using, say, chat GPT, latency, whether it's two or four or five seconds to get an answer means that data center that had the Nvidia chips could be anywhere in this country. So we are now scouting for places in the country that have electrical generation capacity where the state taxes data centers, the same way they tax manufacturing and others, where the local community is just itching to get hold of a data center, which will generate billions of investment jobs for multiple years of construction, will hire at least 200 high school trade school grads to operate the data center when it's done. And each building has about that many. And these are guys that average six figures. So a small rural community where high school grads were moving away to find jobs, they can suddenly raise families and build a career right there at home. And the third major benefit is to the taxpayers. Data centers themselves pay massive amounts of property taxes to the local county. The county can then reduce rates for everyone else as well as add infrastructure that the county needs or services the county needs. We spoke about loud and county earlier and over half of the property taxes in loud county are paid by my industry. They're paid by data centers. That's enabled loud and county for 10 straight years to reduce the property tax rate on the rest of us who pay property taxes there. So there are counties in every corner of the country. And if they're lucky enough to be in a place where the utility has available power that our industry will pay for the substations and will pay for all the interconnect will commit to long-term power purchase agreements. And utilities like to say that data center burns the same amount of power all day every day. And that's the favorite customer for utility. They can build base load to serve us and they can shift a huge chunk of their fixed costs onto the data center customers because we are large load customers. That's why rates are lower in Virginia than they were five years ago, adjusted for inflation. And the same thing is true in Ohio and a handful of other states that have a lot of data centers. - There's two other things that affect the location that I think a lot of people probably don't realize. One is that Steve generally alluded to but I want to dig into a little bit more the nature of the workload that goes on in the data center. And I think a lot of people just think it's the same. You know, we're just gonna be cranking these chips 24/7, 365 all day every day. But there is a lot more gradation in terms of the workload. One difference for an AI data center is gonna be the difference between training and inference. And the training is where you're training the model. And that is not gonna be particularly location sensitive and this gets to my second point, which is that folks in this industry are innovating on the fly. And so right now they're coming up with new ways to manage the training workload that are in my mind are kinda like the way international financial markets operate where if you have a trading company, you know, you basically are trading 24/7, you're just basically shuddling your book around from London to New York to Tokyo to, you're just basically keeping your book going and you're just shifting it from one to the other to the other. And so there's some work on some models to basically be able to pause the training work and move it to someplace else. And I'm fascinated by that. But then the other is the inference which is gonna be more latency sensitive. And but if you think about it on kind of human time scales, you know, we don't we don't notice it. The more we have an agent economy and the more that we use cloud code to build agents to do stuff for us, they'll notice the latency more than we do. - Yeah, that's interesting. Goldman Sachs had a report I think it came out about two weeks ago that said the rapid expansion of artificial intelligence is driving massive surgeon like Tristan demand. But they made a point of saying, "Well, the shift may cause a minor drag on national GDP growth." The analyst concluded that long-term productivity gains or may I will far outweigh the economic cost of the increased power consumption. So I guess part of it is not enough for people are gonna read that report to feel better about this maybe this little hiccup. So what I'm keep thinking in my head is explaining to what I'm gonna go to my book club tomorrow night and a bunch of women are gonna be upset about this because they think that they, and then actually it's gonna, the tax situation, I think of it as an anxiety tax. There were like charging people because they're just not sure how or when this levels out. So kind of walk us through how this cycle works. 'Cause you mentioned jobs on the front end would never really get spoken about people talk about the lack of jobs and the tail end when it becomes just maintaining the servers. So how does this cycle out? - I think so three to four years from starting to scout for a site to having a data center that's open for business. And I told you earlier that the scouting is a net item that starts with power and makes it sway down to the community that is desperate to get a data center. We have communities in South Dakota for example that have declining population, aging population, the lack of jobs, infrastructure that's decaying and the property tax is keep going up. For a community like that, they are desperate to get data centers. And we've had a lot of success stories in rural parts of the country. And once it's open, those 200 folks making six figures end up creating a new labor pool, right? And the ability to work locally. You talked earlier about the cost of the electricity. Well, let's be clear, this is all costs that are paid by my companies. When we build a data center, we build a substation, we arrange for the purchase of enough capacity from the utility, all of those costs fall upon us, our industry, our investors, which includes the two of you, right? Because most of your retirement portfolio is full of the magnificent seven companies. You're welcome. And those companies are the ones who were taking shareholder capital and redeploying it towards the construction of data centers and paying for the electricity to serve up AI at all of us over on the country. So that cycle is not one that imposes a tax on anyone else. It's paid by our industry. And as Lynn will show, the track record's been very good, is that when we create