You’re out of free articles.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Sign In or Create an Account.
By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy
Welcome to Heatmap
Thank you for registering with Heatmap. Climate change is one of the greatest challenges of our lives, a force reshaping our economy, our politics, and our culture. We hope to be your trusted, friendly, and insightful guide to that transformation. Please enjoy your free articles. You can check your profile here .
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Subscribe to get unlimited Access
Hey, you are out of free articles but you are only a few clicks away from full access. Subscribe below and take advantage of our introductory offer.
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Create Your Account
Please Enter Your Password
Forgot your password?
Please enter the email address you use for your account so we can send you a link to reset your password:

Robinson Meyer: Secretary Granholm, welcome to Shift Key.
Jennifer Granholm: Thank you. Glad to be on.
Meyer: So let’s start here. Can you just start by setting the scene for us for what the Department of Energy has done over the past few years, and what it has become? Because as I think about it, and as we think about Trump again, I remember back in 2017 when, I believe your direct predecessor, Secretary Perry, took over the office, and he was very excited to talk about energy and to talk about fossil fuels and to evangelize oil and gas. And then he had to learn that at the time, DOE was really more of a national-labs-and-nuclear-weapons type of agency. I feel like that’s less true today. So can you just set the scene for us on how far we’ve come?
Granholm: I mean, I can talk about the department, yeah. But I’d love to also, at some point, talk about the outcomes.
Meyer: Let’s talk about outcomes first, and then let’s talk about the department.
Granholm: Great, great. So just at a level, I hope — I know your listeners are savvy in this space, but just to put it in perspective: It is astonishing what the Inflation Reduction Act and the Bipartisan Infrastructure Law have been able to do to build out clean energy in the United States. It is mind-blowing. If you had told me — I’m a former governor of Michigan, and when I was governor, we saw all these jobs leave, all this manufacturing gone. It was just — and the federal government just sat on its hands and did nothing, although when the Obama-Biden administration came in, they saved the auto industry. But every supply chain, every — and they were all going to Mexico, or China, or whatever.
And so the hollowing out of much of the middle class and the manufacturing, and I think this is one of the things that Trump really picked up on. But across the nation, in industrial communities, small towns that had a factory that left … So the Inflation Reduction Act and the Bipartisan Infrastructure Law have caused now, as of today — and this is so far, since the Inflation Reduction Act was only passed a couple years ago — over 950 factories to come to America or expand in America, building these products for this clean energy transition. That is just mind-blowing to me. And those announcements are announcements. Then there’s the groundbreakings, and then there’s the ribbon cuttings. And so there’s been some groundbreakings and some ribbon cuttings, but all of those are going to bear fruit over the next few years.
Meyer: So during the Trump administration.
Granholm: Exactly. And that’s, I mean, regardless of who’s in the White House, this is really great for America, that we have, now, industrial strategy that has reversed that track, that tide. So that’s, in terms of outcomes, that is a huge outcome.
The second outcome that I think is really important is how much clean generation that has happened. So this year, we will add 60 gigawatts of clean electricity to the grid. That is more than double what we have ever done — 30 Hoover Dams’ worth, is what I like for everyday folks to understand the magnitude of what we are adding. And that trajectory is going to continue as long as those incentives are in place, and it’s going to be higher next year and next year. And so the goal that the President came in with, with 100% clean electricity on the grid by 2035, and to get to net zero by 2050, by 2030 we will have — just because of these two laws, the Bipartisan Infrastructure Law and the Inflation Reduction Act — we will have 80% clean electricity on the grid by 2030, and we will have reduced GHG emissions by 40%. Now that doesn’t speak to … the reduction in GHG doesn’t even speak to what the cities and states and private sector is doing. So we’re hopeful that we can get to the half, 50% reduction by 2030.
Meyer: Do you think those trends will continue through and despite Trump?
Granholm: Yeah, this is the question of the hour. Here’s —
Meyer: And we can get more into this later, I just wanted to stress test that.
Granholm: I’ll just say, I have hope that this will continue, and I don’t think it’s irrational exuberance because of the fact that those 950 factories, 85% are in red counties, red districts. And because, you know, governors and members of Congress and, you know, President-elect Trump will be able to oversee the fruition of people hiring, steel going in the ground in these factories, it really would be political malpractice to undo those incentives when it’s creating everything that both parties have asked for, which is to reshore supply chains, critical minerals for electric vehicles and other products. It’s just, it’s doing the thing that America should do.
And so I am, I’m hopeful because of that, and I’m hopeful because of the technologies that there’s bipartisan support for. So we know there’s bipartisan support for nuclear, for example. We know there’s bipartisan support for geothermal. We know there’s bipartisan support for hydroelectric power. We know there’s bipartisan support for getting critical minerals reshored, or at least the extraction and responsible extraction and processing of them. So there’s a lot of areas that I think there’s bipartisan support for, and therefore it will be difficult for Congress to overturn or yank back from the Inflation Reduction Act.
Meyer: Let’s talk about the office. In addition to accomplishing those things — I remember in 2021, looking at the changes that were being made the Department of Energy, some of the changes that were in the Bipartisan Infrastructure Law, and thinking, DOE is going to be a totally different agency by the end of this term, or even just in a few years, than it was a few years ago. Can you talk about how the office has changed under your leadership?
Granholm: This is one of the reasons we’ve been able to achieve those outcomes, is because we restructured the department. We still have our national security mission, of course, and that’s very important. And we still have the mission to clean up the legacy waste sites from the Manhattan Project, as well. But we have completely restructured to deploy, deploy, deploy.
We still have the labs, of course, doing all of the work that leads up to deploying on the technologies that really are significant, and those missions still occur. But we were given over $100 billion through these acts, 60 new programs to restructure and focus the Department of Energy on this clean energy transition, and that has utterly changed — we’ve hired almost 1,000 people to be able to do that. We’ve created a whole new vertical, a whole new Under Secretary for Infrastructure, which is all about the infrastructure of this clean energy economy. So that change, I think, has been significant.
We hired a bunch of people from the private sector to come in to understand how technology — we don’t want to take a risk. For example, through the Loan Programs Office, which I know you’re well familiar with. You know Jigar Shah, who runs that, he doesn’t take a risk, they don’t take a risk on the technology working. You know, the physics are going to work, the chemistry is going to work. The niche that they provide is great technology, do they have a business plan that is going to ensure that they’re successful in the market? So we’ve hired a number of people from the private sector to be the Sherpas on that and make it work. And the results have been really quite astounding.
Meyer: It’s really become more of an industrial policy department, in some ways, which is something that a lot of other countries have. In fact, it’s something that most developed countries have, or many of them have. Japan has one. Germany has one. But it’s something the U.S. has historically not necessarily pursued in a civilian agency. At least we have, we do DoD industrial policy forever.
Granholm: Yes, this is what’s so incredibly great about this. You know, again, I say this with my hat as former governor of Michigan: When we used to, you know, we were competing as states. States had their own economic development agencies, but we didn’t have a federal partner on any of it. It was like bringing a knife to a gunfight, and we would see, you know, we can’t compete with the industrial policy of China as a state, right? So the fact that we now have a federal strategy that enables states to enhance, but that there is a federal decision that we are going to get in the game. We’re not going to be passive and allow other countries to eat us for lunch.
Meyer: So how do we think that … how do you think that these programs will fare under the next president?
Granholm: Well, I do think because — I think they are durable, first of all, because one, we’re going to commit a huge amount of the money we were given — 98% of the programs we were given, 60 new programs under the Bipartisan Infrastructure Law and the Inflation Reduction Act. We will have at least on one round of funding, 98% of them, and we will have gotten commitments for all of these selections by January 20. So there may be a couple that we’re not able to do because they’re multi-year programs, etc., but we want that commitment. So that creates one layer of durability.
And of course, the second layer of durability is that they, these programs, are all going to stuff that Republicans have been calling for, too. So for example, there’s been a lot of curious complaints about the EV strategy of the administration because, it is said, all of the critical minerals, and China produces the batteries and the critical minerals. And that’s true. China has — but that’s why this strategy is reversing that trend. You know, China’s cornered the market on batteries, and now we’re make … I mean, of the 950 companies that I talked about, over 400 of them are batteries or electric vehicle companies that we’re building in the United States, you know, the anode, the cathode, the separator material, the electrolyte, the critical minerals, the processing of those minerals. That’s all coming back.
Meyer: Well, and there is something … So have you talked to Doug Burgum or Chris Wright? Yeah.
Granholm: Well, I hope I get a chance to talk to Chris Wright. I’ve definitely talked to Doug Burgum in my role, and in his role as governor. And he’s been an all-of-the-above guy.
Meyer: I think it’s interesting because one of the big questions I have going into the Trump administration — in addition, I should say, that there’s like, two buckets of questions, right? There’s the, as you were saying, there’s a lot of durability, if you were to look at conventional ways of assessing durability, in the department’s programs, right? A lot of these programs are bipartisan.
Something we saw during the Trump administration was that the Office of Management and Budget from the White House would publish these extremely drastic budget cuts when they proposed a budget to Congress, and there’d be all these news stories about how they wanted to cut the EPA budget by half, and they wanted to cut the DOE budget, and they wanted to cut NASA’s budget. And all these news stories would come out, and then six months later, the Republican majorities in Congress would pass a budget, and it would lightly top off all of the budgets, and all the funding would go up, because it turns out the government largely does things that Congress wants it to do, and there’s a lot of support in Congress.
That being said, it also seems like this administration is going to test the limits of impoundment and what the president might be able to stop. But once you put aside that bucket, when you hear Doug Burgum talk about what his complaints with the Biden administration’s policies are, many of them are not grounded in factual inaccuracy. I would say they’re … for instance, the big point he made during the Republican primary debate was about, if we switch to EVs, we’re going to make China more important. We’re switching from OPEC to Sinopec, is, I think, his line. That is exactly what the suite of policies that have been advanced in this bill are supposed to do.
And so one big question I have is like, are we going to learn in the next few years whether these are actual concerns to these folks and they want to preserve these programs? Because it turns out, this is how you would try to create an EV industry in the U.S. Or do they really just care about oil and gas and their concerns with supply chains, with mineral supply chains, are just kind of a tissue to cover up larger oil and gas concerns?
Granholm: I mean, obviously they’re very pro fossil fuels. We know that. But I will say all of the rhetoric has been about all of the above, an all of the above strategy. I mean everybody from Doug Burgum — I mean, all of these Republicans on the Energy Committee, they’ve all said that. So it just would be strange to turn your back on everything that you have said all of this time.
And I don’t even think, honestly, I can — This is what I would say: I think that the Trump administration, it seems, what do I know? I haven’t talked to them personally, but I, it seems from all I can, they want to reshore manufacturing. So if it’s not, if you eliminate the Inflation Reduction Act, you got to replace it with something that’s going to attract all of that investment. There has to be some industrial strategy. Otherwise you’re just ceding the territory to China, which is the thing you’re complaining about. So if you don’t like an incentive-based strategy, which is really what the Inflation Reduction Act is, I see he wants to put in a tariff-based strategy, if you just want to do all sticks. I’m not sure that’s good for the economy overall, but a blend of carrots and sticks, I’m sure most people would say are important. And so, you know, maybe you call it something else, but you got to get in a game, because otherwise our economic competitors are only too happy to see that happen. Can I tell you a quick story?
Meyer: Yes, we have a little bit of time. This is what this is for.
Granholm: So when I was finished being governor, I went to China to see what China was doing in this clean energy space, because they were cornering the market on a lot of these technologies, solar in particular. At the time, I went with Securing America’s Future Energy, an organization that is focused on competitiveness in this clean space. And we went to a presentation by a bunch of mayors in China, mayors of provinces. And as we were standing there watching the presentation, one of the mayors stands next to me, and he says, “So when do you think the United States is going to get a national clean energy strategy?” And I said, “Oh, I don’t know … Congress …” And he looks at me, and he gets a big grin on his face, and he says, “Take your time.”
Because of course, they see our passivity as their opportunity, right? So this is why what we’ve done, what these laws have done, what the president has done, is so amazing. So I get that you may have to put a different spin on whatever the new administration wants to do, but ultimately, you have to have policies like these if we’re going to be successful in reshoring manufacturing.
Meyer: What do you think is the trajectory — you mentioned 60 gigawatts of clean generation coming online. What do you think is the trajectory of clean energy under Trump?
