The US vs. China Manufacturing Debate

I just got trolled. Why did I not learn Chinese? It's not just that you get a factory, you get a full engineering team that co designs the product for you, and they co design the product around their supply chain, around the tooling they've got.

0:00

If somebody else can make all the stuff that you want, that's not good. We have only so much tribal knowledge left. Once it's gone, it's. It's gone.

0:10

In China, there are 20 elons. Maybe their elans are not as good. You can debate that sort of thing, but, like, if I'm an American entrepreneur that wants to build the most advanced device of its category, I can just go there and do it.

0:18

When you hear these analogies, you kind of want to just like hit your head against the wall. Like, yes, of course they're great at doing all these things, but like, what are we doing?

0:28

Welcome to this US versus China manufacturing debate. In the 1980s, the US produced roughly 30% of the world's manufactured goods. And today it's roughly around 16%. Over that same timeframe, China has gone from almost Nothing to nearly 30%. This is created one of the biggest transfers of power in human history. And so today I'm talking with Aaron Slidov and Sam Di Mikko. We're having a debate about what this means and more importantly, what comes next. So, Aaron, in five minutes, can you please explain how this decline happened and please give a little introduction to yourself.

0:38

Sure.

1:10

So I'm a Rust Belt Midwest boy, grew up in Cleveland, clawed my way out, eventually moved out here to San Francisco, lived in New York, Louisiana, India, a couple other places, strangely. But I have a background in physics, engineering. Most of my pursuits have been chasing moonshot ideas, basically. And part of my early career was actually building Waymo before it was Waymo here. And I have a pretty decent breadth of hardware projects. The kind of, like, genesis of how I got to where I am is trying to make a consumer product by myself. And I had no exposure whatsoever to manufacturing. And the thing that I wanted to make ultimately was a stupid plastic widget. And I found out the hard way what tooling was and how to find suppliers. And all of my buddies who worked, like, still at Google and like other big tech companies on hardware teams were unanimously just like, yeah, go to China. Like, go find a supplier, get on Alibaba, get a WeChat, you know, like, find somebody. I tried that for a little while, and ultimately the overwhelming kind of consensus was that I'd have to, like, fly there and live in a Factory basically for a little bit. And this is kind of like a rite of passage. You know, I wasn't really that willing to do it at the time because this was just like a side project of mine. I was like, like, what is going on with this? Why do I have to go do this? And so ultimately I tried to go find American suppliers and that was an even worse shit show because they don't care. I learned over time the reason they don't care isn't because they don't care about me or my product. It's just they've been squeezed to the absolute margin on what they can do. And this is part of the global manufacturing world stage. This is something to me that I thought was an insane product problem because in my brain what didn't compute was we became so wealthy and smart and powerful that we just kind of offloaded all this work. And obviously over time China became very good at what they do. So in the beginning they weren't. This is like a whole train of people handing off skills and learning and undercutting and building capacity. So it's a fascinating story ultimately, but one that really hit me hard being from a Midwest rust Belltown because growing up you never think about why your city is decimated and what happened. This was the richest place in the country for 100 years. You do learn this stuff obviously in high school, in college when you're taking economics classes in poli sci. So it's fascinating to really see it all kind of like unwind in front of your face. And and my answer to it was trying to build a response with the company that I'm now building. And it's kind of like a marriage of Silicon Valley engineering talent with Detroit world class manufacturing talent. And so to me I think that we can do this. It's really going to come down to a matter of will ultimately. So that's a quick background on how I'm sitting here now.

1:11

And Sam, so hilariously we were, we, I think we both started our careers in different buildings in the Google campus very close by each other. And it's actually really funny because so I, I went to college for electional engineering. I also, I do some, I moonlight as a software engineer occasionally but I was on the Google Glass team which is a little different than WHMO because you know, it's a consumer electronics device versus a like at that point it was, you know, in some sense a research project basically. And so the default assumptions were this is going to scale all this other stuff. So I, I graduate I graduate undergrad in electrical engineering and I'm about to start my master's program and I'm like an intern at Google. I'm Google X. And I'm like, I just got trolled was the realization I, I had. And I'm like, why did I not learn Chinese? Because I didn't understand. Like, if you want to build like scaled consumer electronics, you're going to need to go to China if you want to run as fast as possible. And I actually encourage, you know, people still in college or if you've got a chance to, you know, change your grad school curriculum or whatever to like, you should take a year of Mandarin minimum if you want to get into this world. Which is kind of an aggressive statement, but I think it's, it's, it's factually true. So I started going to China in 2013 after I had become a full time employee there and got like, kind of the download of like where to go in Shanghai, like what to, what to go do in sightsee and all this other stuff. And I basically realized like, okay, I actually like doing this. It's like pretty cool to go to worldclass international cities and like and you get to then have engineering meetings and stuff like that as well. And so there was actually this like travel element to it. And also people like rag on San Francisco for like not having great night life and all this other stuff. You can just immediately inoculate that if you're going to China like once every, once every two months or something like that. So, so basically, basically I got picked up into this universe now where this really accelerated was Google Glass obviously was a little too early for its time. Oculus was not. And so I went and joined the Oculus team as they were getting bought by Facebook. And well, they had to, they, they was interesting is it's kind of the same thing that you mentioned, which is like the transition of China from like a lowend manufacturer to a high end advanced manufacturer. I saw this at Oculus. So like Oculus was. The Oculus dev kits were built by, I believe, the same contract manufacturer that made Guitar Hero and Rock Band. And it was a toy, it's a toy factory. And so in some sense it's like, yeah, they can do injection molding of like let's call it cosmetically. I don't want to rag on them because I think it's like, you know, it's a different era basically. Basically think like damp factory in Guangzhong, Guangdong province looked straight out of like the 80s sort of thing like that. It's probably a pretty similar site to what you would see if you were like going to China in the 90s or something like that. And then the advanced electronics piece was kind of kept into a like okay, they put a computer on the factory line that like does some of the advanced like software defined hardware stuff that like you need to do for a VR headset or whatever. So that was where they started when they got acquired. Then as soon as, as soon as the acquisition went through I was involved with some of the initial contra manufacturer selection because I had worked on the Google Glass team on some of like the inside, you know, factory stuff there. And so got to see kind of the next wave of Chinese manufacturers. And these folks are not mucking around. Like this is like it's not just that you get a factory, a building with some operators that they go and recruit, it's you get a full engineering team that co designs the product for you and they co design the product around their supply chain, around, around their like, like around the tooling they've got, etc. So it's not just that you get a room with some cheap workers, it's you get a machine that effectively can build it for you. I, I think the analogy for kind of the audience that, that you've got is like it's kind of funny to use the Claud code for X sort of joke. But like it's kind of like you have a LLM that can build hardware and you communicate to it through Google Slides and in person meetings and sometimes drinking some baijo at dinner. But like that is how it actually works and you are essentially able to be the high level engineering input to that system. Now there's kind of the split brain because you do that but then you're also doing advanced R and D for like hey, how does my computer vision system work? How like a lot of that stuff ends up has stayed stateside and like you kind of see where, why this is like, you know, this is the, this is the same transition that you've seen kind of from like the United States still does the high value stuff, they still do the industrial design, they still do like, like Johnny, I didn't lose his job yet. You know, it's kind of the point like and that team at Apple for instance has like one of every potential machine that they want to use and they, they piloted different processes and they do it internally. But like who scales it? Not them obviously. And so I got to see kind of this transition from like okay, China is this like cheap labor, toy factory, you know, there's lead paint in the, you know, whatever, you know, all, all the stuff that people dismiss China with to like, oh, this is fundamentally changed. And like in the past, you know, couple years as I've been building impulse, I've seen a different thing which is like these guys are not just really good at building stuff. And like you can, you can scale like an arbitrary VR headset or product through them. It's like they are leaders on the R and D side as well. And if you want the best battery cells, if you want the best process engineering for like how you build a battery pack, if you want like all this stuff, it's like it's an email away from you getting it. And not only that, they are super eager to go and work with you and they're super eager to go and work with you for a bunch of reasons. And a lot of them are, I would argue, fairly like, fairly well intentioned. It's like, it's like, well, they see, you know, the United States as like the benchmark of, you know, especially Bay Area engineers and stuff like that is like this. We can learn a ton of stuff from this and this will mean that we're ahead of our crosstown rivals in a hyper competitive like, like you could argue that it's like all these companies have been pumped up by state subsidies and it's like, it's like in China there are 20 elons is probably the way to think about it. In the US there's one and the NIMBYs have basically tried to make it so that there's none. There's, there's none. And, and, but in china there's like 20 to 50. Maybe their elons are not as good, you can debate that sort of thing, but like that's how their hardware ecosystem and software ecosystem as well works. And the benefit of that is if I'm an American entrepreneur that wants to build the most advanced device of its category, I can just go there and do it.

