Loading summary
A
AI is going to sort of make dramatically new types of automation possible. It's tasks that used to be having to had to do manually, you can now do automatically by some machine, sort of reduce the labor burden. People are trying to use this basically AI technology to sort of drive robots around. If you could, you know, automate a big fraction of like, you know, construction work, you would be able to sort of drive down the cost of building a building and find these sort of, you know, address construction productivity problem that we've had for decades.
B
My guest today is Brian Potter, a structural engineer by training and a longtime practitioner of the same. He writes on the hidden logic of how things actually get made and the author of the Origins of Efficiency. Please enjoy my conversation with Brian. So Brian, the book the Origins of Efficiency, did you write that before or after your experience at Katera?
A
That was after. That was several years after.
B
I know that. But to me this was all new to me. I haven't really thought in terms of the practicalities of trying to make actual construction more efficient. The, the experience at Katera, I know I followed it somewhat when things were happening, but talk to me about like what happened. You were there and I know you, you wrote a long piece on it, which I read. But I'd rather have you explain the whole concept to our listeners and viewers. Yeah.
A
So Katera, for those who don't know, is this really well funded construction startup that formed in, in the 2000 and tens. I joined there 2018 and their goal, or what their goal had evolved into by the time I was there was just, you know, revolutionize essentially the construction industry by way of, you know, prefabricated factory built methods, which is an idea that is perennially popular. You, every, you know, 10 to 20 years someone comes along and says, oh, I have a brilliant idea. I will revolutionize the industry with this, with this, you know, factory built methods that nobody has ever, ever thought of before. And then they inevitably just not work out the way that they imagine that it will. And so I joined them in 2018 after I had spent, you know, 10 or 2018 after I had spent, you Know, 10 years of my career in the construction industry. I had thought the industry was like, you know, incredibly backward, incredibly inefficient. And I had thought that these Katera guys were basically on the right track to try and fix it. And so I joined them sort of building and managing a team of structural engineers while I was there. And you know, for a while it was like very exciting. We were like growing super fast and Hiring all these people and the company was like getting bigger and bigger and, you know, signing all these deals. And then it just sort of started to kind of go sideways. You know, the cost kept ending up coming just back super high. We were trying to sort of dial in our manufacturing processes and there was all these difficulties, you know, in the factories and stuff. Just kept being too expensive. You had to, you know, design and redesign and re redesign these. These products that we're offering over again. And they just kind of weren't coming together the way that we hoped. And, you know, Covid hit and sort of threw a wrench into everything and started being, you know, series of layoffs. They started before COVID but Covid certainly didn't help. You know, many rounds of layoffs, sort of winnowed, winnowed the group down. My entire team, team was laid off. And they were just, you know, trying to sort of pivot into some sort of model that would work. And they sort of struggled and staggered along for a few years. I eventually left after about two and a half years there and went to another engineering job. And sometime after, they eventually sort of gave up the ghost and declared bankruptcy.
B
It seems like there were multiple causes, right. For the failure because when I was reviewing it and reading your Another day in Cadera dice, I instantly thought of Sears Roebuck and the success that they had. Kind of mid century 1908 through beginning of the century 20th century through 1940, they sold, I think, close to 100,000 mail order houses that became quite popular. And, you know, it was kind of the birth of that. Some of them are actually considered like really beautiful houses today. So I thought maybe that that's where they got the idea. But no, I'm wrong about that.
A
No. So that's like sort of. There's like a whole history of like prefabricated construction. Construction that's been tried, you know, and sometimes done successfully, but attempted multiple times and in multiple ways, you know, in many different periods and places. So the Sears mail order home was one sort of subset of that. There was actually a whole constellation of these mail order home builders that they were. It was actually based on this idea of people would sell like what were called knockdown boats, which is basically they would assemble a boat out of wood and then just disassemble it into the pieces, like mail you the pieces, and you could like assemble it and stitch it together yourself. And basically somebody said, hey, I should do the same idea, but with houses. And they would send you essentially all the parts to a house and you would have to sort of stitch it together yourself. And I. And you know, that, yeah, they were in that business for. Yeah, from the early 20th century up through the 1930s. And then I think they ended up going bad, going out of business, essentially, because they ended up losing huge amounts of money on the mortgages. And I think they ended up losing so much that, like, erased any. The entire history of profits from the. From the. From the business. But there's, like, lots of different. You know, people have tried, like, the prefabrication thing over and over again. It really became popular in the US After Ford had such major success with, like, mass production and, you know, dropping the cost of the Model T by such huge amounts and some. And, you know, they had had, like, large continuous process factories before that. But, like, with Ford and mass production, this was like the first time that you had, like, this really large. This really complex good being produced in large quantities and like, sort of these sort of continuous methods. And that had really, really dropped the price of it. And so people said, hey, this, you know, method. These methods have made, like, this complex big thing, a car, way cheaper. We should be able to say, apply the same ideas to, you know, complex big houses, which are essentially just, you know, stitching parts together in the same way that a car gets stitched together. The same idea should work. So it's in the 1930s where you really see this idea of, like, prefabrication as a way to, like, reduce cost, start to take off. And so it's really. It starts to get more popular in the 30s. It gets, you know, that people try it again in, like, the 50s and the 60s and the 70s, and it just keep reiterating, like, the same basic thesis where it's like, if I move my process into the factory, it will become much more efficient and I'll be able to, like, lower my costs and prices dramatically. And Katera was just like, the latest of, like, a long history of the same basic thesis. Yeah.
B
And as I was reflecting on that, I thought that there was kind of a parallel. The external event that wiped out Sears was, as you said, the mortgages. And the mortgages didn't get paid because, oh, Great Depression. And with Katera, I know that we're going to talk about other factors for why they went down, but, you know, the pandemic hit, and that couldn't have been friendly to the project at hand. Right?
A
Yeah. That certainly hurt things and disrupted things. And I've heard folks there who claim that, like, if the pandemic didn't hit, they would have been able to muddle through and maybe not have, you know, had the transformation effect on the industry that they had hoped, but would have been able to emerge as a successful business. It may even be true, but I don't know for sure.
B
What are the lessons that are that, you know, are applicable beyond engineering and beyond construction for, you know, is it just, you know, when you get 2 billion in funding and you spend like a drunken sailor, that leads to really bad things? That's a pretty obvious one that I've seen in a lot of startups. You know, Weworks being a classic example or are there in reading your stuff, I think you have very good reasons from the engineering point of view why this didn't work. But you're the expert.
