Podcast Episode Summary
The podcast episode features an interview with Chris Power, the founder and CEO of Hadrian, a company that aims to build the factories of the future. The main topic of the episode is advanced manufacturing, specifically the challenges and complexities of the industry. Chris Power explains that advanced manufacturing is an industry term that covers complex and high-precision manufacturing, often in industries such as aerospace, defense, and semiconductors.
Key Points:
Takeaways:
Participants:
Recommendations:
Today, we have an episode with Chris Power, the founder and CEO of Hadrian. Hadrian is a company trying to build the most efficient factories on the planet.
In this conversation, we’ll introduce the audience to advanced manufacturing, and expose them to the reality that it’s a remnant of the first space race. We also cover the challenge of manufacturing, the importance of visibility in complex systems, the killer app for space, how simplifying the world of atoms can be done through bits – and ultimately, what kind of experimentation that may unlock.
Timestamps:
0:00 - Introduction
1:28 - What is advanced manufacturing?
4:18 - When a hatch is jammed at the ISS
6:11 - What’s happened since Space Race 1?
8:56 - A retiring workforce
13:20 - What is at stake?
21:03 - Onshoring manufacturing
24:06 - Is capital enough?
24:43 - Fixing the problem with technology
28:55 - Convincing the old guard
31:20 - Building a new culture
34:45 - Experimenting with hardware
37:24 - The value of observability
41:47 - The cost of timeliness
46:09 - Hadrian’s key risks
49:22 - Why Hadrian pivoted
52:22 - What talent is needed
57:31 - Why focus on space first?
1:03:38 - The killer app of space
1:10:36 - Who inspires Chris?
Resources:
Hadrian’s website: https://www.hadrian.co/
Chris’ Twitter: https://twitter.com/2112Power
Stay Updated:
Find us on Twitter: https://twitter.com/a16z
Find us on LinkedIn: https://www.linkedin.com/company/a16z
Subscribe on your favorite podcast app: https://a16z.simplecast.com/
Follow our host: https://twitter.com/stephsmithio
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. For more details please see a16z.com/disclosures.
Stay Updated:
Find a16z on X
Find a16z on LinkedIn
Listen to the a16z Podcast on Spotify
Listen to the a16z Podcast on Apple Podcasts
Follow our host: https://twitter.com/eriktorenberg
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
Host
They would have had an extra billion dollars in revenue last quarter had more of their parts been delivered on time. Today we have on Chris Power. Chris is the founder and CEO of Hadrian. Hadrian is trying to build the factories of the future. And in today's episode, we talk about what advanced manufacturing is, how advanced it really is, what's happened since the first space race, the complexity of onshore manufacturing, the killer app for space, and how simplifying the world of atoms can actually be done through bits, and ultimately what kind of experimentation that might unlock.
If you like this episode, I have a feeling you'll also like our recently published and first ever American Dynamism 50 list, a list of 50 companies building in the national interests that embody the ethos of American dynamism. And guess who's on that list? Patreon. You can find it at our homepage at a16c.com or at a16c.com slash American-Dynamism-50. Enjoy!
Disclaimer: The content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any a16z fund. For more details, please see a16z.com slash disclosures.
Chris, welcome to the show.
Guest
Thank you for having me.
Host
All right, let's start with a really simple definition. What is advanced manufacturing?
Guest
Advanced manufacturing is an industry term that generally covers the complex and high precision side of all industries, which is usually semiconductor, aerospace, defense. Basically, anything that you think of from like the Jetsons flying car future is generally bucketed as advanced manufacturing.
Host
I think it's interesting that this term "advanced manufacturing" covers an industry which isn't always that advanced, or maybe hasn't kept up with the time. So could you give a couple examples of ways that this space of advanced manufacturing is maybe not the reality that people might expect?
Guest
On the defense side is a great example. Everyone sees defense primes coming out with drone programs or advanced fighter jets and stuff, and really flashy products and marketing. Wow, everything works perfectly. But under the hood, once you get below the assembly level, someone making a fighter jet – really what they're doing is outsourcing all the components. The wing, the engine, everything, to a tiered layer of 10 mini primes, then 20 tier 3 suppliers, and then literally 10,000 small suppliers dotted across the country that are doing everything from a machine component to the circuit board that is one part of the fire control system. That entire layer is basically chaos.
The B-2 bomber, which is one of our most strategic assets in terms of the nuclear program, about a year ago, the government had to issue an RFQ for some of the parts on the B-2. Not just to replace the parts, but because the guy that designed it or engineered it retired and they didn't have it documented anywhere. There's literally no manual or document of how to make that part. So the government had to go out and say, "Hey, we need someone to come up and reverse engineer this entire thing." So there's stuff like that that's really scary.
Another example: we shipped the Ukrainians what we thought was three years of inventory of Stinger and Javelin missiles, which are shoulder mounted, really cheap missiles that can take out a tank. So for providing that to a human force in Taiwan or the Ukraine, it's exactly what we want to be sending people to deter invasions. We shipped them three years worth of inventory. The Ukrainians blew through it in three weeks. And then our AT&T CEO basically came out and said, "Hey, by the administration, you want more Stingers and Javelins. But one, it'll take us a year or two to spin up manufacturing. Secondly, we don't know how to make some of the components anymore."
So basically, below that veneer of product, everything else is a complete disaster. And I think because of the flashy products and you're making a really incredible fighter jet or drone, people think it's super advanced under the hood. In reality, it's a bunch of people in garages making components that look like they're out of *The Fast and the Furious 2*. It's all duct taped and spit.
Host
Yeah, I mean, you use the word "insane," but it really does surprise me to hear that there are these examples. And it also sounds like these examples are not unique. I heard in a couple other interviews you mentioning that 50% of F-16s are grounded because they don't have the requisite parts. Or I think you gave another example of the International Space Station, something happening to it, a hatch getting jammed, and then a similar phenomena happening where they had to go and hunt down a piece of paper. Can you tell that story?
Guest
The F-16 is the scariest. Basically, if you look at the DoD stats, more than 50% of the F-16s are grounded. And it's mostly because they can't maintain them, and it's mostly because they can't get parts. What ends up happening is they end up cannibalizing other planes for parts. And that makes the problem worse and worse and worse, because of this advanced manufacturing supply chain issue. This is not a COVID issue. It's not an inflation issue. It's been going on for years. It's just now that it's coming to the forefront as more and more people are realizing how important defense was and is.
The space station one is hilarious. I'm going to get some of the details wrong here, so if anyone from NASA is listening, I apologize. Basically, the hatch at the space station gets jammed. And astronauts want to know how much force they can apply to unjam the hatch. Because if they apply too much force and they snap it, we're screwed. So they ask NASA, "What's the spec for this hatch component?" And it takes them several days to figure it out because there's no record of that part. Eventually they find that the guy that made it 20 years ago is retired. And as the story goes, they find the drawing in his desk drawer from the paperwork that they managed to save when the business got shut down. And that was how we knew how much force to apply to the bay door to get it open.
For something as advanced and symbolic as the International Space Station, that is the ground floor reality of how the supply chain works. It's completely bonkers.
Host
That is bonkers. But this, as you mentioned, didn't happen overnight, right? We didn't wake up one day and have everything broken. So what's happened in the last couple of decades? If we trace back all the way to when some of these things were being created decades ago during the first space race, what happened between now and then?
Guest
In the US, we really won World War II because of manufacturing and logistics, not because we necessarily had the shiniest missile. What ended up happening is we had a really good aviation industry, a really strong automotive industry, and a couple of strong core manufacturing industries that could very quickly retool to missile production or fighter jet production.
