China is killing the US on energy. Does that mean they’ll win AGI? — Casey Handmer

How will we feed the 100s of GWs of extra energy demand that AI will create over the coming decade? On this episode, Casey Handmer (Caltech PhD, former NASA JPL, founder & CEO of Terraform Industries) walks me through how we can pull it off, and why he thinks a major part of this energy singularity will be powered by solar. His views are contrarian, but he came armed to defend them. Watch on YouTube https://www.youtube.com/watch?v=3cDHx2_QbPE&t=3s; listen on Apple Podcasts https://podcasts.apple.com/us/podcast/china-is-killing-the-us-on-energy-does-that-mean/id1516093381?i=1000722101753 or Spotify https://open.spotify.com/episode/3vV7Vki2qJhO2AyodIaTMo?si=eeb8cb025abe4315. SPONSORS - Lighthouse helps frontier technology companies like Cursor and Physical Intelligence navigate the U. S. immigration system and hire top talent from around the world. Lighthouse handles everything for you, maximizing the probability of visa approval while minimizing the work you have to do. Learn more at lighthousehq.com/employers https://www.lighthousehq.com/employers - To sponsor a future episode, visit dwarkesh.com/advertise https://www.dwarkesh.com/advertise. TIMESTAMPS (00:00:00) – Why doesn’t China win by default? (00:08:28) – Why hyperscalers choose natural gas over solar (00:18:01) – Solar's astonishing learning rates (00:27:02) – How to build 50,000 acre solar-powered data centers (00:40:24) – Environmental regulations blocking clean energy (00:44:04) – Batteries replacing the grid (00:49:14) – GDP is broken, AGI's true value must be measured in total energy use (00:58:45) – Silicon wafers in space with one mind each This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.dwarkesh.com https://www.dwarkesh.com?utm_medium=podcast&utm_campaign=CTA_1

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Dwarkesh Patel?

  Today I'm interviewing Casey Handmer. Casey has worked on a bunch of cool things. Caltech PhD on some gravitational wave, black hole gimmick stuff. Then Hyperloop, then the Jet Propulsion Laboratory at NASA. And now he is founder and CEO of Terraform Industries. Casey, welcome.

• 0:16Speaker 01

  Thank you. It's great to be here finally.

• 0:19Dwarkesh Patel?

  Big question I'm interested in, to the extent that AI just ends up being this big industrial race. Who can build the most solar panels? Who can build the most batteries? Who can build the most GPUs? And transmission lines and transformers and et cetera, et cetera. This is not what the US is known for, at least in recent decades. This is exactly what China is known for in the right, where they have like 20x the amount of yearly solar manufacturing the US has.

• 0:46Speaker 01

  Yep.

• 0:46Dwarkesh Patel?

  Obviously, we have X work controls right now, but over time SMIC will catch up to TSMC's leading edge. So what is a story exactly of how the United States wins this? Like, why does China just not win by default?

• 1:00Speaker 01

  Do you think that China is better at capital allocation than the United States? Do you think the Chinese business environment is better for business than in the United States?

• 1:06Dwarkesh Patel?

  I think you can make this first principles argument about these other industries where they're killing it. But like, doesn't seem to have hampered BYD or cattle.

• 1:12Speaker 01

  Look, they're so much better at building high-speed trains than the United States.

• 1:15Dwarkesh Patel?

  Right.

• 1:16Speaker 01

  Right. I would never like hold up a flag saying I'm really good at building high-speed trains. Like, that is just a sign that you're really bad at capital allocation. Like, why would you devote in 2025 so much industrial effort and money?

• 1:26Dwarkesh Patel?

  They're devoting a lot to solar overcapacity, which in your opinion is the key to future industrial growth.

• 1:34Speaker 01

  Accidentally correct.

• 1:34Dwarkesh Patel?

  They call the most important thing correct, right? Which should count for something.

• 1:37Speaker 01

  Well, they're in a similar situation to Europe, but unlike the United States. So the United States is the luckiest goddamn country on earth. Because it's surrounded on two sides by oceans and on the other two sides by like friendly allies, right? China is surrounded by 15 countries who are mostly hostile to it, right? With no good like mountain ranges or rivers or anything to really separate them. And then they get all the oil, almost all the oil from the Middle East, right? In countries that they don't control, don't have strong diplomatic relationships with, on fleets of oil tankers that they can't defend because their navy doesn't have the ability to operate effectively in the Indian Ocean.

• 2:11Dwarkesh Patel?

  But you're working on this, right? If you get synthetic fuels, working at Terraform.

• 2:14Speaker 01

  Personally, yes.

• 2:14Dwarkesh Patel?

  Doesn't that asymmetrically help China? Which might be fine.

• 2:17Speaker 01

  It does. It absolutely asymmetrically helps China. We're not currently working with China. We don't plan to, but like, the physics is very obvious. You know, and synthetic fuels have been around for 100 years. Like, there are projects in China right now working on synthetic fuels. It would not surprise me if they were. Thinking pretty seriously about this.

• 2:30Dwarkesh Patel?

  Just to spell out for the audience, if China has all this electricity production, and the bottleneck is that only a third of final energy use in a modern economy comes from electricity, the rest, you need gas and whatever to transport things.

• 2:44Speaker 01

  Coal, they use a lot of coal in China.

• 2:45Dwarkesh Patel?

  Right, and what Casey is inventing is a technology to turn that electricity, which only can supply a third of end uses right now, into synthetic fuels, which can supply 100% of the electricity your civilization needs. So then China's energy advantage then becomes overwhelming.

• 3:04Speaker 01

  This technology levels the playing field. It levels the playing field a lot.

• 3:08Dwarkesh Patel?

  Right.

• 3:08Speaker 01

  But at the end of the day, you know, China still contains the poorest Chinese people anywhere on earth.

• 3:13Dwarkesh Patel?

  I mean, but.

• 3:13Speaker 01

  Never underestimate the capacity for an autocratic dictatorship to shoot itself in the foot.

• 3:17Dwarkesh Patel?

  I don't know, I, whenever I, I agree that they've obviously made bad decisions, but even if you have the poorest Chinese people anywhere in the world, they can still be quite rich, you know, like Singapore's rich or whatever.

• 3:28Speaker 01

  Yeah.

• 3:29Dwarkesh Patel?

  Also, there's parts of China which actually contain quite rich Chinese people.

• 3:33Speaker 01

  Oh, yeah.

• 3:33Dwarkesh Patel?

  So you had to compare not all of China against the US, but Shanghai and Guangdong against the United States. So you can have a part of China that is as big as America, and as wealthy as America, and as innovative as America.

• 3:44Speaker 01

  Like the Indian middle class is larger than the US middle class.

• 3:46Dwarkesh Patel?

  But also, it's not, it's not nowhere near as wealthy. Whereas there are parts of China which are humongous, which are actually as wealthy as the United States, and in many cases as innovative, et cetera.

• 3:56Speaker 01

  Yeah, no, I'm saying don't underestimate it, but at the same time, like, you know, we want to find the truth here, right?

• 4:01Dwarkesh Patel?

  Right.

• 4:01Speaker 01

  And the truth is like, we should not count the United States out of the battle and just give up. We're very much still in the race now, provided we don't, you know, take extra effort to shoot ourselves in the foot.

• 4:09Dwarkesh Patel?

  So right now, we are export controlling chips for the purpose of, we want to keep our AI lead, or stay in the lead in AI, and we recognize this is a key input in our ability to compete in AI.

• 4:23Speaker 01

  Yep.

• 4:23Dwarkesh Patel?

  So we are going to export control China's ability to have these chips.

• 4:26Speaker 01

  Mm-hmm.

• 4:27Dwarkesh Patel?

  Energy is also a key input in this AI race. And if China wanted to do the converse of what we're doing to them with these cheap imports, what they would do to us is to export control solar and batteries.

• 4:41Speaker 01

  It would be asymmetrical. It would hurt them worse than us.

• 4:43Dwarkesh Patel?

  If they did terrorists?

• 4:44Speaker 01

  Yeah, well, China obviously depends upon like the US export market for its economic dynamism.

• 4:48Dwarkesh Patel?

  Right.

• 4:48Speaker 01

  Right? And and the United States, it's going to hurt both parties to like sever that link. But if you sever the link completely, China's ability to make advanced chips right now is like basically not there, whereas the United States can make them. And United States' ability to make solar arrays is embryonic, but it's, you know, it's actually not that far behind China's, right? It's maybe five years behind.

• 5:08Dwarkesh Patel?

  If if we decided we want to produce 100 gigawatts of solar capacity every single year.

• 5:13Speaker 01

  We're already on track to do that.

• 5:14Dwarkesh Patel?

  Okay, is it going to be as cheap as it is to do in China?

• 5:18Speaker 01

  My my views on this are somewhat different from mainstream.

• 5:20Dwarkesh Patel?

  Okay.

• 5:21Speaker 01

  Which is great, because this is a podcast. So mainstream view would say, well, China has cheaper labor, which is no longer true, because they compare to Mexico, and it's got lower environmental regulations, which is true, and that it is more business friendly, which is absolutely crazy. I've never heard like, there's, there's no way you could justify that, like, your company having to have a, like, inspector from the CCP on its board who, like, harasses you about Xi Jinping every day, like, helps you do your business. And and also, like, the rule of law is not great. So like, you're constantly having to pay bribes to people in order to stay in business. Right? The idea that the United States cannot compete against that with, like, mostly or fully automated solar panel manufacturing in the United States, which has cheaper natural gas, by far abundant oil, abundant human resources, great financial capacity, you know, world-leading automation, et cetera, et cetera, is crazy. Like, we could literally copyplace solar manufacturing factories.

• 6:05Dwarkesh Patel?

  How much additional solar power capacity do you think we could be putting on that's manufactured in the US by 2028?

• 6:13Speaker 01

  This is a good question. When Russia invaded Ukraine, I thought, ah, finally, the Europeans will see sense. And they'll like, pull the trigger on like, we need to localize production of solar panels from dirt to the finished module, which is like a roughly a four-stage process. They didn't. So they're still paying Russia like a billion dollars a day for the privilege of being invaded. But at the time, I thought they could probably do that in about two years. And I think the United States could probably do that in two years or less, if you started today, like. It's currently eleven o'clock, so like we're gonna start cutting checks by noon. And yeah, I think you could you could ramp up pretty quickly. A lot of technology already exists here. It's not like it has to be reinvented from scratch. It's mostly a case of like putting in phone call to all the different manufacturers here in Germany and so on. They're saying, we need you to ten exercises at your factory, starting today, blank check, go.