A
Yeah, I mean a big one is just that, you know, in startups you're always trying to hit what's called like product market fit. Right? Where you have like a. Basically you have created some sort of product, some sort of good or service that like people are like clamoring to buy. And then once you know, you don't want to, it's very risky to like scale up your operations before you've hit that. Because if you, you know, you don't know before that, you don't really know what it is you're supposed to be building. And if you build this big organization that's devoted to producing X and it turns out we need to produce Y, it's very expensive to sort of make that change. It's expensive. It's especially expensive if you're like building physical things in the physical world, you know, building $100 million factories that you then decide, oh, we actually don't need this. Oh well, you know, shoot, I spent all this money on this factory that happened at Katera at the very early on. They were very big into this material called cross laminated timber clt, which is like these big heavy timber panels, sort of like a super plywood basically. So instead of, you know, plywood is 3/4 of an inch thick, these would be 9 inches thick or something like that used for like the structural floors and walls of a building. They bet, you know, very heavily on this. They had built what I think was like the largest CLT plant in the world. But by the time the CLT plant came up, came online, they realized, hey, this product is very expensive. It's very hard to sort of achieve the goals that we want using this thing. And almost as soon as they got it online, they were sort of trying to get rid of it as I understand it, because they, it no longer fit with what they were trying to do. And there was just, you know, tons of examples of this. They had brought all this stuff, you know, all these trades in house and then they sort of retreat sort of later thought, well, maybe we should be outsourcing more of this work and you know, blah, blah, blah, blah, blah, you know, trying all these different products. But they were trying to sort of, you know, constantly trying to find traction and find their foot in, which is very expensive to do when you have like you know, a 10,000 person company and you have you know, hundreds of millions of dollars in capital equipment that maybe now you're not sure you actually need to use. So you know, many of that they never really truly found product market fit in the sense of here's the thing that we're selling and this is what people want to buy and we're going to scale up our operations to produce this or deliver this service. They were constantly trying to sort of figure out what that needed to be.
B
And another thing that I thought about was maybe there was also a failure in investor company fit. And by that I mean did, were the people running the company getting a ton of pressure from Silicon Valley, you know, the hyper growth strategy that might work with digits but doesn't work so well with Adams?
A
Possibly. They got a lot of their money from SoftBank, which is famously, you know, for like, I just want you to grow as big and crazy as possible and writing just like truly outrageously enormous checks. I certainly wasn't privy to those conversations, so I don't really know, but it certainly fits the pattern.
B
So did, did you as an employee feel from the management and from, or from your boss or whatever? Was, was that a constant pressure, we got to go faster, we got to build bigger? Or, or were you relatively immune from that kind of pressure?
A
The pressure at the early on was just for like growth at all costs basically, or like scaling up our operations in anticipation of growth. So I was, you know, trying to hire as many, find as many good engineers and you know, good CAD operators and stuff as we could and like setting up the department in preparation for, you know, handling like the huge influx of work, you know, setting up standards and getting our software in place and, and all sorts of stuff like that. And you know, it never really quite all materialized. You know, we were doing anticipation of like this huge work and like the volumes of work were just like never really not that high. And people were like, are we ever gonna like really actually build any buildings and we did. You know, it wasn't like we were doing nothing. We were certainly putting lots of buildings up, but it was never like the huge amount that sort of. We were. Had been prepping for. And then. So then there was then like, sort of a. Like, you know, figure out exactly as the company grew, figuring out exactly what the role of our department was in it, which kind of like, changed over time as they started, like, tweaking, you know, how they foresaw the business model evolving.
B
Well, the. The other thing that I'm taking away by going through all your stuff is like, obviously we. We have some obstacles to the idea of American abundance, for example, particularly in this industry. And is it, you know, bad policy, fragmented institutions? Can we lay it all at the feet of Robert Moses, who famously profiled in the. As the villain and the power broker? You know, I. I actually read a lot about that. I've read the book. I read it a long time ago, but I actually read a lot about it in anticipation of talking to you. And. And I'd kind of forgotten that he. And. And the reaction to him. More specifically, the reaction to the book about him, the power broker, you know, how he ruined New York, caused a lot of the NIMBY attitudes and the procedures and everything else. But, like, was that there before? Has this always been a problem in terms of building and. In America?
A
Yeah, it's interesting. So there's two things, I guess, that have shaped my view on this. One is this very good book by a guy, Mark Dunkelman, called why Nothing Works. I'm not sure if you read it, but it's a lot about, like, you know, Robert Moses and, like, the aftermath of him and sort of the reaction to sort of the book and like, how he suddenly started to be perceived so negatively. And he sort of. Dunkelmann kind of sees this, like, huge transition in, like, US Politics overall where there's these sort of two competing tendencies. One is this sort of what he calls the Hamiltonian tennis tendency, after Alexander Hamilton, to sort of have, like, a robust, muscular government that is capable of, like, doing a lot of things successfully. And then there's also this sort of Jeffersonian impulse that's, like, fundamentally suspicious of government power and wants to, like, check it and restrict it and constrain it and prevent it from, like, inflicting harm intentionally or accidentally on, like, the U.S. citizens. It, like, views, like, the government as, like, a big. Basically a big danger to its citizenry. And over, like, the course of history, like these. The relative, like, strengths of these tendencies have Sort of waxed and waned and like, you know, from maybe the 50s through the 7 through up the early 70s, it's like really a sort of Hamiltonian was sort of in ascendancy and like, you wanted like have these like government agencies that could handle. They could, you know, successfully deliver sort of a lot of things us suggest one World War II off the back of like government intervention. And people saw like, you know, oh, the government can come along and like, successfully do all these big major things. But then starting in like the, you know, around like the 1960s, late 1960s, early 1970s, that sort of gave way and people started becoming much more suspicious of and much more worried about like the government's, you know, power and authority and started to put in all these restrictions and laws and rules in place that basically made it like, much harder for the government to sort of, you know, do anything at all. And this comes from like a huge number of, you know, ways and, you know, forms. So a lot of like the environmental rules that sort of started popping up in like the late 60s and early 70s, national environmental policy act and you know, Clean Clean Air act and all these things which are good rules in many, you know, at least in the way that they were originally envisioned, but they also serve to like, you know, really restrict especially things like the National Environmental Policy act, you know, what the government is able to do. They give citizens a lot of ability to like, intrude. Not intrude, but like, you know, halt government efforts by. Through things like litigation and stuff like that. And so that he sort of sees this as like a, you know, we've. We're sort of on the end of several decades of like this Jeffersonian impulse reigning supreme. And so it's left us in this world of like, where it's very, very hard for like government agencies to actually accomplish anything because we've bound their hands in so many different ways based on just sort of these suspicions and reluctance to let the government have the authority.
B
Yeah, I had a guest on who wrote a book contrasting and comparing America and China and his thesis. America is now a lawyerly society, whereas China is an engineering society. And then obviously the conversation took the normal course that you would expect. Very similar to the Hamiltonian, Jeffersonian divide. If America was suddenly faced with a World War II type mobilization for housing or for energy, for whatever reason, reason, could we do it? Do we have the capabilities? What capabilities would we discover, in your opinion, that we still have? And what capabilities do we think we have that are mostly a myth from your Point of view.