After that, through the financialization of everything in the late 70s and the 80s, basically the corporate incentive outsourced everything to lower cost countries like China, where we sacrificed robustness for profit and loss optimization. Through that, we drove a lot of manufacturing out of the country. And what people don't realize is that when you do that, you don't just lose that base layer manufacturing of the cheap stuff. You lose the skill set and the culture that people know how to manufacture things. When we need to retool it to semiconductor, or we need to retool it to defense production or space production, that talent base and that capacity just simply doesn't exist in the system anymore.
Part of the reason is cost-plus manufacturing. Everyone's been so fat and happy in defense for a long time that it flows through the supply base. If the government can't really do anything about it, if you're shipping fighter jets late, does the supply chain really care about it at a base layer? Culturally, the entire industry is tooled up around 50% of our promises we break all the time. Once that sets in and becomes the accepted norm, it's really hard to turn that needle another way, both culturally and systematically.
Then over the last 30 years, because we haven't really had to fight a great power competitor, we've only been going up against people in the hills or some really small countries, basically warlord bands. None of that system has really been tested because we can just show up and look scary and people run for the hills. So none of the stresses in the system have been revealed until now, where it's really started to break down rapidly.
It's a bit like if you don't maintain your personal health for 10 years. You might seem fine on the surface, but until you go try and play a game of touch football with your college buddies and you realize that your knee is completely fucked. It's the same type of situation. We think we're fine, and it takes 10 years to find out we're not fine. But by that point, the rust is kind of set into the system, and we're in a really dangerous strategic position because of that.
Host
Yeah, I think the fitness analogy is a great one because everyone can relate to that idea where they're like, "Oh, I'm totally in shape," and they try sprinting 100 meters. Even that is a shock to the system. So it sounds like we haven't kept up. But I think there's also something you've talked about, which is that the people who were involved several decades ago are also starting to retire. So what does that look like? How many companies are actually involved in this space of advanced manufacturing? And what are we seeing there in terms of that almost lapsing?
Guest
There's something like 20 to 30,000 small businesses that make up the majority of the defense industrial base. To frame this before I go into the detail: if you're making a fighter jet, you're actually outsourcing most of the assemblies, like a wing or an engine, to another company like Pratt and Whitney. Then Pratt and Whitney has a network of 2,000 suppliers that give them the components. They might do some engineering, and then what these companies are actually doing is taking the Lego kits and assembling them, and then throwing the assembly up another layer. And Lockheed finally stitches the whole thing together.
There's about 30,000 of these small manufacturers that in aggregate make up the entire defense industrial base, but individually they are small businesses. Maybe they have 10 to 20 million in revenue max, and there's maybe 15 or 20 employees. The danger there is because we outsourced manufacturing and because we culturally decided that if you don't have a college degree, you're worthless to society, there's been no new entrance to the manufacturing workforce. It's not sexy and it's not rewarded in the culture. And because we've had no shocks to the system, no one's realized what a big problem it is.
That entire band of owner-operators at that 20,000 to 30,000 small business level – the average age of a worker is like 55. The average age of an owner-operator who is a small business owner that is actually needed to function in the business is 62. If you switch back to the first space race or the Cold War, there were a bunch of 30-year-olds that went inside their businesses. And now 30 years later, they're 60, 62. And because we're not rewarding culturally being in manufacturing, all those people were relatively successful, so their sons and daughters got a college degree. They don't want to take over the business.
We have this problem where the entire defense and space and semiconductor industrial base is on this house of cards of 62-year-olds that, through no fault of their own, are retiring. There's no one to take over those businesses. And also, the knowledge that is used to make those parts is not in some GitHub where they can pass it on to Lockheed and say, "Hey, someone else can go make this part." It may or may not be written down or recorded in their business. It's mostly in some two or three people's heads that have been making that thing for 10 years. It is complete art.
Regardless of capital or talent, the IP or the knowledge of how to make that component is not even fungible or transferable. So you have the situation where it's not just a matter of throwing a billion dollars at the problem and scaling some new factories. It's like, no, literally there's two guys in the country that know how to make this turbine blade for this engine component, and they're all retiring. Once they retire, good luck. Now there's six months of reverse engineering just to figure it out.
Often this is on specialized equipment. One of the Northrop suppliers, this was a couple of years ago, but there are many examples, were trying to move the production of one part from one facility to another. It was in a different state. Part of making this part is to coat it with a coating that makes the component work how it's meant to work. It's a chemical process. They did all the manufacturing engineering, rehoused it in this new facility in a different state, and they couldn't get it to work for like a year. They finally figured out that there was a mineral content of the water in the different state that was slightly different and undocumented. The water as part of the process was different. They literally couldn't figure it out because you've got these minor differences in the localized environment, which can all be controlled for. It is not that scary on an individual basis. But in an aggregate, when you realize it was like $100 billion a year of these parts being produced, and it's all kind of crazy duct tape accidents that it works at all – that's the situation that we're in as a country.
Host
Yeah, it almost reminds me of those memes that are like "expectations versus reality." Expectations are like you have this image of a SpaceX rocket taking off and you're like, "Oh my gosh, look how far we've come." And then the reality behind it is these people about to retire with desktop drawings in their drawers, or in this case, a specific mineral content in the water that's changing their ability to produce these products.
I've actually heard you use the term "dangerous" when you speak to the point that we're at in terms of this pipeline and this particular space of advanced manufacturing. If someone's listening, they might be like, "OK, a bunch of people are retiring. Some people are not very keen on the idea of us continuing to pursue space." But what would you say to the average person? What's at stake here? What are we going to lose if these people retire and we don't have these things documented?
Guest
Unless we solve this problem, I think the country and our way of life is at existential risk. The analogy I give to people is: if you're living in a small town, we've built up 200 layers of abstractions in society so that you can be an artist, or a painter, or in finance, or in crypto, or making video games. The reality of the world is that 200 years ago, we were killing each other over food. It's a miracle in the first place that we're here and that society is relatively stable, that the roads get picked up.
People forget because they grew up in America that it's so successful that they don't have to worry about having a bulletproof car. Otherwise, if you have more than $100,000 net worth, someone in the gang might try and steal your daughter. For the rest of the world, that is a reality. But because America is so isolated culturally, most Americans' view of geopolitics is "Russia bad." You don't learn those lessons. As a younger person, you don't realize that unless we keep the roads paved, this can all fall over very, very quickly.
If you run the scenario: right now we are successful because we are the world's police. Whether we should play that role or not is obviously up for debate, but the reality is that we are fine culturally because everyone understands that if you fuck with America, we will put a missile over your head and you're dead. Or if we go into a great power conflict, we have enough logistics and infrastructure to go and win that conflict, or at least be scary enough that conflict never exists in the first place.
The analogy I like to tell people is: bar fights happen when both people mispredict their ability to win the fight. Bar fights don't happen for two reasons. One is there's two UFC fighters staring down each other and they both know the cost of a conflict and the other person is scary. So that fight never happens. Or there's a bunch of morons, but there's a bouncer who is big and scary enough that the conflict never happens in the first place. That construct of a bar fight relies on impressions and social trade-offs. Enough people have seen their friend getting beaten up by a bouncer, so I'm probably not going to even test that assumption.