• 6:59Dwarkesh Patel?

  Right, I guess a lot of your predictions seem to be not predictions, but more like if we had World War Two levels of motivation, if we had Manhattan level project level intensity around doing a specific thing, how fast could we do it? Which is like Elon, like if Elon was running government, how fast could it happen versus?

• 7:17Speaker 01

  It was for a brief period.

• 7:18Dwarkesh Patel?

  So maybe maybe then we should put it like if Elon ran the government like he ran SpaceX. As opposed to the question of like, okay, what is actually practically likely to happen, given that we are not treating it with?

• 7:28Speaker 01

  Okay.

• 7:28Dwarkesh Patel?

  World War Two level intensity.

• 7:29Speaker 01

  So if you look at XAI, which Elon is involved in, obviously, what are they actually focused on right now? Right, they're focused on the chips, right? Because they understand like the key bottleneck is the chips, not the solar power.

• 7:39Dwarkesh Patel?

  Right.

• 7:40Speaker 01

  Right, because even if Trump puts in a 200% tariff on Chinese solar, right? And we're not able to bypass it via like Vietnam or something, it's still a bargain. It doesn't matter. Like, if you if you need solar to run your data center, it's it doesn't hurt in terms of the overall cost picture. It doesn't matter at all.

• 7:55Dwarkesh Patel?

  Right.

• 7:55Speaker 01

  What matters is having the chips at like competitive capabilities per chip and enough of them, you know, installed in your PCBs, in your data centers, hooked up your liquid cooling, ready to go. Right, and that's actually something that that, you know, Elon and his company is great at. It's like figuring out this mass production, semi automated mass production. They've got this facility in Texas, which is making the Starlink receivers completely automated. But like, at what point does like, oh, we don't have a solar panel factory become on the critical path? Right, I very much doubt it's ever gonna be on the critical path. There's dozens and dozens of manufacturers of solar panels worldwide that are all competing against each other.

• 8:28Dwarkesh Patel?

  So you're a big solar bull. Right now, the hyperscalers are making decisions about with the data centers that they're building, they're gonna be one gigawatt, two gigawatts, in Metis case, five gigawatts, how they're going to be actually powered. And the people with actual money on the line are choosing natural gas. And it's not like they can't see the learning rate and the, I mean, they're building things which will be online in 28 or 30 or something. So why are they wrong and you're right?

• 8:55Speaker 01

  I mean, that's their job. They probably know more about it than I do. But no, in all seriousness, if you're like XAI right now, trying to build Colossus One data center in downtown Memphis, right? You want to get it done super fast. So you're like, what are all the different things we need? What are the factors of production to build this? We need a building. Well, we don't have time to build a building. We'll buy a building. Okay, we'll adapt it. We need power. We need thermal cooling. That stuff you can deliver on a truck. So that's what they did. Need access to gas. They had access to gas. They could tap into a local gas line. And actually, if you can tap into a gas line, generally speaking, you can get enough power, like the energy transmission capacity of like your regular gas delivery pipelines is way, way higher than electricity overhead lines, and it's easy to upgrade or whatever. And so if you're in this situation right now, you say, well, you know, are we constrained by our ability to go and rent gas turbines?

• 9:39Speaker 01

  And no, they're not, because there was enough available once, maybe twice, right? But at a certain point, you realize, well, as you grow, you start to touch all these additional constraints. And some of those constraints include gas availability. So there's a lot of chat about doing this in Pennsylvania, for example, where there's quite a lot of stranded gas in parts of Texas. But at the same time, the United States is gearing up in its ability to export natural gas overseas. So the price will not be infinitely low forever. You start to run into constraints around turbine manufacturing rate, around transformer, you know, transformer production rate, around grid capacity. And also kind of running into problems where the AIs and the humans who depend on like kind of legacy electricity production and delivery utilities are kind of competing with each other.

• 10:23Speaker 01

  We just saw this recent Ford auction in PJM result in

• 10:28Speaker 01

  in like kind of very high, like unsustainably high prices for consumers who depend on cheap electricity to like, you know, heat and cool their houses and have general prosperity. So if you kind of look far enough in the future, you say, well, you know, you can just turn up the dial arbitrarily high. Like you could say, well, we're going to put in a gigawatt a year. Well, we can, we can meet that constraint with gas turbines, right? We're not going to run out of natural gas at one gigawatt per year indefinitely. Okay, what if we're doing five gigawatts per year? What if we're doing 50 gigawatts per year? What if we're doing 100 gigawatts per year? Like, you can, you can just break the situation. Does that make sense?

• 11:00Dwarkesh Patel?

  Yeah.

• 11:01Speaker 01

  But not to reach prematurely for analogies, but, but like Henry Kaiser set up the Shipyard in Richmond, just down the road here in San Francisco, over near Berkeley. And initially making ships for the British and then by the end of the war, four separate shipyards operating in parallel to the point where like, he was bottlenecked on his supply of steel because steel was rare enough in the war because everyone was using it for different things. That Kaiser Industries went off and built not only a steel mill, but also a steel mine. Right. They went and started digging rocks out of the ground to turn into ships.

• 11:31Dwarkesh Patel?

  Yeah, yeah.

• 11:31Speaker 01

  Right. And it's the same sort of situation you have here where these these massive industrial verticals. And here I'm quite bullish on X A in particular because the Elon cinematic universe has just done so much industrial stuff compared to, the Googles or Metas of this world can kind of reach all the way through down into primary material supply if they need to.

• 11:49Dwarkesh Patel?

  The reason that these current plans are being done based on natural gas is that this is a sort of like.

• 11:55Speaker 01

  Well, PJM has all kinds of different sources of power, right? They have nuclear as well. They have gas, they have coal, all kinds of stuff. And actually, this this price here is is probably driven more by the delivery cost growth than by the generation cost growth, if that makes sense. So when you pay your utility bill, the cost is sometimes broken down by like a delivery cost and a generation cost, sometimes inflation costs and things. And so the delivery cost is the cost of, you know, basically what it costs the utility to build and maintain all the power lines that connect all the houses to all the power plants in some gigantic area divided by your marginal usage with all kinds of other complicated rules designed to make it fairer.

• 12:35Speaker 01

  And the problem that we see and the reason that PG&E here in California, for example, is perpetually on the brink of bankruptcy is that even though like the cost of additional solar panel or additional wind turbine or additional gas turbine or whatever is relatively cheap. Getting that power to your house is actually really expensive. Why? Because you've got, you know, generally like. Unionized labor that has to build and maintain power lines in areas that already have built up infrastructure. So you have like multiple collisions, whether this is like a power pole in your own street or like building a new transmission line, which requires you to like eminent domain land. So you're in court for like years and years and years, spending public money, litigating against other people who are also spending public money to litigate against you on behalf of other interest groups and so on and so

• 13:19Speaker 01

  forth. And then you've got wildfires and it's just, um, just like, it's like the poster child for Baumark cost disease. So one of the reasons that like, we're going to see large scale pruning of these grids is that we just can't afford under our current regulatory regime to maintain.

• 13:33Dwarkesh Patel?

  When you say pruning, are you just like, everything will just go off grid?

• 13:36Speaker 01

  Well, I mean, it's fairly clear to me that for really like large captive loads, like air data centers or aluminum refineries or whatever, you're going to have to build your own power plant for them, which is how it used to work. Like if you, if you had an aluminum plant back in the day, you would be building your own power plant for it as well.

• 13:53Dwarkesh Patel?

  It seems inefficient to have redundant power plants at every single industrial.

• 13:58Speaker 01

  Let me paint a grand vision for you. It would seem inefficient, but, but actually like, if you are sensitive to the cost of power expressed maybe in like supply elasticity or something like that, you just have to do it. There's no two ways about it. Like, is it inefficient for the XAI Colossus data center to have its own captive power plant, which it does, right? On the backs of a bunch of trucks in their parking lot? No, it's not inefficient. It's, it's the cheapest way for them to get power.

• 14:25Dwarkesh Patel?

  Okay. But a big picture question, across different kinds of ISOs, like from Texas to Pennsylvania to whatever, people are building data centers, which will not be online for many years, and they're choosing natural gas. What's going on?

• 14:40Speaker 01

  Well, I think they, you know, we haven't completely exhausted the supply of turbines relative to GPUs.

• 14:47Dwarkesh Patel?

  And do you have some estimate of when we'll run out of, because we can also make more turbines.

• 14:51Speaker 01

  Well, I think everything before about 2030 has spoken for at this point. Yeah, you could make more turbines. The funny thing is like, it's actually relatively expensive, I think, to like spool up additional production of these turbines, for example. So, so actually, here's one, here's one thing you have to grapple with sooner or later. Conventional power generation is a steam engine.

• 15:09Dwarkesh Patel?

  Right.

• 15:09Speaker 01

  Right. So you, you have some kind of chemical that you find inside the earth that is out of chemical equilibrium with the atmosphere and you burn it and it makes heat. Could be coal, could be gas, oil. You're me, they're true birds and bees here. Yeah, exactly. And it makes heat and you boil water and the water goes through some kind of mechanical contrivance that creates motion and then that motion twists the magnet and generates an electrical field, which then pushes electrons down wires, which then push electrons through a series of gates that then approximate thinking. It's kind of complicated. But the key step in this, which is converting heat into electricity in the most efficient, most common, it is the same if it's a nuclear plant or a gas plant, combined cycle plant or a coal plant or whatever, is what's called a Brayton cycle.

• 15:54Speaker 01

  A jet engine on an aircraft is a Brayton cycle as well. And just anytime you have a Brayton cycle with like a bunch of incandels spinning at high speed, it's just going to cost you like a bunch of money. So because it's inherently inefficient or what? It's just inherently expensive to build. Okay. What is the cost of, so like GE makes these 100 megawatt gas turbines, right? I guess, yeah. I don't actually know what the retail price is. I would suspect that if their price is flexible, it would have gone up a lot. But if I recall correctly, it's like 35 bucks a megawatt hour is just the flag for cost for. How much, right? 35 bucks a megawatt hour, just for, just for the Brayton cycle, right? So we're not talking about the fuel, we're not talking about the heat exchanges, we're not talking about, you know, the cooling ponds or anything like that.

