Main Topic: The conversation revolves around various topics, including engineering, software, economics, investing, and energy. The main topics discussed are:
Key Points:
Participants:
Takeaways:
Austin Vernon is an engineer working on a new method for carbon capture, and he has one of the most interesting blogs on the internet, where he writes about engineering, software, economics, and investing.
We discuss how energy superabundance will change the world, how Starship can be turned into a kinetic weapon, why nuclear is overrated, blockchains, batteries, flying cars, finding alpha, & much more!
Watch on YouTube. Listen on Apple Podcasts, Spotify, or any other podcast platform. Read the full transcript here.
Follow Austin on Twitter. Follow me on Twitter for updates on future episodes.
Timestamps
(0:00:00) - Intro
(0:01:53) - Starship as a Weapon
(0:19:24) - Software Productivity
(0:41:40) - Car Manufacturing
(0:57:39) - Carbon Capture
(1:16:53) - Energy Superabundance
(1:25:09) - Storage for Cheap Energy
(1:31:25) - Travel in Future
(1:33:27) - Future Cities
(1:39:58) - Flying Cars
(1:43:26) - Carbon Shortage
(1:48:03) - Nuclear
(2:12:44) - Solar
(2:14:44) - Alpha & Efficient Markets
(2:22:51) - Conclusion
Dwarkesh Patel
Okay, today I have the pleasure of interviewing Austin Vernon, who writes about engineering, software, economics, and investing on the internet, but not that much else is known about him. So Austin, do you want to give us a bit about your background? I know that the only thing the internet knows about you is this one little JPEG that you had to upload with your recent paper, but what about an identity reveal, or a little bit of a background reveal, to the extent that you're willing to share?
Austin Vernon
My degree is in chemical engineering, and I have a lifelong love of engineering and things like the Toyota production system. I've worked as a chemical engineer in a large processing facility, and I've done a lot of petroleum engineering. Now, I taught myself how to write software and I'm working on research and early commercialization of CO2 electrolysis.
Dwarkesh Patel
Okay, yeah. I'm really interested in talking about all those things. So the first question I have is from Alex Berger, the co-CEO of Open Philanthropy. He asked this question when I asked on Twitter what I should ask you, and he suggested I should ask you: why so shady? You have famously kind of an anonymous, pseudonymous personality on the internet. What's up with that?
Austin Vernon
Yeah. I think he posted a tweet that said, "I don't know who this guy is or if he's credible at all, but his stuff sure is interesting." That really made me laugh. It just doesn't seem necessary. I'm fine with my ideas being well-known and communicating, but I have less desire to be personally famous.
Dwarkesh Patel
I gotcha. I wanted to start off with a sexy topic. Let's talk about using Starship as a kinetic weapon. I thought that was one of the more amusing posts you wrote. Do you want to talk more about how this would be possible?
Austin Vernon
The main thing with Starship is that you're taking a technology and making it about a hundred times cheaper for cargo and a thousand times cheaper for people. When things happen that drastically, you're looking at huge changes, and it's really hard to anticipate what some of those can be. I think there's a lot of moon-based, Mars-based stuff that doesn't catch the regular public's eye, and they have trouble imagining some of the point-to-point travel that could be possible. But when you start talking about it as a weapon, it lets people know they should be paying attention to this technology. We certainly do not want to be second or third getting it, and we should make sure we're going to be first.
Dwarkesh Patel
You mentioned in the post that as recently as the 90s, the cost of sending one kilogram to space was around $20,000. More recently, SpaceX has brought it to $2,000. There are a lot of interesting questions you can ask when you consider what will be possible once we get it down to $200 per kilogram to send into orbit. One of them might be to manufacture these weapons that are not conventional ballistics. Why might this be an advancement over conventional ballistic weapons?
Austin Vernon
Regular conventional ballistic weapons are extremely expensive. This is more like a bomb truck, but even cheaper than the B-52 delivering mass on target. Think about how expensive it is to fly a B-52 from Barksdale, Louisiana, all the way across the world. You can do it from South Texas or Florida with Starship and get more missions per day. When you go orbital, it takes a lot to get to orbit, but once you're in orbit, your fuel consumption is pretty good. So over long distances, it has a lot of advantage. That's why point-to-point works for longer distances. There's a sweet spot with these weapons where you want it to be pretty accurate but also cheap.
You're seeing that problem with Russia right now. They have fancy parade-style weapons that are really expensive, like multi-billion dollar cruise missiles, but they're missing the $5,000 guided artillery shell or the $20,000 JDAM that you can just pit massive amounts of. They're short on all those because they had a limited amount of chips they could get from the US into Russia to make these advanced weapons. Starship gives you a platform to deliver. You could pit JDAMs in a shroud, or you could just drop unguided kinetic projectiles. It becomes impossible for a ship to launch missiles to intercept yours if your cost is so low you can just overwhelm them.
Dwarkesh Patel
There are a few terms there that neither I nor the audience might know. What is a JDAM? What is a shroud? And why are chips a bottleneck here? Why can't it just be any microcontroller?
Austin Vernon
JDAM is a Joint Direct Attack Munition. We took all our Vietnam surplus bombs and put a little fin kit on them. It costs about $20,000, which is cheap for a weapon because the actual bomb costs maybe $3,000. It turns it into a guided weapon. Before, you were probably lucky to get within 500 meters of a target. Now you can get within two meters. The number of missions you have to do with your planes goes down by orders of magnitude. It's a huge advantage in logistics and how much firepower you can put on a target. We didn't even have to make new bombs; we just put these kits on all our old bombs. As for the chips, there's an organization in the UK that has been tearing down Russian weapons found in Ukraine. They all have American chips in them. Technically, they're not supposed to be able to get these chips. Russia can't make a lot of its own chips, especially the specialized kinds you might want for guided weapons, so they've been smuggling them in.
Dwarkesh Patel
What is special about these chips? As far as I'm aware, trade with China is still going on, and we get a lot of our chips manufactured from Taiwan or China. Why can't they do the same?
Austin Vernon
It's the whole integration. It's not just a specific chip, but the board. They're more like PLCs where you have wired-in programming. To be able to do the guidance and all that stuff, it all has to work together. I don't have a really good answer for that one, but they're hard to replicate. That's what matters.
Dwarkesh Patel
That's interesting. That has a lot of downstream effects. For example, India buys a lot of its weaponry from Russia. If Russia doesn't have access to these chips, then other countries that buy from Russia won't have access either. You had an interesting speculation in the post where you suggested you could keep these kinetic weapons in orbit, like a sort of Damocles. That sounds like a really scary scenario where you can have orbital decay and these kinetic weapons falling from the sky and destroying cities. Do you think this is what it could look like in 10 to 20 years?
Austin Vernon
The advantage of having them on orbit is you can hit targets faster. If you're launching the rocket from Florida, you're looking at maybe 30 minutes to get there. The target can move in that time. If you're on orbit, you can have them spaced out to where you're hitting within a few minutes. You really have to have a two-stage system for most weapons. If you have a really aerodynamic rod that gives you good performance in the low atmosphere, it'll get going too fast and just burn up before you get there. Tungsten is maybe the only thing that could go all the way through. That's why the original concept used big tungsten rods the size of a telephone pole. But tungsten is pretty expensive, and the rod concept limits what you can do. A lot of these weapons will have a shroud that slows you down in the upper atmosphere. Once you're at a velocity where you're not going to melt, you open it up and let it go. If you had to fall from the sky, some may make it to the ground, but a lot would burn up. Stuff that makes it to the ground is actually pretty light, with a large surface area. That's the whole thing with Starlink satellites—they were meant to completely fall apart on re-entry.
Dwarkesh Patel
One implication of that is that these may be less powerful than we might fear. If kinetic energy is mass times velocity squared, and there's an upper bound on the velocity, then you can upper bound the power these things will have.
Austin Vernon
Even the tungsten rod—sometimes people don't do the math and think it's going to be like a nuclear weapon. I think even the tungsten rod is like 10 tons of TNT or something. It's a big bomb, but it's not a super weapon. As I said in the post, these are almost more defensive weapons. I can keep you from putting your ship somewhere. I could try to bombard your cities, but I can't take ground with it. I can't even police sea lanes with it. I'd still have to have regular ships to enforce the rules of the sea and board freighters.
Dwarkesh Patel
You speculated in the post that you could load this up with shrapnel and have it explode next to an incoming missile or aircraft. Could these get that accurate? That was surprising speculation to me.
Austin Vernon
For ships, I think it's pretty doable. I watched videos of how fast a ship can turn. If you want to release your shrapnel to try to kill their radars, you'd want to do it above the ceiling of their missiles. How much are they going to move between your release and impact? It's maybe like a thousand feet. That's pretty simple; you just shrapnel the area. For an aircraft, you would be steering all the way in. It's doable, but it'd be pretty hard. You might want to go even slower than you would with a ship attack. You need a specialized package to do the aircraft. But you can see these aircraft with synthetic aperture radar from satellites and guide it in the whole way. You could even load heat-seeking missiles into a package that unfurls right next to them and launch conventional missiles. It would be hard to do, but it's the kind of thing you might be able to do if you put some effort into it.
Dwarkesh Patel
I find this kind of speculation really interesting because when you look at modern weaponry used in conflicts, it often seems directly descended from something you would have seen in World War Two. Warfare changed so much between 1900 and 1940, but now it seems like the entire category of weapons hasn't changed.
Austin Vernon
Our physical technology hasn't changed that much. It has made more sense to put better electronics in the same tanks we have than to build a new physical tank that's way better. We haven't learned enough about tanks to build a new one that's way better, so we just keep upgrading our existing tanks with better electronics. They're much more powerful, more accurate, have longer range weapons, and better sensors. The tank looks the same, but it may have several times more killing power. The Ukraine war right now is using a lot of 40- to 50-year-old weapons, so it especially looks like that.
Dwarkesh Patel
That worries me when I think about our own military's stockpiles. I imagine we don't have the newest of the new thing, just maintained versions of decades-old technology.
Austin Vernon
There's a lot of debate about how ready our military is. For certain situations, it's more ready than others. Most people talking about it have the incentive to downplay our capabilities because they want more defense spending. I think we're probably more capable than what you might see in an editorial in The Hill. Just us sending a few weapons to Ukraine and how successful they've been using them shows a little bit of that. But there's so much uncertainty when it comes to fighting, especially a naval engagement. We just don't have that many ships in general. You can have some bad luck. You always want to be a little wary and not get overconfident.
Dwarkesh Patel
If the offensive tech we sent to Ukraine is potentially better than the defensive tech, it's possible that a ballistic missile from China or Russia could sink a battleship and kill a thousand soldiers. Do you think this opens up avenues for defensive tech as well?
