Founders Fund on Truth-Seeking, Taiwan, and Whether You Can Still Beat the S&P | Ep. 53
Trae Stephens and Delian Asparouhov on what makes the firm different, why board meetings rarely add value, and why today's market feels uncomfortably familiar.
Trae Stephens is a Partner at Founders Fund and co-founder and Executive Chairman of Anduril Industries, a defense technology company building autonomous systems for the U.S. military and its allies. As of May 2026, Anduril is valued at $61 billion. Prior to co-founding Anduril, Trae was an early employee at Palantir Technologies, where he led teams focused on the intelligence and defense space.
Delian Asparouhov is a Partner at Founders Fund and co-founder and President of Varda Space Industries, a company manufacturing pharmaceuticals and advanced materials in microgravity that can’t be made as effectively on Earth. Delian was a Thiel Fellow in 2013 after dropping out of MIT computer science. Prior to Founders Fund, he was Chief of Staff and then Principal at Khosla Ventures under Keith Rabois, and before that founded Nightingale, a healthcare startup backed by Y Combinator.
We discussed what makes Founders Fund’s culture uniquely truth-seeking, from its 1% personal co-invest requirement to a structure with no partner meetings and no residual economics. Trae shared why he believes Series A boards are largely a waste of time, while Delian reflected on working alongside Keith Rabois and eventually striking out on his own. We also explored their concerns about today’s market, how Founders Fund navigated the 2021 bubble by concentrating on its two strongest companies, and why macro timing rarely works in venture. The conversation closed on America’s semiconductor challenge and why a Taiwan conflict may be more likely than many assume.
Timestamps:
(0:00) Intro
(0:54) What makes FF unique
(8:01) Lacking EQ can be an asset
(12:10) Delian on working with Keith
(18:30) Why Trae hates board meetings
(24:13) Benchmark vs. Founders Fund
(26:31) Investing as the exhaust of operating
(28:52) Venture as a micro asset class
(30:42) The infamous 2021 offsite
(34:21) Current market sentiment
(39:18) Can you beat the S&P?
(48:53) America's semiconductor blind spot
(51:56) Taiwan and the 90% probability
(57:58) Robotics, humanoids, and the China question
Watch on YouTube; Listen on Apple Podcasts or Spotify
Transcript
Truth-seeking and the 1% mechanism
Jack Altman
I’m very excited for this. What a crew we have here. Delian, you excited to be here?
Delian Asparouhov
Oh, yeah.
Jack Altman
I’ve been looking forward to this for a while.
Delian Asparouhov
In the lion’s den.
Jack Altman
100%. It’s going to be hard. And, Trae—
Delian Asparouhov
Oh, sorry, tiger den.
Jack Altman
The tiger den. Yeah, that’s right. Can we run that back?
Trae Stephens
There was a lot of discombobulated commentary there.
Jack Altman
We’re going to try to have a serious conversation.
Ev Randle
That’s going to be this whole two hours.
Jack Altman
I got to do this with Trae, and that was amazing, so we wanted to bring it all together. My first question is actually for Ev, and we’re going to start in a really positive place. We really respect what you guys have done and what you do as a firm. I hope that’s the same. But Ev, you’ve worked at both. You’ve talked a bunch about what makes Benchmark unique. I’ve obviously gotten to hear a lot about that from you. I’d be curious for you to talk about what you loved about Founders Fund, and maybe compare it to what’s different about other places, maybe Benchmark.
Ev Randle
When people that aren’t in the investing business think about what investors do—what the day-to-day is, or what an investment committee looks like within a venture firm or any investing firm—they have this idea in their head that the team comes together, they’ve done a bunch of research and a bunch of work, and they have this very intellectually honest debate about the merits and considerations of the investment. Weighing the risk-reward, and then coming to some intellectually perfect decision about that investment given all the information that they’ve gathered. That’s just not how basically any firm actually operates.
There are so many other politics or organizational structures and incentives that push you away from this truth-seeking place. So I think the really unique thing about Founders Fund, that when I came in I thought was just so amazing and unique, it wasn’t only that it was a culture that truly bought into the process of truth-seeking, really putting truth-seeking on a pedestal above the hierarchy of who’s a GP and who’s not a GP or any of these other things. But there were also organizational incentives and structures in place to keep that the same basically no matter who was in the halls of Founders Fund.
The one that I’ve mentioned before is the 1% mechanism, or provision, or whatever it’s called. For any investment that you do at Founders Fund, 1% of that investment on a dollar basis is set aside for the people that worked on the investment to basically angel invest up to 1% of the total dollar amount. It’s not just some workplace benefit. The entire idea of that mechanism is to basically measure the conviction of the partners or the investors that are sponsoring a given investment.
Jack Altman
Do you guys still do that? If somebody doesn’t want to do it… Is that a thing?
Trae Stephens
No, I think it’s really important to be mindful of liquidity constraints, obviously. People on the team have very different personal financial situations. So it’s not super prescriptive, but it is a useful tuning.
You don’t have to do the full 1%. You can split it up between the different parties that work on the deal, or you can say, “Hey, I have five grand or whatever that I want to put towards this.” It’s not a signal of lack of conviction, it’s just based on liquidity constraints.
Delian Asparouhov
I think it’s more helpful on an individual comparative basis. If you’re known for doing lots of 1%’s and on this one you’re choosing not to, people will ask. But if you typically don’t, and then don’t, that’s not necessarily a signal.
Jack Altman
Are there any other incentive-aligning mechanisms like this that you guys have going?
Ev Randle
Obviously, you’ve been at Founders Fund much longer than I was. Do you agree that truth-seeking is the highest calling of the team? What do you ascribe that to? Is it cultural, are there organizational things that lead to it?
Trae Stephens
When I first joined Founders Fund, as you guys know, I didn’t have any interest in venture capital. I just ended up there because Peter basically told me that I was going to do it, and it wasn’t clear that I had an option.
Jack Altman
I love that. That’s the best kind of recruiting process. “You’re going to work here.”
Trae Stephens
So when I showed up, I kind of didn’t know what I was doing. I had this hypothesis that venture capital was going to be more like what you said: It’s like this diligence exercise, you meet with hundreds of companies a year, you do a deep process on each of them, and the hard part is making the literal investment decision. It’s not what it is at all.
There are tens of thousands of companies that are raising at any given point. 99% of them are obviously, knowably bad within a second. So really it’s about access and finding the super high-conviction things that you want to pour your energy into. I think that’s really what makes Founders Fund unique. We’re just very internally honest about that. If you’re not personally super high-conviction—”I really want to do this deal, I’m going to fight for it, I’m going to be willing to argue, I’m going to be willing to put my reputation on the line disagreeing with Peter”—if you’re not willing to do that, you’re just never going to get a deal done. So that’s really all it is at the end of the day.
Delian Asparouhov
I spent a little bit of time early in my venture career at KV. One of the things that I find very uniquely different between the two models is that at FF, so much more of what we do—I’d have to look at it on a dollars basis or deal-count basis—rises from below, if that makes sense. It’s individual team members that have really strong conviction and are steadily presenting that case to other colleagues and working their way up. At FF, every check above a meaningful size still requires Peter’s approval. But it’s almost always that person steadily gaining conviction and convincing others, and then eventually presenting that case to Peter, much more so than top-down Peter saying, “We need to go invest in this thing.”
I think that creates some truth-seeking by default, because there’s such a high activation barrier. Peter’s got many things going on, Trae’s got many things going on. If you’re going to take up their time, especially because venture’s not the only thing that they do, you’re going to want to have a really compelling case as to why. At other places where the GPs are full-time, only-investing GPs, I find that they’re also the ones leading a lot of the investments themselves, which I feel like is less the case at FF.
Jack Altman
The incentives thing is interesting. One of the things I’ve noticed here that Ev and I have talked about is that because there’s nothing at Benchmark to fight over—everybody’s equal, you’re never going to get a promotion, you’re never going to get a raise—because a lot of that stuff is stripped out, it leaves there to be nothing other than trying to make good investments and trying to help your teammates make good investments. At a lot of big firms, that gets harder because there are other competing interests that people have.