• 16:36Speaker 01

  Just, just the amortized cost of the high speed, high temperature spinning components is 35 bucks a megawatt hour. Do you think the hyperscalers are being irrational or do they have some reason? Oh, to be clear, they don't care about cost for cost of power. This is, this seems extremely like, it's very counterintuitive. So for like Grandma Kettle in Pennsylvania, she's very sensitive to electricity cost and we don't really want her to like suffer in her retirement from unaffordable electricity costs and like having to sit there shivering. Like that's not the image that we want. At the same time, like what is the, what is the economic value to you of using, I don't know, Claude or Grok or whatever you use, like on a monthly basis? Yeah. But like, it's obviously much more than the subscription, but like, is it like 10 times more than the subscription maybe? Yeah. Okay, let's say, so let's say the subscription is on the order of 10 bucks, the value is on the order of 100 bucks, no, actually probably like 100 and 1,000.

• 17:21Speaker 01

  How much does it cost? XAI or Anthropic or whatever to serve your usage.

• 17:26Dwarkesh Patel?

  The marginal variable cost of like serving it, like in electricity is like less than 10% of the actual cost of.

• 17:31Speaker 01

  Well, their cost of serving it is is like maybe a buck per million tokens or something like that. And the cost of electricity is about 10% of that.

• 17:37Dwarkesh Patel?

  Yeah.

• 17:37Speaker 01

  So it's like 10 cents of electricity is generating $1,000 worth of economic value.

• 17:42Dwarkesh Patel?

  Right.

• 17:42Speaker 01

  Right. So it's very obvious that Anthropic could be like, oh, electricity cost basis is increased by a factor of 100.

• 17:47Dwarkesh Patel?

  Yeah.

• 17:48Speaker 01

  And now instead of paying 10 cents on your $100 bill for for power, you're paying 10 bucks.

• 17:52Dwarkesh Patel?

  Right.

• 17:52Speaker 01

  Right. So we're putting your your your subscription up to $110 for an electricity capacity charge.

• 17:56Dwarkesh Patel?

  Right.

• 17:57Speaker 01

  And then they could go out and buy turbines for prices that would make your eyes water.

• 18:00Dwarkesh Patel?

  Okay, so then why are we going to get the solar future? Like in 2032, we're going to have hundreds of gigawatts of extra demand for data centers. And at that point, most of it's coming from solar. And why is that?

• 18:14Speaker 01

  Well, you just, there aren't enough turbines being being manufactured.

• 18:17Dwarkesh Patel?

  Right. But also, I think in the early 2000s, if you, we can probably overlay the graph of like how many turbines are being manufactured. Like, right now, we're at historical.

• 18:26Speaker 01

  They've ramped up basically to like the early 2000s rate again.

• 18:28Dwarkesh Patel?

  But I don't know, you had to make more solar panels as well, right? Like, there will be supply elasticities for both solar and natural gases. Is there some reason to think that it's worse for the supply chain involved than having a natural gas powered data center than a solar?

• 18:42Speaker 01

  Yeah, I do. The learning rate for for natural gas is nowhere near as steep as solar, which just tells you that it's easy to make solar panels, much easier to make solar panels. There are actually very few manufactured products which are easier to make. Like, the learning, the learning, the the right solar coefficient is 43%. So every time we double commute production, we get a 43% reduction in cost.

• 18:59Dwarkesh Patel?

  And what is the basis of, like, why are we finding 43% worth of things that can be made cheaper or more efficient every single year?

• 19:07Speaker 01

  Well, I mean, roughly speaking, there's like 10,000 manufacturing process engineers working on this full time.

• 19:12Dwarkesh Patel?

  Right. That could be true of any process, right? But no other process sees the kinds of learning rates.

• 19:17Speaker 01

  That's completely true.

• 19:18Dwarkesh Patel?

  That's all they're saying.

• 19:19Speaker 01

  Yeah, so in order to like sustain this over a long period of time, you obviously need to have like a demand elasticity that exceeds like your learning rate.

• 19:25Dwarkesh Patel?

  Yeah.

• 19:26Speaker 01

  So so otherwise you would, after, you know, a couple of booms, you would you would saturate your market at the current price and you'd have no additional growth. But in this case, like, you know, every roughly every two years, we're doubling production, two, two and a half years doubling production, the price is coming down by a factor of forty, forty or percent. So like roughly twenty percent, fifteen, twenty percent per year. And then just as a result of that price reduction, demand skyrockets by like probably six times more than that additional marginal capacity, like production capacity increase.

• 19:51Dwarkesh Patel?

  Right.

• 19:52Speaker 01

  And this is one point where I'll say actually the pros, so-called pros are definitely wrong. Conventional wisdom is like, oh, solar demand is going to saturate like this this week. It's gonna, it's gonna saturate. We've got a graph here somewhere that's like, this year is it, it's never gonna grow anymore. And instead it's just like blasting out the top of the graph. And this conventional wisdom is wrong. It is not only the case that solar adoption, production, price decreases are continuing, they're accelerating, and the rate that they're accelerating is still accelerating. Right?

• 20:17Dwarkesh Patel?

  Wait, sorry, the rate that is accelerating is accelerating?

• 20:20Speaker 01

  Yes.

• 20:21Dwarkesh Patel?

  As measured in the total fraction of energy that's coming from solar or?

• 20:25Speaker 01

  In the sense that its fitness for the markets that it is being produced for is increasing over time. So it is, it's still extremely early. Right?

• 20:34Speaker 01

  We're still like the Apple two computer era of solar.

• 20:37Dwarkesh Patel?

  Backing up, if the story is that the reason solar is getting cheaper is because there's a lot of demand for more solar and that demand can sustain economies of scale or whatever is going on.

• 20:49Speaker 01

  Yes. I'm gonna go ad lib here and like agree with Elon Musk on this.

• 20:53Dwarkesh Patel?

  Then shouldn't that also be true of gas turbines and transformers and power stations and whatever else that's required for the the non-solar future? Cause like even there, we're expecting AI to drive up demand for power regardless of the source. So

• 21:08Speaker 01

  Yes.

• 21:08Dwarkesh Patel?

  The extent of story for solar becoming cheaper over time is just that like, well, demand will go up and that will drive efficiencies. Why isn't that true for?

• 21:16Speaker 01

  Well, let's say you're a bank and you're trying to decide whether to lend, I don't know, like gee, a bunch of money to like expand production of of their gas turbines.

• 21:23Dwarkesh Patel?

  Yeah.

• 21:23Speaker 01

  You can, you can write them the check today. And then they'll start scaling up their factories. And then they'll, you know, they'll start to see the benefits out in three or four or five years.

• 21:31Dwarkesh Patel?

  Right.

• 21:32Speaker 01

  You don't know if the AI bubble will have burst by then. You don't know if China will have invaded Taiwan by then. You don't know if Siemens or Philips or someone will have outcompeted you. Right? You don't know if G is like major looming structural problems will cause it to be unable to compete as it has in the past. You, you also don't know, like, in order to make that money back, you have to then operate that plant at that capacity for 20 years. Right? And if I was looking at the same charts as they're looking at right now, I'd say, what are the odds that in 25 years' time? We can produce gas turbines at a price that is relevant in a world where solar is already at its current price.

• 22:06Dwarkesh Patel?

  Right.

• 22:06Speaker 01

  And batteries are at a price where they're already like, you cannot win.

• 22:09Dwarkesh Patel?

  I feel like there was actually a similar discussion a couple of years back or a year back. When AI people were like, no, AI is real. This is going to happen. And then SK Hynix, Samsung, et cetera, were like, we're not ramping up HBM production because like, oh, HBM is used largely for AI workloads. And if this demand doesn't continue, then our manufacturing, additional manufacturing capacity for HBM will not have been worth it. And then there was another bottleneck with co-OS.

• 22:33Speaker 01

  What happened after that? Did they end up indeed ramping up their production?

• 22:37Dwarkesh Patel?

  I think so.

• 22:38Speaker 01

  Well, so when someone says, we can't do it, we won't do it, no way, no how.

• 22:41Dwarkesh Patel?

  I mean, not in the US.

• 22:41Speaker 01

  What they're saying is, write me a check.

• 22:43Dwarkesh Patel?

  Right.

• 22:44Speaker 01

  Right? And they did. And now Samsung's coming on board in the States to build T6 with XAI, I think. So like, they all got there in the end.

• 22:51Dwarkesh Patel?

  Maybe, maybe it's worth going into the numbers, right? So right now, 43% of US data center power consumption is from natural gas. And basically you think asymptotically that will be like 100% solar. If you go to like 2040.

• 23:05Speaker 01

  Yeah, so I mean, obviously like legacy production coal and stuff's going to retire over time.

• 23:09Dwarkesh Patel?

  Right.

• 23:10Speaker 01

  And, and if a gas plant is still making money, people will keep operating it. But, but at certain point, like it is the case right now that operating a coal plant costs more than building a new solar plant. So it's just cheaper to.

• 23:19Dwarkesh Patel?

  So yeah, I want to, I want to know what the.

• 23:21Speaker 01

  And then obviously capacity is going to increase a lot, so that helps to dilute the existing.

• 23:25Dwarkesh Patel?

  Yeah.

• 23:25Speaker 01

  Production.

• 23:26Dwarkesh Patel?

  And, and also the, um, the amount of use is like going to increase a bunch, right? Like the, the amount of like data center use of energy will just be.

• 23:34Speaker 01

  Yeah, for sure.

• 23:35Dwarkesh Patel?

  Exponentially higher, so.

• 23:36Speaker 01

  Yeah.

• 23:36Dwarkesh Patel?

  The, the new stock matters a lot as compared to the existing stocks. Anyway, so I, I want to know 2027, what fraction is natural gas? 2030, what fraction is, uh, natural gas versus solar?

• 23:46Speaker 01

  For, for new load or, or for, uh, for the.

• 23:48Dwarkesh Patel?

  Let's say new load.

• 23:49Speaker 01

  For new load.

• 23:50Dwarkesh Patel?