A
Yeah, it's a good question. I have, I guess, a sort of constellation of thoughts on this one is that, you know, in the very. To go back to Covid, in the very early stages of COVID people thought that we would see that sort of like, emergency mobilization type of like marshaling our resources to respond to Covid or something like that. People thought, oh, we're, you know, suddenly going to sort of get our manufacturing gear to manufacture like, you know, PPE and stuff like that or like, you know, really sort of rise to sort of meet this challenge of this, you know, virus. And in some ways we did, you know, operation Warp speed, was able to produce vaccines extreme extremely quickly and, and start manufacturing them quite quickly, far, far faster than anybody thought was possible. But in many ways we didn't like, right. We never really like, oh, we're going to bring, you know, we're going to scale up our manufacturing operations and start producing all these like, PPE stuff that isn't like really short supply. We never, we never, you know, there was never like a huge upswell of American mask manufacturing or anything like that. And so for a lot of like, Covid stuff, even though there's this big emergency, it didn't really inspire like this sort of transformation in capabilities. We really just kind of muddled through in a lot of ways. We didn't end up with like a CDC that was like, massively competent at dealing with this, dealing with this pandemic. If anything, the, the pandemic showed how like, rotted our institution was and how incapable we were of, of fixing it. So, you know, that's one perspective is that we can't, you know, we've really atrophied our ability to sort of respond to these major threats or like, you know, changes in the world. But then I also, I look at the sort of AI build out where we're in the middle of like one of the great infrastructure construction efforts in history. Basically just the amount of like, capital that's being deployed to just build all these data centers and, you know, install all these chips and get all these capabilities online is like really, truly astounding. You know, there's, you can see these like, graphs going around. It's like comparable to like any major like US project basically, you know, Manhattan Project or Apollo or something like that in terms of like, the money and the resources that are being deployed, deployed and like the physical infrastructure that is being built. Only something like, you know, only a few things. Stuff like the railroads are like, really exceeding it. So it you know, it also shows that, like, in the right circumstances, we can like, still deploy, you know, infrastructure and resources to like, build capabilities, like really, really astoundingly quickly if there's motivation to and if stuff is not blocking the way. You know, one big part of it is just, I think for that build out is that for a long time, a very, very long time, local jurisdictions really, really liked having data centers around because they paid a lot in property taxes, but they didn't demand all that much in the way of like, new government services because, you know, it's basically just a big building with computers in it. It wasn't like dramatically raising the population. And so it didn't like, change like, you know, how many schools we need to provide, how much traffic is on the roads and, you know, fire services, police services and stuff like that. It didn't really, like stress all that stuff very much. So you just had this like, big business coming in and paying you a big check for property taxes and not really demanding very much of you. That's really starting to change partly because the new data centers are so big that they are placing stress on the infrastructure. Right. Power and water and all this stuff that has become popular to talk about, but also just because the construction is so vast. And also, I think, you know, the Cape AI capabilities are so like, they're making so many people nervous that people are really starting to oppose data centers in a way that they haven't before. And so, you know, the sort of local restrictions on like, sort of the building these things, even in places that they used to be like, really, really popular, like Virginia, it's just becoming stronger and stronger and stronger. And so, yeah, it kind of the AI sort of shows like, oh, we still have these capabilities when they are not restricted and when there's the incentive is. Is. Is correct. But also shows like, how quickly the forces that can like shut these things sort of down can move.
B
Yeah, and I've done kind of a deep dive on, on the history of innovation. And what I found was really interesting. It happens all the time. Like literally the industrial, the most obvious is the Luddites during the. The weaving explosion of innovation. But it goes on and on and on. And I think a lot of it has to do with just kind of a basic fear of the new and good storytelling. I mean, back to the power broker. He wrote a good book and he crafted a villain and people responded to that particular villain. And there's a villain being crafted right now about AI. And you know, you get people emotionally reacting to it. But on the subject of AI, do you think that with the AI we have right now available to us, what benefits do you think it can bring to your field?
A
But what do you mean by my field? My field?
B
Well, engineering, building, the whole real world, construction, etc. Etc.
A
Yeah, so I, you know, I guess there's a, a few things. I mean, one is this, this, you know, at a high level you can, you know, AI is just some, is a new tool for automation that's, you know, we've been automating work for, for, for, for centuries. AI is going to sort of make dramatically new types of automation possible. But it's, you know, gonna still be the sort of automation that we've seen before. It's like, you know, tasks that used to be having to, had to do manually you can now do automatically by some machine, sort of reduce the labor burden. You know, it used to take 60, 70, 80% of people working to sort of work in agriculture to produce enough, you know, food just to feed society. We've sort of, you know, thanks to automation and other technological improvements that's dropped to like, you know, 1 to 2% or something like that. So, you know, frees people up to do sort of new other things or maybe have to do less, you know, lead more fulfilling lives or having to do just less, you know, back breaking unpleasant labor or whatever. And yeah, you know, so a big part of it is just going to be like tasks that used to be having to do manually are going to be sort of automated. So you know, a lot of design and engineering tasks, I foresee getting sort of automated with that. You know, automated away. You know, people find new, perhaps find new higher leverage tasks that they can do now that these sort of, you know, other specific things are, are, can now be done by machine in terms of like the physical side. You know, people are trying to use this basically AI technology to sort of drive robots around. You know, there's all this investment going into like humanoid robots which are a lot of these things or like, you know, not necessarily humanoids other companies are working on, but the same basic idea of like using these sort of big, huge neural networks essentially to, to drive these sort of like robot arms or humanoid robots or whatever. That's a lot. You know, the progress is not, we're not at the same level of just like, you know, the chatbot AI models or whatever where it's like increasingly capable of like doing any information processing task. These things are not, these things are not, not as good as driving a robot around yet as they are, you know, answering a question about, you know, the history of nuclear power or you know, how government agency works or, you know, where you should shop to find a good parachute, you know, any sort of information processing task where these things are already like extremely, extremely capable. Those capabilities aren't nearly the same place as they are in robots, but they're getting better and there's lots of investment and lots of enthusiasm and people, many people think that like, you know, it's going to be on like a similar sort of evolution where eventually these things are going to be able to move around quite capably in the real world and be capable of like automating a lot of physical, physical, you know, actions and operations the same way that we're, you know, with current AI models we can now, you know, see the possibility of operating, of automatic automating, you know, a huge swath of intellectual work, if not, if not all intellectual work. So.
B
And do you think that AI will have the ability to assist the human designers to, for example, could you think, working with an AI, do you think that they could solve that premanufactured problem that, you know, Sears ran into, you guys ran into like, or, or no, maybe.
A
I, you know, I don't think the problem with like, you know, these, these prefabrication was around like insufficiently clever design of the building basically. I think it has to do more about, you know, I've written quite a bit about this, but has to do with more with like the fundamental constraints of how you put a building up, like difficulties achieving economies of scale and kind of, you know, huge, you know, different, different like jurisdictions with sort of different requirements. The sort of fundamental nature of like building a, you know, putting a building on site which has to meet like your site requirements, all these sorts of things like that. It has very little to do with just like not being smart enough to design a, you know, a pre fabricated building properly or something like that. So I don't think it would, it'll have effect like on that side of it. I do think that if you could, you know, automate a big frac fraction of like, you know, construction work, you would be able to sort of drive down the cost of building a building just by using you know, these automated labor, you know, robots or whatever instead of sort of manual workers and finally sort of, you know, address this construction productivity problem that we've had for decades, which is just construction productivity. Labor productivity never really seems to improve. If you have, you know, sufficiently good AI, sufficiently good robot control, you can find a kind of finally address that problem.