Austin Vernon
Generally, the consensus is that defensive technology has improved much more recently than offensive technology. China has a strategy called Anti-Access/Area Denial, or A2AD. Missiles have gotten better because the sensors on missiles have gotten better, so they can keep our ships from getting close to them. But they can't really challenge us in Hawaii or something. It goes both ways. It's hard for us to get close to China, but Taiwan has a lot of missiles with these new sensors as well. I think it's probably tougher for China to get close to Taiwan than most people would say.
Dwarkesh Patel
Can you talk more about that? Every time I read about this, people say that if China wanted to, they could knock out Taiwan's defenses in a short amount of time and take it over. Why is that not possible?
Austin Vernon
It might be possible, but it's a guess. Taiwan has one of the largest defense budgets in the world. They spend about $25 billion a year and recently added an extra $5 billion. They've been buying a lot of anti-ship missiles and air defense missiles—stuff that Ukraine could only dream of. Ukraine's military budget was like $2 billion. Taiwan has a professional army and they're gearing up. The other thing is they're an island. Russia could just roll over the land border into Ukraine, but there are very few successful amphibious landings in history. The most recent ones were all the Americans in World War II and Korea. The challenge is on China to execute perfectly. If they had perfect execution, then possibly. But if their air defenses on their ships aren't quite as good as we think, they could end up with half their fleet underwater within 10 hours.
Dwarkesh Patel
How has the Ukraine conflict updated your view of Taiwan's defensive capabilities?
Austin Vernon
I didn't really know much about it, and then I started looking at Wikipedia and all this stuff they're doing. Taiwan has a lot of modern platforms like F-16s with our anti-ship missiles. They have a lot of their own indigenous fighter bombers and indigenous anti-ship missiles because they're worried we might not always sell them to them. They've even recently gotten long-range cruise missiles that could possibly target leadership in Beijing. That makes it uncomfortable for the Chinese leadership. If you attack them, you're going to have to live in a bunker. But again, I'm not a full-time military analyst, so there's a lot of uncertainty. It's not a given that China is just going to roll over them.
Dwarkesh Patel
That's comforting to hear. Let's talk about an area where I have a little bit of a point of contact. I thought your blog post about software and the inability of it to increase productivity numbers was super fascinating. Before I ask you questions about it, do you want to lay out the thesis?
Austin Vernon
That post is one where I felt like I caught lightning in a bottle. Everything I wanted to put in it just fit together perfectly, which is usually not the case. The idea is that the world is so complex, and we really underestimate that complexity. If you're going to digitize processes and automate them, you have to capture all that complexity at the bit level, and that's extremely difficult. You also have diminishing returns where the easily automatable stuff goes first, and then it's increasing corner cases to get to the end. You have to write more and more code. That's why we don't see runaway productivity growth from software—because we're fighting all this increasing complexity.
Dwarkesh Patel
Have you heard of the waterbed theory of complexity?
Austin Vernon
I don't think so.
Dwarkesh Patel
It comes up in compiler design. The idea is that there's a fixed amount of complexity in a system. If you try to reduce it, you'll just end up migrating the complexity elsewhere. An example is programming languages that are not type-safe, like Python. You can say it's a less complex language, but you've added complexity when two different types of numbers are interacting, like a float and an int. As your program grows, that complexity exponentially grows of all the things that could go wrong. You can choose where to have your complexity, but you can't get rid of it.
Austin Vernon
There's an interesting thing when you start pairing that with management theory. For a long time, the most complex thing we were doing was high-volume car manufacturing. We got a lot of innovations and organization from car manufacturers like the assembly line. Then you had Sloan at GM creating the way the modern corporation is run. Then you have the Toyota Production System. Arguably, creating software is now the most complex thing we do. There are all these squishy concepts that underlie the Toyota Production System that software has had to learn, reimagine, and adopt. You see that with Agile, where we can't have long release times and need to be releasing every day—that's limiting inventory.
There's a whole thing showing up in software that existed in car manufacturing, where you're talking about reducing communication. Jeff Bezos famously said he wants to reduce communication, which is counterintuitive to a lot of people. This is age-old in car manufacturing. Toyota has these cards that go between workstations and tell you what to do. People normally think of them as limiting inventory, but it also tells the worker exactly what they're supposed to be doing, at what pace, at what time. The assembly line is like that too. You just know what to do because you're standing there, a part is here, and it needs to go on there.
It comes by at the pace you're supposed to work at. There's a famous paper by Levitt and Syverson where they went to a car factory and studied how defects propagated. Once a car factory gets up and running, it doesn't matter what workers you put in there. If workers are sick or you get new workers, the defect rate is the same. All the knowledge is built into the manufacturing line. There are concepts around idiot-proofing that are very similar to what you'll see in software. You had Uncle Bob on the show. Uncle Bob says only put one input into a function because you'll mix them up otherwise. The Japanese call it "poka-yoke"—you make it where you can't mess it up.
That's another way to reduce communication. In software, of course, you have APIs. I'm really interested in this overall concept of reducing communication and reducing how much cooperation we need to run the economy.
Dwarkesh Patel
Speaking of the Toyota Production System, one thing they do to reduce the defect rate is that if there's a problem, all the workers in that chain are forced to go to the place where the defect or problem is and fix it before doing anything else. The idea is that this gives them context to understand what the problem was, how to make sure it doesn't happen again, and also prevents a buildup of inventory that keeps making these defects happen.
Austin Vernon
Exactly.
Dwarkesh Patel
You said a little bit earlier that software is a place where complexity is the highest in the world right now. One interesting thing is that software gives you a choice to interface with the complexity you want to interface with. For example, a machine learning model is really complex, but ideally you get to a place where that's the only kind of complexity you have to deal with. You're not having to deal with the complexity of how the program is compiled or how the libraries are built. You can fine-tune and work on the complexity you need to work on. It's similar with app development. Patrick Collison has a blog post about Stripe as a "solid state" system. The basic idea is that Stripe hides all the complexity of the financial system. It charges a higher fee, but you can treat it as an abstraction of a tithe you have to pay, and it'll take care of that entire process so you can focus on your comparative advantage.
Austin Vernon
It's very similar in car manufacturing and the Toyota Production System. It's the same conceptual framework. In the Toyota Production System, everyone works at the same pace, and your work content is the same. There's no room for not standardizing the way you're going to do things. Everyone gets together and says, "We're going to put this certain part together this certain way at this little micro station, and it's going to be the same way every time." That's part of how they reduce defect rates. If your assembly process is too long, longer than your time allotment to stay in touch with the rest of the process, you keep breaking it down into smaller pieces. Each person only has to know a very small part of it. Even the overall engineering team has all sorts of strategies like this too. They have tools to help them break up all these processes into very small parts and make it all hold together. It's still very hard, but it's a lot of the same ideas. You're taking away the complexity of making a 30,000-part car where everyone is just focusing on their one little part and doesn't care what someone else is doing.
Dwarkesh Patel
The interesting thing there is that you need one person who knows how everything fits together. One of the tenets of the Toyota Production System is that you need to have a global view. In the book, they talked about examples where people would try to optimize for local efficiencies. They pointed to Ford and GM for trying to do this, where they would try to make machines run all the time. Locally, you could say that process is super efficient because it's always outputting stuff, but it ignores how that added inventory or that process had a bad consequence for the whole system. It's interesting if you look at a company like Tesla that is able to do this really well. Tesla is run like a monarchy, and this one guy has a total global view of how the entire process is supposed to run. Where do you have these inefficiencies?
You have some great examples of this in the blog post. I think one of the examples from *The Goal* or *The Machine That Changed the World*—this guy goes to a factory, and the author asks, "Is this an efficient factory?" The guy says, "Yeah, this is totally efficient. There's nothing we can do adopting the Toyota way to make this more efficient." So he says, "Okay, let me look." He finds that in one stage, they're treating steel in some way. It should only take a couple of seconds, and the main process does only take a couple of seconds. But some local manager decided it would be more efficient to ship their parts out to get the next stage of the project processed somewhere else. Locally, it's cheaper, but the result is that it takes weeks to get these parts shipped out and back. The actual time the parts spend getting processed is like 0.1 percent of the time, which makes the whole process super inefficient.
It seems like the implication is you need a very monarchical structure with one person who has a total view in order to run such a system. Am I getting that wrong?
Austin Vernon
Not necessarily. You do have to make sure you're not optimizing locally, but I think it's the same constraint you have in software. A lot of times people just run over it because processing has been getting so much cheaper. People are expensive, so if you can save development time, the trade-offs are different when you're talking about the tyranny of physical items. The constraints get a little more severe, but you still have to fight local optimization. The level you have to fight it is probably different with physical goods.
I was thinking about the smart grid situation from a software perspective. There's this problem where I'm putting my solar farm here, and it's impacting somewhere far away, creating really high upgrade costs that cost two or three times more than my solar farm. The obvious thing if you're doing software is to break all these up into smaller sections so you wouldn't be impacting each other. You could focus on your own little thing. But the problem is, if you're going to disconnect these areas of the grid, the equipment to do that is extremely expensive. It's not like I'm just going to hit a new tab, open a new file, and start writing a new function. Not only that, but you still have to coordinate how this equipment is going to operate. If you just let the grid flow as it does, everyone knows what's going to happen because they can calculate the physics. If you start adding in all these checkpoints where humans are doing stuff, you have to interface with the humans, and the amount of things that can happen really starts going up. It's actually a really bad idea to try to carve all this stuff off just because of the reality of the physical laws and the equipment you need.
Dwarkesh Patel
Okay, interesting. And then I think you have a similar Coasean argument in your software post about why vertically integrating software is beneficial. Do you want to explain that thesis?
Austin Vernon
Yeah, it actually gets to what we're talking about here. It allows you to avoid local optimization. A lot of times, you're trying to build a software MVP and you're tying together a few services. They don't do quite what you need. If you try to scale that, it would just break. If you're going to take a really complex process like car manufacturing, distribution, retail distribution, or the home buying process, you really have to vertically integrate it to be able to create a decent end-to-end experience and avoid that local optimization. It's very hard otherwise. You just can't coordinate effectively if you have 10 different vendors trying to do the same thing. You end up in constant vendor meetings trying to decide what the specs are, instead of giving someone or a team the authority to just go start building stuff.
When you look at these companies, they have to implement somewhat decentralized processes when they get too complex, but at least they have control over how they're interfacing with each other. Walmart has the vendors control their own stock. They don't tell the vendor, "We need X parts." It's just, "It's on you to make sure your shelf is stocked."
Dwarkesh Patel
Yeah. What was really interesting to me about this part of the post was that I had this vision of where software is heading, where everybody will have a software-as-a-service company and they'll all be interfacing with each other in some sort of cycle, calling each other's APIs. The implication from your argument was that, given the necessity of integrating all this complexity vertically in a coherent way, the winners in software should end up being a few big companies that compete with each other.
