AI 热潮将造成大量淘汰者：谁赢谁输？| David Frankel

The bubbles get bigger. This is the wave of our lives. Will there be roadkill from this wave? Oh my God, there's going to be a lot. Well, the problem is the mega-platforms are taking call options.

Pro rata is almost like the original sin. I have never seen secondary markets as liquid. Are we headed for another dot-com crash? Definitely. It is not a question of if; it is a question of when. Nobody knows.

David, last night you sent me a forwarded email, and it was my first-ever email to you 11 years ago.

11 years.

I can't believe that. Do you know what I found so funny? I just had dinner last night with Mamoon, and I look at the people who've been kindest to me, which is you, Mamoon, Josh Kushner, and Neil Mater. It's fascinating that the people who were there when there was nothing are also the greats.

I'm honored to be included in that list. But maybe part of the thing is that they're intoxicated, and you were intoxicating, in my view. You were 19 years old, but you were full-on focus and energy. You just brought it, right?

I think maybe part of the job and part of the fun of the job is recognizing that. It's not all it takes, but you had it. You have it.

1. The Seed Fund Squeeze

It's super kind of you to say. In terms of having it, obviously we both play at the early stages, and I've said before on social media and on X that the hardest part of the market is seed, in many ways. The worst-performing funds will be the $50 million to $100 million funds.

I say this to explain because you're too big to be collaborative, to write those $100K to $250K checks and be a friend, but you're too small to lead an $8 million to $10 million seed round. Why am I wrong, and why will this vintage be great for those funds?

Okay. There is so much to unpack here. You've got this narrowing out in venture, where the bigger you get, almost like it becomes a pyramid. I think there is the business of venture, which is asset management and this channel, right?

So you've got Cambridge Associates and the fund of funds, and all they're doing is selling access. They're fine with it. If, you name it, the top 10 or top 5 names are not in XYZ great company, it's much harder to sell. I would say at this point, if you're not in the top 5, if you missed the $3 trillion companies, you're much harder to sell. So it's not trillion-dollar robust, but if you look at the numbers over the last 25 years of how many companies were created that are over $100 billion, there were fewer than 100 companies over the last 25 years—fewer than 100—that are sustainably over $10 billion.

At that top end, you've got to be in that. The median company—and we've done a lot of work on this very recently—of the top 500 companies created in the last 25 years is $2.6 billion. Now, if you own 5% of one of those companies, you return the fund each time.

I would say what's gone on in seed is that there are a whole bunch of unreasonable bets being taken with loads of funds and loads of money. It's quick because you've got to get the check in, because you've got to get to the next fund. So it's incredibly tough at seed.

What makes this still a great business is a little bit of what I said about you: you can wait and wait and wait and wait if you're patient, and then you just see someone, right? You see a founder or you see a team, and you just go, “I have to be there.” To me, that's code for, “This may just be another Uber, another Suno, another Shield AI.”

A little bit of this, answering personally, is a drug. Finding Harry, right? Finding that is a bit of a drug, so I'm addicted, if that's the case. If you're in early, you still have a chance of returning a fund.

I think it's a totally different business. Do you have to be the full $8 million? Definitely not. We can't be. Can you write a $3 million check? Can you write a half-million-dollar check? Now, valuations and uncapped notes are changing the business, but you don't just have to do that.

So if you're on-piste—and I've been doing this for nearly 18 years—it was always expensive. It was always tough. But you find some of the best people off-piste, always.

2. Big Seeds Become Insurance

On those rounds, when you look at the $8 million to $10 million rounds, or the large seeds that we see today, are you able to participate with the $2 million to $3 million when the multi-stage products provide such an efficient seed product that you might get $100K, but a $3 million check is much harder? Are you able to even do that strategy?

I hesitate to say this, Harry, but I think we're being seen—and I could be over-extrapolating from the last 20 deals that we've been involved in—almost as an insurance policy, where we're side by side. We're putting in $500K or $1 million, and there's been $8 million or $9 million going in.

From smart entrepreneurs, there's almost this knowledge of, “They may abandon me, and then having FC in my back pocket could be useful.”

Sure.

“And I'll use their brand, right? I'll use their distribution network to go out and say, ‘They actually don't suck. We're not doing $10 million in ARR yet, but they're more patient.’ And we'll be patient, and we're the testimonial salesperson.”

So I think there's some recognition of, “Wow, for $500K or $1 million, not a bad insurance policy.”

By the way, we're not doing that many $8 million rounds. We're still finding $3 million to $4 million rounds.

Are you?

Yeah. The valuations there move a lot, right?

Yeah.

They change a lot. By the way, the other thing is there's very little evidence yet that these hot, hot AI companies that are raising huge amounts of money are capital-efficient, right? They're anything but capital-efficient.

The jury’s out on whether that’s going to work still.

Totally get you. Just before we move away, you said it’s not in the hot, hot, hot. You often don’t get paid for being a value investor, and you can sometimes be criticized for being smarter than the market or, whatever contrarian term you want to use. My question is: do you think about whether this is an asset that will get financed in future funding rounds? If it’s not in AI and it’s a traditional enterprise HR company, I can’t get that funded for a good Series A. Does that impede your thinking on whether you’ll do the seed?

3. Every Startup Is AI

Well, everybody’s AI, right? It’s almost like saying that you’re not AI today is like going, “I’m not using the internet.” Why wouldn’t you use the most contemporary tools?

Everyone’s AI. You’ve just got different approaches. You’ve got a second-time entrepreneur who goes, “I know this domain really well. I’ve been doing SAP consulting for 10 or 20 years, and I’ve built a platform, but this part still sucks. And I was playing around with Claude Code.” This is a real situation. “I was playing around with Claude Code. My CTO is unbelievable. We’re now putting 4 or 5 together. Would you be involved?” And it’s a $20 million cap.

Sure.

Right? We see loads of that. So the concept of it’s only a 20-something—I’m not saying we don’t do that. You know very well we do that all day long. But we also see other startups in places that feel off-piste, and then you look at it and it’s worth tens of billions of dollars this time.

The statement that is said to me more than ever is, “Price matters less than ever because the only thing that matters is that you’re in the true winners of the day.” How do you feel when you hear that?

The scale of how much you have to win is different based on your price. It’s pure math. Uncapped notes suck at the seed stage.

Yeah.

I’m not saying we’ve never written one. Unfortunately, I’ve written one, and I think the founders are exceptional. I think they’ll do great.

Do you regret it?

I don’t regret it at all.

Yeah.

Because I love the relationship. But financially, will we do as well there? That’s going to be $100 million to $300 million priced when it happens. Now, you’re in a year in advance and you take that price. From a venture perspective, it doesn’t make much sense.

No.

And access is being sold. The Ivies—Stanford’s done this forever, but MIT and Harvard are doing the same thing. It’s like you just want to be there, right? Sometimes you really have to think that through, and we’ve said no plenty there. By the way, we’ll probably regret the ones that we said no to.

4. Founders Need More Than Funding

We see Y Combinator really professionalize startup founding in a way that turns it into almost a norm for people leaving some colleges in particular, and some programs at certain colleges in particular. Do you worry about how easy it is to be a startup founder today in terms of that normalization of it and what that means for what we do?

I do.

Yeah.

I think there are so many founders, right? It’s like du jour. I think there are fewer entrepreneurs, and when the tide goes out, everybody goes, “I knew. I told you so.” And nobody knows when the tide goes out.