The reality, of course, is that most police officers and most bouncers are probably incompetent, and can probably get taken out by someone relatively competent as a civilian. But we have enough social construct around the concept that it's a really dumb idea that no one wants to take the risk. In Defense Land, I describe it as a "lethality mirage." A lethality mirage is basically everyone else's impression of you is that you're a 10 out of 10 lethal, so they absolutely are not going to fuck with you. Then maybe you lose one small conflict and someone goes, "Well, hold on. Maybe these guys aren't so scary as we thought they were." In reality, I think we're about a 3 out of 10. The real danger comes when a great power competitor finds out before you find out that you're actually a 3 out of 10.
The problem with everyone thinking advanced manufacturing is in a really good place is that we don't go fix the problem. Culturally, everyone thinks it's fine. The roads are getting paved, fighter jets get made, we go into a conflict, we're fine. In reality, we're probably so far away from doing that. If we have one conflict with China where we expend most of our ballistics inventory, we might not be able to remake it for like five years. Then we're basically standing around with our hands tied behind our backs.
When there are not one or two great powers in the globe keeping the peace, that's when you get fragmentations. You get despots and dictators pushing the boundaries of what can be done in their country. They see weakness and then they start moving into invasions of other countries, and we get back to this kind of multipolar fragmented world. So it really is quite serious. To go from the state that we are currently in, the tail end of Pax Americana and Peace Through Strength, to a highly conflicted world is a really scary position. It's not going to be a matter of "oh, we go to war." It's "oh, we go to war and we lose." And also, we don't have food. And the American consumer can't buy an iPhone for less than $10,000 or have laptops at all.
Imagine if GPS went down. All the airlines are on GPS, all the train systems. Our ability to coordinate military conflict relies on GPS. There are less than 50 GPS satellites, and all of those can be shut out of the air pretty easily. All it's preventing some adversary from doing that and basically collapsing Western civilization in one fell swoop, taking out less than 10 satellites for the cost of maybe $200 million in lasers or hypersonic missiles, is the fear that the response from America is so great that we will win any conflict that's put ahead of us. What our adversaries are starting to realize faster than we are realizing is that our lethality is more like a 3 out of 10 versus an 8 or a 10 out of 10.
As VCs know, risk happens slowly and then all at once. By the time you realize the risk and it starts cascading, it's kind of too late to respond. The difference between this time around and the last time around was, yes, risk happened slowly and then very quickly in World War II. But at least we had the fundamental culture and infrastructure of a defense industrial base to be able to go, "We're making great cars, so let's shift that to making fighter jet chassis." And we're kind of okay. The response time might be 18 months to two years. When you don't have that fundamental infrastructure, the response time can be 10 years. You're in a really serious position where you're caught with your hands behind your back and you can't respond, and everyone knows you can't respond. Then basically everyone can run around doing whatever they want. That's how you get the collapse of society.
As soon as someone realizes the police force are incompetent, you get a bunch of bad actors that creep through the system. That's basically the state that we're at, but not just at a localized bar fight level, but how the globe operates in general. So yeah, it's incredibly scary. I think people are going to get a really hard wake up call in the next two years.
Host
I think you brought up a lot of good points there. One of them is this idea of space that many people think is a frivolous endeavor. But what happens in space directly impacts what happens on Earth. You mentioned GPS, but it's also weather, agriculture, us understanding climate change. All of this happens within space. But I think you've also brought up the defense side. And then even if people aren't interested in either of those, there is the direct relationship to manufacturing that happens with iPhones, but also medical devices, semiconductors, which are in so many of the devices that we use. This is all part of advanced manufacturing.
I think it's important for people to realize that even if they don't believe in this idea of American dynamism, which we obviously do, there are direct correlations to their everyday life that will happen if we don't have this infrastructure.
Guest
Even if you're an EA type of person, or a pacifist, or into climate change, pick any industry that you think is important. Good luck running compute on GPT-3 if there are no chips. Or China takes Taiwan and then the only chips you can buy have a bug in them. It's not that hard to draw a line. It's very serious.
Host
I want to understand this idea of onshoring or bringing some of this technology back to the system in the United States. What I want to understand here is it sounds like there is a delta between where we are and where we want to be. How much of that delta is broken down into the investment in the space versus the talent available or the technology that we have available? Or is it just a matter of time for us to catch up? Obviously, it's a multivariate equation. But what would you say are the most important drivers or maybe the things that are underrated by people? Because I would imagine that some folks might just say, "Okay, let's just pour a trillion dollars into this. That'll fix the problem." But will it?
Guest
It won't. The most important bit is talent. If you think about software engineering, and you want to solve the problem of how can we produce more data scientists or people that are capable of working at OpenAI, which is arguably the top tier of software engineering, the way that you get more of those people is having this base layer of application developers, front and back end engineers, more hardcore back end engineers, SREs, and data scientists. Maybe you have a million people and you produce a thousand really incredible people in these domain spaces. Knowledge stacks on each other.
You have this talent competition. You produce amazing people in these fields. The US doesn't have the track team that goes and wins the Olympics unless we have thousands of colleges that are really competitive for track and field. That's how it goes. To take that sports example, you can't just start from a cold start and say, "We want to be competitive. We want to go below a trillion dollars and win the Olympics." It doesn't work like that. That takes years and years of training. You need to get people when they're 14. You need to train them up. It's a cultural thing. Why would a kid want to go do sports in the first place? In America, we have that problem solved because America loves sports like the Romans did.
You think about that from a manufacturing talent level. You want to go solve rocketry, or defense, or hypersonics, or semiconductor. It's not a capital problem. It's a "where are the people that are going to go engineer that system and then run that system?" The reality is they don't exist. The impact of that is time.
Look at the semiconductor problem. How fast can we reshore semiconductor? The Biden administration gives Intel $50 billion. That's arguably the hardest, the heavyweight division of manufacturing. The talent doesn't exist. You can't go and hire 20,000 people, or even a thousand people, that know how to go and execute that construction, how to go engineer that factory, how to go run that factory. They simply don't exist. You cannot simply go and train someone up in even three years to go from junior college track and field to winning the Olympics. It doesn't work like that. You need the coaches, you need the entire cultural infrastructure.
That's really what we lost when we outsourced manufacturing in the 80s. There is no talent base that is generative of the genius types of people that can actually go and do all these advanced things. That is the thing that people don't realize. It's not a capital problem. It's a time and investment problem. You have to have the base layers there to be able to execute the hard stuff.
Host
Yeah, using the fitness analogy, it's like taking the Olympians from 30 years ago and saying, "We're just going to toss them back in and compete in the next Olympics." It doesn't really work that way. Or like, "Hey, we've got these kids and they kind of fit. And by the way, we've got this training program and in two years they'll be winning the Olympics." It doesn't work like that.
Guest
Right, definitely.
Host
Okay, so it sounds like what Hadrian is doing to solve this problem is a mix between fixing the talent pipeline and also using technology. So let's return to the talent, but how are you using technology to fix this problem?
So the way we look at it is there's like five or six core parts of high precision machining. Some are lower skilled than others, but they're all pretty highly skilled. So what we are doing is basically at each part of the factory, developing software, basically grabbing the last smart people in each of those domains that exist and pairing them with the best software engineers that we can find saying, go automate as much as possible that part of the skill. And in reality, you can only automate 60 to 80% of that. Some of it is just like the technology curve doesn't exist. You're waiting for 10 years of machine vision to catch up.
So the job of a startup is to industrialize research or existing technology and apply it to a new domain. It's not going to sit there with 40 PhDs and try and create something that's like high risk. That is like the science pipeline. So we grab and integrate whatever we can and build software around those talent bases to remove the boring stuff, automate what we can. And then the last 20 or 30% where it's impossible to automate, we build software that is highly process driven, where we can throw someone new to that role in there. Within 30 to 90 days, depending on the role, they can be trained up into that and have enough scaffolding that if they just follow the instructions, they will get 90% of the way there.