Austin Vernon
I think that's especially true when you're combining bits and atoms. Maybe less true for pure software. The physical world is just so much more complex, and the constraints it creates are pretty extreme compared to a pure software world where you could maybe get away with more of everyone having an API.
Dwarkesh Patel
Right. I guess you might think that even in the physical world, given that people really need to focus on their comparative advantage, they would just try to outsource the software parts to these APIs. But is there any scenario where the learning curve for people who are not in the firm can be fast enough that they can keep up with the complexity? There are huge gains from specialization and competition that go away if this is the world we're forced to live in. We have a lot of counterexamples of what you're talking about. Apple is the highest market cap in the world, and famously they are super vertically integrated, combining hardware and software. Is there any world in which you can keep that kind of benefit but have it be within multiple firms?
Austin Vernon
This is a post I've got on my list. I want to write about blockchain applications. The one that excites me personally the most is reimagining enterprise software, because the things you're talking about—hard typing and APIs—are basically built into some of these protocols. I think it has a lot of exciting implications for how much you can decentralize software development. But you can still do that within the firm. If the government is placing all these rules on the edge of the firm, it makes transactions with other firms expensive. Internal transactions can be cheaper because they avoid government reporting and taxes. You'd have to think about how these technologies can reduce transaction costs overall and decentralize that, but also what the costs are between firms.
Dwarkesh Patel
It's really interesting if the costs are logistic or if they're based on the knowledge that is housed within the factory. If it's just logistical stuff, like you had to report any outside transactions, then that does imply that a technology like blockchain could help. But if it's that you need to be in the same office and if you're not, you're going to have a hard time keeping up with what the new requirements for the API are, then maybe the inevitability is that you will have these big firms that are able to vertically integrate.
Austin Vernon
For these big firms to survive, they have to be somewhat decentralized within them. You're going to the same place—what does it look like? Even if it's a giant corporation, it's going to have very independent business units as opposed to a 1950s corporation.
Dwarkesh Patel
Burn Hobart, by the way, has this really interesting post that you might enjoy reading while you're writing that post. It's about type-safe communications and that Bezos thing about how strict his style was for how to communicate and how little to communicate. There are many examples in Amazon protocols where the only way you can put in this report is in this place, and you have to give a number. You can't just say, "This is very likely." You have to say, "We project X percent increase." It has to be a percent. There are many other cases where they're strict about what type definition you can have something have in written reports. It has the same consequence that type-strict languages have, which is that you can keep track of what the value is through the entire chain of the flow of control. You've got to keep work content standardized.
We've been hinting at the Coasean analysis for this. For the people who might not know, Coase has this paper called "The Theory of the Firm." He's trying to explain why we have firms at all. Why not just have everybody compete in the open market for employment, for anything? Why do we have jobs? Why not just hire a secretary by the day? The conclusion he comes to is that by having a firm, you're reducing transaction costs. People will have the same knowledge about what needs to get done. You're reducing the transaction cost of contracting, finding labor, etc. The conclusion is that the more transaction costs are reduced within a firm as compared to between different firms, the bigger firms will get. That's why the implication of your argument was that there should be bigger tech firms.
Austin Vernon
Yes, definitely. They can decrease the transaction costs faster within. Even at the limit, if you have large transaction costs outside the firm between other firms that are artificially imposed, it will make firms bigger.
Dwarkesh Patel
So what does the world look like in that scenario? Will we just be like these Japanese companies, these huge conglomerates where you rise through the ranks from the age of 20 until you die? Is that what software will turn into?
Austin Vernon
It could be. I think it will be lots of very large companies unless there's some kind of change in firm transaction costs. That could possibly come from blockchain-like technology, but you probably also need better regulation to make that cheaper. Then you would have smaller firms. But in the end, it doesn't really matter. You'd be working in your little unit of the big bank or corp. I don't know what that would look like on a personal level.
Dwarkesh Patel
Speaking of these Japanese companies, let's talk about car manufacturing and everything involved there. We hinted at a few elements of the Toyota Way and lean production earlier. Do you want to give a brief overview of what that is so we can compare it to potentially other systems?
Austin Vernon
All these lean Toyota process systems have a lot of similarities. Mostly, you want to even out your production so you're producing very consistently. You want to break it into small steps. You want to limit the amount of inventory in your system. When you do this, it makes it easy to see how the process is running and limit defects. The ultimate goal is really to reduce defects because they're very expensive. That's my best shot at a quick summary.
Dwarkesh Patel
The interesting thing about the Toyota system—at least when *The Machine That Changed the World* was released in the 90s, they went into the history of Toyota. There was a brief time after World War II where the company ran into troubles. They needed to lay off people to not go bankrupt. They had much more debt on books than they had assets. They wanted to lay off people, but obviously the people were not happy. There were violent protests. The U.S.-written constitution gave strong protections to labor that they hadn't had before, which gave labor an even stronger hand. Toyota came to an agreement with the unions that they'd be allowed to do this one-time layoff to get the company on the right track, but afterwards, they could never lay somebody off. A person works at Toyota from the time they graduate college or high school till they die.
That's super intense in a culture. In software, the average tenure at a company is like one year. The difference is so much. There are so many potential benefits, but also drawbacks. If you're talking on a time scale of 50 years rather than one year, the incentives are way more aligned between the company and the person. Anything you could do in one year is not going to have a huge impact on your stock options. But if this company is your retirement plan, you have a much stronger incentive to make sure the company runs well. You're probably optimizing for the company's long-term cash flow. There are also benefits to having that knowledge build up in the firm from people who have been there for a long time.
Austin Vernon
I think there are diminishing returns to how long your tenure is going to be. Maybe one year is too short, but there's a certain extent where if you grow faster than your role at the company, then it's time to switch. Maybe five years is a good number. If you're not getting promoted within the firm, your human capital is being wasted because you could go somewhere else and have more responsibility.
Another interesting thing about that story is that almost all lean turnarounds come with no-layoff promises. If you're going to increase productivity, everyone thinks, "Oh gosh, I'm going to get laid off." Instead, you have to increase output and take more market share. It's kind of like burning your bridges. The process really requires complete buy-in because a lot of your ideas for how you're going to standardize work content come from your line workers—that's what they're doing every day. If you don't have their buy-in, it's going to fail. That's why it's really necessary to have those kinds of clauses.
Dwarkesh Patel
That makes sense. Was it in your post or in the book where they talked about how if somebody makes their process more efficient and therefore they're getting more work allotted to them, they're going to stop doing that? Do you have to give more downtime to your best workers or the people who are most creative in your company?
Austin Vernon
For a worker at a plant, small rewards work pretty well. A lot of people on drilling rigs used to get $100 Walmart gift cards if they met certain targets. Sometimes small rewards for new ideas work. But because the whole system has to grow together, if you just improve one part of the process, it may not help. You have to be improving all the right processes. Normally it's much more collaborative. There's some engineer looking at it, saying, "This is where we're struggling or we have our defects here." Then you get together with the supervisor and the workers in that area, and you all figure out what improvements could be made together. Usually the people already know. You see a problem at the top and you're just now realizing it. You go talk to the people doing the work, and they're like, "Oh yeah, I tried to tell you about that two weeks ago, man." Then you figure out a better process from there.
Dwarkesh Patel
Based on your recommendation and Stephen Molina's recommendation, I recently read *The Goal*. After reading the book, I'm much more understanding of the value that consultants bring to companies. Before, you could think, "What is a 21-year-old who just graduated college? What do they know about manufacturing? What are they going to tell this plant that they didn't already know? How could they possibly be adding value?" Afterwards, it occurred to me that there are so many abstract concepts that are necessary to understand in order to increase your throughput. I can see how somebody who's generically smart but doesn't have that much industry knowledge might be able to contribute to a plant.
Austin Vernon
This applies to consultants or young engineers. A lot of times you put young engineers right in the thick of it, working in production or process right on the line where you're talking to the workers the most. There are two advantages. One, the engineer learns faster because they're actually seeing the real process. Two, there are easy opportunities for them to still have a positive impact on the business because there are just $100 bills laying on the ground from going up and talking to your workers, learning about stuff, and figuring out problems they might be having that could help you lower costs.
I think there are a lot of consultants. I know Accenture has like 600,000 employees or something—it's a large number. A lot are doing more basic tasks, and then there are some people doing the higher-level stuff. There's probably a lot less of that.
Dwarkesh Patel
There was a quote from one of those books that said, "We don't consider you an engineer unless you need to wash your hands before you can have lunch."
In your blog post about car manufacturing, you talked about Tesla. What was really interesting is that in a footnote, I think you mentioned that you bought Tesla stock in 2014, which might be interesting to talk about when we go to the market and alpha part. You talk about Tesla using something called "metal manufacturing." First of all, how did you know in 2014 that Tesla was headed here? And what is metal manufacturing, and how does it differ from the Toyota production system?
Austin Vernon
I was just goofing around and made that up. Someone actually emailed me and said, "Hey, what is this metal manufacturing? I want to learn more about this." I had to say, "Sorry, I just kind of made that up. I thought it sounded funny."
The idea is that there's this guy, W. Edwards Deming. He founded a lot of the same ideas that Toyota ended up implementing. Toyota respected his ideas a lot. America, except for the software industry recently, never really got fully on board with them in manufacturing. Now you have software people coming and implementing this in manufacturing. It's the real American way of doing things. When you look at these manufacturing processes, the best place to save money and optimize is before you ever build the process or the plant—very early on.
If there's a criticism of Toyota, it's that they're optimizing too late and they're not creative enough in their production technology. They're very conservative. That's why they have hydrogen cars and not battery cars, even though they invented the Prius, which was the first large-sales hybrid. Tesla is really making Deming's ideas their own and Americanizing it. They say, "We want to cast this because that'd be easier." "Well, we can't because we don't have an alloy." "Well, invent the alloy." I love it. It's great.
Mostly I just like Tesla because I agree with their engineering principles. I didn't know the company would come to be so valuable. I was just always reading their stock reports. I thought, "I need to buy some stock so I have a justification for spending all this time reading their 10-Ks."
Dwarkesh Patel
I want to get a bit more detail on the exact difference. Lean production is able to produce cars without defects and matching demand. What is it about their system that prevents them from making the kinds of innovations that Tesla is able to make?
Austin Vernon
It's just too incremental. It's so hard to get these processes working. The faster you change things, it becomes very difficult to change the whole system. One of the advantages Tesla has is that if you're making electric cars, you have a lot less parts, so that makes it easier. Once you start doing the really hard work of basically digitizing—they don't have speed limit dials—you start removing parts from the thing, and you can actually start increasing your rate of change even faster. It makes it hard to get behind if you have these old dinosaur processes.
