Energy strategist Kingsmill Bond joins me to explain why the transition to “electrotech” is unstoppable, whether or not politicians care about climate change. It’s not the reduced emissions, it’s physics (electrotech is more efficient) and economics (it’s cheaper). Despite political headwinds in the US, China and emerging economies are racing ahead with electrification and sector after sector is seeing peak fossil fuel consumption. There’s no stopping it: electrons will triumph over molecules.
(PDF transcript)
(Active transcript)
Text transcript:
David Roberts
Hello, everyone. This is Volts for November 21, 2025, “Clean electrification is inevitable.” I’m your host, David Roberts. Way back in 2021, I had energy analyst Kingsmill Bond on the podcast to discuss his optimism about what he calls electrotech, the cluster of technologies — including renewables, batteries, heat pumps, EVs, digital control systems — that is enabling the global shift to clean electrification.
Since then, Bond has moved from Carbon Tracker to RMI to the research firm Ember. In the meantime, events have only bolstered his optimism. In September, he and his team at Ember released “The Electrotech Revolution,” a report and an absolutely riveting slideshow arguing that we are on the cusp of a fundamentally new energy age: “a century of evolution is converging into a decade of revolution.”
We are experiencing simultaneous revolutions in the way we generate electricity, the way we transport and store it, and the way we use it. Fundamental physical, economic, and geopolitical forces have rendered these revolutions inexorable and irreversible, or so the report argues. By the way, it’s almost a shame to cover this report on a podcast, since it really is quite beautiful. Any lover of charts and statistics should check it out on the web.
I am thrilled to have Bond back with me today to discuss this revolution, its drivers and consequences, why it is unstoppable, and what we can expect to see in the next decade.
With no further ado, Kingsmill Bond, welcome back to Volts. Thanks for coming.
Kingsmill Bond
Thank you, David. Glad to be here.
David Roberts
I’m thrilled about all this. I’ve got so many questions. This whole report might as well be a mission statement for Volts. This is what Volts is all about. I was just wallowing in these slides, absolutely loving it. Let’s start here. One of the things I think, one of the services you’re doing here, even aside from all the data and information, is just giving this cluster of trends and technologies a distinct name and identity. Talk about how you distinguish electrotech from cleantech or climate tech more broadly.
Kingsmill Bond
I should start off by saying that the name was invented by my colleague Dan Wolter.
David Roberts
We’ll thank him then.
Kingsmill Bond
Kudos to him. Exactly. What we’re trying to do here is just to—Carlos Perez, Venezuelan academic, talks about technology clusters coming together, and we just wanted a way to describe, as you said earlier, the fact that these new technologies are coming together around the electron. The reason why electrotech and not cleantech is I think cleantech is a wider or slightly different definition and slightly tarnished because it had pulled into the mix lots of technologies which simply don’t have learning curves and don’t have the efficiency characteristics that we talk about and also are not cheaper.
For example, biomass or carbon capture or hydrogen. These are not really electrotech technologies. We wanted to focus on winning solutions which enjoy the drivers of physics and economics and geopolitics. That’s why we focused on electrotech.
David Roberts
This is a subset of cleantech, you might say the specific subset of cleantech around the electron.
Kingsmill Bond
Exactly. The key point is that by calling a name and pulling it all together, we can draw into the tent AI, which I’m happy to argue is a very clear electrotech technology.
David Roberts
We’ll get to that later.
Kingsmill Bond
We bring in batteries and so on and so forth.
David Roberts
The other thing I think is useful, one is just to distinguish this particular subset of cleantech, but another is losing the “clean” thing and replacing it with electrotech. I think it’s also a reminder that is really important, that I’d love to hear you talk about for just a minute, that the trends and technologies involved in electrotech precede any concern about climate change, or rather the large-scale trend toward electrification that is pulling all these clusters of technologies slowly together. That trend predates climate concern, predates anything clean.
This is something that is going on regardless. Talk a little bit about the history angle.
Kingsmill Bond
As your listeners will know, electrification has been going on since 1880 and the United States was famously responsible for half of all global electricity demand in the 1930s. Absolute global leader in electrification in the past, and I suspect it will be again. These are technologies which have been growing for many decades. What’s interesting is that it’s only in the last few years that a lot of the technologies have crossed key cost tipping points where they can challenge the incumbency on cost. When you get cost tipping points and volume, that’s when you get very exciting things happening.
David Roberts
I just want to drive home the point that the merits, the advantages of these technologies — economically, physically, etc. — stand apart from climate concern. It is good that they’re reducing emissions for obvious reasons that we agree on, but more or less all of this would be happening regardless, don’t you think? Just because of this better technology?
Kingsmill Bond
Exactly. These technologies, as I say, predate our concern about climate, they have been driven by the laptop and the iPhone in recent years, are massive drivers of our capacity to get cheap batteries and cheap solar panels. Solar panels have much more in common with a laptop, for example, than with a barrel of oil. These are technologies which to a large degree are spin-offs from the tech sector.
David Roberts
That’s important to understand. It’s easier to understand this stuff as the next chapter of the digital technology revolution rather than the climate technology revolution. That makes more sense of them.
Kingsmill Bond
What are we doing? Electricity by definition is a digital technology, whilst fossil fuels are analogous, and we are digitizing our energy uses. AI is the electrification of thinking. The more we electrify and digitize and create flexibility, the more we can use the resources. It has a lot in common with the way that the tech revolution evolved. I have a background as a tech analyst back in the 1990s, so this is very familiar territory for me.
David Roberts
Before we jump into the dynamics, let’s talk more fully about what technologies we’re talking about. As you say, there are these clusters converging and you cite three clusters: renewable electricity supply, electrified demand, and digital coordination. Briefly describe what those three clusters are and what the convergence is.
Kingsmill Bond
The first point is that what is unifying all of these classes, apart from the electron, is that they are modular and small and on learning curves. That narrows down the sphere of options for you. On the supply side, the two standout technologies — we are of course very ecumenical, anything could join the tent — but the two standard technologies which do work are obviously solar and wind. On the demand side, the standard technologies are electric vehicles and heat pumps and the electrification of light industry, particularly in China, which is a very big thing right now.
The messy middle is the most interesting. That’s the area where there’s going to be the most innovation. Your friends from Piclo, for example, or Octopus, or the incredible innovation that’s going on — battery storage. AI we’re already claiming as a connection technology — but you’ve all kinds of software solutions and then you’ve got better cables and HVDC lines. All the great folks are on your show, David.
David Roberts
It’s all painting one big picture. You’ve got new, better supply technologies, electrified demand technologies — EVs and heat pumps — and then all this digital stuff, all the software to coordinate how the supply matches up with the demand. That’s electrotech. What I think is the most important part of the report, the part I enjoyed the most, is this first bit about fundamental drivers. Understanding these three fundamental drivers is crucial for understanding the energy world right now and understanding why, despite short-term fluctuations in this or that market or country or specific technology, the long-term arc here is inevitable.
Understanding these drivers helps you understand why that is. Let’s go through them. The first is physics. I love this point. This is a point that I bang on all the time on the pod. I think Saul Griffith was the first person to make the point on the pod, which is just the simple point that electrotech is more efficient than fossil fuels. Talk about that for a minute. What does that translate to? What does that mean?