4:40

Yep. So Aaron, I think Elon on a podcast recently basically explained that right now Most of the US's biggest companies effectively just export files that they send to China and then those files are converted into actual products. With this like massive decline over the last 45 years, how do you actually think the US can reverse this decline?

11:34

So part of, you know, this story and building my company and seeing this stuff kind of firsthand led me to also try to start a national movement around this which is re industrialized. And the idea of that is we have only so much tribal knowledge left in terms of how these systems work, how you can ultimately create a machine like that. But once it's gone, it's gone. And preserving tacit knowledge is very hard because it's not like the Matrix where you just jump in and download.

11:53

I know kung fu.

12:35

Right?

12:36

Yeah.

12:36

No, there's a take on this that's like machine. So I would also argue that like, effectively you machine learn or you reinforcement learn how to manufacture stuff. And China is incredibly good at this process. And so basically it's like, okay, if we can break this down to like a simple example of like, let's build a lithium ion battery cell. It's not really simple, but there's all of these process steps where you're basically like laminating, like laminating films on top of each with different chemical processes and like the, like the makeup of the, the powders you're using and like the settings on the machines that do certain steps. All of that stuff you can imagine. Again, I'm assuming your audience is full of a bunch of aspiring AI engineers or whatever. All those things are like implicit weights into a neur, like a neural network, basically. And you can pre train, you know, you can take all the knowledge in the world and design the best thing, the best single widget. But, but there is a reinforcement learning with factory feedback step that unless you build stuff at high volume, you will never achieve. And this is actually the thing that like, I've been talking to a number of folks doing kind of like new manufacturing things, and it's like that is the key thing that I think a lot of people who have not been close to the process, they don't get basically is this like implicit reinforcement learning aspect of it. And the fact that like so much of the state of what makes advanced manufacturing work is not in the spec.

12:36

Yeah. If you think about it like a stateful system, it's impossible to observe until you have high throughput, which is mass production, basically.

14:01

And he's going to fund you for that.

14:11

Exactly. It's very hard to understand all the parameters basically until you do it. So that's why the Chinese machine is so efficient, because they are just gathering more and more tribal knowledge and it's going to snowball out of control right. Until basically we'll get to this later. But the humanoid thing is an insane problem if it actually plays out.

14:12

What do you mean by that?

14:38

Well, we'll get there, we'll get there. But basically the whole purpose and mo of re industrialize is that I would prefer if we could preserve as much of that knowledge as we can, encapsulate it and then rebuild it here. Right. And be able to have these kind of like, reinforcement feedback loops because we're building better brains and like, compute. And it's, in my opinion, it's not necessarily that, like, we need to live in a unipolar world with a single industrial superpower. But, like, there has to be another option. And I also, I don't believe that, like, autarky is a necessity. Right. Like, you can't. Trade is actually a good thing. However, when you do actually have an adversary on the world stage, you cannot continually empower them. So part of my ethos about this is that we have insanely deep capital markets, we have the talent. And this is where it comes down to. A lot of people criticize this in terms of communist central planning. Basically where you need, you're going to need this kind of capacity one way or another for medicine, for electronics, chips, all of this kind of stuff. We scrambled to cover EV supply chains and chips really fast because those were critical at that time. And it really just turns out that it's all critical. If you forget how to make anything, it's bad, right? So ultimately, my whole worldview is that the toolbox is here. We need the people to actually, like, do it while we're still kind of, you know, working with China and other industrialized nations to produce the things that we want to build.

14:38

And Sam, same kind of question of, with this massive decline, do you think like a, it's reversible and how do we do that? And then if you had $100 million to invest in a factory today, where would you decide to put it and why?

16:37

That's the great set of questions. So first thing is, is this reversible? I think there's a silver lining to like, some of this stuff. And actually, I'll disagree with your point of, like, it's all important.

16:49

It's.

16:57

I think you can distill kind of all of the next set of devices into what I would call like the electric stack. And so this is. And, and if you want a 200 page reading on this, Packy and I put together a really long article called the Electric Slide. But basically it's lithium ion batteries, power electronics, embedded compute, and then electromagnetics. And I would also add a fifth column, which is like computer vision hardware. So like, I, I've worked on came and image sensors and stuff like that. So like. But basically anything from a Waymo to a robot to a stove of all things, you can use those, those, those technologies and actually make stuff, stuff happen. And so basically the idea is like pick devices that kind of give you, let you exercise all of those key, like those four to five key muscles, make sure there's a portfolio of them, some like portfolio of them that you can then selectively onshore and that, and make sure the volumes are high enough that you can get at least one of every process machine needed in that entire supply chain onshored and at 100% capacity. Because the problem is like if, if, if I need like five different types of injection molding machines and I'm using each one of them for like, you know, 10,000 parts a year, well I'm not going to be able to fund those. Like, it's not going to, it's not, it's not going to actually make sense. The flip side is like the only industrialist in the United States that has actually successfully done this recently. Well one, he doesn't share, he does it in his own ecosystem. But two, but two, he actually can bring the volume to bear to actually go and do this. Like starlink has enough, enough customers that they can fully do a lights out manufacturing line for that, that, that system and it makes financial sense to do so. That said, how did he actually do it onshore? At the beginning it was like pandemic era. Probably a pain in the butt to go to Vietnam or China or something like that to go and do this. They went and did it in the parking lot in, in la. And I think that goes back to where I, what I put my factory. And it's like, it's incredibly important and I said this earlier, it's like, it's incredibly important to tie the, like the factory and the engineers need to basically be in sync. And this is the superpower that companies like BYD bring to bear. Like literally the factory and the engineering office are like in the same facility. In most cases, like maybe it's a couple buildings away, but it's not, it's not across town.

16:58

It's in the same campus.