Granholm: Well it, you know, if the Inflation Reduction Act holds, those incentives don’t go away. The developers are going to take advantage of it. I think it is inexorable that you will see continued climb of clean energy generation. I, you know, I know there’s a lot of talk about drill, baby, drill, etc. I mean, they have been drilling, baby, drilling. There hasn’t been limits on that. They may want to open up more auctions and more leases, but the private sector has to be willing to actually put the capital investment into going into those spaces. And the market is a global market, and they’re going to respond to what is happening globally. So I think clean energy … I believe strongly that if the Inflation Reduction Act stays, the clean energy investments will continue to go on that upward trajectory in a way that is astounding.
Meyer: Let’s talk about fossil fuels for a minute. As of when we’re recording this, the LNG study is not yet out. Can you give us a sense of what we might hear when that study comes out, and what the department has aimed to do with it.
Granholm: Well, I’ll just say this. I’m going to wait for the study to come out to speak to what its conclusions are, but know that for everybody listening, we were looking at what the greenhouse gas impacts are of all of the additional authorization that has been, and the current authorization that has already been approved for export. We have, we’re looking at what the impacts are from local communities who bear both the economic, the jobs, as well as the adverse impacts to that. We’re looking at what the global trajectory is of demand, and what our allies might be needing in terms of liquefied natural gas. And importantly, we’re looking at the impacts back home on the costs of natural gas. We, you know, if we are producing little over 100 bcf a day of natural gas, and you take a good chunk of that and export it, what does that do to the prices at home for those who use it? And so we’re looking at the costs at home.
So the study will come out soon. There will be a 60-day comment period, so it will run into the next administration, obviously, but I think it’s important to lay out what the study finds. Just the facts. These are, this is done by our labs. These are just facts. And I think it’s important for those commenters to comment on it so we have it for the record, and potentially a future administration could make policy based upon what the study reveals.
Meyer: LNG has been such a big hot-button issue in this administration, and to some degree it has dominated some of the discourse around these questions. Have you been surprised by the degree to which it’s dominated discourse, or surprised by the degree of the backlash?
Granholm: Yeah, a bit, because of the amount that we have already authorized for export. I mean, the pause on additional authorizations does not impact in any way, shape, or form the amount of exports we are doing. We have already built up and are exporting … we’re the largest exporter in the world, 13 bcf, you know, billion cubic feet per day. We are a powerhouse in exporting LNG, and we’ve authorized a whole bunch more that is either under construction or seeking final investment decisions, etc.
The amount we have approved is massive. And this is … this study was a pause on even more, you know. So we’ve approved an amount that is almost equivalent to what the U.S. the demand will be, you know. So it doesn’t affect any of that. We didn’t pause any exports at all. So have I been surprised? Yes, because of — But I understand politically that some things aren’t necessarily based on the actual facts. But people are afraid, maybe, that there will, that portends some callback in existing authorizations or something, which it does not.
Meyer: So often when you go back, and I’ve been going back in and looking at the climate policy and energy policy from the administration, and what the coverage of it has been at the time. And what’s been striking — and I remember observing this when I was covering these issues, but it has been really striking to go back and see it in retrospect — is that the Biden administration was doing a lot at the same time, right? It was trying to stabilize the fossil fuel industry coming out of the Ukraine energy crisis. It was trying to help clean energy. It was trying to reduce emissions. But there’d be this pattern of, the Biden administration would pass the IRA, and a lot of the coverage on the IRA focused on the fossil fuel leasing provisions in the IRA, to the degree that I was like, did they read the law? You know, are they aware of what the policies are in this law? And then Biden would approve the Willow Project, or there’d be the fight over the LNG, and that would dominate attention. That would get a lot of coverage, and it would drive a lot of attention from climate activists, too.
Did what Biden was trying to do during this administration break through? Did what this administration was trying to do, did it break through with the people who needed to hear it? Meaning on the renewable side, on both sides, on the climate side and the fossil fuel side.
Granholm: I would say no, it definitely didn’t break through. But that’s more of a comment about our media environment, right? I mean, for media, with all due respect …
Meyer: This is an interesting question. Yeah, I mean, it ties into, like, what can the government do in terms of policy? How much do people notice what’s happening in terms of policy?
Granholm: I mean, unless there’s a massive controversy, people are looking at, you know, TikTok and Instagram, and so the things that get the big, you know, generate the most clicks are the ones that have some backlash, right? So those are the things that people notice. Do they notice the good news? No. Do people know, they have any idea about the amount of factories that are coming to America, or expanding in America? No clue. No clue. So that’s a question about, how do you do that? How do you get the word out? I’ve been across the country and almost every state you know, trumpeting this, but it still doesn’t rise to the level of consciousness.
And maybe it’s because people aren’t so interested in it, but I think that people want to see — I mean, the economy was the number one issue, and if you’ve got a factory coming to your area, that should be a positive, positive thing. But people don’t attribute it necessarily to the Biden administration. Is that a failure on our part of not yelling loud enough, etc.? Or is it a failure of the atomized media that we have? It’s just it’s hard to know going forward.
Meyer: Well, I think if it is a failure of the atomized media, it is a rather ominous sign for politics that tries to build things in the future. You’re the governor — you were the governor of a state that had ties to the fossil fuel industry, not in that it was extracting fossil fuels, but in that the main export, it uses them, right? What do you think Democrats should understand about the fossil fuel industry? And like, not only Democrats in government, but rank and file Democrats?
Granholm: Well I think, first of all, I think it’s important to understand the word transition. You know that we are in a transition from a molecule-based fuel system to an electron-based fuel system, to electrify. And that electrification, obviously, is so much more healthy for the planet, and it creates huge job opportunities. But the transition period, you know, we have 26 years between now and 2050, when we need to get to net zero — and notice it says ‘net zero.’ The fossil fuel industry is going to continue to exist. The net means that we have to find the technologies that reduce, obviously, greenhouse gas emissions from that fossil fuel industry.
People, also Democrats, understand, if they live in one of these states, that the fossil fuel industry provides a huge number of jobs, good paying jobs, and so you can’t just come along and say, get rid of all of that, because you’re talking about people and people’s economies and communities and towns. So it, there has to be a vision of build, build, build on the clean, and that market will expand, so that you can have a reduction on the fossil side. But I don’t think that we’re going to be fully rid of all fossil because I, and I think the technology will allow for us to clean up the fossil fuel side of things.
Meyer: Do you think that the voters, that rank and file Democrats — even beyond Democrats, just the type of voters the Democrats want to talk to — do you think they see this link between fossil fuels and their livelihoods and their economies? A link that Republicans are very eager to point to, but that I think Democrats sometimes avoid. Do you think that people see fossil fuels and associate their economic well-being with it?
Granholm: No, I don’t think they ,,, unless they live in one of those communities, right? I don’t think they necessarily link … I think people are really interested in reliable power, yeah, and they’re interested in how much that reliable power costs. That’s, for most people, they’re not looking, when they flip on the light switch, is it something that’s coming from natural gas-powered? Or is it something that’s coming from solar? People like the idea — and I can say this as a former governor, communities love the idea of creating jobs that are future-facing, that their kids will stay for. And the next generation, the Gen Y, Gen Z, they’re, you know, they don’t want to work in an industry that is polluting the planet. Many of them, I mean, some of them — of course, I don’t want to generalize. But for the most part, communities embrace these industries of the future. And so I think that is something that people listen to, and they understand that this transition is happening, and it is not just like flipping a light switch, that there has to be a glide path so that the communities that have powered us for the past 100 years can be part of powering us for the next 100 years, but using clean. But that transition period has to be one where you are sensitive to the fact that these many rural places feel very threatened.
Meyer: Democrats did a lot to help the fossil fuel industry during this administration, to stabilize it — after the pandemic, to stabilize it, after Ukraine, there was a Defense Production Act in the beginning to help the fracking industry. There was the use of the Strategic Petroleum Reserve to stabilize gas prices and oil prices. After Ukraine, the response they got from the fossil fuel industry was nothing but vitriol, nothing but, Biden doesn’t want us to drill. We hate the tone coming out of this administration. We’re not happy with this at all. Did Democrats do too much for the industry? Should they not have done what they did?
Granholm: Well, I don’t think Democrats … I don’t think we did it for the industry. We did it for the people. So for example, on the Strategic Petroleum Reserve, you know, gas prices were bumping up to $5 a gallon because of the dearth of supply globally as a result of Russia’s fuel not being on the market. So we wanted to make sure that people in the United States and around the world, but people in the United States in particular, weren’t paying for Putin’s war, and so releasing more supply obviously brought that down. I mean, today we’re $2 under what it was during that height. So it worked. And we have replenished the Strategic Petroleum Reserve at a price that was much more favorable to taxpayers. So it wasn’t, that wasn’t about saving the industry. That was about saving people’s pocketbooks.
We did help our allies on liquified natural gas exports, and they were reeling as a result, a couple of years ago, as you know, as a result of the war and Putin weaponizing natural gas. So that had an impact on the industry. It helped the industry. But our the strategy really is what’s going to help people here in the United States.
Meyer: There’s two different attitudes toward the fossil fuel industry, as — I’m going to generalize a little bit, but when I look at climate folks, folks who work in this professionally, who think about these issues day to day, whether they’re activists or engineers, I think people’s attitudes break down into two camps. And one of those camps is that the fossil fuel industry is a big enemy. They’re a big enemy of the transition. They’re going to fight the transition at all costs. They’re going to do everything they can to preserve fossil fuels. And just look at their support of Donald Trump, right? What they want from the Trump administration is shoring up demand because they want to lock people into fossil fuels.
And there’s another point of view that says, Well, you know, it’s not great, but the obstacles to the transition are more along the lines of technology development and deployment. And the fossil fuel industry, in the meantime, actually does provide a lot of energy for people, and you can’t ignore that. Where do you come down between those two camps?
Granholm: Well, I definitely think that industries tend to preserve their self-interest, right? And so the fossil fuel industry is all about fossil fuels. And so if they view another type of energy, cleaner energy, as a challenge, then they’re going to do everything, which — I mean, that has proven throughout history to have been the case.
I will say there are a couple of actors, several of the majors, who maybe have a little more of the luxury of being able to do investments in technologies that allow for a clean energy transition. That’s encouraging. I mean, when I … there’s a big oil and gas conference in Houston every year. And the first year I went, I was like, you guys, you should transition to be diversified energy companies, not just oil and gas. You guys have all of the resources. You’re the richest companies around. Why don’t you being … why aren’t you investing more in geothermal, for example, because you have the skill set. You’ve got, you use rigs, you go, you understand about extracting energy from beneath the surface. You’re experts in the subsurface. You got skill set. This began the big blowback. How dare you come and tell us what we should be doing? But it’s just realistic, and to be fair, some have, a lot have not.
Meyer: Some that got into it have now backed off of it, too, is the other thing, too. Because they’re just, the business is so … like, even if some leaders in the company are like, well, electricity looks like the future, we should get into it, the profit pattern’s totally different. The margins are totally different. So they then wind up getting back out, because fossil fuels just can be very profitable.
Granholm: This is why, to me, this transition, we have to almost overbuild on the renewable side, so that it is, so that it were, we just have so much momentum there. And the cost, when you overbuild, you’re scaling, your costs come down, and it becomes a game of cost, right? Our natural gas is so cheap, and so that’s why it’s a difficult thing.
Meyer: Right. This is like, in some ways, the good and the bad, right? Is that we have all this cheap natural gas, which means we’ve had fairly cheap electricity, right? And that means we attract a lot of industry that is not going to Europe. For instance, China also has cheap electricity because they subsidize it. They burn a lot of coal at state subsidy. They do a lot of generation. I know it’s just a tricky … it is tricky.
Granholm: The whole thing. It’s complicated. It’s not binary. I mean, really, we have to think about it: Where do we want to be in 2050? We want to be at net zero. How do we get there? How do we back into that? And I think the electricity grid side of things will come along very, very quickly. You know, it’s going to be difficult to decarbonize heavy industry, and we’re working on strategies to do that. You know, the building sector — really, really important to be able to address that. So, bottom line is, it’s complicated.
Meyer: Clean energy and electricity, let’s move on. So I think, as recently as 18 months ago, the word on electricity demand, on electricity markets, and what was happening in electricity in the U.S. was, yes, these AI companies are going to come along. They’re building a lot of data centers, but they’re not going to slow down the transition. I think that might have changed a little bit, that consensus might have changed a little bit since then, or I think it’s a little less clear, let’s say. What’s your view?
Granholm: I think first of all, we know the demand is going up, and data centers are a part of that. So is electrification of transportation. So are all these 900 factories coming on. So there are a number of reasons —
Meyer: Right, AI remains a relatively small percentage …
Granholm: — at the moment. But it’s, but there is no doubt that it’s a growing part of it. We have managed increases in demand in the past, and we will continue to do that going forward by adding all of this additional power, which is why, you know, we just need to continue to build generation. But what our conversations with the hyperscalers on these data centers — first of all, we believe that it’s really important to have these data centers, the AI data centers, in the United States, for national security reasons. We need to hold on to that. And if that’s true, and they build out these data centers, and they need another, whatever, 20 gigawatts of power for the data centers — bring your own power. BYOP. So bring your own clean power. Your shareholders and your commitments to clean energy, all of them have commitments like that.