Kingsmill Bond
In simple terms, it means you can get two or three times as much useful energy from electrotech solutions as you can from fossil solutions. At the moment we have this remarkably inefficient system. You put into the factory, as it were, at one end, 600 or 650 exajoules, and you get a round of primary energy and you burn it. As you burn it, you have these enormous, inevitable thermodynamic losses of between 40% if you’re a coal-fired power station, or 50%, or if you’re a car, about 75 or 80%. You can’t avoid this stuff.
David Roberts
A car is the worst of all. It’s amazing how little of the energy we put into cars ends up moving them.
Kingsmill Bond
We’re pouring, from our calculations, two thirds of the primary energy into the air and wasting it.
David Roberts
This is one of the facts I want to put on a T-shirt, Kingsmill, because this is something that people in our world are familiar with at this point because they have heard it many times. But this is something that blows people’s minds outside our world when they hear it for the first time, which is that two thirds of the energy we dig up out of the ground in raw form is wasted. That is mind blowing. It is probably not mind blowing to you anymore, you are so used to it.
But I think that blows people’s minds. We are wasting two thirds of the raw energy we have access to. Moving to electrotech doesn’t eliminate waste, but it radically reduces it.
Kingsmill Bond
Your waste levels are going to be around 10%.
Again, it depends how you structure it. But if you think about it from first principles, if you want electricity and you have a house in Japan, for example, you can pay somebody to frack the ground in the middle of America, get the gas out, pipe it to the coast, freeze it, put it on a boat, bring it to Japan, unload it, unfreeze it, bring it to a—
David Roberts
Put it in a pipeline.
Kingsmill Bond
Put it in a pipeline, burn it, lose another two thirds of it, turn it into electricity, get it to your house. You can do all those things, that’s great. Or you just buy a single solar panel, stick it on your roof, and get exactly the same electrons. That’s the point. It’s dramatically inefficient — the fossil fuel system. We, being finance folks and very focused on the numbers, can calculate that’s four and a half trillion dollars of lost energy. It’s also a four and a half trillion dollar opportunity for entrepreneurs such as Piclo to aim at.
For decades we had no other solutions. We couldn’t come up with anything better than fossil fuels. Now we can. That is the exciting point.
David Roberts
You cite Amory Lovins’ distinction between “obedient electrons” and “fiery molecules.” That’s important to understand. Electrons are more manageable. They’re digitally manageable. You can move them around in precise ways. When you’re burning stuff, the molecules fly all over the place, you can’t—there’s a limited amount of control you have over your fiery molecules. Electrons are always going to be tamer, let’s say.
Kingsmill Bond
They are tamer. The big problem with electrons in the past was, first of all, they were very expensive. Until recently, you go back to the start of electrification — obviously the start is zero — but even through to the 1930s, 40s, electricity was about 5% of final demand. This stuff was really expensive. We solved the cost problem. The other massive problem with electricity always was you couldn’t store it. You couldn’t store it. You had to have instantaneous supply and demand everywhere at all times.
That’s why we’re excited about these convergence technologies. Suddenly you have these batteries coming through. $100 per kilowatt hour is the number everyone is talking about right now, but they are doing it for $60 in China. You can store this stuff now. We are just scratching the surface of what that means.
David Roberts
You have a new piece up on Ember that frames this as an agricultural revolution. For the first time, we have granaries. Imagine global agriculture if you couldn’t store grain, you just had to eat it out of the fields. You had to time it to meet everyone’s hunger in real time. Imagine the revolution of being able to store grain. It changes all the systems and all the system dynamics. That’s roughly what’s happening in energy. We have granaries now.
Kingsmill Bond
You have it there in the Bible. Joseph and his Technicolor Dreamcoat is all about granaries. An even more powerful analogy, and more recent, is refrigeration. Until refrigeration, food rotted in the supply chain, and then around 1900, refrigeration comes along — massively resisted by the ice industry. The great line saying “ice is better for you.”
David Roberts
And all those ice jobs lost.
Kingsmill Bond
Refrigeration comes along and it creates a whole new industry. Argentina can supply Europe with meat. You don’t need massive milk trains taking huge amounts of space outside New York because they have to all come in by 6 o’clock in the morning and you have this obsolete infrastructure. You can see where I’m going with this analogy — the capacity to store electrons is going to completely change the rules of the game. At the moment we have around 400 gigawatt hours of storage to 4,000 gigawatts of solar and wind.
It’s completely feasible that within 20 years we will have two or three times as much storage as we have renewable capacity.
David Roberts
One of the things I always say to people, just to spark their imaginations, is to imagine what an electricity system will look like when it doesn’t matter where and when you’re producing your electrons. When we have enough storage that we have rendered that irrelevant. Anybody who understands how the electricity system works, that’s a real mind blower. One point I wanted to make under this particular heading of physics is that, as you said earlier, it’s important to distinguish this is not true of green tech more broadly. It is specifically electrotech that has this feature of being much more intrinsically efficient.
It’s not true of biofuels or hydrogen.
Kingsmill Bond
Hydrogen, as everyone knows, is the most expensive form of electron because you have to turn it into hydrogen then back again. I want to talk about biofuels because biofuels are a fake solution to the climate problem.
David Roberts
They go the wrong way on efficiency.
Kingsmill Bond
The efficiency of biomass is sub-20%. You’re putting in 100 exajoules of biomass in the front end of the machine and all you get out the end is 20. What a waste! It’s a spectacular waste. It’s even more of a waste when you think about biomass conversion efficiency of sunshine. You put biomass in a field and it’s collecting less than 1% of the sunshine. Then you’re wasting half of it in the whole conversion process, while a solar panel in the same field is going to collect 20%. Shameful!
David Roberts
People complain about the land use of solar and I just say, “Why don’t we take a fifth of the land the US uses to grow corn for ethanol, hand that to solar, and we can power the whole country off it.” Another point I wanted to make, distinguishing electrotech in the physics sense from the larger green tech: another efficiency, quasi-efficiency argument people sometimes make against this stuff is about materials. Materials efficiency. I’ve been over this on the pod before, but there’s a great chart in this slideshow just comparing materials used.
Briefly compare the materials efficiency of electrotech versus fossil fuels.
Kingsmill Bond
The key point about fossil fuels, and it is very obvious, but people miss it.
David Roberts
People don’t think about it. It’s worth pointing out.
Kingsmill Bond
What we specialize in is the pretty obvious, but you have to be numerate. We’re burning 16,000 billion tons of fossil fuels every single year. Then you come back in the next year, you burn it all over again. In contrast, the entire build out of electrotech solutions would require a fraction of that in order to build the infrastructure. We’re talking a billion tons. That’s it.
David Roberts
I do think it’s fair to say that just constructing the generator — if you just look at the amount of materials to construct an electricity generator — you are using more materials per kilowatt output for solar than for a natural gas power plant. But then, as you say, when you’ve constructed the solar plant, you’re done spending money. The fuel rolls in for free forever. After you construct the natural gas power plant, you’re going to buy natural gas every day for the rest of your life. That is material, natural gas is material that is digging, that is disruption of the landscape, etc.
If you count the fossil fuels you have to dig up and burn, it’s just wildly more materials, mining disruption, transport, etc. Wildly more in a fossil fuel system.
