Plain English with Derek Thompson - The Energy Story of the Moment: The Unstoppable Rise of Solar Vs. the Unmovable Demand for Global Fossil Fuels

Episode Date: February 7, 2025

Fans of green energy like me face some inconvenient truths about the global energy picture. First, coal sounds like a dirty technology that the rich world is moving on from. But nearly 9 billion tons of coal were burned last year—an all-time high. Second, "peak oil" is a prediction that many analysts have thrown around in the past few years, but oil production is also near its all-time high. Third, we might not even be at peak wood: Global wood fuel production was higher in 2024 than in 1980. At the same time, I think the renewable energy revolution is proving to be its own unstoppable phenomenon. Solar and battery installations are still exploding upward, and whereas some skeptics worried that the earth wouldn’t be able to provide the essential elements and metals to build out a green energy system, those doubts seem, for the moment, overwrought. Lithium, which is one of the most important metals for battery production, has seen its resources double since 2018. So what we have is not a pretty picture but a messy one. A green energy boom matched with an enormous demand for fossil fuels, as billions of people around the world drive and eat and demand the middle-class lifestyle that is their right. Today’s guest is Nat Bullard, an independent energy analyst and the author of a new extraordinary report on the state of energy and decarbonization. We talk about everything: coal, oil, wood, and natural gas; the history of nuclear vs. solar in America; the solar and battery revolution of the 21st century; the political barriers to its growth; the rise of BYD in China; the flatlining of Tesla's growth; and the future of energy technology. If you have questions, observations, or ideas for future episodes, email us at [e-mail]. Host: Derek Thompson Guest: Nat Bullard Producer: Devon Baroldi Nat's report: https://www.nathanielbullard.com/presentations Learn more about your ad choices. Visit podcastchoices.com/adchoices

What's up, everybody. Chris Vernon here and welcome to a new season of the NBA and the mismatch. And huge welcome as well to my new co-host, Dave Jacoby. I can't wait to link with you twice a week every Tuesday and Friday right here on the mismatch to break down everything that's happening in the league. Who's playing well, who we loved, who we loathed, trade rumors, team dysfunction. We've got you covered right here. So follow us, subscribe, and hit us with those five-star ratings on Spotify or wherever you get you. your podcast. And also don't forget to follow us on social media. That's at Ringer NBA and check out the full mismatch episodes with the two handsomest podcasters in the history of podcasting

right on the Ringer NBA YouTube channel. Today, the state of energy in the U.S. and around the world. Energy is good. In fact, it is genuinely difficult to overstate just how good it is. Human progress is in many respects a story of our ability to transform our natural environment into energy that releases us from the scourge of natural suffering. Our food is energy, our clothes and our homes, energy, our air conditioning, and our hot water, our light, and our electricity. You are hearing my voice right now, because jewels of energy pulled from natural gas or oil or the sun itself were transformed into an electric current that flowed into a device that you're currently holding or driving or resting on the counter.

Energy is good. And its absence is bad. According to the historian Vaklav Smil, nearly a billion people today still have no access to electricity. And the average annual per capita energy use, among more than three billion people today, is at levels comparable to those of rich, European countries and North America in the 1850s.

In the next few decades, these billions of people will increase their energy usage as they economically progress, because progress itself is an energy-intensive state. But most of the world's energy relies on burning stuff we find in the ground. Wood, coal, oil, natural gas. The burning of these hydrocarbons automatically releases carbon dioxide into the atmosphere, where it traps heat, warms the planet. And thus, our greatest triumph, which is energy progress, comes with a receipt, is inextricably connected to a great challenge, which is an energy transition toward green or clean or zero-carbon technology. And make no mistake, the challenge

is very real. 2024 was statistically the warmest year on record since at least 1850. The average length of American heat wave seasons has now tripled since the 1960s. In India last year, companies had 320 mentions of heat waves in their earnings called transcripts more than in the previous four decades combined. In one day this past October, Valencia received more rain than it had in the prior three years. Here are some inconvenient truths about climate change and decarbonization. Number one, coal sounds like a dirty tech that the rich world is moving on from, but nearly nine billion tons of coal were burned last year in all-time high. Number two, peak oil is a prediction that many analysts have thrown around the last few years,

but oil production is also near its all-time high. Number three, we might not even be at peak wood. Yes, if you're listing in the U.S., or Europe, you might think, wait, wasn't burning trees for heat, something that we did in like the mid-1700s? Yes, yes, it was. But global wood fuel production

was actually higher in 2024 than it was in 1980. At the same time, I think there's optimism here, room for tremendous optimism. I think renewable energy is proving to be its own unstoppable phenomenon. Solar and battery installations are still exploding upward,

and whereas some skeptics are worried or have been worried, the world wouldn't be able to provide the essential elements and metals to build out a green energy system, those doubts seem, for the moment, a bit overwrought. Lithium, for example, which is one of the most important metals for battery production, has seen its global resources double since 2018. So what we have here is not a pretty picture, but a messy one. A green energy boom matched with an enormous, nearly unsatiable demand. for fossil fuels around the world, as billions of people drive and eat and demand the middle-class

lifestyle that is their right. Today's guest is Nat Belard, an independent energy analyst. Something I really want to do with this show, and just generally as a journalist who's interested in a chaotic diversity of things, is to find individuals that I can trust to give me a masterclass on a big topic. This is my second sort of masterclass episode. of the year. Just as Michael Sembalist from J.P. Morgan did a big, wonderful state of the 2025 economy episode for us a few weeks ago. Nat is cut from the same cloth, just an extremely wide-ranging, articulate, and absolutely no bullshit expert who really bestrides the energy field in a way that few others do. I'm Derek Thompson. This is plain English. Nat Ballard, welcome

at the show. Derek, thanks for having me back. It's great to be here. Your decarbonization deck is cause for celebration in many corners of the world, and I'm very honored that you agreed to bring the party here. You're in Singapore, and I'm in North Carolina. So if folks feel like I sound a little sleepy or Nat sounds a little groggy, just choose instead to be impressed by the technological fact of podcasting in real time across 10,000 miles. Nat, to kick us off, who are you? And what should we know about this deck? So I'm an analyst by training. I spent the last 18 years now, really only looking at the world of energy and how it changes state by technology, by market, and by what we choose to prioritize in terms of what we build to energize our whole world. I did this for a startup. I did this at Bloomberg for more than a decade. I not do this on my own. I do this also through a startup that I've co-founded in the U.S.