So this is why you’re seeing these really interesting announcements coming from Google and Meta and Amazon, etc., partnering with geothermal companies like Sage and Fervo, partnering with small modular reactor companies like Kairos, are turning on power. So it’s, to me, very exciting that, not only that there may be, there’s an additive component to this. And there’s a willingness on the part of the data centers to actually pay a green premium for those, that addition, those additional electrons.
Meyer: Do you hear that they are willing to … is the willingness to pay that green premium durable? Because I do hear concerns that basically, once Trump comes in, or in a year from now, these companies are going to want to keep building AI data centers. They’re going to be starved for power, and they’re just going to build natural gas because it’s not coal, and they need the energy.
Granholm: I think that there will be some temptation to do that. And I think that while nuclear reactors, for example, are being built, they may rely, there may be some reliance on natural gas to be able to power — natural gas combined, perhaps with CCS, for example, carbon capture …
Meyer: Yeah, Exxon just made this announcement. It’s going to try to get into that.
Granholm: Yes, exactly. But these companies themselves have made commitments. I mean, Amazon’s commitment to clean has been part of their ethos. So they can’t just rely upon that. They have got to bring their own.
And let me just say one other thing, is that bringing their own and paying a green premium for that, they recognize the backlash of adding demand on a grid that is socialized across the rate base. They recognize, they do not want to have protests because people’s rates are going up because of this data center. And data centers, of course, they are fantastic for construction jobs, but they don’t employ a huge number of people. There are ripple effects, of course, in an industrial cluster, but they’re worried about the backlash — as they should be, and therefore bringing their own power and paying for it is a part of the strategy.
Meyer: Thinking more broadly, the Biden administration came in, and you came in with an interest in fixing issues in the U.S. innovation and deployment chain, let’s say, and in the chain of innovations, or the chain of scientific research, as it goes from the lab and the national labs to eventually the marketplace. And one big criticism, including from one thing that Jigar talks about the Loan Programs Office is trying to do, is make sure that when we invent technologies here, they don’t get then exported abroad, sold abroad, and then made somewhere else and commercialized somewhere else. Today, in 2024, is that chain working? And what are the biggest problems with that chain as you see them?
Granholm: Well, there’s so many, there’s so many opportunities for solving the biggest problem. So for example, we launched eight Earthshots to be able to reduce the price for this next-generation technologies like enhanced geothermal or floating offshore wind platforms or clean hydrogen. If we can reduce the cost of those technologies by 90% — and that’s what a lot of the labs are working on — that makes them irresistible.
And that’s, making these advanced technologies irresistible is part of the strategy. And that includes, you know, nuclear reactors. So we have a whole office of technology transitions led by Vanessa Chan, who came in from outside, and she has overseen all these liftoff reports for these clean technologies that have, really, industry benchmarks about where we should be and when we should be there. We feel really good about the trajectory of these technologies as a result of that work, and the work that’s being done by the labs.
There are a number of knotty problems, though, that are beyond those particular technologies. One of the knottiest problems, honestly — and I mean knotty, K-N-O-T-T-Y — is the grid getting the connection through. And I think that that is a problem that we’re working on, using AI to solve. But there needs to be a much greater emphasis on how to get the RTOs and the ISOs to get more power online.
Meyer: What technology in the clean energy … actually, let me just ask one follow up, which is, do you think that those Earthshot programs will be sustained through the next administration?
Granholm: I do because, I mean, if there’s a big support for hydrogen, for example. So how do we get the cost of hydrogen down? I think you can cherry pick, I suppose. But you know, carbon management is one of the Earthshots. So how do you reduce the cost of that so a lot of the technologies have support?
Meyer: Thinking across the full spectrum, full portfolio of technologies that the DOE has supported or gotten engaged on in the past four years, what technology are you most excited about?
Granholm: Well, I’m certainly super excited about the deployment of solar that has just been, just a rocket ship, which is very, very cool. I personally love, and would love to see a greater investment by the next — we would have done this if we had the opportunity to serve — I think, a greater investment in geothermal, enhanced geothermal. I think that has huge promise, to be able to do that. And I would also want to see a greater investment, a consistent investment in the grid. So we got, there’s $70 billion that is focused on the grid through these laws. And we were able to do spot efforts, to do reconductoring and get more power across existing lines, to do grid-enhancing technologies, to do undergrounding, all of that. But a consistent investment in the grid, kind of like what we have for the Highway Trust Fund. If we had a, we have a national grid, it’s, you know, the transmission and distribution miles, 7 million miles, and 75% of the grid is over 25 years old. We need a consistent investment for both to deal with these extreme weather events, cyber, and make, just making sure that we are upgrading the poles on which this electricity is carried.
Meyer: It is so different, and it’s interesting compared to highways. Because highways, it is a form of transmit, right? I mean, yeah, we move energy around on the highways. Highways are this big national system. The highways are also publicly owned. The grid is not. Do you think that’s an issue?
Granholm: Well, some of the grid, I mean, you know, it depends, right? I think that we just need to think about it differently. Electricity is just a fundamental human need, like transportation on a road, or even more importantly. So I just think we have to think differently about how we support the grid and this scattershot … and I know it’s, obviously it’s a federalism issue, as well. But we have a system like that for the roads, and we should take a look at that. There’s state roads, and the money goes to the states and they do it, and then there’s federal roads, you know, federal highways. I think we can borrow lessons, I’ll just say that, from what we have done with other infrastructure to support the grid.
Meyer: Looking back, I think one of the big promises, one of the big claims that Biden made coming into office was that government was going to do big things again. It was going to intervene and help the economy in big, strategic, supply-and-demand-oriented ways, right? It was going to bring back factories. It was going to bring back jobs. It could, government could be a major force for good in people’s lives, and that this would not only be good for the country in a kind of broad way, but also help defeat Trumpism.
Obviously, Donald Trump is about to be president again. But I also think the Biden administration did go much further in its economic policy than maybe the Obama administration did outside of healthcare. What did you learn from this go around? What did you learn from this outing at trying to do economic and industrial policy at the federal level, and trying to do big things with the government’s power?
Granholm: You know, when I was done being governor, I wrote a book about this. It’s called The Governor’s Story, where I was lamenting the fact that we didn’t have national strategy to be able to create jobs in our states. And the fact that the troika of laws — I would include the CHIPS and Science Act, as well — but there are three basic laws that are the basis for this table of industrial strategy, and that it has worked so well.
I feel like, you know, all of the folks who have in the past been total free market, “free market” players, government shouldn’t be involved, we shouldn’t be picking, “winners or losers,” or all of that, and allowing all these other countries to take our jobs. I think people have now seen this play out, the fact that you don’t have any hands on and other countries are playing. And I have, I have learned, and I hope others see that had, the government can play an enabling role for actually creating jobs in this country and making us competitive. Who knew policy actually works?
Meyer: I hear that …
Granholm: I hear a “but” coming.
Meyer: Kamala Harris, still, though, did not win the presidential election. So I think, to some degree, the Biden administration has proved you can do these things, and that they elicit a response. I think, unfortunately, what it’s also demonstrated is that this might be somewhat disconnected from political outcomes. So what lesson do you learn from that?
Granholm: I mean, yeah. Lesson A: It works, yeah. Lesson B: How do you, how do you communicate that in a way that people understand that it’s working? How do we, yeah, how does that happen? It’s that. It’s back to the question we started to talk about at the front end. I wish I was a great communications guru and I could tell you how to do it. It’s one of those things that you almost have to, because it was so soon, in the results-bearing part of this strategy, meaning that you haven’t, people haven’t been hired for all those factories yet because they’re still under construction, or they’re still being contracts.
I still feel like it’s not over yet. The story’s not over yet. In the immediate, yes, there was a communications gap because people didn’t believe it. And I think people are angry, and they don’t trust government to begin with. And so if you’re coming in as a candidate saying that your government’s bad and they don’t believe you, and I’m going to fight for you, that’s a more compelling message than, yeah, we we have all these factories that are just starting. You need to wait a few years, but it’s gonna happen, I swear. So how do we communicate that? You’re in the communications business — tell me.
Meyer: Well, do you think it was a mistake to communicate so many of these changes, so many of these policies were communicated through manufacturing — I mean, it’s what we’re doing right now. Not many voters actually work in manufacturing. We have a largely services-dominated economy. Was that a mistake?
Granholm: I’m sort of manufacturing-centric.
Meyer: I was gonna say, in some ways, because of governor of Michigan, you have an anomalous …
Granholm: Totally. This, yeah, it’s not, it’s not how everybody else sees it, for sure. And I totally, I totally hear you on this. So here’s, you know, here’s another data point, is that — or factor: Part of the Inflation Reduction Act and the Bipartisan Infrastructure Law was to give people the ability to weatherize their homes, to have rebates for appliances that made them pay less on their bills. The cool thing about this was that, is that 49 out of the 50 states have said, yes, I want that program, and they’re applying for the funding to make that happen. But it’s just happening now.
Meyer: I was going to say, I remember looking at the timelines for this and being like, they’re not going to get that out by the election.
Granholm: Yeah, I mean, it’s just, the four-year period of time is just such a speck on the timeline that’s necessary to actually see it to fruition.
Meyer: What do Democrats need the government to do that it cannot do, or that it has not been able to do?
Granholm: Besides communicating these, the successes, and making sure that we can move more quickly on …
Meyer: Communicating, or, yeah, maybe it’s that these were great rebate programs, and also, the law passed in August 2022, and we’re just getting them out now, right? Like, it took two years to get through that whole process.
Granholm: Yeah, and that is a, that’s an issue, I think, that government has to deal with. In general, when you have a new program, you’ve got — rightfully so — all of the oversight, and making sure. And then the states have — if you push something through to the states, this is the same thing with the NEVI program, the vehicle program went to the states.
We have oversight on the federal level. States have oversight on the state level. It’s a new program. So, for example, for electric vehicle chargers, the NEVI program was to get these chargers to the hardest spaces, the spaces where no private sector entity would go, where there was no electricity. So the contracts took a while to figure out, and how to get them, and the states had to figure out their own processes. So they’re starting, we have 24,000 coming online from the NEVI program. But again, they will come online over the next couple of years — years. So the slowness of the systems, that is an inherent challenge.
If you move too fast, though, then you end up making mistakes. You don’t want to see some wrongful spending. So it is, I mean, government takes, the government, wheels of government, eventually they’ll get it right. There’s a lot of oversight. You want to make sure there’s no waste, fraud, and abuse, but it may not happen on the convenient timelines of a four-year term.
Meyer: I understand how, you go slow, and it avoids mistakes, and you get your one shot to do things, I will say, looking back, the times when the government has been able to pass policies that resulted in political wins — and specifically, when Democrats were able to expand the government’s powers in ways that result in political wins, to build things, do big things, they did them pretty fast. The 1930s, like, the New Deal was a program, right? Yeah, all the LBJ programs were relatively fast.
Granholm: Yes, I totally agree. And in history, there was less concern about a lot of the stuff that, yeah, that there have been embedded processes. I mean, it’s, I think, immediately, of the DOGE efforts to try to reduce bureaucracy. And honestly, there is truth to that. There should be some reduction of the steps it takes to get something accomplished. And there needs to be sort of a value-stream mapping of the bureaucratic systems. This is true.
We need to speed up permitting. We know we need to do that. And you know, the friction, the litigation, the NIMBY issues, in addition to the the government’s internal processes, it’s just too much. So can there be streamlining to make things move more quickly? There can be, and I think that should be a focus, as well, for Democrats, not just Republicans.
Meyer: After four years at the nerve center of decarbonization in the government, what do you think is the biggest obstacle to decarbonization, and to doing the energy transition?
Granholm: I think right now, I mean, I would say the most immediate big obstacle is that we have 3.7 terawatts of clean energy waiting to get into the queue, to be, build out that generation from the transmission point of view. If I could tackle one thing, one thing quickly, it would be the interconnection queues.
Meyer: Wow. And so you would be willing to pay, you would be willing to make other concessions to handle transmission?
Granholm: I would be. And then the, I would, there’s one other thing I would say, is, I would want to invest much more in baseload power, baseload clean power. So bigger investment in geothermal, bigger investment in hydroelectric — even distributed hydroelectric, you know, making dams where they don’t already exist, stuff like that that really are … you know, I believe nuclear is a really important part of the clean energy future. And those nuclear reactors, small modular reactors, I think are really important. I’d continue to invest in them. I would also continue to invest in fusion.