For some parts of high-precision machining, that is way harder than others. Basically, we're using technology to solve the talent problem. And then when we can't use technology, we have to simplify it so that you can hire in someone who has had no aerospace experience and then making flight hardware for a rocket program in 30 to 60 days. Because for us to scale out to replace the 50 billion, so let's argue that we can get to 10 billion, we can't go and hire a million machinists. We have to automate our way so that we can go train 100,000. And that is barely doable, but it has to be a combination of solving the talent pipeline, using automation to basically lower the burden of that talent pipeline problem, and then focusing on the areas where talent is scarce or harder to come by.
as well as attracting as much as possible entrance to the new workforce, which is why branding and marketing is so important and speaking about this problem over and over and over again because people don't know it's a problem and they realize like, oh, I'm a software engineer at Google. I can actually contribute here or I work in hospitality and I can actually get in the fight and help solve this problem is really important from that aspect as well. And that's why the cultural reward system where manufacturing is not really sexy is a huge problem because unless it is going to sound dumb, But unless some 19-year-old from UCLA can go basically convince some hot girl at a bar that manufacturing is cool, then why would they ever go shoot for manufacturing versus going to work for Google or Goldman Sachs?
Just as I keep going back to the cultural challenge. Because unless it's awesome and celebrated, that's a really, really hard problem to solve.
Yeah. I actually met one of your newly trained or currently trained machinists who used to be a copywriter. Everest? I think last week at LA Tech Week.
Guest
Yeah, Everest was awesome.
Host
Yeah. And so it is a fascinating process that you guys are undertaking to take people who aren't within the sphere and training them up to be machinists. And something I've heard you talk about before is this maybe counterintuitive idea that to actually take the existing systems, you need to simplify them in order to enable these new jobs, right? Because you can't train a previous copywriter to be the machinist from 30 years ago that can do everything that that person did. But if you do simplify those processes, that opens up the job pipeline for many of these people to become that machinist.
Guest
Totally. And this is something that software engineering has done really well, where there's like a clear delineation between SRE, DevOps, data science, front end, back end, and different tiers within that. And traditionally, machining has often been there's one guy that does 90% of that whole pipeline themselves. So there's no clear snap-off points where there's a degradation of skill where you can put someone here and then they learn the next run. It's all one person solo operating the entire thing, which just doesn't scale.
Host
How do you convince the existing machinist to transfer this knowledge? I'd imagine there's some sort of incentive that you need to put into place because, I mean, maybe I'm wrong about this, but I imagine they might feel like their jobs are being automated away or that they're being replaced in some way. And of course, some of them, it sounds like they're planning to retire. But how do you actually convince the person who has all this knowledge to want to share it and be part of this new system?
Guest
I think there's a couple of things. The first one is the smartest and best people know that the industry is due for a transformation and that it's inevitable that someone's going to come along and do this and they want to be part of the winning team, not left behind. So that's a big part of it. Yeah, incentives really matter. And I think culturally in manufacturing, you don't get many opportunities for growth. And there's a bunch of really smart people out there who've been kind of siloed and not given the opportunity to stretch and grow. And I think culturally, they understand that if we're automating one part of the factory, we're not going to suddenly fire 20 people.
No, no, no. Go pick whatever else you want to do. Manufacturing engineering or CMM or a higher skill level of CAM programming. And I think the other thing is, yeah, incentives really matter. And I think we're probably the only factory on the planet outside of SpaceX that is giving everyone in these roles the same tier of equity as the best software engineers have. every technician in Hadrian, whether you're packing boxes or whether you're an absolute master of your trade, and maybe you're one of the only 20 people left that can do it, is being given the opportunity to have probably the only opportunity they've ever had to create generational wealth for them and their families.
And with that, you're an owner of the company, and then you're highly incentivized to contribute to the system instead of being like, well, I'm on $40 an hour, and this guy's on $20 an hour, and I'm on $40 an hour because I have this tribal knowledge why why would i go and share it because that's what you know that's the incentive and that's why like oddly there is no like stack overflow for machining because i think it's basically a function of the fact that manufacturing has been an hourly workforce for such a long time even the highest skill positions because then you get this like weird localized competitive set where even within your facility you're not really wanting to train or contribute to people because you're putting your own job at risk.
Whereas in software engineering, your salary, you're in high demand. It's more about sharing knowledge and going as fast as possible. And I think, yeah, that all comes down to culture. But I think that actually gets driven off like historical being an hourly position for such a long time with no meaningful equity upside. Because then why would you share your knowledge if you're just going to train someone to replace you or the owner of the crappy factory that yells at you every day of why the part is late is no longer incentivized to give you a higher hourly wage, you know?
Host
Yeah, totally. And obviously there's positives and negatives to each of those cultures, but something I've heard you talk about before is you have all of those different roles sit together at lunch or actually meld together in a way so that there isn't like the software engineers here and then the new machine is here and the old machine is elsewhere. And so can you speak a little bit to that and what you're seeing in terms of what aspects of each culture are being brought together and highlighted?
Guest
Yeah, I think, and again, we're not perfect. We're probably the only ones doing it at all, but many, many, many challenges. I think the first one is making sure that the software engineering team understands that the person that's operating is the golden goose and they are the customer. And changing that mindset from, oh, there's this internal team that does some function. It's like the customer service team at a SaaS startup, like who cares, whatever. But if the SaaS customer screams and says, we need this feature, like everyone knows to go and jump. So making sure we understand that our internal team is the customer of what we're building.
And if they have one hour of downtime, it's a really serious problem. And also their happiness as a user is a really serious problem. And that creates this interesting dynamic where you've got technicians that are on hourly wage or whatever, and there's a software engineer from Stripe running around being like, oh shit, this guy thinks my product is terrible. This is a really serious issue. And that means a lot. The big problem to solve is continually reinforcing it. Everyone's working hard and there are different challenges. And I think people not from the software domain are like, why are you building this feature?
And it's like, it's six months of two software engineers. And yes, I understand it's painful, but continually melding those communication streams together and letting people understand each other's worlds is really important. I mean, the ultimate Hadrian employee would be like a person filling out quality paperwork who was also always a software engineer so that like the localized pain, they would just go and fix or vice versa. So yeah, there are many, many challenges to getting that right. But the important thing is to like, have some shared pain both ways and then make sure that the interaction is like, operating gives feedback, software engineer picks it up.
But the really difficult part is like, this is really painful for you, but it's also six months of software engineering. So it's getting deprioritized. but how do you make sure that is actually understood and communicated throughout a rapidly growing organization is an incredibly hard problem. And even really simple things like making sure software engineers are out on the factory floor as much as humanly possible, sitting behind operators and even doing their jobs where possible is incredibly important culturally and also just for decentralizing the product learning. I mean, there's this classic trap of you think the product's good and it's not, and then whatever feedback mechanism you've got is imperfect.
So trying to get it down to single-threaded, like the person that's building the automation for this part of the product is capable of actually doing that job and is forced to do that job every couple of weeks and relearn the real user pain is an enormous challenge, but it's probably the most important thing that we can be doing to maintain that culture at scale as we're rapidly growing.