And my main priority is to try to bring the clearest possible view of the present day. I think we have so much noise and so much sort of energy, no pun intended, floating around the world in terms of thinking about the future, that the best possible thing we can do to help us is to just get a very clear view of the present day. So starting a couple years ago, freed from... the surly bonds of a giant corporation, I was able to go ahead and do what I'd always wanted to do, which is this mega presentation that told a narrative, if you will, driven entirely by facts, but a narrative that talked about where we are and used that to think about where we might be going. And it's, I think, a useful exercise that I do every year for myself largely in terms of helping me understand, but I decided to make it public and publish it.

And, you know, I wouldn't say that there was necessarily a void for this kind of thing in the market. but there is a world that tends to be bucketed into these multi-trillion-dollar sectors on their own. Transportation, the transportation of fuel, the consumption of fuel, the power sector, oil and gas, petrochemicals. And I was like, you know what, no one really gets a chance to wrap this all into one ball. Now, it's by no means comprehensive, by its nature, it can't be. But I try to touch upon everything that I think of as the prime movers or the main drivers of change in the moment. and just get a clear picture of where we are. And hopefully that helps inform people on where we might be going.

We just had Michael Sembalist from J.P. Morgan on the show to give a state of the economy as his eye on the market newsletter just is this dozens of pages that give you a sense of everything that matters that's happening in the U.S. and global economies. I think of this report very similarly. This is a state of global energy report. 200 pages, 200 charts. We'll link to it in the show notes, of course. But here in the show, I want to focus on four major themes of this year's report, which I'm just going to name now to set the table for you and for listeners. First, I want us to be very clear about the challenge.

It's not just that the world's getting hotter. It's also that fossil fuel production is near or at record highs. So the fossil mountain has never been higher by some measures. That's the challenge. Number two, I want to talk about progress. In particular, I want to focus on the progress made in solar energy and battery technology. Number three, I want to talk about bottlenecks to progress. Progress isn't easy.

There are a lot of regulatory and political and market side challenges to, barriers to, solar and battery and other clean tech moving forward. And then finally, I want to talk about China, Tesla, AI, and the future. So let's start with point one. Let's start with a challenge. And to be honest, I think you lay it out beautifully. And let me just give two facts that I gleaned from your report.

In 2000, Europe consumed twice as much coal power as India. Today, India consumes more coal than Europe and North America combined. So why is coal production near or at? an all-time high, it's because of facts like that. Number two, in 2000, Europe and the U.S., both imported 10 times as many barrels of oil as China. Both imported 10 times as many barrels of oil as China. Today, China imports more oil than both of us. I think that offers a taste of why fossil fuel production remains so high, but help us understand from a global standpoint just how robust and sturdy fossil fuel demand still is. We're in a tricky spot right now, which is that it's

possible to sort of see where these things might peak every time you look forward five years. And in fact, it's been that way for at least the last decade in my mind. And we have yet to meet it because energy demand is really sticky. It's really persistent. There are parts of the world that are still energizing, so to speak, like they need to just get access to energy to begin with. There's other places that are sort of under-energized compared to the way we live in Europe, Europe and the U.S. and all of the kind of what we would call the rich countries of the world in the OECD. And then there's also just a lot of industrial activity in particular in a place like China. And in a place like India, there's industry as well as a demand cycle that we went through a long time ago in North America, which is air conditioning,

making sure that we're better prepared for the heat of a warming world. And it's been very hard to displace these trends with anything new yet. There are some signs that, at least in China, this might be coming towards a couple of potential peaks. I'm very low to call like a full on peak in anything right now until we see more years of data. But these two countries on their own are what, 2.6 billion people, 2.7 billion people, with together much less access to energy per capita basis than anything we've got in the West. And also, in particular in China's case, the bulk of manufacturing of almost any good you can think of, whether it's a primary good, whether it's a finished good.

And so that just means that all of that demand has flown to these places. And then there are other places that have to catch up. So, you know, it's important to note, and this is not original from me, but there's no such thing as like a wealthy low-energy country. There's no such thing around the world as a country that has high degree of per capita income and a very low degree of energy consumption. It just does not exist. And so, you know, in the absence of anything better,

we're going to be doing this with fossil fuels. And the sort of urge, the societal and human urge to be, you know, in a high energy and wealthy position is, I think, very, very hard to not only unwind, but hard to even not wish on people, in a sense. It's not just China and India that are producing a ton of fossil-based energy. It's interesting because I'll sometimes watch economic commentators who see the incoming Trump administration and say, finally we have someone who's going to unleash energy abundance in America.

And I sometimes wonder whether these commentators or economic reporters realize just how much oil and gas, the U.S. is producing and exporting. What's a good way to wrap our heads around the state of oil and natural gas production in the U.S. alone? To be very direct about it, most of those comments are giving in a kind of fact-free environment. I mean, the U.S. is the largest oil producer today and ever. We are the largest producer of natural gas. We are the largest exporter of natural gas in liquefied fashion, a trade which did not exist. a decade ago, effectively in the United States. Put it as an aside, when I started doing this,

I did a project for a great big power utility and a utility and energy holding company, and they owned two U.S. liquefied natural gas import terminals at the time. This was 2010. Okay? Their expectation was that the U.S. was going to be importing natural gas in liquefied form from places like Qatar or Australia.

And instead, in that time, blown past that and have buried both of those countries in terms of our total exports. This is very basic data. Like, it's not political. It's not fancy. It's not fussy. We produce more of those two sources of primary energy than anybody in the world. And we now export it.

You know, when I was, when I was a boy, the U.S. was importing as much energy as we now export. Like the U.S. materially exports on a primary energy basis, rolling together all the energy value in coal, oil, and gas. we export more now than we imported when I was when I was little boy. I don't understand how this is not seen as some kind of abundance. To add another wrinkle to it is it's not like we are an export-based energy economy either, which is another unusual thing about that. Most countries that are exporters, that's what they do. Like they are businesses, societies, countries that are built around that as a prime mover of their economy.