B
And another question I had for you when I was reading your stuff was, is it simply a matter of, like, the new rules and regulations, et cetera? I mean, I think of how fast we built the Empire State Building. I think about how fast we built, like, monumental projects in the past. Is it just regulatory environment, social environment, that we can't build them like that anymore?
A
I mean, we can still build stuff fast if we choose to. A lot of these data centers are going up quite fast. You can build certain things, power plants, certain industrial facilities. They can go up quite quickly. So we can build quickly if we need to. Or again, in certain cases where we haven't made it outrageously difficult call. Regulation is certainly a big part of it. I kind of view of it as, like, you know, if you had these, like, steadily encroaching regulations, steadily, like more difficult bureaucracy that needs to be navigated. You know, the rise of this Jeffersonian impulse that, like, makes it just harder to do anything that has to sort of go through a government process. But it's, you know, a big part of it is just on sort of like on the technical side, just like the fundamental nature of building buildings, and it's just difficult to improve for kind of various reasons or has historically been anyway.
B
And you know, what, what efficiency that, that, you know, of, which I might call a forbidden efficiency in the field, one that would actually work, but that would instantly trigger cultural or political rejection.
A
Oh, I don't know. You know, I don't know of any sort of instant things like that. I would say the big theme of my work generally, especially around, like, construction stuff, is that there are not like, super easy solutions and that any sort of thing you like, oh, if we only we could just do this. You can find examples of someone trying that and it not working, or like, oh, here, you know, you can try this. You know, it's. It's not going to do what you intended to do because of these various complicating factors. Or you may think the binding constraint is this. But if, you know, cases where this constraint has been relaxed, oh, it turns out, you know, you still don't get the improvements that you hope to see. So I, you know, fundamentally, a lot of my work is, I would say, pushing back on that idea. And just, you know, a lot of these difficulties are due to sort of the inherent nature of the process and not due to sort of like one difficult roadblock that we've erected that if we just change this one thing all of a Sudden our problems would be fixed. Most our problems are not like that, I don't think.
B
So. What innovations and or procedures have you seen over the last, let's call it 10 or 15 years that either surprised you because they worked or surprised you because they didn't work in sort of
A
construction or more generally.
B
Let's go generally. But then also touch on construction.
A
Sure. I am quite surprised at how quickly you can, you know, to get, take it back to AI. I'm quite surprised at how quickly we've managed to sort of scale up this, you know, the build out of these, you know, these computing power and data centers generally. I thought, you know, basically our sort of, you know, NIMBY sensibilities were so, so strong that like any major infrastructure project was going to be just inevitably like strangled by these sort of things that just made it hard to build like anything really large in really large volumes to deploy very quickly. And we didn't build things quicker than I have anticipated, you know, and now we're starting to see sort of backlash to that. So maybe that will like slow down. But up until now I've been impressed about how sort of quickly stuff is, has come online in terms of like stuff that I'm, you know, was surprised that doesn't work. Like my whole history like, you know, researching construction, you know, I've over and over again, again, you know, run into an idea of like something that's like, oh surely if they did this, this would definitely, like I said, this would definitely work. And in learning of some case where it didn't work. One example is that, you know, we've talked a little bit about prefabrication and you know, if only you could like get like prefabrication done at scale or something like that, then you would be able to sort of achieve these like cost savings that people are constantly hoping that prefabrication will do. It's like only if we will like use this really widely. But then I learned about, you know, countries that have basically deployed prefabricated home building really, really widely and they still don't see cost savings the way that you would sort of hope or expect them to. So like Sweden is like the typical example here where they built some huge fraction of like their single family homes and apartment buildings are built like using factory built construction basically, you know, something like 80, 90% of single family homes and like 40% of apartment buildings or something like that, prefabricated construction. But their costs are not like low. It's not like it's, you know, it's not like they're not producing the model T of, of homes over there, like in, in the sense that like, oh, these homes are just so massively cheaper than anything you could build by hand. Their homes are like more expensive than the homes that we build in the US from what I can see. So yeah, and just over and over again running into things like that, it's like, oh, maybe this idea would work. No, it didn't really. Oh, if we relax this regulation, it would have this big transformative effect. Not really. Yeah, it's very, very tough.
B
And if you had to explain America's building problems with a non construction analogy, would it be like health care, medieval guilds, enterprise software, you know, political?
A
Yeah, yeah, I would. You know, again, kind of two things. We talked about this once, once, a little bit before and I went into half of them. But one half is like this Jeffersonian versus Hamiltonian impulse and sort of the rise of sort of Jeffersonianism. And then the other is just. We've talked about this a little bit ago, but once something becomes like, you know, I'm almost not sure kind of how to characterize it, but once people like the reach some like certain level of like affluence or success or like once something becomes, you know, gets reached to some level of like disruption, people are willing to tolerate it up to some certain point. And then once it goes beyond this like point, people start to get really upset about it. So like in the 1960s I've written this essay about like growth in California and how for a long time California was like a very, very like fast growing state. And up through like the 50s, California like took pride in what like a fast growing, fast growing state. They were for a long time they had like a, this big sign that was on one of their bridges, I think maybe the Bay Bridge in the Bay Area, but I don't remember. But anyway, it had like the population and like, of California next to like the population of New York. And it was tracking when California would like surpass New York as being the most populous state. So for a long time, you know, California was like, oh, we're like a big growing state and that's like good. Basically for our society, growth means like, you know, more people, more jobs, you know, greater division of labor. We're getting wealthier, that's what we want. But then at some level, you know, eventually that kind of like tipped and you know, the consequences. There had been enough growth that people were like wealthy and successful enough that they started to be like very Unhappy by like the consequences of all this growth. So, you know, as this growth was taking place, you had all this like environmental ruination and like, you know, just huge swaths of like, you know, forest being cut down and like water being polluted and all this stuff that people like, you know, oh, what, you know, what hath we wrought? All these sort of, you know, we were rushing, grow, go, go, go, go. And we've made ourselves rich and successful, but now we're sort of ruining the environment that drew us here in the first place. And so in the 60s you see like this very wide scale, very like grassroots, like switch to like being opposed to growth. And you know, jurisdictions started electing like, you know, political leaders that would like enact anti growth policies and all these things. And sort of California still grew after that point, but its growth was like much more checked and it become much more fraught after that. And I sort of, you know, I. So yeah, either, you know, people like when, you know, if there's like a rising tide that's lifting all boats, people are very happy to sort of be a part of that. But eventually if they start seeing the consequences of that, they're, or get to a point like, oh, I've, you know, I'm, you know, this is starting to sort of affect me or the place I live in negative ways. They're gonna start eventually turning against that. I'm thinking of like, you know, China, which I'm not an expert on China, but I understand it. Like a lot of what the Communist party derives its legitimacy from is that they've been able to like successfully deliver really robust economic growth. And you know, these people remember like, what it's like to be like wretchedly poor. And they're very happy to like be in a society that's growing like much, much richer and more successful. And we have all these things that we didn't have before and we're a much more powerful country than it used to be. And so they're, you know, they really rely on, be able to like be able to sort of continuously dive these improvements. And they think that if like those sort of growth maybe runs out, maybe things are going to be start to. The people are going to start to sort of reconsider what they think is important.