But Founders Fund seems to have avoided a bunch of that through a lot of these mechanisms we’re talking about. There’s the 1% thing. There’s the fact that a lot of the senior people are crazy busy. I also feel like you guys have a cultural thing where there seems to be something to it. Ev’s talked about how, at a Founders Fund partner meeting, people can get pretty heated over ideas, or even maybe at each other, but somehow at the end everybody still really likes and cares about each other. I think that’s the inverse of ruinous empathy, or toxic collegiality, which I think is much more common.
Ev Randle
Where everyone’s very polite.
Jack Altman
Everyone’s very polite and then after the meeting they’re like, “Look at that idiot. I can’t believe he thought that.”
Trae Stephens
I just think the EQ of the team isn’t high enough to be able to operate in that way.
Jack Altman
I think we just got a little sprinkle of it, so as a result everybody just says what they think and moves on?
Trae Stephens
Yeah, because I just don’t think people are tuned to even understand how the things they say have an emotional impact on other people.
Jack Altman
Would a super high-EQ person not work particularly well at Founders Fund?
Trae Stephens
Either that, or they’d be really good at manipulating the group.
Jack Altman
Yeah, just take full advantage of it.
Trae Stephens
One or the other. I’m not really sure. It could go either way.
Ev Randle
They’d never know.
Peter Thiel and Founders Fund culture
Jack Altman
Actually, can I ask, what’s Peter’s interpersonal… I could imagine you saying this in either direction, but he clearly is deeply attuned to people in some important way. Is this thing you’re describing about EQ related to him, or is it different from him? Obviously, when you have a founder of a firm, a lot of the DNA is going to go through it. But does much of this come from him?
Trae Stephens
There’s one thing he said to me very early on. We were talking about having kids and what we wanted for our kids. We were talking about intellect as one of these things that you as a parent think about wanting for your kid. He said something along the lines of, “I think 130 is roughly the right IQ, because anything over 130 there are too many trade-offs.”
You’re not hiring people for their EQ or lack thereof. You’re hiring people to have really high IQs that are able to hold their own in these debates. The reality is, when you do that, there are trade-offs. So we’re experiencing the result of those trade-offs at any time.
Jack Altman
Too smart for your own good?
Trae Stephens
Yeah. Or maybe it’s exactly for everyone’s good, because we don’t have these weird harbored politics and animosity. At least I don’t feel… I’ve never felt like that.
Jack Altman
Delian before the pod told me that he can’t stand you, but other than that, I think everyone likes each other.
Trae Stephens
Of all people, I think Delian and I are very cool with one another.
Delian Asparouhov
I think Peter deeply understands people, but from an intellectual framework level. He’s very good at quickly understanding where somebody is biased, where they have blind spots, what they’re good at, what they’re not good at. In terms of how his words will have XYZ emotional impact, as Trae said, I think that he’s poor at. But when it comes to deciding whether or not to make an investment, it’s actually much better to understand somebody’s intellectual capabilities versus the emotional-impact side of it.
I also think right now is probably amongst the best cultural states it’s been in, at least in the time that I’ve been there, but even from the history that I’ve heard. There is a variety of reasons why that is, but definitely one of them is that if you look at the median age, or median tenure, of somebody on the team, it’s probably on the order of seven or eight years, which is reasonably high relative to most venture firms, amongst a spread of call it 12 people that are there.
Jack Altman
Has the culture gone through ups and downs? You’ve been there a long time.
Trae Stephens
Oh, for sure.
Jack Altman
Do you think it matters? Maybe to ask more accurately, what do you think the impact is of good and bad cultures at various points in time? Because obviously you can make a great investment in the context of a terrible culture and vice versa.
Trae Stephens
Primarily, it’s, can you retain the people you have that are doing a good job? That’s ultimately the thing. Because you’re right, you could have… Bridgewater as an example, great performance, but I think it’s a really hard place to work for a long period of time. That’s the trade-off you’re making at some of these funds. It’s usually not a systemic problem. Usually there are individuals inside of these systems that tend to buck against what the general expectation is from the team for what they signed up for, if that makes sense. And then when they feel like there’s been some sort of violation, things get dicey.
When I first came into Founders Fund, people were definitely in their offices, staying very quiet, staying away from each other. I was walking through the halls trying to give people high fives, and they were like, “What’s going on right now? This is really weird.” But then things got really collegial. It kind of goes in waves depending on how people are vibing, how we’re adding to the team, whether or not we’re exiting rapidly when people aren’t working out, and just being really honest with ourselves about what we want in the long run.
Working with Keith Rabois
Jack Altman
I was just thinking as you were saying that. You obviously worked really closely with Keith for a long time, and then you didn’t. I’m actually interested to hear, we’ve never talked about this, but what are your reflections? You had this apprentice relationship for, I don’t know what it was, five, seven years? It’s wild.
Delian Asparouhov
Four and a half years, six years.
Jack Altman
You’re kind of doing your own thing now, with him going back to Khosla. Has that been seamless and smooth? Do you reflect on that a lot? Because I think it was an uncommonly tight apprenticeship model. You don’t see it all the time.
Delian Asparouhov
Yeah. Obviously the single biggest influence I’ve ever had on my career. I think that’ll probably be the case for the rest of my life, because there’s nothing that compares to that early period where he was the only person I basically worked with for a really extended period of time. Versus now, even though Trae and I work together a lot—
Jack Altman
But Trae’s not that good.
Delian Asparouhov
I’m just not spending 12 hours a day with Trae, even when we’re most collaborative. There’s always this talking point that I give to founders or investors, which is that within an operating company there’s actually a decent set of constraints you have to live by. You have to serve your customers, you have to fundraise, you have to actually hire talent, et cetera. So if you were to examine the cultures of Anduril and Varda, for an example, there are for sure differences, but in the grand scheme of distribution curves they’re probably not that different relative to the distribution curve you see in venture.
Because in venture, there’s effectively only one rule, which is IRR and make money, and then the rest is totally free-form. So if you were to compare the Benchmark versus Founders Fund cultures, my guess is you’d see way more differences there, whereas XYZ top-tier startups would actually be much more similar. Seeing Keith operate in these different environments very closely, my net reflection—and I think he would agree with this—is that his work style, personality, and how he likes to do things actually fit into how Khosla Ventures’ culture operates a lot better than I think it did at Founders Fund.
I’m not sure that would’ve been obvious without him literally doing what he did and in some ways sampling the two. I also think him coming over to FF helped fix some of the things he wanted to get fixed at KV. So I think the whole process and outcome was the best for everyone all around. Obviously, now not working with him day in, day out is an interesting experience. It definitely took me some time.
I remember Trae and I went on this super long walk, maybe a week after it was clear that’s where things were headed. It had already steadily been happening over the prior year or two, where I was doing more deep tech, more Varda things, which were very much not the Keith fintech type of things. The thing that made me comfortable in some ways staying was actually looking back at the prior year’s investments and realizing that, of the people I worked with on votes on a check, the prior year was actually 80% Trae and Scott Nolan, and only 20% Keith. So I was like, “Oh, I’m already doing this job.”
Jack Altman
You’re out of the wing.
Delian Asparouhov
Yeah, I’m out of the wing already, basically.
Trae Stephens
The other thing that’s worth mentioning about this is that Keith, stylistically, is a really good coach or mentor because he’s very hands-on. He wants to do updates. He wants to get in the weeds with how something is going.
That’s the opposite of Peter and myself. We’re not like that at all. I think I’m a terrible lead… I don’t mean lead investor.
Delian Asparouhov
Board member, manager.
Trae Stephens
I’m not a good board member, I’m not a good manager, I’m not a good mentor. I’m just not wired that way.
Jack Altman
Are you sure that means you’re not a good board member? Because I think a lot of great board members are… I agree about the manager thing, but…
Trae Stephens
I think it depends on what you’re looking for out of a board member. If you’re looking for someone that’s going to get their hands dirty and help you with things tactically, then I’m probably not going to be very good at that. If it’s “protect the founder, be an advocate for the founder, stay out of the way, and don’t create unnecessary friction”, yeah I’m great at that. I’m great at not creating additional friction. But it’s just stylistically super different.
Delian will reach out to me and be like, “Hey, there’s a company that I would like your read on,” or, “There’s a question that I have that’s hyper tactical.” But am I going to reach out to Delian and be like, “Let’s review…”?
Jack Altman
You’re like, “Am I going to reply to that email? No I’m not.”
Trae Stephens
It’s just not me. I think that maybe you got to the point where the thing that you legitimately got in droves from Keith was not something that made sense in the long run.