It is not a prime mover of our economy. important element of it. But yeah, we live in an energy world that I think we should acknowledge is really abundant to put it in terms that I know you like. And unequivocally so. Like there's no parsing you need to do with this. There's no caveat or qualification. We have an awful lot of this. Does it need more unleashing? This is a really a fascinating question. The way that I always think about it is that if you're a business executive, you of course want the call option to be able to do more. That is reflected to some degree at a policy level. But we're also not only this giant energy economy, but we are market-based. You cannot, unlike in some other countries,

call up the chief executive of a big producer and say, you are going to do more. Maybe you can do that under the guise of national defense, but you cannot force publicly listed companies to do something like this. It's the flip side of these companies burning $40 to $50 billion a year when they were building the capacity we now have, is that you also can't make them produce when the time comes. I think that's important to note. We have a great deal of everything. We now export.

We have this nimble private sector that drives almost all of this, and it will respond to a market. It will respond to price signals. It will respond to supply and demand. And it's unusual in that sense, if not unique, globally. I think everything you said,

gives us a really good sense of what the energy transition is up against. All the energy technologies that we've named so far. We didn't mention wood, but by some measures, according to your report, global wood fuel production was higher in

2024 than in 1980, just to give people a sense of how much carbon-intensive energy is being produced, how much we're burning. Wood and oil, natural gas, Coal, certainly, these are carbon-intensive technologies.

And when we use these technologies to cool people's homes, heat people's homes, make people food, obviously that's a mitzvah. But we all know what the cost is. We know what the receipt is. It's carbon into the atmosphere that's creating a hotter planet. And that's part of the motivation for the energy transition. But I want to give people a sense of just how tall that fossil mountain is, because it is daunting. So let's get to the energy transition.

Let's start with solar. By percent growth, how fast is solar growing right now compared to other energy generation sources in modern history? And to what do you credit its growth at the moment? So solar is a unique beast in the energy system. Everything that we've developed over the last couple of centuries has generally thrived by becoming bigger. Efficiency comes by greater individual scale of a small number of very large things. There are worldwide counts only in the thousands to then tens of thousands

for typical big energy assets. And they become more efficient by getting bigger. Solar and wind in an allied fashion are different. You build them by the tens of thousands every year in the case of wind turbines, and comfortably these days by the billion in terms of solar panels. And solar right now is growing faster after the kind of liftoff moment, which takes a long time for all technologies, but after the kind of liftoff moment that is arbitrary in this case, but it's about 100 terawatt hours a year, which is like what a medium-sized US state would consume. We're in a position where solar grows faster than anything else has ever grown in absolute

terms. In relative terms, okay, nuclear power was growing about the same rate in the 70s when the energy economy was much smaller. But it's just the fastest thing that we've ever added in absolute terms to the power grid once it hits a takeoff point. And that's a bit of a brainbreaker for a lot of people because it doesn't match this, it doesn't match this expectation of, well, it's thousands of giant coal plants with, you know, with exhaust stacks and cooling towers, or it's, you know, dozens or hundreds of nuclear power plants being built. It's assets when you count them up financially in the like tens or hundreds of thousands every year,

and it's individual deployments of these technologies by the billion. But it's that deployment and that total amount of stuff that happens every year that's being built that gives us the kind of improvement in cost and efficiency and performance that you tend to expect with some kind of mass manufactured good as opposed to something that's made out of tons of steel. I'll give you a quick aside here as a comparison. So a great big natural gas plant that you

would buy these days. The turbine sets in those way more than an Airbus A380 empty. Okay. So that's one technology that you, if you think about the, that's the achievement of greater efficiency through greater scale. Somer on the other hand creates greater efficiency and lower cost through greater volume, which in turn gives you a reason to, gives you costs, drives you costs, down and efficiencies up, which gives you reason to do more of it. It has a way of sort of inducing its own good fortune, to put it that way. I think it's really fascinating as a piece of energy history that if you go back to the mid-1950s, the solar cell, the efficient photovoltaic cell is invented at Bell Labs just around at the same time that the first nuclear power plant

is built in America. And I sometimes think about being, you know, Nat Ballard in that moment, trying to be like an energy analyst, thinking which of these technologies is going to take over the future. Is it the photovoltaic cell, the solar cell, or is it a nuclear power plant? And if you had bet on nuclear power plant, you would have been right for decades. Nuclear power is built like crazy in the U.S. the 1960s, early 1970s. It's built around the world. It still, today, accounts for what, 20% of U.S. electricity generation. Nats nodding his head. It's about 20%. But famously for the nuclear heads out there, we essentially stopped building new nuclear power plants in the 1980s, 1990s, 2000s. I think only one plant, the Voltal plant, has been finished in the 21st century that

was started in the 20th century. I mean, it's just extraordinary, the degree to which we've essentially stopped building nuclear in this country. Meanwhile, like the tortoise winning the race against the hair, Ironically, nuclear is the hair here. The tortoise winning the race is solar, which is getting off the ground and fits and starts in the 1960s, 1970s, kind of has a shutdown moment under the Reagan administration, the 1980s, is basically dormant for decades, and then boom in the 2010s takes off and is now, according to your words and your reporting and other reporting that I've seen, by many accounts, the fastest growing energy generation source in human history.

It's an extraordinary thing to compare those technologies, which again had sort of their birth dates in a way were right around the same time in the mid-1950s. You're bouncing on in your chair like you have something to say. So before I go to the next bit, which I think is going to be batteries, why don't you jump in there in my solar history? I love the way you've captured this because it's something I hadn't even thought about that much, but the way that they both emerged at the same time and at the same kind of moment in American history too. And what we have to think of is there's kind of like a cognitive dissonance for most people who made their bones in the energy business anytime between the 70s and today, which is like you had a big thing that got better by getting bigger, right? And then you have a tiny thing that gets better by doing more of it. And these are very, they're very kind of different modalities that you have to sort of embrace. And they don't care about each other.