B
Yeah, of course, the irony about China's rapid growth is it was probably mostly driven by Deng's opening up China to limited capitalism. Right. For those 20 years when he took over. And that was basically the, the engine that drove much of China's growth. If we return to a more Hamiltonian aspect here in the United States. And, and you got named like Head of Housing and, and you could flex your, flex your muscles a little bit, a little bit more than you could under a Jeffersonian administration. Like what, what, what would you do? What would be your, your platform?
A
It's, it's tough because you need a lot of this sort of restrictions on like, you know, building things are at like the state or like regional or like local level. A lot of it is local, like city supervisors or whatever or like design review boards that like are listening to like local residents unhappy about, about like you know, some new apartment project or, or whatever. So a lot of this, you know, has to be done at like, and it's, you know, it's in such a way that's like, it's, it's hard for like, you know, federal tools or federal policies to change that in like huge ways. There's sort of carrots and sticks that they can do, but it's hard for them to sort of like mandate things. So a bit, but a big one is just like encouraging sort of state level, state level, you know, setting of like, or like reducing sort of a lot of these like growth restrictions that maybe like local jurisdictions have and such like that. What federal has a lot more ability to influence is stuff that goes, you know, like large scale, certain large scale infrastructure construction projects, sort of like a lot of like energy, energy building, transmission lines, pipelines, you know, you know, big solar installations. Federal government, you know, has a lot of like leverage there that it can do. And so like big wide scale reform and how, you know, permitting by of you know, stuff that goes through like this federal permitting process, you know, reducing like the burden of environmental review and things like that could like really have a huge build out or huge impact in terms of like how much energy infrastructure that we can build and things like that which is really going to becoming very, very important.
B
And you know, the federal government owns a lot of disused land that you know, nobody, there's nothing on it. Why haven't they pivoted to some of these, like you brought up solar. Why, why not use some of that federally owned land? Again, you're the czar of the, of this building thing. Would that be something that could be generative?
A
They do use that land for a lot of energy stuff. A lot of oil and gas drilling is done on federal land. Part of it is like, you know, again we have this like, you know, all this like burdensome process for like doing all these things and like the oil and gas industry is like very large and successful and they've had like a long time to kind of like work the process in a way that some of these, like newer, more nascent industries haven't had time to quite work it as well. So there's a lot of like favorable environmental law and rules around permitting for like oil and gas stuff that maybe doesn't exist for like other stuff yet. But some of that, that's changing. I think they've, I think recently changed some of these rules that makes it easier for sort of like geothermal energy drilling which like use it, which there's like these novel geothermal energy technologies which basically use this oil and gas drilling methods to sort of drill into the earth and manually create these fracture networks and then inject hot water, hot liquid underground and basically extract thermal energy from underground and use that to drive a turbine or whatever. And there's some promising companies that are using this technology and I think they've recently changed some of the rules or they're in the process of changing them. I have to look up the specifics, but to sort of give some of the benefit, you know, the, the permitting benefits that maybe oil and gas have been able to take advantage of to, to apply to sort of these, these geothermal methods. So you know, it's, you know, some of it's, it's, it's changing and evolving, but all these, you know, this stuff invariably moves quite slowly, never quite as quick as you'd like it to.
B
And again, listening, it seems to me that the mismatch here is a lot of what is throttling. Our ability to, you know, build more houses, et cetera is all happening at the local level, right where they have endless review, endless ability to challenge, to sue, variety of reasons. So basically what I'm hearing you say is even if you had that Hamiltonian power at the federal level, you would be bound in some way or face a lot of bottlenecks from what's going on locally. I mean, is that just a permanent problem? Is there a fix for that?
A
You know, that's a good question. I wish I knew like of a ye easy answer to that. But you know, hearts, you know, other than like, you know, long term change of like hearts and minds. You know, I work for the Institute for Progress and there's this sort of broader interest, interest in sort of progress studies more generally of which the, the group I work with is a part. And I think you've, you've talked with many people sort of in that, in that Vein. And I think a big part of, like, what those people are interested in is like, cultivating this sense of, like, oh, progress is good, and expanding our capabilities is good, and economic growth is. Is good, is part of, like, you know, changing these, like, attitudes in sort of a broad way. Because a lot of it just stems from, like, this. These, you know, sort of ideas that have been inculcated into people's heads. And if you can sort of change how people, like, think about these things or, like, recognize, like, what actually is responsible for, like, this world of incredible plenty that we've created and how, like, if we sort of, you know, may we could continue to sort of improve that if only we were allowed to. You know, there's like, a big emphasis on sort of, like, that sort of work.
B
And I'm just speculating and would love your opinion. Like, aren't there pockets where you could do a project. You mentioned earlier that people were in California particularly really pro growth because it was seen as the rising tide, lifting all the boats, et cetera. You know, are there pilot projects that could be done somewhere where there was a significantly disadvantaged population, where you could do some form of building, some form of different way of building a village or what? I mean, like, I'm not the expert here. I'm just thinking out loud where you could say, see, look, this was a very disadvantaged area. And now because we've done this, look at it thriving.
A
Yeah, this is like a perennially popular idea. And you see, you know, variations of this concept show up in a bunch of ways. There's a big interest nowadays in, like, finding news, founding new cities. Right? And like, finding these new cities that would, like, designed to be, like, you know, growth or prosperity engines. There's a lot of, like, existing cities and places where in the US where like, maybe they would, like, they would be fine with like, a little more growth that offer, like, have offered, like, bonuses to, like, tech workers, like remote tech workers. Like, hey, if you're a tech worker that can work remotely, we will pay you, you know, 50, $50,000 or something to relocate to Tulsa, Oklahoma or Delaware. I forget exactly where they were, but there's a variety of these. Of these places. They're trying to sort of, you know, encourage people to come and into. Into this city. You know, for a long time, people were, like, so fed up with, like, the difficulties of changing things. In San Francisco that were, like, you know, a lot of, like, the venture capitalists and tech population, there's, like, you know, we're just gonna, like, relocate all of this to Austin or to Miami or something like that. For a long time people were like, you know, really trying to like, like Miami, like the new tech hub. And I think it proved like really difficult. Like Austin has been like a major success story. But I think Miami basically didn't really work out. And basically it proved pretty hard to sort of dislodge, you know, the ecosystem that had grown up around San Francisco and move it anywhere else. All these things, you know, there's like network effects and like built in advantages. Right. It's like once something's like successful and there's all these like people and places that are located in this place that you want to be, there's a very hard marketing problem of like getting everybody to just, you know, no individual person has like an incentive to leave. Right. Because everything here is already like where I want it unless everybody else goes all at once. But if everybody else sees it, everybody else is staying here, then they're gonna stay here. Here too. These, you know, these sort of, you know, things have proved to be quite durable and difficult to change. I'm not, you know, same with like the cities thing. I think it's hard to sort of spin up like a new city from scratch because if nobody else is there, there's no reason to go there. And if there's no reason to go there, there's no reason for anybody to like say, start. You know, it's, it's that it's these like sort of difficult chicken and egg or, you know, problems that are kind of hard to, you know, break out of.
