Welcome to the fifth episode of the Palladium Podcast, where we explore the future of governance and society. This week, Wolf Tivy interviews Nick Pinkston, founder of Plethora, a rapid manufacturing company in San Francisco, about manufacturing and industrial policy.
For suggestions or questions, email firstname.lastname@example.org.
0:00:31 Wolf Tivy: Hello, and welcome to the Palladium Podcast. I’m your host, Wolf Tivy, Editor of Palladium Magazine. Jonah, our editor-in-chief, has been stuffed in a closet somewhere, so… And we lost Ash somewhere in the cold wastes of Canada. So I’m in charge now. Today we have a very special guest, Nick Pinkston, founder of Plethora, a rapid manufacturing company in San Fransisco. He’s on to answer our questions, talk about manufacturing industry and what these things look like in the future. So Nick, welcome to the show.
0:01:00 Nick Pinkston: Yeah, good to be on. I’m super excited to be on. I’ve been a fan of the magazine since I’ve had the paper version early on…
0:01:06 WT: Right, yes.
0:01:06 NP: With the binding, yeah. [chuckle]
0:01:07 WT: Yeah, yeah. That was at our first Palladium party, at the launch party, we had a very limited edition…
0:01:14 NP: Yes, yes. [chuckle]
0:01:14 WT: Paper version of the first articles, and that has, I guess, become something of a collector’s item. Yeah, so we always have a question of the week. This week, the question is, “If you could only use one social media platform forever, what would it be? But email doesn’t count; email’s too easy of an answer.
0:01:30 NP: Yeah, yeah.
0:01:32 WT: So for me, I think it would be Twitter for obvious reasons, or possibly Urbit. I’m pretty bullish on Urbit.
0:01:36 NP: Yeah, [chuckle] nice, nice.
0:01:37 WT: And Urbit could be the future of social media. And so that’s like the big hedge answer. That’s for me. What about you?
0:01:43 NP: Yeah, I know, nice. It’s funny, I feel like I aspirationally wanna say Twitter, but I just end up on Facebook all the time.
0:01:48 WT: Oh man.
0:01:49 NP: I have these private Facebook groups that are actually awesome, and one of the things about Twitter is you can’t say everything on Twitter, or you piss everyone off or something. So I really like these private Facebook groups. I’ve been trying to be like, shit, instead of having these screeds in Facebook that no one will see but 20 people, I’m trying to actually… Actually, how we found this topic was actually on Twitter, when I put my first big tweet storm out there and it got a good reception.
0:02:15 WT: Yeah, yeah, great tweet storm.
0:02:16 NP: Yeah. And so I think I’m trying to get more on Twitter actually, just because there’s so many more interesting people to engage with.
0:02:22 WT: Yeah. I have a Facebook, but I really don’t use it, and partially that’s deliberate. I just… [chuckle] Facebook doesn’t feel right. But Twitter… I like Twitter. Twitter is where a lot of important conversations happen. It’s really the public forum. That’s why I think it’s great.
0:02:38 NP: Totally. Totally.
0:02:39 WT: Anyways, let’s move on to more interesting topics, like Palladium stuff. So usually we talk about our recent articles, and we’ve been publishing some interesting stuff, notably the Octopus article, I think is a very important topic that we’ll have to dig into more depth at some time, but Nick is even more interesting than that, so we’ll focus on him. So let’s get started. Nick, I’ve got a few questions to get us started, and then we’ll take it where it leads us. So first of all, can you explain your company, Plethora?
0:03:12 NP: Sure. Yeah, essentially, the point of Plethora, our mission is to make it very easy to go from an idea in your brain to shipping product at scale. And so there’s two big pieces of this. So we’re making hardware products, anything from a self-driving car to a AR headset, all the different stuff in-between. So we’ve done… I can’t name the companies, but basically, the biggest tech companies on earth, all the way down to people in their garage making some new stuff. And so there’s two big pieces of this, and we draw a lot from software analogies. So in the same way that in software where you type code, you debug it, then you deploy it to a system, it compiles, and then it runs, we take all of those things and make factory and engineering versions of them.
0:03:51 WT: Yeah, I looked at some of your talk notes…
0:03:54 NP: Oh yeah, nice.
0:03:54 WT: And that was good. And definitely, I saw an emphasis on the software methodology, where it’s focused on type, iteration…
0:04:02 NP: Exactly.
0:04:03 WT: Debugging, just very quickly… Fast feedback, the ability to just get your parts quickly and also get information about how it’s gonna work even before you get it.
0:04:14 NP: Yeah, totally. So one of the things I look at is engineering metabolism. So if you think about CS, computer science, as fast as you can type and hit “enter”, the code evaluates and you know if it works. That cycle, one, you don’t get infinite of them, so in CS it’s like as much as you can think. That’s why we’re so into Agile and all these team methodologies, ’cause the team is actually the unit of work that’s the hardest. In manufacturing, it’s in a Gantt chart mentality. So you’re gonna design a bunch of stuff and that’s gonna take a while, but then the execution of manufacturing does too.
0:04:44 NP: So that design-build-test loop is a week or a month or a year, or whatever your complexity level is. And so what Plethora looked at is, “Okay, what are the major issues in that?” And so the overall goal is if I can make a thing faster than you can design it, then you’re now the bottleneck. And we want people to be the bottleneck ’cause then we can make better teams, we can augment their engineering…
0:05:03 WT: You get all you need from just running those parts.
0:05:04 NP: Exactly. So at first we just looked at, “How can we make a better compiler that will take a 3D model and then put it through a factory and turn it into a real thing?”
0:05:13 WT: Yeah, so I remember… So I used to work in engineering R&D. We were building fuel cells. So I have some familiarity with this process. I remember making out the Gantt charts, it’s like, “Alright, the next six weeks I’m focusing on this prototype that will test this concept that will go into the final product.”
0:05:27 NP: Sure.
0:05:28 WT: And yeah, the bottleneck really was… A lot of the time, you can get up a model relatively quickly, as fast as you can think, basically, in a CAD program, but then you have to produce engineering drawings, you have to send it out to the manufacturing guys, they take a while to get it back to you, it costs a whole bunch. And so that really drags out your iteration times. And so I was always… I spent a lot of my time thinking about, “Alright, exactly which tests do we have to do?”, and really having to plan that process out.
0:06:04 WT: Whereas in software, I’ve also designed some software stuff, it’s just like you say, you have an idea, you figure out how it’s gonna work, you type it in, you compile, you see if it works, you write some tests for it if you’re up to that, if you’re into that kind of thing, and it’s much faster and it’s much more focused on the thinking rather than having to point out this material process.
0:06:25 NP: Yeah, and I think that actually, it mirrors the creative process more. You actually are able to… ‘Cause when you’re in such a complex system, you actually can’t deterministically know everything. So in that case, iteration is the only way you can actually explore a complex system. So your iteration time is the unit of which your innovation can be.
0:06:43 WT: Yeah, and even if… Actually, this is something that I was thinking about in thinking about how to prepare for this. Even if in the long term something is gonna end up being simple and well-characterized and with mature technology such that you can just design it all upfront, at the time that it’s innovation, the physics is uncharacterized… Yeah, often you’re dealing with new phenomena, you’re dealing with new technologies and design so you don’t know how they behave, you don’t know what the optimal thing is, you haven’t reduced it to the simplest possible form. So you really are dealing with this complex system that you don’t fully understand, and so you have to do it empirically.
0:07:18 NP: And that actually… So this is, you’re talking about the complex system of the product space, but there’s also the the market space. So you’re trying to make a thing that succeeds in the market, and so it might physically work, but it might suck in people’s minds. So to me, iteration is always reacting to what market space actually looks like. And so we actually…
0:07:37 NP: And John Boyd, he talked about these cycles of the OODA loop. So the observe-orient-decide-act loop, and you wanna be inside the loop of your competition going faster. So the market loop, you have to be inside too, so you can react faster when the market is a good thing, but also your competitors. So if you wanna out-compete your competition by engineering, you have to figure out how to get your iterations faster and how you learn better in order to do that.
0:08:00 WT: Yeah. So you guys are basically shortening that OODA loop in hardware manufacturing.
0:08:06 NP: Exactly. And really, what we’ve done right now is 1% of the overall vision. So there are other companies that I’m going to start that are also working on this stuff. So this is my third company in the space and each one of them keeps getting more and more broad in its application. But I think that the main issue is that the engineering process has evolved with, basically, in service of the tools being slow in a certain way. It’s not actually on thought of, “How do we make a better engineering process to do this and then make tools for itself?”
0:08:37 NP: And one of the issues is, is the tools are so damn hard to make. All of CAD software was developed by military and huge corporations to serve their interests, which back in the day were giant machines that needed Gantt charts and we didn’t have the stuff to keep track of that. So CAD was made a certain way, and computing power wasn’t even enough to do any more than that.
0:08:56 WT: Yeah, this is something I’ve thought about in my various past lives and hobby projects. I’ve noticed in CAD software, you often have this notion of the sketch where you can just sketch up a bunch of geometry, you put constraints on it, just solves the constraints and makes it whatever shape it should be.
0:09:14 NP: Yeah, in a smaller way. Yeah.
0:09:15 WT: Sometimes it doesn’t quite do it, but for the most part, it just works. And thinking about that process abstracted out to a much larger set of engineering domain. So rather than just geometry, what if you could do that with equations and systems and all this? And that was something that really fascinated me for a long time, was the fact that that would really accelerate the engineering design process at the design calculations level, and there wasn’t really good software to do it.
0:09:48 NP: Any software.
0:09:49 WT: Yeah. And so that’s something I imagine could be massively improved. But like you said in your big tweet storm, this CAD software is actually very high investment to get a bunch of people to actually figure out the algorithms that could sort that out and write it out and build a good user interface for it, and all that stuff.
0:10:08 NP: There’s another interesting thing with the CAD software, is the CAD software is used by people who cannot make the CAD software. So there’s actually an open source kernel that was basically a company that went bankrupt and became Open Cascade, and that was never able to be made into real CAD. And my theory here is that unlike, say, Apache or one of the classic code bases that was open source, people who use Apache can improve Apache. People who use CAD software can’t write CAD software.
0:10:34 WT: Yeah, they’re not gonna write patches for it. They’re not gonna be like, “Oh, actually, I need this feature so I’ll just that put that out.” [chuckle]
0:10:38 NP: Exactly. So even having open source community wanting to do this, and everyone wants this, it just isn’t going to happen. You need a large concentration of people who have PhDs in Mathematics all coordinated very closely to make good CAD software. And that’s why there’s only two kernels that exist now, basically.
0:10:55 WT: Right. So it’s inherently the Cathedral model rather than the Bazaar.
0:11:00 NP: Yeah, [chuckle] yeah, totally, totally.
0:11:01 WT: Alright. Yeah, so that’s… Okay, so you’re looking at basically… You’re talking about this much larger vision where you’re starting to look at problems like that, but you’re starting with this quickening the iteration loops?
0:11:16 NP: Yeah, totally. And how do we… So I had a company previous to this where I was actually making software that would just make the quoting parties here, and then we found the vendors who were the factories to make stuff, but the issue was when we looked at the time savings, yeah, the quoting process is on the way and finding good vendors is long, but once you get vendor relationship, it’s okay. But the vendor is still the big problem and it’s because that manufacturing, even though it’s like “automated”… Like, CNC machining technology is what we use and that’s already very widespread.
0:11:44 NP: And for people listening who don’t know about that, it’s basically, you take a block of metal or plastic and you use the thing that looks like drill bit to cut away the stuff you don’t want and what’s left is the part. So that’s been around since the ’50s.
0:11:56 WT: And the tool path can be generated automatically by the software.
0:12:00 NP: Not really, not really. So in limited cases, that’s true, but actually, the hard part is actually the 3D model in CAD. It will take you a day to get the tool paths, the code that the machine uses to actually work. And it’s even beyond the tool path. It’s actually how you hold it and all these other different things that what a machinist does. So our compiler system, that’s its whole point is how do you look at a part, characterize its geometry, figure out what machines should be in there, what cutting tools should be in there, what work holding. It’s the whole thing.
0:12:27 WT: So you guys are figuring out how to automate that process?
0:12:29 NP: Yeah.
0:12:29 WT: I didn’t know a lot of visibility into that when I was…
0:12:31 NP: Sure, sure.