Host
Yeah, I mean, you hear of some companies doing this, like if you work at Airbnb, you're encouraged to stay at Airbnb and understand the product. But this is obviously a different level of that. Something that stood out to me, though, is you speaking to obviously, this is a really hard problem to solve. And I think, you know, people say that anything that integrates into some form of hardware, not just software and bytes, is like playing on hard mode. And I can speak to that as in I've mostly worked in the arena of bytes, and even within the marketing sphere. And so within that sphere, you can just kind of A-B test the shit out of everything. If you're unsure, you test it, right? If you have the right amount of traffic. With hardware, you can't really do that. And so I'm curious to know, as you're building this company, how are you deciding what bets to take? And also, what signals are you looking for to make sure that you're on the right track? Because you don't have millions of page views that you can just test to get the signal from that noise.
Guest
Yeah, that's a really good question. So there are like very obvious projects that from the outset everyone knows are the biggest points of pain, whether it's time or annoyance or whatever it is. And you can't sit in a room for three years and try and build an automated factory because you only really know what actually matters and what doesn't matter at all is by running the factory and then being very prepared to rip up your roadmap and realize that what you've been doing for the last year was an assumption and not reality. So yeah, you can observe a lot of pain, you can observe where bottlenecks appear that you didn't expect them to.
And there are two layers to that. One is like a localized layer, like, okay, what is causing this team this amount of downtime or pain or effort? But then the factory level, what is producing costs outside those localized things? Because in the real world, you can have a highly optimized process over here. But if it's spitting out unclear results, or there's some stochastic variability, then very quickly, the factory can go from smooth to chaos. and all of a sudden 30% of the people in their downstream roles are dealing with the poor outputs of this other thing. So constantly going up and down those layers and making sure each team has the ability to screen for what they need, but then also having observability of the factory itself and going like, yeah, you guys think the cost is here, and it is, but we're not solving it here.
We're going to solve it here because that was actually the original root cause of why this issue came across here. And some of that is system. Some of that is just really good foundational operational leadership and being intellectually honest enough to go like, yeah, this is not like a paperwork problem. This is because a customer did this and we need to build a product around here because that's actually what prevents all these downstream issues. I think the other problem that people don't realize when you were talking about marketing data, no factory on the planet has observability. So you can't simply...
And maybe Foxconn does, but you can't simply throw open a dashboard and go, what is my uptime? What are the costs for this job? How much labor did we spend on it? It all kind of swags and fuzzy mats and people standing around trying to do time studies which have their own psychological flaws. So a big part of the technology that we building is a factory data platform where all of the people and all of the hardware are hooked up to one internal data service. And it took us like a year to build this and we're still three months away from having it operationalized so that everyone in the automation team and operations can look at it and go, oh shit, we actually thought we were really efficient over here, but we're not.
And then it becomes much more clear where the costs and the systems are. But even getting to that point is a bunch of software engineering that no one has ever done purely to solve the observability problem. And it becomes much more obvious to everybody where to point resources and where the pain is coming from.
Host
Yeah, that's fascinating. Because, again, bringing my experience from marketing, there's so many times where you think something is an issue or some article is going to work or some landing page copy is going to be best and you're wrong all the time. And so if you really don't have that data layer. It's so hard to tell what's truly going on. And I'm curious to know how that impacts the end customer, right? So if you're a SpaceX or an Android buying parts through Hadrian, do they get access to that data? And if so, how does that impact that relationship compared to what they're used to experiencing?
Guest
So those are two good assumptions of potential Hadrian customers. Secondly, in terms of the data, what all customers want is transparency and observability. So we had the CEO of a large defense prime in last week. His number one problem for their supply chain team is observability. And there's this metric that they use internally, which is average number of days before the part is meant to land their shipping and receiving facility that they get notified it's going to be late. And the average is one. So out of a 6 to 12-week purchase order or something like that in terms of lead time, the average number of days that they're being notified that something's potentially off track is basically the day that it's due.
And this is a huge cultural problem. It's a huge observability problem. So yeah, what customers want is effectively DevOps-level visibility that's not muddied by human aggressiveness or, hey, we'll get the job done, don't worry. We'll pull it out of the fire. So the world's best Gantt chart, basically. To build that, though, you need the entire factory data system to get that data because if you just have... The Flexport strategy in the early days was build the customer portal first and then have a bunch of analysts just typing in shipping and receiving data. You can't do that in manufacturing because it's so complicated and you have all these cultural issues of people are always going to say, I'm going to get the job done.
And that's not because they're lying. It's because they got white line fever, they're going to get the job done. We're a couple of weeks-ish away from launching this to customers because it took so long to build that internal scaffolding to give people the data. But yes, even now we're sending just Excel spreadsheet summaries of like, okay, here's where we're at in this production run. And people are literally responding like, this is amazing. No one's ever given this data or whatever, which is from like software land is completely insane. It would be like AWS not having a dashboard of like, what's your SLA uptime, which is insane.
That's a really hard problem to solve. That's what customers want is like, no, this is the platform, it's the service, and we have observability. And then when things do go wrong, you're learning about it an hour after it goes wrong, not six weeks down the track when we have to face up to the problem when we can't ship parts, which seems crazy from right on outside the industry, but it's like a magic wand to everybody else in the industry.
Host
Yeah, I mean, even extrapolating past advanced manufacturing, I think it's just like a cultural phenomenon that we expect these updates. Like even if you talk about like ordering from Amazon, people have come to expect my package is going to arrive on this date. And if it's not, I'm going to get a notification and only some small fraction of packages get delayed. And of course, in-bets, manufacturing is much more complex. There's many more reasons why you can't have that level of precision. But it is, I think, like, again, a cultural phenomena that you have come to expect that kind of system.
And it's interesting that this is now being implemented within advanced manufacturing. I'm curious to know, within the spectrum of Hadrian and your customers, with this implemented, let's say you do get to the point where you have that precision update, what happens? The obvious thing is like, okay, a rocket gets shipped on time. But how does that actually influence the wider industry? Does this mean that because things are being shipped on time, companies have better margins because they operate better and therefore more companies can enter the space? Or can you speak to the downstream or you could say second, third order effects of actually having that system in place?
Guest
Yeah. So if you skip a couple of years ahead and everything's being done through Hadrian, whether we're building it first party, but either everything is being shipped on time or you've got forward warning where something's going wrong, two big things happen. One is something like 50% of the total cost of a product, like a satellite, for example, is through delays or inventory levels, which you shouldn't need to have. So as an example, if you think about a gap chart to build a satellite and it takes 90 days and you do this piece first and this piece first and all the parts have to arrive on time.
Firstly, because the supply chain is super unreliable, no one is building that gap chart super tightly. They're building it with about 50% of slack in it because you can't rely on those inputs coming in at the right time. So once we get customers to a point where they know that if we say, here is the date, it's reliable, everyone can compress their schedules around that reliability. And because manufacturing time is payroll costs, it's working capital of inventory, it's a build rate. If you just compress time in manufacturing, you just win on cost. So I think that most of the products will be able to drop their manufacturing costs by like 30% to 50%, not because of cheaper parts, but because of the reliability and the ability to compress schedule around that reliability.
The second point is Ford notice of errors. So if you have a manufacturing line down because some vendor was giving you a bunch of satellite parts and you expected them on Monday, and then you find out on Monday, they're not going to come for two weeks, which happens like 30% to 40% of the purchase orders, that dynamic happens. Then you've literally got people standing around, they can't do anything, which is a huge cost in and unto itself. and another CEO I was talking to last week. Basically, and the way Aerospace works is you have an order book and then you can't book revenue until you actually deliver your satellite or your product or whatever.