B
Yeah. And of course the example of the famous ghost cities in China would be another example of, yeah, we'll build the city and then no one goes.
A
Yeah. Although I think they actually, I think actually China's somewhat of a counter example. I think they were having like such big urban migration that a lot, they sort of, a lot of these were like cities were like built in anticipation of population growth. And people said, oh, you're building this stupid city in the middle of nowhere. That's wrong. What's wrong with you? But there was such huge migration. A lot of them did end up basically filling up as they anticipated. But that's in like, you know, sort of a different situation because they were having this like, they were like in this big transition from like rural population to urban population. They were having all these people move from the countryside and move into the cities when they're in the middle of a situation like that. You can like build these new cities in a way that you can't, if ever, if you're already in like, a highly urbanized population with all these sort of like, existing industries and sort of networks in place that are like, are going to be resistant to being dislodged.
B
I like you leaning into the network effect because I think you're absolutely right. Do you think that there's any mitigating thing you could do if you were trying? You mentioned Austin succeeded where MIAM failed. I'm curious. I'm certainly not an expert in this at all. But why. Why did us. Was it simply that Austin was so much closer to the place they were drawing population from?
A
I don't have a super deep knowledge of like, the what's on, you know, the situation on the ground in these two cities. And, you know, I don't know how much Austin has been. They've been, you know, it's grown quite a lot. I don't know how much success it's had in like, relocating, you know, the venture capital ecosystem. As far as I know, the San Francisco Bay area is still like, by far the biggest and largest, and nothing else in the US Is really close. But I do think that Austin has been very successful in growing just because they've made it very easy to sort of build housing there. I think Miami, for all the. For all the people that were trying to like, you know, turn it into the next San Francisco or whatever, I think they actually are like somewhat nimby, and it's not actually amazingly easy to sort of build new housing there. So I think, you know, a lot of it can probably, if you look
B
historically, I guess you can also see a waxing and a waning of trends. Like, for example, for a long time immigration was desired in the United States. You'd let everybody in, and then you had this backlash in the beginning of the 20th century with, or even the late 19th. You can also see a waxing and a waning of trends. Like, for example, for a long time immigration was desired in the United States. You'd let everybody in, and then you had this backlash in the beginning of the 20th century or even the late 19th century with the Know Nothing party and all the people who, like, didn't want anyone to come. Is this kind of sinusoidal thing affecting what, you know, the, the building and. And whatnot in the country today?
A
Yeah, that's a good question. I don't. I actually, you know, I should know more about this because IFB has a very robust and strong immigration team that tries to work on, like, encouraging policies that Would you know, make maximum use of like high skilled immigration. Because the US has been like, that's been such a huge story around the US successes, right, is that we've been able to attack the, attract like the best and brightest talent from around the world and bring them into an environment where they can really like make a maximum use of those talents. So it's like, you know, it's great for the U.S. it's great for the people that, that come here too. And I really don't have a great sense of like how those perceptions have evolved over time and, and how it's like changed on like the love, you know, how that's been reflected in like the level of policy and how versus like what the sort of typical citizen thinks about it. I'm just not informed enough to, to, to know about it. But yeah, I, you know, I don't like the late 19th and early 20th century. It wasn't necessarily amazingly popular even as we're like getting like a really large number of immigrants. So I don't, you know, I just don't know enough about it to speak unfortunately.
B
Yeah, I know maybe a little bit more about it because I'm Irish and one of one, one of the central aims of a lot of these movements was to keep the Irish people out. And I just sometimes see a parallel there to like not in my backyard that type of attitude. It being prevalent in, in the discussions today about the fact that we need more houses. The houses that we have are very costly right now. What's the solution? Why can't we come up with one?
A
Yeah, a lot, a lot of this, you know, comes down to sort of ideas of like concentrated harms and like diffuse benefits. So like a city, a city overall or a state overall will like benefit from like population growth or other bigger, stronger economy, more division of labor. It's like nicer to live in like a bigger city. But like if you're building like a big apartment building next to, next to a, you know, a housing development or whatever, the people, you know, even though that built big new apartment building might like have like a small impact on like the overall rents in the city and overall contribute making the, you know, over the city slightly more affordable. The cost of that thing like the disruption and like the increased traffic are all going to be concentrated right next to that like one spot. So the people there are going to like rationally oppose it because they're getting, you know, you're talking like a small diffuse benefit over the entire city versus the concentrated harm that one you know, one group of people really does not, does not like, so they rationally oppose it. The benefits are like diffuse enough that like, it's hard to marshal like a lot of support in favor of it. And so, you know, you have this sort of fundamental asymmetry. I ran into this like in my neighborhood not too long ago where some developer wanted to build like a, you know, retirement community sort of kind of near our neighborhood. And it was going to be like a multi story building or whatever. And you know, that's, you know, people, people need place to live, including older people that if you don't build those things, like the cost and the difficulty of like finding of being living in a retirement community goes up. And so, you know, the more you build, the more like affordable these things become. But that benefit is like, you know, spread very broadly and then the harems are all concentrated, you know, of, you know, right in this like one spot. And so all a bunch of the local neighbors were like, very opposed to this new building that was going to get put up. So again, there's these fundamental comes back to sort of like these incentives or the perception of what the incentives are.
B
I think that framework is an excellent one to look at a lot of problems. The diffuse benefits versus the upfront costs. And what we notice is it's hard, right, to think about the diffuse benefits that a particular project. And it's really easy to look at the harm, right? Like it reminds me of the idea of learning via negativia. And by that I mean it's really easy. Let's take a drug, right? Like if there's a drug available in Europe and not here, you tend to look at it from a. It isn't available here. So you're not thinking about it, you're not worried about it, but you're also not thinking about all the lives that drug that is available in Europe could have saved in the country. Right. So we, it just seems our human os has a really hard time dealing with that kind of construction, right? Like, how am I supposed to have an opinion on something that isn't available here? You want me to think about all the benefits of what it would be, all the positive benefits of if it was available here. Yeah, and we kind of go on fritz. So I definitely like that framework of sort of the obvious and immediate costs that present themselves versus the diffused benefits that might also temporarily happen at a different rate and a different amount of progress. Do you think was, was your time at Katera one that basically like no more moonshots or do you think, do you have an idea for some moonshots that might actually work?