Meyer: There’s a big tradeoff that the administration has kept dealing with, I think, or that it has kept running up against during the past four years, and to some degree, is really inherent in the design of the Inflation Reduction Act, which is, when you decarbonize, when you seek to make this transition happen, do you optimize for deployment and commercialization, or do you optimize for reducing emissions?
I think the clearest version of this tradeoff is with the hydrogen tax credit, where there is a tradeoff between getting as many electrolyzers out there as possible, trying to build as much electrolyte electrolysis capacity on the grid as you can, versus making sure that those electrolyzers don’t indirectly increase emissions across the national economy. In that case, but even more broadly, how do you come down? If there’s a tradeoff between deployment and reducing emissions, where do you come down?
Granholm: I don’t like the binary choice. I will say this: I’m very big on actual deployment, knowing that there will be continuous improvement on the technologies that are being deployed, including GHG emissions. I just think that people need to get used to, I mean, you know, when you put out a product for the first time, there’s always going to be things that you want to improve on, and because the overall goal is to reduce emissions, that’s got to be a primary thing of the product that is being put to market. So I think both have to happen, but the deployment is is utterly priority for me. Get it out, and then improve.
Meyer: Do you think … are the technologies ready, at this point, to think about reducing emissions? Or do we just need to deploy a lot?
Granholm: Well, if you’re talking about electrolyzers, we still have to get these hubs up and running. And this is another thing, waiting for guidance on the tax credit associated with hydrogen is, is a whole ‘nother thing. But, but the bottom line is, once they’re, I mean, the whole point about doing hubs, for example, in hydrogen, and using fuels that are either renewable or natural gas with carbon capture or nuclear — the bottom line is, we’ve got to get these hubs, are the places where we’re going to experiment about what works and what doesn’t work. We’ve got to get the products in the hands of those who are going to be using them so that we can learn from them. New technology requires deployment in order to learn, in order to improve.
Meyer: Looking back on the IRA — and to some degree, this troika of laws, but really the IRA — what is your biggest regret about how they played out?
Granholm: My biggest regret is that there wasn’t, we weren’t able to get consistent investment in the grid through the IRA.
Meyer: From the, from the point of view of the law being designed …
Granholm: Yeah, from the design of it. That could be something that that is worked on in the years to come.
Meyer: What’s your biggest regret on implementation of the law? Speed, clearly,
Granholm: Speed, yeah. Trying to get through all of the hoops that you need to get through to do it, right? Why was it so slow? Again, because of all of — let me just say, I’m super proud of the work that our team has done, and the fact that we will have 98% of the funding out from these rounds of funding. That is amazing.
Meyer: I’m going to ask you what you’re proudest of.
Granholm: But it is, I regret we weren’t given the opportunity to serve in a second term to make sure that we were able to carry this forward, and that people can see that the benefits of these laws actually create an industrial strategy that reshores manufacturing, and it generates clean electricity and creates opportunity for communities all across the country.
Meyer: Ultimately, the IRA did a ton of industrial policy through the tax code, and then it did a ton of grants and loans through DOE. And I think there could be criticisms of the speed of DOE getting loans out, let’s say.
I mean, something I’ve heard from companies is like, look, LPO is — $400 billion of loan authority, they only got out $54 billion. That number might go up a little bit in the next few days. But I’ve also heard criticism that we tried to do all this stuff through the tax code, and it just took a long time to get that tax guidance out, and we were trying to jam a lot through a relatively small tax policy office. What would you change … if there were to be another attempt at industrial policy, be it on climate or anything else, right? Because we want to do … like, Democrats do want to have goals for other sectors of the economy. What would you change about that implementation, or that way of writing the law, to make sure that industrial policy is effective in the U.S. government as it exists?
Granholm: It’s hard for me to sa,y this is a process that I would remove from the system, because it, we have so many technologies, so many agencies involved, and I think that there needs to be a comprehensive look at it. I do think that in a next iteration, incentivizing or removing waste streams from permitting has got to be a priority to get this stuff deployed. As I say, incentivizing investment in the grid has got to be a priority. Incentivizing investment, specifically, in baseload clean technologies.
Meyer: How would you incentivize those things without just winding up in another slow process.
Granholm: So tax code’s a little more efficient, yeah. So the tax code, I think, gets the private sector to sit up and take notice, and move, yeah. And so doing it through the tax code, as opposed to grants or loans, that happens, assuming that folks are but the guidance —
Meyer: This time it did take long, a long time for the guidance to come.
Granholm: Yeah, for some. I mean, for like, solar stuff that we’ve, you were used to, instead of being incentivized in the tax code, that all obviously went very quickly and will continue to move. But it’s just, it’s the startup of these things where we haven’t written guidance before, where it’s a technology that has subsets of supply chains, what’s involved that, you know, and you have a small office. So I’d invest in the office to be able to expand the ability to get these through a little more quickly. But I think that ultimately, even though it takes a little bit of time, it works, and it will work, and we will see the benefits of this over the next few years, and I hope people recognize the incredible courage that it took for Joe Biden to act, create industrial policy in America where we had not done it before.
Meyer: What are you proudest of?
Granholm: I’m proud to work for this administration, for this president. And I’m really proud of our team, that we have been able to do so much in so short of amount of time, with all new programs. It is an amazing, amazing thing to have been in this position at this time. I feel so utterly lucky, and I hope the next guy sees it for the jewel that it is.
Meyer: I think you’re in this role for another 35, 36 days.
Granholm: I think it’s like 26 working days.
Meyer: After that, what sort of problems are you thinking of next?
Granholm: Oh, I’m still thinking of all of these problems. I’ll still be active in this space in some way, shape, or form.
Meyer: Is there a particular set of problems you’re eager to tackle?
Granholm: I don’t know what’s going to … I have no idea what I’m going to be doing next, but it’s got to be in this basic work. In the clean space is where I will be.
Meyer: So at the end of every episode, we ask — Jesse and I, my co-host, do an upshift or a downshift. You just pick one. And an upshift, you pick something from the news or something that you’ve encountered recently that either is making you feel more upbeat about the energy transition, or more downbeat about the energy transition, about our ability to make it happen, about its ability to, our ability to decarbonize the economy. And so to conclude, I wanted to ask you, do you have an upshift or a downshift to share with us?
Granholm: My upshift is that I am optimistic about the durability of these programs that I think, regardless of who’s in the White House, it is inexorable, this clean energy transition — because, because the Inflation Reduction Act has made investing in America irresistible, because there’s bipartisan support, and because of all of the mayors and the governors and the members of the, members of the private sector who have raised their hand and said, We got this. We’re taking the baton.
Meyer: Secretary Granholm, thank you so much for joining us today.
Granholm: I appreciate it. Thank you, Rob.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
On Cybertruck deaths, Texas wind waste, and American aluminum
Current conditions: Yet more snow is dusting New York City with at least an inch fallen already, though that’s set to turn into rain later in the morning • Authorities in Saudi Arabia issued a red alert over a major sandstorm blasting broad swaths of the desert nation • Heavy snow blanketed Romania, halting transportation and taking down power lines.

In his State of the Union address Tuesday night, President Donald Trump unveiled what he called the new “ratepayer protection pledge.” Under the effort, the White House will tell “major tech companies that they have the obligation to provide for their own power needs.” By mandating the bring-your-own-generation approach, the Trump administration is endorsing a push that’s been ongoing for months. The North American Electric Reliability Corporation, the U.S. grid watchdog, called for data centers to build their own generators. An industry-backed proposal in the nation’s largest power grid would do something similar. “This is a unique strategy,” Trump said. “We have an old grid that could never handle the [amount] of electricity that’s needed.” With tech companies constructing new power plants, Trump said, towns should welcome data center projects that could end up lowering electricity rates by inviting more power onto the local grid.
The political blowback to data centers is gaining strength. It is, as my colleague Jael Holzman wrote recently, “swallowing American politics.” On the right, Senator Josh Hawley, the populist Republican from Missouri, introduced legislation this month to restrict data center construction. On the left, Senator Bernie Sanders, the democratic socialist from Vermont, reiterated his proposal this week to halt all data center projects. In the center, Pennsylvania Governor Josh Shapiro, a Democrat with unusually strong support among his state’s GOP voters, recently outlined plans for a more “selective” approach to data centers, as I reported in this newsletter.
Trump isn’t the only Republican pushing back against the data center blowback. On Tuesday, Mississippi Governor Tate Reeves delivered an impassioned defense of his state’s data center buildout. “I understand individuals who would rather not have any industrial project in their backyard. We all choose where to live, whether it’s urban, suburban, agrarian, or industrial. I do not understand the impulse to prevent our country from advancing technologically — except as civilizational suicide,” Reeves wrote in a post on X. “I don’t want to go gently. I love this country, and want her to rise. That’s why Mississippi has become the home of the world’s most impressive supercomputers. We are committed to America and American power. We know that being the hub of the world’s most awesome technology will inevitably bring prosperity and authority to our state. There is nobody better than Mississippians to wield it.”
Replying to Sanders’ proposal, Reeves said he’s “tempted to sit back and let other states fritter away the generational chance to build. To laugh at their short-sightedness. But the best path for all of us would be to see America dominate.”
Sign up to receive Heatmap AM in your inbox every morning:
The subcompact Ford Pinto gained infamy in the 1970s for its tendency to explode when the gas tank ruptured in a crash. The Ford Motor Company sold just under 3.2 million Pintos. By the official death toll, 27 people died as a result of fires from the vehicles exploding. Tesla has sold more than 34,000 Cybertrucks; already, five people have died in fire fatalities.
That, according to a calculation by the automotive blog Fuel Arc, means the Tesla Cybertruck has 14.52 deaths per 100,000 units, compared to the Ford Pinto’s 0.85 deaths. “The Cybertruck is far more dangerous (by volume) than the historic poster child for corporate greed and grossly antagonistic design,” Fuel Arc’s Kay Leadfoot wrote. “I look forward to the Cybertruck being governmentally crash-tested by the NHTSA, which it has not been thus far. Until then, I can’t recommend sitting in one.” That is, however, based on the lower death toll figure for the Pinto. Back in 1977, Mother Jones published a blockbuster cover story under the headline “Pinto Madness” claiming that the number of deaths could be as high as 900.
Texas accused the recycling company Global Fiberglass Solutions of illegally dumping thousands of wind turbine blades near the central town of Sweetgrass. The company allegedly hired several subcontractors to break down, transport and recycle the blades, but failed to properly dispose of the waste and instead created what Windpower Monthly called a “stockpile” of more than 3,000 blades across two sites in the town. Attorney General Ken Paxton, a Republican candidate for U.S. Senate, seized on a Trumpian critique of the energy source, saying the dumps damage “beautiful Texas land and threaten surrounding communities.”
Off the Atlantic Coast, meanwhile, Orsted is at a transitional moment for two of its offshore wind projects. The Danish developer just brought the vessel Wind Scylla to port after completing the installation of turbines at its Revolution Wind project in New England. The boat is headed to New York next to start installing the first wind turbine at Sunrise Wind, according to OffshoreWIND.biz.
Last month, I told you that Century Aluminum inked a deal with Emirates Global Aluminum to build the first smelter in the U.S. in half a century in Oklahoma. On Tuesday, the U.S. Aluminum Company, a local firm in the state, joined the project, signing an agreement to “explore the development of an aluminum fabrication plant near the new smelter.” If completed, the project — already dubbed Oklahoma Primary Aluminum — would roughly double U.S. primary production of the metal.
The Biden administration had placed what Heatmap’s Matthew Zeitlin called “a big bet on aluminum” back in 2024. By spring of last year, our colleague Katie Brigham was chronicling the confusion over how Trump’s tariffs on aluminum would work. With the recent Supreme Court ruling upending Trump’s trade policies, that one may remain a headscratcher for a little while longer.
Another day, another landmark energy investment from Google. This time, the tech giant has made a deal with the long-duration energy storage startup Form Energy to deploy what Katie wrote “would be the largest battery in the world by energy capacity: an iron-air system capable of delivering 300 megawatts of power at once while storage 30 gigawatt-hours of energy, enabling continuous discharge for 100 hours straight.” The project will power a data center in Minnesota. “For all of 2025, I believe the installed capacity [added to the grid] in the entire U.S. was 57 gigawatt-hours. And in one project, we’re going to install 30 gigawatt-hours,” Form CEO Mateo Jaramillo told Katie. “What it highlights is, once you get to the 100-hour duration, you can really stop thinking about energy to some extent. “
Rob checks in with Near Horizon Group’s Peter Freed about the AI boom’s power needs.