  2035, etc. Basically, like, okay, if eventually you're right that we'll pave the earth in solar panels to sustain our quadrillions of AI souls, what is the pace of that?

• 24:00Speaker 01

  Well, I think the question to ask is like, what is the major constraint on that ramp up?

• 24:04Dwarkesh Patel?

  Right.

• 24:04Speaker 01

  Right? And then everything else is just drafted in behind. And I suspect that the hardest thing to make will always be the silicon, like the GPUs. Right? So the question is really, how quickly does TSMC ramp up its production of GPUs? And that's a question for you, not for me.

• 24:20Dwarkesh Patel?

  I'll just use some numbers that AI 2027 used for their compute forecast, which even if you don't buy their singularity thing, I think they did a reasonably good job with crunching the numbers on their compute forecast. And I think they said there's on the order of 10 million H100 equivalents in the world today. And I think they said by 2028, there'd be 100 million. So basically 10x more H100 equivalents in the world.

• 24:43Speaker 01

  About a kilowatt H, something like that?

• 24:45Dwarkesh Patel?

  Yeah.

• 24:45Speaker 01

  Yeah.

• 24:45Dwarkesh Patel?

  Yeah.

• 24:46Speaker 01

  Okay, so it's like 100 gigawatts.

• 24:47Dwarkesh Patel?

  Right, okay.

• 24:48Speaker 01

  And that's, I mean, that sounds roughly right. You're not the first person to give me a call and ask me about this, I'll put it that way. I'm not going to name names, but like, pretty much all the names you've heard of have given me a call and said like, we know that you're a minority voice on the paper that came out recently with Scale Microgrids, talking about how you could do 90% solar, 10% gas. And I said, you can go all the way 100% solar, I wrote a blog post about it. And so they always called me up and said, what about this? And they're all talking like 5 gigawatts in the next few years. So that's just like 90 plus percent solar for just those. So I think within a few years, we'll probably see that like the majority of new DCs that are going in will be mostly solar.

• 25:21Dwarkesh Patel?

  Within, sorry, how long?

• 25:23Speaker 01

  Let's say by, let's see, what's it, 2027, majority of new DCs going in at that point would be mostly solar.

• 25:29Dwarkesh Patel?

  As in like new?

• 25:30Speaker 01

  As in groundbreaking at that point.

• 25:31Dwarkesh Patel?

  But if you're groundbreaking in 27, you're probably like planning it now, right?

• 25:35Speaker 01

  Oh, that's why they're calling me.

• 25:39Speaker 01

  My consulting fees are extremely affordable.

• 25:43Speaker 01

  But I really, I don't have deep visibility because I'm not in the same room with like, with the meta people as to like when we're going to hit the wall on transformers, when we're going to hit the wall on like, just how much like municipal like peak load can we shave off, which is the latest thing that's been doing around. So which it turns out like, there's a handful of places in the United States, and by handful, I mean like literally handful, where like there might have used to be an aluminum smelt, there's a bunch of like latent capacity in the grid. And there's also a bunch of generators on the grid that are notionally turned down and they're like operators, a 40, 50% capacity factor. But they max out at about eighty percent capacity factor because you've got to bring them down for maintenance pretty often, right? Especially if they're old. And so they're saying, well, you know, we could pay you just to like operate this old coal plant or something at high capacity. It'll go down this power line to this place where the smelter used to be.

• 26:27Speaker 01

  We'll set up there and then we promise to like curtail when you need the power.

• 26:30Dwarkesh Patel?

  Right.

• 26:31Speaker 01

  Which basically means they just have like a massive captive battery plant as well, which is fine. You just buy that until it arrives on a truck. The major advantage of doing that over the pure solar play is that the power is already there, so there's no risk there. There's no, and then you don't need a massive amount of land. Like, the problem with the solar approach is that there's no two ways about it. It's just, it's a farming operation. You need a huge amount of land.

• 26:52Dwarkesh Patel?

  Yeah.

• 26:52Speaker 01

  Right? The total amount of land that you're using, less than one percent is under batteries, under roads, under data center structures, et cetera, et cetera. It's mostly solar.

• 27:00Dwarkesh Patel?

  Right, okay. So let's get into what this, if you've got a five gigawatt plant you want to build.

• 27:07Speaker 01

  Yes.

• 27:08Dwarkesh Patel?

  Break down the numbers for me in how much land in terms of solar you need to farm this out. And especially, I was talking to somebody in this space and they said, look, the big problem is not, obviously, cost for, the cost of energy for these data centers is a small fraction of the total cost. Most of the cost is going towards chips. So then the issue is just, can you, can you make the energy available? And they were saying, even those solar panels themselves, you can acquire. The issue is getting that much contiguous land and getting the permitting to interconnect it or whatever the word is, is like apparently a big hassle.

• 27:45Speaker 01

  Yep. It's kind of a nightmare.

• 27:47Dwarkesh Patel?

  Yeah. And so they're like, well, at that point, is it actually easier than just getting on the grid or, but yeah, what's your, like, if you need tens of thousands of acres of solar, where can you do that without, and get like?

• 27:58Speaker 01

  Basically in Texas. I mean, like, there's this very popular misconception that like, there's not enough land to do solar.

• 28:04Dwarkesh Patel?

  Right.

• 28:05Speaker 01

  Right. This is garbage. If you've ever flown in an aircraft in the United States and you've ever looked out the window, you'd be like, oh, wow, look, there's a lot of land you could put solar on, especially west of like one hundred and ten.

• 28:15Dwarkesh Patel?

  Yeah.

• 28:15Speaker 01

  Right?

• 28:15Dwarkesh Patel?

  Does it need to be flat or no?

• 28:16Speaker 01

  No.

• 28:17Dwarkesh Patel?

  Okay.

• 28:18Speaker 01

  Doesn't matter.

• 28:20Speaker 01

  Like, do trees grow on mountain slopes? So it doesn't matter.

• 28:24Dwarkesh Patel?

  Right.

• 28:25Speaker 01

  And the total amount of land, so just for reference, Nevada is something like 80 million acres or something like that. So just Nevada, which is like 90% federal land.

• 28:31Dwarkesh Patel?

  Yeah.

• 28:31Speaker 01

  Is 80 million acres. And I would, I would never say that we should, we should sacrifice Nevada to the AI and, and pave the entirety of Nevada from one wall to the other. But I just saw a bunch of things in my feed the last couple of days that like, you know, Vegas is, is falling apart and.

• 28:46Dwarkesh Patel?

  Really?

• 28:47Speaker 01

  Well, like no one, the boomers are retiring, no one goes there anymore. People would go to see like the 100 million acres of solar. Yeah, okay. So five gigawatt plant, so.

• 28:55Dwarkesh Patel?

  So even if you did it in Nevada?

• 28:56Speaker 01

  We can do it anywhere. You can do it anywhere you can find the land.

• 28:58Dwarkesh Patel?

  Yeah.

• 28:59Speaker 01

  Right? People say, well, you can't do this in Europe because Europe doesn't have solar power. Europe has solar power. I've been to Europe in summer. It's like sunny for like 20 hours of the day.

• 29:05Dwarkesh Patel?

  Right.

• 29:06Speaker 01

  It's a bit seasonal, right? But that's not a big deal.

• 29:08Dwarkesh Patel?

  Wait, I mean, it is if, because you, because energy is a small fraction of the cost, you care more about making sure the ships are running all the time, right?

• 29:14Speaker 01

  Yeah, so in practice, what happens is, let's say Europe hypothetically like awakens from its slumber and decides it wants to participate in AI. Okay, I hope it does. They say, well, we're gonna have to put, you know, 100 gigawatts down of solar at some point to build these data centers and most likely be in Southern Europe. Spain is not particularly heavily populated. That's a great place to start. So we put in 100 gigawatts of solar, solar data center in Spain. And then in order to achieve, you know, basically, if you're spending like AI hyperscaling money on your GPUs, you want to have like four nines uptime in order to maximize your like tokens per dollar spent.

• 29:49Dwarkesh Patel?

  Yep.

• 29:49Speaker 01

  Okay, tokens per dollar spent on the entire project, not just on that.

• 29:54Speaker 01

  This is a very subtle point. I can go into vast detail on it later on, maybe. But let's just say you need four nines uptime. In order to achieve four nines uptime in like the middle of winter, you need to have a lot of solar overbuilt, right? Is solar overbuilt a bad thing? No. Is the fact that we produce 40% more food than we need a bad thing? No. It's much better than producing 40% less than we need. Okay? And it just means that effectively you have a giant captive power plant attached to a data center that. Ninety nine point nine percent of the time produces more power than it needs. And ninety nine percent of the time produces much more power than it needs. And that can now actually be the source of power that the local utility, instead of being like naughty naughty data center, you must disconnect. When we tell you to, they say, hey, data center, I notice you've got like a bunch of power you're not using. Three hundred and sixty days of the year.

• 30:39Speaker 01

  Would you mind ever so much if we threw a power cable over the wall and we powered our entire town? Off your spare power at like essentially zero marginal cost. Plus, you know, whatever residential batteries that we need in addition to local power supply.

• 30:51Dwarkesh Patel?

  Brian Potter had a good analogy in his blog post about this, where he's like, I don't know, my MacBook has a terabyte of storage. Hmm. And I use like a hundred gigabytes. And I just got the terabyte version because it's cheap enough and I might need it at some point that it's worth it. And so you're saying solar gets gets so cheap that it's it's the way we'll treat hard drive space. There's like get a bunch of excess.

• 31:12Speaker 01

  Yeah, but also like the the market will be made. At the the place like the the new the new marginal like consumption and production. Right? So like all the people who are working in the space right now are like. Oh, I'm in I'm in the business of delivering power or storing power. I'm going to serve the AI market because that's where all the growth is occurring.

• 31:32Dwarkesh Patel?

  Right, right.

• 31:33Speaker 01

  That's where all of US GDP growth is occurring right now.

• 31:35Dwarkesh Patel?

  So I guess you didn't answer the question of, yes, theoretically, we could do this. But is it going to be possible to get the permitting to have tens of thousands of acres of contiguous?

• 31:47Speaker 01

  It doesn't need to be contiguous. Well, I mean, it helps if it's contiguous, but doesn't if it's convex. Um, so this is, you know, you can have a bit over here and a bit over there and you can wire them together relatively easily. In fact, in the limit, in the limit, you have fields upon fields of solar arrays with.