Delian Asparouhov
I remember when I first started working with him. I would describe it as: you get to work with this person that’s one of the greatest of all time. He has this knowledge tree through all the situations that he’d been in, the different business models, startups, founder conflict, lawsuits, et cetera. In each situation that I’d get to observe him in, I got to see one branch of that tree. And then as you get to observe him over the course of six years, you actually create the meta-layer tree, and it’s like, “Ah, this is the tree of knowledge that works for Keith.”
I’ve definitely adapted a lot. I would say my equivalent is probably 60-70% overlap with Keith. And then my 30% is mostly that I like a little bit more of the hardware-y, physics-y materials, those types of things. But 70% of how I get excited about a seed round or how I operate with a founder afterwards… I definitely still largely take the Keith-style venture assistance role, balanced with the fact that I’m deeply involved with Varda, so I maybe can’t quite do what Keith sometimes does in terms of depth of involvement with companies. But I definitely take a different approach to it than Trae and Peter do.
Jack Altman
Like what you said, Keith fits beautifully at KV. He’s a particularly distinct personality, so maybe you see it more sharply with him. But I think we probably all are great in one environment and terrible in another. There are people who’d be an amazing founder for one thing, horrific for the next. I think so much of life is getting yourself into the place that you’re supposed to be in.
Trae Stephens
If I was at a venture fund where there was an expectation that I take a board seat on the deals that I lead—
Jack Altman
You’d quit.
Trae Stephens
I would quit. There’s just no way, I couldn’t do it. I don’t like being on boards at all, even for companies that I started myself.
Delian Asparouhov
Even Monday partner meetings, if we had to do those.
Jack Altman
Oh yeah, you guys have no partner meeting.
Delian Asparouhov
Yeah. If we had to sit through four or six hours of taking pitches all together, then discussing them all together, and discussing the portfolio every single week, I think the entire Founders Fund team would quit.
Board work and venture theater
Jack Altman
Do you not like the board member stuff because you think they’re board member theatrics, or because you’re like, “I made the investment and you should figure it out, and I don’t want to be part of it”?
Trae Stephens
I’ve thought about this a lot. There are three things. The first reason I hate it is the board member theatrics. It’s people who—
Jack Altman
It’s exhausting.
Trae Stephens
—who believe that somehow they are better at running the company than the founder. They stand on a soapbox and try to convince everyone in the room that they’re really smart. I can’t deal with that.
Jack Altman
I hate it when you’re in a board meeting, and someone’s there, and you can tell they’re like, “This is my Super Bowl. I’m about to show off in front of everybody.”
Trae Stephens
Yeah, “I’m about to show off”. Exactly.
Jack Altman
I’m like, “come on.”
Trae Stephens
If you invested in the company, you should probably believe that the founders are the right people to run the company, otherwise you shouldn’t have invested. So that’s the first thing.
The second thing is that information density is at an all-time low. In a pitch meeting, in 30 or 45 minutes I can do a full dump of everything that I believe I need to know. If I have follow-up questions, I can do that asynchronously. But I can’t tell you how many times I’m sitting in a board meeting that’s scheduled for three hours, and it’s like, “Dude, we could have done this in an email.” Just tell me what the numbers are, tell me what your struggles are, answer a couple questions, let’s get out of here.
Jack Altman
And then you think about how expensive that three-hour meeting is. They had their whole exec team there. Oh, man.
Trae Stephens
Yeah, it’s crazy how low the information density is. The third reason is that I don’t actually think I’m adding a lot of value, and I don’t like being in situations where I don’t feel like I’m adding a lot of value. So when I sit through this thing where I sponge up a bunch of information, maybe interject one or two things, this is not only a waste of my time, it was a waste of everyone else’s time to have me sit here and pretend that I’m going to know the right things for you to do in all these moments.
Jack Altman
Okay, so you don’t want to do it, but do you think venture should have people who do want to do it and are engaged and do the whole board member thing? Or do you think the whole construct’s wrong?
Trae Stephens
At some stage, I think you need to have governance. There should be a board that serves a function. I think these boards are pulled together way earlier than they should be. A Series A company should not have a board. It’s just really stupid. “What are you reporting on?” “Oh, we built additional features into our product.” It’s like, “Who cares, dude?” I just don’t understand.
Jack Altman
I can’t imagine you at a Series A board meeting and they’re like, “Yeah, we had a great month,” and you’re like, “I don’t give a shit, man.”
Trae Stephens
Yeah, I really don’t care. It’s exhausting. There’s almost a perfect correlation between the funds or people that think you should have a board at a seed or Series A and the type of people you don’t want on your board long-term.
Delian Asparouhov
I think the core thing at FF is, for a decent chunk of us, especially on the venture team, we’re constantly comparing it against just going to work on our own companies. So the way that we view venture is very different. We are working on this because we’re explicitly choosing to not work on the company, and so this better be very high value. Relative to somebody who’s doing it full time where it’s, “I need to figure out some way of making myself feel valuable in this world,” and so board member theatrics is one way of feeling good about it.
Ev Randle
So then you both have boards at Varda and at Anduril. Do you feel like you’ve been able to take these things that you really don’t like about boards and… Not to put you on the spot, but how would you grade the work that you’ve done to make the boardroom at Varda and Anduril?
Trae Stephens
Massively condense the information density. That’s the number one thing. There’s no such thing as a useful three-hour board meeting. It’s 90 minutes tops. Only run through the most important stuff. Obviously you can have additional materials if people want to dig in more deeply async, but keep it really tight.
At Anduril, one other thing is that there are only three board members: me, Palmer, and Brian, and we’ve controlled that. We don’t want to have seven investors as board members, because it just increases friction in a way that isn’t actually useful for the company.
On the third point—my ability to add value—I feel differently, because of what Delian said, which is that this is also my full-time job. I actually feel like I am doing the things that are most valuable for me to be doing personally with the company, as chairman, as a co-founder of the business.
My biggest point of feedback to most companies is just on the information density side. It is crazy to schedule a two-hour board meeting or a three-hour or a four-hour board meeting. If you can’t do this in 60 to 90 minutes for an early-stage company, what are we even doing? It’s just absolutely out of control.
Delian Asparouhov
Trae’s actually on the Varda board, so he can maybe grade how we do, independent of my assessment of it. I do think early on we did a pretty decent job of keeping it very dense. Now in the later stages, I think of it as three sections of the board meeting. Because we have completely different customer sets from our DoW business and our pharma business, we typically end up doing a DoW update, and it’s basically just for Trae.
Because, look, he’s one of the best DoW people of all time, and definitely way better than anybody else on the board by many orders of magnitude. And then even so, we just did our last one yesterday, and I’m sure Trae, midway through, was just… We were asking, “Do you know this customer group? Do you know this?” Trae’s like, “Nope, completely different from where Anduril works and does, et cetera. I can give high-level guidance, but you guys are the ones that know this.” That’s kind of what I imagine was going through the back of your head.
Trae Stephens
We also did the whole thing in 30 minutes.
Delian Asparouhov
Yeah, it was 30 minutes, so we ran through it quickly. And then on pharma, we do find it incredibly valuable to have some of our external board members, including Samir from KV. On any particular deal, pharma asset, et cetera, there are just so many dynamics that it’s honestly hard to have all of that internal expertise. So there I actually find it deeply valuable.
And then the third section is ops, finance, et cetera. At our current stage of amount of capital raised, deployment, et cetera, it is helpful to have some external governance, even down to founder comp, executive comp, and just getting external perspectives.
Benchmark vs Founders Fund
Jack Altman
This is another area where I would say Benchmark and Founders Fund are pretty opposite. For you guys, as much of the work as possible is on building. In some ways the investing is, we’ll do as much of it as we can, but it’s competing with operating in a lot of ways.
Versus at Benchmark, everything is designed so that there’s literally nothing else to do but invest. The goal almost is that Benchmark wants there to be nothing for the partners to do but make investments and work with those companies. You want it to actually be very hard to make investments.
Delian Asparouhov
You want there to be lots of things that compete for your attention that aren’t investments, that are better things to do than investing.
Trae Stephens
Yeah, it’s like being professional at anything. If you’re a professional baseball player, you’re going to hit a lot in the batting cage, you’re going to take a lot of ground balls, because that’s what you’re being paid to do. You have to do it. So if you’re a professional investor, you’re going to feel like, “I have to be doing deals all the time.” Now, the challenge is having discipline around that, because sometimes the right decision is not to invest.