And they would have had an extra billion dollars in revenue last quarter had more of their parts been delivered on time. Growth is slowing because they can't meet their order book, not because they're not working hard because the supply chain is a complete mess. The example I would give is like, if you're a back-end software engineer and you weren't really sure whether you could spin up a new AC2 instance or not. And it might take three months, it might take a week, you don't know. So all of your sprint planning is out the door. All of your like, can I get this product out to the customer and compete is out the door.
And potentially you're sitting around as a backend software engineer, like twiddling your thumbs for three weeks because you're blocked on this core piece of infrastructure. So imagine if there was no supply chain team and all these smart people could be doing more value-added activity versus calling Bob's machine shop three times a week, being like, Bob, have you put the parts on the machine yet? And getting lied to. Or the aerospace engineers themselves, not waiting for parts for six weeks and being able to have parts every two weeks so that they can make a test satellite and blow it up and then iterate really, really fast.
When I say we can let companies move 10x faster and lower their cost of manufacturing by 50%, I don't mean because we're cheaper and we are cheaper and we can be, But the main benefit is that you have this high reliability infrastructure layer so you can compress schedule around it. And that's the huge win is you can make a fighter jet in six months, not 12 months. And that's what makes it $150 million, not $300 million.
Host
I'm coming back to this meme of expectations versus reality. And you imagine these very, very talented machinists and you picture what their job is from the outside. And really, it's calling Bob's Machine Shop three times a week being like, where's my part? So I think, yeah, a lot of that is illuminating to realize that the supply chain, of course, is complex, but with the use of technology and training the right people can be simplified. Within Hadrian itself, what would you say the key risks are? So this paints a really, really interesting picture of what we can do to solve the problem. But if Hadrian were to fail, why would that be? Would it be financing? Would it be a miscalculation on how complex some of these systems are? What do you think some of the key risks are?
Guest
Yeah, I think the broad risk of Hadrian is that it's just unbelievably complicated. And if one piece doesn't work, then the whole thing doesn't work. So the execution, it's a complex coordination problem to take a teal. I mean, I think the second problem is as the business grows, the reality, the truth seeking of like, is this automation possible within X timeframe and how we're planning and kind of around that is really, really hard. because you have to make sure that people are being realistic around, are we actually saving time with this? Or did we just build scaffolding for this process? Which is totally fine.
You can't really build automation until you have the scaffolding in the process. But recruiting cycle of someone highly skilled, it might take three months. And as you're growing revenue at an exponential monthly rate, and you're booking customer sales on your ability to deliver that, and you're expecting the automation curve to go here and intersect your hiring plan here, and the automation curve turns out to be here and you have this gap, it's not as simple as, hey, we're breaking even on this part of the project. Let's just put more labor into it for another 12 months because you can't hire into that gap for three months and then you're behind the eight ball and then customers hate you.
So the prediction of when projects are going to land and when things may or may not be industrialized as a new piece of automation or process is really, really hard. And then I think the third big risk is in manufacturing, if you have a bad process, and a bad process maybe is something that generates an error more than 1 out of 100 times, that's not a mistake. It's like, okay, now this has downstream impacts on schedule, or this machine is down, or whatever. If you scale before there is a level of error-proofing that lets you scale, you end up scaling to a point. Something that's happening once a week now happens 100 times a week, and there's so much negative work in the system that, yes, not only do you make no money, but then you're chasing your tail, and now you're behind on everything, and you start this really dangerous negative spiral of negative work.
And I think having the discipline to understand where that point is and continually say no to customers, even though they really want stuff, which is really exciting. And I want to grow a massive business, so I also want to scale as fast as possible. But if you go too fast before you're really ready to scale and have that level of error reduction, you can basically blow yourself up and you won't even know what's happening until the last minute. once you've told customers you're ready to go and you're ready to scale, and then you blow up a production order and you break the promise that, hey, we won't ever have a launch delay because of us, then you lose the magic.
And startups can basically survive infinitely as long as you hold the magic. So that is something that I think about a lot is when is the right time to push those buttons and what level of scaling can we do now versus how do we need to be disciplined around error reduction versus shooting ourselves in the foot?
Host
Yeah. I mean, every company deals with some amount of technical debt, But you're right that hitting that scale button too soon can really exponentially increase that amount. One of the things that I think you explored originally when you discovered this problem was, okay, a bunch of people are retiring. These machine shops may go out of business. Why don't I just acquire them? Why don't I do like a private equity play and just acquire these shops instead of doing what you've done now, which is more so build the technology, train individuals to service them or almost like absorb that intelligence. So can you tell us a little bit more about why you pivoted there and why you didn't go with that original approach?
One of the things that you don't really learn until you're in the weeds is every piece of hardware variability—whether that's a different machine, a different cutting tool, a different vice, or piece of work holding—every new combination of those things that you add to the search space of the problem has a non-linear impact on the software engineering complexity you need to build to get to a level of automation that actually solves the problem, versus just being 20% better than your competitors so you have a great private equity roll-up.
Guest
It wasn't until I was in the weeds of that problem that I realized most legacy machine shops have one of every machine ever made. Even within 20 machines, five might be the same, but even those five are set up a completely different way. That adds enormous nonlinear complexity in terms of the software automation you have to build on top of those systems—to the point where it's close enough to impossible that you can call it impossible. You can squeeze out some margin improvement, but you would never get it to the point where it was a scalable system or repeatable enough to solve the problem at the multi-billion-dollar level I want to solve it at.
That was the core reason why you have to build this from scratch—hardware hand-in-hand with software hand-in-hand with processes. You make localized trade-offs, standardize the physical world to abstract it into software, and then abstract that into process. Now that we have the base of that with clear line of sight to what it looks like at scale, we're actually going back into acquisitions. We can reliably say to customers and machine shop owners who want to exit: one, we actually know what we're doing now; two, we have this standardized system. To transfer a legacy machine shop's parts to Hadrian—whether it's a new customer order or an acquisition—there is now a clean funnel for how this can be done.
In terms of training and reskilling, some of the people we acquire might want to stay on for the journey. That's great. We have this integrated system where we can plug them in, they can learn new skills, and they can be part of the winning team. But without building all of that from scratch, you never have that core of what great looks like. Merging people into that starting from an incredibly divergent hardware base and process base is almost impossible to clean up both culturally and systematically. You end up doing 200 hardware integrations instead of three. It's impossible.
Host
The complexity is hard enough within one of those shops. Something I noticed on your jobs page is that you have around 20 open jobs at the moment—it's always changing—but there are so many different types of jobs: data scientists, mechatronics engineers, security officers, salespeople. We talked about the importance of marketing in this space as well. Across the spectrum you're working in, what talent is most needed right now? Are there specific types of folks you wish there was 10x more of?
Guest
Everyone. The reality is it's everybody. If I had to wave my hand, I'd say give me 100 software engineers who worked at a startup, then worked at big tech, but actually have an aerospace engineering degree and did an internship at Boeing. Where we see people go the fastest is when they have high context around what the real problem is, versus having to learn the ropes on what manufacturing or aerospace engineering is. That would be the magic wand, which obviously doesn't exist.
Host
How many people fit that bill?
Guest
Three? Handfuls. The factory talent—yes, we always need more people, but there's no one critical talent base where we're really struggling. In terms of automation, we just need all hands on deck. What people don't realize—and we struggle with in recruiting—is that you don't need a hardware or manufacturing background to come into a business like Hadrian, SpaceX, or Anduril to be a competitive candidate or create value with your software skills. A lot of what we're building looks and smells like enterprise SaaS, except our customer is internal users, not external users.
