Sean Liu of Founders Fund on evaluating technical risk in hard tech — and why most VC frameworks don't apply

deploy it. Uh, speaking of capital, we have venture capitalist Sean Lou from Founders Fund in the studio. Welcome to the stream. How you doing, Sean? Doing well. Let's give it up. I'm great. Uh, good to see you. Um, we just talked to uh Gro from Solen. Uh, is it solen? I keep messing up. Solugen.

I I I I want to hear your side of the story. How did you initially wind up working there? What did you learn? Uh, and then we'll go into some of your venture career. Um, so I'll start with uh how I ended up there. It was a reference from NAP and uh Oh, cool. Brian Singup.

I'm I'm kind of like a crazy tech guy that likes capital expenditures. Any company that has tech risk, bring it on. Crazy. The more dollars the better, right? Um and uh you know, I was thinking about getting my hands dirty with something early, something fun and solid came up.

Met the guys, love the vision, love the team. And uh it was a hard job. We were small in the middle of COVID and needs help and I felt like CFO COO is one of those underappreciated roles. So, I took a took a chance on myself and the company and it's been a lot of fun. Yeah. Second question, what did I learn?

Um, I think probably the it's all management theory is Any book you read is complete You you just you go in and you uh it's uh you kind of you learn on the job and uh you know just what whatever those advice they give you is all it's all it's all fake news. So, yeah. I mean, what what's an example of that?

What was the first moment where you were like, "Oh, this this completely I got to throw out the playbook on this one. " You know, I think there's there's this whole thing about you got to be nice to your employees.

It's just like, "Hey, the HR gurus and the team is going to tell you, you know, you can't work your employees after 5:00 p. m. You're just like, "Look, we're a startup. Let's go. " Right? Like, you work on Saturdays. You run you run you run night shifts. There's there's something called a Dupont schedule for a reason.

you run 247 and uh factories don't go down just because you need to take a break.

So yeah and uh I think uh I think I think it's just setting those like cultural guidelines and you know sometimes you when you hire people you hear like people say hey you know you just you got to define your culture as you know where we have all these missions and values which is true and at the end of the day for startup do you have the do you have the IQ the grit and the grind to uh to to to suffer through a marathon that's a that's a startup so it's it's almost like a different profile how much of the work at that time was focused on fundraising and kind of educating venture capitalists about underwriting what is kind of a firstofits-kind company to really go through the Silicon Valley pipeline uh versus kind of the operational role of the CFO COO actually deploying the capital.

I I would say it's it's I mean it's both your your firefighting every day, every minute. Um, I remember when we joined there was no one no one no one no one in the company to to do finance and um we had a really bad head of FP and I fired him after two weeks.

So I went to count inventory myself and uh you know yeah we you know we're not a SAS company. We got you got to count inventory. You got to know what is your warehouse. So it's it's it's important. You got to know what every piece of equipment cost.

And on the flip side of that, yes, there's a huge and uh you know, capital makes a difference, right? Cost of capital. We we haven't had a new chemical startup that scaled in, you know, more than 30 40 years. There's a reason for that.

Partially is because there's a tech barrier where R&D is centralized amongst these um petrol giants and uh they don't really want to release IP. And the second piece of this is cost of capital for startups is so high. You build a factory probably cost you you know 300 million bucks on the low end.

We have a revolutionary technology. You still got to bridge through the valley of death and you know for hard tech capital intensive business you you have more of them than you typically see in SAS.

So educating them on how to think about risk um and them venture capital um how to think about risk at what point should you see inflection in value and what is that actually transcend into is uh is you know I would say is other half of the battle. Yeah.

On the IP side, uh are you like were you watching for I for patents to expire and then implement them in immediately? Are there other ways around these problems? Like like what was the IP landscape at the time?

Are are there still like trade secrets that are just uh important that you could just focus on maybe reverse engineering or finding a different pathway? Like what what was that like at the at the company at the time? I mean I think you know there it's really interesting, right? Like I think patents only get you so far.

That's why you know if we if we have faith in patents, China wouldn't be where it is. Um but I think uh but I think the the key the key is what I always say if we take a big step back a lot of the formulation though how still sits with the petrol giants.

They have a lock in and they have a lot of resources to continue to be patent trolls.

I remember one of the, you know, probably two months on the job, you know, right about when we were about to close down, one of our competitors sent a assist and you know, cease and desist to us that was like, "Oh, you stole our formulation.