But what it takes to be an entrepreneur is very different in terms of fortitude, the ability to energize, and the ability to go up that learning curve. The number of times I’ve seen the difference in the trajectory between the CEO and the CTO: the CTO, at some point up to 50 people, you’re golden. Then at some point you go, “Actually, we could bring in better technical skills.” If you’ve got a good co-founding CTO, that person becomes like a Swiss Army knife and is deployed in different ways.

The CEO goes on this serious journey where the learning curve is steep, and they’ve got to learn to manage, and they’ve got to learn to put bums on seats. I think of people like TJ at Pullpak or Jack at SeatGeek, and they’re changed individuals.

I had coffee a week ago. We had an hour with Mikey Shulman from Suno. I said, “What are you doing?” And he said, “Thirty to 40% of my time is just recruiting.”

I had lunch years ago, decades ago, with Jeff Bezos. I was invited to a lunch, and someone smarter than me said, “What do you spend your time doing?” And he said, “50% of my time is bums on seats.” That’s never left me. That’s the CEO journey. That’s the entrepreneur’s journey, and there are many founders that don’t cut it.

I think one of the biggest mistakes that I see investors make, though, is when they turn down a company because they don’t like the other co-founder. The truth is, the other co-founder most often isn’t there in 3 years. You don’t like them because you don’t think they’re good enough and not as good as the CEO.

Will you invest if you think the CEO’s amazing, but you don’t think the CTO is up to scratch? Or the head of sales, who’s also the co-founder, isn’t as good? What are your lessons on that?

Rarely. We do that less, and I think your logic is correct. But so early on, we’re looking for this package. I’m looking for this CEO-CTO kind of magic. In some ways, I literally use that word. I’m looking for the CTO to be a bit of a magician and the CEO to be a good salesperson. That’s my favorite combination.

I agree with you. The CEO being a good salesperson and being a real entrepreneur is actually more important because the CTO role can be fungible, depending on how complicated it is. But I have said no more times than yes in those situations, and I regret some of them.

The dynamic between those founders matters. Early on, I look at the dynamic, and in some ways I think I want to replicate the partnerships that I’ve loved and go, “I’m looking for some kind of alchemy here.”

You don’t have to be identical. You don’t have to finish each other’s sentences. In fact, I prefer that you are different. But how aligned are you, and how much do you trust each other’s competence? In a career, I’ve seen alchemy maybe one time—five times, 4 or 5 times. But when that alchemy happens, it’s because of that interplay between those 2 people. So I’m watching that pretty carefully.

Has the type of founder that you like changed, especially in the last few years? I think our team has definitely oriented much more toward deeply engineering-specific people who come out of DeepMind, who come out of Gemini, or—has that changed?

Less than you’d think.

Mm-hmm.

I would say the youth, the energy, the focus, the smarts—you put that package together, and it’s an intoxicating package.

Mm-hmm.

I look at experience and go, “What are we going to need to package with that experience?” There are certain situations—SaaS and enterprise SaaS certainly looked like that—where you’d learned the lessons. You understood the market. You understood who the buyers were, although that’s very fluid too. But did you have the focus and the energy?

I see these 20-somethings, and by the way, it’s a psychographic in a way, so I’m not saying that—I don’t want to sound ageist. The psychographic of that focus and intensity can last for decades, but there’s something about it at that early stage that is just, “Wow, I want to be part of that.” And that still turns me on a lot.

The theory of the relationship, et cetera, going one more time is great in theory, but man, to go on this journey, you need so much energy.

5. Triple Triple Double Double Still Works

When we look at the scaling journey, and we talked about how founders have changed—or not changed—for you, one thing for me that’s changed, and I get in so much trouble for this—VC Brags, this Twitter account, killed me for it the other day. I very candidly said, “I turned down a company the other day that went from $1.5 million, and they were going to go to $5 million, and then they were going to go from $5 million to $15 million, and it’s just not enough anymore. It’s not interesting. I’m sorry. For venture, we have an opportunity cost of capital where we can deploy, and that’s not fast enough.”

Has triple, triple, double, double gone? Is that still a venture path in today’s landscape?

$1.5 million to $5 million?

$1.5 million to $5 million in ARR. You’re looking at this company going, “Okay, you’re going to be $1.5 million to $5 million, $5 million to $15 million, $15 million to $30 million. David, 4 or 5 years down, we might be at $70 million.” Is that still a venture pathway?

These 10-year funds are taking 18 years. The one thing you learn is loads of patience. It’s such an opportunity when people go, “It has to be $1.5 million to $10 million to $15 million,” and then reality sets in. Sometimes it’s twice as expensive and it takes twice as long.

Harry, we still own every last share in SeatGeek. That was an investment I made in 2010. It’s become one of the top 3 ticketing businesses in the world. It just takes a really, really long time.

Some of our greatest companies were showing tremendous promise, but that $1.5 million to $10 million to $20 million—I just think, are they executing?

The other side is, is revenue the only metric? Sometimes there is traction on dimensions that the market is not necessarily recognizing, but you’re an insider. The retention in that account is really good, and that one account is now spending 4 times what they spent a year ago. There are more DAUs, and—

There’s got to be traction. Frankly, a lot of what we do to try to tell an entrepreneurial story to get more funding is show the different dimensions of traction.

But I think this go-go-go overnight or you’re bust—I think there are a lot of orphans out there for that. Sometimes, frankly, I look at those funding rounds, and they’re called seed plus or seed extensions, and I go, “That may be the opportune moment.” When they’re being abandoned and they can’t get the capital because the bigger funds have moved on, maybe that’s the opportunity. It’s not what we’d really do, but I can see it as a capital-markets opportunity.

Do you remember Bullpen, where it was like their business—

Yes, of course.

—to do exactly those rounds?

Yeah, yeah.

I always thought that was an interesting business.

Yeah. I don’t know how they’ve done, but they priced those rounds. They priced them for bigger, bigger players.

I think the thing is, you’re so paid for the risk that you’re taking there. I mean, they really were aggressive in terms of ownership. I think they did ipsy, which was a big business.

Yeah.

And so you have one and it pays for the rest. Can I ask, when you look at this, David, you’ve been doing this for 18 years, and you hear people like me say, “Oh, 1 to 5, like triple, triple, double, double is dead. Is it really a home run if it’s, you know, we need $1 billion in revenue?” Jason Lemkin says on our show, “Billion valuation? Come on. That’s not venture anymore.” Is this peak bubble when you review the 18-year journey that you’ve had?

6. The AI Boom Creates Roadkill

The historical or anachronistic view on this would be: the bubbles get bigger. This is the wave of our lives. I feel that way, by the way. If I look at the internet, SaaS, mobile, and AI, nothing looks the same.

And will there be roadkill from this wave? Oh my God, there’s going to be a lot. You look at those stats of 500 companies, less than 100 over $10 billion in the last 25 years. How many times, Harry, over the last 11 years have you heard, “This is different. This is different”? It doesn’t mean that there aren’t survivors and companies that are going to change the trajectory of technology forever, and I think in OpenAI, Anthropic, and SpaceX, we’re seeing that already. These are the Metas and the Googles of our era, highly likely.

But wow, it’s Hollywood, man. Ninety-five percent are not going to be there. And it goes back to: why is seed interesting? I don’t have to be in V1.