19:21

It's in the same campus and there's all of that tacit knowledge of like, yeah, you can't really design it that way because this doesn't fit kind of thing. Well that's like immediately short circuited and the context is already present. And in the US we decided, oh, you got to like go do a bake off with a bunch of governors and like, I mean you're going to watch me, you know, be a hypocrite. On this when I. Whenever I open a US Factory at some point, but it's like, okay, I got to go. I got to go. I'm trying to make a stereotypical Republican governor name, but basically, like, I got a base. I got to basically, like, go to some red state, tell him how much I love Donald Trump, you know, all this other stuff. And then I get my tax breaks and I get my free campus, and I get all this stuff. Meanwhile, my engineers are over in Hipsterville in California, and they don't want to move over there. They don't want to move to, you know, real Midwestern small town that I am re. Trademark. Re industrializing. And so this is actually. I think this is actually a problem. So it's like, I think there's like, two solutions, one of which is like, we should convince some smart engineers to go move to move to move to these actually kind of awesome Midwestern towns in Ohio. And like, if we can basically rebuild those, it's like, they're going to be great places to live for smart people, because those smart people lived there before, like, these were walkable communities with, like, you know, with great amenities and like. And again, the engineers were next to the factory. The flip side is, like, where would I do it today? Well, where are the smartest engineers on the planet right now? The Bay Area. So you should put it in the Bay Area. How do you do that? Well, there's a guy who successfully expanded a factory in the Bay Area by putting a tent in the parking lot, because that was the only way you can get around all the permit constraints and all this other stuff of, like, needing. I need to. Need to expand facility in Fremont. So I think that basically the way I'm approaching it from my perspective is the. The Yimby movement was a good start for, like, let's unlock housing production to, like, make it so it's actually affordable to live in places like in the most popular places to live on the planet. But there's going to need to be an industrial movement that is kind of its counterpart if we want to be able to, like, you know, make this stuff work. And that that movement is going to need to be kind of like a. Yeah, it's okay to build a factory now. And the contrarian take is that, like, okay, we basically built up scaffolding of, like, environmental laws from the 60s onward that have basically made it illegal to do stuff. I'll give you an example of, like, California, bit by bit, has restricted what I would basically call chemistry at like, a foundational level where it's like basically illegal to do chemistry with metal. And if it's illegal to do chemistry with metals, you can't build anything with the electric stack because fundamentally the electric stack, that's batteries. Also. If you want to coat any metal, like you want to anodize anything. Yeah, good luck. You got to do that in, you got to do that over in Reno.

19:22

Can you explain what you mean by California made chemistry with metals illegal?

22:05

Yeah. So basically California was an industrial superpower. We built like, I don't know what like half the world's battleships. At one point we built like the like Kaiser shipyards. Like we had all this awesome stuff going on also this whole facility by the way, like this was all, this is the center of like manufacturing of like, like the bridge superstructures for aircraft carriers were built like right around here and stuff like that. Awesome, awesome. You know, industrial, you know, industrial spot. But obviously if you do chemistry with metal it does cause pollution. And so there are sites throughout the Bay Area and this is like within human memory that have like trichloroethylene contamination like the Google campus. There are I think teflon barriers below some of the buildings. And because essentially like, like the polluted pollution from both metal, like metal working and semiconductor production, because we also used to have tons of fat, all the, most of the fabs were here too. And effectively that, that pollution is still with us today. And you see controversies ranging from like hundreds point hasn't put new housing in where the old naval shipyard because they keep finding like little radium painted markers that were like on like if you worked on an aircraft carrier at night, you'd basically put these little glowing green dots on you so that you don't get hit by you know, you know like a plane landing or whatever, they go find one of these and then like that shuts down housing production for like five years. And so but basically because of the past memory of like you know, I would argue kind of reckless environmental stuff, we put in a bunch of scaffolding of new laws and, and these were all well intentioned and we built them over many years. But now it's effectively illegal to do chemistry with metal. And so, and if you. And, and so that seemed fine, you know, back you know, you know back into the 90s and 2000s, it's like hey, we're, we're, we're. There still was consumer electronics manufacturing in the Bay Area. Like the like Google Glass for instance was like a consumer device manufactured in California. And they had a flex that Said they, they had flex on the box. It was like designed by Google, assembled in California to basically tell Apple like, you know, you know, a little bit. We know and I've been in that factory. It was pretty cool to actually see an American factory building the most advanced consumer electronics device at the time. And they did a good job. But where were the plastics injection molded? I think it was China.

22:08

Yeah.

24:27

Where were, where were the flex circuits made? I think that, yeah, like, it's like you think of, like you think of all the core components. Where was I going for camera module manufacturing? Yantai, China. So like you could already see the component. The, the most advanced pieces of that were getting pulled out. Where were the chips packaged? Probably Malaysia. And so we built these global supply chains over the years for, you know, again, well intentioned reasons. But like then when the EV revolution hit when like basically it became very obvious that like everything will be built like a smartphone, everything will be with the electric stack. It'll be like it doesn't matter if it's an iPhone or a BYD Han EV or A, or even like most of the like actuators and stuff in like Starship, you could argue Starship is like much more an electric vehicle than any rocket prior to it. And so all of these things use the electric stack. And California really prides itself on being in the forefront of, well, it was in the forefront of the EV revolution, it was in the forefront of the solar revolution. It was like, like all of these technologies were kind of like really scaled here and we've like built scaffolding of environmental laws over the past 50 years that have effectively banned us being able to compete at a vertically integrated level with these like on these technologies despite, you know, a lot of the best engineers in these technologies being here.

24:27

Aaron. The first like two factories I believe with Tesla were here in the United States. And then when Elon went to China, the Shanghai factory was like the fastest ramp of any factory that he'd ever built. And I remember him saying that the Chinese were just significantly better at basically producing these cars in that factory. Like that factory is just fundamentally more efficient. So what does that kind of say to you? And like how do you think about that?

25:42

I mean, you can't refute it really. The, the way that I think about that is you hear him kind of talking in most of his appearances about manufacturing, about, you know, just like the level of urgency and seriousness that they have about what they do. And you know, that's honestly kind of like a, a moral dilemma in some Ways for. For Americans, because we, we really like to. With environmental laws like this. These are all reactionary snapshots of something that happened in time. And we don't revisit those things and we don't do it in a more proactive way. Everything is always reactive. And this is.

26:06

And it's also how our legal system works too.

26:52

Exactly. Yeah. Americans are. They don't really do anything until they have to.

26:53

And we do it the best.

26:58

Exactly. So this is kind of like the way I think about this is like, yeah, Elon is very, like, enamored by how good they are at this because we, we aren't doing the same. Right. Like, we don't have. We don't have that same level of urgency. Right. And financialization of everything has really driven us to chase, you know, the higher levels of this. And I think that where I don't necessarily disagree with the electric stack, but I think in a lot of my work, we work with a lot of the largest OEMs and manufacturers on the planet, and it's a constant whack, a mole game of where do you want to do your final assembly of the thing? And then it's kind of like, now I need to assemble a bill of materials of what's going into this, and then I have to go build the supply chain and chase all those components around so that I can assemble it efficiently somewhere. And all of the, like, everything that's, you know, kind of like ramping to a higher level of technology requires all these components that we have just, you know, totally shit the bed on, hands down, everywhere. And so my job now at Atomic is finding, right, like all of these places where the lower level componentry, like metals and plastics, have to chase the electric stack components around to get assembled somewhere else.

26:59

I've seen delays on modules where it's like we need the enclosure for our battery pack was a long lead time versus the welding of the cells, which is like crazy.