• 32:01Dwarkesh Patel?

  Tell me your dream case.

• 32:02Speaker 01

  Okay, fields, just just solar arrays as far as I can see. And then within the solar arrays, roughly in the middle of them, you have your your batteries and your.

• 32:11Dwarkesh Patel?

  I play factorial, I remember this.

• 32:12Speaker 01

  Okay.

• 32:13Dwarkesh Patel?

  I remember this optimal layout of like batteries and.

• 32:16Speaker 01

  Yeah, your batteries and you've got your data centers. So in terms of ground, like floor area, you know, it's roughly, let's say, ten percent racks, ten percent access to the racks. Um, maybe like fifty percent batteries stacked up on top of each other, and there's just cooling, something like that in terms of what what sits in the in the like the centralized node. Right? And that could be a hundred megawatts or it could be ten gigawatts, depending on how you want to scale this. But then all you need to connect that to the outside world is like an optical fiber. Right, an optical fiber cable, which you can string up on poles, you can run underground, you could even use microwave links if you really wanted to. You could use Starlink if you really wanted to. I don't know what the Starlink would be fast enough. I'm not sure if it's like, capacity is high enough. You could use laser links if you really needed to. And that's, that's it.

• 33:00Speaker 01

  It's like this completely self-contained world of like-

• 33:04Dwarkesh Patel?

  Because it's off-grid.

• 33:05Speaker 01

  Yeah, of computation that occurs off-grid, like on private land, somewhere in the backwoods of Texas, where like, no one lives and no one will ever live, because it's completely inhospitable to humans.

• 33:12Dwarkesh Patel?

  In terms of the ratios, it's one trend that was impressed upon me is that the power density of racks is increasing a lot.

• 33:20Speaker 01

  Yes.

• 33:20Dwarkesh Patel?

  As the flops for GPU are increasing, so we've-

• 33:23Speaker 01

  It's like a megawatt per rack is what they're heading to now, which just seems bananas to me.

• 33:27Dwarkesh Patel?

  I think it was even more than that, but yeah, yeah.

• 33:32Speaker 01

  So let's, let's get concrete here for a second. You got, let's say you got one rack and it's one megawatt. And I'll leave the calling to like, someone who specializes in air conditioners, but it's, it's basically throw air conditioners at the problem. And then you have batteries, so in order to get four nines of uptime on this, you need, in, in South Texas, you're actually in less than this, but let's just say 24 hours worth of, of battery storage. Right? Because that means it'll get you through two of the, two, two bad nights in a row, basically. And, and actually, it turns out that you can significantly decrease power consumption with a very small reduction in overall compute. So if you, if you've got like three really bad days in a row or something, you can actually just like, you can dial back your power usage quite a lot without compromising your inference or, or, or training. Okay, so you've got, say, a Tesla Powerwall, something like four megawatt hours.

• 34:14Speaker 01

  So one, one megawatt rack and then six Tesla mega packs, each of which is roughly one truckload worth of stuff. So like one truckload worth of rack and then like six truckloads worth of batteries. And then in order to operate this at an average power of, of one megawatt, your solar arrays in Texas would be something like 25% utilization. So on average, you know, if, if the sun came up every day and the day was the same length all the time, you would need four megawatts of solar arrays, which is about four acres of land. But in practice, because you're aiming for four nines instead of like one nine, you need an overbuild of about two and a half X. So you've got about 10 acres of solar, so 10 acres of solar, six truckloads of batteries, one truckload of data center and some cooling stuff.

• 34:56Dwarkesh Patel?

  For how big of a

• 34:59Speaker 01

  One megawatt, that's just one megawatt, right? So 10 acres, one megawatt kind of situation at four nines. Yeah, so if you want five gigawatts, and that's 5,000 times 10, so 50,000 acres. And actually, at larger scale, you can probably cut all those numbers down by 10, 20%, but like on that order. And like 50,000 acres sounds like a lot.

• 35:17Dwarkesh Patel?

  It does sound like a lot.

• 35:19Speaker 01

  The amount of land, the amount of land put aside for Oak Ridge was 100,000 acres. The amount of land put aside for Hanford was about 100,000 acres.

• 35:25Dwarkesh Patel?

  With Hanford?

• 35:26Speaker 01

  Hanford was where they made plutonium in the Manhattan Project.

• 35:28Dwarkesh Patel?

  Okay, I don't know how big that was.

• 35:31Speaker 01

  About 100,000 acres.

• 35:32Dwarkesh Patel?

  Okay, but I mean, isn't like, was it like, oh, this is so small? And then you're like, oh, but it's 100,000 acres or?

• 35:37Speaker 01

  Well, so the reason they, and it's still largely unpopulated now because it's the National Laboratory. But the reason they did that was they thought, oh, we're going to need four piles to produce plutonium. These are not, these are not nuclear reactors that produce excess thermal energy. So you can't actually make nuclear power with them, but you're making plutonium with them. Then in the end, they needed two, I think. And they wanted them spaced out because they thought like they might just like spontaneously like explode. And a bunch of other facilities and plants and stuff as well.

• 36:00Dwarkesh Patel?

  Austin Vernon had an interesting blog post where he said that if you allow for, if you have like diesel generators or something, which can take over the generation for like 10% of the generation during these, you know, with the winter or something, then you can have 60% reduction in the amount of solar panels you need to install because you don't need to have, you don't need to plan for that contingency.

• 36:25Speaker 01

  Yeah, yeah, so basically there's a balance here. So this is not a very complicated optimization problem. Like for people who do optimization problems for fun, this is how you do it. You start off with a bunch of like NREL data on like what your solar abundance is in this particular part of the world. And then you just start throwing solar panels and batteries at it over like the course of a one year simulation until you hit the number of nines you want, right? And then you, to an extent, you can trade the amount of panels and the amount of batteries you've got back and forth. And there's like a very, very broad optimum, right? Or you can throw in a third thing like a diesel backup or a gas turbine or whatever.

• 36:56Dwarkesh Patel?

  The issue here is Look, if Meta or Microsoft or whoever just wants to get something off the ground, this might be low OpEx to have this huge solar farm, but high CapEx, where you need to hire like 30,000 people to go in the middle of a desert and install 50,000 acres worth of solar panels. And they're like, why would I not just buy like 50 gas turbines instead?

• 37:23Speaker 01

  Why don't just outbid, like Microsoft or Meta outbids Google or something for the last gas turbine that's available that year? Yeah, I mean, totally. The thing that I think Meta has realized is like, Zuck is running out of time to spend his money to win. The CapEx is not crazy high, just to be clear. Like, the CapEx is still dominated by just the GPUs.

• 37:40Dwarkesh Patel?

  Right.

• 37:40Speaker 01

  Right? So like, how much does 5 gigawatts worth of GPUs cost?

• 37:44Dwarkesh Patel?

  I don't know if my numbers will be wrong, but like 250 billion or something?

• 37:47Speaker 01

  Yeah, 250 billion. Sounds about right.

• 37:49Dwarkesh Patel?

  Yeah.

• 37:49Speaker 01

  Okay. So like, is 50,000 acres going to cost 250 billion dollars in Texas?

• 37:55Dwarkesh Patel?

  That's so much money. Wait, I didn't, I did the math in my head. I'm like, that's a lot of money.

• 38:00Speaker 01

  Like, maybe hundreds of millions of dollars, something like that. So it's like, literally, 0.1% of the cost is land, right? How much does a megawatt of solar cost? Well, if you go and ask the usual suspects, they'll tell you million dollars. But this is one of the things that breaks my brain at Terraform, which is my day job, which is that the modules themselves, without tariffs, would be like 8 cents a watt. So that's like $80,000.

• 38:20Dwarkesh Patel?

  8 cents a watt?

• 38:21Speaker 01

  Yeah.

• 38:22Dwarkesh Patel?

  But they're like a dollar a watt, including installation and everything?

• 38:26Speaker 01

  Including installation and everything. But like, the panels are the magic part. They're the thing that turns sunlight into pure electrical energy at 25% efficiency. Everything else should be like less than that. That's what we're, if you want to work on that project, come and work with us at Terraform, because we're very cost sensitive.

• 38:42Dwarkesh Patel?

  We'll give you an opportunity to show, don't worry.

• 38:44Speaker 01

  Constantly. But, no, in all seriousness, like, the central takeaway is that, is that the hyperscales are not power cost sensitive, they are power availability sensitive. And for all these things, you just run into this like supply elasticity wall at the sort of rates of increase that we're talking about. And solar is by far the best option for like... Firehosing energy at a given problem. Right. Because it rains down from the sky.

• 39:14Dwarkesh Patel?

  If you're running a frontier technology company, you know how essential it is to recruit the world's best talent. But this requires navigating the Byzantine US immigration system. Not only do you not have the time to deal with this yourself, you just don't have the tacit knowledge to maximize the probability of success. But given how critical exceptional talent is, you can't take any risks with visa approval. You need to work with the best. Lighthouse handles everything better and faster than you possibly could, and they do it all with minimal input from you. Lighthouse knows the nooks and crannies of the immigration machine. They know how to frame publications and awards, or how to structure comp comparisons versus benchmarks,

• 39:55Dwarkesh Patel?

  and how to put everything together into the most compelling story possible. And they've even optimized the tiny details, like the tone they use when they draft support letters to help US immigration officials understand the importance of tech and startups. Companies like Cursor, Together AI, and Physical Intelligence are all already working with Lighthouse. You can join them by visiting lighthousehq.com/employers. All right, back to Casey. Between the fact that the, um, maybe solar prices will go down and the fact that demand is going to go up, do you think electricity prices are likely to rise?

• 40:32Speaker 01

  Yes, but electricity prices at this point are a reflection of a regulatory irrationality, right? And it's the same situation in Europe, and Australia if that matter. Like, your prices will rise until you've had enough, and you say, no, we demand that you allow us to take advantage of power technology that's been invented in the last 50 years. In terms of things that are causing us to lose to China, tariffs are neither here nor there because, as we've discussed, we're not sensitive to cost on power. But the environmental regulations that are actively preventing us from deploying renewable energy.