new demand and pay our share of the costs, regular utility rates go down. And that's partly because every PUC in Public Service Commission is mandated by law to never allow a customer to shift their costs onto other customers. And that's the law, but if we need belts and suspenders, let's make new large load tariffs to give people more assurance that that will be the case. Why is there a tax on the folks at your book club tonight? - Well, I think it's, as I say, it's more of a feeling of anxiety, but they also feel like they're subsidizing. And that's, I don't know that that's true. I just think that we see a lot of media reports about that. - And it's gonna be, the challenge is to do, Steve, what you just said is absolutely true in theory, so regulatory, regulatory economics and cost allocation theory all line up exactly with that. Our ability to do that in practice is something that I've been working on quite a bit in some of my own research recently. And in part, I think over the past few decades, the regulatory practice has departed from that good sound regulatory cost allocation theory in ways that obscure the allocation of the fixed costs and may generate in cross subsidies that I think this data center moment that we're in is really gonna catalyze. I think it should catalyze a real re-examination of regulatory practice. We have to go back and open the black box of cost allocation theory and say, okay, I mean, and I'm saying we can go back to John Stuart Mill, 1848, the economics of joint production and roll forward and see the theoretical arguments for how to do cost allocation better. - Our industry has shown up in that economic theory conversation and say, as President Trump has said, that our industry, if we build a data center, will pay all of the costs for interconnect. We will pay our share of whatever new generate of capacity the utility has to put on. So we are committed to paying our share of all of our generative and grid costs. They've always been that way. And if the utilities a little slow in implementing the tariff will shame on them. If the public service commission is subject to a committee and a state legislature that demands they retire a coal nuclear plant and stand up solar and wind, will shame on them. Is that isn't going to power my industry and is not going to power our economy we need base. and not necessarily the kind of intermittent power that solar and wind provide. Let's talk about where we need to make the change here and really this is probably a little something you've studied a lot is. So we have some federal laws, but mostly this is a state and local issue when it comes to regulations. And you know, the lawyer you think PC did back in November, he just said we just we have we're not we haven't thought about where we're headed on this. We're just continuing to perpetuate my words on his on this, you know, the how we used energy in the past and there just hasn't been really the forecasting of understanding some of this technology on realize AI came on pretty quickly. But so if you are to spend your time trying to do corrections in the market, are you spending a lot of time in front of state, PUCs, local state, you know, city councils, the legislature, where does the energy need to be done from the human capacity to fix the energy problem for the data centers? I think it's probably all of the above, but you know, in my work tends to focus most on public utility regulations. So that's at the state level, but then also has some some stuff at the federal energy regulatory commission. Because data centers and it's complicated for several reasons, number one data centers tend to connect to the grid at the transmission level. And so technically speaking, the transmission grid across the state boundaries, right. So so there's organized organization, organized wholesale power markets. So for example, in Virginia, Virginia is in PJM as is, you know, Maryland, Pennsylvania, West Virginia, etc. And so that transmission, that market and the grid that allows for that market to transmit and deliver on the transactions across the state lines. And so it's under federal jurisdiction, but at the same time, the investment that the that the utility is making, because the utilities own transmission and distribution, you know, transmission and delivery wires networks. And so so it's both a state regulatory jurisdiction and a federal regulatory jurisdiction jurisdiction. And right now there's a bit of a territory fight going on because in November that the DOE issued an an offer announced for advance notice of proposed rulemaking to direct perk to do some some rulemaking around large load interconnection to try to standardize it and streamline it. And this is going to have a big jurisdictional fight between the federal and the state regulators, because the state regulators very jealous, Lee guard their federalism. And so so that's one place where it's going to it's going to show up. And so it's really complicated, though, because because the incentives are just so murky in ways that we probably don't want to get into because it's deep in the weeds. And so it's it's really, really difficult to get new transmission capacity built building new transmission lines is very difficult, whether it's citing or permitting or just the utility regulation parts of it. You know, in my work what I look at is the more demand side, right, how can how can we get better pricing, how can we use markets to enable better utilization on the existing grid that we have because the average you know average grid utilization rate across the country is only like 60% or something there's huge amount of access has the on average. And we're just not set up to take advantage of that. Yeah, similar like the tech, the text, it and say that that Texas example you gave us where they said, you know, we've got gas is coming off that's just going to go evaporator in the air. So we're going to grab that and utilize it and putting a data center right next to it. And so having demand flexibility and more use of pricing and markets is the way to coordinate that better. The longer the money is then doesn't seem to because in Kansas, the utility, even without a data center in sight, couldn't even build a new substation that it needed. And it was because you had a collection of opposition, not just your nimbie not in my backyard, but we had far left and far right kind of joining together at the bottom of the circle and complaining