28:34

Yeah. And in cars, the two longest lead time single components on a car are the bumper and where your doors close into the car, the sheet metal piece. Those come out of molds. So one comes out of a sheet metal stamping die and the bumper is injection molded. There were stories back in the early days of Tesla of them getting their bumper molds cut in Japan and flying them to California on 747s just to cut two weeks off of production time, if that gives you any indication of just the lower level components and how urgent they are to get something final assembled somewhere. So, yeah, I mean, it's like supply chains are a game of whack a mole because of the way we've constructed them like this. And we are to kind of blame for that because we built them intentionally that way.

28:43

I mean, and you go to BYD and then you have a very integrated campus where it's like, oh, like they have much of the dependencies of the electric stack all in house. In a lot of cases now they, they also go to sub suppliers. They also do do all this other stuff. But like you have the optionality of being able to go full vertical. And then you see like the, you see what Elon's been doing lately, especially with like Starlink and, and, and, and, and stuff like that, and you're like, okay, this guy is trying to go full vertical where he has his own industrial concern. Now I would argue you need to start from demand and pull everything up because you can't finance this with venture dollars. Like I, I don't think you can. Now some of our friends have successfully financed large manufacturing concerns with venture dollars. But, but like you're, that's not going to like, that's not the way that the capital markets work in the United States. And there's actually, I think one of the other pieces of kind of deindustrialization that people forget is we didn't have national bank branches in, in the like, you know, like in 1900 or whatever, like when, you know, this facility was built. But if you had a local bank in your town, it could only invest in like the, the, the, the crazy people in the town. And so it was a lot actually in some ways easier to get capital because the markets weren't as mature. So they're like, I guess we have to invest in, you know, stove guy and injection molding guy because they're the only game in town doing, doing stuff that actually can, you know, make me returns. Now we've replaced some of those systems with ever bigger venture funds that are kind of more like private equity firms, especially at the top level. But like, I, I think that you like what Palmer Lucky's doing with Erebor and all this other stuff. It's like, there's definitely a reaction to, to like we, we, we, we, we forgot how to do the capital market side of this thing. But there's another thing that I, I think is really important to just highlight is we're in an AI boom right now. And I think it's like there's an enormous opportunity from the AI boom and people are missing what it Is so the, the missing piece is people think, oh, we can onshore stuff because we can get humanoid robots. And like, it's going to magically solve these things. And it's like you can just automate the factories, whatever. Like, you know, you know, honestly, that's stupid because if you look at what, how Starlink onshored their consumer electronics line into the millions of units, it was a manual production line that they later automated like Elon's little algorithm or whatever, he says the last step is automating. Like, he learned that the hard way because he laid out an automated line at Tesla the wrong way and almost killed the company. Almost killed the company. And so, but the point is, like, you can actually start manual with final assembly and be totally fine. And this stuff works. So, so going back to the AI, boom, what we're missing, what we're clearly missing, currently we are building, like the hyperscalers are ordering tens of thousands of units of like, let's call it three to four million dollars data center racks. And these things have like, you know, hundreds of GPUs in them. In some cases, like 100 GPUs. Something kind of crazy. They've got all of the various pieces. They've got energy storage built in. They've got like, they've got all of these different, like, they've got advanced power supplies, they have the full electric stack at the most advanced node. They have all this photonic stuff. They have like all of these key technologies. And the demand is there. The, and the demand is there installing it in the United States and until,

29:36

until all these states shut it down.

32:47

Well, and then, and, well, the enemy of manufacturing is nimby. But yeah, and, you know, some of them could be. I'll go into conspiracy theories later. But basically, basically the, the point is, like, there is massive demand for the most advanced computers in the world. And these are in some sense, electromechanical systems. Like, they've got, you know, they've got all of the. They've got injection molding, they've got stamping, they've got all of this stuff. And why is, why aren't there not American supply chains spinning up to service this demand at scale? Like, it seems very straightforward to be like, this is possible. Instead, what's happening, and this is a very spicy take, is there's no tariffs on this stuff. So meanwhile, like, you know, meanwhile, like, I gotta pay the full, full tariffs on, you know, stoves or whatever. But the other, like, you know, the other Sam down the street, he doesn't have to pay anything. And so, and I think this is actually like, I, I don't think people won't like have talked about this in like a, in a, in a very clear way. But like I just see this immense opportunity to use the data center boom to pull onshore a bunch of these advanced capabilities. Just like how the EV revolution and like, and like the automotive companies going EV has been able to pull some of the battery technologies and other things onshore and we're not doing it. And I think, I think, I think that's going to be a big regret that we'll have five to ten years from now because it's like, cool. There's this immense capex boom that we could muster, we could muster to basically onshore advanced manufacturing.

32:48

How do you think about that? Because if right now we're in the early stages of basically putting in all the weights and training, you know, this, this new supply chain, how do we basically reverse that so that we do actually end up with the, you know, most advanced manufacturing?

34:17

I mean, I mean if you had me run industrial policy, it would be, things would be a little different. But like, but basically I don't want to nominate myself, you know, in future administration. But basically I want to run businesses. But basically, basically I, I think that we need to be thinking about this from the systemic level of like you need to have a baseline capability of being able to do this stuff. And to, and military demand is not enough to sustain. Like I got someone was yelling at me about like, like how I didn't understand American PCB manufacturing. And there's like this one company that basically makes like all the US military's PCBs. And I'm like, I actually have never heard of this company because they don't do anything for consumer electronics except for Huawei apparently, which is a little interesting. And so we basically don't work that muscle at all now for the reasons that like I think you've described, which is like the American vendors are like, have been squeezed so much they don't want to play ball with like crazy stove guy or whatever.

34:28

Yeah.

35:24

And frankly like then you go, go to China and it's like red carpet rolled out. Hey, do you want to work on some cool stuff? It'd be really cool to build that because it's just cool.

35:25

There's a South park clip about that.

35:34

And, and it's like, it's like, and it's awesome. And also, also, also like it's such a flex when these companies go and like pick you up in a car or whatever. And it's literally a car they make. I mean there's another. The only other company can do this is basically like, you know, in the Elon universe, obviously, obviously that's, that that exists. But it's kind of like, oh, but In China there's 20 of these and there's 20 of these and you can hire them to do advanced manufacturing projects for you. So like if I could go hire Tesla to go build the best induction stove in the world and they gave me their Bay Area engineers to do it, would it be worth it? Probably. I think it would actually, I think it would actually work great. Now are they set up as a business to do that? Absolutely not. And that's the other missing piece of the US ecosystem is like, I don't think that's a venture scale business. Or maybe it is, but like, maybe I need to be more thoughtful about what the financial structure ends up looking like. But we don't have the incentives in the capital markets for businesses like BYD or Gore Tech or like Longshear or. I'm thinking of a couple. Or, or, or yeah, two exist. Basically.

35:38

I can push back on this a little bit just because I've been trying to make a case that this level. Basically let's just build Chinese level capacity here. Right, Good.

36:51

Take

37:02

the venture backwell piece is really interesting because this is where the rubber really meets the road. So it's really challenging. If your narrative is I just need a factory in America. We're going back to the systemic level here. So you can't walk into a bank and just say I want to go get like a machine shop, you know, like hook me up with 10 million years ago.

37:05

You could totally do that.

37:27

Exactly. Now no, in China you can do that. Yeah. And it's just like the kind of time and place of the shop and the SMB. We might be looking at like the end of that. Right. And so the idea that. But you could, if you actually do this in an intelligent way where you have, you might start in a more manual capacity kind of setting. But the dollars that you need for the R and D and for market capture is very well suited for Venture, especially if you just bum rush it. Right. It's probably the only hope that venture has at building these companies at any scale because otherwise it's like you're hopping on one leg with your arms tied behind your back, you know, and you have to kind of like build a little factory and show some good progress in unit economics when like the bigger Prize is just like building you know, multiple hundred million dollar factories and like can founders actually, you know, do that?