If there were 5 companies that were worth $5 trillion with exits—if you look at SpaceX, Tesla, and Meta, that’s trillions of dollars already. You take then Nvidia. I think Nvidia started more than 25 years ago, but even if you look at the last 25 years, you can add Palantir to that and Palo Alto Networks. That’s about $5 trillion of market cap. And then the other 495 at a $2.6 billion average.

Some of those are—we hope everything looks like Shield AI and Suno. But if you have 5% of a $2.6 billion outcome, you’ve returned your fund. If you have a $500 million outcome, it’s incredible still. And that’s why I think seed isn’t dead.

I think seed is crowded, to some degree very commoditized. I feel commoditized. I’ve said this many times. I feel like brand and, in some regard, distribution—as in your portfolio and people saying nice things about you—get you to the table.

7. Mega Platforms Take Call Options

But if it’s commoditized, does price not just become the separator? And if price is the separator, the mega-platforms win.

Well, the problem is the mega-platforms are taking call options. Is this good for the mega-platforms? Is this good for the LPs, or is this good for the entrepreneurs? Well, probably for 95% of the entrepreneurs, it’s not good.

Why? You get more money at a higher price with mostly a more junior VC who will let you do your work and not get in the way.

Mm-hmm.

Isn’t that what all entrepreneurs want?

I mean, sounds amazing, right?

Doesn’t it?

Yeah. The more junior entrepreneur moves on, right? You’re orphaned.

Uh, more junior in venture invest, you mean?

Yeah.

Yeah, sorry.

And the more junior principal at that big fund moves on. They start their own fund. They move to another fund. It happens all the time. So the person who invested doesn’t have a mandate. They can’t sit around with a partnership and say, “Look, let’s just put another 5 to 10. Let’s turn over another card,” because your champion’s gone.

Yeah.

By the way, I’m being contrarian here. This does not always happen this way. I’m just giving you the other side to this. And then you haven’t made the kind of 1, 5, 10, 15 ARR, whatever you want to call it. You just haven’t made that. So it’s like you’re overlooked because it’s like, let’s focus on our real winners, and that thing’s worth $2 or $3 billion.

So, for 95%, the mandate for further funding is dead, is gone. Now, this is the beautiful thing about most entrepreneurs: they just don’t think about themselves in that category. “I’m the 5%. I’m the 2%.” And that’s why we love entrepreneurs.

Mm-hmm.

But the stats are so far against you. It goes back to—I hate to think of ourselves as their insurance policy, but I think a few entrepreneurs have thought about that. And I think there’s a little bit out in the zeitgeist going, “FC’s a great insurance policy. You want them in the round.” And it costs very little to have Harry or David in for $500K or $1 million.

8. Fund Size Stays Disciplined

Are you really not tempted to raise more? Every single constrained fund, including Benchmark, historically the central figure in discipline in venture, has raised a $1.5 billion growth fund. I was with another great growth fund that is very disciplined as well, but we’re raising billions too. Everyone who was disciplined is like, “No, we realize the game on the field is you need money.” Are you really not raising more?

It would be disingenuous to say to you that we don’t have the discussion, that it isn’t a tension, that we go back to it. It’s hard to be contrarian. When there’s so much money going around, it’s hard to say no.

Yeah.

And then here’s how we come out: the GP has been the biggest LP, and we’re greedy for returns, not management fees.

What percent of the fund are you now?

In the last few funds, we’re certainly the largest LP. There’s no LP that is bigger than the GP. We’re seriously aligned with our LPs, but what are we seeking? And this is the answer to your question. It may be wrong. Literally, if you do the analysis, you may go like, “That was crazy. You left so much on the table.” We’ve been very disciplined about strategy and very disciplined about DPI.

But if I just look at you—I’m sorry, I’m playing devil’s advocate again. I mean, you had Coupang, you had Uber, you had The Trade Desk, you’ve got Shield AI, you’ve got Suno. Tack on another $300 to $500 million vehicle and keep going. I’m sure you knew Mikey was amazing. I’m sure you knew TJ was great. I’m sure you knew that these were great. Surely that is a conversation that has rationality.

Because it’s a rational conversation, it comes up. And then you come back to saying, “Okay, who wants to do this?” You’re at an off-site for the partnership, and someone says, “Who wants to do this?” And I go, “Oh my, I love the early stage,” right? I may do it, right?

And by the way, I am an opportunist as well. I think of myself as some kind of value investor. So the interesting times for that for me have been when nobody’s funding them and I think that person’s great. Or it’s a consumer play, and I know consumer multiples are lower, but this is an internet acquisition device, and these founders are better at acquisition. So it’s not in the hype, hype, hype, go, go, go. I’m kind of immune to that.

Ugh, I’m in pain. I love you so much because you’re so much shorter than me, but I’m just like, the market can stay irrational longer than you can stay solvent. And when I look at Wix today trading at 2.1 billion on $2.1 billion of revenue, it’s a great example where there’s obviously rationality at play, but it doesn’t matter. The market’s the market.

And if consumer, say, is getting the pricing that it’s getting, I can’t change that no matter how good the acquisition machine is. And so don’t fight the tide that’s against you—that’s my thesis or ethos. Fucking swim in the swim lane that’s swimming in your favor. Am I wrong, and am I just missing a contrarian beat?

No. There are so many ways to do this. People have done so well. There are big funds, right, that have returned very well, right? You’ve got to be in the right vintage, but if you look at Thrive or a16z, they’ve had some big funds that have returned very, very well.

Yeah.

A little less since 2020. If you look at the DPI analysis, the jury’s out from 2020 onward. Now, of course, if you’re like Josh and you’re in SpaceX and OpenAI, that’s going to be the most ridiculous fund, but wow, you are in the most rarefied air.

And then there’s just something that’s competitive and unique, and it is potentially economically irrational. But if I was in that company, I was first, I wrote the biggest check. Somehow, for me, being competitive with me—that is the biggest thrill. I was with that founder from the beginning, and we literally backed up the truck and gave them everything they wanted.

And by the way, does that mean that we’re not writing $3–$4 million checks now? We are, right? Because if you want to get a percentage ownership in something that you think is extraordinary, you’re writing much bigger checks than we wrote before. So the fund is growing faster than it used to.

What is your average ownership now? Has it gone down over time? Because I look at ours, and our biggest mistake—and I can look at Deel, ElevenLabs, Granola, Starcloud, and Fractile—we could have done them all, but would have had 1% to 2%. We turned all of them down purely for ownership. That is hundreds and hundreds of millions of lost returns for ownership.

I've never thought about that.

Bugger.

I mean, all things being equal, I'm a capitalist, right? All things being equal, I'd love to own more upfront than less.

But it wouldn't be the reason you turn it down.

I've never turned it down. Never. And Mikey—I wanted to give him every last cent. He reached a point where he said, “Look, that is the dilution I'm willing to take. I'm not willing to take another iota of dilution.” We gave him what we gave him, which was literally every single cent in his first round. We showed it to other people, by the way. I showed it to you.

Thanks, David.

No problem.

We were going to bring that up.

And then when Matrix led—which was not a popular round; lots of people said no in that round—we asked for every last cent. But would I have said no to Mikey because of percentage ownership? When you meet the right people and you're all in, you get what you get.

And so you will do the 1% to 2% and take it, even though you can't size up in subsequent rounds?