Kingsmill Bond
In the particular case I remember, I did the calculations a few years ago: if you do solar versus coal over the lifetime of producing electricity, it’s 2,000 times as much stuff from solar panels versus a coal plant. Because coal is very heavy and you can only use it once.
Incidentally, the latest IEA WEO is out today and it has its advantages.
David Roberts
That’s the World Energy Outlook.
Kingsmill Bond
It has its advantages and disadvantages. One of the things they did is look at an issue we often get as pushback against the argument that fossil fuels are very heavy. People say, “But minerals require lots of ore.” The IEA did the calculation and said, “Coal and oil also require lots of rock and stuff to be moved.” When you do the total calculation, they worked out that the total amount of stuff that needs to be moved for the fossil system is about 100 billion tons versus about 10 billion tons for all of the minerals.
That particular fake argument we had coming down the pipe, that’s also not true.
David Roberts
The other important fact that you note in the report that’s worth pointing out is once you dig the materials to build a solar panel out of the ground, you can recover them from the solar panel. We have, at least in our line of sight, something like a closed loop where you’re done digging, where you have the materials you need and you’re just recycling them again and again, which is not something you’re ever going to get out of fossil fuels.
Kingsmill Bond
This idea of continuously recycling valuable minerals does require a change in mindset because we’re used to just chucking stuff out. If it’s valuable enough, and some of these minerals are very valuable, then that’s what you’re going to do. I’m sure you’ve had the folks from Redwood on the pod. That’s what they do. They take these valuable minerals, they recycle them. Let’s imagine you’re losing 10% in that supply chain.
If, over 15 years, you get 10% efficiency gains, then effectively you’re recycling these minerals or the use of these minerals forever. The next time you build a car, it’s going to be probably 40% more efficient and therefore you can do with 40% less mineral.
David Roberts
Your efficiency gains are keeping up with your materials losses. You get to a closed loop, which is never going to be possible with fossil fuels. These are three fundamental drivers. The first one is physics — electrotech is intrinsically much more efficient. Much less waste, much less disruption, material, transport, etc. Pick your metric. It’s just a lower lift and more power with less digging. That’s the physics. I’ve hit this point on the pod a couple of times. We had Michael Liebreich on. I think Saul Griffith made this point.
You often see these primary energy comparisons, I’m sure you’re very familiar with them, that show that out of the primary energy entering the economy, only a tiny sliver is renewable. That is deceptive for precisely this reason. Electrotech doesn’t have to replace that whole chunk. It only has to replace a third of it — the third that’s actually going to useful energy. We don’t have to replace the waste. The target is much smaller than that graph makes it look.
Kingsmill Bond
Your target is three times smaller than it looks. It’s also a fake argument that you have to replace everything. Of course, you don’t replace everything overnight. One of the very first things they teach you in finance is you don’t think about the size, you think about the growth or the acceleration even. That’s the point — these technologies, you don’t want to look at their size, you want to look at the share of the growth they’re taking. That’s the exciting point — we’re taking 90% of the growth in electricity generation.
By the end of this decade, we’ll be taking all of the growth in final energy demand as well, coming from electrotech. We’re there, we’re eating your lunch right now.
David Roberts
That’s a good segue to the second fundamental driver here. The first is physics, the second is economics. The basic insight here, which I’ve hit on the pod many times, is electrotech specifically — those technologies are on learning curves which mean they get cheaper as they scale and fossil fuel doesn’t. Another thing that’s remarkable to people who are not familiar with this stuff is that the inflation-adjusted prices of fossil fuels today are roughly what they were 100 years ago, 200 years ago. That’s not true of electrotech.
Electrotech goes down. Talk about why that is.
Kingsmill Bond
In its simplest form, it’s technologies versus commodities.
Technologies have the capacity for people to innovate, particularly if they are small and modular. As they get bigger and more people get involved, there are more brains and areas of talent figuring out ways to reduce costs and make it better. As a result, costs fall over time. As they get bigger, they get cheaper. With fossil fuels, it is the opposite. You extract the cheapest stuff first. As they get bigger, there is constant pressure for costs to rise because you have to go further and further offshore and deeper and deeper.
David Roberts
Digging farther and farther down, you are going to more and more extreme environments. That is, all things being equal, pushing the price up. Fossil fuels also have technologies involved which are pushing the price down. That upward pressure and the downward pressure more or less over time balance out, which is why the prices are, even though they fluctuate a ton in the short term, eerily steady over long periods of time.
Kingsmill Bond
Nobody can explain why this is the case. It has happened the way it is. There is lots of innovation going on in fossil fuels, but it is balanced by the fact that you take cheap stuff out first and, net — net, the price in 2025 is roughly what it was in 1900 or 1925.
David Roberts
That’s wild to me.
Kingsmill Bond
Can you imagine the solar panel back in 1976? $100 a watt. Now it’s 10 cents a watt.
David Roberts
Because the sun notably does not get harder to get the more you harvest it. It just remains just as easy to get on the 5 millionth unit as it is on the first.
Kingsmill Bond
You were saying to Michael Liebreich on your pod that you need inspiration in this world, but this is surely the most inspiring: that we could use live sunshine, which gives us as much energy in five days as the entirety of all fossil fuel reserves — dead sunshine or fossilized sunshine — can do if you burn the whole lot. That’s extraordinary. We have this incredible opportunity to harvest the sun alive every day and then store it in batteries every night and wake up the next morning and do it all over again.
David Roberts
Because of the association of climate tech with environmentalism, people still associate it with efficiency and doing less and pulling back our aspirations and becoming smaller and degrowth, all that stuff. For very obvious reasons that we’re talking about here, if you have big grand ambitions for humanity, you don’t just want to 2x the amount of energy humanity uses, you want to 5x it or 10x it, you want to desalinate the oceans and go to Mars. All these big things. You simply physically can’t do that with fossil fuels.
You’re going to reduce your home to a cinder and burn all the fuels before you can get there. The only way to do really big futuristic things is with the sun’s energy. That’s the only source big enough.
Kingsmill Bond
This is again why, as you say, it’s a hundredfold increase in potential energy capture just by harnessing the sun. For me, the best analogy to what’s going on here is the biomass to fossil transition of 200 years ago. The other one, of course, is famously, as Ian Morris says, the hunter-gatherer to farming transition 10,000 years ago. Nobody remembers that, no data for that, but you get the point. Every time humanity stumbles upon or finds or innovates a dramatically better way of harnessing energy, a lot changes.
David Roberts
A lot of growth. A couple of questions about these economics, the learning curves. One is, and I think this is a question everybody has: these learning curves, there’s something almost supernatural about them because you just look at solar — what are the forces that are making solar cheaper? It’s not one thing, it’s some research over here, Germany passed these laws, China invested this money. It’s a grab bag of things. Yet the net outcome is this eerily consistent drop in prices every time global deployment doubles.
You’re seeing the same thing with all these other technologies. There’s something almost eerie about it. One of the things everybody wonders is, are these learning curves going to continue? Are they going to go on forever? Do we know anything about whether there are limits to these things?
Kingsmill Bond
To your point about whether it’s eerie or not, don’t forget we’ve been living with a famous learning curve — which is Moore’s Law — for decades and nobody ever says that’s particularly eerie.
David Roberts
I think it’s eerie, but it’s really cool.