There's a YouTube channel called The Limiting Factor that went into the detailed numbers on what it costs for Tesla to do their giga casting, which saves tons of parts and deletes thousands of robots from their process. If you already have an existing stamping line where you're just changing the dies based on your model, it doesn't make sense to switch to casting. But if you're building new factories like Tesla is, it makes sense to do the casting. You can build new factories very cheaply and comparatively much easier. They have lots of technical debt, I guess you could say in a software sense.
Dwarkesh Patel
That's super interesting. The analogy is quite apt. Microsoft probably has tens of thousands of software engineers who are basically servicing its technical debt and making sure the old system runs properly, whereas a new company like Tesla doesn't have to deal with that.
The thing that's super interesting about Tesla is that its market cap is way over a trillion, and Toyota's is like 300 billion. Tesla is such a new company. The fact that Toyota, which is legendary for its production capacity and production system, is worth many times less than this company that's less than two decades old—it's kind of funny.
Austin Vernon
In that measure, I don't like market cap. You need to use enterprise value. These old car companies have so much debt that if you look at enterprise value, it's not so jarring. GM is worth like 40 billion, and they have like 120 billion dollars in debt. Their enterprise value is like five times more than their market cap.
Dwarkesh Patel
What is enterprise value?
Austin Vernon
Enterprise value is basically the value of the actual company before you have any claims on it. It's the market cap plus your debt, most simply. If you're the equity holder and the company gets sold, you have to pay the debt first. You only get the value of what's left over after the debt. That's why market cap is... Tesla has very little debt and a lot of market cap, and these other guys have a lot of debt with less market cap. It excuses the comparison.
Dwarkesh Patel
One of the interesting themes across your work is that automating processes often leads to decreased eventual throughput because you're probably adding capacity in a place that has excess capacity, and you're also making the money-making part of your operation less efficient by having it interface with this automated part. It sounds like there's a similar story there with car manufacturing.
Austin Vernon
If we tie it back to what we were talking about earlier, automation promotes local optimization and premature optimization. A lot of times, instead of automating a process to make a really hard-to-make part, you should just figure out how to make that part easy to make. After you do that, it may not even make sense to automate it anymore, or you get rid of it altogether. Then you just delete all those robots.
Dwarkesh Patel
Okay, let's talk about the project you're working on right now: CO2 electrolysis. Do you want to explain what this is and what your current approach is? What is going on here?
Yeah, so I think just overall, electrofuels right now are super underrated because you're about to get, hopefully, some very cheap electricity from solar, or maybe some wind, possibly even nuclear or geothermal if we get really lucky. It will make sense to make liquid fuels or natural gas or something just from electricity and air essentially. There's a whole spectrum of ways to do this.
Austin Vernon
CO2 electrolysis is one of those. Basically, you take water, electricity, CO2, and a catalyst, and you make more complex molecules like carbon monoxide, formic acid, ethylene, ethanol, or methane. Those are all options. But it's important to point out that right now, if you added up all the CO2 electrolyzers in the world, you'd be measuring their output in kilograms per day. The products I just mentioned, we make millions of tons per day. So there's a massive scale-up needed if it's going to have a wider impact.
There's some debate for the whole electrofuels sector about how much you're going to do in the electrolyzer. One company I really like their approach, which is different than mine, is Terraform Industries. They want to make methane, the main constituent of natural gas. But they're just making hydrogen in their electrolyzer, then capturing the CO2 and putting it into a methanation reaction. Everything they're doing is already world scale. We've had hydrogen electrolyzers power fertilizer plants, and methanation happens in all ammonia plants. It's well known, very old. Methanation is hydrogen and CO2 combining to make water and methane.
Their approach is more conservative. But if you do more in the electrolyzer—like I'm going to make the methane actually in the electrolyzer instead of adding this other process—you could potentially have a much simpler process with less capex that scales downward better. You don't need traditional chemical engineering that heavily favors scaling. With the more Terraform-like processes, they're playing with absolutely ginormous factories, and those can take a long time to build. They're having to fight the complexity that creeps into chemical engineering every step of the way, because if they don't, they'll end up with a plant that takes 10 years to build. That's not their goal. It takes 10 years to build a new refinery because they're so complex.
So that's kind of where I am. I'm more on the speculative edge. It's not clear yet which products will be favorable for which approaches.
Dwarkesh Patel
Okay, yeah. And you're building this out of your garage, correct?
Austin Vernon
Yeah. That's where the electrolyzers come in. Everything with electrochemistry is a flat plate instead of a vessel, so it scales down. I can have a pretty good idea of what my 100 square centimeter electrolyzer is going to do if I make it quite a bit bigger. I just have to worry about how my flow might interact in the larger one and make sure the mixing is good. But it's pretty straightforward because you're just making your flat plate a larger area. The scale is different than scaling a traditional chemical process.
Dwarkesh Patel
I'm curious how cheap energy has to get before this is efficient. If you're turning it into methane or something for fuel, is the entire process energy positive? How cheap would electricity need to get before that's the case?
Austin Vernon
The different products and different methods have different crossovers. Terraform Industries is shooting for $10 a megawatt hour for electricity. Their process is simpler, a little less efficient, and a lot of the other products have better premiums—they're worth more per ton than methane. So your crossover happens somewhere between $10 and $20 a megawatt hour. Solar right now is maybe $25, maybe a little higher because panel prices have gone up in the last year. I think the expectation is they'll come back down. Getting down to $15 where you start having crossovers for some of these products like ethanol, ethylene, or methanol is not science fiction.
Dwarkesh Patel
I think in Texas where I live, that's where it's at. The cost of energy is like $20 or something for a megawatt hour. Well, not this summer, but recently. A lot of times in Texas, the wholesale prices are around $25 to $30.
Austin Vernon
Gotcha.
Dwarkesh Patel
Okay. A lot of the actual details about how this works went over my head. What is a flat plate? Before you answer that, can you just generally describe the approach? What are you doing to convert CO2 into these other compounds?
Austin Vernon
It just looks like an electrolyzer. You have two sides, an anode and a cathode, and they're just smushed together like this because the electrical resistance, if you put them far apart, uses up a lot of energy. So you smush them together as close as you can, and you're basically just trading electrons back and forth. On one side, you're turning CO2 into a more complex molecule, and on the other side, you're taking apart water. When you take apart the water, it balances out the equation and your electrons. I probably need to work on that elevator pitch.
The basic idea is you need to put power in to convert CO2 into these other compounds. The inputs are electricity, water, and CO2. The output is usually oxygen and whatever chemical you're trying to create, along with some side reactions.
Dwarkesh Patel
These chemicals you mentioned—ethanol, methane, formic acid—are these all just fuels, or what are the other uses for them?
Austin Vernon
A lot of people are taking a hybrid approach with carbon monoxide. Twelve Co has raised a lot of money to do this, has like a hundred employees. You can take that carbon monoxide and make hydrogen, and then you have syngas to make liquid fuels. They want to make all sorts of chemicals, but one of the main volume ones would be jet fuel. Formic acid is the small fry of all these. It's an additive in a lot of things, like preserving hay for animals. Ethanol—there are people that want to use it. There's a company that makes ethylene, which goes into plastics, like polyethylene, the most produced plastic. Or you can burn it in your car, although I think ethanol is a terrible vehicle fuel. You can also just make ethylene straight in the electrolyzer. There are many paths. Which path wins is an interesting race to see.
Dwarkesh Patel
The ability to produce jet fuel is really interesting. In your energy superabundance paper, you talk about how even if we can electrify everything with solar and it becomes super cheap, that's not going to have an impact on the prices to go to space, for example. But if a process like this is possible, it's some way to add liquidity and turn this cheap solar and wind into jet fuel through this indirect process. So the price to send stuff to space, or to have cheap plane flights, all of that goes down as well.
Austin Vernon
It basically sets a price ceiling on the price of oil. Whatever you can produce this for is the ceiling now. That's the way I think about it.
Dwarkesh Patel
Do you want to talk a little bit about how your background led into this project? This is your full-time thing, right? Where did you get this idea, how long have you been pursuing it, and what's the progress?
Austin Vernon
I've always loved chemical engineering. I loved working at the big processing plant because it was like a kid in a candy store. If I had extra time, I'd just walk around and look at the plant. That was so cool. But the plant where I worked had an uptime of like 99.7%. If you wanted to change anything or do anything new, it terrified everyone because they earned their bonuses by running the plant at 100% uptime all the time. That just wasn't a good fit for me.
I always wanted my own chemical plant, but it's billions of dollars to build plants. That's a pretty big step. With this new technology, there's a window where you might be able to build smaller plants until it optimizes to be hard to enter again.
Dwarkesh Patel
What will happen when it becomes hard to enter again?
Austin Vernon
If someone figures out how to build a really cheap electrolyzer and they keep it as intellectual property, it would be hard to rediscover that and compete with them.
Dwarkesh Patel
How long have you been working on this?
Austin Vernon
About not quite a year. I actually got the idea to work on it from writing my blog. When I wrote the heating fuel post, I didn't really know much about another company in the space, Prometheus Fuels. I thought, "Oh, this is an interesting idea." Then I got talking to a guy named Brian Helgman, and he said I should do this, but not like what Prometheus is doing. So I started looking into it and liked it. I've been working on it since.
Dwarkesh Patel
It's interesting because if energy does become as cheap as you suspect, and if this process works, then this is like a trillion dollar company, probably. If you're going to get the patents and everything.
Austin Vernon
With chemical plants, there's a certain limitation. There are only so many places that can have good locations for chemical plants. You start getting hit by transportation and all that. You can't just produce all the chemicals for the entire world in Texas and transport it all around. It wouldn't work. So you're talking about a full globe-spanning thing. If you're building factories all over the world, someone is going to figure out what your intellectual property is. You would have to keep innovating to stay ahead of competitors. Ultimately, it's a commodity, so you don't have the same kind of defensibility that other sectors do.
Dwarkesh Patel
I see. There's not like network effects.
Austin Vernon
Yeah. So what happens is the scale starts increasing as you go on. Even though this is easier to scale down, there are certain elements that are very hard to scale, and the organization as well. You only need a few competitors. Early on, a few competitors will continue to grow against each other and limit the margins. It would be hard to be the fifth competitor 30 years down the line.
Dwarkesh Patel
What is the state of this project right now? Are you planning on starting a company? What are the milestones you're shooting for?
Austin Vernon
It is just me. I have a family of engineers, so it's loosely supported by other people in my family who are participating some. Basically, I've already done a lot of the theoretical design work at a very cursory level to make sure it makes sense and the cost will be reasonable. Now I'm working on the electrolyzer to meet the targets for reliability, product concentration, and energy costs. Also, is it manufacturable? Right now, a lot of electrolyzers they use in the labs are literally smaller than a postage stamp and very difficult to make.
Dwarkesh Patel
Had you started working on this before or after you quit your job?
Austin Vernon
Oh, after. I quit my job like five years ago. I was doing software stuff in between.
Dwarkesh Patel
What did you work on?