9. Pro Rata Is The Original Sin

Well, again, I think pro rata is almost like the original sin. But if others have it, I don't think that we should be excluded if others have that pro rata. We're seeing rounds now where there isn't pro rata for anyone but the lead, or the most major shareholder. So it's not pro rata for all major shareholders; it's for the lead shareholder. I'm not sure I agree with that either in this environment.

I kind of think that there should be a universal approach to treating your investors equally. But I think pro rata is generally not great for entrepreneurs. It's a call option against you. We feel like we've had to work every time to put in a bit more money. We've never, ever led another round. So we have this view that it would be negative-correlation bias; it would be unfair to everybody if we weren't somewhat uniform.

Do you think it's harder than ever to accurately concentrate dollars effectively, given the rise of such preemptive rounds? We've had them where we haven't even wired the money and there's a new term sheet.

At different valuations.

Yeah. And that happens quite often now. Is it harder than ever to concentrate effectively when it's just so fast?

Some kind of framework is really, really necessary, and I credit my partners over the years with that—saying, “We may be writing bigger checks, but above that post-money valuation, it's really not our opportunity anymore.”

You can look in the rearview mirror and say, “Man, I should have done Uber. I should have done Shield AI.” Huge kudos and power to the people who did. But a framework lets you act very quickly.

I would say credit to Eric Paley in this case: he always created some kind of discipline. So the post-money went up and up and up as rounds, the momentum, the size of money, and the environment changed. But we would never lead another round. We've never done that in our entire history. So we haven't been preemptive, and we haven't been like, “We'll lead your Series A, and we like you more than others.” But our ability to participate has always been there.

Peter Thiel said before that if he'd just done every round that anyone else had done at an up round and it was a good brand, he would have done much better. Have you found that to be true?

Given the era, this has been the golden, golden era. From a data-driven approach, it's probably true. If we'd followed on in Uber, Coupang, Shield AI, you name it—

Suno.

Suno. If we'd just followed on, probably the data would show that we've done pretty well. Our view would be that we would have had to have followed on in everything, and I think the absolute return would be better. I don't think that the multiple would necessarily be better on the fund.

I'm not being rude. A framework isn't the enemy of this venture cycle. But I think it's so easy to be rigid in your mentality around, “Oh, we won't do anything over $1 billion.” But—and you're going to absolutely wince at me here—are you ready for real shit?

Go.

I think a $1 billion valuation is the new Series A. And you're like, “Whoa, Harry. Whoa, whoa, kiddo. Calm down. Listen to the facts.” We used to do a $50 million post-money and hope it would become $1 billion—20X without dilution, like Blunt. Now you enter at $1 billion and you hope it becomes $20 billion. We have Mercor at $20 billion. We have Cognition at $26 billion. Cursor got sold for $60 billion. Sold. This is liquid. Well, maybe $1 billion is the new Series A, no?

I think you may be looking at the top 2 or 300 companies.

Is that not our business?

I don't think so. I think that's the momentum business, and I think knowing how and when to get out quickly with some of those really, really matters. That's not really my business. My business is value: getting involved early and trying to find value opportunities.

There are times, again, when it's an intoxicating founder and being on that journey together, but I'm not sure that those are your fund returners. The difficulty with some of those momentum assets is what we were talking about earlier: you've got to be able to run for the exits when you can.

It's exactly what you were saying: you didn't think that founder was all that great, or you thought that the valuation was so far ahead of the reality of the business. But you're asking a question—you're asking a momentum question.

Yeah.

And is it all momentum? I've got to be careful not to be too anachronistic in this, because we have invested in momentum. It would be so disingenuous for me to say that we haven't.

When you say that, what do you mean, “we have invested in momentum”?

Our knee-jerk tends to be, when this has gotten across a certain point, “We're out of here.” And credit to Eric at a point for going, “We've captured 80% of the value. We could capture another 20% if we did Uber at Series A or if we did Suno at Series A.”

By the way, it's not just on paper. I think there would be buyers for that position. So, in hindsight, I look at that and I go, “Were you anachronistic?” By the way, we didn't even seek to participate in that round. We said, “We built our ownership position, and we're done.” This is not the kind of investor we are. We're looking for the next seed-stage round.

I think, Harry, what we've done is we've drunk the Kool-Aid to such a large extent now. You and I are so different: you're going, “This is hot. Let me go, go, go.” I'm going, “I've got a smaller fund. Where else can I really maximize my ownership versus getting, I don't know, a 5X or a 10X?” But, of course, the environment makes you look quite silly in retrospect. The question is, how long does this environment go on for?

And it's also about—you have unbelievable returns, and you've made a phenomenal amount of money for your investors. But the quantum of cash that you move matters. Josh, Elad, and the multistage funds moving hundreds of millions and billions make a larger quantum of cash. So I get you in terms of your multiple going down when you lead the Series A.

Look, there are so many different ways to play this, and I think when you talk about Josh and, you know, a handful of others, they've killed it. They've absolutely killed it.

A lot of LPs, very wrongly, I think, don't like the large platforms and always just come back to this very basic, rudimentary thought that as you scale fund size, returns always get worse. Always. Whenever someone says “always,” be careful. But I think with the outcome expansion that we've seen—Cursor at $60 billion, trillion-dollar companies in a matter of years with OpenAI and Anthropic—you will see venture returns with mega-platform sizes. Do you agree?

Largely, no. So, largely, I would say: who are their LPs? Who are they working for? And in some of these cases, not even endowments anymore.

Exactly.

It's sovereign wealth funds. And sovereign wealth funds and public investment corporations are looking for IRR.

Yeah.

They're not measuring this in how many times you X the fund. That doesn't mean that a16z and Thrive haven't X'd a few of their funds really, really nicely. Again, subsequent to 2020, the TVPI is there, and in some cases, they're on steroids. The DPI is less there if you look at the actual stats.

But they're working for these sovereign wealth funds, and they're giving great IRR. Some of the endowments—some of the biggest endowments—are rounding errors now. The question is, who are you working for?

I, again, obsess over this alignment with the entrepreneur. We're working for ourselves as well, right? And we're working for DPI, and the bigger we make the fund, the tougher it is on the DPI. What am I doing this for? Fund of the fund of the fund.

If we look back on Fund II, it's all about applied AI. If you look at the winners in Fund II, it's Shield AI—which, by the way, in 2016 was called Shield AI. It's Verkada. It's Whoop. Now, all of these things are commoditized. All of the hardware is commoditized. Video cameras are commoditized.

Drones—I mean, you can buy a drone for—forget DJI—you can buy a drone for $20 now, right? $50. It's about putting AI around these completely commoditized platforms. It was 10 years ago, but it wasn't the theme.

So the one thing that you're talking about is momentum around a theme, and I'm going, “In 10 years' time or in 5 years' time, there will be a new theme.” The job will have been to get into that theme ahead. I don't even know what it is. I hope I've got some on goal.

And those weren't the expensive ones. They never are. So using AI is really important. I guarantee you all of these things are called applied AI businesses today, or physical AI. The job is to be in there 5 years or 10 years ahead, and it's not where the momentum is. It never is.

When you look at Fund II, and you said a couple of names there with Verkada, Whoop, and Shield AI, how concentrated are the returns in your funds? I spent time with Ho Nam from Altos, and he's spoken about the return concentration with Roblox. Mind-blowing to me. How concentrated are yours, and what lessons do you have from that?

The amazing thing is they've been way less concentrated than you would expect.