" And it's like tiny little variance and something was complete and they they can still sue you out of oblivion, right? Um, but I think we've kind of learned over time to in some ways you kind of you do have to fight some of the battles similar to what they do.

you you file all these umbrella patents as general as possible um to protect your core IP and on the other side you you kind of have a lot of trade secret that's very difficult to duplicate um but at the core because how different solen approach is to thinking about chemistry and how to scale this right it's a generalized enzyatic based platform that you know it makes that that side of it much easier for us to navigate talked about uh talk about the transition to founders fund how' you wind up there and then I want to go into some of the deals that you've underwritten and how you think about uh some of what's going on in the hard tech landscape broadly.

Yeah. Um it was it was similarly you know I I I was you know my my family is in the Bay Area you know want to start a family and I was thinking about what are the places where I get to work with the smartest people building investing the craziest things and still be able to help solen founders.

So transparently was only choice I couldn't see myself working in any other VC firm. I don't think I'd fit very well in any other VC firm and you know I'm very grateful that the guys took me under their umbrella.

So uh obviously you've you've looked at a lot of hard tech companies at the same time and uh valuations feel very high like walk me through how you're actually underwriting a growth deal in a somewhat capital intensive business. How is it different than looking at a software company?

what what are like the major pitfalls or landmines that you need to be looking at when you're building a investment thesis? Um, and is there anything that's kind of like uh underrated that most uh companies or VCs might be missing when they're underwriting these uh hard techch companies in this like kind of new era?

Yeah. Um, man, it's a it's a broad question you asked and I think it's it's a hard one to answer because within hard and capital intensity, those are two different things, right? Sure. Last year, you know, NAP and me and I did a did a deal deal in Crusoe and it's more asset intensive.

There's technology involved, but the way we think about something like a Crusoe, which is more of an energy thesis and long-term bet on, you know, the infrastructure shifts that's happening in the US, the energy gaps that we see is is quite different than what you typically see in something like a Neuralink or SpaceX, right?

where where where if we take a step back if you think about what is the technical risk you feel you face and the first question we always ask is do you need to invent science and if you do then that's a very different question and very transparently you know founder son is is in the business of backing revolutionary technologies and companies right the end part is important so how does this science transfer into a company is sometimes missing in those discussions you see really brilliant technologists building these cool tech and you realize that, you know, if you believe you need five, six, seven new fields of science to all come together and have that to be true, then then we are probably not the right place to be, right?

Um, but then as you kind of progress towards the true like growth stage, are we thinking about scale at risk? If so, what are we dealing with? Is it like no unknowns or unknown unknowns? And what are what are the different ways of thinking about that?

Sometimes when you look at let's say a a a new generation technology a lot of the question we say look the science is proven we don't need to invent science let's look at what does it take for this to materialize in such a way that we as investors transparently can actually generate a return that makes sense in a time frame that makes sense right and a lot of that hinges on do you have the supply chain ready the the upstream fuel sources turbine generators and everything else that's that's associated with building something like a a new generation technology can can You do you believe that if you you if you you figure this out can the rest of the supply chain also figure this out with you and a lot of the time the answer is no and that makes it that much harder and the last thing is I do feel like in most hard tech spaces founder have a tendency to oversell the technology and they have a deep founded belief that the moment you know the technology is mature people will come and that is not true right and I think a lot of the questions I think Trey Peter Nap and others ask on the team is right like how are you going to sell this thing?

Just because you make it doesn't mean people want to buy it. Um and I think some of the best investments we made are you know in hard techch are the technology company that doesn't need to invent science. They're making incremental improvements to scale.

They understand the bottlenecks as a supply chain have a realistic plan to deal with the bottlenecks they see. And then lastly the market is already there right if you think about SpaceX launches a market. Yes, it's expanded, but they have a really strong anchor to kind of take very to take market share very quickly.

So, that commercial-mindedness on day one is super important. Yeah, it's the same thing with Starlink where people are already buying internet and then they just switch their provider or or add on a secondary provider. Uh Jordy, do you have a question?

Do you see hard techch as a category becoming something that goes the way of maybe the creator economy was in 2020 where a lot of you know maybe there's opportunities but a lot of investors get so burned that um that it that it becomes kind of this flash moment in time uh or is it something that you see the average venture capitalist continuing to invest in consistently over the next 10 years?