Kingsmill Bond
It’s really cool. We run with it. The world expert in this stuff is Doyne Farmer in Oxford. He makes a very good point. He says, “We do not know where the innovation will come from, but we can be very confident that small modular technologies, once they get onto a learning curve, it keeps going for a very long time.” The great hope of the fossil fuel system is, and you see this in the IEA’s World Energy Outlook.
David Roberts
They’re still doing it?
Kingsmill Bond
They’re still doing it. Still doing it. Still flat. Still flat solar. It’s embarrassing.
David Roberts
It’s not they haven’t heard the critique at this point. We should say what we’re talking about. The listeners are very familiar with official predictions of the price of these electrotechs — solar panels. What the official predictions say over and over again is that the price is going to plateau, that it’s going to stop falling and plateau. They have been saying this every year for about 20 years, decades now.
At this point they’ve heard the critique, the predictions are theatrically wrong over and over again. It’s not that they haven’t heard feedback. Have they seen that they’re wrong and yet they keep doing it? Speaking of mysterious, this is mysterious. What is going on with that dynamic? Is it that if you take the learning curve seriously and draw the line so that it keeps doing what it’s doing, you end up with some really crazy projections, projections that sound crazy and utopian. I think that scares people off. What do you make of that dynamic?
Kingsmill Bond
In the case of the International Energy Agency and the CPS scenario, the carbon profit scenario we call it, that they released this morning, it is described as an explanatory scenario, but in fact it is normative. It is a normative scenario to please the fossil fuel system in the same way as the net zero scenario was a normative scenario to please the green warriors. These are both normative scenarios to cater to particular groups of stakeholders. You want to maximize the profits of the fossil fuel system? Go for the CPS. You want to deal with reality? Look elsewhere, please.
David Roberts
It’s not just the IEA, it’s the EIA. Even a lot of big private consultancies and research firms — it’s a culture almost in the analyst community that they refuse to accept the good news here, they refuse to believe the good news in front of their eyes and they keep thinking it’s going to end.
Kingsmill Bond
The reason why is not rocket science. When you’re paid by the fossil fuel industry to come up with their scenarios, then you are going to come up with very conservative scenarios. That’s what has been happening right across the board. What the IEA has been able — or had been able — to do for the last four years was to crack open that story and tell the truth. It’s been fantastic. Of course they’re still doing great analysis in lots of different areas.
The baton is now being taken up by other folks who can see a much more optimistic future.
David Roberts
A couple of individual questions about this learning curve dynamic. When I talked to you in 2021, you included electrolyzers in your list of electrotech technologies that are on learning curves. When I talked to Michael Liebreich a while back, he was poo-pooing electrolyzers, saying “they’re not falling in price, they’re not proceeding.” What’s your take on electrolyzers? Are they on a learning curve, are they part of this electrotech package, and more broadly, hydrogen too?
Kingsmill Bond
We don’t get everything right, David. Four years ago we were more enthusiastic about this stuff and felt that they would get onto learning curves. It turns out that, as Michael Liebreich has said, whilst the electrolyzer is small and modular, all of the kit around it is not. Therefore, we haven’t seen that learning curve effect. We haven’t seen inherently within hydrogen the same levels of efficiency as in all these other electrotech technologies. We’ve moved on from that, focusing on the stuff that’s working.
David Roberts
You, Michael, have downgraded your expectations for hydrogen and hydrogen technologies in this whole package?
Kingsmill Bond
There’s something important to be said here. We all got, myself included, four or five years ago, a little gaslit — not so much by the hydrogen industry as by the incumbency. The incumbency telling us we have to figure out how to solve the last 10%.
David Roberts
“We need a liquid fuel. We have to have a liquid fuel.”
Kingsmill Bond
Where else in your life do you try and figure out what you’re going to be doing on a Tuesday in 25 years’ time? Of course you don’t figure that one out. You just get on with your life and you try and look for the next five or ten years. The Chinese have this famous saying, “crossing the river by feeling the stones.” That has been and will continue to be the right solution — we need to figure out where we can deploy these technologies at speed and scale with reasonable cost and do it with massive scale.
The more we do it, the better we get at it. We had this other piece in the report saying that the ceiling of the possible is now at 75% and we’re at about 20% electrotech. Who cares about the last 25%? Put some of your brilliant scientists onto it and they can figure it out. In the meantime, the vast majority of our time and our talent and our force and our genius needs to be poured into getting from 20% to 75%. By the time we get there, we’re going to be aiming for 95%.
David Roberts
This is the baffling thing about Bill Gates’ recent intervention in this whole thing. He’s still on that, which feels like vintage 2018 to me. He’s still on that, “Oh, we got to figure out the last 10 to 15%.” Can we just do what we know how to do? Can we just do what’s right in front of us? We’ll figure that out in the process.
Kingsmill Bond
Poor Bill. He’s obviously much richer than I am, but he’s not very good at energy and he’s vastly influenced by the fossil apologist Vaclav Smil with his book, “How the World Really Works.” What does Vaclav do? He takes these four sectors, the last 25%, and says because we can’t solve that, there is no energy transition. The analogy is a bit like saying in 1900, “Because I can’t fly to America, you can’t drive a car.” No, of course you can. You get there eventually.
David Roberts
One other question about technology and learning curves, which I didn’t see talked about much in the report, is grid-forming inverters. I guess that’s more one of these digital software products than a physical product technically, but I didn’t see a lot about them in the report. Are they on your mind and are they also getting cheaper or have you done much thinking about them?
Kingsmill Bond
We love this stuff. Anything — we love all these technologies, these clever technologies in the middle. While it may be a 110-page deck, that’s also quite small for the entire system. We are planning to do much more work on these technologies in the messy middle. The whole point is that this is going to unleash this cornucopia of innovation to enable people to figure out solutions in lots of different ways. Grid-forming inverters are a fantastic solution to enable you to solve some of the very specific problems.
I remember being beaten up 10 years ago by my ignorance about not understanding all these issues. Now, thank goodness, somebody has come up with some answers.
David Roberts
The whole idea that giant multi-ton, constantly spinning masses are going to be preferable to what is a little computer has always amused me. This is obviously a step up. The first fundamental driver is physics. These technologies are more efficient. The second driver is economics in that these technologies are all getting cheaper while fossil fuel prices more or less stay the same, which means their economic triumph is inevitable. We’re in a fuzzy spot right now where individual electrotechs are passing up individual fossil applications in individual places.
In the fullness of time, these lines are all moving in one direction and fossil fuels are all staying the same. In the fullness of time, these are going to win on economics.
Kingsmill Bond
They’re going to win. The other obvious point to make is this is why I find it funny. You talk to people being very expert at, say, gas in Malaysia, and “here’s 100 reasons why your ideas are wrong.” Just stand back for a second and think about this from a system perspective. All that’s happening here is Economics 101. You’ve got a giant cheap source of energy coming into the system. I don’t know exactly how it’s going to play out. I don’t know exactly where it’s going to end up, but I do know it’s going to drag down the cost curve.
If you are at the top end of the cost curve and think that it’s going to stay there — the LNG industry, for example — forever, then you’re going to be very disturbed by the reality of all of this cheap energy coming into the system. It will find a place.
David Roberts
As the report says. This is not some arcane, secret knowledge we have developed. The graphs are right there in front of them too. The graphs with the line pointing down are right in front of them too. I don’t know what they think is going to happen between now and five years from now that stops that line from going down. This is all very foreseeable. The third fundamental driver — we have physics, we have economics. The third one is geopolitics, which is such a vast, huge topic.