Jack Altman
By the way, one good way to implement that is by saying if you want to make an investment, you have to be on the board for the next 10 years.
Trae Stephens
I would never make an investment. I would just immediately come out. I’m out. You’re going to pay a 1,000X multiple on revenue, and you’re going to be on the board for 10 years.
Jack Altman
And you have 2,000 hours of board meetings.
Ev Randle
Delian, you said something recently that did really resonate, which is that the investing side of things should almost be an exhaust. You see it almost as a side hobby, or an exhaust of what you already do in your operating day-to-day job at Varda.
Even as a mental framing, I was trying to frame that for myself. In some ways, depending on the stage of investing that you’re doing, you could see your job as: I’m a researcher and a networker with this group of people—let’s say you’re in cybersecurity, cybersecurity experts or whatever—and then the exhaust is that once in a while I do an investment.
Delian Asparouhov
There’s this investment that we were just talking about earlier—that you guys met this week, that we just led the seed round on—that I think is a perfect example of this. Over the past couple years, I’ve just gotten deeply frustrated watching the Cognitions and Cursors completely transform how a backend software engineer at a Ramp or a Meta, or name-your-favorite-software-company, significantly improves their productivity.
You just look at the sets of tools that we use internally at Varda for our firmware engineers, thermal modeling, et cetera. It’s the same stuff that I was using in 2012 when I was building robots, and there’s effectively no change. The software’s slightly updated, but there’s no improvement in productivity. So I just started spending a ton of time looking at all the tools in that category. Some are now later-stage companies in our portfolio. Nominal, I think, is one of these that was an early example of this.
The one that you guys met this week was an example where I was just so frustrated that there’s this very rote manual work that our junior mechanical engineers do that felt like the classic type of thing that AI should clearly be able to, not just slightly improve the workflow on, but just completely remove. I’d probably met at that point maybe 15 different companies that were in this space, frustrated by all of them. I met the 16th one and I was like, “Great, this is the one.” It’s just the best founder, the best product, the best idea, and I was like, “I want to invest.”
I’m partially doing it because I like doing the early stage, get 10%, start to be a little bit of that founder coach. But then I also specifically like it because it solves a problem at work. Another one that we did together way back in the day that solves a problem personally: we both did Eight Sleep together back in the day. It was not because we had some deep hypothesis on mattress markets or anything like that.
Jack Altman
You were just like, “I’m so hot in the middle of the night.”
Delian Asparouhov
I’m just so hot in the fucking middle of the night. And I’m just like, “I don’t want this to happen anymore, and this is a good product.” I remember at the time we were like, “We have no idea if this is going to be a good business, but it is a really good product.” Typically over time, good products turn into good businesses.
Trae Stephens
My memory is that I was actually crapping all over this deal. I was like, “Guys, this is so dumb. Why are we investing in a mattress company?” I think it was Keith that actually said—
Delian Asparouhov
Just try it.
Trae Stephens
I slept on it for one night and I was like, “Let’s frigging go. Let’s invest in this company.” But yeah, I think that’s totally true.
On the being-a-professional-investor side of things, if you look at different vintages, like 2022, you could argue that actually the best investors were the investors that stopped investing. How do you deal with that, if you’ve tuned your whole system for building a track record by making good deals?
Jack Altman
I think there’s a couple. I wasn’t here, obviously, but there was a fund that had 2022 included in it. It’s hard to be excited about much of what happened in 2022, but it’s a three-year fund. It also has 2023 and 2024 in it. So one answer is that you’re consistent through vintages, you have some time diversification per fund, and you hope that the 2022 fund also includes 2023, and you get through some of it that way.
Trae Stephens
Brian Singerman used to always say that venture is not a macro asset class, it’s a micro asset class. As long as you’re in the right companies, it doesn’t really matter. But if you look at the data around the 2022 vintage, or the window around that, it’s massively underperforming other vintages. I think there’s some truth to that, and then there’s also some risk.
Concentration and conviction
Jack Altman
There is. There is risk to it, but unless you’re Peter Thiel, trying to call the market moment is also very risky. Take public market index investors. People who just buy and hold the S&P destroy people who try to time it almost every time. So if you’re not one of a small number of investors that you guys might have on your team, you can mess it up in a big way.
There were a lot of people who were like, “Ah, 2017, this is a bad time. We’re at the top. Don’t invest for the next few years,” and then you miss some good companies. Or in 2023/2024, if you still don’t have your foot back on the gas, and you can miss it. So my counter would just be that most people can’t call the macro very easily. If you can’t call the macro very easily, and then you miss the good vintages you miss it all.
Trae Stephens
If you can’t call the macro, there’s at least enough data that you should be able to remember prior cycles, which it doesn’t seem like we have the ability to do at all. Memory is so short. 2022 is not that long ago.
Jack Altman
It was four years ago, and everyone’s like, “Man, we can’t do that again.” Maybe if this happens in 20 years, our kids won’t remember. But it was like 34 months ago. It’s crazy.
Ev Randle
A core investing memory of mine was actually an off-site at your house, Trae. This was peak 2021. The crazy thing was that the consensus in the room was that the bubble was about to pop. One, that we were in a bubble, and then two, that it was definitively about to pop. The conclusion we came out of that off-site with was—well Peter wanted us to go to the beach for a year and do a sabbatical—let’s invest in our very best companies. We made two investments. I won’t say them, in case that’s weird. But we did two investments, and both of them have been, even vintage-adjusted, really good.
So to the micro versus macro point, the winter of 2021 was probably the worst possible time to make an investment in the last 15 or 20 years in venture. Yet if you still get the micro right… Maybe you would’ve made more money if you’d done those investments in the summer of 2022, because you still probably could’ve done them, they’re still really good investments, both of them.
Trae Stephens
This is the nature of Peter. No matter how annoyed you are with him, he’s always right on some timeline. He’s always right. I can’t tell you how many times I’ve been like, “Oh, this is so ridiculous. He’s getting so emotional about this thing, and I don’t understand why he’s so charged about it.” And then 12 months later, I’ll look back and I’ll be like—
Ev Randle
“Shit.”
Trae Stephens
“Wow. Can’t believe how well he nailed that.”
Jack Altman
But I think it’s really dangerous to try to do this if you don’t have someone like that.
Trae Stephens
Yeah, if you’re not Peter, I couldn’t do that. There’s no way I could.
Jack Altman
I think almost nobody can do that.
Trae Stephens
I think there’s something really unique about him. He understands that it’s less important to be exactly true than it is to be directionally true. Because it’s really hard to shift the Overton window. Most people try to moderate their approach to something, and because of that, they’re not directionally pulling hard enough to shift any strategy.
Whereas Peter is just like, “I’m going to dig in my heels on this,” just to move the window a little bit. That tiny adjustment that ends up being made is really powerful. I come up with this all the time with Anduril, with our executive team. Everyone wants to be reasonable. I feel like I’m constantly coming in over the top being completely unreasonable. I think it’s just a learned—
Jack Altman
Not actually wanting the outcome that you’re saying.
Trae Stephens
Right. It’s just a learned behavior from Peter. I’m going to say something where even if I don’t believe that we should do exactly what I’m saying, I’m just trying to pull people out of the unintentional stasis that they’ve locked themselves into.
Delian Asparouhov
The quote that I remember from that off-site was basically Peter saying something along the lines of, “I know that you guys are going to want to deploy. There’s going to be people on this team that want to make investments right now because the category’s so hot. Everybody thinks everything is the best, the bubble’s about to pop. So the rule is, go figure out what the two or three best companies are. We’ll invest at whatever the market-clearing price is, and the bubble will pop. But because we’ve invested in the truly best ones and gone hands-off on everything else, the best ones will make it out the other side.”
So one of those investments was the $9 billion post Ramp round. Nine months later, the company did a down round at basically $6 billion in valuation. So in theory that late-’21, early-’22 investment looked bad. The company’s now obviously at a $40 billion valuation, and so even the IRR with the down and up still made sense.
So I think it was Peter pushing this: “You guys are going to deploy. I can’t stop you. Even if I wanted to stop you guys from fully doing it, it’s not going to happen. So therefore at least just do the very best ones and then stop.” And then we did feel good, because we made two really big investments and we’re like, “Okay, we can stop now.”