Austin Vernon
I worked on several products. I have one that's a data service for oil and gas that's somewhat successful. It has paying customers but is still relatively small.
Dwarkesh Patel
Your blog is pretty recent. You started about a year ago. What encouraged you to do that?
Austin Vernon
I was curious about cryptography in general, specifically for blockchains. I wanted to be able to read the Bitcoin white paper and understand some of this, like IPFS. I figured the best way to do it was to write. People talk about how you should write, so I created an IPFS blog. I learned a lot. It was not the most reliable blog when I was running it on my own droplet, so thankfully I migrated to a service with much more uptime.
I wrote several posts to learn about hash functions and private key cryptography so I could understand the white papers and what they were doing with the math. Eventually, I had this blog. My first non-crypto topic was on how to build a cheaper house and how it's difficult to reduce home construction costs. It made it on Hacker News, and I thought, "Maybe people actually want to read this stuff." So I've just been writing since then in my spare time.
I interned for Protocol Labs, which built IPFS. I got a chance to learn a lot about it. Trying to learn how Filecoin exactly works threw me into a whirl for a while. It's really interesting. I actually had a blog on IPFS, kind of a toy thing, not the one I ended up writing on. Obviously, at the moment, nobody else is going to seed it for you, so you have to use a centralized service like Pinata. But it's a fun exercise.
I was running it off a droplet on Digital Ocean. If you use the direct content hash, it works pretty well even if you're linking through your ENS name. The problem was when I was first doing this, the fees on Ethereum were so high that I didn't want to change that link all the time. I tried to use the pinning feature with IPNS and going through CloudFlare's eth.link. They look up your IPNS name and try to find it. The part that was breaking for me was CloudFlare couldn't always find my server using IPNS. The service I'm using now, Fleek, goes directly to the content hash. On DNS, it's cheap to change. You can change it in one minute. If Ethereum fees got lower, I might switch back. If it's one cent transactions, it would be no big deal to just change the content hash every time you update your website.
Dwarkesh Patel
What is the reason for having it on Ethereum?
Austin Vernon
Just for fun. It is inconvenient. If your content hash is changing every time you update the website, you have to keep re-updating where people can find the site or use some other service to take care of it. If transactions are cheap, you can automate it all and it costs you a little bit each time. It would be fine. But it was like $50. I'm not going to pay $50 to post a blog post. Then you find that typo and it's like, "Oh gosh."
Dwarkesh Patel
Let's talk about your paper with Eli Dorado on energy superabundance. You have lots of interesting speculation in there for what might be possible if energy gets a lot cheaper. Let's jump into it. Per capita energy use since the 1970s has not gone up. Before that, there was this thing called the Henry Adams curve where per capita energy use would increase 2% a year. After 1970, that was no longer the case. Ironically, right after the Department of Energy was created. But we've still had economic growth since the 1970s. It's been slower, but even though per capita energy hasn't increased, per capita GDP has increased. In the paper's introduction, you talk about why increasing energy use is necessary for increasing economic growth. But doesn't that pattern suggest you can still have decent economic growth without having to use energy? Or have we just not come across the constraints yet?
Austin Vernon
I think you just have diminishing returns. There are physical limits to how efficient things can be, and as you get closer to that efficiency limit, it gets harder and harder and takes more and more effort. So there are diminishing returns.
A perfect example is what we were just talking about with oil and natural gas. Oil is quite expensive, but easy to transport. You can produce it anywhere in the world and get it anywhere else pretty cheaply. Natural gas is extremely expensive to transport, but it's a very useful fuel for making fertilizer or anything else. Not everyone has natural gas or the economic capability to extract it using traditional processes.
If you can just build these natural gas factories where you're using sunshine and water and air, then all of a sudden everyone has access to natural gas, even if you weren't blessed with easily obtainable natural gas reserves. There's really a whole story about the tyranny of geography when it comes to energy. Some countries have extreme electricity use per capita, like Iceland and Norway, where they have crazy amounts of hydropower and people build aluminum plants there. But then you have places in Africa where they have no coal, very little gas. They're just energy starved. Their transportation system sucks. You can't transport coal in. The hydropower is there, but only so much of it, and it may not be close to where their cities are.
If you start adding solar to the mix for them, along with some of these other technologies, it could be an incredible increase in energy availability. They aren't even meeting what we talked about in the paper. We were looking at doubling rich world use, but it would be more like 10X for someone in Africa.
Dwarkesh Patel
If energy becomes that abundant, does the bottleneck for civilization become the resources that are the backbone of the things the energy is doing? The actual raw materials necessary to build the factories? Or to what extent can even that be overcome?
Austin Vernon
I would argue the ultimate limit is human capital. What more abundant energy does is allow you to redeploy human capital away from trying to figure out how to use scarce energy sources and just let you waste some of it now.
Here's an example I love about trucking. I love trucks, not as big a fan of freight trains. Freight trains are extremely efficient. They use very little fuel. But if you're going to use a train, the train doesn't come by all the time. They may not hold to the schedule. You have to aggregate your product with other stuff or your raw materials, and it adds a lot of cost to your production. The Toyota production system runs on trucks, not trains, because the truck is extremely flexible. It comes when you need it, it goes when you need it.
When you have cheaper energy, with electrification and automation of trucking, you are going to shift a huge amount of goods from trains to trucks. It's going to have huge knock-on effects all across the economy. It enables more specialization. There are a lot of products where you're limited on your suppliers because transportation is expensive. It reduces working capital because a lot of times it takes longer on trains. Similar stuff with smaller ships, more air freight.
One thing that shocked me is Eli was telling me about how the elasticity of demand for air freight is just insane. When you decrease the cost a little bit, demand goes to the roof. You always think, "Oh, we can't do this with batteries," and then someone comes up with a more clever idea. Even if you have a 500-mile range limit for your freight plane, the freight doesn't care if you have to stop every 500 miles to refuel or recharge. You can go over land on almost all these routes. You could go up through Japan and the Aleutian Islands, or you could go over land from China to Europe and charge wherever's convenient.
If that electric plane has half the operating cost of the jet plane, the amount of freight you're moving on airplanes will go way up. It'll go down on ships, and everyone will be better off. Right now, if you're a shipping company, you have real working capital problems because your stuff sits on a boat for a month. You've got to finance that, and if things change in the meantime, you don't really want that product anymore. Air freight is an absolute economic booster. If you can make it cheaper, it's really exciting, but it uses way more energy.
Dwarkesh Patel
An analogy that just occurred to me: if Moore's Law had stopped in 2005, we would still have a lot of interesting applications using compute, and the effects of the computer would still have permeated society. But obviously, a lot of things that are possible today with computers just wouldn't have been tried or been possible in that kind of world.
Austin Vernon
All your engineers would be working on optimization instead of building new products.
Dwarkesh Patel
In Jason Statham's new book—we're not putting you at this point—but in his book *Where's My Flying Car?*, one of the points he makes is that GDP growth has probably been overstated because a lot of what constitutes growth has just been increasing the efficiency of existing machines to make them use less energy. That doesn't result in more total resources or goods or services being produced.
Austin Vernon
Instead of making the laundry machine more efficient, you can just create a new kind of machine that may need to use more energy.
Dwarkesh Patel
For this vision to come to pass, do you need energy to just become super cheap, or do you need advances in the ability to store that energy as well?
Austin Vernon
If lithium batteries are the bottleneck, it doesn't matter if you can get energy super cheap if you can't put them in appliances, cars, planes, or whatever. The important thing to think about is that our current energy is so expensive, especially electricity. With our energy resources, which are basically third-world, it's quite difficult to make electricity comparatively. So what we use electricity for is stuff we really want to use it for. It's hard to imagine we're not going to turn our air conditioner off, so we're willing to pay a lot of money for that electricity.
If you look closely at a lot of the use cases that use tons of extra energy, they're much more flexible in how they use the energy. There's not a whole lot of storage involved. If you're growing crops, making methane for rocket fuel, or making chemicals, you can design these processes to run when the energy is available. The batteries are really going to be for keeping your air conditioner on, where you're willing to pay a lot of money. I don't really see the batteries and storage as a limit.
Dwarkesh Patel
What about something like air freight? Wouldn't you need some way to store that electricity for air freight, or can you just convert it to jet fuel?
Austin Vernon
I was thinking more like grid storage. In transportation, that's going to dominate battery demand. Grid storage is tiny in comparison. But you're basically getting to the point where you're making batteries out of dirt because that's how you scale it. If you're making batteries out of carbon, iron, and phosphate, it's just how many battery factories you want to build. There's plenty of lithium, you just have to build the lithium mines. I don't see any hard limits there. Once you build all the factories, you're pretty much ready to go.
Dwarkesh Patel
Is the point you're making with alternative batteries that even if they're worse than lithium batteries, we'll have so much energy that it doesn't matter? Even if we lose a lot of it, that's fine, we'll just use whatever we can take? Or are you saying we can produce batteries with other chemistries that are as good or better?
Austin Vernon
Right now, the shortage is really nickel. In the very short term, lithium is starting to be kind of a shortage, but there's plenty of lithium. Lithium iron phosphate is seeing a huge amount of substitution right now because it avoids nickel. It's not quite as good as some of the nickel chemistries, but for a lot of applications like cars, it just doesn't matter. The aircraft and stuff will pay the premium for high energy density batteries.
Eventually, there are technologies that use less and less materials because they're just better batteries, like some of the concepts around solid state. I'm not sure if those will come to fruition or if they'll be that much better when they do, but there are lots of opportunities for substitution down the line.
Dwarkesh Patel
What is solid state by the way?
Austin Vernon
Right now, all our batteries charge and discharge through the lithium ion going back and forth between the cathode and the anode. It travels through a liquid, and that liquid is an electrolyte, which means ions can travel through it. Solid electrolytes are a little more challenging, which is why we don't have them. You get rid of the liquid, and the ion has to travel through a solid instead. The promise is it could be a much higher energy density and theoretically cheaper too, just because it uses way less stuff. But there are all sorts of problems. They degrade faster. Batteries have six different areas you have to hit the requirements on, and if you miss one, it's no good. They're kind of hard to improve in that sense.
Dwarkesh Patel
If energy superabundance is going to come from solar and wind, which are intermittent sources, you would need progress in battery storage, right? That's contingent on that?
Austin Vernon
A lot of the extra energy uses we talk about don't really require many batteries, if any at all. Transportation yes, you have batteries in the vehicles. But if you're going to have abundant nuclear electricity or abundant geothermal electricity, you still have to build all those electric vehicles and you still need the batteries for that. The extra batteries that solar and wind require over geothermal, I think it ends up being pretty minimal.
The way I think about it is, if you can have a solar farm that's going to give you $10 per megawatt hour electricity, you just have to figure out how to utilize that. If you do, you'll be very rich. You'll beat the guy who's paying $40 per megawatt hour from the more expensive traditional generators.