Huh.

Look at Fund II. Forget Fund One now. Fund II: Verkada, Shield AI, Whoop, and PillPack. In all of them, for the most part, we were one of—or the single largest—investors in the first institutional round. It's not concentrated.

If you look at Fund One, we always talk about The Trade Desk, Uber, and Coupang. Fund One still has Airtable in at the very beginning. There are challenges in the SaaS environment, but Simply is the biggest piano-teaching and musical-instrument-teaching company in the world. SeatGeek—I haven't sold a single share in SeatGeek. That's still in Fund One.

Why haven't you sold a share in SeatGeek?

I think it's spiritual at this point.

It's a religion. Beneath the shirt, you've got Jack's face, right?

No.

That would be epic, brother.

I did that at my LP meeting. Before Jack and Mikey both presented, I literally said to them, “Get me T-shirts.” And I had ripped open my shirt.

That would be very funny.

But I'm actually worried about this, which is—and I'm not positioning this at Airtable at all—I think Howie is wonderful and brilliant, and has a brilliant product team. But you're seeing the cannibalization of leaders in a space like Airtable, respectfully, and Snyk, the cybersecurity company, which in a similar vein is going through challenging times in terms of growth rates and everything involved. There hasn't been a liquidity event, but the cannibalization has already started. It's like the innovation cycle's taken steroids and gone too quickly to allow liquidity events to even happen. Does that worry you, too?

By now, Harry, it's very hard to play around with Claude or something like it and not have the revelation that we've all had. But then you look at some of these SaaS companies, and you look at the SaaSpocalypse.

When we were on the Olo board, when it was listed, we'd look at companies like Veeva, which is, I think, at a $30 billion market cap now. It's come down at least 50% or more. And we'd go, “We want to be this company.” It's hard not to look at some of that market-cap erosion and go, “Is the baby being thrown out with the bathwater?”

And it's about the last 5%, I think. I would say, if you're looking at Airtable—and Veeva or Olo look very different—the more embedded you are, the more difficult you are to displace, because billions of orders are being run on your system in real time, or mission-critical biotech research is being run on your system. The more embedded you are, I think the more overdone that SaaSpocalypse may be.

Mm-hmm.

The less embedded, clearly, the easier you are to turf out and play around with Claude, you name it. But I think we're underestimating that last 5%. The contrarian in me—this is not what I do—would say, “Buy a basket of the top SaaS stocks that have all lost huge market cap. You're going to do okay.”

You are. Rory O'Driscoll, who we do the show with every week, has done that, and I put my money into Palantir and said, “I'm a momentum surfer.” I did better.

And that's the hard part: the opportunity cost of cash is so real that you can be in one and try to be smart, but you're probably right long term, or you can just be a momentum trader and you'll be right, actually, in the short term. If you can time it well, it makes a difference.

Yeah. The difference, in a way, between our styles is it comes back to concentration. Every single company I invest in, I invest in with the hope that it could be another Suno or Uber. I literally do. I don't invest in companies and go, “I'm investing in you, Harry, because I think you can be a 10X outcome.” I don't do that.

You don't?

No.

Wow.

For every company we're investing in, we think, “Wow, this could be ginormous.”

Jason Lampkin just told me a very similar thing. He's like, “I'm not smart enough to predict the future. What I look for is, can I get a 3X on my next funding round? And if I can get a 3X on my next funding round and I really believe in a great entrepreneur CEO and a great CTO, I'm in.”

So we use the same logic, but it's always been 10X. I will not invest in this if I'm not sure there's a 10X. At our team meeting, I love it because—dot, dot, dot. If you can't complete that sentence, you can't invest. That's how we start the team meeting. That's how we start talking about a portfolio company.

What's your greatest-ever answer to that?

In more recent times, I've gone, “I love it because I'm obsessed with Harry. Every question I ask, I get a better answer than I expected. Every time I press, there's no evasion of the facts. He never says to me, ‘Oh, we're the only one in this business.’ He always says, ‘It's so much harder than you think it's going to be. It's so much tougher,’ and, ‘This person's leaving me.’ And I love it because they're obsessive, they're all over it, they're so deep in this, and I just can't get this out of my—”

I will not say I love it because of valuation. By the way, we've always come to valuation last. We've always gone opportunity, market, founders—founders first and foremost, it's in our name—and we come to valuation last. I cannot say that every single time we've invested we've gone, “This is a perfect valuation.” In fact—

Rarely is.

No, it rarely is.

The best deals—

It really, really rarely is.

Both sides feel uncomfortable, I find.

Of course.

Yeah.

Of course.

All right.

Exactly right. By the way, it can go, “I love it because of insight into the vertical.” I love it because of an edge that nobody else can match in a commoditized business. I'm writing this piece on nepo babies, and I'm going, “I love to fund nepo babies.”

What?

I'm writing this piece right now.

Why do you love to fund nepo babies?

TJ Parker was working in his dad's pharmacy when he was 15. He has got more edge in that vertical than he knows. Mikey comes to voice AI, to music, to audio, right? They've come out of Kensho. That's all they did at Kensho. So you take Mikey, Georg, and Martin Camacho—that's all they did. Martin was the CTO of Kensho.

Now, they're not the nepo baby, but Evan at Rebar is. Evan at Rebar is HVAC preparation and HVAC quoting. There are over 100,000 mechanical engineers in the U.S. who are making at least $100,000 each when they graduate, and all they're doing is sitting with this blueprint process so that they can quote on new commercial.

Evan worked for his uncle's company, which was rolled up in a private-equity firm with 10 of these things, and they said, “Go out and find the AI for this.” And Evan goes out and goes, “There's no AI for this.” And he goes, “I'm starting Rebar.” They're folks who have been in these verticals since they were kids. He watched his uncle in this vertical. It's like there was nothing else he was going to do. They have more edge than they know what to do with.

I get you. Sorry, just to be clear, for me, a nepo baby is a trust-fund baby who has billions of dollars. I was like, “Dude, I do not want to be funding the kid who's at Scorpios in Mykonos spraying Dad's money.”

I—

We're using nepo babies with different definitions.

We have very different definitions.

I'm talking about folks who've—

Historical background.

—been in a vertical—

Yeah, yeah, yeah, I—

—and have lots of edge in that vertical.

That I totally get. You said you haven't sold a share of SeatGeek. The timing of when you get out matters a lot. Do you have any lessons on when to get out, given I think this generation of seed managers will be defined by their ability to access and navigate secondary markets effectively?

It's interesting you're asking this at a time when I have never seen secondary markets as liquid. It's probably not that surprising given—

Fewer IPOs, fewer M&A up till this moment. There's an IPO market that will probably be open for the remainder of this year, and then these IPO markets always close. So, in the top 100 names, wow, the secondary liquidity is incredible.

And you can price your position reasonably efficiently. You can look at a round and go, “Okay, the secondary market is offering me a 25% discount. It’s probably worth 7.5, not 10.” Then you can look at a number in the top 50, at least, where you’re being offered at least the price per share of the last round.

100%.

Because loads of folks, loads of big folks like Blackstone, didn’t get their pro rata, and then they’re sucking it up.

Most I’m seeing do not have a discount, for sure.

Yeah. We’ve seen a premium where insiders know there’s another round. Talking to your point about momentum, you were talking about momentum in the early stage. We’ve seen situations in our multibillion-dollar names where the round goes down in December and the board’s already talking about the March round. We sometimes see it when we’re not on the board, but we just see it in the momentum in the secondary market.