We keep circling this idea. We might as well to say it. Bell Labs invented the solar cell. They also invented the transistor, which ended up becoming famous because of Moore's law. Like what you're saying is to a certain extent, nuclear power has the opposite of Moore's law. Oil does not have a Moore's law. Coal does not have a Moore's law. Chopping down trees to burn them does not have a Moore's law, which is the idea in transistors

where every 18-month, the transistor density on these chips seems to double in history. Photovotex cells have something like a Moore's law. There's something like an incredible ability to have a learning curve over time that allows of technology to become better, smaller, more efficient. We've never had this for technology, really, before solar. Is that right? We haven't had something that is in a position where you can make them by the billion. The closest thing I can think of would probably be the internal combustion engine in a car.

Like the kind of thing that we now, we make by, I don't know, we probably make 80 million of those a year new for cars. I don't know how many we make, you know, in related things like generator sets or things like that. But it's, you have to think about it as something like that. The difference with those, though, is that they still require fuel. Like, the cost of energy that you get from a gen set is not buying the gen set. It's what it costs to fuel the thing. The difference in the solar is that, like, you pay up front and that's it. You know, I have friends who have early solar panels from the 70s as home decorations that are still energized. They're like, yeah, like, if I plug anything into

it, it will still make power. weakly passively, less efficient than it did in 1976, but it still works. Whereas the gen set, if you're not fueling it, if you don't put anything in that generator, it will do nothing. So there's a big difference that I think, you know, again, we sort of inherently know this, but sometimes it's worth making it more explicit in terms of how we think. I asked you to pick the slide out of the 200 slides in your deck that makes you most bullish about the energy transition.

What is it? and why don't you tell me a story about it? So it's interesting. I spent all this time covering solar for years. Like, you usually can tell. Once a solar analyst, always a solar analyst. And covered wind as well, have looked at all of the technologies that generate power. And they're all nifty, and they've all improved a great deal.

It's the battery that I find most fascinating right now. So installations of battery energy storage, so not in cars, but attached to the grid, went up almost 70% last year. And the battery is actually the thing that allows these other technologies to kind of escape the range-bound nature of how much they can integrate into an energy economy. Without some kind of energy storage to serve all kinds of purposes in the grid. Sometimes it's like down to the intrastect time. Others, it's out to four, eight, 12 hours. Eventually it will be over multiple weeks is the thing that allows something that is never –

never always there to become reliable enough that you can count on it. And the battery is like connective tissue or enabler or something for all of these other things to happen. And the great part is that the batteries themselves follow a similar kind of dynamic to manufacturing solar or manufacturing consumer electronics. They all derived originally from the consumer electronics business. They now have their own logic behind them in terms of who makes what and what purposes they're optimized for.

But the battery is what allows us to blast through what used to be viewed as pretty consistent range limits in terms of how much you can integrate wind or solar into a system. And in a wonderful way, they've got their own industrial logic behind them. They've got a lot of highly competitive companies. They've got a ton of innovation and not just kind of pie in the sky innovation, but like the really kind of important grind it fine process innovation. to make these things by the billions. And it's the battery that allows us to go and build 40% solar in a market and 25% wind and have things work, along with more macro stuff.

But batteries are also going to, like, they're going to revolutionize the way that systems get built full stop where there is no grid. Like if you're going de novo to a place that has really under-energized, let's say, you'll find people building sort of expanding grids of solar, wind, and batteries together. that maybe make it less likely that you're building a lot of big centralized stuff. And they're fast. And they're interoperable. And the markets are really highly competitive. These are all good things. Again, back to a book that I hear someone's writing from an abundance perspective. Sorry, has written, not is writing. Has written. I'm going to take the bait there in this next question. Let me just sum up where I think we are so far. We've talked about the Goliath of sticky fossil fuel demand. We've talked about the David of

solar electrons and batteries that move them across time and space and make that solar powered electricity useful. As you said, I have a book coming out with Ezra Klein on, among other things, the future of liberalism and technology in America. And there is, in fact, a section about the challenges of building the energy infrastructure of the future here in America and why the U.S. doesn't entirely seem up to task to that challenge at the moment. You know, we have questions about interconnection cues. We build solar power, but how do we connect it to the grid? There's issues with long-distance transmission, where we fall in miles behind the pace on building the power lines that carry electrons from where they're generated to where they're used. In the biggest picture,

Nat, what do you see is the most significant impediments to decarbonization in America? and don't mention incoming policy just yet, because that is the next question. And I appreciate doing this as a two-parter because I think it's actually important to disaggregate different things because they can be impacted at different levels and with different approaches. The first is this, you mentioned the interconnection queue. For those and this, that is the volume of energy capacity in the U.S. that people want to build, and are waiting to connect to a wire, put it very, very basically. And it's just a silly number.

It's like twice. For folks, I'm sorry, I don't, I really, I'm not going to interrupt you again. I just, here I do find there some people for just like, for whom energy doesn't quite click without a metaphor they know more visually. Is it a little bit like a huge reservoir of water in your local town, but no one building the pipes to get that water to your house to use for cooking, to use for bathing.

Is that the kind of thing we're looking at? I'll give you a demand-related one, which is this. Somebody's built a bunch of subdivisions and they're not yet connected to that water. They're waiting for the water sewer power hookup. Right. ingo, got it. So it's the inverse of that, right?

It's the supply that's waiting to be connected, not the other way around. And, look, everybody wants to do this. Like, everybody wants to go and build. People are ready to do this. They have built their businesses, their lives. They hope to build their fortunes in building new energy infrastructure for the U.S. The things that keep that from happening are a mindset that is very slow in terms of what is required to approve one of these things.

There's a lot of, I would say, willful delay on the part of people on both the American left and the American right in terms of saying, I don't want this thing to happen. and so I'm going to sort of gum it up and keep it from going. There's a lot of proceduralism. I mean, you can't imagine, actually you probably can imagine, the amount of paperwork you have to do to get one of these things built. And that's millions of dollars, many, many years. It could be completely a sunk cost. So it's largely a mentality.

It's a decision. And I don't think of it as policy or politics. It's societal. It's a decision. Are we going to build things or not? and I try to strip it down as very simply as possible to just that. Like, that's a decision. Do we decide that we're going to build or not?