Dwarkesh Patel
Before we get into which sources of energy are most promising, let's talk about some of the other implications of energy superabundance. We talked a little bit about travel. One thing that might be concerning with air travel for passengers is if the bottleneck step is TSA and other regulations. To what extent will reducing travel time or increasing flight speed or number of flights have an impact on how much time you have to spend in an airport or in transit?
Austin Vernon
Right now, Airbus has this super jumbo thing. None of the airlines really loved it because it's too big. It's too hard to get everyone loaded and unloaded, and you hit diseconomies of scale. Electric planes are likely to be tiny in comparison, like 10 passengers. That's easier to load and unload, and you're going to fly out of smaller airports. You won't be going to this giant regional airport with all the parking problems and all the security. You'll be driving to your neighborhood general aviation airport where there's a small line to get through. A lot of these small aircraft, under certain situations, even avoid some of the screening requirements because they're just not as dangerous. If you have a small plane, there's only so much damage you can do with it.
Dwarkesh Patel
I did not know that. I've got to start booking planes from the small airports to avoid the TSA.
Austin Vernon
It's very nascent, but there are some business models coming down from the NetJets style to a little more commercial. They're trying to hit a price point similar to first class, but you get to avoid all the airport craziness. I'm a believer that if that existed, people would get angry enough that they would loosen up a lot of the rules. It seems impossible to change those rules now, but the average person it costs them no time because most people don't even fly very much.
Dwarkesh Patel
What is your vision for what a city could look like if energy got a lot cheaper? In the paper, you have all kinds of interesting projections about drones, electric deliveries, and the entire congestion of the 3D space, and with tunnels as well. What does the city look like with energy superabundance?
Austin Vernon
Basically, you disaggregate the car to a certain extent. Inner city car trips become less common because cheap flying is going to be cheaper and more convenient. The bots will deliver your stuff. I love the tunnels because I don't like taking people's land. With tunnels, you can run new roads without imminent domaining and taking people's land away from them when they don't want to lose it. It makes people so angry when you take their land that it's very expensive to eminent domain people because they will fight you until the sheriff has to show up and haul them away.
If you can go around that with tunnels using existing right-of-way, it makes that societal cost significantly less expensive. Then it's just the engineering challenge. I think there's really an opportunity now. The Boring Company is the famous one, but recently I think there's a Hacker News company that wants to do tunnels for electricity. They have this plasma boring machine concept. It seems pretty crazy right now, but it's one of those solutions that reduces coordination costs across the whole economy and improves property rights. People should really try to build it.
Dwarkesh Patel
You mentioned one of these machines in your blog post on tunneling. It was the SpaceX one. I forgot the name of it.
Austin Vernon
Proof Rock, yeah. It's pretty big, but it's apparently all electric, which is kind of insane. It can just do it in one step.
Dwarkesh Patel
How is it getting the material out?
Austin Vernon
The problem with most tunneling is in soft soil. It's kind of difficult to drill through soft soil because of the materials handling. When you first start drilling an oil well through this stuff, you actually have to limit your drilling speed. You don't even have to put any weight on the bit, just pumping fluid around basically jets out the fluid. That's what you're doing with the boring machine in soft soil. Managing the spoils, they have muck carts a lot of times. Maybe it's basically trying to use a conveyor belt, but you could also just make it a full liquid and pump it out like in the oil field. We carry our cuttings in mud and we pump it.
The other big challenge is they have to keep the walls from caving in on them. Current boring machines in soft soil spend an enormous amount of time erecting these tunnel supports that keep it from collapsing. It's kind of counterintuitive, but it's dramatically faster to bore in hard rock than soft soil because in soft soil you spend so much time on non-productive time. In hard rock, you're just blowing it going.
Dwarkesh Patel
To get back to cities, you mentioned something in the paper about Marchetti's Constant, which is that the amount of time people spend in transport per day is the same. If you increase the speed at which they can move, then they have a wider surface area in which they can explore.
Austin Vernon
Yeah. I don't know if physically the cities will look that much different, but their effective economic size will be much larger. You could live in Cedar Rapids and commute to Minneapolis with some of these technologies. Your city in Cedar Rapids still looks the same, but you don't have to work there. If you have a better job in Minneapolis, you could commute there three times a week or five days a week.
Dwarkesh Patel
If commute time stays the same and people just get more spread out when energy becomes cheaper, do neighborhoods and cities become this unwalkable mess out of a Jane Jacobs nightmare if the conglomeration goes away?
Austin Vernon
I think it's actually the opposite. If you have tunnels and alternative methods to cars, then you use cars less. In many cities, they never made sense for cars anyway because they were built before cars. In New York City, you're never going to move everyone around in a car unless you build tunnels. But even then, there are other technologies that make a lot of sense. People like walkable neighborhoods. Even though I live in a city that requires a car, some of the hottest neighborhoods are walkable neighborhoods. The neighborhood is walkable itself, and then you just drive your car to wherever else you need. The car is hidden within the neighborhood.
Dwarkesh Patel
Maybe we'll see more segregation, not in the racial sense, but in the sense that people will prefer to live in these walkable neighborhoods but have no problem commuting to work using a VTOL or something. You would end up seeing these walkable neighborhoods and industrial zones that are way far away distance-wise, but not that far away time-wise.
Austin Vernon
Right. It's the same if you want to live in a small town that would be too far to commute to a city now, but you could in the future. More choice.
Dwarkesh Patel
What is holding back VTOLs? VTOL is vertical takeoff and landing. The reason you need to go to the airport is because you need a large landing pad. The hope is that if you could vertically take off, you could lift off from your roof. We've had prototypes of this kind of stuff since the 30s. Why don't we have these widely available? Is energy the constraint, or is it something else?
Austin Vernon
In the past, liquid fuels are theoretically dense enough, but they're too complex and too expensive. When you're turning heat energy into mechanical energy, a lot of weight and complexity comes with that. Some of these old concepts have all these engines. If you electrify them, it really changes the game. Just now, because it's not just batteries, it's the motors and the inverters that are now getting dense enough and small enough to make sense. But it takes time to get this stuff through the FAA, for better or worse. The technology hasn't been good enough long enough to get stuff through the FAA.
There are some limitations. Right now, a lot of people want to use batteries, but the batteries are just on the edge of good enough. You're going to have a 50-mile VTOL, not a couple hundred mile VTOL. My dream VTOL application is a nuclear-powered quadcopter that carries a container. You could take the container directly from the factory in Vietnam directly to the people who are using it or the warehouse in Arkansas.
Dwarkesh Patel
Theoretically, you could have these drones carrying huge payloads weight-wise.
Austin Vernon
You wouldn't necessarily want a large payload. You just want whatever the customer wants. You want to size your vehicle to deliver that payload, that's the most efficient.
Dwarkesh Patel
What does this mean for computing? If energy gets a lot cheaper, Bitcoin mining becomes... well, it doesn't necessarily become more profitable because other people's energy is cheaper too. But what are the other consequences? Does spinning up an AWS server become trivial? Does building a deep learning model cost nothing in terms of GPU time? What impact does it have on computing?
Austin Vernon
I think the limitation would probably still be chips for a while, until you figure out a better production process for that. It'd be a while before it becomes energy. Smartphones really worry about energy. There could be some interesting things with smartphones if you have a very power-dense betavoltaic battery, like a nuclear battery, where you don't have to worry about running down your battery. But outside of smartphones, I'm not sure that energy is the limit for a lot of this computing.
One of the interesting things to speculate about at the end of the paper is a potential carbon shortage. In an email to Tyler Cowen that he published on his blog, you said by the end of the century we'll have a carbon shortage. Presumably because of the process you talked about earlier, the thing you're working on. If you can take CO2 out of the atmosphere, what is the probability that this ends up happening? Do you think it's more than 50% by the end of the century, or is it just speculation?
I think it's extremely high that it happens, and it's harder to put the timeline on it. By the end of the century, it might be a little... I think I ran some numbers in there. If you 10x current plastic production and you're just putting it all in landfills, I think it was a little over a hundred years to get through all the carbon, and you're assuming your rest of your carbon output is zero in that scenario. But it's probably pretty hard to do it by the end of the century without a lot of growth.
But it's kind of the exponential thing. A large number of carbon emissions have happened in the last 20 years, and it was very small before 1950. You could get surprised at the back half. The last 10 years go crazy. It makes it hard to predict.
Dwarkesh Patel
Yeah. By the way, in Will MacAskill's new book on longtermism, one of the things he speculates about is if society collapses and we need to restart, one of the things we'll need is coal or some other sort of dense, easy-to-use fuel. The problem is we've been burning up easily accessible coal, coal in places where you just dig up and find it. So one of the things he's concerned about is making sure we leave some easily accessible coal silos around so that in case society collapses, we can restart and use these to power up a second industrial revolution.
I wonder if you could use a process like this with carbon sequestration to actually just build up these kinds of reserves. If a longtermist is really interested in making sure we have that kind of resource, could they just use this process to do that?
Austin Vernon
Actually, there's a company called Charm Industrial, and they're basically doing that. They take trees and do a process called fast pyrolysis, where you burn biomass without oxygen in an anoxic environment, and it makes this bio-oil. Then they're injecting the bio-oil down into wells and selling carbon credits. So it's already happening.
You could say, "Oh wow, I knew that is easy to burn." You could burn it for heat, it's okay, but it's hard to refine. There were a lot of people that tried to do bio-oil as an alternative for petroleum like 20 years ago, Cleantech 1.0, and they all failed. So it makes me laugh that they're reimagining the process to sell what are right now very expensive carbon credits. But you could do something similar. You could actually just make straight carbon if you wanted to.
Dwarkesh Patel
I see. The thing I find interesting about this is, often in the case of global problems, people will early on identify that a thing is going to be a problem, but it often ends up being the case that they get the direction of the problem opposite. If you think about population, in the 70s people were correct that global population was going to be a problem. The thing is, it seems like now the problem is going to be that the population might decline too fast, not that it's going to grow exponentially.
I think this is another example of that kind of thing, where CO2 is going to be a problem either way. I'm not sure if it's going to be a problem that we overproduce it or we have shortages.
Austin Vernon
Yeah. If you just think out at the large scale, if you're going to be a Kardashev-scale civilization or you're using condensed amounts of energy, that's going to have side effects, and you're going to have to figure out how to manage that one way or the other. One of those is eventually Earth may just be a nature preserve and we all live in space or something.
Dwarkesh Patel
Yeah. Okay, let's talk about nuclear. It seems like you're much less optimistic about nuclear than you are about solar and wind. Do you want to explain why that's the case?
Austin Vernon
Yeah, well, especially solar more than wind. Wind, I think it's limiting because of transmission problems. If you want to build out huge amounts of wind, like some of these zero-carbon plans call for, you're going to have to take a lot of people's land to build transmission lines and stuff, and that really pisses people off. They fight hard, and it becomes expensive.