0:12:32 WT: So it’s just like, yeah, we send off the STEP file and the drawings to the guys at the CAD shop and they figured out somehow…
0:12:37 NP: And that’s what it should be. The only thing you weren’t into is if you just throw a 3D file there… Often, people don’t know the limitations of either the process or the particular shop that you have. Maybe they don’t have a machine big enough or a drill long enough.
0:12:49 WT: Yeah, and that was something, I remember I would do a lot of thinking about, “Okay, how is this actually going to be done?” When I was designing the part, it’s like, “Alright, make sure all the corners are round so that you don’t have stress concentrations, you can actually get a tool in there, etcetera, etcetera.”
0:13:04 NP: Yeah.
0:13:06 WT: But some people would not do that right or they don’t necessarily know that intuitively and it doesn’t work that way.
0:13:12 NP: Totally. And then the thing is, is how would you even know what the tooling they had was? So you get into weird things where the thing holding the drill bit is crashing into the part before the drill fully makes the hole. So you actually need to know this. So what we did is we made a system before the compiler which is sort of analogous to a debugger, which knows every single limitation of our thing. So, every drill bit is in there, every machine limitation is in there, and all the limitations of physics are in there.
0:13:37 NP: So when you’re designing, we’ll say like, “Hey, that hole is not accessible for this cutting tool, so yada, yada, here’s how to fix it.” So that’s where the two-part thing we do is, in real-time, you can debug your component as you’re making it, and then you can say, “Bye”, and then the compiler spins up and makes one. And that’s really the whole core of it, is being able to go through the whole process quickly.
0:13:56 WT: And so your system works by embedding into the CAD software with your piece of software so you can watch what they’re doing and give feedback in real-time.
0:14:05 NP: Exactly, exactly, yeah. So they’re just getting analyzed, in a second it gives you results, and then you can adjust.
0:14:10 WT: Yeah, cool. And so this, I guess, leads us to once you have this process where the process that the engineer is interfacing with is no longer them having to think about the individual machines, then you guys get to think about the individual machines and design better machines. Is that the idea?
0:14:29 NP: Yeah, I mean ultimately… It’s funny, this is just one process of CNC manufacturing. So if you look at… There’s a great book called, what is it, Manufacturing Processes for Design Professionals, beautiful book, has like a hundred different manufacturing processes. Everyone of those will then be automated. And maybe 10 to 20 are how most of the stuff you would do and at least prototyping might be made. So it’s a huge thing. I’m actually an advisor of a company called Path Robotics, which is sort of like the Plethora of welding.
0:14:57 WT: Okay, cool.
0:14:57 NP: And so they’re working on analogous things in the welding industry, which I also find very exciting. So there’s just, to me, there’s a fundamental limit of not many people know about manufacturing and start startups. So there’s actually a heavy founder limitation for this. And ’cause I’m from Pittsburgh, Pennsylvania and my grandfather was a machinist and my dad was a production manager and I just grew up around factories, I knew those problems, but no one else cares.
0:15:19 WT: Yeah, yeah. No, this is something… When I was thinking about products in the space, one of the things I realized is that there are very few people who are simultaneously familiar with the atom side and the bit side.
0:15:31 NP: Yeah.
0:15:31 WT: And these things really do need to come together to get progress in this field.
0:15:36 NP: Yeah, totally.
0:15:37 WT: You mentioned the problem of machine tools are hard to make, and this is what your recent tweet storm was about. Basically, Samo asked this question publicly ’cause he’s been looking into the industrial stuff. He, Samo Burja, we can provide a link, he asked a question on Twitter, “What’s going on with the concentration of machine tools? Why are they only built in Germany and Japan? And what happened to the American industry? What happened to the British industry, etcetera?” And then so you went in with a bunch of history on the topic, and discussion of the big capital investments that are required to do that kind of stuff, and where it all went, what all happened.
0:16:24 NP: Sure.
0:16:24 WT: So if you could go over…
0:16:26 NP: Yeah, sure.
0:16:26 WT: Your thesis in that area, that would be great.
0:16:28 NP: Yeah, to kind of run through it, it was sort of looking at, so what even is a machine tool? It’s interesting, there’s this thing called the product space, that I really love, and it actually looks back at, what are the core industries to make everything? And so most things that you encounter with are made using some molded processes or something, and molds are cut using machine tools, which is what we’re talking about.
0:16:50 WT: Yeah. So machine tools here are like lathes and CNC metal…
0:16:53 NP: Technically, it’s anything that’s a machine that does the work. So machining is the thing that are lathes and mills that we had to invent, and basically, it allows us to make precision parts, which before, there were only blacksmiths and you either pounded shit with a hammer or you poured molten metal into a mold, and then after that you would file it. And so you used to… For old guns, for instance, some of the most mechanical things, you would hand-fit things with a file, and you would basically, you’d file a little bit, check it, file it, check it. And so each one was made uniquely.
0:17:23 WT: Yeah, extremely expensive.
0:17:23 NP: Yeah, extremely expensive and you couldn’t have repeatability, so you couldn’t make interchangeable parts. So this whole thing we heard of interchangeable parts, that was, you had to have precision and consistency to do that. So machine tools, those are repeatable.
0:17:36 WT: That was a big thing in the 19th century, basically, like the…
0:17:38 NP: And before, yeah.
0:17:39 WT: The replaceable parts was this Holy Grail in manufacturing. I don’t know the history very well in here, but there was sort of the American system of manufacturing, which I think was related to that. But it was like this big effort. It wasn’t…
0:17:53 NP: What’s interesting…
0:17:53 WT: It wasn’t just something that happened accidentally, basically.
0:17:56 NP: Yeah, it was certainly deliberate, and a lot of people tried. And actually, the American system, and I forget the guy who was the one who invented this, actually was a fraud and the rifles never did have interchangeable parts from his system.
0:18:07 WT: Interesting. [chuckle]
0:18:07 NP: Yeah, so the government ordered a bunch of muskets or something, and he made 10 that you could do this with, and it was extremely painstaking and manual. It was not using this process.
0:18:15 WT: Right, ’cause he was manually fitting them across each other and so on.
0:18:18 NP: Exactly, yeah, yeah. So that’s what happened. But, yeah, over time… A big thesis of this is that a lot of government investment vis-a vis military usually investment in the American case was put into this to make interchangeable parts for war in various implements in that way, and then it leaked down to everything else.
0:18:35 WT: As is often the case. Basically, some things need really high capital investment. The biggest entities around are states, so they’re gonna be able to make the biggest investments. States in terms of their engineering needs, the largest engineering need of a state is basically war. So yeah, a lot of stuff comes from war.
0:18:53 NP: Yeah. The first production system that made anything like interchangeable parts actually, I believe was in Florence, maybe Venice, and it was one of the arms production systems of maybe the 1400s was kind of mass producing muskets in a certain consistent way. And then there were block and tackle systems in the 1600s in England where there were purpose-built machine tools just to cut the pulley assemblies for big ships. And then we got into muskets and stuff in the Americas. But we gotta fast-forward it through the creation of machine tools and precision, which was like how you take it by hand and use a file to… There’s all these ways of making. But once you have accurate machine tools, the issue was you needed very good machinists to do it. So you could even say that the public education system focuses on calculus at the end ’cause you’re supposed to make steam engines and machine tools. And you knew that… You need to know that math so you can make the mechanisms to do that.
0:19:42 WT: Right, yeah, yeah.
0:19:43 NP: And so we needed people who could run those machines. And so that’s why high school kind of did that.
0:19:47 WT: Yeah. Of course, these things have kind of drifted apart since then.
0:19:49 NP: Yeah, of course…
0:19:50 WT: And no longer really…
0:19:52 NP: Yeah. Calculus is almost like a vestigial math that we learn that is like maybe not even more applicable for most people’s lives, than it was previously. But then afterwards, I think World War II was something where you can look at like really interesting things in capital investment. There was a kind of conversation on Twitter of, what was the effect of so much money being thrown at machine tools in World War II and yeah, actually, I believe it’s in, The Rise and Fall of American Growth, that book, has this track that shows the modernization, like how old, how the machine tools over time, and in World War II, you see the youth of the machine tools go way up. So, far fewer and fewer… More and more machine tools are new, basically. Because the war, and then a higher productivity and so some people have linked that to the higher post war productivity.
0:20:38 WT: Oh ’cause all of the stuff got replaced, as part of a big capital investment.
0:20:43 NP: Exactly. Yeah, so it just bought more new stuff, but right after that, the Air Force is trying to beat the Russians in the airplane game and so Operation Paperclip took a bunch of engineers and shit from the Germans, and so one of those things…
0:20:54 WT: Yeah, that’s how we got to the moon.
0:20:56 NP: Yeah, yeah, von Braun, famously. One of the guys, I don’t know who it was, but there were people who built these giant heavy presses in the Heavy Press Program.
0:21:03 WT: Yeah, you mentioned that in the tweet.
0:21:04 NP: Yeah and so they have a bunch that are still in Cleveland, from the 50s or something, they were made… They basically, every airplane commercially you flown in, has parts from these things in Cleveland. They make giant aluminum forgings, basically.
0:21:16 WT: And that’s for like wing spars or something?
0:21:18 NP: Yeah exactly, exactly. It’s like the big parts that are made all at once, and they cost like a million bucks each. And Boeing got really good at working on them once they made them. Yeah, it’s crazy, but that was direct government knowledge that had to do that. And then to make the dyes for those things, it was then very complicated geometry. And so the Air Force asked MIT, and then there was this kind of guy who went around famously, like trying to get up support for a computer-controlled machine tools, and so MIT and this guy went together and made the first computer-controlled machine tool, which the program…
0:21:47 WT: The CNC machine.
0:21:48 NP: Exactly, or what they called “NC machine” at the time, because it was just a numerical control. And it’s kind of funny, there’s a good book called Forces of Production, which talks about it, it’s called, A Social History of Industrial Automation, and it talks about the race between numerical control, which is programming, which he’s kind of a Marxist, and he talks about like, “Oh, well the engineers would program and then operators would execute those programs on the machine tools, not knowing how to do their own code.” Which he paints as a de-humanizing thing. And there was another approach people took, called the playback system, where you have a manual machine tool, and a professional union machinist, would make the part once, and then you could put on repeat mode. Interestingly, any error that guy would have put into the part would be maintained through every piece. So, it wasn’t as repeatable.
0:22:35 WT: Yeah, I remember that reminds me of Kurt Vonnegut’s novel, Player Piano.
0:22:40 NP: Okay, I haven’t read that.
0:22:40 WT: It’s pretty obscure but it’s…
0:22:43 NP: People tell me to read it all the time. Actually…
0:22:44 WT: It’s about the automation of manufacturing and all its social effects and everyone gets unemployed and blah blah blah.
0:22:52 NP: Okay.
0:22:52 WT: I don’t think the sociology was that great for that one but it had that system where things were played back by tape…
0:23:00 NP: Oh, interesting.
0:23:01 WT: To run the machines.
0:23:02 NP: And the NC stuff was on tape, as well. It just was, who is generating that tape? Was it an engineer or was it a guy in…
0:23:11 WT: Yeah. It was a recorded tape…
0:23:12 NP: And weirdly, these ideas came together in what’s called CAM software now. So like, I believe the Air Force again, made a system where instead of programming individual lines of code, that looks like just regular programming language, you could just click on stuff in 3D to program it, and that’s known as CAM software.
0:23:29 WT: Yeah, and that’s how we generate the CNC G-code, I guess you call it.
0:23:34 NP: Yes, yeah, people used to just program in G-code directly.
0:23:37 WT: Yeah, that’s crazy.
0:23:38 NP: Yeah, I know, I know. You what’s crazy? Even today, 70% of G-code is generated by people who are just right next to machine.
0:23:44 WT: Really?
0:23:45 NP: The reason for that is most parts are actually just cleaning up castings. So if you think about an engine block, it comes out of a foundry as a rough near net shape, as they would call it, piece, and then you have to bore the cylinders and tap the holes and so you really just have to know what the bore to bore dimensions are, and you just had to program that into the machine and people kind of guess and check the way to it working a lot of the times.
0:24:08 WT: That’s kind of crazy. That’s really neat.
0:24:09 NP: Yeah.
0:24:11 WT: Okay, so then that’s sort of the big capital investments in the 50s. And then in the 70s, and 80s, we had a lot of off-shoring and Japan and Germany ended up with these industries.