The state of venture
Jack Altman
Is there any rising discomfort in Founders Fund at this current moment that’s like 2021, or is that not the tone?
Trae Stephens
I think it’s probably different across different people. I am very uncomfortable. I am not enjoying this moment at all. Maybe some other people are.
Jack Altman
Why aren’t you enjoying it?
Trae Stephens
I feel like we’re getting back to this point where the prices are untethered from reality, and it reminds me a lot of 2021.
Jack Altman
Obviously there’s AI, but I know you spend a lot of your time in hard tech, which is now probably just as hot as AI, it seems.
Trae Stephens
Totally, yeah, if not hotter in some ways. I feel like there’s almost been a bit of a reset, particularly on vertical AI SaaS. People are like, “Oh, crap, the labs are going to do a bunch of this stuff.” So maybe there’s even been a bit of a reset there, but there hasn’t been any form of a reset on hard tech.
Founders Fund for over a decade has had this Founders Fund Science concept of, “we wanted flying cars and instead we got 140 characters.” We’ve probably been more active than anyone in investing in these hard tech companies. Did they work out? They did not, for the most part. They did not work out. Most did not.
Jack Altman
But you got the ones that mattered.
Trae Stephens
Yeah, we got the ones that mattered. But if you think about where most of those gains came from, it was actually the PayPal mafia. It wasn’t a thesis that turned into great returns. It was a network that turned into great returns.
Jack Altman
It’s funny, because most people in your position would obviously say, “Oh, our thesis played out, and yeah, we had some losers, but we had Anduril and SpaceX, and so it was basically right.” But you’re saying that actually the thesis didn’t really serve us, and we would’ve made those investments anyway?
Trae Stephens
And I think they were actually hard to make even with the thesis. Were it not for Luke Nosek basically putting his career on the line, I don’t know that Founders Fund would’ve done the early SpaceX investments. When I pitched the Founders Fund team on Anduril, they were basically like, “Who’s running this company? It’s not you and it’s not Palmer, so you need to come up with a better solution.” So these things weren’t even obvious even when they were direct network.
So I think on the thesis side of this, we’re just way more careful around it now than we were historically. It’s not sufficient to have a really strong technical founder. You have to have somebody that’s equally good on the business side, otherwise it’s just not going to work. What we’re seeing a lot of today—whether it’s hard tech or defense tech or manufacturing, industrial automation, whatever it is—is that there are a lot of people with great ideas and interesting technology, but significantly accelerated rounds relative to the commercial progress. The belief I guess is that if the product works, the company will just obviously work. I don’t think that’s actually obvious at all, that the company will work.
Delian Asparouhov
I admit I don’t mind the current moment as much, but a part of it is that probably 50-60% of the investments that I do are these sub-$30 million post rounds. Sometimes some hot founder from some XYZ spin-out comes and says, “I have offers at $75-100 million post,” and I’m like, “Great. That’s just not the business that I’m in, and you’re welcome to go raise that.” So I mostly ignore it.
Not to speak on behalf of the entire firm, but maybe if I were to give the closest that we have to the 2022 take, it’s that there are a set of generational companies that we’re involved in—OpenAI, Anthropic, SpaceX, Anduril, et cetera—that are all now largely priced in, so late-stage, all IPO-ing soon, that it’s not very obvious what that next cohort of companies is that could even accept a billion dollars of investment right now.
So I think that’s probably the biggest question. We’ve had a pretty wild deployment rate over the last three, four years, especially with some of our growth checks. That maybe won’t be as possible over the next three or four years if there aren’t logos like that.
Trae Stephens
For Delian to make these investments at under a $30 million valuation, he has to spend a lot of time at preschools, finding young founders who are willing to accept such low valuations for their seed round.
Ev Randle
Is that why you’re not allowed within fifty yards of a school? I’ve always been wondering why that was the case.
Jack Altman
We’ll run this segment by your PR team before we air that. It’s a little bit on the line for me.
Delian Asparouhov
It’s not even close to the line for FF.
Trae Stephens
The idea of a $30 million seed round valuation is, “Oh, wow, that’s reasonable.”
Jack Altman
It’s reasonable.
Trae Stephens
It’s 3X what reasonable would’ve been ten years ago.
Jack Altman
It was so recent that it was unreasonable. Five years ago people were like, “Whoa, these YC companies think they’re raising at $25 million? Who do they think they are?”
Delian Asparouhov
Like, I did a $6 million post seed round two months ago.
Jack Altman
Six post? Wow.
Trae Stephens
I think the thing that the founders are missing in this equation is that the economics at these different prices at the early stages shift significantly. There’s so much money in venture. There are all these new funds that are setting up. It’s way too aspirational. There’s way too much money in the system.
Jack Altman
Forget the new funds. It’s really just the mega platforms consolidating $40 billion or whatever.
Trae Stephens
Yeah, even that. Because those terms are shifting, the deals that you could do 10-15 years ago that led to these hundred-baggers—
Jack Altman
They’re gone.
Trae Stephens
They’re totally gone. Then the question is, can you actually, over a venture fund cycle, beat the S&P 500 with this new unit economic, even if you’re hitting the monopoly players? I don’t know. It seems like it’s way harder than it has been historically.
Jack Altman
I think it is way harder than it has been. You’d expect multiples across the industry to compress, but the power law is still super strong, obviously.
Delian Asparouhov
One thing we’ve noticed is that in our latest venture fund, which we kicked off maybe six or seven months ago, the average entry price right now is something like $700 million in valuation. If you compare it to the prior five, those would probably be on average—because we do some mid-stage rounds in there—but it would probably be like $80-100 million entry point average price. It probably speaks to the fact that even the seed rounds we’re doing are so high, and then you add in some of the mid-stage things that are also high…
Jack Altman
We don’t talk about this much. What’s your current comfort with the temperature in venture and funding and everything like that?
Ev Randle
I’m also deeply uncomfortable. The similarity to 2021 that I draw is that in 2021 there was this unanimous sense… Because in 2021, investors were making a lot of money. There were a lot of things that were IPO-ing. Everyone was getting a great cohort in 2019, 2020, and 2021 of really big IPOs that gave everyone a big taste of liquidity, at the same time that everyone was getting markups very quickly and very high markups very rapidly.
There just started to become this sense where you’d have a company that would be at, say, $100 million of ARR, and it would get priced at $10 billion. So you’d have this 100X multiple on it, and the investor would be like, “But obviously it’s going to go 100 to 300, and then 300 to 700, and then 700 to 1.4, and that’s going to be worth $30 billion, and I’ll make a 3X in four years or whatever.” It was just this sense of inevitability without really digging deep into the immense difficulty and the base rates of how many companies at $100 million of ARR ever make it to $1.4 billion of ARR, in that arbitrary example. But there were a lot of situations that looked like that, where everyone was like, “Of course it’s here, and the trend’s good, so it’s going to be there. And wow, if it does get there, it’s a 3X from where you’re investing.”
Historically, the amount of companies that have gone from here to there is 5% of all companies. So you’re over-underwriting relative to historical norms by 20X.
Trae Stephens
Not to mention all the dilution along the way.
Ev Randle
And you’re not incorporating dilution in that equation. So you only have these situations when people are making a lot of money. I also think this duality is the key to a bubble. These last three years unquestionably are going to be the period where venture capitalists have made the most amount of money in history. Anduril, SpaceX, OpenAI, and Anthropic are going to produce the most returns that we’ve ever seen, probably by two orders of magnitude, of any batch of companies.
But then when you strip out those four or five or six companies, you start, one, playing with house money, and two, you start assuming that everything’s going to be daisies and roses because of the success that’s been baked in other parts of the market. I’m feeling that investors have that feeling of inevitability, where it’s like, “Yeah, I’m going to invest at $5 billion, but it’s going to be $15 billion in two years, and then we’re going to IPO or we’re going to sell it at that price.” We’ve forgotten how hard, how brutally hard, it is to go from one space to another.
Trae Stephens
Even if you go down from the $100 million ARR line, one of the things that I’ve noticed in my 12 years doing this is that there’s this weird glass ceiling between $20 and $30 million, where a reasonably managed enterprise SaaS company is going to get to $20 or $30 million in revenue. But then getting through that glass ceiling is super freaking hard.
Jack Altman
Yeah, the percent that make it from $20 million to $100 million is a huge drop-off.