It's not like the wind turbines are easy to site. You pay people, and you'll actually see they never put above-ground power lines on the people's land where they put the wind turbines. They're always underground, so at least they get to the county right away. But when you get these giant transmission lines, like in the grain belt or something, they almost inevitably have to go across a lot of people's lands, and you can't just stuff them all in county and state right-of-ways because the pylons are so big.
Dwarkesh Patel
Sorry, what is a pylon?
Austin Vernon
The pylon is what holds the wire up, the tall tower.
So solar is much more flexible about where it can go. I think with solar getting cheaper, the obstacles are pretty simple. It's like, "Well, gosh, this is expensive to fill with racking. Why don't we just lay the panels on the ground?" Or, "Gosh, this glass we're encasing with is getting expensive, and we don't need it to last 80 years or 50 years. We can just put some plastic on it instead." Or, "We've gotten the actual photovoltaic cells so cheap, and all the other labor and stuff is getting more expensive. Well, why don't we just add another layer and make more energy?"
Those are your solar solutions to get down to $10 a megawatt hour, and they're pretty straightforward. Whereas with nuclear, it's like, "Well, the light water reactor can't get us there. Let's instead cool our reactor with sodium, which catches fire when it reacts with water and catches fire when it reacts with air." Or you could cool it with liquid lead. That's an option. Helium, which leaks a lot. Or you could do molten salts that corrode everything. We don't really have anything that...
So I think those solutions for very large reactors are pretty hard. There's a lot of reasons why. A lot of stuff just reacts poorly when you expose it to neutrons. They each have their own features that make them possibly good candidates.
I actually think regulation is kind of oversold a little bit. To the extent that people were internally consistent, they would see the NRC as a regulatory success story. The background on this is my wife's mom and stepfather are nuclear engineers. I've worked at all levels of the nuclear power industry, so I get to ask them general questions and learn a lot about it.
Back in the 80s, the nuclear power industry was in real trouble because their competitors in coal and natural gas got deregulated. Most of the cost of coal is the rail getting there, and the rail industry got deregulated. Then the natural gas industry got deregulated. So the cost of their alternatives was falling. And they had to build more safety into their plants because of all these incidents. It wasn't just Three Mile Island. It was Brown's Ferry, Rancho Seco, all these things that could have been really scary. To a certain extent, we got a little bit lucky that we didn't have a worse disaster. They were just relatively limited accidents at their sites.
There was actually a time where nuclear power plants were selling for less than what the fuel they had on their plant was worth. So what the industry did and what the NRC did is they moved to probabilistic risk assessment, which is usually the gold standard. People are really happy that we use probabilistic risk assessment for commercial crew with NASA and SpaceX. They want the FDA to use more probability, more expected value. What this allowed was basically rolling up some of the rules and moving into the risk assessment.
Around 1980, nuclear power plants only ran about 60% of the time. They weren't very reliable. They had all sorts of unplanned outages. The safest mode of operation is just running as designed, so the more consistent nuclear power is, the safer it is. The probabilistic risk assessment allows you to do repairs while you're running, which was discouraged before. Before, if your main cooling pump was leaking, you'd be like, "Oh gosh, I hope we can make it," and then it just fails and you shut down the reactor. Now it's like, "Alright, we have backups. The safest thing to do is actually repair it now while the plant's still running, get it repaired, and put it back online."
To give you an idea of the safety standards the NRC has, I think for the plant taking damage, it's like one time in 10,000 reactor years. For a large release, it's one in 100,000 reactor years. There are 93 operating reactors at less than 93 sites. So we should only see a Three Mile Island under the current standards once every hundred years or so, and a large release like a Fukushima type situation once every thousand years. But in just a few years in the 1970s, the industry had three or four of these damage events. So the safety is incredible. Now the operating capacity is up to over 90%. The plants are extremely reliable, and it lowers their cost because their costs are so fixed.
Dwarkesh Patel
Can you compare it to a country like France? They've had a lot of liability problems with their nuclear fleet in the last couple years. This year, their capacity factor might barely be over 60%. We have like 90 gigawatts of nuclear, they have 60 gigawatts. That makes a huge difference for Europe that those plants aren't running full out. You see a lot of charts about, "If Germany didn't shut down its reactors, what would the energy balance be?" But you don't see as many like, "If the French could run their reactors like American reactors, what would the energy balance be?"
I could go on about how that integrates into new plants if you want.
Yeah, I do. Because the line I've always heard on this from my bubble is, "Oh, they haven't approved a new plant. The NRC has not approved a new plant since it was created." I guess they just approved the design for the new small modular reactors, which I would love to hear your opinion on as well. So I'm very curious to hear this perspective.
Austin Vernon
Okay. So think about it. In the 1980s, you had new sources of fuel, new competitors. You also, by the end of the decade, increased the amount your nuclear power plants ran by a lot. A lot of these new power plants people were thinking about building were at existing sites, like an extra reactor at Watts Bar or whatever. You basically got a "buy two, get one free" by running your plant better. So you don't really need them as much. All those contributed to just not making sense to build new nuclear power plants because the existing fleet ran better, you had more competitors, and electricity demand slowed down.
Is it hard to get through NRC approval? Yes. That last one, the mini reactor you're talking about, took like 10 years or something. But when you think about a probabilistic risk assessment, no one ever says, "Well, gosh, the NRC's current standards of a large release that would happen once every thousand years, is that reasonable?" We're not arguing over that. We're just talking past each other. To me, that's a pretty reasonable risk level.
If you're going to 10x your reactors, that means almost certainly you'd have a Fukushima within your lifetime if you go with NRC standards. But it turns out it's pretty cheap to do way better. A lot of the reason why the plants weren't built may not have been because of regulation, but because market conditions changed. You had more competitors with coal and gas being deregulated. You also had increased production from the existing nuclear plants. If you were going to build an extra reactor at an existing site, you might not have needed to anymore because you got so much more production out of your existing plants. They shortened the fueling time, just a lot of all-around improvements paired with electricity demand flattening that really made new plants not economic or not necessary.
When we think about the probabilistic risk assessment, it just takes a lot of engineering time to get it through. If you look at how hard it was for SpaceX to get Falcon 9 and Dragon through NASA's loss of crew risk calculations, it took years, took hundreds of millions of dollars. It's kind of funny that people see that as a success, especially when the stakes were only a few lives for people that volunteered for danger. Then you have a nuclear power plant going through the same probabilistic risk assessment, and it could impact many more people's lives, and it's like, "Oh, that's not good enough."
I think it would make more sense to argue about the actual risk factor numbers, as opposed to just, "I'm mad that we're not building nuclear power plants." It actually becomes very inexpensive to improve the risk probabilities. The old plants we're running now have active safety systems, which means you have to maintain them and they have to work. If you want to move the control rods back into the reactor, there's a mechanism and a motor that does that that can fail. When you're calculating your risk, you have to calculate, "Oh gosh, what if this motor fails? What if my control fails? What if I don't have a properly trained operator to do it?" It's the same for the cooling systems.
But this new generation of plants has passive safety systems. Natural convection can cool the reactor in an emergency, or the rods are more like a dead man switch where if something happens, they just drop in from gravity. The new power plants, like the one that just got approved or the one they're building down in Georgia, can be orders of magnitude safer than the running plants. It's not really a huge cost increase. You're just changing how you do these things. In fact, all the literature for them says they're supposed to be less complex and easier to build.
But you're talking about a project so complicated it takes thousands of workers years and years to build, working every day. If you're going to go through and do the engineering in great detail to prove that your plan is safe under the probabilistic risk assessment, it's going to take hundreds of thousands of hours of engineering time. It's going to take a long time. That's why you see investors willing to pay at this point, but after you build it, because of the Rankine cycle and all that, is it going to generate economic power? It's not necessarily going to.
One way to think about this is my father-in-law always says, when people ask him about why we're not doing more nuclear, he says, "You got to think about the politics first, and the economics second, and those are the important ones." People are submitting designs and wanting to build plants that are big enough to impact lots of people's lives. Even if that risk is very low, some people are still bothered by that. Also, they're selling an easily substitutable commodity in most cases. On the political side, if you can substitute nuclear power, people will, even if it's coal or whatever. People don't really care that much about the emissions. They just care about their electricity turning on.
You see opinion change very fast when nuclear power is no longer a substitute. All of a sudden, Germany is like, "Well, we could turn our reactors back on." Japan, same way. They've had these reactors off for years, but now that there's an energy crunch, they're like, "Let's turn them back on."
I think the future for nuclear power, which would be a better future, is you create products that impact fewer people's lives or have the potential to impact fewer people's lives, and also are not substitutable. I think that means small reactors. If you have a battery that can power your phone, or a little battery out in your garage that can power your house... these are hard to make. There are a lot of problems, especially on power density. Nuclear is very energy dense but not necessarily power dense, so you have to do a lot of work to get there.
One of the most exciting examples of recent nuclear technology is these people at some national labs and NASA got together and created this KRUSTY reactor. It's only one kilowatt, so it's small. I think it weighs like 400 kilograms. It fits in a room. They got the whole project done in a couple years for less than $20 million, and it worked great. It's very safe, partly because it's so small, but it has almost no moving parts. The whole thing has a Stirling engine on top, and that's the only moving part.
There are several startups now working on improving that technology and commercializing it. That's the kind of nuclear stuff that is really exciting to me. Micronuclear is really exciting because it has so much potential. When you start putting nuclear in that small form factor, there's no other energy source that can compete with it on energy density. You can do things you could never do before. Whereas selling to the grid in a large power plant is like, "Well, I can do that lots of ways."
If you think through this lens, you see the entire nuclear debate is the nuclear proponents trying to claim that nuclear is not substitutable and that we should pay more, accept the risk, or whatever. Maybe we should, but it makes it hard to promote that technology. If you could have a phone that you never had to charge, people would love that. They'd be like, "I don't care if it's nuclear. I just have a phone that never goes dead."
Dwarkesh Patel
I guess the question is, to what extent is that lengthy and expensive process necessary for the probabilistic risk assessment?
If there's a way you could just have the process be more streamlined and have the same effectiveness in evaluating the harm? And then another thing is, if we haven't seen it—like zero people or very few people have directly died from nuclear—is it just that we've gotten lucky? Or are you saying that could have been way more and we're just in a lucky timeline?
Austin Vernon
I'll go backwards a little bit here on answering those questions. More so than what people are responding to is just that Fukushima didn't have airborne radiation that was very dangerous, but people still got removed from their homes. There's a lot of costs associated with that. It's hard for me to believe that if we had a similar thing in the US, there wouldn't be some type of mandatory evacuations that were really unpleasant. If you could get your power from coal or natural gas without that risk, a lot of people would make that trade-off.