0:24:20 NP: Yeah. And so it’s interesting. One of the things that we found in that tweet thread is the like the… I believe it was in the 60s or 70s, the top 10 automated machine tools, CNC machine tool companies were American. So, the entire industry of this era, and we had invented it, and then by the ’80s, FANUC, a single company in Japan had 50% market share in machine tools globally.
0:24:39 WT: Yeah, that’s crazy.
0:24:41 NP: So what happened, yeah, yeah, it’s crazy. And FANUC, actually, they make basically, the controller or the brain for 70% of the machine tools today.
0:24:49 WT: Wow.
0:24:49 NP: And they don’t make the tool, necessarily, but they make the hard part, the thing that converts the computer stuff into motor movements and they also make the motors and the linear motion systems, they’re the hard bits.
0:25:00 WT: That’s the tricky part, ’cause there’s this backlash in the system, right? You can’t just get the motor to take a certain amount of steps and have an easy translation between that and actual motions.
0:25:08 NP: There’s so much.
0:25:09 WT: Because there’s swapping the ball bearings, there’s swapping the gears, there’s all kinds of stuff.
0:25:14 NP: Sure. And it’s crazy, so the people who actually make the machine tools, just buy everything from FANUC, basically, or people like FANUC. And so, they’re basically mechanical guys or just machinists, and they are buying off the shelf components and they know how to make metal. And even that is crazy, because the metal actually is… Everything is actually jello in the world. Nothing is hard. You might think cast iron is hard but actually, it’s drooping all the time. Like at Plethora, we’re on the water, and there are actually times when the tide comes in or something, and the machines go out, because the concrete is moving slightly and the machines move slightly, and that means they go slightly out of square, this kind of stuff.
0:25:51 WT: Oh man.
0:25:51 NP: It’s actually crazy how much you have to do, so there’s all this computational stuff you do to compensate for that.
0:25:58 WT: Right, oh it’s crazy.
0:26:00 NP: And so one of the questions was, “Why did America stop investing in machine tool companies?” And I don’t know if I’ve ever seen an actual study that says, but my anecdotal thing here is like, one, it was cool in the ’50s and ’60s to do manufacturing, and it was a big competitive advantage in the country. We had an industrial policy, such that we invested a lot upfront. These initial companies were actually profitable because there was a lot of innovation to do, and so people just hadn’t commoditized it yet. By the time it got to the ’70s and ’80s, everyone else like Japan was like, “We really want some of this stuff.” And so they made their own industrial policies that made these companies, and then supported them with low interest loans and all this stuff to make them that way. I think it was Fujitsu in the story that actually was the one who made FANUC, actually, internally.
0:26:43 WT: Made FANUC, yeah.
0:26:44 NP: So that’s was probably because Fujitsu is getting a bunch of money from the government too, ’cause it’s one of the like, [0:26:49] ____ or whatever, that are, you know, one of the big companies in Japan.
0:26:51 WT: Yeah. And so I looked at these economic transitions over the latter half of the 20th century before, and of the big things that strikes me is how much of an economic transition there was in about 1973.
0:27:02 NP: Well, that’s the classic productivity wage thing on the chat.
0:27:05 WT: Yeah, so that’s one of the things that happens there, is productivity continues to go up, but the wages don’t after ’73. You get the oil shock, you get the gold standard, you know, you get Watergate. There’s actually on many graphs, like about half of graphs you might expect to be correlated with the health of society, or the health of the economy.
0:27:25 NP: Totally. Yeah.
0:27:25 WT: Or whatever. There’s something going on and I think in ’72. I’ve always wondered what happened there. And so, one of the things that, anecdotally, looking through this stuff you get things like Fun Brown retires around then. Kelly, from… What’s his first name?
0:27:43 NP: Which Kelly?
0:27:44 WT: The skunkworks guy.
0:27:45 NP: Oh, yeah, I forget his name, but yeah okay, interesting.
0:27:48 WT: He retires around then. Mueller from NASA also retires around then. And basically, what you have is the entire generation that was educated before the war, and was active through the war, retires in the early ’70s. And then you simultaneously have a bunch of political upheaval, changes in how the economy is run, etcetera. Something changed there, basically, and it changed how we do the industrial policy, basically.
0:28:17 NP: Yeah, yeah, it’s interesting. And also, I wonder what the bureaucracies were doing at that time. It feels like post World War II, we had really good institutions and people were ambitious and you could like… There’s a great book called, “American Genesis” which actually is the industrial history from 1870-1970, and I think it’s funny that it’s bounded by 1970. And you know, throughout that book, there’s kinda two big periods of say, FDR time, and maybe like 18… I don’t know, not to say FDR, but this whole thing of when industrial…
0:28:45 NP: The second industrial revolution, when trains and railroads and everything happened, and there were people who were bureaucratic entrepreneurs who were building out pieces of the federal bureaucracy, when you could actually have like, “Here’s a person who hired people and do it.” And it was almost like a startup, where we just can’t do that anymore.
0:29:00 WT: Yeah, so that’s the FDR regime and then after the war as well. They’re building up the CIA, they’re building up all these…
0:29:04 NP: Yeah, NASA.
0:29:05 WT: Yeah, NASA. They’re building all these big parts of the government, and it’s like… There’s a lot of founding going on, founding of institutions.
0:29:15 NP: Yes. Totally.
0:29:16 WT: But then, a lot of the guys who did that retired in the early ’70s. And I guess they failed the succession problem.
0:29:22 NP: And ossification might be the natural order of things.
0:29:24 WT: Yeah, or whatever it was, the energy ran out and now we’re kind of cruising.
0:29:29 NP: Yeah, also what’s interesting, I’ve been thinking about this a lot, is when you have such a run. Like I think right now, China’s in a run, where it’s like there’s the development curve and it’s an S-curve, and they’re in the steepest part of the S-curve, or just got through the steepest part. And at that time, we just did to, it was from 1870-1970, that was our steepest part of the S-curve. And it’s easy to be an optimist when your material conditions are changing all the time, for the better. You’re like, “Yeah, of course we’ll make future investments. We just did that for 50 years and it was all great.” And I think at a certain point… Remember it was the 1964 World’s Fair or something?
0:30:03 WT: Yeah.
0:30:03 NP: Was like, you know, GM had the flying cars exhibit and everything. People were very optimistic and thought the future was definitely gonna be moon bases or something.
0:30:11 WT: Yeah, if you projected all this on an exponential curve, it’s like, “Yeah, of course, you have the moon bases and flying cars in 2000s.” Right?
0:30:17 NP: And I don’t think there’s any material reason why we couldn’t have done that, actually. Yeah, maybe… Maybe wouldn’t have been quite as fast but I… We ran out of optimism, it seemed like.
0:30:26 WT: No one has figured out how to make a flying car work, just like energy wise, but…
0:30:30 NP: Well, I mean, it depends. We built several flying cars, sort of from people at Plethora at various states of working. It’s interesting, some of them exist. Yeah, they don’t have a lot of range yet.
0:30:41 WT: Even if you’re running on gas, it’s a lot of gas.
0:30:44 NP: Yeah totally, totally.
0:30:46 WT: And I guess… But I guess in the ’50s and ’60s, they were also doing things like nuclear reactors in your car, stuff like that, so you know.
0:30:54 NP: Yeah, RTG would just fly forever.
0:30:56 WT: Yeah, with the energy densities on that one… It’s not quite good enough.
0:31:00 NP: Yes, yes. I’m for Betavoltaics and Tritium batteries, but…
0:31:04 WT: Yeah. Yeah.
0:31:08 WT: Yeah. I guess the trouble with Tritium is that you can’t turn it off. Anyways, we are getting too deep… Yeah, so we lost all the stuff in the ’70s and ’80s somehow, through just…
0:31:23 NP: And there’s another piece here actually.
0:31:24 WT: Okay, go on.
0:31:25 NP: So I think that not only did like… There was this thing that came out in the ’50s and ’60s, where it was still innovative, like machine tools were like the web or something. And so, there’s another thing which is, I think, is that the capitalist system after a while, once it’s commoditized and your profits go to nothing, the government wasn’t propping up the actual end-state companies; ’cause that’s not like what America, at least maybe typically, wants to do, is support a bunch of companies.
0:31:50 NP: Whereas, actually in Japan, they’re cool with that. Or say, France with Citroen Peugeot, they may never actually exist without government help, because the unions vote the money because of so many jobs or something. I assume, I don’t have the deep history regarding that, but I kind of feel like that’s what those companies are actually doing. Sort of like when Thatcher privatized stuff, and then all those British car companies, like British Leyland and Enfield, they all… They all died, because those companies were worthless.
0:32:15 NP: And it was really good, they got churned out. So the question is like, did we actually lose anything by American machine tools going out? And I think that it’s interesting to think about what we did lose, ’cause in a way, we did lose a lot of people that knew all that stuff.
0:32:28 WT: Yeah, we lost a lot of capital in capability, that might be useful at some point.
0:32:32 NP: But it’s interesting though, is that we today can buy… There are 50 vendors of the same type of machine tool. I go to the show called IMTS in Chicago every other year, and it’s basically a bunch of clones of everything. Everyone has the compact car, everyone has the SUV version of these machines.
0:32:47 WT: Yeah, yeah yeah, the cars all look the same, and seen people commenting on this, we’ve optimized cars closer and closer to the fundamental limits. It’s like they are all egg-shaped, they are all slightly different variations of the same idea.
0:32:57 NP: And imagine, with FANUC… And it was actually, it’s FANUC, Siemens, there’s only a few different people that make all different electronics and motors to these things, so of course they’d all be the same, if all the fundamental components are all from the same people. You are building the same car with the same engine, basically, every time. In many ways, maybe we didn’t actually lose that much from those businesses that died, but you could say that…
0:33:17 WT: Yeah, economically, no. I don’t think…
0:33:20 NP: Yeah, so the knowledge of it is like, okay, this is one type of the machine tool, we had to move on to other types of machine tools, and the one we did move into, interestingly, was semi-conductors. We still have Lam Research, Applied Materials. There’s a bunch of these companies that are the leaders, top five companies, I think four of the five are US companies, that make the machine tools to make chips. I think that’s really interesting, although what we’ve increasingly lost is that people like Taiwan Semiconductor, are the people who do the actual manufacturing of the chips. We’re still making the fancy equipment, but we’re losing it, the actual production to Taiwan Semi, which is interesting, so the people who are really good at setting up the chip fabs are maybe over there now.
0:34:00 WT: Yeah, and this is something where the classic economic arguments for free trade, it’s like, “Oh well, let whoever wants to do it, do it. Let whoever is good at it, do it.” And that’s more prosperous for everyone, but once you start to get into the geo-politics and the competition between spheres of influence, that’s when it matters where the stuff is.
0:34:18 NP: Yeah. And then China has been really, really interesting in how they’ve done very aggressive policy. I don’t know if you saw the sort of prisoner’s dilemma, that was not passed by the high-speed trains of China. So basically, it was like, I think it’s someone in France, I forget who, but one of the people in France, it was Bombardier of Canada, it was Siemens, and it was whoever from Japan who does that. There were like five companies, or four companies or something, that have all high speed trains. And so, China basically played them all against each other and they’re like, “Okay, we’re gonna give the contract to whoever basically moves the plant over here and does slowly import substitution of Chinese parts.” And then eventually, they just figured out how to make the trains and just made them in China.
0:34:57 NP: And so, there was no coordination among, say, like state departments in each of those respective countries to say, “Hey, we have an ability to coordinate here, and not give up this technology to what would be the biggest high speed rail user in the world.” And there could have been a lot of transfer of wealth from our technical know-how, and they just took it. Which is really smart, it totally makes sense.
0:35:16 WT: Yeah yeah, they’re definitely… The Chinese are very shrewd these days. They’re definitely playing the game well, and I think, until quite recently, no one was worried about that.
0:35:27 NP: It’s insane. It’s really… Because I remember asking people this, like, 10 years ago, “It seems like these graphs in China are gonna yield something really crazy.” And I was just thinking like, “I must be missing something. I’m not some scholar of this.”
0:35:43 WT: Yeah, but that’s the thing, is no one had the confidence in it. And you couldn’t decisively demonstrate it, right? But then 2018, last few years, basically, people have started to realize, “Okay, they’re actually, they’re way ahead of us on a bunch of dimensions,” at least in the projection. Right?
0:36:00 NP: And of trade, in how they do it, certainly.