Trae Stephens
Huge drop-off. Early on in my time in venture, I would meet with these companies and think, “In three years they’ve gone from zero to 15 million in ARR. That’s a really good trajectory. We should dig in.” Now I’m just like, I don’t know.
Jack Altman
It’s tough because the way VCs are underwriting is forcing founders at $1-5 million of ARR to tell a story that they’re going $2 million to $30 million to $150 million, all the time. And then it forces people to build companies in weird ways. It’s all distorting.
Delian Asparouhov
I think what’s different about this cycle relative to ‘21 is that in 2021, not even the best companies were able to beat Google’s early years of revenue growth. And then Anthropic, in this cycle, is literally the first company since Google to actually show an even steeper curve. That feeds into both the “hey, you’re playing with house money, there’s so much liquidity coming,” but then also the assumption that anybody else will now be able to match that same revenue trajectory.
There are things even in our portfolio. Cognition is obviously sounding like a wild growth rate. Then I think it just goes down the stack, where people assume even the seed-stage $1 million ARR company is going to be able to follow that.
Jack Altman
Totally. But it goosed up a lot of businesses that could raise at a quarter of the valuation that they’re raising at. But because the market’s where it is, rather than just saying, “I’m going to go from 1 to 4 to 11…” There is a way to venture-finance that business. It just looks different.
Trae Stephens
How do you guys respond to the pitch at a seed or a Series A or a Series B, where the founder is super confident that they’re going to go from 1 to 10 to 100 to 500 over the course of three years?
Ev Randle
The running joke is that it doesn’t matter what market you’re in or what company you’re running. At the end of your plan, for any company under $5 million of ARR, this year’s 30.
Jack Altman
It’s always 30.
Ev Randle
It’s always 30. You see a deck, and it’s like, “We’re at four.” You say, “Are you going to go to 30?” And they’re like, “Yeah, how’d you know?”
Jack Altman
It’s crazy. We had nine of them in a row.
Ev Randle
There’s a little bit of “don’t hate the player, hate the game” in there as well. You have Hemant on a podcast being like, “If my partners bring me something that’s growing 3X, I tell them don’t bring it to me.” The new bar is 10X. That’s nothing against Hemant. That’s where the market is. Obviously, it’s a game of opportunity cost, not absolutes. You’re trying to do the best thing in any given year, not something that just meets some bar.
So part of being a good founder is being able to raise capital and play the capital-raise game well. But I think you do have to mix that with a large amount of being genuine as well, and you have to ideally have worked out the idea maze in order to show a path to actually doing that. It can’t all be bluster.
Jack Altman
Yeah, and I think also, our model is that we’re going to work closely with them for a long time. That’s just what it is. So it’d be very frustrating for us to work with somebody who says, “My plan is 10 to 100, but I’ve got no plan beyond how I’m going to manage it. I haven’t thought about what team I need to hire. I haven’t thought about what that product service area looks like. I haven’t thought about…”
If you’re going to take on the role of working with this person, in order for that to be an enjoyable plan as you work with them on it, you’d want to believe that they’re going to think, “Okay, I actually know what’s going to go into it, and it’s going to be really hard.”
Trae Stephens
This is sort of Peter’s quad chart from Zero to One around indefinite optimism versus definite optimism. If you’re going to be optimistic, you should at least have a plan.
Jack Altman
Yes. It’s too frustrating otherwise.
Ev Randle
I feel like you would probably be more allergic to any version of that, Trae. I just remember, there was some company we were meeting back in the day, and they were like, “I’ll be coming from a dinner with Marc Andreessen.” You just email back, “Don’t do that. That’s not a game that we should be playing.”
Trae Stephens
Oh, the competitive game of “look at me, I’m peacocking all of my connections.”
Ev Randle
Yeah. It’s “Oh, you guys better move quickly because afterwards I have a dinner with this famous person.” You’re like, “This is just founder-to-founder feedback. Don’t do that.”
Delian Asparouhov
Sounds like Trae.
Trae Stephens
I don’t remember doing that, but it sounds like something I would say. The thing that I feel is so triggering to me about these crazy plans is not the optimism. I believe that they should have the ambition. It’s the willful removal from all historical examples.
Jack Altman
It’s an easy time to be like, “Things are different now. The rules have changed. You can have billion-dollar companies with four-person teams.” I think you can convince yourself that just because the rules have changed, there are no rules at all anymore.
Trae Stephens
You can kind of do that with software, but with hardware, you have to make real things. This idea that you can 100X production in a 12-month period… I’m sorry, you literally can’t. I’m just going to say that.
Hard tech and US chip fabs
Jack Altman
Okay, I want to go to a hardware topic. Before we all sat down, we were talking about chips. I’ve spent some time with some public-market investors recently. Everybody knows that people love semis right now. People don’t like software, they like semis.
We obviously have huge compute shortages. We have the situation with Taiwan and China. We’ve got America without its own real ability to fabricate anything. There’s a lot of startups working in chips and semis in various ways.
I’d just be curious to hear your lay of the land on the situation, where we are as a tech industry. Maybe then we can get into country positioning and some of the broader things, and maybe that connects to Anduril and whatever. But maybe just starting with chips as a point in time, what’s the lay of the land? What are people excited about? What are we missing?
Trae Stephens
I think the biggest challenge is just around fabrication. There’s a lot of energy going into chip design, making more efficient AI chips, or more leading-edge chips, or whatever it is, for memory or for inference. There are all of these different things that people are working on, and that’s great. I’m glad that we’re doing that. But at the end of the day, these chips are all coming from the same place in terms of where they’re being made.
For decades, the United States was really the only player in semiconductor manufacturing. That’s actually where Silicon Valley came from, and we’ve completely lost our way with this. It almost entirely happens outside of the United States. To the extent that we were doing anything in the United States over the last couple of decades, it’s not leading edge anymore. It’s some of the more mass-scale stuff, like automotive chip fab and things like that, or low-scale research-and-development fabs.
We haven’t really figured out how to reshore the leading-edge fabrication. This is what the CHIPS Act was intended to do. It doesn’t seem like there’s a whole lot of traction coming out of the money that was spent there. The government has made this big bet on Intel, taking an equity position in the company to try to get them to the point where they can be not only a manufacturer of their own chips, but also a merchant supplier of chips to other people. Their yield rates are still really low. Lip-Bu is doing everything he can to level that up, and has the Terafab project that he’s working on with Elon now as well. So it’s our best bet, but not at all certain that it’s going to work.
The other companies that have leading-edge processes, like TSMC and Samsung, have been getting attention from the government to try to get them to build domestic fabs in the United States. But the incentive structure for those is totally wrecked. It’s not clear why they would build any of these fabs. It’s not clear that the talent even exists to do it domestically anyway. So I feel like our focus has just been wrong.
I will say this, and I actually mean it: we’re focusing on the easy part, design. I’m not saying it’s easy on an absolute basis. I’m saying on a relative basis, we’re focusing on the easy part, and there’s a lot of money to be made there. But if we don’t fix the hard part, none of this is actually going to work.
Delian Asparouhov
There’s not even an optimistic plan right now. If you were to just look at all the chess pieces today and how they’re playing for the next five years, there’s nothing that would indicate it would be any different five years from now. In 2015, we lost access to space because we shut down the Space Shuttle and were dependent on the Soyuz for taking people up to space. At least then you could be like, “Elon is trying.” There’s a path to something where those will be brought back, and by 2020 we have it.
If you were to analyze what everyone is doing today and say, in 2030 or whatever, are there going to be leading-edge fabs in the United States at scale? No. There’s not even anything that anybody’s working on that would lead toward that today.
Ev Randle
Over the next decade—and if you haven’t thought about this, no need to give an answer—what do you think the probability is that Taiwan is quarantined or otherwise?
Trae Stephens
I think it’s high probability that it’s not what it is today. It could be a Hong Kong-like situation, or some sort of détente where the United States continues to have access to the critical supply chain and we’ve given up something in that arrangement. Or it could be a full-scale invasion where we have to make a really hard decision about whether or not we want to actively support Taiwan. One of those outcomes is wildly probable. I think it’s +90% in the next ten years.
Jack Altman
I feel like your view must be something along the lines of, we need US chip fabs, right?