And the unfortunate thing is for a lot of the country, the decision is we're not going to. And look, there are parallels to. We're not also going to build housing in inexpensive coastal cities. We're not going to invest in expanding all kinds of other elements of local state capacity because we don't want to spend the money on it. But the reality is if you do that, you force all the future. right? And sometimes you could force all it completely. You could prevent the future. And I don't particularly care for that as an American. Let's put it that way. Bro, wow, that was, that was

a more effective summary of my book than I was prepared to offer on this show. So I really appreciate you doing that. Let's move to politics. You've got an incoming administration that has made noises and even plausible threats about shutting off a lot of the subsidies that came with the Inflation Reduction Act, ironically named, because a lot of it had to do with clean energy subsidizing. What do you currently consider the biggest political risk to the energy transition? Now, we're moving from the mentality change that we need in order to actually build energy to what's the biggest political risk to the energy transition at either the regulatory or legislative level? I think of these things as one thing being downstream of another. There's politics,

there's policy and then there's regime, there's inaction, like there's stuff that's actually happening. So the challenge is that policy, and it should be to some extent downstream of policy. You know, people are elected to enact the will of the voter. And that has to move from politics into policy. And then what comes after policy is ideally stability too. A mix their end of like, yes, we want to do what the voters want, but we also want to ensure a multi-political and ideally even multi-business cycle stability to give people a reason to invest. And you make it very, very difficult for people with long planning cycles who need to devote capital up front on the, not the expectation, but the necessity of having it paid back over years to just constantly change the underlying regime. provide not enough stability across political and business cycles for people to be able to build. That's something that we didn't use to have. I'll give you great analogy.

When I used to live in Washington, D.C., I took some interns to see an old coal plant that's just down the river from the airport on the Potomac. And the engineers there are something like, oh, yeah, we built this in 1950, and the two state utility commissions here and in Maryland, one across the river from each other, we're like, you know what? our demand growth is fast. Why build one of these when we can build three for thrice the price? And we had the state commissions got together with GE and we built three identical power plants. Like that's what a regime of stability looks like where the expectation is, yes, I can build something and keep it going. And when you rugpole people, regardless of their politics, in terms of what they want to build, it makes it very difficult from long-term commitments to come in. The other challenge in the U.S. is the not quite the guarantor, but the underwriter to an extent of so much of this real transformational stuff, as always in the United States, is the government.

Through grants, through long-term loans, through long-term contracts, through long-term stable policy, to plan for things that have to go on for years and decades. And look, you know, I can tell you that the world of the world of other parts of domains, like building things such as, as highways or providing for national defense are generally able to operate on longer timelines like that, that gives people reasons to invest, that gives people reasons, a reason to commit capital on a long-term basis. We're talking about the impediments in the bottlenecks to the energy transition. You've named checked an anti-build mentality. You've named checked political instability and lack of contiguous political will in order to make this transition. There's another thing that we have to throw into

the mix here, which is your deck points to evidence of what I have to just call a climate tech bubble or a climate investment bubble. There are big institutions, Goldman Sachs, Morgan Stanley, BlackRock, JPMorgan, City, Bank of America, all of those financial institutions have said they're going to leave the global net zero alliance. Now, what exactly that means? I'm going to let you tell me. But in addition to that, climate and sustainability references in CEO letters are plummeting. The S&P Global Clean Energy Index peaked now three and a half years ago. Messages of ESG in corporate announcements and earnings calls are down 60%. Something seems to have popped here.

What I'm less clear on is what exactly has popped and what it means. So what has popped and what does it mean? It's a really good question and it's an important one. ESG I've always thought of as a complex and kind of unwieldy framework in the sense that these are their ideals, but they're not necessarily an investment driver. And just tell people what it is. Yes, ESG, sorry. Environmental, social and governance have been lumped together into this acronym.

And it's funny, it's really important that you ask me to do that because when you split them out and say them together, I would hope that your listeners are saying, what do these have to do with each other? And the answer is nothing necessarily. And I say that because you can have companies with fantastic governance at a corporate level and a terrible environmental record. You can have companies that have terrible governance and a great environmental record. And they've been sort of lumped together as a set of things that you should view as lenses on investment. And they can be a lens on thinking about a company. But that doesn't mean that they're necessarily a driver of investment. And unless you are an environmental services company, then environment is probably not a prime mover

of your business. It's an element of how you measure what your business does. You could apply the same thing to everything else. Governance is governance. Like every company that's public in theory should care about governance. And in fact, we all do. Like you can go on to any kind of financial information service and find it all you need to know about who does what, who's linked to whom, So they've always been a kind of an unwieldy construct. And I think it got very, it got very overbought in 2020 and especially 2021 when it was a kind of commitment that people could make in the absence of having to enact that commitment yet. The same thing with, yeah, this alliance of the net zero banking alliance that got announced, which is, to be honest, a huge ask to make of institutions to make a commitment, you're going to do this particular thing to zero out a large. section of your business on this timeline. Imagine trying to do that for any other business.

You're going to zero out your loans to consumer packaged goods companies. You are going to zero out your investment in private sector housing in the next couple of years. That would sound like a big ask. And I think, I think, you know, the theory of these things was a lot easier than the practice. And there's a reason I say that it's like a 2021 thing. These all got announced all around the same time in 2021. And I've actually had. had people ask me, and I think not in a snarky way, and I tend to agree with the question, is this all something that happened as a sort of pandemic effect? And there is an element of that, I think, that this is, especially at a time when you're like, look, the world is materially

different now than it was a couple of years ago. We're in two years now where energy uses sort of declined and, you know, where we've still seen, we've still seen a lot of, like, energy and attention going to the clean side of the ledger on things. Let's make this commitment. But then it becomes very challenging to enact. And it would be challenging, regardless of politics, policy strategy, or anything. It's just a challenge. And it's another question beyond our scope here, but what is the role of these institutions? Like, what is it their job to do? Is it just a rhetorical bubble that's popped? Or something more material popped as well? Because among the things that I listed, a lot of it was literally words. Like global net zero

alliance, that's just words. Climate Sustainability references and CEO letters. Also words. mentions of ESG down 60%, also just words. But words mean something. And sometimes words are connected to material outcomes. So is the popping of the ESG bubble rhetorically connected to something that's more materially significant? So what's really fascinating about this is that, well, I should step back a little bit before I say it's fascinating. There are other parts of the world besides the United States. You know, your listeners may be aware that there are institutional and retail investment.