Much is going to happen and we understand little. I feel less able to predict this than almost any of these other areas. The one thing we know for sure — just talk a little bit about the balance in the world between fossil fuel exporters and importers. I think that is a very basic fact that people need to understand.
Kingsmill Bond
The starting point here is that three quarters of people live in countries that import fossil fuels. That means Europe, famously dependent for 60% of its primary energy on imported fossil fuels.
It means China, where it’s about 20%, it means India, about 30%, much of Southeast Asia, and very large swathes of Africa and South America as well. Most people in the world are importers of fossil fuels. Under the Pax Americana, 80 years since the Second World War, that was fine. Americans, thank you very much, protected the seaways, and we could source from lots of different places and it was all fine. That whole framework really started to break down after 2022 and particularly after President Trump started to weaponize fossil fuels in 2025.
The fossil fuel folks in the US thought they were being brilliant, but you should be careful what you wish for, because what have you done? You’ve just panicked the rest of the world into thinking, “Okay, enough’s enough. We need to get our own energy.”
David Roberts
We think we have them over a barrel. But we don’t. They have alternatives now. This has not sunk in perhaps with the Trump people. If you squeeze these fossil fuel importers hard enough, they are going to look around for alternatives.
Kingsmill Bond
That’s precisely what’s happening now. You show up in Europe, you force the Europeans to sign this $750 billion deal. What do you think they’re going to do the minute the door is closed? Do you really think they’re going to do nothing? Of course they are. All across Europe now, people are saying, “We have to electrify.”
David Roberts
We cannot ever again be in the position of being bribable, manipulable — we have been in the last decade.
Kingsmill Bond
The famous analogy I always think of is the 12 young men and women that were sent to the Minotaur to be eaten alive every year until Theseus saved them. This is how we feel in Europe — just really aggrieved. The Indians feel exactly the same way. Every Indian will tell you that the country imports $150 billion a year of fossil fuels. This fossil dependency was okay in the past. Not great, but okay. Now it’s not okay anymore.
People are looking for solutions. It’s highly complex geopolitics. There’s lots of great work done, for example by Thijs Van de Graaf at the Brussels Institute of Geopolitics, identifying how people are going to change the way they operate for this new world. This brings us on to why Electrotech is the solution to this. The reason it’s the solution is because everyone has got — everyone, apart from Singapore and Belgium and a couple of other countries.
David Roberts
Arctic Circle, I don’t know what their capacity is —
Kingsmill Bond
But everyone’s got a huge amount of renewable potential. If you focus your renewable potential on the easy-to-solve areas, you strip out farmland and forests and cities and lakes, everything else — the places you don’t want to put these solar panels — and focus on wastelands and low-quality land, we have 100 times as much renewable potential. The great challenge of this generation is to figure out how to bring that to bear. There are two ways of doing it. The first is to put up the solar panels and the wind turbines and stuff.
The second, and the one that people have not focused enough on, is to electrify everything. The electrification angle has stagnated in the West since 2008. We got leapfrogged by China, but we have also been leapfrogged by Southeast Asia. There is a massive electrification story going on all across Asia. Here we are sitting in England, twiddling our thumbs, fiddling around and failing to do it.
David Roberts
One of the counterarguments here, which is a little glib, but we should address it, is some people say, “You’re just trading dependence on fossil fuel producers for dependence on China. China’s the one manufacturing all this stuff, so you’re just shifting your dependencies.” Why is that wrong?
Kingsmill Bond
I wonder if this argument could be seeded by our friends in the fossil fuel sector. The main reason this argument has no merit is that it’s the difference between renting and buying. If you buy a solar panel, you’ve got it for 30 years and you can use it and generate electricity for 30 years. If you rent a barrel of oil, you burn it the next morning and then you have to get another one the next day. Think how scared you feel in certain jurisdictions as a renter — versus owning your own house.
That’s the fundamental difference between the two. It’s even better than that because China does indeed dominate solar panel production and battery production and minerals refining. If you look more widely at the electrotech sector, you discover lots of areas where other countries play a huge role. In this messy middle, most software, if you look at stats, comes from the US.
Germans sell more cables than they buy solar panels. The Japanese make transformers. There are lots of pieces inside the system that other countries can get involved in making. Furthermore, it is a challenge to all of us. Stop moaning about the fact that you are not making this stuff and make it. There is no one stopping you.
David Roberts
You can’t just choose to dig up fossil fuels if you want to.
You can only dig up fossil fuels if you’re sitting on top of them. Any country can decide to make this stuff. One of the things people miss, which is a very intriguing bit in this report, is that China is not just exporting these technologies, it is also exporting the capacity to manufacture these technologies. Then you’re not, I don’t even know what the analogy is with the renting and owning, but then you’re buying a house-building company, and then you’re really not dependent.
Kingsmill Bond
There is this famous number that Chinese FDI on electrotech is $200 billion, which is the same size as the Marshall Plan. There are these fantastic maps of Chinese-built fabs being built in Hungary and Vietnam and Uruguay and Chile and all over the world — they’re putting up fabs so that people can make it for themselves.
David Roberts
Someone linked it on Twitter. It was in a Chinese catalog of some kind, where you could literally just click to buy a solar panel factory — a modular solar panel factory — which is mind blowing to me. The manufacturing of these things is only going to get easier and any country can do it. As you say, I forget the exact statistic in the report, but something like 92% of countries in the world have 10x the renewable potential of their current demand. There are not a lot of limitations here.
You can imagine a world where all or most countries are energy self-sufficient. That is a fundamentally new state for humanity to be in. We’ve never experienced anything like that. The energy flows and energy dependencies have shaped human history and society and politics from the jump. A whole new chapter of our time on this earth. We’re entering it and we have no idea how that is going to play out. Very interesting to watch.
Kingsmill Bond
It is interesting and we are blessed to be alive at the moment. The other point you are touching on is the real opportunity in the sun belt, for the 80% of people who live in countries within a reasonable distance of the equator. That is 80% of the global population. If you look — sadly England is not one of them.
David Roberts
I looked at the map of renewable potential and the middle of Europe is the lowest. You guys got screwed there. It’s the lowest renewable potential.
Kingsmill Bond
All it means is we have to work harder. The point is it’s a profound driver of justice and opportunity and growth and development for the people of the global south and emerging markets who don’t have or don’t use a lot of energy right now. Suddenly they can harness and store sunshine. That’s got to drive development because in the old fossil fuel punctiform world, wealth and power went to the small number of countries — the petrostates, 5% of the global population who were fortunate enough to sit on these massive fossil fuel deposits. That’s why they can build skyscrapers a kilometer high, but that’s where all your money’s been going. That’s $2,000 billion a year of free money going to build this stuff. Now that’s going to come to you and you could get your own energy. How cool is that?
David Roberts
Let’s talk a little bit about China and its role here. It seems specifically what people don’t see coming is road transport, specifically in China, is getting at what you call the soft underbelly of oil demand. Talk about China’s role here, what it’s doing, and what effect that’s having on global markets.
Kingsmill Bond
China is doing this not necessarily for the sake of the planet, but for the sake of their own interests, as most other countries. Ten or twenty years ago they identified an opportunity in leaning into electrotech. It has paid off. There was no guarantee 20 years ago that it would, but it has paid off. Since 2008, China and the US and Europe were running neck and neck in terms of electrotech patents. Now China completely dominates — 75% of patents.