0:36:02 WT: Yeah. And in government capacity and all these things, and they are playing the game that we are not even playing. And so that started to get people worried, so the alarms were starting to sound. Some people were paying attention maybe a bit more to that.
0:36:14 NP: Yeah.
0:36:14 WT: But to respond to that, requires… To get the state departments of the various countries to cooperate and hold on to technology, that in itself is an institutional task, that we may not be up to yet, and it may take a while to spin.
0:36:30 NP: Yeah. I think we were up to it. This is the thing, is we’ve actually regressed from the 60s, when the US was the clear western dominance.
0:36:36 WT: But if you’re not using a capability, it kind of atrophies away, it goes way.
0:36:39 NP: Fully. We’ve also voted it away, in some ways. There were actually… I remember looking at this of like, where does the long term planning of the US government actually sit? If you look at the only one that’s left is the office of Net Assessment in the Pentagon, which has a 40-year time horizon. The Pentagon still does have longterm ability, but there was one in the state department, there was one in commerce, those were both defunded.
0:37:01 WT: Interesting.
0:37:01 NP: And actually, I believe… I don’t know if this is true, but I’d heard something about Newt, back in 90s or whenever, voted this out and he had his own private organizations that did think tank stuff like this, ’cause he’s bit of nerd, and he wanted the money too. That’s a cynical way of reading it. And then the congressional offices that used to do research was apparently defunded too. So it’s interesting, is like, when I talk to my he’s like deep into this and the government, I was like, “Hey, where does this live?” And he was like, “Oh, it doesn’t. There’s a four year cycle of Presidents, you have fun.”
0:37:32 NP: And so, when you look at that, okay, it doesn’t seem like we have this long-term ability anymore, outside the military. And it’s interesting cause the military are the people, weirdly, who are the almost still the best bureaucracy, in a sense.
0:37:43 WT: Yeah, you do see them still thinking long-term, like, “Where is this all going?”
0:37:47 NP: Yeah. DARPA. The only effective VC or whatever in the government is DARPA, who actually has a really interesting model, where the partners, in effect, the program managers, can only stay between three to five years max. And so you cannot get ossification in that system ’cause everyone’s churning out so much. That was designed explicitly to do that.
0:38:03 WT: How do they keep institutional knowledge then, in the thing?
0:38:07 NP: Yeah, it’s interesting, so I don’t know all the ins and outs of DARPA. I’m not even sure they’re public. I’ve been hosting some events with them, actually, to try to better connect to… It’s crazy there’s so many people in Silicon Valley who are trans-humanists. And so, I had my friend, Justin, who’s the head of the bio-technology office at DARPA. He was out here and I so I hosted the thing at this bio-tech incubator, IndieBio in San Francisco, and he gave a half hour talk that was like brain computer interfaces, flying drones through your mind, vaccines that make anything… Can kill any disease. These things they are working on, where it’s like, “This is the trans-humanist shit. If you wanna fund that shit, the military is already doing that.”
0:38:43 WT: Yeah, yeah. This is something like a few years ago, it’s… People are interested in all these futurism stuff. I was in some futurism crowds.
0:38:51 NP: Yeah, sure.
0:38:52 WT: And the only sort of official anything that we could find, looking at some of these big problems, was the military. It’s like, “Okay, the military years ago had presentations on where’s AI gonna take the military? Where is this gonna… How is this gonna affect what we’re doing?”
0:39:08 NP: Sure.
0:39:08 WT: Yeah. And so it’s like that’s cool to see, but again, it’s just this one part of government, then the rest of it should be thinking this big way too.
0:39:17 NP: Surely.
0:39:17 WT: And this is… I mean, partially, this is kind of what we’re trying to do with Palladium. It’s like, “Alright, the discourse has gotten focused on the elections and short-term stuff and in fighting, we’re not taking the big picture long-term like where’s it all going kind of stuff.” And so what we’re trying to do with Palladium is like, “Alright, what does the ruling class have to know about? How do we look at the world?” Let’s build up at least a small culture here of people actually looking at these big questions and trying to think over that long term.
0:39:47 NP: Sure.
0:39:47 WT: And ideally, we can help spark some of these changes that could create these longer-term institutions.
0:39:52 NP: Sure.
0:39:54 WT: Yeah, so this is all really interesting to us. Where’s it all going? What does it all mean? How do you do this stuff?
0:40:00 NP: Totally. Yeah, it’s interesting, when I think about the interaction between the short term and the long term in this way. I think one of the issues is that the people who are supposed to do, all the think tanks, your Brookings and your Aspen, and these people that are thinking about these things in the long term. In a sense, like my friend in the government was like, “No, that’s a think tank job, is to do long-term shit. We’re here to just pass laws that they write for us,” basically.
0:40:24 NP: And so it’s like, the issue is that we don’t actually have good coordination on either side, for long term stuff. It’s become very short-termist. And it does feel like maybe the Republicans are better at it right now. The last big push for this was actually probably the installation of neo-liberalism from Mont Pelerin. Mont Pelerin is… Are you familiar with Mont Pelerin? So it’s…
0:40:43 WT: Can you go into that?
0:40:44 NP: Yeah, for people who haven’t realized, basically, a lot of philosophers, like your Hayek and Friedman and these guys, kinda got together in the, I think it’s the ’50s, and even still happens today, but it’s not the… The arch dudes of that era. And they basically put together what we would call neo-liberalism now. And then there were people, I think it was the du Pont brothers were the people who bankrolled the whole thing and they were reacting to FDR, and this whole thing. And so they basically had a very deliberate plan of like, “Okay, we need to make the Cato Institute and the Federalist Society, and all these think tanks.”
0:41:17 WT: And that got in with Reagan and Thatcher externally?
0:41:18 NP: Ultimately, yeah. ‘Cause remember there’s that Milton Friedman quote that says, “You… ” What is it? “You solve the problems of today with the ideas lying around.” And you’re basically always waiting for the crisis and then you pick up the ideas lying around. So our idea is to have… Or our method is to have good ideas lying around. And if you look, there’s a really good book called, Invisible Hands, which tracks this thing from maybe the ’30s when FDR was the trigger for a lot of this backlash, all the way to the ’80s when it was fully implemented.
0:41:46 NP: And so you can see a lot of neoliberal policies did clear some of that and it seemingly the bureaucracy crystallized in a bad way and Reagan was like, “Big government’s bad, blow that shit away, and then let the market do it.” Maybe that was an over-correction in some ways, but it did blow away some sclerosis perhaps, that was a useful…
0:42:02 WT: Yeah, yeah, it was… Yeah, it clears out some of the brush, even if it…
0:42:08 NP: Exactly, yeah.
0:42:09 WT: Even if it didn’t fully revitalize the system.
0:42:11 NP: Yeah, so now we have to react and revitalize. To me it was just…
0:42:14 WT: Yeah, revitalization is really an important thing right now.
0:42:16 NP: Yeah.
0:42:16 WT: And both parties, the big parties, the Democrats and Republicans are both kind of in disarray right now.
0:42:22 NP: Fully, yeah.
0:42:23 WT: You got Trump. Trump, just completely blew up the Republican Party.
0:42:25 NP: It’s amazing, yeah.
0:42:26 WT: And the Democrat Party is just like…
0:42:28 NP: It’s never been coherent.
0:42:29 WT: There’s factions within. It’s not going anywhere.
0:42:33 NP: They at least… The whole Green New Deal, they’re kinda hearkening back to a FDR-esque, “Let’s do a big thing.”
0:42:39 WT: Which is good, right?
0:42:40 NP: Yeah.
0:42:40 WT: Let’s think about state capacity. Let’s build up big visions and so on. But unfortunately, the big visions they have are not…
0:42:49 NP: They’re half-baked at this point.
0:42:51 WT: They’re half-baked. They’re half-baked.
0:42:51 NP: Yeah.
0:42:51 WT: And yeah, it would be nice to see some more fully baked things. And that’s why we’re trying to kinda do what we’re doing.
0:42:58 NP: Of course.
0:42:58 WT: Let’s understand these things. Have some of these good ideas sitting around, of like, how do we analyze this? How should the ruling class look at this?
0:43:04 NP: Yes.
0:43:05 WT: So what… More broadly, I’m interested in the state of manufacturing and industry and so on in America, and where’s it going right now? Are we losing more or is there a resurgence coming? Or are we looking at a wider sphere beyond America or like…
0:43:36 NP: Sure.
0:43:38 WT: I’d like to know your thoughts on that, ’cause you’re right in it.
0:43:40 NP: So yeah, it’s interesting. So I would say that even though the sort of CNC stuff and all that traditional stuff has, went to Japan, and then the actual assembly work went to China and whatnot. It’s interesting to see where the new stuff is coming from. So in the thread actually, so I CC’d a few of my friends who run the most relevant 3D printing companies now. And so they were talking about… So this would be your desktop metal, working on 3D printing and metal, Formlabs which has a very affordable 3D printer, and so on.
0:44:10 NP: And so they were basically like, “Hey, actually, most of the interesting stuff is actually coming from America and it actually from the same academic system.” Both of them are in Boston. Boston is America’s CAD capital. So MIT has made a lot of robotics stuff, a lot of CAD stuff. And so there’s just a good talent pool. So we have very good human capital in that area. And then in San Francisco, we have carbon robotics which is very well funded and doing a totally different way of 3D printing. So you can see 3D printing in the machine side is… Actually, the technology’s great and is an American thing.
0:44:39 WT: Yeah, 10 years ago it was just a bunch of nerds printing stuff out in ABS and nowadays it’s 3D printed rockets out of metal. They’re doing high quality stuff.
0:44:47 NP: Yeah, it’s interesting to see… So the first 3D printers were actually in the mid ’80s and super early.
0:44:53 WT: Yeah, that’s with the starch and so on.
0:44:55 NP: So that’s, yeah, that’s the Z Corp style. And so that was one, there was stereolithography which is a UV photopolymer where you shoot a laser and it cures, and then there’s the FDM, which is squirting basically hot blue plastic and build it up into the thing. And then hobbyists found that because some patents lapsed in 2007 or something.
0:45:12 WT: Okay, that’s what caused that.
0:45:13 NP: And then MakerBot came out… Yeah, it’s funny ’cause I saw all this history ’cause I know the MakerBot, and I was building with one of the co-founders these [0:45:20] ____ before they had MakerBots. It’s so funny. Then that resurgent blew up and MakerBot was acquired and nothing really came of a lot of those machines. And then it was in the low of the pipe cycle. It was like everyone hated 3D printing companies. And then out of the ashes…
0:45:34 WT: When was that?
0:45:35 NP: That was probably five years ago, I would say. So MakerBot was acquired and nothing happened. All of these ones that came out and nothing came of them. It really was because, at a certain point, it was like, “Cool, they’re cheap now.” 3D printing is cheap. And so the real question is, what are the capabilities do you have for us?” So the metal 3D printing, the faster 3D printing of carbon, the more affordable one, new processes, which is what Formlabs looks at, and how do you democratize the old processes that are $100,000 bucks or $500,000 bucks, that’s happening.
0:46:02 NP: And then on the software side, we’re seeing also American companies, like nTopology is making CAD just for 3D printing because these new capabilities in 3D printing are limited by regular CAD systems that are really designed in the mid ’70s.
0:46:14 WT: Yeah, and the CAD is designed off for machining-type stuff. It’s flat surfaces or curved only, one direction and that kind of stuff.
0:46:21 NP: Yeah, like sheet metal… There’s a couple… There’s probably three or four major packages in CAD of the different manufacturing types. And so this is another one of those things that will probably be absorbed into CAD over the next five to 10 years. So that’s on the cutting edge of American manufacturing. It still seems strong. It is still supported by our industry, university, government complex still seems to be pretty good in that.
0:46:44 WT: Okay, that’s good to know.
0:46:46 NP: Yeah. And actually, the book that inspired me came out of MIT. It was called Fab, by Neil Gershenfeld. It still leads the center of bits and atoms which is some of the fundamental work. And they’re working on programmable matter right now very theoretically and somewhat applicationally. So the cutting edge is there. The part that everyone sees is automobiles going offshore or this kind of stuff.
0:47:07 WT: Yeah, and steel and all these burned out towns, used to be doing something and now aren’t.
0:47:11 NP: Totally, totally. And I grew up north of Pittsburgh and rusting was so everywhere. And I think the major issue was… I think that you want some industrial base for innovation. You look at Hyundai, the car company, bought a steel mill just so it could get innovate in steel to make the cars light.