Just a handful of tech companies plan to spend nearly $700 billion combined this year investing in artificial intelligence — and much of that money will go to data centers and the energy used to keep them on. How is this boom transforming the American energy system, and what does it mean for clean energy?
On this episode of Shift Key, Rob is joined by Peter Freed, a founding partner at the Near Horizon Group and the former director of energy strategy at Meta from 2014 to 2024. They discuss why data center developers opt for certain energy sources over others, why AI is driving an unprecedented off-grid natural gas boom, and why batteries now pair especially well with gas. Yikes!
This conversation was originally recorded for a webinar hosted by Heatmap Pro. Shift Key is hosted by Robinson Meyer, the founding executive editor of Heatmap News.
Subscribe to “Shift Key” and find this episode on Apple Podcasts, Spotify, Amazon, or wherever you get your podcasts.
You can also add the show’s RSS feed to your podcast app to follow us directly.
Here is an excerpt from their conversation:
Robinson Meyer: We know there’s this giant capex surge coming from the hyperscalers. I mean, it’s reached the point now where tech companies’ stocks suffer when they announce investment because they seem to be in an arms race of spending on data centers. We were just talking about the behind the meter gas boom. There’s a lot of renewable energy developers in this audience, or battery developers. How should they be thinking about this moment and what do they need to be doing to make their projects or to work with data center developers in the most attractive way?
Peter Freed: I’ll bring us back a couple of minutes to when I said, look, if you’re a data center developer and you’re building gas plus storage and you’re thinking of that as a pretty complicated thing, someone is really going to have to work out on explaining why the introduction of a variable renewable resource into that configuration is worthwhile.
And obviously there are people that believe that that’s true. Intersect believed that that was true and it worked out really, really well for them. There are ways to tell that story. And I think that the renewable energy development community probably still has some work to do to help explain that. So that’s sort of thing number one — like, the closer you get to the operations of the data center facility, the more work you’re going to have to do to explain why you believe that the integration of renewables into that makes sense.
Now, you can remove yourself somewhat from the actual operations of the facility. And this is where we get into bring your own capacity conversations. And you know, there’s been some really interesting stuff sort of talking about, okay, maybe there is a utility which has sufficient wires capacity as — and like, there’s enough room on the transmission lines to plug a data center in and turn the lights on, but they don’t have enough market capacity. Like, they don’t have enough of the financial products required by the RTO that they operate in to serve that facility. And so that can become an interesting opportunity for renewables in particular, storage in particular, trying to figure out how to put together these bring your own capacity products to serve data centers.
And I’ll say, you know, when I first heard about these bring your own capacity opportunities, I thought that they were pretty niche. I was like, okay, well, you know, a utility has sufficient wires capacity to serve a giant data center, but they don’t have capacity in the market. Like, that feels like something that’s not going to happen that often. But apparently, I mean, I was incorrect.
You can find a full transcript of the episode here.
Mentioned:
Breaking Down the Doomsday AI Memo That Spooked Markets
Inside Form Energy’s Big Google Data Center Deal
The New York Times on AI’s polling problems
Previously on Shift Key: What’s Really Holding Back New Data Centers
This episode of Shift Key is sponsored by …
Heatmap Pro brings all of our research, reporting, and insights down to the local level. The software platform tracks all local opposition to clean energy and data centers, forecasts community sentiment, and guides data-driven engagement campaigns. today to see the premier intelligence platform for project permitting and community engagement. Book a demo today to see the premier intelligence platform for project permitting and community engagement.
Music for Shift Key is by Adam Kromelow.
This transcript has been automatically generated.
Subscribe to “Shift Key” and find this episode on Apple Podcasts, Spotify, Amazon, or wherever you get your podcasts.
You can also add the show’s RSS feed to your podcast app to follow us directly.
Robinson Meyer:
[0:08] Hi, I’m Robinson Meyer, the founding executive editor of Heatmap News. It is Wednesday, February 25th. I don’t know if you were paying attention, but on Monday, the S&P 500 dropped over 1%, as did the Nasdaq, because of a memo from an investment research firm that was basically a science fiction short story. We’ll put it in the show notes, but it argued that AI is going to work, is going to be so successful that it will in fact cause mass unemployment and trigger a recession. Now, it wasn’t the first memo or argument that there might be bad effects for the economy if huge amounts of white collar workers are put out of work.
Robinson Meyer:
[0:46] But it was the first that argued it may be in a persuasive way, in a mechanical way. And for whatever reason, it caused absolute carnage in the stock market. And I think it showed not only that artificial intelligence is the biggest story in the American economy right now, something we already knew, but that nobody knows what’s going on with it. And that holds true for data centers, which are the biggest topic in energy and climate and electricity, but that are changing in a way that is very hard to track. And so on this episode of Shift Key, we’re going to talk to someone who is tracking the way that data centers are changing. A year ago, Jesse and I spoke with Peter Freed about exactly what the data center buildout was doing for electricity and renewables. Peter is great. He’s a founding partner at the Near Horizon Group. He has over 20 years working at the nexus of clean energy, climate, and computing.
Robinson Meyer:
[1:31] And I think most saliently for 10 years from 2014 to 2024, he was director of energy strategy at Meta. He was right up against the coalface of energy procurement for data centers. A year ago, we talked to him. We learned a lot about what data centers meant for electricity and renewables at that moment. But a lot has changed since then. And so on Monday, the same day the stock market was down 1%, I talked to Peter Freed for a Heatmap Pro webinar about how data centers have changed over the past year and what he thinks they’ll mean for electricity and renewables and energy and emissions going forward into 2026. It was a really educational if disconcerting conversation. We talked about the enormous surge, the $600 billion to $700 billion of investment in data centers that are going to happen this year, why data centers are driving a huge off-grid natural gas build out now, and why data centers are also splitting the battery industry from the renewable industry. It was really, really interesting. I learned a lot, and we liked our conversation
Robinson Meyer:
[2:30] so much that we’re releasing it now as a special Shift Key episode. Thanks as always for listening, and let’s go to that conversation now.
Robinson Meyer:
[2:41] Hello, everyone. I am Robinson Meyer, the founding executive editor of Heatmap News, and you are watching Heatmap Pro’s special live webinar presentation of Shift Key. It is Monday, February 23rd, and I’m going to welcome our guest in for a moment. You can already see he’s joined us here in the virtual recording studio. But before we get started, I just want to put a word in for our sponsor today, keeping it all in the family. Our sponsor today is none other than our very own Heatmap Pro. If you want to keep following how data centers are reshaping the energy transition in practice, that is the work we do every day at Heatmap. Our pro team tracks where energy and data centers projects move forward, where they get delayed or canceled, and how community response shapes outcomes. We have data nobody else has on renewable energy projects as well as data center projects and the political risk that they’re facing. If you’re interested in learning more about the software, visit heatmap.news/pro. That’s heatmap.news/pro. Okay, let’s get started. Peter Freed is our guest today. He’s a founding partner at the Near Horizon Group. He was director of energy strategy at Meta from 2014 to 2024. And he brings to this conversation over 20 years of working at the nexus of
Robinson Meyer:
[3:51] clean energy, climate, and recently data centers. It’s always great to talk to him. Peter, welcome to the Heatmap Pro live Shift Key.
Peter Freed:
[4:00] I’m so excited to be here. It’s been like a year since we did one of these. So there’s a lot to talk about.
Robinson Meyer:
[4:07] There’s so much to talk about. Okay. So I want to actually start exactly there. So the last time we talked, it was February 2025. It’s been exactly a year. Let me start by asking, are you busier now than you were a year ago?
Peter Freed:
[4:21] That’s a great question. So, you know, I have built my current professional life on trying not to be too busy. So I’m doing okay at that. I could certainly be unlimitedly busy right now. There is a lot going on. The market has certainly not settled down. In fact, if anything, it’s gotten crazier. So yeah, it’s pretty nuts right now.
Robinson Meyer:
[4:41] So a year ago when we spoke, one of the key themes is the amount of speculative froth in the market that I think your term was, you know, two guys in a pickup truck announced they would be data center developers, and then they were treated as data center developers, and that was entering the math. Has any of that froth started to fall out? Or is it more intense now than it was a year ago?
Peter Freed:
[5:04] So the answer is yes and no. I think a couple of different things are happening. And I always find myself when I’m talking to you, Rob, referring to your own reporting, but I’m actually going to start there. Rob does not pay me to do this. I just really like the work that Heatmap does. So I think like a lot of the speculative activity that began in 2024, like that we were talking about a year ago in early 2025.
Peter Freed:
[5:26] Especially from inexperienced developers, well, we are now starting to see some of those projects kind of falling out and maybe the less polite way to say it would be falling apart. Like, you know, these are very complicated projects. They are getting bogged down in permitting processes. They’re getting bogged down in community opposition and all of this other stuff. And so I think that, you know, Heatmap Pro had done some really interesting reporting around the number of projects that had failed in 2025 and that that was growing. And I’d provided some comments on that then. But I think that’s what we’re starting to see, right? Like, I am someone who believes that maybe 10% of the projects that have been proposed around the country will actually end up being fully built out data center campuses. And part of what falling apart means or falling out of that process means is just that like projects will be canceled in various ways. Commercially, you won’t really see that. Maybe in the regulatory process or utility interconnection proceeding, the timelines will be so long that someone will fail to do it. The most public stuff, of course, is what we’re seeing around community opposition and these, you know, town hall meetings where dozens of people will show up and be protesting a facility for whatever reason. And then I think ultimately the financing, you know, people have to figure out how to secure the money to build these projects.
Peter Freed:
[6:41] And as it turns out, it’s a lot of money. And while there is capital available, like figuring out how to put those pieces together has been challenging. So I think that’s the yes stuff has fallen apart side.
Peter Freed:
[6:52] On the other side, there is still a ton of activity going on, people bringing new things to the market. And, you know, I can’t tell you how many like
Peter Freed:
[7:02] calls I get where someone’s like, hey, I’ve got a refrigerated warehouse that has electrical capacity. Can I turn that into a data center? So that sort of stuff is still happening. And I think maybe the difference is people have realized a little bit more that if you have a refrigerated warehouse that you want to turn into a data center, that maybe you shouldn’t do that yourself. And so I think what I’m seeing a little bit more of is partnership between the more established and experienced data center developers with people that bring other pieces of the puzzle to the table. And so there’s still a lot going on, but I do think that maybe the level of complete ridiculousness where it’s just very random people trying to do very random things, like maybe that’s coming down a little bit.
Robinson Meyer:
[7:47] We’re going to talk about every facet of this question, but what is the biggest bottleneck right now for projects?
Peter Freed:
[7:54] So I bet if you asked five people in the industry that question, you get five different answers. I have always said, and I think I said this to you the last time we talked, that I am very confident in the ability of the power sector, not necessarily the utility sector, but the ability of the power sector to meet the needs of the AI demand signal. And I do think that that is continuing to be the case. Last time we spoke, you all had done some really good reporting around the constraints on gas turbine supply chains. And those supply chains were indeed constrained. They’re still relatively constrained. But what we’re seeing is people are building data centers with a whole bunch of different thermal generation technologies, very small reciprocating engines, simple cycles, aeroderivatives, like all of the different flavors of put gas into it, power comes out. And so I think, you know, we’re definitely seeing that sort of thing happening more and more. And so, you know, the power sector is figuring out how to do this. So there are still plenty of people that say, oh, the power sector is the big bottleneck. I’ve never really thought that the power sector would be the key bottleneck. It might just look different than we thought it was going to look.
Peter Freed:
[9:04] I am hearing on the ground that now this is something that I’ve been wondering about slash worried about for a long time. Construction labor is becoming very constrained in many markets, in particular skilled trades. So skilled electricians have been a group of workers that have always been challenging to get enough of a data center construction. And now we’re ballooning the industry. So skilled electricians are an issue. Long lead equipment has and will continue to be an issue. So this is not necessarily the generators themselves, but on the electrical side, we’re talking circuit breakers, we’re talking transformers, and, you know, people are ramping up production capability, people are trying, you know, there’s a couple of startups that are doing like new cool power electronics for that kind of stuff. So that’ll need to catch up. But, you know, these are sort of the ups and downs. Is it power? Is it labor? Is it long lead equipment? Is it chips? Is it, you know, you get all of those answers even today.
Robinson Meyer:
[9:59] I feel like a year ago, there was this idea that only renewables had speed to power. And then really, X.ai and Elon Musk demonstrated that you could still get gas turbines if you didn’t care about the pollution that you pumped into an urban neighborhood or any of the other externalities. Do renewables still have an edge when powering data centers? Or has X.ai and this idea that there’s all sorts of gas available to you if you care far less about the externalities or are willing to pay to not care about
Robinson Meyer:
[10:30] them. Has that kind of changed the game?