The other thing with Fukushima is, as I understand it, they were able to use a lot of seawater to keep the reactor from getting too out of control because it was on the ocean with fast currents. They were just dumping radioactive stuff into the ocean, but it was dispersing quickly and wasn't a big deal. If you're on a freshwater reservoir, like most US power plants are, your risk equation might have been different there. I don't know enough about it to know if that really matters, but I think the main thing is because of the precautionary principle, people are still going to get removed from their homes, and people don't like that.
You can always streamline processes, but the thing is, people are submitting designs that are extremely complex. Whether your design is ultra safe or not safe at all, to do all the engineering to prove it costs about the same either way. That's part of why these new plants are so much safer than the NRC standards. It's just not that hard to make them that much safer. A lot of your licensing is going to spend the same engineering resources no matter what, based on your plant complexity.
That's the difference why KRUSTY was able to go through so fast. Their thing is very simple. They don't have very many moving parts. There's only so many things that can go wrong with it. That's what's exciting to me about these other startups. They have the potential to get through faster with less money, and then there are real markets in remote power and space and military where people are willing to pay the premium for these initial models.
Dwarkesh Patel
Okay, so you're not bearish on nuclear in the future given the new designs with passive cooling and stuff like that? It was more like the old designs that you're pessimistic about. Is that correct?
Austin Vernon
Yeah. If you look at the cost of electricity from the reactors that just got approved, it's quite expensive. I think it's usually around $40 to $50 a megawatt hour best case, but more likely it could be up to $80 a megawatt hour. They're not building it in deregulated power markets because you lose money. But there are places where it could make sense. Some places in Europe have very expensive electricity, and Japan and Singapore. There could be some markets in there, but that technology still has to compete with those places building solar panels or all these other technologies.
Then there's the whole argument that nuclear can do this and that, but the people building the reactors clearly don't want to build them in deregulated power markets because it's not economic. That's why I'm excited about the small ones, because there are alternative markets other than selling this substitutable commodity that's very cheap.
Dwarkesh Patel
Have you talked to Eli about this? What is his opinion?
Austin Vernon
Yeah, Eli finds out about these new startups that fit this bill and sends me the information because he knows I'm excited about it. His specialty is governmental affairs, so there's still lots of opportunity to improve the process at the NRC. Recently, NEI, which is the industry group that's very powerful, their goal with the NRC was to reduce the nuclear rules by like one third. You also have the NRC writing new standards for Gen 4 reactors, which is supposed to be done in a couple of years because Congress instructed them to do it. There's lots of opportunity to try to improve the process, but it's very complex.
I'll give one example: the Browns Ferry accident. The main thing that came out of that was you can't have control cables for safety systems on redundant safety systems on the same cable tray, because that cable tray catches on fire and you lose both systems. It's very expensive to run extra cable trays and do all this cable separation. That's actually one of the problems delaying Vogtle in Georgia right now. They had like 500 issues of sharing the same cable tray for safety systems, so they have to build all new cable trays and clean out the mess of the stuff they already built and redo it. Super expensive.
The NRC tried a pilot program where they did performance-based safety instead of the strict cable separation rule. I think Oconee was one of the power plants that tried it, and it ended up being more expensive than just the simple rule. The reality is often very complex. When you have these complex plants, it's just hard to do. It can always be improved, but I think the small ones could end up greatly outcompeting the large ones because they have less complexity.
Dwarkesh Patel
You had a small section in that piece about fusion where you were especially pessimistic about fusion. What is your take on fusion?
Austin Vernon
I'm not pessimistic about fusion. I'm pessimistic about fusion technologies that heat up water to make steam and run it through a steam turbine, because they're not efficient. It's just so expensive to do the steam turbine and the condenser and all that kind of stuff. That basically makes you uncompetitive in most deregulated power markets.
There are startups who have plans to do direct energy conversion. I don't know how feasible those plans are, but presumably you think in those cases fusion could have a big future.
Dwarkesh Patel
Yeah, if you're pursuing a direct conversion technology, you actually have a chance of success. A lot of people I've talked to in the fusion space say, "I can make electricity for $50 a megawatt hour, and because I'm fusion, people should pay me $50." It's like, well, not everyone may want to pay you $50.
Austin Vernon
It might involve an initial period of large subsidy, like we had to give to electric vehicles and even solar. We had to give huge subsidies for solar in the beginning when we were at the beginning of the learning curve. That might be necessary.
Dwarkesh Patel
I really disagree with the subsidy solar has had. If you actually look at the numbers, it proves the point. People say, "Because Germany did the feed-in tariffs, that made solar cheap." If you have a country that's one percent of the population and they spent a tiny portion of their GDP, and that was enough to scale the technology, you should just let some other fool do that and reap the benefits. I would be supportive of taking away most of the subsidies for energy in general.
Austin Vernon
You're saying it's unlikely that the small subsidies that Germany gave were enough to actually make the difference?
Dwarkesh Patel
I'm just saying if it took such a small amount of subsidy to do it, someone will be foolish enough to do that. In this case it was Germany. They spent a lot of money doing that, and they're not reaping the benefit because it's not compatible with their environment or climate. We benefited from them doing that. We didn't have to. We still do spend some subsidies on solar, and I think they're very poorly designed, so it would be better just to get rid of them.
But the thing with fusion, if you're just heating up water to make steam, there's no learning curve anymore for steam engines. That technology is so mature. That's why some people are looking at supercritical CO2 cycles, because maybe this could be a little cheaper than doing steam turbines. There's some possibility there. There are other technologies like thermoelectric generators. But the direct conversion technologies have a massive advantage, not only in initial costs but in ongoing operating costs.
There's one more topic I really want to talk about. You have an interesting post on where you can actually expect to find alpha, given that at least public markets are efficient. Do you want to explain the basic thesis of that post before I ask you specific questions about it?
Austin Vernon
If I was going to sum up that post, I love Fama's original paper where he lays out the efficient market hypothesis thesis. He says there are multiple types of information. The first is if you just have pricing data for stocks or whatever securities, you can be the smartest person in the world and you're not going to make any money doing that because it's just random. But if you start incorporating more information, like what's in 10-Ks and all that, if you're super smart you might be able to make a little bit of money there. We see that with people like Renaissance Technologies, and you can debate about Warren Buffett and all that.
Then there's the third category, which is the strong type information. Basically, you have legally acquired private information, and you can make money that way and be significantly less smart. If you want to just take Fama's paper and figure out how to make money, it's like, okay, I should find legal ways to acquire this information, and then I don't have to be a super genius to make money on it.
Dwarkesh Patel
What I thought was really interesting in your post was this point about how one of the ways you can actually earn excess returns is through labor. Buffett in the early years would go into these factories or companies and interrogate every single piece of operations. I thought that was an interesting twist on Piketty's thesis. He has this claim that not only does capital earn more than labor, but the more capital you have, the higher returns you can earn. Harvard has access to hedge funds that may be able to earn excess returns.
I thought yours was basically the inversion of Piketty. Over time, as Buffett has gotten wealthier, his returns have gone down because it's harder to invest the marginal dollar more effectively. As you said with the Medallion Fund, they no longer accept outside money. The interesting thing about labor is that the reason Buffett was able to earn those excess returns in the beginning was because of the labor he put in. Capital is just fungible with other capital, so capital doesn't enjoy as high returns as really good labor, really smart labor. That's the opposite of the Piketty thesis.
Austin Vernon
I think there was actually a paper on Marginal Revolution a couple years back. I'm pulling from my memory here, so I could be missing a little bit, but basically it studied all these businesses and what happened to the business after a founder unexpectedly died. The earnings just dropped like a rock because they lost some irreplaceable human capital there. They didn't spend any time training them because they died unexpectedly.
That also has an interesting implication for CEO pay. In the Marxist sense, your pay is what it costs to replace you. If Steve Jobs is so irreplaceable that if he goes away, earnings are going to drop like a rock and stock prices are going to drop like a rock, that means he should get paid a lot. That's how expensive it is to replace him. He may be irreplaceable. It's actually worth whatever dozens of millions of dollars you're paying him. I'm generally a proponent for letting the market decide that.
Dwarkesh Patel
Another way you suggested that firms could earn excess returns is by developing a unique brand. Y Combinator is probably able to earn excess returns from a venture capitalist because of their unique brand. Do you want to talk more about that?
Austin Vernon
A lot of this intangible capital and labor are complements for regular capital. If you build a brand around being a good investor, you can raise money from other people and charge more on it than if you're just a no-name. Building a brand or building relationships is extremely valuable and can juice your returns, just like specific knowledge.
Dwarkesh Patel
What do you mean it's a type of specific knowledge?
Austin Vernon
To build a brand like Y Combinator, you have to understand what tech founders want. They use that knowledge to create a place that's great to go do your startup.
Dwarkesh Patel
Is the market for blogging efficient? There are now financial rewards to blogging. The Effective Ideas blog prize, other kinds of grants, the Open Philanthropy contest. If you're really good at blogging, you could earn six figures given the regularity and size of these prizes. Is this a market that we should expect to be efficient?
Austin Vernon
It would be hard to measure. Given my own experience, I'm blogging for free, but the benefits I've gotten from learning about what I'm blogging about and then a few connections I've made that helped me with my projects I'm working on, there are huge returns. If my project is successful, it could be almost immeasurable. I would guess it's very hard to measure and probably inefficient. More people could blog because it's hard to predict the returns to what your blogging might have.
Dwarkesh Patel
Let's take that part of it out. Let's just talk about the factor you mentioned, which is that the gains people get are huge. That's also the case in my case. It's kind of interesting. Efficient markets doesn't imply that everybody will put their money into the stock market. That's not the implication. It's very possible that you have irrational actors who are not invested or writing. But the question is, given that you write something high quality, will it get noticed by the market and get the attention and broadcasting it deserves?
In my experience, some of your first posts ended up on Hacker News. In that sense, that market was efficient. It seems to me that when somebody finds a good blogger, it's not hard for their initial post, or at least their subsequent posts as they get better, to gain an audience.
Austin Vernon
I do think that. I don't know what the counterfactual is. We don't know about the people that didn't have posts go to Hacker News. The alternative for me is I just would have blogged way less if one of those early posts hadn't gotten more attention. It's hard to know what the counterfactual is. How many people have abandoned blogs that did three posts and would have written one more? Maybe it would have been better.
Dwarkesh Patel
Awesome. This has been a lot of fun. Thank you so much. I think we're two hours over at this point. Thank you so much for your time.
Austin Vernon
Thank you.
Dwarkesh Patel
Do you have any other final thoughts or any other subjects that we should hit on?
Austin Vernon
No, I think we covered everything.
Dwarkesh Patel
Cool. People can find you at austinvernon.site. Your Twitter is @Vernon3Austin. It'll also be in the description. Thanks so much for coming on, man. This was a lot of fun.
Austin Vernon
All right, thank you.