• 41:07Speaker 01

  In the United States, like this is the reason Texas is winning. Texas is out deploying California 10 to 1. The regulatory environment around solar is just insane. It's insane. The regulatory environment. Okay. So in the United States, part of the reason that solar has not been deployed at massive scale yet is that a bunch of laws went into action in the early 1970s that were intended to protect our environment. And that makes a lot of sense. And our environment's a great thing we should protect.

• 41:30Dwarkesh Patel?

  But how, I mean, I think people will be familiar with NEPA and whatever. But like, how is it especially impacting solar?

• 41:34Speaker 01

  Well, let's say you've got a bunch of private land out in the middle of nowhere, right? And you want to build it, solar on it. You'll probably end up triggering NEPA, right? At which point you now have to do what is not in the law, but considered necessary under current regulations, which is like your four-year environmental impact review, which generates like so much paper that just the environmental impact of producing the report, because you have to cut down trees to make paper, is more than the environmental impact of just deploying the solar. Like, this is bonkers. It is crazy town, right? The thing that drives me particularly crazy in Southern California is that just because solar is kind of new and off-grid solar is very new, unless you're very, very careful, you end up getting regulated as though you're trying to build a chemical plant, even though it's a solar array. And the impact of solar around desert is arguably positive because it shades the ground

• 42:19Speaker 01

  and improves like soil moisture retention.

• 42:23Speaker 01

  They're like, if you wanted to reverse desertification, you would basically just deploy solar panels on it. And that would pay for the process.

• 42:29Speaker 01

  But you, you end up having to go through more stringent environmental review process than if you just wanted to like grade the whole thing and cover it in concrete. Or grade it and then park a bunch of like old rusting cars that are dropping oil into the aquifer.

• 42:40Dwarkesh Patel?

  Right.

• 42:41Speaker 01

  Which in many cases, you don't need a permit for at all. Right? But to build the solar, you have to go through this whole process. And, and like, if there's one thing that anyone listening to this can do, it would be like, have a categorical exemption for solar deployment. Or, or just like, if I put money in escrow account that says like, if after 20 years we have to pull this out, like, we'll pull the solar out of the desert and it goes back to being desert. Right? I will do that in a heartbeat. But like, if I have to hire another biologist $10,000 to be like, well, on that 40-acre plot, we found a tuft of grass, which we believe might be a, a critical, you know, one of the 20 species that this particular species of bee sometimes. Eats and the species of bee is not technically endangered, but it might be at some point in the future. Therefore, you can't deploy there, even though it's like zoned unrestricted industrial

• 43:25Speaker 01

  and it's sandwiched between a rocket test stand and like a chemical plant, for example, in an industrial part of the desert. I will like, I'm going to become the Joker. It is insane. It is like, I just think we need to be a bit balanced about this. It's like, I don't want to like drive species into extinction. But like the meta problem here is if we don't move our industrial stack off fossil fuels in 10 or 20 years. First of all, we'll get poor the same way you UK did, right? Because they ran out of coal, basically. And the second thing is we'll get poor because we'll flood our coastal cities in Florida underneath climate change.

• 43:59Speaker 01

  We need, we need solar synthetics for that part. We also need to do sulfur injection and a couple of things.

• 44:04Dwarkesh Patel?

  People will point out that transmission line growth has been stuck in a rut for decades and we have all these bottlenecks instead of in terms of substations and transformers, et cetera, et cetera. Why will this not hamper this abundant solar future?

• 44:21Speaker 01

  That's a really great question. So actually, you and I had a conversation along these lines almost two years ago when we first met and, and it caused me to go and write a blog post. So this is, this is a good way of thinking about it.

• 44:34Dwarkesh Patel?

  I think there's another blog post you wrote, though, which was also to our conversation we had, which is how to feed the AIs.

• 44:38Speaker 01

  That's much more recent. Yeah, that was after dinner, I think.

• 44:42Speaker 01

  And to be fair, like, I usually am fairly clear in my blog posts if I'm posting or if I'm serious. But, but this one, actually, I'm dead serious on. It's actually the one where I'm, it's like the most out of the money bet as well. Like, everyone else that I consider to be like a respectable forecaster in this area disagrees with me on it. So that's one side. Well, so the grid, we know why the grid is expensive. It's a lot of wires strung up in like hard to reach places that are hard to maintain, especially as, you know, the workforce ages and, and regulations and all the rest and eminent domain and so on and so forth. Okay. So grid's not going to get cheaper anytime soon or easier to build. And if you look at the projections of like how much grid, you know, Dewey would have us needing to build in the next 10 years versus how much is actually being built. It's like, it's not even in the same order of magnitude. So you say, well,

• 45:25Speaker 01

  are we totally screwed? The answer is no, we're not totally screwed because batteries actually do the same job that the grid does. But this is kind of weird, seeming out. The grid transports power from one place to another. It transports almost instantaneously at the speed of light. So it's actually performing a spatial arbitrage. Right. Right. So the idea being that right outside the local nuclear power plant, power is really cheap because they make a lot of it. And in your house, power is really expensive because you don't have a power plant in your house. And you you pay the intermediary a small fee and they allow this this trade to take place. And that's basically how the grid works. And until quite recently, the only way we had of meaningfully storing energy, like storing electricity on the grid was pumped hydro. And that only works in a handful of places in limited, limited capacity. And it doesn't work all that well either.

• 46:10Speaker 01

  The efficiency is not great. Now we have batteries. Batteries store power at one time of day and they release it at another time of day. So batteries are performing a temporal arbitrage, an arbitrage over time. But they can be local or they can be more remote. I think we'll end up seeing batteries next to the solar arrays and batteries in the middle of the grid, substations and batteries on the sites of existing power plants that get turned off and batteries in your house and batteries everywhere in between. One way of thinking of this is what is your per capita allocation of batteries in kilograms per head? And like when you and I were much younger, you know, the lithium ion battery was in a cellphone, say. So you got like 10 grams per person or something. And nowadays, half the people in this town drive Teslas.

• 46:54Speaker 01

  So your per capita allocation of lithium ion battery is 100 kilograms or something like that. So we're talking like four or five ooms of increase of total, total battery per person. That trend is only going to continue. And then you say, well, okay, we've got batteries that are performing this temporal arbitrage and the sun comes up every day. Right? So like the, the, the power swing from like midday, you're otherwise curtailing the solar array to dusk when everyone's watching TV and cooking dinner or running the air conditioners to cool off in the evening is very predictable. Right? Whereas like, oh, we had like really bad weather. So we had to use the power line that runs to the power, extra power plants over by Hoover Dam or something. Doesn't get used nearly as much or like its, its peak utilization is almost never. Which means that the utilization of the batteries is on average, like let's say 300 days a year.

• 47:39Speaker 01

  And the utilization of your most expensive grid, most highest voltage grid assets is much, much lower. And that includes like the, the substations and transformers and stuff that serve that. So it's really bad position to be in if you're a grid operator. Because you've got this aging existing thing that the batteries are cannibalizing, the batteries are being installed behind the meter. You don't have a say in whether they're being installed, how they're being used. All you know is that your utilization of your asset where you get to charge top dollar for it is just dropping year after year after year, as the same time as your operating costs are increasing year after year after year. So it's just very clear that like the average distance the electron is going to travel between generation and consumption is going to decrease in the future pretty radically.

• 48:23Dwarkesh Patel?

  Yeah, I guess the.

• 48:24Speaker 01

  It's already decreasing, it's going to continue to decrease.

• 48:26Dwarkesh Patel?

  The difference is that, I mean, it's especially helpful for solar, but like solar is the one that's most intermittent.

• 48:33Speaker 01

  You can predict the amount of solar power you're going to get in three days pretty accurately because of weather prediction.

• 48:38Dwarkesh Patel?

  But you can't like buy more batteries, you can't like change the amount of batteries you have.

• 48:41Speaker 01

  Well, actually, in the limit you can, because you can put them on trucks and drive them around. Right, so there could be a capacity market for batteries where you drive them around to people who need them in a pinch. In practice, it's going to be cheaper just to double the size of your battery. Because batteries can continue to get cheaper and cheaper and cheaper and cheaper. But what it does mean is you can say, well, I know that I'm going to have, you know, three low days, so I will start curtailing now by 5%, so I don't have to curtail by 50% in three days. And then overall, for the whole year, I'll only curtail, you know, five hours, so I'm still at four nines, instead of having to curtail 24 hours because I can't predict the weather.

• 49:14Dwarkesh Patel?

  Okay, so let's assume you're right. And then, I mean, I think at some point you will be right, like maybe we disagree about, sorry, I'm not qualified to disagree. Maybe you and some other person might disagree about what year it happens, but I think it's hard to deny that in the asymptote.

• 49:27Speaker 01

  Oh, yeah.

• 49:28Dwarkesh Patel?

  Our civilization is headed towards lots of energy use for AI and a lot of that coming from solar. In that asymptote, I've just like, I want to get like the crazy nerd sci fi, like what does our civilization look like? What is happening in this, you know?

• 49:42Speaker 01

  Katachev level one.

• 49:43Dwarkesh Patel?

  Yeah, let's wait till get to turning the entire Earth into an AI factory, but more like, I don't know, the 2030s, where you've got multiple people who are building on the order of five gigawatt or 10 gigawatt sites. The value of the hardware is dependent on its complement, which is the software, right? Like right now, AI models are fine. And so the hardware they're running on. The economic value they can generate is sort of bottlenecked by how good the software is. But if you actually had AGI, if you had like a human level intelligence, or maybe even better.

• 50:16Speaker 01

  Ideally better, yeah.

• 50:16Dwarkesh Patel?

  Yeah, running on an H100. That H100 is worth a lot, right? Like, we're paying a lot for humans to do work. Right now, I don't think AI is that valuable. Like, the models themselves aren't super, super valuable in terms of just pure economic value, right? OpenAI is generating on the order of 10 billion ARR, or some, you know, 20 billion ARR.

• 50:37Speaker 01

  The socks is terrible. How can they sleep at night?

• 50:39Dwarkesh Patel?

  I know. But for context, McDonald's and Kohl's generate more yearly revenue than that. But I think the promise of AGI

• 50:47Speaker 01

  Yeah.