Even the deep tech side is very close to—if you're a developer who's been working on Unity and building video game geometry engines, that is very close to the software engineering we're doing internally. Integrating with the hardware, if you're the type of software engineer who is okay with reverse engineering some crappy API—anyone from a Rippling, for example, has done that with some crappy payroll API from a vendor that doesn't have it fully documented—that's the exact same problem space. Tons of software engineers need to get it through their head that it's all just regular software engineering problems. Hard problems, don't get me wrong, but you don't need a manufacturing background or an aerospace background to get in the mix and start adding a lot of value very quickly.
Host
I love that you mentioned Rippling because there are so many examples of successful companies that ventured into a surprisingly complex space and just documented and simplified it. I'm a Canadian, and using a product like Rippling is so nice. The idea is that I just click a bunch of buttons and say, "Are you an immigrant alien? Are you not? What state are you in?" and it outlines the process for you. That's a nice parallel because many people have used tools like that before. Imagining a parallel within advanced manufacturing—of course, this stuff is incredibly complex—but if you can simplify it for the end customer to just understand, "Okay, I'm going to get it exactly on these dates," and then internally map out the key steps and insert automation where possible...
Guest
That's a great example. A HR rep doing that without Rippling is 20 to 40 hours. You still need a HR rep to do that job, but it takes them 30 minutes. They're not filling out forms for 40 hours a week and drinking a bottle of wine at night because of it. It's the same here. We're not replacing the human in machining. We're just saying, "Hey, click three buttons, not 20 dumb buttons a bajillion times a week." Click buttons in a nice workflow. Don't have to go read the paper manual on what immigration policy is for Canadian BZ visa holders. It's all workflow logic. It's all a process. It's very gnarly to wrap your hands around that entire problem simultaneously, but that's the game we're playing.
Host
Using that example, to your point, you're not the experts from the space. If I have a question like "What the hell does a resident alien mean?" I go to my lawyer and ask, "How should I respond to this?" But I only involve them when necessary. It's the same with machining. You only bring in the expert machinist when absolutely necessary and simplify the rest.
Guest
Yes. Then the expert machinist can go learn software engineering, or they can go learn harder and harder levels of machining. They can spend 100% of their time on really gnarly problems that you need an expert for—not how to program a threaded hole, which they've done successfully four times a week for the last 10 years. You don't ever have to do that again. Go solve this other problem and figure out how to automate that.
Host
There are a bunch of different problems you can attack within advanced manufacturing: medical, semiconductors, defense, etc. Sounds like Hadrian is focused at least at the moment on space. As I mentioned before, there are many people who think space is a frivolous industry. I'll read you one tweet from someone I saw recently: "No offense, but I 100% think space colonization is a childish desire." Feel free to respond to that. But really, within the sphere of things you could pursue within Hadrian, why specifically space to start?
Guest
The really short version is that Stripe doesn't get to sell to AT&T until they sell to a bunch of Y Combinator startups. There's a huge amount of capital flowing into commercial space, which means there's a bunch of net new spend. These companies are being run by 30-year-olds, not 50-year-olds. You want to sell to startups first and hit that early adopter spectrum. Commercial space has all the complexity of every other industry we would want to serve in the future, so the automation we're building is almost completely transferable. But there's a real need because these companies are trying to go super fast, so they will pay for speed. They're sick of calling Bob's machine shop, but they're young enough from all these rocket and satellite companies that they're willing to take a shot on something new and work with us to develop the system. It's just your classic early adopter problem.
Then holistically, why go to space at all? Why put humanity's resources out there? We could go through the argument that space is a war-fighting domain—you need observability of the planet to stop nuclear launches, so you need satellites. GPS is a great example. We can go through the time-worn arguments that most medical advances on the planet are downstream of NASA and the ISS and how that flows through society. But all of that is an abstraction of why Columbus jumped on a shitty raft to go look at a new island in Polynesia as a tribal leader. In some ways, it's manifest destiny. Humans are built to expand and explore. That is a core drive of the species.
With that argument, it comes down to growth versus degrowth. That is the eternal cultural fight. You can call it capitalism or communism, the state versus the people, or decentralization versus centralization. But what it comes down to is: are we going to use the limited resources on a planet to go get new resources so we can continue this magical species-wide journey of settling the solar system and finding out what's out there? Or are we going to accept the status quo and sit there, continuing to shoot in a hole instead of inventing the toilet?
People who say, "By stretching towards the future, you need to concentrate on the problems of today," don't realize that historically, you solve poverty by building farms, not by trying to optimize a shitty hodgepodge hunter-gatherer system. Yes, that rewards people non-linearly—the people breaching towards that future—and you leave some people behind. But that is what takes the band from here to here. Everyone's quality of life goes up. At some point, the Earth's resources are going to die out. We have one shot in the next couple of decades to expand so we can get new resources, make ourselves more efficient, and start solving all these problems.
But these are all platitudes and economic arguments over the internal cultural problem of humanity: do you want to go do cool shit and grow as fast as possible and see what's out there? Or do you want to sit there twiddling your thumbs because you're scared? That is the eternal growth versus degrowth fight. Everything we see in the media—legacy media, politics, wars, communism versus capitalism, crony capitalism versus true John Galt capitalism—it's all a poor abstraction over growth versus degrowth. Ultimately, you've got to pick a side. I pick growth.
If we're picking growth, then let's go invent new technologies, settle the stars, find out what's out there, and mine asteroids for resources instead of ripping up forests in the Amazon. It comes down to expansion or population control. At a certain point, we're either going to cut down trees or mine asteroids. I don't want to cut down trees. So let's go do something cooler and more sustainable. The third option of restricting resources is a stupid argument over nuclear. Let's not incur some risk of getting totally clean energy that is net-net way less morally and economically impactful than solar panels, where 90% of the solar panels in the country are made in China by basically slave labor, with millions of minority and immigrant women being chemically castrated every year and forced into servitude.
Ultimately, it's growth versus degrowth. The case for space is that it is the last frontier we have. You can think of the moon as the eighth continent. As a species, we're going to go out and get it, continue to grow, and improve things—or we're going to stagnate. If anyone is choosing to stagnate, you're on the wrong side of history. You're going to get blown past by everyone else. I wish people would frame the problem correctly. It's growth versus degrowth. There is no economic argument. You're for humanity as an expansive moral species, or you're against it. You want to corral us in this bubble where we're just going to compound the problems. I think it's ridiculous.
Host
One other way you could put growth versus degrowth is zero-sum versus positive-sum mindset. We can pursue space, and the technologies will hopefully return back to things on Earth. But we can also pursue space and pursue improving climate change and pursue AI and pursue other exciting things. It's not one necessarily versus the other. I do think it's important to point out what we've talked about already: there are several advancements from space. I was looking this up because we're doing an episode with Privateer, but whether it's LASIK, limb replacements, or tires on your car, all of that has been impacted by the very tough engineering challenge of building things in space and returning them to Earth because it's much easier to apply them here.
Without going too far down that rabbit hole, I think one thing that might be missing is the idea of a killer app—something people can understand within the context of space. Some people might argue that satellites are already the killer application of space that we use on Earth. Do you have any thoughts on whether we already have this "killer app" within space that helps people recognize the value?
Guest
That's a really hard problem because a lot of the benefits of space are not the end product. It's the infrastructure. It's GPS. Space invented Google Maps or Apple Maps. It's recognizable by a consumer, and they kind of get it viscerally. I don't think there's a killer app, and I don't think there will be, because it's like—does your average consumer know about AWS? Not really, outside of our technology vein. But can any of the stuff they love exist without it? Probably not.