They have very high shares of manufacturing of these technologies. Possibly the most interesting story is that China is electrifying while we are not. Electrification in the West is stuck at just over 20%. China is electrifying at 10 percentage points a decade. They’re now at 30%, they’re heading for 35%. As a result, they’re reaping the geopolitical benefits of being able to sell these technologies abroad and so on. The reason we say this is because it’s not an option to do nothing.
China is inventing technologies which will make them stronger and stronger because they are more efficient and superior to burning stuff. Ultimately, if other countries wish to compete with China, they are going to have to get involved. I’m looking at you, America. You can’t just sit around burning limited supplies of shale gas and hoping for the best. This is another reason why I’m very optimistic. The tech bros get this. They get the fact that electrotech is a tech story. They know what to do, they are doing it. AI is an electrotech solution.
The great thing about America is that eventually it always gets the right answer. I can be very confident that, led by folks in California and other states, this will happen in the US too.
David Roberts
The Trump people seem to be very taken with these forecasts which you often see these days, which say fossil fuel demand is going to be strong and robust out through 2050, 2060, 2070. I think that’s what they’re banking on. Even if a shrinking pie, we have the biggest piece and that will bring us prosperity for the next several decades. I think that’s a disastrous gamble to take. I suspect you agree.
Kingsmill Bond
As a professional analyst, it’s easy to identify the errors that are being made in these models. The biggest error is, and I can critique the latest model that just came out from the IEA called the carbon profit scenario, the CPS: solar and batteries have been falling at costs of 10 to 15% a year for decades and then suddenly they stop falling. The annual cost fall drops to about 3%. That’s the first error they make. Suddenly cost falls are going to stop. The second error they make, as ever, is that the growth stops.
The third error they make is they go, “Okay, I get that I can’t block the Chinese, sadly, and the Europeans haven’t got any fossil fuels, but fortunately I can force my LNG, my oil onto these folks in the emerging markets, they have to take it.” But guess what? They don’t. They’re not. You’re seeing this incredible leapfrog. Two thirds of the emerging markets have leapfrogged America in terms of solar as a share of generation. It’s crazy.
David Roberts
That’s a good segue from China because part of the China story, or a piece of the China story, is these other emerging markets, the non-China emerging markets. The big worry is China developed on the back of coal, famously.
Kingsmill Bond
So did the UK and so did America.
David Roberts
And so did everyone so far. The worry has always been that these emerging markets are going to follow China’s pathway. If they did that, we’re all screwed. There’s been hope for years about leapfrogging. I’m sure you’ve been hearing the term leapfrogging for decades now and it really seems like it is happening. It’s mainly happening because China is overproducing all these electrotech technologies and making them extraordinarily cheap. They’re just leaking out into, “I did a pod with some guys in Pakistan where they’ve imported 40% of their total country’s demand worth of solar panels in two years.”
As you say in the report, that’s happening in a bunch of these markets.
Kingsmill Bond
It’s far more prevalent than people realize. Two thirds of the emerging markets have already leapfrogged the US in terms of solar as a share of generation.
David Roberts
That means solar is a bigger chunk of their total generation pie than it is of the US’s — two thirds of emerging markets.
Kingsmill Bond
That’s two thirds by demand, not by number of countries — by demand is even more powerful. I’m talking about India and Brazil and Mexico, Vietnam. I don’t see the drivers as Chinese overcapacity leaking out. It’s more that China’s cheap stuff and these guys are buying it because it suits them. That’s going on in solar and there’s no ideological commitment against solar. Why would you? It’s sunshine. The other one, which is even more interesting arguably, is this electrification vector: people are electrifying their industries, electrifying their transport systems.
US EV share of sales is around 10%. Vietnam is 30%. Nepal, if I take the export stats, is 75%. This wasn’t meant to happen for decades — that the emerging markets leapfrog to these superior technologies. The reason they’re doing it is because it’s cheap. You can buy an electric car for $10,000, you can buy a solar panel for $50. Why wouldn’t you?
David Roberts
Do you think the consequence of that is that, as you say, the US is betting on LNG exports? That’s the fossil fuel industry’s big bet in the US right now. They are betting that these emerging markets are going to need LNG. Furthermore, they’re trying to tell us this LNG is cleaner than the coal these emerging markets would otherwise be burning. Do you think electrification in emerging markets is going to moot that story? Consequently, do you think there’s going to be a bunch of overbuild and stranded assets in the LNG market?
Kingsmill Bond
There’s a very obvious battle of overcapacity between LNG and solar right now. The US industry is tooling up for 300 BCMs of LNG exports. Chinese are tooling up for about 1,200 gigawatts of solar exports. Those two numbers generate roughly similar amounts of electricity. The difference is the Chinese one does it in year one and it does it for the next 30 years. The point is it’s a battle of overcapacity between the two. Just simple economics: the break-even level that you require to make LNG projects work, it depends which one it is, but let’s call it around $10 per MMBTU is $80 per megawatt hour electricity in Vietnam. You’re getting it at $50 already today. You stick a battery on top and it’s $70, and by 2035 it’ll be $40.
David Roberts
These arrows are all pointing down. Whatever it is now, it is less than that five years from now.
Kingsmill Bond
The IEA says, “What that means is you’re just going to have very cheap gas going out there.” What that means for the people producing is they’re doing all of this stuff for nothing. Thank you for pouring billions of dollars into this export capacity, this highly complex technology where you’re not even going to benefit from it. It makes no sense. The tragic thing here is that these are the last hurrah of the fossil fuel system.
What is the point in doubling down on the technologies of the last millennium when you could be dominating the technologies of this millennium?
David Roberts
In a perverse way, the fact that weve fallen so far behind on electrotech, people are using that as an argument for why we should double down on LNG. “Hey, China already dominates this stuff. Let’s do what we’re good at.” I dont think that’s going to work out long term. You mentioned already that three quarters of total energy demand can now be supplied by electrotech. We know how to do transport, we know how to do buildings, building heating and cooling. This has been one of the fun things about the last decade or two — just watching electricity conquer things that a lot of people confidently said it would never conquer.
It’s just moving up the ladder doing things people said it couldn’t. Consequently, you’re seeing peak fossil fuel demand in various industries, sectors, countries. You make a lot of this in the report, these various peaks. Talk about these local peaks we’re seeing. It blew my mind a little bit that global fossil fuel demand for various things has already peaked. Go through a little bit — what’s peaked and what hasn’t.
Kingsmill Bond
What we’ve done is not very sophisticated. We’ve just taken the International Energy Agency’s World Energy Balance database, which is behind a paywall, but it’s not very sophisticated. We just take a look at that data and split the world, the energy system, into two parts: making electricity and then using electricity — final energy, they call it. Then we just take a look at what’s going on. In the making of electricity, you’ve got solar, wind, growing and growing. If you continue to extrapolate this growth, at some stage they’ll supply all the growth of demand. That moment was the first half of 2025, when global electricity demand grew about 380 terawatt hours and solar and wind alone were 400. In the biggest sector of fossil fuel demand, responsible for 40% of fossil fuels — electricity — we’re at the peak right now. These peaks are things that can happen, you can have outage of hydro in China and nuclear in France and you get bounces, but we’re there or thereabouts in the electricity system.