0:47:26 WT: Yeah, and you see stuff going on with SpaceX insourcing everything because stuff out there just isn’t good enough for what they were trying to do. And this is also… When I think about what is the nature of the city or nature of what causes it to be possible to do startups, it’s like you have this ecosystem of capabilities around where you have people who are experts, you have all the machines that are just there, whether you can purchase them or rent them, or use them, or whatever. All the parts are around. You can just go and ask the guy next door what he thinks about how how he solve this problem. This is so critical. People overlook this. With a big company, they have enough capital to kind insource a lot of stuff.
0:48:08 NP: Somewhat, yeah.
0:48:08 WT: With the startups, you really need this ecosystem to be able to do innovative stuff.
0:48:13 NP: What’s interesting is the best ecosystem for hardware on earth, is Shenzhen. And so before I started Plethora, I went over to Shenzhen, and there’s an area in downtown Shenzhen called Hua Cheng Bei, and that is… The electronics market is a whole district of 10-storey buildings where you’re just buying capacitors and motors and stuff. It’s crazy.
0:48:30 WT: If you’re building something, you just go out there and get the stuff.
0:48:32 NP: Yeah. And you’re like, “Hey, I need a million of these.” “Cool. We have a warehouse in the outside of town. We’ll truck that to you today.” It’s that kind of town. And there are people who are like… The depth of which there is expertise in medical fasteners, water-proofing, the people exist. That’s part of it. I would say that probably one thing that China benefits from, and I don’t have data on this, but I just think it’s true, is that more smart people go into manufacturing in China.
0:48:57 NP: I would say, here, it’s not been cool probably since ’70s or ’80s to do that. I know a lot of the people who are the heads of these things, and they’re from an older generation. I feel like they wanna retire. They were beat up by the outsourcing thing. A lot of people got beat up. There was the high times of manufacturing, people made a bunch of money, and then it was the low times. And I think they never really got over that.
0:49:19 NP: And so when you look at… So I was talking about the World Economic Forum actually, I don’t know, maybe last year, and they had this whole thing on the Fourth Industrial Revolution. And it’s driven by their members, which are big old companies who have a bunch of factories that are doing it the old way. And so the Industry 4.0 stuff is basically like measure shit and tweak what you find, which is obvious and people are doing that already. So it’s, to me, mostly hype. But one of the things that was funny is, is that the manufacturers see it as just manufacturing, and that’s actually the wrong view for innovation. Manufacturing has to be coupled with engineering.
0:49:53 NP: So if you think of the internet, we would never think about like, “Let’s just think about it in data centers. It’s all about the data centers, build giant data centers.” No. The engineers are directly on the middle of AWS. They’re programming a terminal that is directly talking to the factory, which is a data center.
0:50:09 NP: So, for me, the fact that it’s framed as manufacturing and their numbers make them frame it that way, ’cause I was like, “Hey, where is the mechanical engineering, or any of the engineering side?” And they’re like, “Yeah, that’s a good point. Our members didn’t ask for that.” And I think this is actually part of the problem is that manufacturing… There isn’t a lot of vision right now. It’s like, “What do you mean? We got… GE was like, “Oh, yeah, we have the machines in the ’50s. I’m gonna have computers. How do we even do this?” That’s the kind of thing. It’s like wayward looking like, “Let’s make the legacy stuff still work because I’m trying to get another 2% net income out of this thing.”
0:50:38 NP: And famously, I think it was Jamie Miller, who was the CFO, still the CFO, was like, “1% a year net income returns.” And it’s like that’s not an innovation kind of environment. [chuckle]
0:50:49 WT: Yeah, it certainly isn’t. And this gets me thinking of the nature of more innovative stuff versus more established stuff. So in the innovative stuff, you really have to be… What you’re saying is so you have to be very holistic in everything. The engineers designing from… Because you have to find those bottlenecks and the bottlenecks might be outside of your traditional little sphere. You have to have a very holistic view of, “Alright, that’s where the bottleneck is. We can fix that to get this part better.” And it’s like you build the new thing by looking at all of the inputs and all of the processes.
0:51:23 NP: And you’re on the line.
0:51:24 WT: Yeah, you’re right in there.
0:51:25 NP: It’s hard to work for Apple, because they know this really well, and so you’re always flying to China to basically tweak every little thing.
0:51:32 WT: Yeah, you’ve gotta be there. And this is the… And even you fly to the factory, that’s pretty distant, you compare that to skunkworks and how we had them. It’s like, “Okay, we’ve got a hanger and the offices are in the hanger.”
0:51:45 NP: Yeah, yeah, that’s right, for sure.
0:51:47 WT: Yeah, and so, this is very integrated, holistic process to be able to do new things, but that’s also inefficient, right? In the long… Once you have an established process that’s inefficient and you as an industry matures, it breaks out into its individual components. You’ve got like, “Alright, we now know how to do things, now let’s do each part well.” Right?
0:52:13 NP: Yeah, they specialized after a while which should provide inter-firm competition, which will lead to better services.
0:52:19 WT: Yeah, but it doesn’t break the paradigm. It optimizes the paradigm, but it doesn’t bring a new paradigm.
0:52:26 NP: Yeah, certainly, yeah.
0:52:27 WT: And so that’s, I think the track you’re describing is like, “Okay, you’ve got maturity of an industry, but it’s mature within a particular paradigm, and this competition between service providers and commodity service providers, that doesn’t provide paradigm innovation. What provides paradigm innovation is the vertical integration and the… Having the engineers right on the line and so on.”
0:52:48 NP: Totally.
0:52:49 WT: And so, it occurs to me that that actually can’t really be done without huge capital investment or otherwise some kind of big subsidy to make that work. And in China, they’ve got a lots of capital investment and they do tricks with the currency to keep the interest low.
0:53:04 NP: Yeah, yeah. [chuckle]
0:53:06 WT: Even so they’re building a lot of stuff and shipping it to everyone, the wages are not rising as fast as you might think.
0:53:12 NP: Though they are actually more than Mexico now in the coasts. So if you look at it, they still want standard of living increases, but now China is now outsourcing to Africa and other Asian countries, to do this, right?
0:53:24 WT: Yeah, yeah, with [0:53:25] ____ and so on, which we’ve talked about before.
0:53:27 NP: Yes, yes.
0:53:27 WT: They’re trying to outsource, but they are… My point is they are managing the problem from the perspective of like, “Alright, what is actually the level of innovation versus optimization?”, kind of.
0:53:42 NP: Yeah, and they’re subsidizing it directly. You can actually get subsidies to put robots in your system now and it’s kind of funny to think about that for China, but they wanna see these productivity increases and they’re actually lagging on robot deployment. I remember seeing numbers the other day that was basically, South Korea has four or five times a robot per industrial worker per capita than the United States, and United States is maybe two, three times China. And so it’s like China really far behind in robot deployment.
0:54:08 WT: It’s probably because it’s very human-heavy.
0:54:10 NP: Yeah, exactly, which kind of makes sense. They wanna keep a lot of full employment and keep [0:54:14] ____ the provinces, being able to onboard to modern society. That all makes sense to me, but… Yeah, I think they’re now like, “Okay, cool. Now switch gears and then throw robots in everything.”
0:54:24 WT: They do a lot of long-term planning. They’re thinking about, “Okay, we need to have a transition to a more service-style economy in the future,” or like people not just doing these raw labor jobs as much.
0:54:39 NP: Yeah, fully, yeah.
0:54:40 WT: And so they’re like, “Alright, well this is actually going to change the shape of cities and so on.” So they’re building cities in anticipation of that kind of transition. It’s crazy, the forward-looking stuff they’re doing. On the other hand, of course, a lot of it ends up being sort of misapplied and badly founded and they build ghost cities that are essentially real estate scams and stuff.
0:55:04 NP: Yeah, I feel like there are… Yeah, totally. And I feel like in some ways, China in maybe a joke way, is almost like the biggest practitioner of MMT, where it’s like, they’re like, “Oh, that didn’t work? Cool, just write it off. That’s our own bank.” And so they just wasted a bunch of limestone and natural gas they dug out of the ground, but that’s it, in a sense.
0:55:24 NP: Yeah, totally, but they make the food that they feed those people too, so it’s in a sense they’ve drawn their balance sheet down on a resources perspective, but they… You kinda can, in a sense. So I remember seeing recently they’re solving for effectiveness, not efficiency right now.
0:55:40 WT: Yeah, and that’s I think really what you do wanna do. I remember… There is one of these sort of abstract principles I’ve learned over the years, is like raw throughput is often much more important than efficiency.
0:55:52 NP: Yeah, this is like the fat startup. People talk about lean in the beginning, but actually at a certain point you have to go fat. And first you’re doing “un-scalable” things, and then you figure out how to throw money at engineers. Zenefits tried to do this to an explosive degree, and it blew up because of some other issues, but like…
0:56:08 WT: Yeah, sometimes it doesn’t work out. [chuckle]
0:56:09 NP: Yeah, sometimes it doesn’t work, but I think that in a sense, you kinda do wanna do that and this is what the blitz scaling thing is in the startup world.
0:56:17 WT: Yeah, interesting, yeah.
0:56:19 NP: So we were saying, back to US policy, what it is now, and then maybe where do we go with it industrially.
0:56:24 WT: Yeah, where would you like to see it go? Given, I imagine you’ve thought about this a lot, thinking from a strategic perspective of, “Okay, we’re in competition with China. We’ve gotta think about how to keep all our capabilities at least somewhat in-house, we have to keep everyone fed, etcetera.” What do you… How are you looking at this?
0:56:44 NP: Well, I think there’s a couple things to say of what will happen. One, I think it’s true that… So there’s a weird thing in the long term where once technology exists, then the question is like, “Well, how would you produce it?” So, say that China were to surpass us in 20 years, in at least certain products. They already, they already have 5G.
0:57:02 WT: They have high-speed trains. We don’t have high-speed trains. [chuckle]
0:57:05 NP: Yeah, exactly. Although we could build them now. We could build them.
0:57:07 WT: Theoretically, if we had the political will. [chuckle]
0:57:09 NP: Exactly. So we have… That’s certainly the institutional and bureaucratic improvements we have to have, certainly need to happen for any of this to really work. But let’s assume that we can get our shit together over the next couple of decades, which, I don’t know, I kind of feel like we are.
0:57:24 WT: No, I see a lot of people not… I went into this, what we’re doing here with Palladium, I’m in engineering because I realized the big problems that we’re facing as a society are on the human side and not on the technology side.
0:57:37 NP: Yeah, fully.
0:57:38 WT: And yeah, so it’s like, “Okay, I’m working on that problem,” but then I also meet other people who are looking at this. And so I am actually optimistic that we can sort these things out.
0:57:48 NP: Totally. So let’s assume we can. So one thing is, I’m interested in… First of all, I fully believe in, if I was king of the United States or whatever, I’d totally increase more funding to DARPA, NSF. I would probably make NSF into more of a competitive structure where it’s like… Right now, I think how it works is there’s like, “Okay, we’re the people who study E. Coli. There’s a board of people, we approve all the grants.”
0:58:10 NP: There’s not like three boards and you look at what they did. There’s not like venture capital funds where the worst ones die. So I would like the NSF actually to have a lot more funding, but actually some kind of competition in the outcomes. This is actually what DARPA does really well.
0:58:24 WT: Competition between the…
0:58:26 NP: Rival groups, yeah, who are trying to achieve something. And so they’re right now more like banks in how they think about it, and it’s like, okay incremental steps in E. Coli, so you can show certain benefits, but in the DARPA sense, they actually are like… So the biotechnology office, I know that what Justin was saying is like, “Okay, any of my program managers could just cut checks up to a certain amount, and I just say, do it, pull the trigger and just do it.” And so they’re like angel investors. They’re like government angel investors. And then he’s like, “Okay, this one’s doing really good. That guy gets a higher limit or gets to stay longer or whatever.”
0:58:58 NP: And so I love that kind of management. I would love to see our funding ability at the top to fund fundamental R&D, ’cause that’s where new companies are gonna take that R&D. It’s like the CNC thing, it was the Air Force, then it turned into this great industry, and then it gets commoditized, and then who cares? That’s how the economy goes, and hopefully along the way, you’ll build other capabilities and build on those capabilities and you’ll still have companies, or you don’t want those companies and they leave. So that’s the one thing is R&D. Another thing…
0:59:21 WT: Yeah, and if I can comment on that…
0:59:24 NP: Yeah, yeah, yeah.