How skewed is the power law? Because I mean, if you look at the the history of like dev tools or like databases, it feels like there's like 10 companies that produce venture returns in most of those categories because a bunch of them got out and IPO, bunch of them hit multi-billion valuations. It felt less. Yeah.

Like there was a point in time where people were saying, I'm going to start investing a lot of money in software as a service.

And then those people if they were decent investors 10 years later they still said I'm going to be I'm still a software as a service investor the question I have is like every VC over the last two years decided I'm also a hard tech investor and I I uh rationalizing it by saying you don't need to be technical to invest in hard tech and and um there's definitely evidence that that that that can be true but I'm curious if you think the average venture capitalist even 3 four years from now will still be deploying into uh broadly hard tech.

Um you know history doesn't repeat but it rhymes. I feel like people don't learn and will continue to do that. I think you know what's interesting is we probably see a more I would say hard techch as a category is probably is a misnomer. I think you'll see more finely parsed views of heart techch. Are you symbio?

Are you, you know, biotech pharma? Are you biotech infra? Right? Like I think that that should that that we shouldn't look at hard as a category in its totality, right? I think defense tech historically was grouped in as hard tech if you look at this, you know, 10 years ago, right? Android was hard techch today.

It's its own category. Sure. Um, and I do feel like people continue to to to invest, right? Like the nature of technology is is hard tech. Very transparent. All new technology that emerges hard techch. So you'll continue to see that.

But the moment you create one power lock company, you'll see you know the whole host of siblings and children come out and then it becomes its own category and people invest in it. You'll continue to see this you know cycle of life and death.

Symbio was super hot in 2020 2021 and I feel like that's now a taboo term but at the same time you feel like there's tremendous amount of progress in technology and I feel like a lot of the company we meet on the AI for science side is pretty much I would say the next generation from a fundamental tech perspective addressing what biology can do right you see a lot of funding going into it um but but but to reframe your your question another way how do you see hard generating returns for venture capitalist I think hardc more than a software is de facto power law, right?

There's there's one SpaceX that's, you know, 99% of the market. I mean, space is not a market. Space in many ways is SpaceX with a lot of different derivatives that opened up. Um, you know, Andro today captures most of the mining share for defense and you see that, you know, in in in other parts of the heart ecosystem.

There's very few pharma companies and they're when they're, you know, and they're they're they're massive in in comparison. Um and I think the reason why I think this is more defensible than than software is because the you know c you do have huge amount of capital advantage and cost of capital is a huge advantage here.

When you're Dupont you have hundreds of billions of dollars of capex on the ground. The marginal cost to produce is so low that there it's very very hard for an incumbent to come in and displace you.

Um, and that's why we feel like there there there's fewer power law companies here, but if you bet on the right ones and double down on the right ones, you can create massive short term. Do you have an do do you have a take or like a post-mortem on what happened when venture capitalists went really big into clean tech?

I think this was like the late 2000s, early 2010s. Uh my kind of naive take on it was that uh it was well-intentioned based on you know environmental concerns in many ways uh but the category definition was too narrow and so they didn't include Tesla as a clean tech company.

Tesla obviously if you had included it in the clean tech bucket would have been the power law winner and driven massive venture returns.

But instead Tesla was seen as a car company and as a result wound up you know uh not being included in that and so all the clean tech funds underperformed just the general software funds.

In terms of like uh in terms of like category definition and just postmortem on the clean tech you know investment thesis that we went through a decade ago. What what's your take away from that era? Yeah.

Um, you know, Garb had this really good saying where he was like, you know, if you're one step too early, you're a genius. Two steps too early, you're a martyr. Um, and I feel like clean tech 1. 0 was the martyr category. Mhm. Um, it's kind of like building Dropbox in the 1990s.

You didn't have enough, you know, bandwidth wasn't wasn't high enough. Internet wasn't quite a thing. And in clean techch 1. 0, the industrial base is not quite there, right?

polysilicon processing um capabilities was not there and um and and and and you know the utilities had the utilities ability to manage um non-firm systems ability is also not there.

So there's a whole host of factors that that that kind of made the the cost structure was not competitive, manufacturing scale up was not there. You need to invent new science to make this work and you also need to then do sales and marketing to convince consumers as it's actually made sense, right?

And you're also coupling that with selling to utilities which has capped ROI.

So fundamentally a bad customers was bad setup all around and and and to your point like I think the the the pitch of a narrower um definition of which is purely focused on generation technology is also I think misplaced right Tesla is not a generation technology it is a car it's you know you group it with car technology that shifts it um and I think if you think about nuclear today it is clean technology I mean yes we're rebranding it but nuclear is the cleanest of all technologies Mhm.