about any changes to the landscape. They disbelieve anything that business tells them they disbelieve and distrust the government. They don't trust media. They seem to only rely on well social media going down some rabbit hole on Facebook that tells them that data centers and electricity are going to make you sterile and rot your brain. I mean, these are the things that come up in hearings. So we've got nimbieism that's translated into banana build absolutely nothing anywhere near anyone. They don't want to change the landscape in any way and they distrust all of us who are trying to tell them that this is going to be good for you and the nation needs it. So that is really increased the challenge of just constructing a transmission line, a gas pipeline, a sequestration pipeline and a substation for electricity. We have some real challenges in rural America right now. So when you are talking to the locals about the potential productivity gained with the whole new air of technology coming on that doesn't balance against the fear they have right now of they're saying, you know, earlier, it's a couple jobs some people are going to show build these things and then they're going to go away and then it doesn't take that much to maintain it. So they just see it as a not a net it's not a net positive thing for their community as it's being explained to them right now. Is there are there better how do we get around that because it seems like there's just each audience is different Shane and you're so right. When I testify a local hearing the folks that show up in the pink t shirts and complain about data centers causing diabetes or digital smog shortening lifespans in northern Virginia. I'm not kidding. It came up last week in South Dakota. Those people will never be convinced they don't trust me. They don't trust you. They don't trust the government. So they're not going to become convinced. So the real audience becomes the taxpayers. And they can rely on the evidence that property taxes go down when a data center comes to town. The other is the high school guidance counselors and job teachers and the parents of high school grads for whom those 200 jobs paying six figures is 200 more jobs than they even had in the county. So if we needed 2000 people, I don't know where we would get them. We have a population that is flat and a workforce that's declining. So where would we hire more people if we needed them right. So we end up talking to the commissioners, the county commissioners, the zoning commissioners, they understand the numbers. They look at what their opportunity to have additional property tax revenue and employment base. And they will end up basing their decisions on actual facts. They'll take a visit to a data center in nearby state. And suddenly come back with a new appreciation that they are noisy. They are not on site. They're not disruptive, but in fact, they add to the economic and tax framework of a county. For that reason, we really focus our attention on the lawmakers, but there are some fringes who don't trust anything and don't want anything to change. So give me an example of who's doing it right. I have a broken record on this. I've always been a big fan of the Texas model in Texas. Their investment climate is very conducive, right. They've got essentially transparent contracting terms, low entry barriers and the from the power side, right. And the structure of the regulation of the electricity industry is they've done what I like to say. And they're the only state in the US that has done this. They've quarantined the monopoly. And for, you know, Shane, I think for you and your listeners who are old telecom folks, you will recognize the phrase quarantine the monopoly from the bill backster, the assistant attorney general for antitrust who wrote the settlement. In the AT&T, the vestiture in 1982, 83. And that this was his principle. If you have a natural monopoly that is sandwiched in the kind of vertically integrated stack between competitive and a competitive industry. And that regulated industry can have anti competitive effects on an adjacent competitive market that you need to quarantine the monopoly. And that's essentially what Texas has done. The regulated footprint. In for the regulated utilities in Texas are just the wires, the transmission and distribution wires and they're integrated into one one utility. And so that means you have open competitive wholesale markets for energy on different timescales, you know, day ahead, day of hourly 15 minutes. And so that's what the market market markets for what are called ancillary services, the voltage regulation, the frequency regulation. All the stuff you need to do in order to operate the grid. And that's provided through markets in ways that it used to just be done through the utility. markets. And so what that means from a data center perspective. is that it's pretty straightforward to come in and propose and get permitted to do this project. The other thing that it's easy to do in Texas and that other states are, because of this regulatory structure, and that other states are starting to explore being able to make happen, goes by the name of bring your own generation. In economics, we call this the Make or Buy decision, where instead of buying from the utility, you can on the property of your data center that you're building, you can hire an independent power producer to come and basically build you a power plan on your property. And in lots of states, and historically, this has been illegal, that if you go back through the history of regulated utilities, the electric utility has the legally government-granted legal monopoly over the generation transmission and delivery of electricity. They're the only ones who are allowed to stream wires over public rights away. And the move to wholesale power markets over the past 30 years has kind of chipped away at the generation portion of that, but we still have the what are called franchise rights. The utility still have franchise rights for the wires, but in Texas, it's fairly straightforward to just have an IPP come in and build you some generation on the same thought of land that your data centers on. So I keep wondering though, who people sometimes need an example to move forward? And there's a lot of change we're looking to have happen at multiple tiers of government here. And you, the three of us can see the vision, right? Because we're users of artificial intelligence. We see