Now, Harry, you’re in very rare air there. Let me just say, I don’t want to in any way make it sound like we’re in that with all of our companies. We’re in that with a handful of companies at any one point. But in those situations, I think the difference in fund management is when you take secondary and the ability to give DPI.

Even in your top names, sometimes taking 20% off the table can return 25% of the fund, particularly if it’s a new-ish fund. So if it’s a 2024 fund and you can give back 25%, why wouldn’t you do that? You’re still long. You still own 80% of that company.

I just think we don’t think about the velocity of cash enough. What I mean by that is, yes, there might be another double, but if I have to wait 5 years, then the IPO, and then an 18-month lockup—Jesus, give me 50% of that now. I’d way rather have the certainty and the DPI now than maybe a double from here with 6.5 years.

This is not a precise science. I’ve looked back in every direction. By the way, the best is you sell 20% and you were wrong. Awesome.

Did you do a good job of the sell-down on Uber?

In retrospect, we probably sold a little too early. This was early on. This was a business getting close to a $10 billion valuation, and there was an opportunity to take some off the table. We were very new.

Also, at the time—I’m so sorry if this sounds awful, and again, you can chastise me—$10 billion at that time was so much more than it is today.

Yeah.

Did you sell all at $10 billion?

No, definitely not. We were net long at the IPO.

One thing that’s very sad or challenging is when an exit event happens and then you look at the number that comes back to you and you’re like, “What? Where did it go?” I think you’re seeing a normalization of incredible levels of dilution today, more than ever before. Do you see that and worry about that?

Yeah. Look, dilution is interesting. I look at Whoop versus Suno. We’re so proud to be in both, but Suno has been a very quick journey. If you look at how much lower the dilution is, part of it is just how quick the momentum has been versus Whoop, which is hardware. It took a long time and raised a lot of money along the way.

I’m unbelievably proud of this company, but with some of these companies, it’s incredible how little dilution there is because the pre just goes through the roof.

We’re also seeing a lot of very low dilution but large rounds. You’re raising $500 million at a $40 billion price, and, actually, seemingly no dilution—$50 million rounds at a $1 billion price. How do you think about and reflect on those? Is it just a brilliant product for founders that they should absolutely take advantage of? Is it a normalization of continuous funding because they do more, more frequently?

This goes in every single direction. You’ve got to be producing, and you’ve got to get into the rarest of air there. There’s probably a secondary opportunity in that kind of situation for us. We look at that, and again, we’re in so early that at those kinds of numbers and with that kind of momentum, we’re trying to sell a little bit of our position.

Do you find LPs have changed? What I mean by that is, I speak to a lot of LPs now, but you know what? Honestly, we can say what we want. They’ve gone back to wanting TVPI. They’ve gone back to wanting big numbers. Yes, they want DPI. Of course, they always want DPI. But they are still very impressed by TVPI, and they’re very impressed by, “Oh, wow, you’re in this glossy name: Lovable, Legora, McCaw.” There’s still that. Do you find they’ve changed, or are they still the same animal?

There’s lots of change because of who was doing this 15 years ago and who’s doing it now. You have to have some allocation, and the big funds provide these containers for the large endowments and the large public investment corporations. If I think of the same LPs that have been with us for a long time, a lot of them have minimum-size checks now, so we’re too small for quite a few of them. It’s like, “If I can’t put $50 million to it…”

I think it just reflects the inflation of the entire environment. There are a bunch that really do need the TVPI, particularly the fund of funds.

Mm.

Because of who they’re selling to. By the way, we’ve seen funds of funds do secondaries of their entire fund. We go, “Oh my God, you’re in fund 2 or you’re in fund 4. Why? You should never sell,” right? This is what’s happening. And it’s like, it’s not about you. You’re a rounding error in this fund, and it’s got 3 or 4 good names.

I think what they’re trying to do is give liquidity to their LPs for the next fund. When I talk to you about secondaries in a particular name, we’ve seen an entire fund—a billion-dollar fund—easily just sell the whole fund or sell a vertical slice of the fund.

What’s going on here is the finance around VC has become so much more sophisticated. I don’t know if this is good for the entrepreneurs. It could be, because it just means there’s way more liquidity in every direction. If you’re a winner, it’s great because you can manage the secondary to some degree. If you’re not on the winning side, in terms of the entire ecosystem, it can be very tricky.

By the way, Harry, I talked to you about this on this podcast. Other than thinking about some secondary in our very mature portfolio, I spend very little time on this. The beauty of this is I am not a financial animal. Ultimately, I’m much more of an entrepreneurial, curious animal. I’m looking again for these wizards, these wayfinders.

This is the problem for me, in a way: I’m looking to repeat a success. I’m looking for the next high. I’m looking for a Noah Glass. I’m looking for someone who is that focused, has a vision, and will not take no for an answer. That’s how I’m spending 90% of my time. I’m not spending much time even on LP management.

Do you think we have less loyalty than ever? You said you focus there on the founder side. You see founders with angel investment portfolios that are as big as our fund portfolios. You have them doing side funds. You have them doing 2 companies at once. You have them leaving very quickly, often in 6, 12, or 18 months. Is there less focus and loyalty than ever?

We’ve definitely seen evidence of that. We’ve also seen people who stick it out way beyond what is rational just because they’re obsessed. So I think on the margin, you see some of these actors. We’ve seen founders, so-called founders, who were kind of the founder, but they got a CEO involved, then they became executive chairman, and they used their brand power.

I think shame on us, because we did get involved in some of these situations where we were dazzled. Then it was like, second time around, is that person going to stick around? Some of it is that they just didn’t get big enough fast enough.

Mm.

Howard Lindzon

So there’s some abandonment. I still see that the vast minority of the time. I think it’s easy to extrapolate and go, “That’s a trend.” I could be very Pollyannish about this, but for the most part, I see founders wanting to make it work.

Second-time founders are a little bit embedded in that question. If you’ve done really well, what does it take to move the needle? Overall, we’ve done a little better with second-time founders who didn’t do that great upfront. They did okay. It’s life-changing. The first million dollars is so life-changing, but they’re really hungry. They’ve learned some lessons, and they’ve got 1 or 2 people who will join them on the next journey.

They’re hungry, and they’re in a hurry as well. We’ve done better there than generally with folks who had great outcomes and kind of said, “Let’s go again.” Those are the folks who got bored and went, “Not big enough, not fast enough.”

What does no one know? You’ve been very successful. What does no one know about making money that you wish you had been told earlier?

Howard Lindzon

You start to realize that the stuff that really matters is kindness, how we interact with each other, how I left you, how you made me feel, and all the rest is fluff. At some level or another, our phones have become these remote controls for our lives.

Actually, the entry price to get what you want when you want it—if you want a vehicle there, your food there, or to book a flight or a train ride—is insane. Earlier last week, my plane was delayed. Literally, on the Amtrak app two seconds later, I asked the Uber to go to Penn Station instead.

The degree to which we can get what we want when we want it, at any level—you don’t have to be that wealthy to get it—is insane. So what’s happened is our level of expectations has just gone through the roof. I don’t think that’s just about you and me. I think that’s the perennial equation: satisfaction equals perception minus expectation.