and say Europe. And they have priorities that are really not aligned with, nor do they particularly care about U.S. politics. So while BlackRock's mentions of ESG in its chairman's letter that he publishes every year have really plummeted, its sustainable assets under management are a trillion dollars. They have not gone anywhere. They have more than $400 billion worth of just retail ESG mutual funds

that had not gone anywhere. Now, they rose significantly from 2020 to like 2022 and then have bounced around, but they've not gone down. There's still interest in going there. And so this is where you get from kind of theory and rhetoric to practice and what actually investors want. And this is where I say that it's a sort of really complex ask to make of something. It's like if these are institutions that are just meant to be making smart, stable allocations on behalf of their investors, then they still, on that allocation, you know, side of things devote money to all of these kind of ideas. ESG exchange-traded funds and mutual funds are almost 5% of the total mutual fund assets

under management in the world. So that number just keeps going up, not growing the way it used to, but it hasn't vanished. It hasn't gone away. So a lot of what you see pulling back are these grand, but also quite unformed plans about the deep future versus the portfolio allocation. present day that is still continuing to go on. Let's hold this up against the solar buildout in the United States specifically. So you have this momentum, clearly, for building solar.

You have problems with interconnection. You have problems with long-distance transmission. You maybe have problems in terms of the vibes being a bit off in corporate America about being seen as a climate person, whereas, you know, four years ago it was cool. Now it's no longer cool. I know there's more to economics than coolness, but investors are people too, and they feel the vibes. Is it your view that all these things together,

and you can add IRA uncertainty, all of these things present an enormous challenge to the continuing momentum of solar in America, or is it your view that solar has reached a kind of escape velocity that makes it a little more, impervious to all of these challenges that we've just listed. I really like this question.

It's my hope and my medium strongly held conviction view that there's been a bit of an escape velocity here. In particular, it's an energy source that moves extremely fast when given the chance, like where you can find that area to connect, it goes very quickly. It builds very fast. It's increasingly reliable. I will tell you, I know loads of people in the development business, and they generally want nothing to do with an ESG label because it doesn't help them. They're in the business of moving

metal into the ground and providing a service for people over the course of 20 to 25 years. Like, that's really what they do, right? In that sense, they're no different than building industrial facilities or building real estate. And so for them, it's like, it's not particularly helpful to have any of that kind of ESG association right now. But the main thing they need is stability. You know, they need, or a sense that, like, if promises are made by the federal government, of a fiduciary nature that those promises will be honored, like that, I would say, is very important. They're not so much probably worried about their counterparty in Microsoft being able to pay for the energy for a data center. They're more worried that something has been yanked post facto from an

agreement that was already made. Last question in the U.S. before we go abroad. I'm fascinated by what's happening in artificial intelligence right now. And my hypothesis is that at the federal level, at the moment, AI policy is in large part, energy policy. Data Center construction investment last year, according to your report, exceeded construction spending in hospital buildings. So the data centers powering AI or other internet technologies just a few years ago

or half a third of what we were spending on hospital buildings. Now they're about to be in excess of what we're spending on hospital buildings just to give people a sense of the extraordinary growth of the sector. IT power demand is expected to double by 2028, the end of this administration. You have a really interesting comment about AI and energy in this report. You write, quote, when it comes to AI, energy is everything and energy is nothing. What does that mean? I love this one.

And to give full credit, I have the data from a group called E.Poc AI. And I, in turn, borrowed the idea from an investing partner and a researcher at energy impact partners named Andy Lubershane. And it's this awesome chart of some foundation models that have been trained. and it breaks down the cost of training those models by their inputs. And those inputs are what they call R&D staff, so basically, you know, software engineers and the cost of equity for those software engineers, which is non-zero, shall we say.

The AI accelerator chip, the other server components that go in, the interconnection cost, and then at the very end of that, all of the energy. So this is an older model, but it's a great example. Gemini 1, almost 50% of the cost of that is software engineers and the equity to pay them. 4.5% of the cost of that model is the energy.

It just doesn't matter. Like it's a tiny, tiny rounding error on an operational basis for what it costs to run things. But you've got to have it. It's not like you can make do with 80% of it. It's entirely binary. You either have it 0% of it or you have 100% of it. And you also have it at either 0% reliability effectively or 100% reliability. And this is really unusual in the world of really massive energy consumers.

If you run a steel electric arc furnace, you are so highly sensitive to the price of electricity. I can't even tell you how important it is for your business. Right. And like you either, you live or die on that. Your margin depends upon that entirely. For a data center, if you double the cost of electricity to train a model, they're like, fine, like, I need it. Like, I need the electricity in order to be able to do this.

And it's way more important that I have it than at what price. That's, okay, that's the train models. This is kind of a special case. It will change over time when this becomes highly operationalized, what they call inference, basically doing stuff for you and me, as opposed to training a model. But it just gives you a sense for why people are like, sure, like I'm going to build this data center.

And in the meantime, I will also build, a power plant that could run a medium-sized city, because that's much less of a part of a cost stack to me in terms of cash outlay, but it is a thousand percent essential to my operations to have it. Just one more question here. As we're moving from the first generation models to the deep reasoning models, and as we're moving, as you said, from expensive training to inference, which is just a fancy AI term for consumer use, right, powering the ultimate consumer use case.

Should we expect the energy share of AI cost input to go up dramatically? Like at that point has a certain amount of, say, engineering work been baked into the pie? Or you still have to keep those engineers on staff in order to run the best models and keep pushing forward the frontier. So overall, just costs keep going up and energy remains about 3 to 5% share of the total price. So I think that when it gets to inference, we will see much more sensitivity to the cost of of energy, right? Because then it becomes something that has a marginal cost that you start charging people for. You know, the challenge right now with training the models is that we don't really directly pay for that. That is effectively subsidized by a giant company

or by venture investors or a combination of both. Whereas inference is going to be like a service that you pay for, it has a marginal cost once you start doing it. I mean, it always has a marginal cost. The question is who is bearing that marginal cost. And there, I think, you will find lots of incentive to be as efficient as possible. For a couple, in a couple of different ways. One, you could either charge people less. I find that unlikely.