Right. So, I think I think there's this this whole wave of how do we think about um how do we think about creating and broadening that definition so that you see where the cleanest the clearest commercial pole is coming from is probably more important than the definition of what the technology is or is not.

On the flip side of that too, I think the last generation was highly dependent on government subsidy. Then no like if you look at the underwriting whether it's from the DOE or others like all of them assumed certain level of government subsidy that lasted for decades on it and it just was not going to work.

And I think that's important to to you know as as we think about investing in hard techch and and you know a lot of these new technologies when does the subsidy go away at what point are you competitive? Yeah.

Uh that that reminds me of the story of Celindra heavily uh like a picked winner by the government and uh it feels like the Celindra example in in 0ero to1 at least is kind of like I don't know I call it like a spreadsheet error.

It's like they built their spreadsheet for the economics of that business and they just missed one variable.

And it feels like in hard techch there's a number of companies uh that are in that category today where uh it sounds really good in a pitch deck but once you actually run the economics there if there if there's one number that's off by you know some small factor you wind up with completely upside down economics.

And so in terms of underwriting hard techch companies like how how important it is is it to go through the long-term economics is are like do you need to bring in like um like experts or run like the first principles analyses on these companies?

I'm just seeing a lot of companies that are like we're going to do this on the moon or this in space and I'm just like oh if if launch costs don't fall by 10x like you've assumed you're completely upside down.

um that seems like an important part of underwriting or is that usually less of a factor because you're focused on the growth stage? Um it's still very important, right?

Like I think when we think about that the the question we always debate at founders on is what is how many like need to believes do you have to make the thesis work? Yep. And if you need to believe, you know, if you believe A and B and C by the point like this codependency, Yep. you multiply them, you get zero, right?

And I think of this very much as like a semiconductor manufacturing process. 7,000 steps. Each step has 0. 1% yield loss. By the end of it, you get no chips out.

So, so it's it's it's similar like do do you need to believe you know to your point if you let's take like you know some crazy thing like asteroid mining in you know Alpha Centuri.

Do you believe need to create like nuclear fusion engines, launch costs, you know, robotics and then you're like maybe I just back all these company and then wait for that to happen. Right? So that's that that that that is a very very important thing.

I think the more ant statements there is the worse it is and very transparent we don't like to make those bets right like the cleanest companies are the ones with one or two of these and very much it's you you you don't need to invent science with very little dependencies um you know when we look at general matter when you look at solen or or or others in our portfolio you do see that neurolink even for that matter it's you you don't see that many and statements a lot of those have been proven out and what we need to understand is really like at what point like what do you need to for that to scale.

Um and um and I think I think to your point on like how do we do due diligence I mean yes I think it's it's we do need to understand the science there is a zero one question like do we do we take on binary risk and where does that exist and then there is a you know going back to founders roots what's the what's our read on the founder how how do we see them in tackling all the risks coming up yeah and the execution side um well thank you so much for joining this was a fantastic conversation we'd love to have you back enjoy for the rest of your day and thanks for breaking it down for us.

Love how cam always fun. Cheers. We'll talk to you soon. Bye. Uh let's check out let's check in on Tyler in the cam. He's got the glasses on. You know that means he's towards final assembly. I'm pretty locked in right now. Come back to me in maybe 15 20 minutes. I'm I'm I'm almost done. Okay. Almost done. Almost done.

That's great. He is cooking. I think he has another maybe hour in the show. It looks like he's making progress. We'll check in with him uh at the next break. Uh, first let me talk to you about eight sleep. Get the Pod 5 Ultra. How did you sleep last night, Jordy? Use code TBPN if you're going to get one.

They're fantastic. 5-year warranty, 30 night risk-free trial, risk uh free returns, free shipping. I got a 91. How'd you do, Jordy? I'm sick. So, oh. Oh, excuses, excuses, excuses. Oh, John beats you one time and all of a sudden the excuses come out. What'd you do? What'd you do? Give us the number.

Sorry, the the Wi-Fi is not working. Yeah. Okay. The number is horrid 6 63. Let's hear it for me. That's maybe the worst. I feel like I I come out strong in the beginning of the week. I usually beat you Monday, Tuesday, but then you catch up and then you just go on a go on a run. Uh anyway, uh get go toleep.

com/tbpn and we will bring in our