37:28

So I have, I have, I have, there's a spicy take that is like hiding in plain sight which is like related to like Android for instance. And so Andrew has announced a lot of like, hey, we're going to build this like advanced manufacturing campus in Ohio. We're going to do all this other stuff I think like, you know, like by the way like they'll IPO and be very successful. So like, like I'm going to lead with that piece. We'll lead with that piece. Like I, I'm very bullish on Andrew, the stock. But the, there's a, there's a flip side of it that I think they even know. Like they, they like, they know this based on kind of, you know, this is not confidential information or anything like that. But like the, the spicy thing is if you build a business that is essentially like designed to service the US Government as a customer, the US Government until very recently has not been the customer that would incentivize anduril to build base level capacity of like the base capabilities. They're much more of a, like okay, SOCOM wants this type of drone helicopter, the Marines want this type of anti drone interceptor. There's some high volume but really low end requests coming from Ukraine. There's like, you know, like basically that you end up being a product portfolio company and in some sense like, and it matches venture in a way because it's like they're not, you're not building advanced factory, you're building, you're building a, like you're, you're basically building a bunch of options that could turn into the next F35 program or whatever. But what that means is you have a company that's almost 10 years old that like hasn't ramped volume production of any one item and then use that to teach itself to do the next thing. And you can contrast this with like SpaceX for instance, which is like much more limited in terms of product scope. Now SpaceX found a very deep market that they uniquely could dominate, which I don't think the DOD has provided yet. But I do think that like this is a really important thing of like this is where I'm also getting obsessive about the data center things. I'm like that's a very high volume deep market that you could just go and you could go and just do. And so I think, I think that's something that's kind of worth describing where like Andrew now has to learn this muscle now because okay, we've got the drone dominance program. We've got all these other things where it's like oh, will DOD order or do, I'm sorry, order, order a million of one, a million of anything. And yes, they order millions of bullets, but they only, only until the Ukraine war were they ordering millions of artillery shells. And now they're going to be ordering millions of drones which again use the electric stack. And so you then have to be good at that. And then if I go have an FPV drone and put it on the hypothetical table here, the lithium ion battery no one really makes in high volume yet high discharge rate lithium ion batteries, pouch cells in the US at scale. You could argue you could, you know, repurpose the Tesla Panasonic line for something like this. But again, not current. Like Elon's not going to share necessarily. You then look at the, the PCB, probably doable in the U.S. but the chips there. Are those packaged in the United States? No, probably packaged in Malaysia. The, the camera module on that thing. I believe there's basically zero camera module manufacturing in the United States. If you wanted a cheap thermal camera, that sensor is probably made in China, not made by Fleur. You can go down the list and basically be like, okay, there's a lot of tech that's going to have to land stateside to kind of, you know, you know, or, or friend short basically to, to do this. The US can totally do this with, with their allies. Like you know, you know, like, like US plus Japan plus Singapore gets you a good chunk of, plus South Korea gets you most of this. But I don't think like, my point is like I, I think the like defense procurement world has been so separate from the consumer electronics world for like 30 years. Basically like since the end fall of the Soviet Union these have been like we've been split brand here that I think that it's going to be a challenge.

38:32

I posted about this this morning, I think where it's just like, you know, less than 5% of everything in defense is produced at higher than 10,000 of anything per year.

42:34

And this is actually going to cause reinforcement. Learning into stupidity is the problem because basically like if you're, if you're like, oh cool, like I'll give an example, like the sat, like if you're building a satellite, so like Astronis across the street, right, so you're building a satellite, I can pick on them because I'm an investor. So I'll just pick on all of These, I'll pick on all these Neoprimes. I think Astranis has like the supply chain for this. They end up wanting to vertical. You end up vertically integrating your supply chain just like SpaceX did. Because you're finding out that basically the suppliers that claim they can go and give you something, whether it's like, you know, ion propulsion, solar panels, like, like there's a whole list of stuff I'll pick on star trackers because that's like a, that's a great one because it's just a camera. And so. But then if you go to Ball Aerospace, they're going to charge you several hundred grand for a camera or something like that. Maybe they've gone cheaper but, but the idea is like, because there was like only one game in town, we were only launching tens of satellites a year or whatever we were doing back, you know, back before, you know, back in the before times, these vendors basically were like the only game in town. And they essentially learned to, you know, they're not consumer electronics companies, they're consulting, consulting firms. And so it's like I go and order a like verified space grade part and the thing is basically like the guy who designed it has been dead for 20 years and it's manufactured by a company that bought another company that bought another company. And you're, and maybe they, maybe their technician forgot how to, you know, do a certain process step or something like that. And you don't know any of that because this is all black box to you and it's itar and all this other stuff and like they all claim, oh, this is the most advanced thing ever. And it's like meanwhile like you go ask me to go to the Shenzhen electronics market and I can be like, cool. That plus that, plus that gets you a thing with like 10x the performance. And this example that I kind of just made up appears to exist across a huge swath of American capabilities. And so if you're trying to rebuild American manufacturing, you're building like getting your supply chains to actually be capable of volume, you end up falling into the same solution that Elon fell into, which is like, oh yeah, like Starlink makes their own printed circuit boards. And then you run into the chemistry being a legal problem and you give up. Most, most people give up.

42:45

And Sam, when you were setting up the supply chain for impulse, why did you decide to go to China? And then what was the like main limiting factor stopping you from just building the same supply chain in the United States and vertically integrating?

45:06

Yeah, so let me Start from the demand side. So we're now at the point where, like, there are. It seems like we're on track for tens of thousands of units a year of, of cooktops and other things. But even with that capacity, let's start with like. And that's a. You know, like, you, like, you could be like 100 million-plus in revenue of your end product and you would not need even one. Like, you would not be able to fully saturate the demand, the, the capacity. You'd not be able to fully saturate the capacity of like an injection molding machine in some cases. So then it's like, why are you buying this machine that is sitting idle 75 of the time or something like that? And, and like, even if you can go finance all the capex, you can do all this other stuff, but it's like, okay, you're. You're literally using something not at full capacity. So the incentive is to go and fractionally do that. Like, just like for the same reason that, like, if I'm a mobile app company, I'll use aws. Exactly. Like the exact same reason you'll want to fractionally use these resources. And so I think there's like a logical incentive to kind of split the. Split the universe between design and build, where it's like, okay, and hilariously, the Soviets were really good at this model. They had, would have like a design agency and then they'd pair it with one or more factories that were often in the same town, but that factory would be like, building. And you see this in like, basically you have the design agency that knows how to design tanks, and they're like the best tank designers in the world. And that the output of those tank designs go to the factory across the street in Ukraine and then the factory across the Urals in, in Russia. And that's like how the Soviets built, you know, a lot of their military infrastructure, their space infrastructure, all this other stuff. It's basically Foxconn. And Apple is almost the same, almost the same analogy because like, Foxconn taps different types of capital structures than what Apple can do. And when Apple's ramping production of a new product, like, they're clearly not using all of this, all of these machines. So like, they can fractionally, like, they're not paying for like the, the next iPhone is not paying for all the complete equipment that is needed, you know, to produce for it two years ahead of time before, while they're still prototyping it, basically. And so you back this up and you're like okay, well also like how much money can I raise as an entrepreneur? It's like this is my first go at it. You know, I'm not, I'm not getting a billion dollar seed round or whatever like these AI lab people. If I had a billion dollar seed round, maybe the trade off's a little different. But, but basically, but basically it's like what can you do with the money you can raise and the leverage you can get? And it's like, well you definitely want to use a contract manufacturer and you definitely want one that ideally comes with a lot of engineering team. So that way like complicated design stuff, you know, you can, you can leverage them for that. And then the last piece is you also want one that comes with supply chain, a lot of the supply chain. So like you give them a complicated thing, they figure out what, how in their network to source it. And then the last thing is you're a startup so who's going to work with you? And, and once you kind of go and bake all this stuff off, you kind of end up with like a couple options. You can, you can go to firms in Singapore, you can go to, they exist, there's just not as many of them. You can go to firms in Taiwan if your thing is pretty much a standard consumer electronics device. If you want to build a laptop, you want to build a phone, you can go to the Taiwanese ODMs and do it. South Korea and Japan have all the capabilities to do this but like LG is not a country manufacturer in a lot of cases. Japan's got a lot of the sub suppliers at excellent quality, but they're not, there's not really, there's not really a final assembly house that will play ball with you. And so yeah, you're in China and so that's kind of, that's kind of the, that's kind of the play. And then why do the Chinese want to work with you? It's like they view, they really respect Silicon Valley and they think that, you know, next generation products come out of there and they want to play ball and helping create them in a lot of cases. And they've got their own, you know, they've got their own hyper powered companies doing similar stuff as well. So I think that's, that's frankly how you end up in these situations.