I find much more likely that you just do gobs more of it and offer that service much, much more widely. Like, this is something that apparently everybody, everybody has now had to dust off an 18, 59 book called The Cole Question and learn about the Jevons paradox. But I will put it in a more plain way, which is you essentially induce demand at a price point. The more of this, the more of this is available in a lower price, the more people are going to want it and use it, which is absolutely the case. Come on. You and I, you and I are not going to look at the incredible advances in deep reasoning and say, well, if we got 10% more electricity efficient, so I'm going to use 10% less of it and spend the same money. No way. Like that's this not really what's going to happen.

We're going to want to do more with it. But the way in which the system builds to accommodate that could be very different. We might not be building $5 billion data centers to do just that. This could propagate outward. This could be, you know, listen, at some point, these could reach the point where they're hosted on a device. That's certainly the dream for all of these sorts of things. And then the equation gets much more complex, much more interesting. But for right now, it's really build and build and build, right? It's like, do as much as you can. Very, very last question before we go to China. Can you put all this together? Because there's a puzzle here for someone to figure out. You have an energy transition happening in the U.S. alongside administration that wants to put its foot in the

accelerator of oil and nat gas production, at the same time that you have this absolute hydra monster of energy demand with AI coming up, is there something that we should be looking forward to, something around the bend, some prediction that you are looking at, or even just a place that we should all have our eyes on? Because it seems to me like these are three cars headed towards intersection and something's going to happen. Um, what, do you think that something is? So I think the main thing is that everything is, everything works in cycles, right? Like, we, we see when everything is going all up and up at the same time, I think it can give us a false sense of correlation. And I think we should be, we should consider that

industries have their own boom and bust cycles, which is really important to note. So I did a lot of digging around to come up with some slides for an analogy here. But you know, you and I are old enough to remember when everybody had to be laying 10 million kilometers of fiber a year in the U.S. Because the new thing was going to be that internet traffic was doubling every whatever, I can't even remember whatever silly number it was. That was the new priority we had to accommodate. The reality is that that bubble burst and then we built almost no new fiber for quite some time. We did the same thing with combined cycle, natural gas power plants at almost the same time. We have massive store of demand.

This is the new thing that we need to accommodate. Let's all go build and we overbuild. And this is in every resource economy, you should think of oil and gas as part of that. They tend to go in cycles. Although in the U.S., I would say that might be something that's easier to mute because you control the capital tap so much more nimbly in a world. of shale oil and shale gas, where the unit is much smaller than building deep water offshore oil. If you're building $5 billion data centers and you've built dozens of those and those

end up being overbuilt, then that is another, that is another boom that will sort of subside. But as with everything, that doesn't mean it goes away. It becomes a new source of capacity for other uses. You know, it becomes a new thing that people use for another purpose. It does not go dark. There is no way on earth that if you managed to build a multi-billion dollar data center, it's going to not be used for something at the moment. You know, the resource that it is is this collection of attributes where it's located, does it have electricity, is it permitted, you know, is there cooling available? You know, that becomes the asset. It's kind of fungible in a healthy way in terms of what people exactly are going to do with it.

Let's move on from energy in America to cars in China. Tesla posted its first annual sales drop in more than a dozen years last year. Tesla sales are basically flat year over year. Meanwhile, China's B-Y-D automaker didn't just outsell Tesla last quarter. According to your deck, I had no idea this was the case. it matched Ford's estimated sales in 2024. Did I read that right? You did read that right.

So BYD, which is an auto company that's not quite 30 years old, is now almost as big as Ford Motor Company. And it does that with not entirely, but with largely an EV fleet. Like it makes loads of electric vehicles. And it is a really extraordinary company. I'm here in Singapore. I live in a great big apartment complex. The garage downstairs me is full of BYDs.

And incidentally, other Chinese electric cars. But I'll give you a more dialed in sense. To be fair, this is not the biggest auto market. It's about 40-something thousand new cars a year. So take that with a grain of salt. But in December of 2021, BYD had 0.2% market share in Singapore. And in December 2020, it had 14.4% market share, which means that it had passed Tesla, Honda, Nissan, Nissan, Toyota, Mercedes-Benz, BMW.

Sorry, Toyota's the only one that didn't pass. It's passed every company except Toyota. In four years. Yeah, in four years. And that's pretty remarkable. And it's because it ships a great product that is constantly improving, refreshed all the time, a lot of new models, very competitive pricing. I travel all over the world.

I was in Brazil in August of last year. There were two giant billboards for two giant Chinese carmakers there, and they have factories there as well. There are ads for Chinese pickup trucks and pretty cool ones in Thailand, all over the place. Same in Malaysia. So there's a thing that I think a lot of Americans are wrestling with. We still have this vision of what happened when Japan arrived on the world's auto scene. Nothing compared to the scale of what's happened to China. China in 2023 made almost 40% of the world's automobiles.

It's unbelievable. It's the biggest maker of autos, the biggest market, it's the biggest exporter, it's the biggest exporter of EVs. All of those things are true. For people who are familiar with the EV fleet in the U.S. They know their Teslas, they know they're big automakers that have their own EV cars or their plug-in hybrids, what makes BYD so competitive? You mentioned quality. you mentioned updates.

From a price standpoint, what are we talking about here? So it's a good question. It varies market by market. So unfortunately, I can't give like a great reading on it, but it's very much competitively priced. If we were to have them in the U.S.,

they'd be competitively priced with anything. In China, which is the best probably example, the EV model underpriced the equivalent internal combustion engine model about 60% of the time. About 60% of the like from like, models are cheaper up front as an electric than the competing internal combustion engine model. So that, that I think is probably the best view I can give on it. And also, they're just new.