0:59:24 WT: So one thing you need to allocate a lot of money well is you need to find the people who have the good judgment and you can just… And let them kinda do it unconstrained.
0:59:35 NP: Yeah.
0:59:35 WT: You can’t just put together a committee of five PhDs total selected. They’re gonna end up being too conservative.
0:59:42 NP: Totally.
0:59:42 WT: You need these like crazies or [0:59:44] ____ green men…
0:59:44 NP: I agree, yeah, and I think we need to figure out how to actually compensate them well and allow all the crazy people to get into government in the first place like…
0:59:51 WT: Yeah, and this is…
0:59:53 NP: ‘Cause people who do this were nuts, all of them.
0:59:54 WT: Yeah, and this is I think a really important thing is you do in these times where you’re gonna get this massive growth in capability, you need these guys who are these big risk takers, you don’t want the conservative people.
1:00:06 NP: Yeah, at least not all of them.
1:00:07 WT: Yeah.
1:00:07 NP: Let’s have both.
1:00:07 WT: You need a bunch of conservative people to keep the critical stuff running. But to really get the innovation, you need a bunch of crazies. But there’s… That reminds me of Lee Kuan Yew’s government in Singapore, they would pay bureaucrats, the top bureaucrats more, much more than the private sector.
1:00:27 NP: Yeah, makes total sense.
1:00:29 WT: And so the… And they… Not just on average that are paid more, which is sort of the case in a lot of western countries now, but they would pay the top guys what they’re actually worth.
1:00:41 NP: Yeah, cool.
1:00:42 WT: Which is a lot.
1:00:43 NP: Yeah, yeah.
1:00:45 WT: And so that’s an interesting thing you have to do. But the one thing I wanted to get into here a little bit is, a lot of these money allocation things are fundamentally political patronage or political conflict and so on.
1:00:58 NP: Surely.
1:01:00 WT: And so you have this interesting balance in society that you have to strike, between productive competition and destructive political competition.
1:01:07 NP: Yeah.
1:01:08 WT: So they are in fact going to be kind of representatives of different factions, like fighting over the resources and allocating them and so on. And you have to make sure… This is one of the big jobs of government actually, especially the very top of government. The big job is to make sure that all the fighting going on, which there always is, is over how to be better, rather than how to take the other guy’s stuff destructively.
1:01:33 NP: Totally. So, there’s an interesting structure that actually is just coming out. So there was a thing issued maybe two years ago called Eye Bonds. And so I’m also really into financial instruments, in new ones. So one of the ones that I found was basically they wanted… There’s this one woman was, she’s blind, and a researcher, and she somehow developed a bond that was basically chartering a, I think a non-profit trust, I think, and then that trust was funded by a bond issue, that only for this entire organization was just for eye research.
1:02:01 NP: And so there was a group of people who are the trustees who would allocate this research, and the bonds would pay out based on the success of that research. And the government was actually backstopping if it didn’t work. So basically, private investors partially fronted money, and it’s like, “If any of these Eye things work in terms of companies, we get an upside.” And the government’s like, “We’ll get an upside too, but we’ll also protect your downside to make market rate returns so investors actually wanna put money into these bonds.”
1:02:24 NP: And so this is one particular structure where it’s sort of like a social impact bond, was another structure like this, and what the social impact bond was is like, for a very basic example, imagine you wanted to vaccinate a bunch of kids in some country and you wanted the market to figure out how to get the cheapest way, and so you would issue these bonds and people would be able to say, “Oh, I’m the person who knows how to make vaccines, the government has money, and then there’s this group of people who are poor who don’t have vaccines. Let’s clear this market.” And so the government would say, “Okay, we’ll pay for success. So we issue these bonds and if you can actually do it cheaper than we say, then you get to keep all that money.”
1:02:58 WT: Yeah, yeah, this…
1:02:58 NP: The auditing was hard though.
1:03:00 WT: Yeah, that’s the trick, is you have to be able to measure it well or it doesn’t work.
1:03:04 NP: Yeah, so in reacting to this hard audit problem, social impact bonds, they said, “Okay, maybe let’s just say the auditing is almost impossible.” So that’s actually just saying the Eye Bond, that any wealth made, here’s how we’re going to allocate it and here’s how we’ll backstop it if we don’t get enough wealth to make a market rate return, and then the investors will get the money. So it’s almost like, how can we encourage capital to come in in a market rate…
1:03:26 WT: Just do something in the space, not auditing what it did necessarily except that it made money.
1:03:32 NP: Exactly. It’s like, well capital won’t invest if it’s bullshit. So first of all, it’s like we don’t know how they’re doing it. They might have their own experts…
1:03:38 WT: They don’t even necessarily know what they’re doing.
1:03:40 NP: Yeah, so that’s the thing is, it’s actually like you don’t have the audit, just make it a price model. So then that’s what this thing does. And so then the government’s just like, “We’re gonna let you guys do your thing. We’re gonna get a piece ’cause we’re taking some risk off the table for you.”
1:03:51 WT: Yeah.
1:03:51 WT: And I think structures like that where they actually chartered a trust to pursue eye research, we could just repeat that, so then you avoid the political game because it’s a pseudo-private, but a public/private partnership, but it’s… Unless the trustees are the political issue here, but I don’t think people would invest in it that way. ‘Cause if I’m doing due diligence thing as a big company, I’m not gonna be like “Oh these are a bunch of sycophants. So maybe there are institutions that could get around that.
1:04:17 WT: Yeah, and I think there are definitely a lot of such things and each structure is gonna work in some limited…
1:04:25 NP: Exactly, yeah.
1:04:26 WT: That only works when you can have certain types of characterization of the impact, a characterization of what you’re trying to do or like the space you’re operating in and so on.
1:04:34 NP: Yeah.
1:04:35 WT: Yeah. And then there’s this, of course, larger issue where you can’t always do that, and so someone at the government has to just kind of do it on judgment at some point.
1:04:43 NP: Yeah.
1:04:45 WT: But…
1:04:47 NP: There’s another piece of this policy that I think this is the most radical thing that I think in this space is basically like… So there’s nothing stopping the US from changing how intellectual property works vis-a-vis our response to China.
1:05:00 WT: Right.
1:05:00 NP: So, say, the 5G thing, we can reverse engineer that shit and make our own 5G based on theirs. What’s stopping us from doing that? Oh we don’t like it. We should just open the chips and dip it. So that may actually be a thing where it’s like, “Oh you’re not playing ball with us, we won’t play ball with you. We’ll just absorb all your good technology.” And the question is, it’s like, okay, so what would prevent us from doing that? So, circuit boards are easy. We have our own chip people. Maybe they have better semiconductor fads ultimately. They’ve got their six-nanometer technology and now we’re doing, intelligence community is acquiring shit in the same way theirs is doing it to us.
1:05:28 NP: And maybe that’s just the game. And there’s another interesting thing which is around how reverse engineering occurs. So we’re actually, to me, coming into a golden era of reverse engineering. What we have now is the ability… A lot of the equations are one way. It’s really hard to develop something, but once you see it, like a drug, it’s easy to copy. And so most things are that. Most consumer electronics, there’s not some secret sauce. It’s like if you open up the Apple chip, you poured nitric acid on it and looked at it, you’d be able to take that and make your own because its just a bunch of ARM cores and some glue.
1:06:00 WT: Yeah.
1:06:00 NP: If you wanted to spend the money. And even then it’s not actually that special so no one does it. But if you did have some nuclear reactor inside, it’s pretty rare. The things that are nuclear reactors though, you can reverse engineer even better as well. So maybe we should actually be funding an entire apparatus technology on how to reverse engineer, and just absorb all the best technology and take it ourselves.
1:06:19 WT: Yeah yeah yeah, that’s interesting.
1:06:20 NP: So they’re not playing within the WTO, [1:06:23] ____, whatever. It’s like “Oh cool, we’re just gonna take your shit ’cause that’s how the world works now.” And maybe the people who are these open source components, there’s, Against Intellectual Monopoly, is like the book that goes over all the issues with intellectual monopoly and property, and maybe we want to actually support that system and undercut them from that way.
1:06:42 WT: Yeah, and this is something where you look at patent law in the past, and it was during wars, these eras of competition, may or may not be a war, but during these eras of competition, governments often find the patents are actually really, like the patents end up really locked up and they have to force a bunch of people to give up the patents.
1:07:02 NP: Yeah, there’s a law that let them do that. Yep, totally.
1:07:04 WT: Like what happened in World War I with aviation, like all the different companies who held different parts of the…
1:07:09 NP: Interesting, huh.
1:07:11 WT: For building airplanes.
1:07:12 NP: Yeah.
1:07:14 WT: And so the government basically had to say like “Alright, look, you guys are gonna share the patents.” [chuckle]
1:07:18 NP: Yeah, that’s cool.
1:07:20 WT: And the patents and the copyright and all this stuff ends up getting, gets a lot of… Captured by the industry basically…
1:07:29 NP: Fully, fully.
1:07:29 WT: To protect monopoly interests, unfortunately. And so maybe some competition can spur reform there.
1:07:37 NP: There’s a legal thing here, but there’s also the technical thing of reverse engineering. So what’s interesting is that ML systems are great at figuring out what’s going on. So for instance, they’re now, I believe, Fraunhofer in Germany is working on CAT scanning technology that can get to a micron resolution. Fully 3D. So you could put like an iPhone in there and it would tell you the chipset, all the things in the circuit board. It actually has a spectral analysis that will tell you what the materials of it are. It’s actually nearly a copy machine for reality. So like that… Right now it’s like thousandths of an inch, so it’s not quite detailed enough for everything.
1:08:12 NP: But if you take those fuzzy models that you get out of CAT scans and put them through an ML system, you would then have a CAD file fully. So what I always thought of is like if you put a CAD file into the Plethora in its ultimate conception, it would just make it, right? So it’s almost like…
1:08:27 WT: Like you could actually do replicator.
1:08:29 NP: Yeah, basically, you have a copy machine for reality. And that technology is like maybe 10 to 20 years away from maturity.
1:08:35 WT: Okay, that’s crazy.
1:08:35 NP: And if that’s true, it’s like any invention can just be reverse-engineered instantly, in a sense…
1:08:40 WT: Yeah.
1:08:40 NP: Unless there’s some super secret thing. The only thing that goes back on is actually quantum shit, because simulation technology, vis-a-vis like if you can take the nuclear reactor in the product, ’cause it’s gotta be in something, right?
1:08:52 WT: Yeah.
1:08:53 NP: And you can figure out like, okay, in effect, all hardware is open source ’cause you get the thing. And so you’re able to open it up and figure that out, and then the simulator can actually guess and check also using ML situations to do that. It’s just simulators are gated on their computational complexity. So a lot of people predicted that quantum systems will be able to solve these simulation things really well.
1:09:12 NP: DNA and protein, and this is like the classic problems that, I remember Rigetti quantum, I think talks a lot about like protein folding. And that may be actually technology in the future that are the most relevant or bio-technologies anyway, but even physical technologies, ML and quantum will help make the simulations faster and less power-intensive right? And those resources…
1:09:32 WT: Yeah, that’s interesting.
1:09:32 NP: Which will also drive reverse engineering.
1:09:35 WT: Let’s suppose you could do full reverse engineering, copy anything you could find, then the game… Strategic advantage becomes not in the hardware itself, the manufacturer, but in the hardware that manufactured the hardware.
1:09:49 NP: And even beyond this. So yeah, I agree with you. So that’s the machine tool thing.
1:09:52 WT: The factories…
1:09:53 NP: Yeah.
1:09:53 WT: And the machine tools.
1:09:54 NP: And the tacit knowledge of how to set that up. So let’s circle that as like the big thing of the people and machines tacit knowledge engine, like why China hasn’t replicated Taiwan Semiconductor, even with infinite money to do so.
1:10:06 WT: Yeah, right.
1:10:06 NP: ‘Cause they’ve gotta build that brain over time. So another funny thing is the adoption characteristics of the consumer culture of your society. So what’s interesting is like China is actually way more into buying shit. They’re a nouveau riche country and a lot of people went from poverty to having something and they’re like “Cool, a cheap VR system that my friends will think is cool and breaks in like 10 days? Cool. That’s so cool.” And that actually lets people sell into markets in a more innovative way. There’s a reason why the first genetically modified baby was in China.
1:10:34 WT: Right.