45:18

And Aaron, Apple is like a $3 trillion company and yet they manufacture virtually nothing in the United States. And Tim Cook has famously said that the skills and all the supply chain required to build production at that scale is all in China and it doesn't exist in the United States. How do we basically reverse that trend?

49:10

He has a really famous clip that I think is it's got to be close to at least 12 or 13 years old now at this point where he's like, I can't even fill a conference room full of tooling engineers in the United States and I can fill a football field full of them in China. So it's very representative of just the collective of tacit knowledge, the people that actually understand how to do the job, the process, all of it. And people have been sounding the re industrialized alarm for 40 years since we started doing this. And it's not an easy trend to reverse, but it's something that's going to take decades to accomplish. This is where I think this ultimately just comes down to how serious we really treat the matter. And it's being able to absorb a lot of this kind of demand. Right. Like manufacturing is very inflexible ultimately when you come down to I'm trying, like from the very lowest level, from refining, you know, raw material and pumping it into illegal in California. Yeah. Another factory that forms it into some, you know, other raw thing that can be pumped into a component stack or whatever it is. There's really in my brain and my estimation these kind of feedback loops of learning all the processes and capturing that knowledge. There's such an immense amount of opportunity there. It seems very insurmountable because there are literally thousands of industrial processes that you could do something like this with. We have to stack rank them. And so the idea of re industrialization is if you want to stack rank, that's fine, but there's still an ocean of things just sitting here that we could go through the same set of motions with. And yeah, I mean, if we don't do this now, what does the future look like? So I guess in a lot of ways I live very much in the future. Right. Because it's obvious that we should be doing this stuff up and down the entire industrial base. And if we don't do this, what does that mean? Are we going to be relegated to some kind of dependency model that allows China to basically take power because they have all this industrial might? And this is where it goes back to how we ultimately stack rank these things, what the timelines look like, where the capital is going to come from. And so if you go back to the demand side model, this is why it's really hard to do this on a shop by shop basis. And this is why you're seeing a BYD city being built. I don't know if that's actually real or not, but it is. Yeah. That in my brain is just a level of seriousness that we need to step up to. Right. And yeah, it's a huge systemic problem across regulation, across capital formation and labor. Right. Because another piece of the systemic issue too is you know, everything is a jobs program apparently. And you know, unions are, they could be a force for good. And I think that, you know, we've seen blunder after blunder of attempting to you know, stand up capacity, get things built here. And they're getting built but we're learning the hard way.

49:29

There's actually a highlight on the union side in California. There's actually two, there's, I think there's a split, there's a split brand in California on the union side because the carpenters unions have actually really embraced the Yemi movement because realizing that this is additive versus like okay, we don't really care about growing our membership. We're just going to make sure that we can, we can rent, seek as much as possible. Obviously that's bad but the carpenters union is like no, this is great. We, yeah, we can build more houses and get paid more money. And so the incentive alignment there with like you can see with California forever they actually got the unions on board with it which is crazy and an awesome achievement. You can see that like the, the yimby California YIMBY managed to align with enough of the union interests even though like DSA and some of, like the, there's kind of the left NIMBY movement basically because I, I, I know why that's the case and it's, it's interesting but it's basically because like the nimby, the NIMBY to win political power, become whatever political party can win in, in their jurisdiction. So in, in San Francisco they're leftists and in, in Florida their rightis. And so it's, it's kind of, it's, you'll see, you'll see, you'll see the chameleon of the NIMBY thing because that's a separate axis I would argue from the left right political axis. But you can see that like okay, if let's call it California, the anecdote of California forever. But also the YIMBY movement had figured out how to make collaborate with the unions in a positive way. Then the flip side is like okay, then there's the like Berkeley academics using SEIU to push all the billionaires to Miami thing. That's going on. And that's where I think we're gonna like, we're gonna run into again the split brain situation of like, I think we have to make sure that like whatever movement here has to be kind of pro labor, but it doesn't necessarily have to like it has to be pro labor in an incentive aligned way. Basically.

53:38

Yeah. I mean to be honest, if you go back and do a re listen of like JD's speech at the A16 summit like last year, it was basically that it's basically like we need to uplift and dignify people's work again. And there's a huge argument to be made that's not zero sum, right. That we don't have to be working in 50s or 80s factories anymore. It's not this terrible, awful thing where you get paid less than an Uber driver. You don't have to be.