• 50:48Dwarkesh Patel?

  is to automate human labor. Human labor generates on the order of $60 trillion of economic value. Or like, that's how much is paid out in wages to labor around the world, right? So that's what AGI can do. And even if you curtail it to just white collar work, there's still tens of trillions of dollars of value. So once we have models which are actually human level, they will be worth at least that, pending the fact that you can build them.

• 51:12Speaker 01

  Well, I don't think we should constrain ourselves to being like, oh, well, maybe there'll be some fraction of current payroll. Right? Because like, that's kind of a very contingent on like, humans being humans thing.

• 51:21Dwarkesh Patel?

  Yeah, I think it's a lower bound, to be clear.

• 51:24Speaker 01

  Oh, yeah, lower bound for sure. But like, if you think about like, someone trying to like, estimate the upper bound for the market value, the market cap of like Caterpillar, right? Based on like, well, it takes, you know, this many men and like, burrows and wheelbarrows to like, dig a trench. And so, you know, it couldn't be more than that, right? But actually, you know, one way to think about the industrial revolutions is every time you figure out industrial revolution, what you're doing is you're finding some way of like, bypassing a constraint, or bypassing a bottleneck. And the bottleneck prior to what we call the industrial revolution was metabolism, right? It's just like, how much like oats can a human or a horse physically digest and then convert into useful mechanical output for, you know, their peasant overlord or whatever?

• 52:07Speaker 01

  And nowadays, we would giggle to think that like, oh, the amount of food we produce is meaningful. In the context of like the economic power of a particular country.

• 52:15Dwarkesh Patel?

  Right.

• 52:15Speaker 01

  Because ninety nine percent of the energy that we consume routes around our guts, you know, through the gas tanks of our cars and through our aircraft and, you know, grids and stuff like that.

• 52:24Dwarkesh Patel?

  Interesting.

• 52:24Speaker 01

  And so right now, the AI revolution is about routing around cognitive constraints.

• 52:31Dwarkesh Patel?

  Yeah.

• 52:32Speaker 01

  That in some ways, writing, you know, like writing, printing press, computers, the internet have already allowed us to do to some extent. A credit card is a good example of something that routes around a cognitive constraint of like building up network of trust, right? It's decentralized trust.

• 52:45Dwarkesh Patel?

  Yeah, that's interesting. It's also really interesting, something that came up, and I already credit James Bradbury and Gorn with making this interesting point when I was talking with them a couple of days ago is, if you measured by GDP, AI's outputs might be underwhelming, right? One of the complaints that economists have about the internet is that it's hard to measure the consumer surplus that's created by the internet because a lot of the, a lot of the goods and services that are made available, you pay zero for them and so they don't show up in GDP.

• 53:13Speaker 01

  Well, it's the same with oil.

• 53:14Dwarkesh Patel?

  Yeah, in that sense that it's only one percent of energy is like one percent of GDP.

• 53:17Speaker 01

  Oil is like eight trillion dollars a year or something, right?

• 53:19Dwarkesh Patel?

  Yeah.

• 53:20Speaker 01

  But if you, if you, if you said, well, one day we're going to consume a hundred times more energy in the form of oil than in the form of food. And the per jewel cost of food is, you know, whatever it is, the cost of a Big Mac, then oil should be like eight hundred trillion dollars a year, right? It's a hundred times, like per, per unit energy, oil, like gasoline is a hundred times cheaper than the cheapest food that humans can digest. So, so, so like, does that mean that we've like shot ourselves in the foot by using oil to run our economy?

• 53:46Dwarkesh Patel?

  Right.

• 53:47Speaker 01

  Because it's so cheap, like, no.

• 53:49Dwarkesh Patel?

  Right, right. And also the, it's fraction of GDP also doesn't correspond to how important it is. For example, oil is like one percent of GDP or something, but if you don't have oil, then you have these oil shocks, which cause double digit decreases in GDP. So the elasticity of demand often matters more than its like raw fraction contribution to GDP. But anyways, on the original point about AI, so you're going to have this huge deflation of, so Gorn put it this way, he's like, if, if you imagine Dario's data center of geniuses.

• 54:24Speaker 01

  Yes.

• 54:24Dwarkesh Patel?

  How is that showing up in GDP? Well, it would be the inputs, which are the chips, the energy, et cetera, and the outputs, which are just the tokens. And neither of those is going to be that astronomical in comparison to the value of those, that data center of geniuses is producing. So in terms of GDP numbers, like. If that data center of geniuses automates a bunch of, or at least complements a bunch of human work, et cetera, it might actually cause like a nominal decrease in GDP, while at the same time contributing massively to what we might think of as the valuable stuff humanity or human civilization can produce. And so in the long run, it might make more sense to think of the size of our economy or the size of our civilization

• 55:06Dwarkesh Patel?

  as the raw energy use that we do rather than GDP. Because again, GDP will see this huge deflation.

• 55:14Speaker 01

  Yeah.

• 55:14Dwarkesh Patel?

  Because the variable cost of running AI.

• 55:17Speaker 01

  Yeah.

• 55:17Dwarkesh Patel?

  Will just be pretty cheap as compared to like paying humans wages, et cetera.

• 55:22Speaker 01

  I think, I think when the, we're at the point where like, you've got a mixed economy with like, an AI doing my job and also a human doing my job.

• 55:31Dwarkesh Patel?

  I love how this is, this is the new way we use the phrase mixed economy.

• 55:34Speaker 01

  Yeah. Then obviously, like, I still have some pricing power relative to humans and the AI does has pricing power, right? But, but if it was the case that like a new, a new kind of job emerges that AI is really well adapted to, like the, because it's not competing against humans for most of those roles, it'd be competing against the other labs, right? And so you'd actually see like the cost pushed down to, you know, a small multiple of whatever the marginal production cost of those tokens is. That would be my guess. So I think, I think it might be a mistake to assume that like, well, if, if we're going to pay, you know, a top AI researcher $200,000, $200,000 a year. Well, let's say the sort of AI researcher that I could be, $200,000 a year, that if an AI comes along that's as good as me at that, even taking into account the fact that realistically speaking,

• 56:18Speaker 01

  I only get, you know, maybe 10 hours of really top cognitive work done a week, that it would also be worth $200,000.

• 56:24Dwarkesh Patel?

  Yeah.

• 56:24Speaker 01

  Obviously, it's, it'd be worth much more than that in, in the sense of like, you can copy paste its output and much less than that in the sense that, you know, whatever the marginal additional cost of spooling up H100s is. And so if some, if some kind of role comes along that like the AIs are really well specialized at and outcompete the humans quickly, then we'd also expect to see that there, both the, the cost of providing that service to drop drastically, at the same time as the overall value generated in the economy by that service would increase a lot.

• 56:50Dwarkesh Patel?

  Yeah, exactly. If we think that the value of cognition is going to be unbounded and the way to derive cognition, you can just, to the extent you think it's still a little eventually when you can just, you can derive it from how much land it takes to power. In H100 using solar panels. That is a very interesting derivation that like, okay, well, this is at a minimum, we're going to just fill up all the land. I mean, at some point, you might have like declining marginal value of cognition or something, but.

• 57:19Speaker 01

  We kind of discussed this earlier, but if you have like a megawatt of 10 acres, 10 acres of land feeding one megawatt H100 or something.

• 57:27Dwarkesh Patel?

  Right.

• 57:28Speaker 01

  That's generating like, let's say a megawatt is 1,000 humans. So one acre is 1,000 humans worth of.

• 57:33Dwarkesh Patel?

  Right.

• 57:34Speaker 01

  Cognition. The implicit land value there is a lot higher than it is as like undeveloped desert.

• 57:39Dwarkesh Patel?

  Right?

• 57:41Speaker 01

  It's also a lot higher than it is as like the most productive farmland that humanity has ever had.

• 57:46Dwarkesh Patel?

  Right. At current hardware efficiencies, I don't know if it's worth spelling out. Basically, H100 has the same amount of flops as a human brain, but also uses way more energy than a human brain.

• 57:56Speaker 01

  Yes.

• 57:56Dwarkesh Patel?

  It uses like 50x more energy.

• 57:58Speaker 01

  Is that right?

• 57:59Dwarkesh Patel?

  So 20 watts versus 1,000 watts?

• 58:00Speaker 01

  Human brain, 20 watts, yeah, could be.

• 58:03Dwarkesh Patel?

  We know hardware can be at least as efficient as the human brain and the human brain can generate this many flops on 20 watts.

• 58:09Speaker 01

  Yep.

• 58:09Dwarkesh Patel?

  So if you do that calculation.

• 58:12Speaker 01

  Yep.

• 58:12Dwarkesh Patel?

  Then that's 50x 1,000, so 50,000.

• 58:15Speaker 01

  Yep.

• 58:16Dwarkesh Patel?

  AI souls off of.

• 58:17Speaker 01

  And it could easily be much more than that because neurons are much slower than than transistors, obviously. So probably 10 years ago, one of my friends reminded me, you know, like the way your phone saves power is it goes to sleep between you like tapping out hello.

• 58:31Dwarkesh Patel?

  Right.

• 58:31Speaker 01

  Like H goes to takes a nap for like 10,000 cycles, you know? It's kind of nuts. So just that like humans, I think Elon's talked about this in the context of self-driving cars as well, which is like anything humans do is like glacially slow from the perspective of a computer.

• 58:45Dwarkesh Patel?

  Let's actually go back to the original point of like, I was explaining why I think it's plausible that there could be more than hundreds of gigawatts of extra demand from AI in the 2030s. I want to understand what that looks like in the real world. Like at that point, it's become basically this industrial problem of can you generate enough solar panels and solar modules and batteries and.

• 59:07Speaker 01

  Yep.

• 59:07Dwarkesh Patel?

  Not to mention the chips themselves.

• 59:08Speaker 01

  That's the industrial point, and then there's a cultural point as well, which is like.

• 59:11Dwarkesh Patel?

  Let's start with the industrial point. I want.

• 59:14Speaker 01

  Okay.

• 59:14Dwarkesh Patel?

  I want to know what the year 2035 looks like.

• 59:16Speaker 01

  Yeah, yeah, yeah.

• 59:17Dwarkesh Patel?

  If we've got AGI and we're just bottlenecked by the ability to deploy it.