1:10:35 NP: I think it’s just more permissive in many ways.
1:10:37 WT: Yeah, I wonder…
1:10:38 NP: Adoption may actually increase their ability to make innovation in a consumer sense that will drive adoption.
1:10:44 WT: Right, because the consumer is this big subsidy on just like can you deploy something?
1:10:49 NP: Yeah, yeah, and I think it might also just be western values as well. I think in some ways, we in the States, and maybe definitely in Europe, are maybe a little more snobbish in this way. We want the iPhone to be perfect, and if your phone breaks and it’s cheap, then you’re considered a rube or something for getting screwed.
1:11:05 WT: Right, right.
1:11:06 NP: Where in China, it’s like, “Yeah, everyone’s just buying the shit, it’s like cool, right?” I feel like they’re a little bit more permissive. [chuckle]
1:11:11 WT: That’s interesting. That’s interesting. Of course it’s related to the… The flip side of that is the… What’s that word, like [1:11:19] ____ or something, like good enough.
1:11:22 NP: Yeah, yeah, yeah.
1:11:22 WT: I don’t know if I got the right word, but where they… You hear these horror stories of tin cans instead of rebar and…
1:11:28 NP: [1:11:28] ____. Yeah, yeah, yeah, yeah.
1:11:29 WT: Concrete and stuff.
1:11:31 NP: And then there’s Shanzhai too, which is actually their, I think it means “green mountain” or something. And basically, it’s the culture of pirating, where people that… You just pay the guy to give you the design and he does it ’cause that’s more than his annual salary.
1:11:44 WT: Right, right, right.
1:11:45 NP: You have like five grand or something, and then you can do all your own stuff using their technology. And that is pretty rampant over there and sort of does lead to innovation in many ways.
1:11:56 WT: And this is interesting. It’s like, how strong do you actually want your IP law regime to be? In China, they have gone with this model where people are just stealing IP from each other…
1:12:09 NP: For certain stuff, yeah.
1:12:09 WT: And copying each other all over the place. So, the startup eco-system in China is extremely cut-throat…
1:12:14 NP: Yeah.
1:12:14 WT: For that reason. Whereas in the west, we’ve kind of like… We’ve got pretty strong protections there, and so it leads to, I guess, a different set of problems. You can’t do this big private investment in technology without good IP law. On the other hand, they’ve got other mechanisms for doing that, but without… But with strong IP law, you get these sort of patent standoffs and stuff.
1:12:44 NP: Mm-hmm.
1:12:44 WT: Something could be possible except no one can actually…
1:12:46 NP: Yeah.
1:12:47 WT: Get all the patents together to do it.
1:12:48 NP: It does seem though, that the major fang style, big tech companies are actually not patent-gated. They’re actually more fair-use, copyright-gated.
1:13:00 WT: Mm-hmm.
1:13:00 NP: One of the things I’ve always wondered is, the real thing on Facebook, I forget what the social network was a few years ago, that they moved off of it, but it basically let you copy all your shit and share between all of ’em. So your Twitter and Facebook and everything were connected.
1:13:12 WT: Right, yeah.
1:13:12 NP: And they hated that. [chuckle]
1:13:15 WT: Yeah, Twitter and Facebook both changed their API interface to not allow that.
1:13:22 NP: And what’s interesting though is that, in the FCC for phones, kind of famously, if you flip over a receiver phone, it says “This device must accept all interference and generate no interference.” And that’s a very basic rule for how the phone has to operate. But in an internet context, I’ve always thought that there should just be a rule that’s like, “Anything a human can do on this website, a machine is allowed to do on this website.” So if I give my password to something and it sucks down all my friends and network and posts or whatever by just clicking everything, that’s totally legal. And I think that alone would actually maybe increase the competition rate, without a lot of, breaking the companies up or something.
1:13:56 WT: Yeah. I’d…
1:13:56 NP: Which might still be useful at some point.
1:13:58 WT: I’d love to see more competition and more decentralization of some of these… The big rent… Basically like rent-seeking companies…
1:14:07 NP: Totally.
1:14:07 WT: In social media and so on.
1:14:08 NP: It’s funny that China also has the opposite thing. They actually like these Alibaba and WeChat, and these things are super-centralized ’cause then they can regulate them more easily.
1:14:17 WT: Yeah, well then it’s just like…
1:14:18 NP: They also pick the winners in that way too.
1:14:19 WT: They pick the winners and then they send in some party men and they’re like, “Alright, now you’re gonna play ball with the government.”
1:14:23 NP: Yeah, any company, any sufficient size has a party person, I think.
1:14:26 WT: Yeah, and that would be something… In America, we could do that as well, and I thought that’s a decent system. I don’t know, there’s obviously ups and downs to it, but you could imagine you have every company past a certain size has a more formal party office, in the thing, from the government. ‘Cause right now, it’s not like we don’t have…
1:14:46 NP: Yeah, stuff like that.
1:14:48 WT: The companies play ball with the State Department and…
1:14:50 NP: Yeah, fully.
1:14:51 WT: And the Department of Defense and…
1:14:52 NP: NSA. [chuckle]
1:14:53 WT: All the political factions and so on.
1:14:56 NP: Yeah, sure.
1:14:56 WT: They have to pick sides and play ball.
1:14:58 NP: Surely.
1:15:00 WT: And that could all be a lot more… Really, one of the big central jobs of the state in society is to coordinate all that stuff.
1:15:05 NP: Yeah.
1:15:06 WT: And right now it’s kind of a mess in America.
1:15:10 NP: Totally.
1:15:12 WT: But yeah, you could imagine that all being much more formalized and more efficient. But that, this kind of brings me back, it’s something we mentioned earlier, why couldn’t we just start stealing technology, as America? And like…
1:15:24 NP: Borrowing.
1:15:25 WT: Or yeah, I guess it’s like it can be copied, so technically it’s…
1:15:30 NP: Yeah.
1:15:31 WT: Whatever.
1:15:31 NP: [1:15:31] ____.
1:15:33 WT: It’s folk technology. It’s like folk music.
1:15:38 WT: But Trump tried to do some of this stuff, in terms of starting to play economic hard ball with China, but of course, big political problems there ’cause he’s trying to do all this other stuff, a lot of people don’t like it, a lot of factions.
1:15:51 NP: And maybe the tools were blunt and weren’t the right tools. No one invented newer tools either.
1:15:55 WT: Yeah, well… And part of the problem there is just there’s no faction coordinated and secure enough to actually make the investments into how you would do that, and then have the ability to actually deploy that in government.
1:16:09 NP: Yeah.
1:16:10 WT: And so we kind of have this big mess on the government policy, government’s angle.
1:16:15 NP: Yeah.
1:16:16 WT: And then the other thing is the spy asymmetry. We have a massive spy gap with China.
1:16:21 NP: Yeah.
1:16:22 WT: They killed one of the CIA people and they have tons of people…
1:16:25 NP: Yeah.
1:16:26 WT: In America.
1:16:27 NP: Yeah, and I think that there’s something about authoritarian states that, especially ethno-authoritarian states, where it’s not like… I’m guessing you can get white people to go to China, and for some reason… But at a certain point, it’s like the… [1:16:40] ____ actually is a weird thing where it’s easier for them to find us than vice-versa, and we’re a permissive culture that just lets that work.
1:16:47 WT: Yeah. And so that’s gonna be one of the big challenges is like, “Okay, given the model that we’re running on here, how do we actually deal with the spy gap or something? I do suspect, actually, there’s going to be… Over the next few years, the competition between America and China is going to bleed down into the cultural factions and the…
1:17:16 NP: Cold War II.
1:17:17 WT: Yeah, and it’s like in World War II, we had obviously a big Japanese sub-population on the west coast and so on…
1:17:26 NP: Yeah, true, true.
1:17:28 WT: And we got into a war with them and it becomes hard to tell what the loyalties are for everyone, and unfortunately, they put them in internment camps.
1:17:37 NP: Of course. Yeah, yeah.
1:17:38 WT: It was really bad, basically, for a lot of people. And I could see that kind of thing happening with the Chinese.
1:17:46 NP: People are already scared. When I was in Pennsylvania and dated an Asian girl from China and we would go… She lives in the suburbs and she would get yelled at like, “Taking our jobs,” or stuff like this. And she was a university student.
1:18:00 WT: There’s these very working class concerns, like taking our jobs, etcetera, which, those are valid concerns, but they’re not bubbling up to the top levels right now, whereas, the national security concerns are gonna be something that bubbles up.
1:18:13 NP: Well, it already is. I think that… I’ve run into CFIUS, which is the financial thing about Chinese venture capital and what you can take or not. But not even just the government stuff from CFIUS, but there’s also… When I have VCs from China or something like this, and I’m like, “Huh.” I’m a little bit freaked out ’cause I’m like, “Hey.” People ask a very specific questions.
1:18:35 WT: Yeah. “Are these guys VCs or spies?”
1:18:37 NP: Yeah. And I don’t know. Sometimes I wonder. It’s like, yeah, there’s a lot of money in China. But I don’t know about manufacturing, so yeah, totally. There’s probably a lot of experts in China. It would be great. And even if they were those people, who’s to say that the government is like, “Oh, you’re gonna do this because you must.”
1:18:54 WT: Yeah, yeah, yeah. Like they call them up and they say [1:18:57] ____.
1:18:57 NP: Yeah, they couldn’t even say no. Right? It’s not even like the most ethical people in the world, but if your livelihood is at stake, you have to do it.
1:19:04 WT: Yeah, yeah. I know. Right.
1:19:05 NP: Yeah. It’s a really unfortunate thing, at least like the… I think, we’re getting our shit together in the West as far as… And the rule of democracies, whatever, in general, and we’re gonna get our institutions such that, that people look at us and they’re like, “Yeah, that is actually the correct way of doing things.” Because no one is voting for tyranny. No one actually wants tyranny. The only time you get popular dictatorships are like heavy duress times.
1:19:28 WT: Yeah, when everything is going really sour, people will take the tyranny over the chaos, basically.
1:19:35 NP: Completely. Yeah, that’s the only time it happens. Where everywhere else, “Of course, we went self-rule.”
1:19:39 WT: Yeah. So then, the question is, what does that look like? Is it working? I’m obviously… I’m not very interested in exploring alternative models, like more technocratic or even like aristocratic or other forms of democracy, or whatever. Let’s look at those and let’s think about them.
1:19:57 NP: Yeah. Another podcast. [chuckle]
1:20:02 WT: But… Oh, you said something that I wanted to follow up on. I’ve lost my…
1:20:11 NP: Pre-China?
1:20:13 WT: No, it was just in the last few sentences. There was something interesting. [chuckle] Sorry, I’ve just completely lost my train of thought, but maybe I’ll get it back.
1:20:24 NP: Okay.
1:20:27 WT: But let’s see, what else is there to discuss? Ah, man, it’s just… It’s bothering me ’cause I had a really good question. [chuckle]
1:20:37 NP: I’m trying go to replay what it was.
1:20:38 WT: No. Yeah. We were talking about the the spy issue, the symmetry, the institution stuff.
1:20:48 NP: Yeah.
1:20:51 WT: Yeah, I don’t know. I’ll drop it.
1:20:54 WT: We might come back. We’ll see.
1:20:55 WT: Yeah, yeah. We’ll see. But yeah, there’s a lot of these scary kind of possibilities on the horizon with how the competition with China ends up playing out industrially, and so on. So, yeah, you had some good ideas in terms of how to kick the innovation into gear. And then the challenge with the innovation, of course, is just like, how do you actually hold onto it and maintain an advantage and not just have people who see it similarly.
1:21:28 NP: And ultimately, what’s interesting, though, is that it’s really the deployment of technology that actually affects people’s livelihood, so I think we have to always view things through what are the material conditions of the people on the ground?
1:21:40 WT: Correct.
1:21:40 NP: So everyone in Europe has cellphones because we invented all of that stuff. We invented GPS, the chips, everything, so the whole world can have this stuff. And so, we don’t even have the better versions of those things, necessarily. Yeah, maybe we all have the same iPhones, whatever, but we did invest a lot to do that and we do get to reap a lot of money because Apple pulls in money from all over the world, so great.
1:22:04 NP: But even if we didn’t tap all the money in the world for Apple, we could still buy iPhones, and it’s a massive productivity increase. The question really is from a company, or a country point of view, is like, “How do we take any good shit in the world and then deploy it in our our country?” That’s where it’s like, whether we develop it, they’re developing it and we take it and make it here. This is a question of how the material get transformed.