Trae Stephens
We need to create a credible deterrent threat for a full-scale invasion of Taiwan. That’s the Anduril side of the problem that we’re focusing on, which has led to our co-founders being sanctioned in China. So that’s that. On the other side of it, we need to figure out some way to reshore the manufacturing credibly, with a real plan.
Jack Altman
Is it basically impossible to start a new company? Are there any entrepreneurs with any amount of capital who could just start a net-new chip fab, or is it too big of a problem?
Trae Stephens
I wrote this article about choosing good quests. I made some joke about how celebrities monetize their brand by selling consumer goods, tequila or popcorn or whatever it is. It’s a commodity. They just slap their brand on it, and they make a bunch of money. Good for them.
Our version of this in Silicon Valley is starting a venture fund. All these people who have the ability to raise gobs of cash and do really hard things instead sit on the roof like Big Head from HBO’s Silicon Valley, raise a venture fund, and piddle around with AI. We really need them to be founders. We need them to be heroes. It’s a big ask, but I’m shocked that no one has taken it more seriously, because it feels like the most important thing that people could work on right now.
Jack Altman
What does it take? Is this Elon plus $100 billion? What actually is required for somebody to start one net-new?
Trae Stephens
I think it’s Elon plus tens of billions of dollars. I don’t know that it would require hundreds of billions, but definitely in the mid-tens to do it. There’s all sorts of levers that you have to pull. You would have to get the US government involved in pulling you to the front of the line with ASML to get the EUV. You would have to figure out the supply chain stuff. You would have to build out the process, which is literally thousands of tiny steps that you have to get high yield rates on so that you don’t fail out as you make them. That’s a talent problem. We don’t have enough skilled labor that understands how that works.
It’s really hard, because it’s death by a thousand cuts. It will require someone who has the ability to manage that complex process as a leader.
Jack Altman
You should incubate it.
Trae Stephens
As I said, I’m not a good manager, so I’m not sure I’m the right person to do it.
Jack Altman
Delian can help.
Delian Asparouhov
I need to get Varda a little further along before I sign up for another one.
Jack Altman
You can do two at once.
Trae Stephens
But I think there are ways that we can shortcut some of this system. We’ve invested in a company called Substrate that’s building a new form of lithography that simplifies the process for the rest of the steps involved. It’s really important that things like that get pushed down the road so that we can actually pull this off. But this is civilizationally important. It’s not just a technology problem, it’s a geopolitical problem.
Jack Altman
I was going to say this before. The China-Taiwan thing on some level feels like we’re all willfully looking the other way about it. If we’re to believe what you said—that it is wildly probable, and it’s a big problem, and we have no plan—we’re all just “That’s all right, just keep doing…”
Trae Stephens
Yeah. Every time I meet with one of these chip-design companies, I think of them as pinch runners. They’re like, “All right, check this out. I’m really good at stealing second base.” And I’m like, “Yeah, but how did you get on first base? Who hit the ball?” They’re just really focused on that pinch running—”I’m the greatest base runner in history”—but they don’t really know how to solve the problem of getting the bat on the ball. I feel like that’s what we’re doing over and over again with semiconductors. We’ve solved for it with hundreds of pinch runners, and we have nobody who can actually make contact.
Delian Asparouhov
Trump also visited China this past week for the first time in basically a decade. Xi Jinping was extraordinarily clear in all of the communications that our number one problem and issue is Taiwan. He references the Thucydides Trap and says, “We’re clearly the rising power. You’re the decaying hegemony. The only way to avoid kinetic war is that you need to accommodate our needs. Our number one need that we’ve made very clear throughout this entire week is that Taiwan reunification is something that we deeply care about.”
If we end up letting that happen, it’s just a huge betrayal of all American values, where we’re the world police and you’re letting a Westernized democracy get taken over by an adversary.
Trae Stephens
And I think you have to take Xi seriously. When he says that, he’s saying, “I am staking my reputation as an authoritarian on my ability to execute on my plan.” He is held accountable by his populace for following through on the things he said he’s going to follow through on. No one in the West takes that seriously, because in our mind it’s a four-year cycle. Politicians come and go. They say stuff, and they very rarely follow through on it. So we have this tendency to just not take it seriously. But he has to follow through. There’s no other way for him to stay in power.
Jack Altman
We need some fabs.
Trae Stephens
We need some fabs.
Robotics and the China question
Jack Altman
Also in hardware—Vinod said this on the podcast, that robotics will probably be the biggest thing of all time. Obviously Elon’s thinking about this a lot with Tesla. Have you spent a bunch of time meeting with robotics companies? It seems like robotics would be something square in your guys’ interest. Are you guys looking at it a lot?
Trae Stephens
All the time.
Jack Altman
It seems like it should work, right? Why wouldn’t it work?
Trae Stephens
On some timeline.
Delian Asparouhov
We haven’t yet made any huge humanoid or generalized-model investment. But we do have a whole set of portfolio companies that use robotics a lot. Anduril has plenty of robots in its foundry. Hadrian, one of our portfolio companies, uses robotics all the time. We’ve made some vertical-specific robotics plays.
I think we’re a little more bearish on the humanoid as a venture category. We’re comparing it to Unitree in China and thinking, is there even an American ecosystem that can support something like this? The hope that at least I personally have is that humanoid robots play out somewhat similar to phones, where it’s Chinese hardware but American software, and that forces this détente that neither side is happy with. China would much prefer that it was Chinese software running everywhere. America would much prefer that iPhones are made in America.
Jack Altman
Did you like Jensen Huang’s point on the Dwarkesh podcast about—
Delian Asparouhov
He’s got to win.
Jack Altman
—how America should go win every market, including the Chinese market?
Delian Asparouhov
My view is somewhere in between. I don’t think we should be helping them, but I do think that some of the chip sanctions actually heavily incentivize them to set up their own fab ecosystem. So sometimes winning the Chinese market is a way to weaken them. I agree with him on that narrow point, but I do think it requires looking at it on a market-by-market basis.
I don’t agree that American venture capital should be funding Chinese companies, because that feels like you’re just strictly helping the Chinese ecosystem. Versus Jensen selling chips to China feels like you’re just giving them the opioid addiction, but in the form of chips, and it’s probably good for them to be addicted to us.
Trae Stephens
Provided that there actually is such a thing as a US company owning a market in China. This is not how it works. I think that’s probably the thesis that Tesla had as well, before they realized that BYD was just going to copy all their best stuff and cut the price. I think that’s the much more likely outcome.
It’s like the super naive ramblings of a monopoly-hungry American who ignores geopolitical reality. “Obviously, the right thing to do is just to do business with them because they’re mercantilist, and this is just going to work, obviously.” Until your IP is gone and they’ve taken the market that you thought existed for you and they’ve exported it to the rest of the world at a much lower cost.
Ev Randle
I have a hypothetical mostly for Delian. If there are Chinese citizens who aren’t CCP members and who deeply want to come to America or come to a Western country and build Western AI—build AI that enriches the Western world, away from a China that they might not want to be citizens of anymore, that they want to leave, that they don’t want to build for—do we have a responsibility, or even should we, invest in those very hypothetical companies? Or should we stay away from those teams that actually deeply want to contribute to and advance Western AI, but because of where they were born, have to go through different means to do it?
Delian Asparouhov
I think if somebody is working on national security technologies, which I’d argue leading-edge AI effectively is, it’s not an unreasonable argument to make that you should have to roughly abide by the regulations around that, which is ITAR compliance. At Varda, at Anduril, we would not be allowed to hire someone like that. If they renounced their Chinese citizenship, if they fully became an American citizen, and if there’s also strong confidence that the CCP doesn’t have some of their family members under gunpoint and can’t leverage that fact to get them to do things on behalf of China, then sure.
But the really difficult part is if it’s just an individual citizen that has made it over, China still controls their life. Even if they hate China and don’t want that to be the case, if the CCP has their grandma at gunpoint and they convince that person to fly back to China and then detain them, it doesn’t matter what that person’s personal beliefs are. What matters is the system that you’re interacting with.
Trae Stephens
The alternate version of this is suicidal empathy. Look, none of us want to believe that every one of the multi-billion Chinese people in the world is bad. I don’t believe that. I don’t think Delian does. It’s crazy to believe that there’s some racist tendency in having a concern around this. But there are people in our society who have that level of suicidal empathy, where they’re like, you can’t say anything negative about the obvious logical fall-through of investing in Chinese nationals with family ties back in the country and stuff.