Peter Freed:
[10:32] I think that there are certainly companies which still care about the clean aspect and the hyperscalers themselves. So really, we’re talking, you know, Meta, Microsoft, Google, and Amazon who have public clean energy commitments.
Peter Freed:
[10:46] You know, you look at a company like Google, which I think very philosophically has been chasing this for a long time. We’ve seen them do all sorts of interesting things. They appear to still be quite interested in using renewables as part of the actual power solution for the data center. So they recently acquired Intersect Power. That is a company which was formed around the idea that you could do these campuses that included solar and storage and some gas. But like mostly it was a clean orientation on that. Google has also announced a CCS project. So that’s sort of one bookend. The other bookend, hopefully, is that people continue with their existing renewable energy commitments, which just say like, all right, however much power we’re going to use, we’re going to make sure that a similar quantity of power is deployed onto the system. And so, you know, that that is still happening. The whole like speed to power equals the fastest way to get data centers on and that we’re going to do renewables as part of that. I’m not seeing a huge amount of that happening right now. Now, you know, Cloverleaf Infrastructure is another fantastic company that has done some really, really creative things working with utilities to put portfolios together that are largely clean portfolios and bring data center capacity on. But the other thing that I’ve been saying to people for a while now is like, if you figure out how to do ...
Peter Freed:
[12:09] Very rapid deployments with gas. And that in and of itself is actually extremely complicated. Like the engineering required to get all of the power quality and reliability characteristics that you would get from a grid interconnection from a standalone power plant. You’re going to do that with gas turbines and storage in most cases. Adding the additional layer of renewables to that is mostly seen as a complicating factor right now. So, you know, there’s a bit of interest, but it’s not universal interest.
Robinson Meyer:
[12:39] But you’re also describing, I think, a pairing that might be normal for data center developers to think about, but a pairing that I think in climate world, we don’t think about that much, which is gas plus batteries. Oh, yeah. Which we normally talk about solar plus batteries as being what, you know, the kind of classic pairing, but gas plus batteries is developers value it because it basically adds reliability onto their gas.
Peter Freed:
[13:01] That’s right. So, So, you know, data center reliability, the industry parlance is like nines of reliability, right? So that means like 99.999, five nines, 99.99. And so if you want reliability that is nines coming after your 99% uptime, you typically do have to add some battery storage. And data centers, by the way, I always used to say like data centers are the biggest battery storage installations on the grid and have been basically forever because they have uninterrupted power supplies inside of the facility. So these are batteries that are inside of the data centers. They usually have a minute and a half to five minutes of backup capacity. And this is for little blips in power quality or, you know, something happens. We’re talking about utility scale batteries, multi-hour batteries now, where people are adding that to the design of the facility in order to get you
Peter Freed:
[13:56] those higher reliability numbers. And some people are requiring that, some people aren’t, and it’s driven as much by cost as anything else.
Robinson Meyer:
[14:04] Are most of the batteries like lithium-ion, standard, multi-hour ... You know, battery installations, or are we seeing these more exotic or kind of frontier batteries with like multi-day storage coming online at projects?
Peter Freed:
[14:16] Yeah. I mean, so right now it’s off the shelf lithium ion, either two- or four-hour, and there’s a lot you can do with that. There’s a much more interesting conversation, which I think we will have at some point in the next hour, 45 remaining minutes around like what you might do with those longer duration batteries. You know the simple fact is there just aren’t enough of those yet to sort of see more deployments like you know i was chatting with some of the folks at form not so long ago i was like oh my god like such a great technology like how do we get a 100X your current supply chain onto the market because think what we get you know and it it just takes time it.
Robinson Meyer:
[14:56] Takes time yeah yeah before we move on from kind of where we’ve been where we’re going last year when we talked it was right after DeepSeek. And actually, there’s been reporting today that DeepSeek is about to come out with its next model. The last time this happened, it was shocking to people because the DeepSeek model seemed to use so much less electricity than any of the current American frontier models. And there was this discussion of Jevons Paradox, which to vastly simplify, is the idea that as something gets cheaper and less energy intensive,
Robinson Meyer:
[15:28] we tend to use more of it rather than less. We tend to enjoy those energy savings by using more of the thing rather than, you know, decreasing our overall energy usage. So a year after DeepSeek with another model on the horizon, has the idea that AI is going to be subject to Jevons Paradox borne out, or has all the use been at these frontier models?
Peter Freed:
[15:52] I am so glad that you asked this question, because actually, like, Jevons was getting so much discussion through sort of the back half of 2024 and into 2025. And then I feel like people have sort of stopped talking about it. I remember, actually, when we talked about it on the last shift, because Jesse did such a nice job of explaining it very professorially. And I was like, oh, that’s that’s why you’re a professor. And I’m not. But yeah, people get it at this point. And the short answer is, yes, Jevons is holding. It’s happening. I think if you look at the efficiency gains that are coming in, facility design to some extent, but mostly algorithmic design and chip design. I don’t remember what the stats are, but if you talk to someone from Nvidia about the efficiency of their chips, it’s remarkable how much efficiency they have brought into the design of those chips. That’s true across the board on all of these new custom silicon and all of these other things. And by the way, AI demand continues to surge through the roof. So, you know, Jevons is something which has applied through history since it was identified pretty selectively. Like it definitely doesn’t apply to everything, but by all accounts, it applies to this. I’m not an economist, but, you know, we are seeing efficiency gains in all of the places that people thought we would see efficiency gains and AI consumption only continues to climb. And in fact, if anything, we’re seeing an acceleration of the climb.
Robinson Meyer:
[17:12] Back two years ago, when we really started talking about the data center story, the idea was, well, we don’t know whether these power, this huge amount of load growth is going to show up or not on the data center side, because at the same time we were beginning to talk about massive load growth, Nvidia was also announcing, you know, its new chip, which uses 99% less electricity or 90% less electricity. And it’s now been several years since then. Those chips are deploying. I mean, they were frontier chips, so they’re still very expensive and supply constrained. But it’s not like the load growth story has gone away, even as those chips has diffused.
Robinson Meyer:
[17:48] What is the biggest trend you’re seeing right now that you feel like nobody else or or the market or the rest of the industry hasn’t caught up to yet? And if you want to affirm another trend on the way here and say, ah, people know about this trend, but it’s actually a big deal. You should do that, too. But like, what’s the trend nobody’s talking about yet?
Peter Freed:
[18:06] Yeah, I mean, OK, so I think we’ve already been talking about this a little bit, but but going back to 2025 to now, so call it the last year, I do think gas is like the idea that we’re going to see behind the meter gas of one sort or another that has arrived. So, you know, it was probably I feel for companies that were really chasing that hard in early 2024 because deals weren’t getting done. But now there are a lot of deals that are either done and not talked about yet or in the works right now. So we are seeing a ton of behind the meter gas. All of these are almost all of them are what we would call bridge to grid. So it’s you know, you’re you’re building a gas fired power plant with the notion that at some point you get connected to the power grid. So I think that’s the trend that that has certainly been affirmed. It just took a little bit longer than I think people thought.
Peter Freed:
[18:55] The one maybe that that I think is being talked about in a way that I think
Peter Freed:
[19:00] of it somewhat differently is flexibility. And, you know, there was all of this conversation around like flexibility is going to be the key to all of these different things. And, you know, a bunch of different companies were pursuing it. Google has demonstrated repeatedly that technologically it’s possible to work with the compute loads inside of a data center. And basically you modulate compute as a means of adjusting power consumption. So we know that technologically it’s feasible. We also haven’t seen outside of Google a lot of companies successfully doing that at scale. There’s a bunch of startups that are sort of chasing different flavors of it. I think what we are also finding is that you can achieve, back to the battery conversation, you can achieve a lot of that flexibility synthetically with batteries.
Peter Freed:
[19:49] And where I feel like we are right now, and we can and should talk more about this, is that all of the pieces are on the table. All the puzzle pieces are out on the table now. We don’t need new puzzle pieces, but we do need people to sort of put those puzzle pieces together in a repeatable and scalable way that we haven’t seen yet. And that means, you know, basically.
Peter Freed:
[20:11] Technology solutions, potentially, you know, touching the operations of the data center, although ideally not, at least in my view, like if you’re going to work on one of these, make it look the same as it always looked at the data center operator and you just sort of handle it. Then you need a software solution that finds the actual configuration and makes that work. And there’s some startups that are chasing that. And ultimately you need the utility or the grid operator to sit in the middle of all of that and say, yes, maybe I couldn’t find this myself, but I concur with what you are saying. And therefore I am validating the underlying value proposition to the data center operator or owner that’s going to have to pay for that in order to unlock new capacity. So I think we’re getting closer to the cusp and I’m hoping that 2026 is the year where we see more of that. So we’ll see.
Robinson Meyer:
[21:00] I want to go back to something you said at the very beginning of that answer. That was all so useful. And we’re going to touch on so many parts of that, but I want to go back to something you said at the very beginning, which was that this is that behind the meter gas has showed up, that new data center projects are building large scale gas installations as they wait for a hookup to the grid. We’ve been observing this and covering this at Heatmap, but to describe it as an industry-wide trend does seem like a huge deal because we’ve talked about data centers for the past two years as an electricity story. And when we look at the electricity statistics, what we see is 90 percent, 95 percent of new capacity coming online is renewables. And there’s a reason why renewables would like perform especially well in capacity statistics rather than kind of power statistics. But. This huge amount of capacity coming onto the grid nationally according to eia is renewables but it sounds like kind of what you’re saying is that data centers are now a gas demand story and are hooking deep into the gas system and we should expect to see that demand showing up in the gas system even before we see it in the power system and so the fact that there’s all these data centers using electricity we might not see their electricity demand in grid statistics that’s
Peter Freed:
[22:18] Right because if you’re about if you’re a fully private behind the meter gas project you’re not reporting no one’s reporting that to anybody now you know, Maybe on the gas demand side, yes, although many of these projects are basically being built on top of the big gas plays, right? You know, that’s why we’re seeing so much of this in West Texas and some of this stuff in Ohio. And so where that would show up in reported statistics and when is a big question. One could imagine a universe in which, you know, you’ve built this behind the meter gas-fired solution and your grid interconnection is going to take five years or seven years.
Peter Freed:
[22:59] Maybe at the end of that time when that plugs into the system, then it shows up somewhere. But there’s actually a bunch of really interesting questions about what happens to those gas-generating assets when the grid connection comes along, right? So, you know, there’s like the VoltaGrid model, which is a company that X.ai has worked with. They’ve got a pretty long track record where a lot of times they literally take the generators, they put them on the back of a truck and they take them on to the next project. And so, you know, some of these projects that people are doing with relatively small, like reciprocating engines, which don’t have the world’s greatest heat rates and they’re not that efficient. Maybe they’ll go on to the next project. In other cases, maybe those generators become the backup generation solution to the grid connection and suddenly go from being a prime power source, so running almost all of the time, to running 20 or 30 hours a year when you’re testing, you know, just for testing. And then the middle ground, I think we’ll see some of this too, is those resources become grid resources. So either they’re being dispatched economically as a grid resource, or it goes into sort of the planning paradigm of the local utility. And the costs have already been borne by the data center customer. And so maybe that works from kind of an affordability perspective or otherwise. And I think we’re at the earliest stages of thinking about that. And maybe just because I like to throw out.
Peter Freed:
[24:23] Ideas for people. Like one of the things that I’ve been wondering about having now seen so much of this happening is, could we be a little bit more thoughtful or strategic about how we’re doing this? I’ll put a little shout out into the utility practice at McKinsey. I really like that team. I find them very thoughtful. And so we’ve just been starting to like virtual whiteboard. Like if you wanted to bring an overall grid orientation to like all of these random behind the meter data centers, which mostly just work in where is there land with gas available? That’s sort of the driving factor, right? I need a big piece of land that has gas availability. Could we be more strategic about that from eventually all of this gets hooked into the grid seven years from now, whatever it is like, is there a way to add like one little thumb on the scale of strategy and planning that says I would optimize this location versus that location or I would do it this way versus that way? And it’s really early days, but that’s one of the things that I’ve been sort of pondering a bunch lately is like, is there a better way to do this that feels more strategically oriented than just like finding the biggest piece of land you can with gas available?
Robinson Meyer:
[25:30] Well, it sounds like what you’re also talking about is the massive gas buildout happening totally outside the reach of any utility IRP, totally outside of any utility planning process, massive gas electricity capacity buildout that right now we don’t see on the grid, but at some point in the future could surge online as this bottleneck works its way through. There’s a great question in the chat that I want to ask, which is what percentage
Robinson Meyer:
[25:53] of data centers total expected demand is the bridge being built for? So if it’s a 100-megawatt project, I mean, are developers building 100 megawatts on site of gas and then waiting for 100-megawatt interconnect? Or is it like they build 50 megawatts and then they wait for the rest, the next 50 to show up?