When you look at the three pieces of the final energy demand, which are buildings and industry and transport — the three biggest pieces — I’ll come on to the other piece, which is petchem, but the three biggest pieces: in buildings, it’s a 2018 peak, we haven’t grown since 2018, fossil fuel demand. In the industrial sector — not widely appreciated in energy for industry, that is to say, not with feedstock, which is petchem — you got a 2014 peak, we’ve plateaued since then.
David Roberts
This is one of the things people hold out as one of these difficult to decarbonize sectors. This is something Bill Gates is always talking about. All these people who are electrotech skeptics hold out industry as, “But what about industry?” But as you show, globally, fossil fuel demand in industry generally peaked in 2014.
Kingsmill Bond
The reason why is not rocket science. China stopped massive build-out of its energy-intensive industries and commenced electrification of its light industry. People love to obsess about the hardest-to-solve last 20%, that bit of steel, whatever it is. But how about all of the textile industry and the pulp and paper industry, machinery industry, and the food industry?
David Roberts
I did a pod recently on those mid-temperature, mid-sized industries. It’s just big heat pumps — we know how to do that.
Kingsmill Bond
In our calculation, we’re at a clear peak of 32% of industrial sectors because the IEA disaggregates into about 15 separate ones. 32% had a peak, about 60% are on a plateau, and 6% of sectors have a little bit of growth still coming through. That’s the industrial sector. Then you look at the transport sector, and transport is particularly interesting because people are always saying, “You can’t possibly decarbonize flying and flying to Australia, can’t do boats.” Flying is not the largest piece of the puzzle. The largest piece of the puzzle by far is road transport. We are decarbonizing road transport such that the latest data that came out this morning from the IEA suggests that there is no growth left in road transport for oil. The largest piece of the puzzle has reached a peak and is going to start to enter into decline. It takes time — these things last 15, 20 years.
David Roberts
I found the quote from your website: “Fossil demand has been flat for industrial energy since 2014, for buildings since 2018, for road transport since 2019, and may peak for electricity this year.” These peaks are not in the future, they are happening. A lot of them are already in the past. One of the things I remembered about our first conversation is you talk about the business dynamics that apply to an industry that is in decline, which are very different than the business dynamics that apply to a business that is expanding. Even if the amount you’re using the year before the peak and the amount you’re using the year after the peak are pretty close to the same, the dynamics shift dramatically.
Talk a little bit about what happens to an industry when it is past its peak.
Kingsmill Bond
We don’t need to look very far to come up with a classic example of what happens. Look at the US coal industry after 2014. The US coal industry tools up for growth, borrows for growth, builds for growth, analyzes for growth. All of their analysts were talking about growth coming forever and it doesn’t come through. You get minor declines and within two years half the industry is gone bust. Within five years the industry gets delisted from the S&P completely. That’s what happens to capital-intensive industries dependent upon continuous growth.
This was the inspiration for Carbon Tracker, this observation that as capital-intensive industries — and this is the difference between fossil fuels and tobacco. Tobacco is highly capital unintensive and sells its products to consumers. Fossil fuels are highly capital intensive and they’re competing with a superior energy source. Once these capital-intensive industries reach a clear peak and struggle to get any new capital because people can see the writing on the wall, then they go very rapidly into significant shock. That is one of the reasons why the fossil fuel industry has been fighting hard to obscure what’s going on.
They just want another five years or another two years, another round, another couple of bonus rounds — so they can retire.
David Roberts
You see investment in that industry starting to shift toward dividends and enriching executives and shareholders while they can. There’s a “let’s party while we can” vibe happening in that industry. Do you think five years from now these peaks will become visible enough that these almost mechanical economic consequences — but there’s a lot of psychology here too, as you say, just the investors’ disposition toward an industry that’s not growing anymore is very different. Do you think that’s going to be visible and obvious to everyone soonish?
Kingsmill Bond
The whole point about peaks is that they’re not particularly visible at the moment. Famously in financial markets, nobody rings a bell at the top of the market. It’s not really that hard if you assume continuous growth of the new, then inevitably it first of all takes growth and then pushes the old technology down. That’s been the story, not just in energy, but in every sector, in our own experience, for decades. It’s totally normal for the growth they need to push the old one out. Because it’s bumpy and choppy and there are lots of variables involved at the moment, it’s a little hard.
By 2030, when you’ve got over a thousand gigawatts of solar coming onto the system every year, when you’ve got 90% EV penetration in China and very high EV penetration across the emerging markets, at that point, it just becomes harder and harder to sustain this argument. This is where we bring a little bit of market savvy to the debate. At the moment, it is slightly debatable, but by the end of this decade, if these growth trajectories continue, it’ll be undeniable. This kind of framing — that we have a peak and then decline — we’re not picking it out of thin air.
It’s exactly what’s happened in country after country. Go and have a look at the numbers. Germany and the United States as well, for what it’s worth, Brazil — these are countries where fossil demand peaked decades ago, and then it does start to decline. This is completely normal. From our calculations, half the world’s past peak fossil demand in electricity generation, two thirds in final energy demand. The only country which is holding up the sky for the fossil system was China. China was so enormous that they had continuous growth in fossil fuel demand.
95% of the growth since 2018 on a net basis came out of China. That’s why the fact that China, you’re getting more and more signals, as we’ve been hearing on your show and others, that Chinese fossil fuel demand is peaking because of the huge amount of solar and wind they put into the system and the electrification of transport, that’s a big moment. It’s the pivot moment in the system.
David Roberts
Mainly the road transport thing is the huge growth of EVs in China. That is the body blow for global oil demand.
Kingsmill Bond
Coming through to trucks now — these famous stats now, 20% of trucks. You and I have both been in this game long enough to remember the people who said, “It’ll never happen in trucks.” This is a similar story to the car story. You start small, gets big.
David Roberts
One of the questions everybody has about all this stuff is there have been big political reversals in the US, and not only in the US. There has been a sharp 180 reversal in the US, but not only that. There has been a pall cast on the whole thing. The US is now in UN meetings about international treaties, bullying other countries not to sign them. They are bullying other countries into taking our LNG. There has been a real distinct pivot in the US. How highly do you rate that as a risk for electrotech? Do you think that is going to stop anything or slow anything down?
Kingsmill Bond
No. This is the tide. As King Canute found out 1,000 years ago, you can’t stand in the way of progress and tell it to go back. Maybe they could have done this 10 years ago, but it’s just too late. To substantiate that point, take the recent forecasts made again by the IEA. They downgrade the forecast for solar and wind deployment in the US by 40 gigawatts a year. They upgrade the number for Asia by the same amount. They downgrade US EV sales, they upgrade emerging market EV sales.
It’s a messy story, but you have to step back for a second. All of us are conditioned, myself included, to look up to America. Imagine America leads in everything and you copy America, do well. Copy California is always my mantra. In this particular instance at the moment, for a while it has been China leading — China’s 50% of everything, 50 to 80% of everything. The United States is about 10%. For 10% of the system, for a few years, and of course it’s not all of the 10% because in the US you have states and cities and individuals who are doing it anyway.
For that 10% which was aspiring to China and it’s not going to China, China for the next three years. That’s a shame. The Chinese aren’t sitting back and saying, “That means we’re not going to do anything.” On the contrary, the Chinese are pouring their technology into the emerging markets.