55:25

No, you work at an advanced engineering facility basically and maybe you're a service tech for robots. Like the, it's, it's a cool job. Actually this is the other point is people misunderstand. It's like factories like this were like emitting pollution and all other stuff because like that's what you did. What does a consumer electronics factory emit? It's like an Amazon warehouse. And by the way, we allow those in our comm. Like we have figured out how to streamline permitting for, for, for stuff that like allows the consumer economy to work. Like all of that stuff just works as is today. Like there's no issues and, and because like if we got rid of that there would be a consumer rebellion. And so you know where these former factories in literally in the dog patch in San Francisco. Well some of them are Amazon warehouses, some of them are cloud kitchens. We're all, all okay with that stuff. And like literally a consumer electronics factory, same, same industrial profile. And I think this is very misunderstood by kind of the regulatory. Like I, I think the people who know know. But it's like this is misunderstood by like the regulatory apparatus and we have to kind of update the priors there in a, in a key way because like you could have consumer electronics factors all over the place here. Then the next bottleneck is well if you want to employ people making 30 bucks an hour in San Francisco, you need to figure out where the housing is. And you'd have to put some housing here. And if you go to a. So, so for instance, like if you walk outside of BYD's headquarters and you, you, you see all of the vehicles they make like you know in the parking lot or whatever. Like it shown up in a lineup, it's pretty cool. But directly across high rise housing for the workforce and looks super nice, pretty good amenities, all this other stuff. And like that is the case across every factory in China. They will basically go and say like okay, there's a housing development or a dorm or whatever depending different levels of, you know, luxury and stuff like that with the manufacturing facility. And there's only one case I know of of this happening in the US right now that's like recent. It's just Starbase. Yep. And so anyway I was there like probably they're building, they're building, they're building mid rise apartment buildings in Starbase. And so I think the level of seriousness is you need to treat this as a full system solve. And the full system solve includes the built environment and making sure that we streamline that and we should treat. And there's actually, we've already have precedence for how to do this in a way that works. It's called like how we've enabled a huge build out of like Amazon warehouses and I mean arguably data centers as well. It's like Amazon warehouses, data centers and like cloud kitchen type stuff. Those have no problem. There's another thing that Tulsa look at that's super important is there's actually a part of the United States that has not de industrialized and it's the consumer goods economy. It's like your potato chip bag, like your, your bag of potato chips, your Pringles can, your Gillette razor, like the stuff that you see in CVS and also our food supply chain and stuff like that. You can't really move it very easily. And it automated in the 60s before we had any of the outsourcing boom. And so like I was told by, I was, I think, I think Drew Baglino said this on other panel we were on. But basically it was like there's really good talent if you pull them out of the consumer goods industry in advanced manufacturing. And that talent was a great source for Tesla making battery cells for instance. And so I think there's things that we, we can look at as examples of like well clearly you can build facilities on the scale and operational complexity of a factory in cities like San Francisco today. And it's like look at Amazon, look at the cloud kitchens, look at all this other stuff. And the second thing is like where is the talent to staff them? And it's like well who's bagging the potato chips? And maybe you can cross train them to actually make chips. I don't know.

56:02

Sam, when we were talking off camera, you mentioned something very interesting that I want to touch on. Does Apple export files or if not, what do they export?

59:42

So there is, there was a picture that was sent or it was posted on Twitter I think a couple of years ago. And it kind of made its rounds around the office when I was, when I was at Facebook or the company formerly known as Facebook. And it showed United's top. It was like at a conference like that, it showed United's top customers. And the number one customer for United was Apple. And it was like to the tune of like 100 million bucks or like it would order magnitude like $100 million in basically business, exclusively business class travel spend that they were, you know, spending on their engineers, getting them, getting them to go all over the world. So I would actually argue that what Apple exports is engineering expertise, often in person. And they actually bring to bear a lot of the automation hardware that ends up powering their factories much more than you would expect. Like Foxconn in many cases kind of supplies a room that Apple fills and a labor force that, that, that works than Apple actually, you know, exporting concepts to Foxconn where they then design it like it's much more of a CM model than a contract manufacturer model, than a JDM or joint development model than you would think for a lot, especially for their primary programs and stuff like that. So yes, while you're exporting files, the factory has been micromanaged by Apple to such a degree that it is effectively an Apple facility. The other thing is where do they invent the process steps required to build these products that are so crazy. Like I'll give an example of like I would be so jealous to be able to basically do this sort of like this sort of like ceramic to, to metal like, like interface where there's no gap. Like, you know, I'm jealous because I have a product that does something similar and we, we, we have, we have silicone on, on, on the gap like appliances would. But like Apple likely pioneered processes for all of their various, especially cosmetic surfaces at their ID offices in Cupertino. And what I've been told is at those, at those offices which are restricted to even like, you know, it's the inner core of the Apple brain trust. They've got copies of like every core process machine that they end up tuning to figure out how to make them work. So it's interesting that like you could think of manufacturing as reinforcement learning, but like Apple ends up doing a lot of that Core work and kind of building, getting those core weights themselves and then pushing it through a scaling partner. So it's really like, yeah, they're not exporting files, they're exporting engineers.

59:50

Google intended to basically reshore consumer electronics with Google Glass and Motorola. What attempts outside of Elon and his companies have been made to basically reshore consumer electronics.

1:02:19

Yeah, so I mean, I think we can also talk about Apple as well here where like Apple did the, the tr. The trash can Mac Pro was done in Texas. My understanding was like huge chunks of the dependencies to build that were all like the tooling was essentially built in China and then it was designed in China, assembled in California, which is kind of a, kind of a, kind of a joke statement. But like that's, that's actually what really happened with, with, with, with, with the Mac Pro. And it was in some sense a political project to make sure that like consumer electronics tariffs didn't get raised during the first Trump administration. Going back to Google. So like I was there across the street from Aaron working on Google Glass and they did. I think it's one of these cases of like Sergey was right about this AI stuff and if you go back to consumer electronics, he was right about that too. It was just early and the, the story here is like they had realized that a bunch of the IP was going to be in like precision optical alignment of the, of the optical elements. And like it was hard and kind of a first, like it would be like you had to do it for the first time anyway. So like if you were going to invent the process that made like the waveguide for the, for the optics yourself, like, why not just do it yourself now? And they were willing and at the time Google was flush with cash, willing to, you know, willing to spend the capex to do it. I think what you saw. And then likewise they bought Motorola because they're like, we need to get into the phone business with a vertically integrated phone partner, not just Android basically Android and Google Play services and all other stuff. So they realized that these pieces needed to exist. And I think we're directionally correct, but the problem was they didn't exist at the scale needed to kind of like again fully subsidize the supply chain. That was that, that, that like needed to also be brought along with it. So like you end up building Google Glass in San Jose and where the components coming from? Well, like I was flying to China, I was flying to. I was flying, I was, I was flying to, I was flying to Asia for a Lot of the component engineering work and if you kind of do the gradient descent of like, well what's the best way to build this thing? You end up back to manufacturing stuff in China basically. And you know, organizations change over the years. Different people come in. There's not like focus of vision of like this is the directionally right thing to do and we have to just pay the tax until it works out. Like obviously Elon is really good at that, but you can see like Google like went through management changes, you know, different senior directors in organizations, you know, think, think these things, these things are all natural evolutions of organizations over 10 years. But like keeping that coherency of like this initial assumption was correct and carrying it through 10 years later, you can actually objectively do this stuff that didn't happen in that organization. I do think that they'll probably make another shot at it though. And I think it's also worth saying like I think there's a similar story happened with AI but like you can see right now founders coming back into play. It's getting, you know, they've been locked in pretty hard with Gemini and all this other stuff and like I think they're doing a good job. So we'll see what happens. We'll see what happens. I'm also like be, I'd be, I'd be excited to see if that, something like that, you know, they, they try to run it back again on that side.

1:02:31

So there's just so many, so many layers to this, right? I mean philosophically if you go back to wealth of nations, basically the idea of the invisible hand versus a more Listian or Hamiltonian kind of worldview on your own sovereignty and just like if somebody else can make all the stuff that you want that's not good and ultimately there's some kind of push and pull that comes down to market efficiency. And I think you can probably argue that debasing the dollar to have capability is an actual real argument. And this is kind of like just look at Chinese industrial policy. It's cohesive and it doesn't break between a four year administration. Right. They have consistent five year plans that they've been executing on for you know, a very long time. And even since like I don't know, 2024, the kind of like priority items in their five year plan, they've dumped a trillion dollars from the state into it. Like we're just, we're not at that level, you know. And so when you like hear these analogies it's, you kind of want to Just like hit your head against the wall like, yes, of course they're like, they're obviously they're great at doing all these things, but like what are we doing? And this, this is ultimately just the kind of question. It's like everybody knows this is going to be a monumental undertaking. It's not something that happens overnight. You really do have to stack rank these things. Ultimately. The things I've Learned Dealing with OEMs at that scale now just through my company, are mind boggling because there's an economic decision to make and you can't break that chain very easily. Yeah, that's why I think this is just ultimately a matter of will. But more so a structural problem of can we have some kind of cohesive industrial policy that's durable and puts us in a position in the future that we want to be in.