Peter Freed:
[26:12] You know, I can’t speak for every project, but within the projects that I am familiar with, it’s 100% of the projected project load. You know, these projects ramp through time, right? It’s not like if you want a one gigawatt data center, you don’t just turn on a 1-gigawatt data center. In fact, a gigawatt data center could take years to ramp up, not just because of construction time, but because you have to put the computers in, you have to turn them on, you have to make sure they all work together. The old industry rule of thumb used to be 50 megawatts every 90 days or so at a given site. Now it’s accelerated some, but it’s not like you’re just turning on 100 megawatts every couple of weeks, you know, like it takes time. So anyway, what I would say is at least the projects that I’m familiar with, people are... Not deploying as much gas as they possibly can. And then when the grid connection shows up, you get into that conversation that we were just having, what do you do with those generators? Or depending on the size of the site, in some cases, they think of them as additive. So if you’re going to do 500 megawatts of behind the meter generation, and you’ve got a 500-megawatt grid connection, in some jurisdictions, you might say, well, that’s 1,000 megawatts of capacity, You’re actually probably more likely you’re going to work with the utility and it’s 750, right? The grid connection will take more and suddenly now the utility has a peaking asset sitting there that helps them manage it.
Peter Freed:
[27:38] Whereas in Texas, interestingly, which is the hottest, you know, this is where a great significant percentage of all of this is happening. With SB6, you can’t, that’s the Senate Bill 6 that, you know, was related to data centers coming online and how you’re going to think about that from the perspective of the grid. And basically, you can’t get a new grid connection without somewhere adding new generation associated with it. And so interestingly, like, there are some proposals and projects that I’ve seen or heard people talking about, where actually the idea is that you build a bridge project to a large scale gas fire generator and still get a grid connection. And that large scale gas fire generator, usually a combined cycle unit, is helping accelerate the interconnection timeline to the grid, because basically what you’re saying to the grid operator is like, I’ve got enough generating capacity sitting here that I can cover this entire load if I need to, but it would still be more efficient. For everybody to have this thing connected. And, you know, so one of the questions that’s really interesting is, do those folks get interconnected to the grid faster because there’s some big resource sitting there? And I think probably the answer is going to be yes.
Robinson Meyer:
[28:52] We know there’s this giant capex surge coming from the hyperscalers. I mean, it’s reached the point now where tech companies’ stocks suffer when they announce investment because they seem to be in an arms race of spending on data centers. We were just talking about the behind the meter gas boom. There’s a lot of renewable energy developers in this audience or battery developers. How should they be thinking about this moment and what do they need to be doing to make their projects or to work with data center developers in the most attractive way?
Peter Freed:
[29:22] I’ll bring us back a couple of minutes to when I said, look, if you’re a data center developer and you’re building gas plus storage and you’re thinking of that as a pretty complicated thing, someone is really going to have to work out on explaining why the introduction of a variable renewable resource into that configuration is worthwhile.
Peter Freed:
[29:45] And obviously there are people that believe that that’s true. Intersect believed that that was true and it worked out really, really well for them. There are ways to tell that story. And I think that the renewable energy development community probably still has some work to do to help explain that. So that’s sort of thing number one, like the closer you get to the operations of the.
Peter Freed:
[30:08] the data center facility, the more work you’re going to have to do to explain why you believe that the integration of renewables into that makes sense. Now, you can remove yourself somewhat from the actual operations of the facility. And this is where we get into that, what, bring your own capacity conversations. And, you know, there’s been some really interesting stuff sort of talking about, okay, maybe there is a utility which has sufficient wires capacity as — and like, there’s enough room on the transmission lines to plug a data center in and turn the lights on, but they don’t have enough market capacity. Like, they don’t have enough of the financial products required by the RTO that they operate in to serve that facility. And so that can become an interesting opportunity for renewables in particular, storage in particular, trying to figure out how to put together these bring your own capacity products to serve data centers. And I’ll say, you know, when I first heard about these bring your own capacity opportunities, I thought that they were pretty niche. I was like, okay, well, you know, a utility has sufficient wires capacity to serve a giant data center, but they don’t have capacity in the market. Like that, that feels like something that’s not going to happen that often, but apparently I mean, I was incorrect.
Robinson Meyer:
[31:30] How would this happen?
Peter Freed:
[31:31] Yeah. I mean, basically like the generation that they’ve built or contracted with is, is sufficient to serve the needs of their existing customer base. And also there is room on the physical infrastructure that underpins that for growth. You know, a lot of, you know, utilities are rarely building exactly to what they need in that moment, right? There’s always some anticipation of growth. And so, you know, particularly with a utility which maybe doesn’t have a huge generation fleet, which is procuring capacity from the market, you could find yourself in a situation where their physical infrastructure could accommodate new load, but they don’t actually have the market product to serve it. And so, you know, either the answer to that is the utility itself goes out and builds a whole bunch of new, probably gas-fired generation to get sufficient capacity to serve, or the customer can bring its own capacity. And if you look at, for example, like some of the nuclear deals that have happened with existing plants or with upgrades or some of those things, like assuredly, the capacity elements of nuclear service are useful in putting together a solution to serve new data center loads across those markets.
Robinson Meyer:
[32:44] If we’re talking about data centers now being major gas fire generator, you know, if you’re talking about putting 500 megawatts or gigawatt of gas generation on a data center site, then suddenly these things shift from having the footprint of a classic cloud data center or even an Amazon warehouse to being a major, I mean, having literally a gigawatt scale power plant in your backyard. So we don’t have these polling results out yet, but at Heatmap Pro, we do periodic polling on the community support for data centers. And we ask questions like, would you welcome a data center in your community? The most recent poll, which is not out yet, we’re going to publish it later this week, is net 24, is negative 24. And more than half of respondents would oppose a data center in their community. It’s a huge decrease in our data set. But do you see backlash to the pollution aspect of data centers or is it to the whole package?
Peter Freed:
[33:43] I think it’s the whole package. And my guess would be that the pollution element makes it worse. So, you know, public sentiment polling for AI as a technology category is extremely poor. Like, I think there was a really interesting article in The New York Times this weekend about how, like, of all of the technology booms where there’s polling data, this one is by far the worst. Like people historically have gotten very enthusiastic about this kind of stuff. And like the public is just very skeptical about the benefit.
Robinson Meyer:
[34:11] In part because the CEOs of the companies tell us it’s going to be bad. Mark Zuckerberg told us that he was going to connect the world, you know, 20 years ago. People had concerns, but it seemed like people were. I mean, they turned out, of course, that technology had massive downsides. But it seems like everyone was pretty jazzed about it. But now even the CEOs are like, this is a major problem. We aren’t able to stop it. And that’s why we’ve adopted these arcane corporate structures in order to contain the technology that we are. Right.
Peter Freed:
[34:40] So with that said, you know, I think. A data center is the closest physical manifestation of AI that a person can encounter right now outside of like robots in your house. And so I think I think in general, public opposition to data centers tends to reflect that public sentiment. And on top of that, pollution elements. Yeah, I don’t think that that that is going to be particularly helpful from a public sentiment standpoint. Now, the question is, does the introduction of renewables to that power solution really obviate any of the concern? And my guess is that the answer is no, because, you know, if it’s a gigawatt data center with a gigawatt of gas-fired power generation or 500 megawatts because there’s a big solar and storage, you know, I don’t think the public generally cares. And so, you know, what I expect that we will be seeing is more and more of these gas-fired projects being in places with favorable permitting regulations. So, you know, think...
Robinson Meyer:
[35:43] Unincorporated county land in this which yes
Peter Freed:
[35:46] We’re already seeing a lot of that places where there’s already industrial zoning and like people you know places and communities that are used to this kind of stuff i think that’s much more likely.
Robinson Meyer:
[35:55] I was going to ask if we’re seeing new hot spots because the hot spot previously was the mid-atlantic you know georgia northern virginia and texas and you already mentioned texas but are there different hot spots emerging slash is construction of PJM, the Mid-Atlantic and the Upper Midwest fading?
Peter Freed:
[36:13] No, nothing is fading. I think West Texas is really like the main story. You know, Meta announced a data center in El Paso and, you know, people were sort of like, oh, El Paso, so interesting. But now the whole Abilene, Midland, Waco, like I would not be at all surprised if within a year or two, if not a true availability zone for the cloud, But that will become a new major computing region. Like enough activity is happening there that we’re just going to see that spring up. You know, I’m hearing a lot of rumblings around Wyoming. Again, not the world’s most populous state. A lot of gas infrastructure runs through it. There have been, you know, Microsoft has got a project there. Meta’s got a project there. I think OpenAI has been rumored to have a project there. So, or Crusoe, you know, maybe with OpenAI attached. Anyway, so, you know, there’s I think Wyoming is going to be a place and then, you know, really anywhere else where you have less dense population zones and the availability of gas may eventually become something.
Robinson Meyer:
[37:15] When we talk about these hyperscalers, so we’re talking about oh, Meta has a project there, Oracle has a project there, OpenAI has a project there. Is that because it’s the hyperscalers that are still driving most of the demand? Or do they act as water buffalo? And then, you know, the flock of independent data center developers kind of follow them because they’re assumed to have detected some price signal. And so you get these big projects that are like the big Meta gigawatt-scale project. And then you get all these little AI data centers around it, which are like independent developers hopping on to what they think might be good price action or something.
Peter Freed:
[37:53] So it’s somewhere in the middle. So all of the major hyperscalers are still self-performing. So they’re building their own data centers. And then there are large developers who are building data centers to also serve those companies. So basically the demand signal for large-scale compute is the four hyperscalers that we’ve already talked about, plus OpenAI and Anthropic, basically. Like that’s pretty much it. Oracle, to the extent that they’re a player here, is largely just a sleeve for OpenAI demand. You know, some exceptions, like they have their own businesses, but a lot of what they’re doing is that. And so almost everything we’re seeing is just consolidating around six companies, which is a very weird market when you think about it.
Robinson Meyer:
[38:41] Is there is someone who’s doing Anthropic’s?
Peter Freed:
[38:44] Actually, the information just had some really interesting reporting on this over the weekend. So right now, like the expectation has been that Anthropic is doing leasing. Either they’re doing direct leasing. So, you know, where they would have more control over how a data center was constructed and built, or they’re buying cloud capacity from the likes of Google Cloud or Amazon Web Services or what have you. So they are not building any of their own data centers, at least nothing that’s been publicly reported. And actually, interestingly, that article that I mentioned that came out over the weekend identified that OpenAI had had these ambitions to build their own data centers and has largely shelved those because for a variety of reasons, including securing financing, they’re mostly also going this leasing route, at least right now.
Robinson Meyer:
[39:28] Well, that bleeds me. So how are hyperscalers navigating, you know, a number of hyperscalers or a number of these big AI companies, notably OpenAI and Anthropic, are at least reportedly trying to IPO this year? That means they need a lot of money. They’re also facing ballooning computing needs. How is that affecting the real world of data center development?
Peter Freed:
[39:50] So we’ve just talked about, in a sense, how the demand signal, I wouldn’t even say consolidated. It was never like, if anything, it’s grown a little bit because now you have two frontier labs along with the four hyperscalers. But you’ve got a very small pool of very large demand. And also what’s happening is I think we are starting to see some consolidation on the capital availability side because the numbers are so large. right? You know, we’re talking hundreds of billions to a trillion plus dollars in the next handful of years. There are very few capital providers that have that kind of money. So the hyperscalers have been using their own balance sheets. They’re flush. But even that is not enough. They’re starting to go out and look for creative sources of financing. There’s large private equity funds that have been playing in this space for a long time. They continue to play. And then ultimately, and we’re seeing this, right, like sovereign wealth funds, that’s kind of the type of money that we’re talking about. And a lot of Middle Eastern money has been getting involved here, other sovereign wealth from around the world, Southeast Asia, etc. So in my view, we are actually seeing both a consolidation of capital supply and demand on the other side and the credit that goes along with that. And so it’s going to be really, really interesting to sort of see how this all shakes out, in part because...
Peter Freed:
[41:16] A lot of the capital providers are relatively risk averse. Like, you know, last year there was a bit of a fear of missing out flavor. And so we saw people at least making big announcements about money that they were going to earmark for stuff. I would say this year, it seems to me that people are really asking real questions around risk. Like, if you are OpenAI or Anthropic who are not public companies that don’t have any credit rating, like, you know, everyone is going to evaluate the risk associated with doing business with those companies. But ultimately, like they don’t have a publicly available credit rating associated with them. How do they go get?