It’s just much easier to not be satisfied anymore because our expectations are so high. Our perception is one thing when you go into a five-star hotel and you have this huge expectation. You walk into a three-star hotel, you have a much lower expectation. Well, extrapolate that equation for life now. So it’s easy to get pissed off.

The antidote to that is stopping for a second and saying, “How will Harry feel when I leave him today? Did he feel like I gave him a real hug and I was kind?” I think what’s going on in my 50s now is: How do I leave people? How do I leave the world? How do I leave the entrepreneur? Was it like we squabbled over the last percentage point, or was it like, “This journey’s been awesome”?

I always think there are energy drains and energy gains, and how you leave someone is how you’re remembered. Going back, before we do a quick fire, I do have to ask: How does this landscape change with OpenAI and Anthropic? They are so seismic in terms of sheer size. Both will be trillion-dollar-plus companies, potentially close to $2 trillion. How does that change the landscape, do you think?

Howard Lindzon

For the better. I remember the Microsoft Google case going on forever and Gates saying, “We are disruptable.” At the time, people were saying, “Who could disrupt Microsoft?” It turns out Google was Google. Then you go, “Who can possibly disrupt Google?” And then you look at OpenAI and Anthropic and go, “Wow.”

If nothing else—and there’s so much else—look at the top of the funnel in terms of where you start your search. When did you last start a search on Google? It’s just mind-blowing, that displacement. The good news in this environment and ecosystem is that they, too, will be displaced. So the platform has changed tremendously.

Do you think Google has been displaced in what way?

Howard Lindzon

No, I don’t. I think Google’s a net winner.

Do you think Microsoft has been displaced?

Howard Lindzon

I think Microsoft has done a crappy job of AI generally. Google is actually, if anything, in pole position because they come from that environment, and the ability to search with context, the ability to apply AI with context, is just incredible. But they’re having to fight like crazy for it.

Microsoft, it’s not clear to me that they can get back because their AI feels second-rate compared to the top three or four. So there’s a platform change. There’s always been a platform. You could argue that radio, television, and the internet were platforms. Can you do well on that platform? Oh, hell yes.

Do you think that will lead to a ton more venture money coming in, with a huge amount of money going back to LPs from the returns that are generated? Will they plow those back into venture?

Howard Lindzon

The returns at the top are going to be incredible. They have to be. I think that capital is going to spill over into venture and all sorts of investing. You alluded to it earlier, Angel. I think luxury is one area. If you own a luxury property, I think—

San Francisco property prices.

Howard Lindzon

Oh, my God. San Francisco is Rome. I was there six weeks ago. San Francisco and the Bay Area—and it’s more San Francisco than the Bay Area—is back on steroids. It’s like going to Rome.

When people write off the United States, which is, to me, still the greatest country in the world for venture capital, I say, “When were you last in San Francisco or the Bay Area?” Because it is insane at the moment. What’s going to happen is there’s always boom and bust, so a lot is going to come out of the system at some point.

Are we headed for another dot-com crash? Definitely. If is not a question; when, nobody knows. But is there a lot of capital and a lot of gain coming out of the system that will be reinvested in venture? It may not be in classic, structural venture. It may just be angels putting money all over the place, and some of those angels are going to know people who worked with them or for them.

I think you can bypass traditional venture to a great extent, and that’s the challenge for us. That’s the challenge of how you stay relevant in this environment when there are so many alternatives.

Final one before we do a quick fire. Do you buy the commonly stated concern about smaller teams, job displacement, and a concerning future for human participation in labor forces?

Howard Lindzon

The underpins to Endeavor are getting better and better and better. I remember when we went from servers to the cloud, and that was like, wow. I get all of this for free. I don’t have to do any of that—security and servers. Forget that. I just do cloud.

If you look at where AI and this foundational platform layer kick you off, it is incredible what you can do with very few people, and we are looking at certainly sub-10-person companies achieving a lot. Do I think that we’re going to have mass unemployment because of AI, and that you’re seeing a lot of leadership now agree with that view? No. I think we’re going to see tremendous productivity gains.

Like every wave, there are the haves and the have-nots. If you’re not training and playing, it’s a little bit why youth has an advantage, because if you’re out of college and tinkering and playing, you are familiar with the tools. You can use the tools.

It used to be that the haves and the have-nots were about having data. I talked about this with Noah Glass and Olo all day long, saying, “The value of having data and using that data…” And, by the way, it’s yours to lose if you don’t enrich that data. Now, the value of having these tools is becoming more and more binary.

But I do believe you’ll see swaths of people retrained on this, and I think you’re seeing it globally. I think you’re seeing this as an opportunity in low-cost environments, in places that are not Europe, not the U.S., not the North, where you can skill people up and provide these skills to the rest of the world at a tremendous cost advantage.

My worry is that it’s much easier to train than it is to retrain. The 22-year-olds coming out of university who are tinkering in dorm rooms with Claude—they’re not super-AI-pilled, but they’re mentally plastic to it. They’re going to be pretty good, versus Simon or Claire, who are 45. They’ve always done their job in accounting, and they’re just not so mentally plastic.

The only advantage that Simon and Claire have is that they are very vertically knowledgeable and relevant. Sometimes, in terms of sales, if you’re selling to yourself, they’ll actually be very good salespeople.

This is a theme that I’m interested in: services businesses, where you won’t buy that from... You want to see your auditor at some point. You’re prepared to say, “I know AI will do an amazing job, but at some point, I want you to come see me and say, ‘You know, I haven’t left this whole thing to AI. I actually know what I’m doing.’”

I think there will be people who are vertically relevant and will be able to sell, and there are many industries where the relationship still matters. At a certain point, if you’ve got litigation involving $100 million, you can get AI to write that legal contract for you when there’s $1,000 on the line. But if you’ve got $100 million in litigation, you want to look at me and say, “Dave, your 10 years of experience—I need it right now.”

So I think there are times when knowing a vertical and being relevant in that place matters. In the service industries, I think it’s good for the U.K., by the way. I think there will be a ton of people who are still needed for the human interface. That’s not going away. I think a lot of the work that was grunt work, with a human behind it, is going away.

On the services side, I think it’s just a TAM expansion play. So much of what you couldn’t afford a lawyer for, you’ll use, and you’ll get great benefits from.

Yeah.

That is just a TAM expansion play.

I think insurance, lots of admin, and life insurance. I want to—there’s been a lot of direct life insurance sales anyway. But I think that in bigger-ticket items, having a human who gets it as the interface still has a place.

There was something interesting about it. I had this incredible founder, June, who’s the founder of a company called Simuli, which does simulation markets. He was like, “We will have companies spend $100 million to $200 million on one model result because that model result is so important—the output of one query.” I was like, “Wow, that’s a really interesting world where you will spend—”

$100 million on Anthropic telling you the answer to one question?

What’s the size of that organization that would spend that kind of money?

P&G, Coca-Cola, NVIDIA, Visa—you name it. “Is it worth us sponsoring the World Cup for a 10-year exclusivity period?” Visa.

Right.

I think that governments and defense organizations, some kind of speculation with data about the future—I think that’s a very interesting play.

Do you worry that Trump’s been good for business but bad for everything else? Is that a hard balance to hold in your head? I ask as an outsider, genuinely curious.

I think you have to hold many truths at one point in time. The question is, did Trump create this environment, or is he presiding over this environment and getting credit for it? I think with all presidents, they arrive and get credit for the environment as it is, and yet it was created many years ago.