1:05:37

Elon has mentioned that he basically believes that humanoids are going to be the biggest product ever created. How do you guys think the advent of humanoids is going to shift the balance of power and our ability to be the manufacturing powerhouse of the world?

1:07:55

So I'm going to put myself in Elon's brain for a minute here.

1:08:09

I don't want to be there.

1:08:13

I mean, if you just go back to the economic decision and kind of like the urgency level, it is the only place that you can do this right now. Right. We don't have the end to end supply chains to do that here now. So the worry that I have is that if I'm in Elon's brain and I think that getting v1.0 of a massive Optimae where it's like 100,000 of these things, does that actually set up this kind of training ground where they can start learning and then self perpetuate and start replicating? Basically whoever gets to that point first, it's game over. So do we want to hand all that over to China? Because they can do it? Because if we do, it's probably not a good decision.

1:08:16

Yeah. And I think Elon has specifically talked about basically the humanoid supply chain doesn't exist and so they're rebuilding it from scratch.

1:09:08

Yeah, it's, I mean it's a, it's a really challenging place to be in because if you, if you take all these language models and other foundational models that are being architected right now, they need better scope atoms in. This is me talking now out of Elon's brain. But like, yeah, the world of atoms is the last frontier of AI. And yeah, I mean like we've seen A lot of really cool robot companies with like folding laundry and stuff. But I do think that this is a really critical decision, if not even maybe more critical than drone production at mass scale too. So I, I don't think it's a good place to, to be right now.

1:09:15

And I, I would also argue that like, we are seeing with the data center, build out immense opportunity to like onshore, the electric stack, onshore, advanced chips, like all this other stuff. And we've excluded these products from the tariffs because there's a cross pressure of, well, if we tariffed it, we just build the data centers in Canada and Mexico and like, you know, or in space or whatever. And so we literally are sitting on an opportunity to like pull in a bunch of capabilities. And I think it's kind of like the, the battle between the invisible hand and like the great man theory is like the way to kind of think, way to kind of think about it where it's like, I'm. Maybe I, I'm. I'm less of a believer in the gray man theory more. And a belief that like you can just white knuckle and you should just do it.

1:10:00

Yeah, I mean like, if somebody came over here and gave each of us $10 billion to do something, it would

1:10:43

be, yeah, we'd just do it.

1:10:48

It would be amazing.

1:10:49

Yeah, we just do it.

1:10:50

Be like, oh, there's two more elons.

1:10:51

But the, but, but, but the invisible hand. Ye, exactly, the invisible hand versus just white knuckling it. You know, it's like, is, I think the, I think the, the situation here. And so, yeah, I would. And, and then the other piece is like, where are the capital markets right now? And it's like the incentives are to build a software company. The incentives are like. And so. And so. And it's weird because I have a strong theory that, you know, as someone who's written hundreds of thousands of lines of software over the past six months thanks to these AI models. It's not slop, I swear. Basically, like, we're now seeing the situation of like, where's the physical moats? It's like the fact that I've got a better temperature sensor on my stove than everyone else means that like, oh, I can, like, I've got the first layer of autonomy for service cooking salt. And then you can start layering on this other stuff. And it's like, am I really an energy storage battery company? I'm like, yeah, I guess that's true. But like, what are we actually doing here? And it's like, we're building the robot. Like people stop calling robots robots and they work. And I think that basically the US is actually squandering some interesting opportunities even ahead of the humanoid side of like, well, why aren't we building the data center racks here? Why aren't we doing all this other stuff? We've essentially decided that that's not interesting and it's not critical. And maybe the only thing that is interesting and critical is like low volume defense stuff. And thus we don't have these capabilities. And then it's like, and then who is getting the wealth of reinforcement learning knowledge from building stuff? And it's, you know, it's, it's, it's the folks actually building the humanoid actuators for people like Optimus and, and, and figure and all these other guys and they're not here.

1:10:52

So let's end on. Apple has spent hundreds of billions of dollars over the past, you know, 15, 20 years basically building out the entire supply chain in China. How do we as a nation become comparatively competitive against that backdrop?

1:12:29

So I will caveat that with. It wasn't really a forced error in some ways to do that. They did try to do some of it here.

1:12:43

They did it here in the 80s.

1:12:56

Yeah.

1:12:57

And they kind of had to do what they did, which is very unfortunate. And it's because of the systemic problems that we architected around making it easy for them to do what they wanted to do here. So again, I think a huge amount of this will ultimately come down to good industrial policy. And I think unfortunately you can't just unwind what they did. They did it. They have all of the process and tacit knowledge behind that. And the question is, you know, at this point do we want to answer back somehow and like what form is that going to take and how are we executing on it?

1:12:58

And I would say a big part of the story too is American, let's call it American policy minds. I think as you go west you're more mysterious and unknowable to dc. So if you go to California, they think we're a bunch of wackos and hippies. But if you go to, but if you go to, if you go to China, it's like unknowable. You think of like a totalitarian, like you know, monolith, even though it's like each individual city is run totally differently by different, like different, different people. And like it's a, it's a very, it's much more diverse place than people think. And so I think the, like not understanding how China works is a systemic problem both in D.C. and for Western technologists as well. Like, you could argue that like one of Elon's biggest assets is that he actually understands how China works. He understands how to manufacture stuff at scale in China. And like, I would also argue that a lot of people, and I don't want to call out, you know, some of our friends and stuff like that, but basically it's like people being like, well, I don't need to learn that because they're bad or something like that. It's like, no, China is an amazing place with a lot of great people and you can learn how to make, make stuff at scale there. And I would strongly encourage people to go visit and like, you know, learn, learn how this stuff works. Because it's like, I don't think it, I think people, it hasn't dawned on enough people how America is not the center of the known universe for technology in the physical world anymore. Now I think we still are in the digital world, but in the physical world that's not the case. And there is like core technologies and ways of doing things. And like the, the, the idea of like your engineers are next to your manufacturing line. Like a lot of this stuff is like, we kind of talk about it in the abstract and kind of like the, the Gundo sort of context and like, obviously there's well intentioned people trying to do this in the United States, but the scale of the operation and the like understanding of how stuff works, it's like, in some sense it's like, I just strongly encourage people who like, want to learn how to build advanced stuff to like go there and build advanced stuff there, absorb the tacit knowledge that is, you know, it's very hard to gain that tacit knowledge in the United States. You need to probably spend a billion dollars of, you know, money that is under your control to do it. But you could do it much cheaper if you went to Shenzhen and like lived at a factory and got a consumer product out the door. So I think that that's something that like some people have shut off their brains to that reality of like, of like, okay, I'm going to basically say like, you know, I'm not going to play ball with the Chinese system, you know, all this other stuff. And it's like, no, this is the way that advanced hardware is made today. You can choose to participate in that world or you can, or you can, you know, go Ted Kaczynski and live in a cabin in the woods. But you know, that's the situation. And so. And I think that people have kind of like, plugged their ears, blindfolded themselves to this reality where it's like, no, like, the most advanced manufacturers and manufacturers in the world are there. And if you want to build stuff fast, cheap and at scale, that's where. That's basically where you end up going. And if you wanted to fix that problem, it's the way you have to fix that problem is to understand how it works. How it works. Yeah, yeah, exactly.

1:13:48