I think there will be something incredible in terms of first-world countries not having access to the internet for the first time through things like Starlink. That will be really incredible for those parts of the Earth. It will become a killer app because it's consumer-focused, clearly linked to space, and has real benefits in terms of education and teaching kids things versus having crappy internet in the Philippines.
For the American population, we're already buried behind so many layers of abstraction. That's really hard. I think potentially it's something in pharma. So much cancer or genetic research can only be performed—or can be performed far better—in terms of materials or chemical research in space. If there's a major pharma breakthrough where we cured lupus or leukemia doesn't exist anymore, and it's because of research that happened on the ISS and it's very publicly done, that could be a real killer app. But the rest of it is all infrastructure. Pharma would be an amazing one. I can see a core consumer link where they really viscerally feel the impact.
Host
That reminds me of my final question: painting a future where Hadrian is successful, or rather painting a future where there are 10 or 30 Hadrians that are successful. What does that world look like? Does that enable us to go cure cancer on a satellite or on the ISS? What are the types of things we might get?
Guest
The analogy I use is for software engineering. We have so many incredible software engineers and software products. That is downstream of companies like Stripe, Twilio, or Amazon because they lowered the barrier to entry for creating new companies and running more experiments on what the world needs—from a million dollars to $100. Because that infrastructure exists, you get this Cambrian explosion of randomness. The people in the ecosystem get a hundred more tries at building something incredible. New people come into the ecosystem because the cost of entry is so low. You get more flood of talent and randomness.
Who knows what comes out of it? But it's obviously true that if you have this cheap infrastructure layer that enables rapid iteration and lower barrier costs to do an experiment or launch a product, you get this Cambrian explosion of madness. Amazing things come out of it. Who could have predicted that downstream of Microsoft Azure is Dolly 2? Without that compute layer existing, none of that would exist. Self-driving cars are all downstream of cloud computing or elastic infrastructure layers.
What I hope happens with Hadrian is, apart from speeding up the current companies making rockets, satellites, jets, and drones an order of magnitude so they can move faster, we automate this so much that it's like flicking the switch on AWS and spinning up a new East Coast instance to start tooling around with something. By lowering the barrier to entry of complex manufacturing, making it cheaper, and making it accessible through an API, we should see two things: the smart people in the space get 100 times more experiments, and new entrants don't have to go work at SpaceX for 10 years to figure out what's going on before starting something. They can drop into it straight out of college and have this manufacturing platform that enables them to rapidly iterate on whatever they want.
All of a sudden, we see this ridiculous explosion of who knows what. Hopefully, the FAA is just chasing their tail trying to stop kids launching satellites off the roof of their houses. That generative property would be a real success case for us: we've built this Archimedes lever of a company that, by building it, generates all these huge second- and third-order impacts in the world. Apart from helping the defense primes scale and butting heads with the CCP, it will be amazing in a couple of years to see a bunch of engineering grads tooling around in their garage, seeing Google-style startups happen without $40 million in funding just to get off the ground. They tap a button, parts show up the next day through an API, and they're experimenting. Who knows what happens after that? That would be the huge success case for us: we are the enabler of that Cambrian explosion of talent and randomness that produces all these wild and crazy experiments in the physical world, lowering the barrier to entry so we get closer and closer to the Jetsons flying car future, versus getting stuck where we are today.
Host
I love that. It might be hard for people to imagine that Joe Smith in his garage is going to go create a space rocket on his own. But if we look back a couple of decades ago, the idea of someone publishing online—which we all do now as long as we have an internet connection—was not democratized. The first blog was in the 90s, and you had to spin up your own server and understand web development. Today, you pull up your phone, open your Twitter app or Substack, and we can all participate. It creates this infinite ecosystem on both supply and demand. It'll be fascinating to see if we can achieve that on the hardware side, specifically within space.
That leads me to my final question. I think this work is really inspiring. It inspires me to imagine that future where we have democratized hardware, where people are excited to participate even if they don't have a doctorate in aerospace engineering. I want to know from you: who is someone that you're inspired by? And what are they working on?
Guest
Inspired by or have learned from?
Host
Could be either. I'll share where this question came from. There was an interview that Alex Honnold did on Tim Ferriss years ago. A lot of people see Alex Honnold as this superhuman who defied the laws of gravity or at least fear, free soloing up mountains. It's interesting to imagine that there may be someone that Alex looks up to in a similar domain. His answer was this guy, Marc-Andre Leclerc. If you've heard of *The Alpinist*, it's a wonderful movie. I won't spoil the ending. It was fascinating to understand that this person I saw as superhuman saw someone else in a similar light. I'm curious if there's someone you look at and think, "Wow, I can't believe this person is building something," and no one's ever heard of them.
Guest
What really inspires me are musicians who tool away by themselves in bedrooms or caves for five years and then produce works of art that are orders of magnitude better than anything publishing studios get out. Whether that was Boston in the 70s, where everyone thought they were a real rock band, but it was one guy who not only wrote the music and played all the instruments, but half the recording equipment he custom built himself purely to get his work of art out there. Stuff like that inspires me, not because it's a single human—I'm nothing without the team—but because it's a reminder that these mythical heroes exist. People from non-college-educated backgrounds or completely outside the system are still toiling away. Humanity has still got it. It's not a cattle mill of people coming out of education and going to click buttons on computers at Google or Goldman Sachs. We've still got this incredible ability to go off and do random things.
Other examples: I'm obsessed with random YouTube building videos where some nutcase convinces his wife that they're going to build a cabin out in the woods, he has no construction experience, and he does it, and it's awesome. Those things inspire me, not because of the singular genius, but because of the doggedness of getting the job done and the sheer idiocy of being like, "Yeah, I've never done this before. I didn't even know what the music industry is. I'm just going to do this album and toil away at it for eight years, hold myself to an incredibly high standard, and create something from scratch."
Why that's important to me is that all of what we're doing is incredibly hard, but it's not *that* hard. The way we've built credentialism into our society is really bad because most of the barrier to entry to doing something amazing is psychological. I try to tell people constantly: "No, I'm a moron. Trust me. This is hard. But most of the trick is being able to take the emotional pain of just hitting the wall with a sledgehammer until it breaks down." That's not what we teach people. We teach people you have to have this degree or be born a certain way. 99% of the time, except for advanced maths, that's just not true. You just have to be willing to get after it, be truthful to yourself, take good feedback, and run at it. Examples like that are really inspiring. Those are the people I look up to, versus someone who's had massive success for one reason or another.
Host
It's about perseverance instead of getting it right the first time, which is generally what we're taught—get an A on the test instead of taking the test 10 times but acing it by the end and really understanding the material. It reminds me of that famous Steve Jobs quote—I'm going to butcher it—but it's something along the lines of: everything that is built around us has been built by people no smarter than you. I think that's a wonderful place to end off on. Chris, is there somewhere you'd like to direct people? How do they find out more about Hadrian, what you're working on, or anything we've talked about today?
Guest
You can follow me on Twitter at @2112power. And if you want to come work for us and manufacture the future, you can email us at [email protected].
Host
Awesome. Well, thanks for doing this. I appreciate it.
Guest
It's great.
Host
Thanks for listening to the a16z podcast. If you like this episode, don't forget to subscribe, leave a review, or tell a friend. We also recently launched on YouTube at youtube.com/a16z_video, where you'll find exclusive video content. We'll see you next time.