Jack Altman
Do you think there’s a de-escalation path between America and China? It’s implied in a lot of this conversation that we’re just going to a tenser and tenser place over time. Is that just what happens when you have a rising power and you have to plan for that to be the case? Or do you think there’s any de-escalation path?
Trae Stephens
I’m certainly hopeful that there’s a de-escalation path, that there are diplomatic solutions and hard-power deterrence that will prevent us from going into war with one another. I think that’s the worst possible outcome. I don’t think the American people have any memory of what it’s like to go into great power conflict and the scale of destruction that would result from doing that.
Jack Altman
Also now versus the last one, I think it would just be a lot worse, right?
Trae Stephens
It would be devastating. We can’t afford to do that. So part of it is we have to hope that China has that same belief, that it’s unthinkable and that we should do everything we can to stay out of it. But they’ve also spent the last 50 years with a very intentional strategy to box the West out of the whole supply chain, to make their success inevitable. If you’ve read the book The Hundred-Year Marathon, it talks about this in terms of Chinese fables. There’s this story that gets cited a lot in Chinese fairy tales, essentially: hide your strength, bide your time.
In Western literature, we have this tradition of chest-thumping, where it’s all about showing strength. We expect that other people will do the same, because we’re like, “Show us what you can do, and we’ll show you what we can do, and then we’ll decide how that gets litigated.” Whereas the Chinese strategy is always to cower. “Oh, no, we’re so weak. We don’t have any power. Our GDP is so low.”
They’re just going to keep repeating these messages until the moment where the flip is inevitable, and then we’re going to look back and be like, “Wait, we have no leverage.” We have no leverage. The Belt and Road just completely eviscerated our whole ability to control the supply chain. We don’t own any of the ships that are required for global trade. We don’t have any military advantages because we haven’t invested in it. We have no access to semiconductor manufacturing. And then they’re standing there with the entire deck of cards in their hands, and where does that leave us?
I think this is the fallacy about the moralizing around this, where it’s, look at all these things that the US government could do with technology, we’re so concerned about how this technology is going to be abused for all of these unethical use cases. Guys, if you’re concerned about a Western democracy, imagine a world in which we have no say in how these technologies are governed. We’re walking ourselves willingly into this situation. You think it’s bad now? Wait until you see what’s on the other side of a dominant China, because it’s not going to look like the effective altruist version that people believe it’s going to look like. It’s just complete banana pants.
Delian Asparouhov
The chest-thumping versus being quiet was exemplified the day that we basically started striking Iran. You had Trump obviously everywhere, thumping his chest. On that day, in the South China Morning Post, there wasn’t basically a single mention in any of the headlines about anything that was happening in Iran. For them, it was just, this is not a thing that we even pay attention to, categorize, or have opinions on. Xi Jinping effectively decided—I forget his exact quote that day—but it was something like, “That’s America’s issue, America’s challenge that they’ve chosen to step into.”
Whereas if you imagine the equivalent of that in the Cold War with the Soviet Union, all of the headlines within the Soviet Union would have been about, “Here’s what America’s doing. Here’s what we’re doing to counter it.” It was all this soft power through proxy wars, et cetera. China’s just not. Even though a lot of their oil comes from Iran, they’re effectively just not engaging in this proxy war. It’s the difference between being against a great power that is another chest-thumper versus the Chinese approach.
Lightning round: founder personality
Jack Altman
Fascinating. I know you guys have to go in a minute, so I want to end with a couple of light topics, just to cleanse the palate. One is our boy Sam Blond. He worked with you guys at Founders Fund, and then you guys led a prior round and we did a recent one. I’m curious—obviously we’re both really excited about the company—but what was your experience with him like as a teammate? Did that inform why you’re like, “I want to back him”? What was the experience there for you guys?
Delian Asparouhov
I worked with him a lot in the Miami office, where he was based. It was incredibly clear that he is just a phenomenal salesperson. He is somebody who knows how to close a deal, wine and dine, et cetera. Unfortunately, as we’ve talked about over the course of this podcast, the venture job is not really necessarily a sales job.
Once you’ve come to a conclusion you want to invest, there is for sure a little bit of sales that happens in terms of convincing somebody to take your capital, but the vast majority of it is: don’t try and close every 10,000th investment that is out there in the world. You need to discern which is the one to five that matter a lot. So I think that was partially why he even came to the conclusion after a year at Founders Fund. He was like, “This is not the world for me. I’m glad that I learned it very quickly. I’m going to go do the world that is for me.” Part of why we love Monaco is it literally just feels like Sam embodied in a company.
Jack Altman
It really does.
Delian Asparouhov
I’ve never seen it so clearly in a company. People talk about, “Oh, the founder personality shows up.”
Jack Altman
Usually you can kind of smell the founder everywhere in a company. This is—
Delian Asparouhov
This is something else. Monaco is Sam. Sam is Monaco.
Jack Altman
Yeah, totally. I think you guys just closed a big new fund. We talked a lot about discomfort. But then there’s obviously a lot that we’re excited about. I realize some of it’s for early, some of it’s for growth.
As you guys think ahead—and this is not Peter in 2022 or whatever saying, “Everybody take a break”, I expect you guys are planning to deploy it—is this going to the same type of stuff you’ve been doing recently? Or are you like, “This cycle has actually played out. We need to find the next trillion-dollar companies that we’re not currently in”? What’s the head space right now?
Trae Stephens
We have a separation between funds. We have the venture fund on one side and growth on the other. The venture fund we actually raised years ago, 2021.
Jack Altman
You’re still deploying out of the same fund?
Trae Stephens
What we did is we realized that we raised it, and it was too big. Again, the economics of these things are really tricky. So we ended up splitting the fund in half, and then we just backed them chronologically up against each other. So we’re currently investing out of the second of those two funds that we created. Then growth is focused almost entirely on companies in the portfolio already that are seeing continued growth. I think only about 20% of the fund is net new.
Delian Asparouhov
80% is doubling down.
Ev Randle
Was OpenAI already in the venture fund at one point?
Trae Stephens
No, that was first in growth.
Delian Asparouhov
But OpenAI and Anthropic are probably two of the very few that both originated from growth, and we’ve doubled down through growth. Almost every other meaningful position or check across the fund… Sometimes it’s in a totally different category. Crusoe is probably one of my favorite examples of this. It originally started off as a Bitcoin miner, and now obviously turned into a data center producer. But yeah, we’ve doubled down throughout.
Ev Randle
It’s funny, I was talking about this with the partners here. I was talking about the Founders Fund growth strategy and how this doubling down within your own portfolio being 80% of the portfolio can only work at a place like Founders Fund. Because it’s maybe the only firm that’s invested in the companies that have compounded at like a +30% rate per year.
Jack Altman
This only works when you have SpaceX and Anduril.
Ev Randle
For 30 years. You have SpaceX, you have Stripe, you have Anduril. There’s probably three or four more of them I’m forgetting. Like Palantir, Ramp. It was just as good of an incremental IRR at $30 billion as it was at $500 million 10 years later. That’s only if you’re collecting those and partnering with those types of companies early.
Jack Altman
It’s hard, because you have to be so rational to say, “We’re not going to just invest in our three best companies.”
Ev Randle
Yes. It has to meet the absolute bar, not just a bar within your own portfolio. And the Venn diagram of those also just happen to be the best investments at growth.
Delian Asparouhov
But then there’s the advantage of being involved from the early stages. The fact that it’s Anduril incubation, Ramp seed, Stripe Series A…
Trae Stephens
Palantir incubation.
Delian Asparouhov
Palantir incubation. It’s not to say that we have perfect access or perfect pricing, but we almost always are able to invest the amount that we want to. So if you just look at the sizing that we’re able to get into these companies, that’s also part of what makes the growth fund work.
Jack Altman
It’s got to be a little hard to match the time allocation to the dollar allocation, though. Because how much time do you need to spend deciding on these companies that you already work on? And then you’re going to meet a lot of these net new companies, but you’re only going to deploy 20% of the fund there. So it just has to be the case that a bunch of time’s going into net new stuff and you just do very little of it.
Delian Asparouhov
Why do a net new defense tech thing when you can just deploy another billion dollars in Anduril? It just doesn’t make sense to do that.
Jack Altman
Yeah. Thank you guys for doing this with us. This was super fun.
Delian Asparouhov
Of course. Thanks for having us.