• 59:20Speaker 01

  Yeah, so you can ask a question like. What do you need in order to run? Like, what is the minimum amount of matter that you need in order to perform this? These calculations? So, so, so right now, we're talking about like air racking and grid and transmission and a bunch of like ISOs and all that rest. You don't need any of that stuff, right? And like, obviously, XAI is on top of this because the first thing that Elon will always ask is like, delete anything you don't absolutely need. So what you actually need is a big slab of relatively cheap silicon to make the power. And then a small slab of relatively expensive silicon to do the thinking. And if it's in space, that's all you need, right? Because it's in sun all the time, so you don't need a battery, right? But if you're on the earth, you need a battery as well. So you need some interconnects. And you don't need a transformer, right? You don't even need a DCTC converter.

• 1:00:05Speaker 01

  You can actually make do with buck converter or with relays or whatever to basically match the, the current output of your solar array with the charge state of your batteries and the power consumption of the, of your, you know, GPU or something. But like a solar array about the size of this desk, for example, will generate about 500 watts in, in full sun. So you could actually imagine like aliens who have different silicon technology stack, building their systems as like an integrated solar array with, you know, a bit of, a bit of, you know, computranium in the middle, for example, on the same wafer. But that's, that's basically all you need.

• 1:00:38Dwarkesh Patel?

  On the same wafer, because it's all silicon.

• 1:00:40Speaker 01

  It's all silicon. It's all silicon all the way down. What's silicon made of? Well, it's an element, it's chemically in the crust. There's no shortage of it.

• 1:00:46Dwarkesh Patel?

  This is a great prompt for a sci fi exercise because especially in space, you don't need batteries. So you're just like, the, the prompt is the, the future TSMC just manufactures integrated solar.

• 1:00:58Speaker 01

  And they can fly around, right? They're, they're solar sails, right? And they don't, they're, they're relatively dense, so they don't fly crazy fast, but they don't need to because they're immortal.

• 1:01:05Dwarkesh Patel?

  Is this what the dice is, you're a little bit of Casey? Is it just, is this going to be computronium at the center of a solar sail?

• 1:01:09Speaker 01

  They can fly closer to the sun to get more power.

• 1:01:11Dwarkesh Patel?

  Right.

• 1:01:11Speaker 01

  Right, up to the thermal limit. And they can fly further from the sun to go and explore or like fly to other planets or something. And they can, they can adjust the orientation of the solar sail with LCD panels that it could be integrated into the, into the wafer itself. So. I mean, that's actually, if you say, what's the post-human state? That's it. We will eventually be-

• 1:01:29Dwarkesh Patel?

  A solar cell with a silicon die in the middle for compute?

• 1:01:32Speaker 01

  Yeah, one human's worth of computation, one human brain can be simulated and roughly a square meter of silicon floating in space.

• 1:01:37Dwarkesh Patel?

  How much are you?

• 1:01:39Speaker 01

  One square meter of silicon, like the thickness of a sheet of paper floating in space.

• 1:01:42Dwarkesh Patel?

  Yeah, interesting.

• 1:01:43Speaker 01

  That's it. That's a future human form. That's my final form.

• 1:01:45Dwarkesh Patel?

  That's the extraterrestrials.

• 1:01:46Speaker 01

  Yeah. That's assuming a little bit of software improvement, but I don't think that's-

• 1:01:50Dwarkesh Patel?

  All that's assuming is software improvement. The Dyson Sphere, that's all it needs, a little bit of software, a little bit of tweaking on the algorithm.

• 1:01:58Speaker 01

  If the software, like the area of the panel is kind of the variable there.

• 1:02:02Dwarkesh Patel?

  Right.

• 1:02:03Speaker 01

  So you can ask, well, what do you need in order to make the silicon making solar rays, making chips is a kind of multi-stage process, but basically you start off with silicates, which are rocks, ideally in a relatively pure form. You chemically reduce them, a couple of different processes that can do that, and then purify them into ideally like six nines of purity for solar ray, maybe nine nines for like really nice computers, silicon purity, and grow crystals, cut wafers, et cetera, et cetera. So then the constraint is like, well, how quickly can you convert the crust into enough silicon to support silicon thought?

• 1:02:34Dwarkesh Patel?

  Yeah.

• 1:02:35Speaker 01

  What does the silicon ecosystem look like?

• 1:02:37Dwarkesh Patel?

  Any thoughts?

• 1:02:38Speaker 01

  Well, it's pretty quick. Like if you're like one kilowatt per square meter and then you use that just to like rip oxygens off the underlying dirt, it doesn't take all that long to, you need about 20 microns of silicon to make a solar PV array. So it's-

• 1:02:52Dwarkesh Patel?

  Like from dirt, as you mean like actual dirt?

• 1:02:55Speaker 01

  Yeah, actual dirt has plenty of silicon in it. So for example, setting up a brand new silicon refinery takes about 18 months.

• 1:03:00Dwarkesh Patel?

  Right.

• 1:03:00Speaker 01

  Right. But that's just with the current technology, right? I actually think we may find ways. So one of the nice things is if you have infinite free solar power, approximately free solar power, you can revisit a bunch of legacy industrial processes that have been optimized for efficiency and say, well, what if we just like use twice as much power and we just want to do them faster and cheaper, right? Like less capex, less lead time, more power, right? Well, you can start solving problems. It turns out that if you want to chemically reduce silicon, you can do it electrolytically, like with less efficiency and a bunch like under a hydrogen-rich atmosphere or something. And then, so one of the ways that silicon can be refined is by turning it into silane, which is a silicon tetrahydride.

• 1:03:41Speaker 01

  I'm not really a chemist, but I think that's right. So SIH4, which is a gas, it's like, it's actually like methane, but one down on the periodic table. And don't breathe it though. And yeah, and then, and then, once it's a gas, you can filter it from all the contaminants which don't form gases or can be separated by density. Much like how uranium is sometimes enriched, but much, much less difficult. And then heat it up to separate it back into pure silicon, where you can then precipitate out a.

• 1:04:09Dwarkesh Patel?

  I mean, the reason I think it's interesting is because whenever people who are talking about the AI singularity, often their expertise is not in energy or physics or whatever. So they focus only on the cognitive elements of the singularity, which is like, how much faster can we make AI smarter, et cetera? I think this is really interesting because if we have unbounded cognition, which sets up both the ability to supply and to demand more energy, I'm very curious, like, what does the energy singularity look like? Where, you know, we're just trying to saturate as much energy that the earth receives and turn it into cognition.

• 1:04:48Speaker 01

  I hadn't thought about that before, but like, there's this idea that like, evolution resulted in this like, continual like ramification and complexification of the thermodynamic gradient, right? So you start with like very simple, like RNA-based organisms, and now you get this like industrial economy. But it may be the case, and I don't have a strong reason to suspect one or the other, that like, what we're seeing is like, the beginning stages of a collapse back towards the simplest possible thermodynamic to cognition stack, which is, we have fusion in stars and the inky blackness of space, and that provides our temperature gradient. And then the most efficient way to convert that into usable cognition is silicon. Like literally electrons being pushed across the Fermi gap in a solar ray, and then taking the return path through some like

• 1:05:31Speaker 01

  set of gates, making decisions about things, and then like beaming lasers to their friends, saying, hey, I just made up a new meme.

• 1:05:37Dwarkesh Patel?

  That is an interesting concept, that for four billion years, we've been increasing this sort of like variance in complexity. Of creatures and then you might see like this big collapse. Should I give you the opportunity to plug why, uh, why people should work for Terraform?

• 1:05:54Speaker 01

  Yeah. So actually, just to give you an introduction, Terraform is my day job. It's a company I founded almost four years ago. We are making synthetic natural gas from sunlight and air. We are also working on other core primary materials stuff. We also have a methanol process, methanol and methane together. Procares to every hydrocarbon you could possibly want. Another chemical ammonia process, steel, desalination. And we can also make cement and a few other things. So basically everything that the primary, uh, primary industry does, except for glass and paper. We are hiring, uh, jobs are available on terraformindustries.com. Yes, the website's meant to look like that because we're very cool. We are some very special people, you know? I know a lot of smart people and I'm privileged to work with some of the smartest people I know. We are mostly mechanical engineers. I will never hire anyone who can't do math.

• 1:06:39Speaker 01

  I will never have the problem at astronomer because we don't have a head of HR.

• 1:06:44Dwarkesh Patel?

  And also the CEO is not having an affair.

• 1:06:48Speaker 01

  Yeah, I mean, like step one.

• 1:06:50Dwarkesh Patel?

  I think that was a more crucial issue, Casey.

• 1:06:53Speaker 01

  But like heads of HR can get into trouble. I'm just saying, everyone does math. Terraform, it's very important to me that Terraform is the place that ambitious hardware people go to become the best they can be. Right? That is really important. I'm not, it's not here to like check in and get your paycheck and like optimize some shiny widget. It's still a small team. It's still like one project per person kind of situation. And I will level you up, like maybe not quite like Jensen taught you into greatness kind of situation. But like at times it's going to feel that way. And you get to work with the, the best people that there are, at least on the west coast of the United States on, on this sort of thing. And it's also a unique company. I thought years ago, like by now I'll have competition. We don't, no one else is doing this except for a small smart startup in the UK.

• 1:07:34Speaker 01

  And, uh, and so, you know, you get on the grand floor and it's going to be super cool technology. And, uh, we, you know, eventually we get to go and build it all on Mars as well and like, help our robot overlords make more of themselves out of dirt. It's pretty cool.

• 1:07:47Dwarkesh Patel?

  Nice.

• 1:07:47Speaker 01

  Come, come work for us.

• 1:07:48Dwarkesh Patel?

  Casey, thank you so much for coming on the podcast.

• 1:07:50Speaker 01

  Oh, thank you.

• 1:07:50Dwarkesh Patel?

  It's fun.

• 1:07:50Speaker 01

  Yeah.

• 1:07:51Dwarkesh Patel?

  I hope you enjoyed this episode. If you did, the most helpful thing you can do is just share it with other people who you think might enjoy it. Send it to your friends, your group chats, Twitter, wherever else. Just let the word go forth. Other than that, super helpful if you can subscribe on YouTube and leave a five-star review on Apple Podcasts and Spotify. Check out the sponsors in the description below. If you want to sponsor a future episode, go to dwarkesh.com/adtise. Thank you for tuning in. I'll see you on the next one.