A
Oh, good question. No, I don't think soured me on the idea of moonshots. You know, I wish it had been like, what, what. What the, you know, what Katera inspired me was like, wow, it's really shocking how, like, little people understand about, like, what actually is, like, required to like, make this process work better, you know, that people have been trying essentially the same idea over and over again and it. And, you know, oh, move our process at a factory. And every single time they tried it, it like, didn't work in the way that they thought that it would in the sense of, like, oh, I'm going to beat the Henry Ford of housing. That never happened. But people somehow people. The lesson never, like, stuck and people just kept trying it kind of over and over again. But it actually, you know, at a high level it makes me feel like, you know, this is one of the major strengths of like, us society, right, is that we're in an environment where people are willing to invest, like, enormous amounts of money on this, like, speculative idea because they think it will, you know, be a success and that, you know, it will be sort of transformative in the ways that they hope for. I'm glad that we live in a society that's able to, like, willing to take those, like, really big swings, you know, for any individual one. You can argue that. And of course, Katera, you would argue, you know, correctly, like, you know, this one is not very well thought through if you, you know, you should sort of maybe retarget what you're doing. But, you know, I really, like, at the, at the high level that it's a, you know, a place where those things occur. And I would like sort of, you know, many, many more moonshots and more investments in sort of like these, like, speculative, like, you know, transformative ideas that maybe aren't like, you know, you know, the popular, like the hip thing. You tend to see, like, a lot of clustering around, like, common ideas or whatever. The. The same time right now all the money is going to like, you know, robots and AI. And a few years ago, all the money was like, oh, I need like, cryptocurrency or whatever. I would like to see sort of a robot, a more robust ecosystem of. Of funding that can, like, fund projects which, like, don't, like, aren't necessarily, like, attractive to like, venture capitalists for whatever reason, but still have the potential to sort of be quite transformative. And you're Starting to see like, more of these things spring up. So I would like to see like many more, many more moonshot type projects of all sorts.
B
Give me an example of one that you would like, that you've seen and you're like, wow, that would make a
A
great moonshot one that I've heard about. And it's, you know, it's not even one that I like, I've like, strongly advocate for because I think it will probably happen eventually, but I've read about it from some robotics experts, is that we don't necessarily have yet have like the robot equivalent of like a GPT3 moment where like this new model comes out and all of a sudden it is like massively capable because it's trained on like such a huge, like, corpus of data or whatever. But then there's speculation that like actually doing this maybe wouldn't be that much work. You know, you get some like, you know, you train, you get, you pay enough humans to like, you know, do teleoperation of like, you know, various tasks and you feed it enough video data or whatever. And maybe you could actually get like a robot GPT3 and have this like, general purpose, highly capable robot model. And it would take, you know, many millions of dollars to do, but maybe like less than you might expect. And that seems like a really, you know, valuable thing, at least in terms of like potentially advancing robot capabilities. But there's so much money that's getting put into these like, robot companies that I have to imagine that somebody, if not multiple somebodies are like, working that on that right now. It's not amazingly easy to find. Yeah, you know, these, these moonshot ideas. It's not amazingly ideas I easy to find ones that are like, promising enough that, you know, you think somewhat, you know, someone should fund, but not so obviously promising and that people have like missed it already. You tend to have to, at least in my experience, you have to have like some sort of like discipline specific knowledge or whatever. And I'm a little bit too, I'm a little bit too much of a generalist. And you know, I, you know, just stay a little bit too close to the surface level of things to have like a really deep understanding of like a lot of what these needs are, a lot of these. So a lot of these like, operations that are like finding these like missed sources of, you know, valuable potential things are they're doing so by like finding a lot of like, domains, experts and like asking them, hey, what do you need in like this specific area? What would be like really transformative. But it's not amazingly easy for someone without that like, you know, domain specific knowledge to like articulate what they are
B
and, and what's it, what's a normal day look like in, in your day job as Senior Infrastructure Fellow at the Institute for Progress?
A
Yeah, almost all of what I do is, you know, researching and, and reading and writing things. So yeah, typically on a normal day I will get up and the morning will be spent working on whatever the current writing project is on my docket. So usually that's like whatever newsletter project that I'm working out. And so mornings are almost always dedicated to writing. I find that like I can really only write for about three or four hours effectively in a day. So I try to block out my time in the morning to basically do that and then in the afternoon I'm basically just doing research on whatever upcoming project I'm working on. So, you know, this is like reading a book or like looking up sources or doing some sort of data analysis or increasingly it's, you know, asking AI to like dig up sources for me or do some analysis for me or write some script for me on that. But usually at a high level it's you know, writing in the morning and like reading and research and the afternoon basically pretty much every day.
B
Do you have an example where working with a colleague at the Institute from a completely different section of, you know, what they're working on, where you guys having lunch together, chatting about things where the cognitive diversity between the two of you and you were like, ah, that's a great idea. I never thought about that or no.
A
So I would. I work from home, so I don't actually work in work in their office. I'm working out of my home office basically every day. Institute for Progress is in Washington D.C. and, and I am located outside of Atlanta, so, you know, a little bit of a commute. But often I will like other people like suggest, you know, topics of essays I should write or topics that I should research. And I'm always like open to those. And some of those suggestions have been some of my like best and most popular and most interesting, you know, essays that I've written. So yeah, some of those suggestions for like topics that I should like look into have been very valuable.
B
So you got to give us at least one that we can put in the show notes. Which one along those lines that got very popular.
A
So I think my most popular essay that I've ever written is about the history of airplane manufacturing by the US during World War II. And basically how we built, like, the huge number of airplanes that we needed to produce, which is not like, you know, necessarily my normal beat. I write a lot about, like, manufacturing stuff in general, so it's not like, totally outside of it, but it's not about, like, buildings and infrastructure. But I think somebody. I don't actually remember the specifics of, but I think somebody basically suggested that would be a good topic to write about. I said, oh, that is something good. You know, I have a sense of, like, what a topic would be good to write about. If there's like, this combination of, like, again, it's kind of the same. It's kind of the same thing with, like, you know, investments. You want it to be, like, promising enough in the sense that there's, like, a lot of, like, sources and research that I can dig up about it to read and learn about it, but not, like, so promising that somebody has already written something really, really good about it. So I'm sorry to, you know, I'm trying to sort of hit that, like, relatively narrow target of, like, investable writing and research topics. And so again, I often draw. I can, you know, it's. It's often useful for me to draw on, like, what other people can. Can see because I can't just necessarily see everything myself. And so that was like, I think a particular good one. I'm working on one that'll. Or I'm going to start working on one a couple weeks that was suggested by somebody else that I will think be similarly good for kind of similar reasons. It's like that right combination of, like, you know, neglected, but still enough to like, really do a good and thorough, interesting research about it and it'll be quite interesting. But I can't reveal that one yet.