Robinson Meyer:
[41:52] They don’t even have GAAP out. I mean, they don’t have accounting statistics out yet, right? Yeah. Only no open AI and Anthropic’s cash flow from periodic stories, correct?
Peter Freed:
[42:00] And so if you’re a company like that, how do you go get $100 billion to build data center capacity? And the answer is you have to start getting really creative. And that’s where we see, you know, Nvidia throwing some of its credit around to help people go buy their chips. And AMD is doing this and Google is rumored to be doing it. So, you know, there’s it’s this is all financial engineering stuff, which is probably beyond the scope of this podcast. But suffice it to say that people are really working out, trying to figure out how to get the amount of capital that they need to do what they’re proposing to do.
Robinson Meyer:
[42:35] Let’s just lean into this for one more question, because I do want I want to
Robinson Meyer:
[42:40] talk about a few different technologies, and this is an energy and news podcast. But why this matters in a broader sense is when we talk about the scale of organizations that are getting involved, the Saudis, you know, sovereign wealth fund, the Norwegian sovereign wealth fund, insurance companies, I mean, these are the big credit providers, reinsurers, right? These are the big, big wells of money at the basis of the global economy. And when we talk about them getting roped into AI at such a broad scale. We are talking in part, it seems to me, about a situation where
Robinson Meyer:
[43:14] Even though there are various parts of froth in the AI economy, but without anyone feeling like this is a dot-com style story where enormous amounts of wealth are going to be created in a kind of magic way, so you must put all your money in it. Without anyone feeling completely like that or any of the big capital providers feeling completely like that, enough money gets soaked into these projects and has to get involved in these projects just to build them that a huge portion of the economy and business investment and therefore final demand gets tied up in data center projects. And so can you walk us through, from your point of view, like what are the next temporal milestones in the data center story going forward? You know, it sounds like you don’t think power is really ever going to be a bottleneck, partially because of gas. But like what what points this year or in future years will we start to understand, you know, these projects are working or they’re not working, they’re coming together, they’re generating the cash flow they need or they’re not.
Peter Freed:
[44:27] I’m going to answer this question in two ways. One is as like a old energy guy, like how part of the conversation we were talking. And then I’m going to indulge for like half a second in, I live in Menlo Park, California. So like right in the heart of Silicon Valley. Perfect. And I am just like, it’s the only thing that anyone really talks about around here.
Peter Freed:
[44:49] And I do want to give you that answer to that question. So let’s do both. So it’s sort of the traditional energy project finance. Like, how do we think about all of this? Like, you know, I think that there are sufficient pools of capital available to move these projects forward. The question that everyone is going to be looking for is like, is there sufficient revenue quality to justify the amount of money that’s getting put in? And that revenue quality is going to be driven by long-term leases on the data centers and long-term power purchase agreements on the power plants. And I think very likely that people are going to figure enough of that out to move the ball forward at least through 2030. There’s going to have to be some interesting, like particularly The data center industry has historically had at least some speculative development, not the two guys and a dog and a pickup truck that we were talking about earlier, but people doing some amount of work to advance projects to a point that they could begin marketing those projects. And the money associated with that early stage development...
Peter Freed:
[45:49] Was big, but not huge. Now it’s huge. And so I think that’s one of the things that people are still going to have to figure out is like, if you’re a sovereign wealth fund that is used to just really stable returns, because that’s what you do for the citizens of the country whose money you’re managing, you don’t give people billions of dollars to go do speculative development work. So how we address that, I think, is one of the things that people are still trying to figure out. Does it get super disaggregated and you sort of see different pools of capital doing that or, you know, do we just figure out different ways to address that? But that to me is sort of the big question is how do you advance projects to a stage where they can move relatively quickly, but prior to them having some credit-worthy long-term thing locked in? That to me is the thing that the industry is struggling with. And I think people are probably going to figure that out. Now, the like AI-pilled, as they would call it around here, like the AI-pilled answer to that question if you look at claude and claude code so this is like the latest models from anthropic.
Peter Freed:
[46:52] 80% of the code, 80% of the code that is Claude is written by Claude. The expectation by the end of this year is that 90-plus, 95% of the code in Claude will be written by Claude. Which means that we are approaching a point of something that industry hands call recursive self-improvement. Which is basically to say that like the AIs are able to improve themselves and therefore the trajectory and acceleration of improvement will begin to look asymptotic, at least exponential and potentially asymptotic. And the reason that I am particularly interested in this is because all of the work that has happened on AI so far is still not on the giant pieces of infrastructure that we’re constructing right now. So there is really not a gigawatt scale data center. I think the biggest GPU training cluster is probably X.ai is with Colossus and it’s three or 400,000 GPUs. Like no one’s got a million-GPU training cluster. And so if you expect even a linear scaling of performance of the models with the infrastructure that are currently under construction, like one can begin to wrap one’s head around the idea that we will hit this point of like takeoff.
Peter Freed:
[48:11] In which case, all of these things begin becoming better very, very quickly. And maybe data center capacity is actually the limiting factor in terms of how quickly that takes off. That’s what I mean, that’s what people believe, which is why they’re trying to build more of this. So in that universe...
Peter Freed:
[48:28] Who knows what anything looks like, but the ability to run compute at scale will be one of the most significant throttles on that takeoff trajectory. And so I think, you know, I think there is a world in which things look so radically different by 2030 that we can’t even really speculate on what that is. And so that’s the other answer to that question, which is just like, maybe AI figures all this out and we don’t have to worry about it. Like, I don’t know.
Robinson Meyer:
[48:59] And I should say that the day we’re recording this, the market’s down. I mean, the S&P is down 1% because investors have become convinced that were that to happen, it would negatively impact a final demand in the economy rather than positively impact. That being said, let’s assume that it’s 2030 and electricity markets still work roughly like they work. And we might have amazing agents to do all of this stuff, but the actual physical technologies are still binding. And let’s talk about a few technologies that could matter here. The first is batteries. You mentioned that batteries are now being paired with gas to drive reliability. Should we think about flexible data center demand? As some, including current Google employee, Tyler Norris, have talked about as a function of algorithmic, you know, and, you know, actually AI activity flexing up and down. Or at this point, should we think about data center flexibility in the real world as just something created by the batteries you have either on site or behind the substation?
Peter Freed:
[50:01] Yeah, again, there’s nothing technologically infeasible about modulating compute. I just think that generally the economics don’t make sense. I mean, Google has figured out a way to make this make sense, which which actually might mean that other people can figure it out. I think, though, if we can do it synthetically with storage, let’s just say at the fence line and, you know, we can debate what that actually means. That’s probably going to be the fastest, easiest way for most people, like most projects to happen. And again, what does that actually mean? Right. So if we’re if we’re bringing flexibility to the conversation, that means that you are on a system which has sufficient energy in most hours to serve the facility. You’re really addressing a resource adequacy constraint on an August afternoon or a December morning or what have you. And so one of the things that’s actually pretty interesting for the renewables folks is if that flywheel gets going, right, like suddenly we’re effectively addressing these resource adequacy things. We’re using surplus energy, the extra energy that’s available in most hours. If the flywheel gets going, then you need to keep adding more energy to the system. And that’s actually something that is very good for renewables because that will be the fastest, least expensive way to add energy to the system to inject enough surplus energy to keep the like flexibility flywheel moving. Like if you do enough flexibility, flexibility stops working unless you add more energy to the system.
Robinson Meyer:
[51:22] The other way you could do flexibility is through these things that are sometimes called virtual power plants. It’s the idea that you could flex consumer demand household demand up and down you people would get paid to have their air conditioner be a little higher a little lower to run certain processes and energy intensive processes in their house when the grid can withstand it there’s been discussion that ai companies the hyperskillers would be able to pay the amount of money that could actually make these projects feasible is that something you believe will happen and what needs to happen to have a gigawatt of VPPs show up, you know, in, say, the mid-Atlantic or in Texas?
Peter Freed:
[52:02] Yeah, I think it’s possible. Look, I think it’s part of a portfolio. So this bring your own capacity concept, we were talking about it with solar and storage, but there’s no reason why other products couldn’t serve that capacity. And so I think what’s likely to happen is that we will see, you know, consumer demand response or what have you is like a small portion of a portfolio until people get more comfortable with the idea. And then maybe it’ll grow. And ultimately, hopefully the market just says, whatever the most cost-efficient way to provide this particular capacity product at this time, like we’ll do that. So I think, you know, hopefully we’ll see.
Robinson Meyer:
[52:41] Which market?
Peter Freed:
[52:42] Well, I mean, that’s the other question.
Robinson Meyer:
[52:44] Right? In ERCOT, I bet the right market would show up. And PJM, I don’t know.
Peter Freed:
[52:47] It’s hard to say.
Robinson Meyer:
[52:48] I’m just in rapid fire now. One of the things I’m hearing from you is that behind the meter gas, is going to bridge to 2030. Data centers are going to have the electricity they need via behind the meter gas and batteries, and it’s going to get to 2030, at which point more gas turbine, you know, power plant scale, gas turbine capacity comes online. Do you see a role for other technologies at that point? I mean, that’s when I would think about like geothermal become salient to this kind of this kind of story.
Peter Freed:
[53:19] Yeah, I think the broad perception is that post 2030, the aperture of opportunity is wider. And I would agree with that with the following caveat. It won’t be any wider unless people are working on it now. And so that is one of the, like, we’ve been talking about this since 2024. And I wouldn’t necessarily say I’ve seen like the intensive level of investment and preparing for a post 2030 world in the last two years. So, you know, we’ve seen some cool stuff on geothermal. We’ve seen some interesting nuclear stuff. But I wouldn’t necessarily say that like a groundswell of investment and innovation that’s required to have a radically different slate of generating technologies in a post 2030, like I’m not seeing that. And so if the world is going to look really different or we’re going to see a lot more opportunity post 2030, I think people need to be doing more.
Robinson Meyer:
[54:12] What’s the one thing?
Robinson Meyer:
[54:15] Public policy or investment decision that would need to happen to slow down or change the makeup of the huge amount of behind the meter gas that it seems like we’re about to get across the country like if you were a democratic policy maker or an energy policy maker someone who wants
Peter Freed:
[54:35] I mean I’m a I’m a carrots not sticks guy, Rob, so I think rather than slowing that down, I would think about speeding the other stuff up and no great surprise. It’s figuring out how to get more transmission built, just permitting reform. It’s stuff that we’ve been aware of for a long time. And we are now seeing significant levels of government subsidy kind of going into all kinds of different things and in ways that maybe we haven’t seen before. So you could imagine a lot of what we’ve seen in the nuclear industry is probably going to help it have a moment that it wouldn’t have had a couple of years ago.
Robinson Meyer:
[55:10] That’s kind of what I’m asking. Is it that there needs to be investment in these public investment or either permitting reform for transmission or some kind of fixed transmission build out or nuclear build out in order to just like, if you care about emissions, this is what maybe needs to happen. There’s a whole set of questions here that I want to hit that we’re right at time for. But if you had advice to data center developers at this moment, renewable developers for a long time have been in a world where Our communities can give them all sorts of feedback. For data center developers, what separates a good developer from a bad developer? What should good data center developers be doing and what do they look like?
Peter Freed:
[55:47] Yeah, I mean, look, at the end of the day, it’s getting engaged with communities, and it has to be that. And that’s always been the case. And people are just slowly getting better at it.
Robinson Meyer:
[55:56] Great. Well, we’re going to have to leave it there. Peter Freed, thank you so much for joining us. It was great, as always.
Peter Freed:
[56:00] Thank you for having me. A real pleasure.
Robinson Meyer:
[56:02] Thank you so much for listening. If you enjoyed this conversation or conversations like it, please subscribe to Shift Key. In any podcast app you use, you can also find us at heatmap.news. And remember, if you want to keep following how data centers are reshaping the energy transition, because they’re doing quite a bit, then that is the work we do every day at Heatmap in our newsroom and also in our pro team. Our pro team tracks where energy and data center projects move forward, where they get delayed or canceled, and how community response shapes outcomes. You can find the link to that in the chat. If you’re interested in learning more about the software, visit heatmap.news/pro. That’s heatmap.news/pro. If you enjoyed Shift Key, then leave us a review or send this episode to your friends. You can follow me on X or Bluesky or LinkedIn, all of the above, under my name, Robinson Meyer. Shift Key is a production of Heatmap News. Our editors are Jillian Goodman and Nico Lauricella. Multimedia editing and audio engineering is by Jacob Lambert and by Nick Woodbury. Our music is by Adam Kromelow. Thanks so much for listening and see you next week.