David Roberts
We’re leaving the playing field. They must be cracking the champagne and high-fiving in party headquarters there because we’re stepping out of the way and letting them dominate this stuff.
Kingsmill Bond
It’s a shame for the US and for us in Europe as well, but the game’s not over. We’re at 20% electrotech in the system and we’re in a marathon. As many people have said, we’re a few miles into the marathon, but you don’t just give up because your competitor’s a few hundred meters ahead. There are lots of innovations going on everywhere. The Chinese don’t have a monopoly on genius. America’s got lots of universities, so does Europe. There are lots of ideas coming up from all over the place.
It’s absolutely worth engaging with this story and we’re going to have to get to terms with it, like it or not.
David Roberts
Final question. Let’s talk a little bit about AI, because this is on everyone’s mind and there are two stories to tell about AI. One is AI itself and how it becomes part of electrotech, as you said earlier, and I’d love to hear your thoughts about that. The other side of the AI story that we’re always hearing about is this massive spike in electricity demand which a lot of people in the US are worried is going to prompt a dash for gas. It seems to me that’s the easiest thing for these utilities that are lazy.
If you’re lazy, that’s the easiest thing to do because all the systems are set up. I wonder on both sides of that what your thoughts are about AI.
Kingsmill Bond
Let me state it boldly. AI is an electrotech solution to match supply and demand of electrons. It is the electrification of thinking and it is a tool to enable higher levels of electrification. That is the starting point. AI is great and it enables stuff to happen. Specifically for electrotech, because it’s only AI that can match the supply and demand of electrons in real time. It is a key way to solve one of the most complex problems facing humanity. Let’s focus on that for a second. When it comes to this second argument, I find it bizarre because it’s completely data free. If you look at the expected growth of AI demand to 2030, it’s 10% of the growth in global electricity demand. It’s 50% of the growth of electricity demand in America. It’s a big deal in America because electricity demand is very low. It has been since 2008. It’s a big deal in the US but it’s not a big deal at a global level. It’s 10% of the growth.
It’s smaller than space cooling, it’s smaller than appliances and fridges. Not to belittle it, but it’s incredibly efficient, AI. You get an awful lot of stuff done for a click, which is why we use it all the time. The other point is that people are obsessing about how much gas is going to be used. We did the math and it’s about 200 or 300 terawatt hours of gas that will be used for AI. Global gas demand is 8,000 terawatt hours, so we’re talking about 3% of the gas.
If you’re in Ireland or North Virginia, it’s a big deal. I get that. But it’s absolutely irrelevant to sustaining the fossil fuel system. Even more exciting, this came out in the report this morning. The IEA calculates that you need about 500 terawatt hours — about two exajoules — of electricity to run AI and you’re going to get about 13.5 exajoules of savings from it, particularly in the industrial and transport sectors. It’s a 6 to 1 reduction in energy demand coming out of using stuff more efficiently as a result of AI.
That number, that 13.5 that everyone’s been hunting for, it’s a holy grail. Amory Lovins wrote a brilliant paper making exactly this point several months ago: that AI was highly likely to result in more efficiency gains than it was going to take in energy demand.
David Roberts
That’s a little bit of a guess, isn’t it? That’s an educated guess. We have no idea what’s—
Kingsmill Bond
We have no idea. Do we have any idea how much — where does this number come from for 500 terawatt hours of electricity demand? That’s the other point. Sorry to be boring, but look at Koomey’s Law. Dramatic potential for reducing the amount of energy demand per unit of processing power. Incredible numbers coming out. It’s 33 times, I think, the number Google talks about, of improvement in a single year. It’s been a wonderful hype story, but I think we can now claim AI as an electrification technology.
Thanks very much, folks. Now let us just get on with it.
David Roberts
You think that AI’s contribution to spreading electrotech and making it more efficient is going to vastly outweigh whatever new fossil fuels it pulls onto the grid in the short term?
Kingsmill Bond
There you go. In terms of the numbers, in terms of the total domestic on, it’s 6 to 1. Gas is only, say, 200 out of the 500 terawatt hours. We’re already talking 14 to 1. Call it 14 times as much or 10 times as much reduction in energy demand as you’re going to have in supply. It’s not surprising that we’ve been worried about it. We should be worried about it. We shouldn’t just do this stuff without due care and attention.
We count what we can see. We can see these vast data centers going up. There must be huge amounts of electricity demand. But you can’t count all of the reduction just coming through.
David Roberts
It’ll be interesting to see how those numbers balance in coming years. I’ve taken too much of your time. We’ve gone over almost everything. As a way of taking us out, I wonder: we’ve talked about all these little mini peaks that have happened here and there. What do you think is going to be the next headline moment you’re anticipating? Is it a clear peak in global fossil demand for electricity? Is it a clear peak for China? Is there a news event that you’re looking forward to that’s going to be symbolic of what’s happening?
Kingsmill Bond
That China peak, because we get very frequent data on China and it will be total fossil demand in China — because they have this official target, 2030, and as everybody knows, the Chinese tend to hit their targets five years early. That is very low quality analysis, but it is also substantiated by all the data coming out of China, saying that coal demand in China is down year to date by 3 or 4%. Coal demand in China for electricity, to be clear, has been responsible for the vast majority of the growth in energy demand over the last decade, sorry, since 2018.
Peaking demand in China — clarity on peaking demand in China — will change the debate because people go, “I get it now. We had a peak in the OECD in 2007, peak in China in 2024 or 2025, and then it’s going to reverse.” Who’s left?
David Roberts
Two biggies.
Kingsmill Bond
Those are the two big ones. Those two together, as they decline, will outweigh any residual growth coming from those small parts of emerging markets which are still growing.
David Roberts
This is a slightly quasi-technical question, but is there an agreed — how many years of falling demand is enough to convince analysts that we’ve passed a peak? Is there any shared answer to that question, or is it just a judgment call?
Kingsmill Bond
Ultimately it’s going to be financial markets. That’s where you’re going to see it first, where you start to see some of these sectors falling out of bed and then it just filters its way through to public consciousness. In financial markets, you don’t wait for the peak to be clear. You act before, because you want to get out before the top. It’s financial markets. You’ll see this. That’s one reason why people need to pay attention right now, because it’s very surprising. There’s so much geopolitical tension out there. The price of oil is still relatively low. When you start seeing it coming through in stock prices and surprising things happen and M&A taking place, which you wouldn’t necessarily expect because companies are far weaker than you think they are, that’s the classic lead indicator.
David Roberts
When the Titanic deck chairs start being frantically rearranged, that is our signal. Kingsmill, this has been awesome. It was awesome the first time. It was awesome this time. I’d love to have you back in five years and we can once again wallow in all the delightful progress. It is always a thrill to talk to you.
Kingsmill Bond
I will take you up on that, David. Assuming we’re still going, we should do it.
David Roberts
Thank you for listening to Volts. It takes a village to make this podcast work. Shout out, especially, to my super producer, Kyle McDonald, who makes me and my guests sound smart every week. And it is all supported entirely by listeners like you. So, if you value conversations like this, please consider joining our community of paid subscribers at volts.wtf. Or, leaving a nice review, or telling a friend about Volts. Or all three. Thanks so much, and I’ll see you next time.