Like, like, there are tons of new designs. There are companies that I see in, in my garage here that I was dimly aware of a year ago. You know, there's, like, they're in my garage. There's a Zeker and an XPeng SUV. And I would be delighted if any of your listeners have ever heard of these companies. They're not. Right? But they're great. They are, and I hate to be too analogical here, but it is very similar to thinking about mobile telephony and what happened with this proliferation of models downstream from Apple and Samsung that are highly competitive, pretty ruthless, pretty innovative and willing to fight it out. And, you know, designed too in a way,

Tesla and BYD, both to their immense credit, are incredibly integrated in their supply chains in a way that is also, I think, breaking expectations and it's important to note. You know, they, they, I think BYD does something like 80% of stuff in-house, which is not the way that the North American auto complex or the Japanese auto complex does its work.

That's everything from batteries to seats. I remember Tesla did this years ago. They're like, we need to make our own seats. It's more efficient. And a lot of people in sort of legacy automakers are like, that seems ridiculous. But there's a logic to it if you can do it right. And BYD in particular has this like full integration of stuff that is really,

really impressive. It allows you to control costs throughout the value chain. It also allows you to integrate innovations much more quickly. Yeah, it's a company to work. watch, shall we say? If we didn't have a lot of structural reasons in North America for the auto sectors we do have, i.e. not importing a lot from certain other places, we would have loads of these around, I think. Now, last question. We didn't have a chance to talk about your charts on carbon removal. There are folks who are investing in and interested in fusion, including Sam Altman,

the head of Open AI. China, I just read, is building, feel free to edit this description, a laser to channel the power of the sun to create energy, question mark? You're struggling, too. Okay, so you read the same vague description of that technology that I did.

With the full fleet of energy technologies on the horizon at your offer, what are you most excited about? I want you to talk about something that is not a biggie today. It's not anywhere close to electric vehicles, but it's something that 15 years from now might be, or 20 years from now, might exist at the level that we now talk about these technologies that are increasing, like electric vehicles and solar.

What are you looking at? So one of them is energy storage, but of longer duration. So going out weeks or even months that enable you to, really build an ultra-high percentage renewable power grid because you can actually account for what the Germans, the Germans have this lovely word, don't cool flauta, which basically means no wind.

I'm not going to bother to translate the actual German, but effectively it means that there's no wind, and you have to back that up with power some other way. But the ability to build a grid sort of on the first principles of being renewed, power that you can then store for a much longer duration of time. Right now, we really only do this kind of operationally in terms of let's make sure that things are operating well and you arbitraging certain periods intraday. The ability to do this over the course of weeks or even

months would be amazing in terms of expanding the realm of the possible for what people can build. it will be, especially in a world if we can't, sadly, if we can't build a whole bunch of new transmission, this will be a part of what we do. Related to transmission, I'm super excited about things that allow us to having literally more dynamic understanding. This is nothing dramatically new, but there's something called dynamic line rating, which just means we have a better sense of how things perform. They're rated one way, but they're rated that way essentially in a laboratory. This is how it works under these ideal conditions, if we don't want to,

get our faces ripped off by some activity with high volatility. We don't want to go outside that. If you can dynamically rate the amount of power that goes through your existing infrastructure, it's like magic, right? You know, it's like you had a, imagine it being a plumbing analogy. You had a one-a-quarter-inch pipe. Now you've got a 1.75-inch pipe without changing the pipe. Like, you're able to put that much more through a system that already exists. I'm very excited about what I think will be coming in geothermal technology, which is receiving the attention that it is long deserved and hopefully achieving a liftoff that it deserves and that we deserve, which is an ability to provide very nearly always on highly

reliable power with a much more broad capability in terms of where it can be cited. I think if geothermal really, if the new geothermal technologies and approaches really work, we'll find something really neat, which is people go to where that resource is. Like if I'm building a data center for training, I'm less sensitive to location. Latency doesn't matter as much. So you build the energy park and then you build the assets to attach to them as opposed to the other way around. And look, I'm always excited to see what happens with nuclear.

My challenge with anything is that they're a little outside my kind of aperture for the next three to even five years. And I think one thing that it's important to remember, and this is a very neutral statement, is that every company, and I mean this, every big buyer wants to make a commitment like this because it's in their interest to make a call option to contract to the nuclear company. And I think they're also genuinely very interested in making these things work. But there's a timeline that takes time. And I'm just not adept enough to sort of be able to handicap or call one of these technologies. they're back a little bit to my earlier thing. Even if we have an absolute explosion

or proliferation expansion of the small modular reactor fleet, it's still only in the range of hundreds in the next, I don't know, five, six years before the end of the decade, as opposed to tens of thousands of long-duration batteries

we hopefully will be building in that same time. Yeah, I feel about nuclear the way the X-Files poster I want to believe expresses. I want to believe. I'm not a nuclear bro. I'm certainly not anti-nuclear. I want to believe, but what I see is

that nuclear power plants are literally the most expensive thing you can possibly build in the world, and they take between, at least, in the West, seven and 50 years to complete. And that's just a long time to spend $10 to $20 billion in something

when the energy that we need, the clean energy that we need is much closer to the here and now. So we'd love to have another podcast with maybe a historian about the rise and fall of nuclear because it's a fascinating story. It's worth doing. I'm not, yeah, I'm not sure that it's going to save us this decade. Nat Ballard, thank you so much. That was wonderful. Thanks, Derek. My one thing to remember from today is that solar is happening. and I am investing in Nat's hope and even his conviction of moderately held faith that this technology has achieved escape velocity in America. This is the first energy technology that operates according to the rules of software in manufacturing.

You know, you look at oil drilling, you look at fracking, certainly you look at burning trees. This is not something that gets tremendously more efficient over the span of a few years, right? oil drilling is more efficient than it used to be, but it's not moving in anything like Moore's Law. But the cost of photovoltaic cells truly is declining at a rate that is similar to the declining cost of software. Now, I think a smart person who knows that of that energy is going to say the cost of those solar cells is not the full cost of the energy itself. You need labor, you need installation, you need integration, into connection. Absolutely, you need all of that. But solar is powered by a different kind of technology that's really important.

And as I said in the open, I am really, really interested in the fact that a lot of people were worried about resource restrictions for the clean energy transition. But lithium, one of the most important metals for the entire energy transition, has seen its resources double since 2018. So I am faithful that, hopeful that on the market side, the technology side, Solar is a revolution that's here to stay. Talk to you next week.