B
Come, come back for more when we'll have. We'll have the one on the airplanes on the show. Notes Brian. This has been absolutely fascinating for me. We are coming to the end of our conversation and we have a tradition here at Infinite Loops where we make you, just for a day, the emperor of the world. You can't kill anyone. You can't put anyone in a reeducation camp. But what you can do is we're going to hand you a magical microphone and you can say two things into it that is going to incept the entire population of the world. Whenever their next morning is, they're going to wake up and they're going to say, I've just had two of the greatest ideas. And unlike all the other times when I wake up with these great ideas and then ignore them. I'm going to start acting on these two today. What are you going to incept in the world's population?
A
Oh, gosh. I mean, one easy one is just, you know, build housing. That's like a really easy one because that stems from, like, this, like, grassroots local opposition to like, building new things or even, you know, like, you know, build more stuff, build more housing, build more infrastructure. If I could find some way to like, communicate that basic idea that would be like, really powerful. And all of a sudden we can build all the house we need to build all the transmission lines and all the solar panels and everything. That would be, like, truly transformative. And I don't know, maybe something about, like, make AI safe. You know, we're. We're in this world where, like, AI capabilities are, like, advancing, like, monstrously. And I think many of the concerns that people have but like, oh, what happens if you, like, you build this thing that is like, massively smarter than everybody else in the world? It's not necessarily going to have your best interests in mind. Just like when humans became massively smarter than other animals, they did not necessarily have the best interest of other animals in mind and it did not necessarily work out for the other animals all that well. Even the ones that, like, humans weren't necessarily, like, interested in hunting or whatever. And so if you could ensure that everybody took that problem really seriously, I think that would be a major win as well.
B
Two great ones to think about. I love in particular, if you get the. If you incept the build more houses, maybe that all that would have to change would be the attitude, right? So that they physically wouldn't have to go out and build the houses, but they could drop their opposition to many, many projects. Brian, where can people find your work?
A
Yeah, I write a newsletter called Construction Physics that you can just search Brian Potter, Construction Physics and it will come up. I'm the author of a book that's called the Origins of Efficiency. It's found on Amazon. If you search Origins of Efficiency, it will come up there. And those are the two main places where. Yeah, my two. Two main, major outputs.
B
Perfect. Brian, thank you so much for being on Infinite Loops.
A
Thank you, Brad.
Infinite Loops Podcast – "Brian Potter: How to Fix America's Building Problem"
Date: April 23, 2026
Host: Jim O’Shaughnessy
Guest: Brian Potter, structural engineer, author of Origins of Efficiency, and writer of the Construction Physics newsletter
This episode delves into the persistent productivity and innovation challenges in America's construction industry and explores deeper institutional, political, and technological roots of why "building" in America has become so fraught. Brian Potter brings his unique perspective as both a practitioner and a writer/researcher, discussing lessons from construction startup failures like Katerra, the long and often disappointing history of prefabrication, and the prospects and limits of AI and automation in changing how we build. The conversation also unpacks the broader cultural and political shifts—from Hamiltonian ambitions to Jeffersonian suspicions—that have shaped America's regulatory landscape and impacted building at scale.
[01:50 – 04:23]
Quote ([03:37])
“We were trying to dial in our manufacturing processes and there was all these difficulties in the factories and stuff. Just kept being too expensive… Products had to be redesigned over and over again. They just kind of weren’t coming together the way that we hoped.” — Brian Potter
[05:06 – 07:51]
Quote ([06:30])
“People have tried the prefabrication thing over and over again. It really became popular in the U.S. after Ford had such major success with mass production… But Katerra was just the latest of a long history of the same basic thesis.” — Brian Potter
[09:27 – 12:00]
Quote ([09:49])
“If you build this big organization that’s devoted to producing X and it turns out you need to produce Y, it’s very expensive… Especially expensive if you’re building $100 million factories.” — Brian Potter
[12:00 – 13:00]
Quote ([12:57])
“The pressure at the early on was just for growth at all costs, scaling up our operations in anticipation of growth ... and it never really all materialized.” — Brian Potter
[14:11 – 19:30]
Quote ([15:20])
“A lot of environmental rules that started popping up... serve to really restrict especially things like the National Environmental Policy Act, what the government is able to do. They give citizens a lot of ability to halt government efforts by litigation...” — Brian Potter
[19:30 – 23:57]
Quote ([22:30])
“AI sort of shows: we still have these capabilities when they are not restricted and when the incentive is correct; but also shows how quickly forces that can shut these things down can move.” — Brian Potter
[25:11 – 30:15]
Quote ([28:49])
“I don’t think the problem with prefabrication was insufficiently clever design. It has more to do with fundamental constraints: difficulties achieving economies of scale, different jurisdictions, site-specific requirements.” — Brian Potter
[30:47 – 31:40]
Quote ([30:47])
“We can still build stuff fast if we choose to… but steadily encroaching regulations and difficult bureaucracy have made it harder for the government to do anything at all.” — Brian Potter
[32:00 – 33:01]
Quote ([32:00])
“There are not super easy solutions. If only we could do X—someone's always tried it, and it doesn't work the way you’d think because of these complicating factors.” — Brian Potter
[33:26 – 36:13]
Quote ([35:10])
“Sweden builds 80–90% of single-family homes with factory methods. Their costs are not low—they’re not producing the Model T of homes. Their homes are more expensive than U.S. homes.” — Brian Potter
[36:26 – 41:14]
Quote ([38:20])
“Once something reaches some level of disruption, people are willing to tolerate it up to some point. Then after that, once they see the consequences, they start to get upset.” — Brian Potter
[41:14 – 47:23]
Quote ([45:15])
“Even if you have Hamiltonian power at the federal level, you’d be bound, or face bottlenecks, because of what’s going on locally.” — Jim O’Shaughnessy
[56:11 – 58:18]
Quote ([56:50])
“The benefit is spread very broadly, and the harms are all concentrated… You have this fundamental asymmetry.” — Brian Potter
[60:12 – 62:48]
Quote ([61:16])
“What Katerra inspired in me was: wow, it’s really shocking how little people understand about what is actually required to make this process work better…But it’s a strength of US society that we’re willing to take these big swings.” — Brian Potter
[65:27 – 67:59]
[70:33 – 71:54]
Quote ([70:44])
“Build more stuff, build more housing, build more infrastructure. If I could find some way to communicate that, it would be truly transformative.” — Brian Potter
On the futility of easy answers:
“Our problems are not like that. Most are due to the inherent nature of the process, not one roadblock that, if removed, everything is fixed.” — Brian Potter ([32:00])
On America’s current constraints:
“Hearts and minds change slowly, but much of it comes down to attitudes about progress, abundance, and what we actually value.” — Brian Potter ([46:05])
On innovation cycles:
“You can also see a waxing and waning of trends, like immigration—first welcomed, then opposed. The same sinusoidal thing affects building and abundance.” — Jim O’Shaughnessy ([53:13])
This episode is a must-listen for anyone interested in technology, infrastructure, economics, and institutional challenges at the heart of American progress. Brian Potter calls for realism, humility, and experimentation—not just tech optimism—while making a passionate case for rolling up our sleeves and, above all, building again.