1:22:24 NP: And then so, from that perspective, I almost think that culture matters more than anything. And like culture almost in a business sense. Like in a business, we talk about organizational culture. We say that, “Culture eats strategy for breakfast.” That’s a thing that they say. Because it’s like, the way you work on these things, is actually how the thing will work in your society. North Korea can buy iPhones, but the thing doesn’t work that well because of their cultural and institutional issues.
1:22:52 WT: Yeah, yeah. You mentioned bringing Westerners over to China. One of the reasons that happens is, in the nuclear power plants, the Germans have people who will obey orders to the letter, and actually be precise enough in their activities, in their culture, to operate a nuclear power plant. And they keep having trouble with this, with the Chinese thing where the contractors will put tin cans in the walls.
1:23:17 NP: Yeah, yeah, yeah.
1:23:18 WT: But it’s just this cultural difference that they’re having trouble with. But, yeah, so this problem of the culture of your country in business and other things, how do you manage those things? How do you maintain them in a good direction so you actually have… You’re sharing the things you need to be sharing and not sharing the things you don’t wanna be sharing?
1:23:45 NP: And this is, to a point earlier that I made, when talk to people in the industry, there is not much of a young guard in manufacturing, who is ascendant. So when I look at things, it’s like… So my friends… I think I know everyone who has a young startup or even a big company but a tech company. And I almost see the Silicon Valley’s culture and business culture is kind of eating the world of business. So it was just on the internet. Now with Uber, and all these, Airbnb… It’s actually just eating everything. And so, if you look at, there’s something that shows that…
1:24:16 NP: Businesses, when you ask their businesses what they fear the most, it’s like, “Disruption by new tech entrants.” So, everything is just, “How did we get disrupted by these new kids?” And so, what that means is they’re just trying get… Like famously, Eric Ries goes to GE and tells them about Lean Startup stuff. And now, he just came out with a book maybe a year ago that was on how you do entrepreneurship to bring this culture into your company. So we’re basically exporting Silicon Valley culture everywhere. And I think it’s interesting to think of it as a cultural export, ’cause now, every city in the world has a startup incubator, an angel scene. They’re all reading Hacker News and TechCrunch every day. It’s actually a global culture.
1:24:49 WT: I wonder how much of that is, it’s Silicon Valley culture versus this is just how young people operate these days and you just need to bring in young people.
1:24:57 NP: I don’t think it’s just young people because I actually think of it as… So this is a little off topic, but I see… You know there was Atlanticism, was a thing. That the East Coast of the United States and Western Europe were in a special bond together. The Washington Consensus, all this stuff. And I actually feel that we’re seeing more of a Pacific movement in thought, where it’s like the innovative people are in Asia and California and everything like this. That’s actually where a lot of the innovation’s happening in the world, where are all the new products and services coming from. They’re in the Pacific now.
1:25:27 NP: And I think that people like New York, London, Berlin, are actually old school in many ways. And those young people are pessimistic hipsters, basically. And hipster-ism is basically, “We don’t know what the future will be so let’s dress like the ’20s and ’30s or in ’50s. Where that was a classic time. Let’s do artisan things because we hate capitalism and industry and automation and technology.” And so, to me, that is not a young person thing… I mean, that is a young person thing but it’s not an optimistic thing.
1:25:52 WT: Yeah, no, that’s a good point. Yeah, and I thought about this before in the context of the succession problem.
1:25:58 NP: Yeah, sure.
1:25:58 WT: You have these big old companies with a lot of older people in them, and they’re not sort of able to bring in the young talent. And then you have these startups that are basically like rooting around that whole power center. They’re full of young people and it’s… It almost feels like something’s going wrong. I don’t know if this is the right condition for things to be in or if it’s something has gone wrong that we somehow didn’t manage to get into the big established places that had a lot of institutional knowledge and then do our thing there.
1:26:27 NP: It’s interesting, too. When I look at this, I think a lot of times… One of my friends, he is from the Mondragon co-ops. So if you’ve ever seen this, it’s the biggest co-op federation in the world. It’s 75,000 people in 260 co-ops called Mondragon. And so, he was talking about, “Nick, the co-ops actually can out-compete traditional businesses and whatever.” I’m like, “Interesting.”
1:26:51 NP: I don’t believe that’s true, but if you really do believe that’s true, you should start a private equity company and you should acquire a company, turn it into a co-op, and you’d actually be able to make money. And scale that. You’d be able to infinitely scale it. So if you know some knowledge, the question is how you deploy it. It could be that you’re like, “Yeah, let’s buy GE and let’s fire all the management at the top, let’s split up… ” You could do anything you wanted in private equity and do this if you actually thought you knew something that was real.
1:27:15 WT: And you could convince enough people to put together the capital to buy GE. [chuckle]
1:27:20 NP: Yeah, but it wouldn’t be that bad. If you really… So this is the thing. It’s not believable, which is why you can’t do that. People fully believe like in the ’90s PE, where you could go in there, fire people, do some synergy stuff, roll them up together and you could get another… Level it up and you’d get 50% out. That was a standard PE model people believed in. And I think that a lot of the stuff that… You do the… Both of these together and you’re like, “Okay.” You could buy GE or you could start 20 startups that full obviate the need for GE.
1:27:49 NP: Which is cheaper and less risky? Probably the startups. And so, at that point, it’s like, “Why do we even need GE?” So their tacit knowledge. Can the startup just learn that or hire those people? ‘Cause corporations are not monolithic things. They’re actually just pieces of paper that coordinate people together. And that tacit knowledge, all those people organized together, they can divest those business units out. Or you could hire whole teams.
1:28:12 WT: So I guess the question then is basically, how much do you need new structures and new forms and new strategies in the market overall, versus just new people?
1:28:26 NP: Yeah.
1:28:27 WT: And whether that trade off goes one way or the other so that decides where the people go.
1:28:33 NP: I think that novel capital intensity is actually a big thing here. So what are the big old companies? So, novel stuff would be like the Boeing Everett facility. They’re giant manufacturing things, so the… You cannot make a 747 without the tooling. And so, even if you hired the entire damn team, you still could not get the tooling or any of the IP, which is basically crystallized inside of Boeing. So, Boeing will exist forever.
1:28:57 WT: For as long as you need those big kind of things.
1:29:00 NP: Exactly. And as opposed to capital intensive things, like say, farms, where it’s basically like commodity. It’s like commodity land, International Harvester or whoever’s been buying your capex from. And you could be just like, “Yeah, let’s just start our ADM today.” Then it would be no different. So it’s a total commodity management thing, and maybe they just have land cheaply and that’s the entire point of ADM. Whereas, say, like a steel mill of a particular kind is capex specifically made and the people specifically trained. That’s gonna be hard to compete with. So, I think, in a sense, a lot of these things are actually not… Like the tacit knowledge is acquirable ’cause it’s in the people.
1:29:35 WT: Yeah, tacit knowledge lives in people. Yeah.
1:29:37 NP: But some tacit knowledge in effect is crystallized in physical things.
1:29:40 WT: Yes.
1:29:41 NP: Or maybe in code bases, which are very hard.
1:29:44 WT: Yeah. This particular thing is… And these particular people are only valuable…
1:29:49 NP: Attached to that thing.
1:29:51 WT: Yeah, attached to that thing. And have the most value attached to that and thus, that thing stays as the center.
1:29:58 NP: Yeah. And that’s why you acquire those companies. So that’s when the PE model works. It’s like, “Oh we think that steel mill actually should be run differently so we’re gonna acquire that really fancy team and technology, blow away the dumb management team, have our own sales channels that are different. You can iterate around that core tacit assemblage of people and tooling or whatever.
1:30:19 WT: And so, this brings us back actually, to part of your diagnosis of what happened in the ’70s and ’80s with the machine tooling, was you kind of had the thesis that those activities were viewed as unprofitable, or they were unprofitable, so they actually got cut.
1:30:37 NP: Yeah, yeah, yep, totally. Yeah, I think that it’s interesting to think that it was probably the right business decision for those companies, in many ways. There’s these big tool companies, they’re trying to be profitable so they do the divestitures to get cash and maybe those things just die. Sometimes they just literally die.
1:30:50 WT: Yeah. Yeah. The right idea for the companies. And I guess, there’s also this larger question of, “Is this the right thing for America? Was it the right thing to have the landscape of the incentive set up that they would do that?”
1:31:03 NP: Yeah, and we’ve lost the human capital, like these assemblages, in a sense, likely that could have produced other things. And we do see a pretty high conservatism in machine tools, and I have a bunch of ideas of how to re-invent various kinds of machine tools.
1:31:16 WT: Awesome. I’m looking forward to it.
1:31:18 NP: Other startups, right? But I think that they weren’t happening because it’s so hard to make these things, there’s no investor appetite for it. Even today, in the beginning, we thought it was crazy to raise money for a factory, and people hated the factory. And only after many years of me actually having to do it and succeeding at it, were they actually able to do it.
1:31:34 WT: Yeah. And this is really one of the big ways in which cultures and countries get stuck in believing that something… Or like, in not being able to do something, is like, people don’t believe in it.
1:31:48 NP: Yeah.
1:31:49 WT: Just ambient-ly, people don’t believe in it, so the most talented people won’t work on it, the biggest capital won’t go into it, etcetera.
1:31:57 NP: Yeah.
1:31:57 WT: And so much of it just comes down to, can you get that crazy guy who makes that investment of time and money to start the thing off, prove it enough that it can start to become self sustaining and prove it to other people?
1:32:10 NP: Yes, and society needs to be optimistic, as a whole, to actually believe that it’s cool to elect these people, right? To me, there’s this link between… I think the material conditions of the society, so say, like China, their generations are said to be like five years, not 25 years, because the difference in lifestyle is so high because they’re in the middle of the S-curve.
1:32:30 WT: Yeah.
1:32:31 NP: So when you see that, it’s a lot easier to think that’ll just keep going. And then now, we’re at some… The west is at the peak of the development curve, and we don’t know where we’re going. It’s very 3% growth or something at the best, and so, it’s harder to think these things will pay off. So you don’t make these long-term investments, ’cause you’re really scarcity minded and conservative.
1:32:47 WT: And I wonder, though, in these environments where everyone’s really scarcity mindset and conservative, you can get people restarting the growth by, if you’re thinking on a longer time scale, you can out-compete that more conservative mindset.
1:33:06 NP: Yeah, and that’s what venture capital did, right?
1:33:09 WT: That’s how it gets picked up again.
1:33:11 NP: Yeah, totally. And the government ossified and didn’t do as good at venture capital as it could’ve. But even though there’s a lot of good things though, we could make it better. The venture capital guys have 10-year fund horizons, and that’s better than Wall Street that has HFT style, trade up the public equities…
1:33:23 WT: That, and add a second fund horizon. [chuckle]
1:33:27 NP: Yeah, or maybe a private equity of the two or three-year fund horizon. There are longer fund horizons and that’s where all the money was made, but the modern economy was only there because we actually got a good venture community to do that.
1:33:39 WT: I’m thinking back now to some report that I saw a few years ago, where some big investor people were expecting over the next couple of decades, like a bull market in politics. That there was gonna be…
1:33:54 NP: Interesting.
1:33:55 WT: A bunch of things blown open in such a way that smart people were gonna start going into politics again. And I think we are seeing the beginning of that, where people are going… And not politics in the narrow sense of elections and stuff, but the larger sense of the big human institutions, the big, central human institutions in our society.
1:34:13 NP: Yeah. Totally, totally.
1:34:14 WT: And I think that’s how we get out of coming off the top of that S-curve. It’s like, how we get out of that back into growth is you have to… Like the bottleneck in many ways, now is those big human institutions at the center. And so, that’s what I’m interested in.
1:34:30 NP: Totally, yeah. I’m with you.
1:34:32 WT: Anyways. It’s been an hour and a half.
1:34:35 NP: Yeah. [chuckle]
1:34:37 WT: Let’s cut this off. This has been awesome.
1:34:38 NP: Yeah, this is great.
1:34:38 WT: Thank you so much for coming on the show. This was a great conversation. I hope the audience learned something. I definitely learned some stuff. So again, for the audience, this is Nick Pinkston, founder of Plethora, and hopefully many more great projects in the future. And I’m Wolf Tivy, Editor of Palladium, and that’s all for now. See you next time. Bye.
1:35:00 NP: Thank you.