where the trajectory of this is if we get this right. So the challenge is getting down, especially with these local people, like using Texas as the example, is there somebody you show the Texas example to and they say we can emulate up to 70, 80, 90% of that in New Mexico or Kansas or South Dakota. Because Texas's market is the most regulated market and the other regional transmission organizations would have to do a complete brain transplant to turn themselves into Texas. That doesn't mean it can't happen though. You remember last summer, President Trump traveled up to Western PA where I'm from and announced that one of the nation's largest mothballed coal plants in over city, Pennsylvania would become a combined cycle gas plant. And that $100 billion of new gas fired electricity generation and data centers hosting AI and other applications would be built by Google and a handful of other companies in an area of Western Pennsylvania that had been mothballed for decades as the steel industry left that area. So that's a situation where state, local and federal leaders came together. My industry showed up with the checkbook and that has moved blazingly fast through the permitting process. So it doesn't have to be the wild west of Texas but Lynn is right that we ought to learn as many lessons as we can from them. Fantastic. Well, I know I have you guys for a limited time and I ask a special favor in the beginning is somebody who's been watching Landman and I love the Billy Bob Thornton Salilquiz. He has a whole thing about using alternate fuel with the wind turbines to help produce more gasoline coming out of the Permian basis there. And Lynn just loves to tell me that there's a lot wrong with that even though I've seen it multiple times and I just love the acting. So let's let's break this apart because I think it's an interesting kind of case study on all the things we don't know about the electricity in industry and what does he get right? What does he get wrong in that? Yeah, I would say yeah, his economics is incomplete. I'll say that way. That's pretty good. And in part, I'll start actually from kind of an engineering point which is that he's making a life cycle argument because you know and he's talking about the situation, you know, they're out in West Texas and this has really been a big move in the past decade to that electrifying the oil fields, oil and gas fields has really been important for their productivity and help to reduce their costs. So electrifying the oil and gas fields has been a really valuable exercise. The way it's happening because it's so windy in West Texas, they're just you know, building wind turbines and the wind turbines are powering or powering the wells. And this is number one, that's you know, I have to throw it clear a little. This is aided by the fact that wind power until recently has had a production tax credit. So there is a tax-based subsidy for to help develop wind power technologies. And that has recently been discontinued, which makes sense because winds are mature technology. And so it should survive on its own at this point. But he starts by making this life cycle argument. And this is what engineers call life cycle analysis where you look at the embodied energy and emissions in the production, installation operation and decommissioning of of a technology. And he gets part of it right. He does miss, he can, you know, he talks about the fact that onshore winds, life cycle emissions are lower for wind than for gas or for coal. But he really only captures a part portion of it. He just captures the construction. He can doesn't capture the actual use of the of the technology. And so his claim that it won't offset the carbon footprint of making it, the wind turbine is not really supported by the the economics. But I will say just to keep it brief, that the one thing that I think they did very trenchantly there is illustrate the extent to which energy transitions are long and slow and incremental and energies are energy technologies are very capital intensive. You end up having this big embedded installed base. They have useful lives of up to 40 years. So it takes a long time for them to get to depreciate and to get replaced by something else. And so I would like to highlight from from his soliloquy the fact that he highlights the fact that there's a coordination problem here. Interesting. Steve, anything you would add to that? Yeah, his initial claim is that there's more carbon footprint from that wind turbine than there is from a gas fire turbine. And if you took over 10 years, the amount of carbon per kilowatt hour isn't even close. That wind turbine over 10 years generates less carbon for all the kilowatt hours it cranks out than would a gas turbine. Just a quick shout out to all of my friends from Lincoln East, June, or high. We did the Ghana Dam project in seventh grade where we learned about hydropower. And being somebody who gets to be in a lot of ubers every time I get in an Uber driver happened yesterday, Guy from Ghana, I said, I learned about the Ghana Dam, that was in high school in junior high actually. And you know, it was interesting that one of them a couple of years he was, oh, we decommissioned that. And I had to think about it. I had to think about how long that had been and that it would make sense at that stage. Maybe that you, you know, you move through it and go to something else. I'd say there's a lot of people who really don't understand how any electricity works. So getting to hear you to talk about it, I'm sure we'll have you on again, there's going to be a lot more about this. But I appreciate the kind of the little baby myth busting on the economics of it. It's I find it to be very important that we get this part of this journey correct because it could be a huge hindrance to our overall economy on where we're headed with all these new amazing things that are coming online. Artificial Intelligence is just the tip of that iceberg. Thank you both for coming on. Please stay in touch. And I really want to thank you for being guest today on Explained Shane. Thank you Shane and Lynn. Yes, thank you very much. Thank you for listening to another episode of Explained to Shane. For more episodes subscribe to the podcast on Spotify, Apple podcasts, or your preferred listening platform. If you enjoyed this episode leave us a review and tell your friends and colleagues to tune in. We'll see you on the next episode of Explained to Shane. [MUSIC]