Letting AI thrive in the US has generally been a good thing for the tech industry in the US. The level, or lack, of safeguards on that could well be problematic, but net-net, if it’s good for business, it’s good for the US—I think Roosevelt said that. That’s what these administrations have said.

By the way, I think that a lot of the tech backlash around Biden was for this reason, whether it was true or not. A lot of insiders say to me it was BS, that, for the most part, Biden was super pro-business. If you look at the subsidies for energy, if you look at a Tesla today, this is the thing that I don’t really get about Elon: the non-dilutive government funding that Musk got for Tesla from the Biden administration was huge.

So, without being political, net-net, government in the US has been pro-business for a long time, and I think that the country is really reaping the rewards of that. There are 2 AI superpowers in the world. By the way, what’s so fascinating is that in the 1820s, China was the economic superpower of the world. I don’t know if you knew that.

No, I didn’t.

Yeah. Great Britain displaced China. A lot of it was the Industrial Revolution, and then the US displaced Great Britain. In The Economist, there was a chart on this.

But in the 1820s, 25% of the world’s economic output was from China. It was the biggest economic machine in the world. Really, what you’re seeing is 2 superpowers emerge for sure, and I think a lot of this is going to be about AI.

AI flows into not just industry but also what’s going on in defense. Having been very, very early—the first check in Shield AI—and watching how that’s played out, the US needs it. Our enemies have access to all of that on steroids.

I’m terrified about China right now, to be honest. When you look at the power and strength of their open models, pfft.

But that goes back to thinking about Microsoft and Google being disrupted. You know that Anthropic and OpenAI are going to be disrupted. It’s unequivocal. Our whole careers are about disruption. Those platforms never, ever stay forever.

Where is it going to come from? There’s an excellent chance it comes from China. It’s coming.

100%. But, God, we haven’t had enough time for them to establish their incumbency before they’re already being taken down by Chinese open-source models.

Yeah, yeah.

It goes to the point on the speed of innovation cycles.

Yeah. By the way, we haven’t even touched on underlying computing. If you look at photonic computing and what’s coming down the line now, you look at Intel at a point when that could never be disrupted, and then NVIDIA—it’s just mind-blowing.

What’s coming against NVIDIA are the photonic computing plays right now, where it’s not electrical anymore; it’s photons. If you look at data centers, everything that can be optical fiber now is—every single connectivity piece of hardware is fiber. The only thing that has not been nailed is the chip.

You’re going to see optical chips, which are very, very energy compliant. So when people talk about data centers and the energy sucks, that’s going to change. In my view, if you say that in 10 years’ time—and I am not a thematic investor, but I am such a deep believer in the status quo always being changed, and that nothing stays the same—I think photonic computing is coming down the line, and I think that’s going to be the NVIDIA disruptor.

Mm.

Or NVIDIA’s going to buy those companies.

Okay. And the capital intensity required to build a photonic company, or an energy company as we’re seeing, is just dramatically more capital-intensive than prior technology. Again, going back to my point, you need more money.

Well, this is where the US could be deficient. If you look at the amount of money that’s being spent in China on energy efficiency and energy research now, I don’t think we’re spending enough. By the way, that’s a negative of the Trump administration: we need much more money being spent on R&D.

I think there was a view that the universities are squandering it. To a large extent, I agree with that. But I think that we tapped a lot of DARPA R&D that finds its way into every nook and cranny of the economy, and we need more of that R&D.

We see some of it. I live in Cambridge, Massachusetts. We have some of the best R&D organizations on the planet. If you look at MIT, Harvard, Northeastern, BU, and BC, what’s going on there, cutting that spend, which goes back into society, I think is problematic.

Totally get that. Another one, though, that is more challenging, I think, to change is just policy and regulation. The Chinese approach to policy and regulation is—it’s almost none.

It’s none. Yeah.

And it means that you can bluntly build and deploy so much faster. I mean, Europe’s the worst. The US is—

Yeah.

What would cause you—final one—what would cause you to increase fund size?

If I am honest about what we did early on, as an angel I had said the risk premium for the seed stage was way overstated. So the premium for experience—I couldn’t get that. A lot of the folks that I got involved with very early were graduating. They were Noah Glass, Jack, you name it, Eric and Micah.

There was a dislocation between the perception of value later versus earlier, and that has been largely narrowed and crowded out. If there was some kind of arbitrage, Harry, we didn’t come at this going, “I’m obsessed with economic arbitrage.” We came at this going, “I’m obsessed with great founders, and I want to vicariously be on that journey.”

But if you had to look at this retrospectively and say, “What did we do in economic terms?” there was an arbitrage. There was a real arbitrage because the risk premium at the seed stage was way overstated. That has changed completely.

What would cause me to raise a bigger fund? If I looked at Series A or Series B or Series C and went, “There is such a value opportunity because everybody’s abandoning this.” I don’t think it’s true at the moment. I think capital and money find their way to everything.

But if you went, “So many Series A companies are orphaned, and there’s amazing value. There hasn’t been a 1-to-10-to-20 increase in ARR in 1 year, but wow, they’re on track, and that looks like—it smells like Olo. It looks like SeatGeek,” I think that would cause me to say, “We should be investing $10 million at that stage.”

So it’s not momentum. It’s a sense of, wow, I can’t believe that others—

I have been very tempted there. I’ve been very tempted to say, “This company is doing incredibly well on the revenue side, and it’s being undervalued.”

Final one. I promise. You say that about Olo. I love Noah. I think Noah’s one of the greatest, most awesome humans. Dude, his 17-year journey to a $1.6 billion, $1.7 billion exit—

$2 billion exit.

$2 billion exit.

Yeah.

I love Noah. I love Olo. It’s an amazing business. It’s an amazing journey. But when you think about utilization of cash most optimally, 17 or 18 years, a $2 billion exit—the IRR is not amazing. How do you reflect on that and justify that versus maybe hot rounds?

Yeah. I mean, the outcome was publicly known. Eventually, Thoma Bravo took the company private for about a $2 billion valuation. So, not a bad result for a few years of work.

If you take it on an IRR basis, you’re probably right. The journey and the fun of it were just enormous. Being involved with Noah, where it was Noah, a few other founders, and me from the beginning, and being on the board until that sale was just the ride of a lifetime.

Okay. Listen, quick-fire round time. What have you changed your mind on in the last 12 months?

What’s been surprising, and what I’ve changed my mind on a little, is where AI should have had a crazy impact, and I’ve seen lesser impact so far. With all the hype and all the momentum, AI changes so much in terms of software, enterprise, and SMB, other than the models themselves and some good stuff around the human interface.

A lot of stuff around voice has gotten a lot better. I don’t know about you, but I would have expected much more around consumer AI. I’ve seen Suno, but in terms of changing how I do stuff, I type much less. I speak much more in terms of communication.

I would say there are so many consumer areas that I feel are not yet played out at all.

When you’re doing Suno at $5 billion, what are you underwriting it to?

I think that the folks investing at that level are going, “This is a Spotify disruptor.” That Spotify and Apple Music—it’s a big bet that because—

They’ve got to go from creation tool to consumption tool.

Oh, totally. Totally. That's why Jack from Snap was brought in. Interestingly, I was at a conference with Martin. Martin Camacho, who's the CTO of Suno, was asked a question: “If a large language model could do what you do better than you do it, would you slot that in?”

You're talking to the CTO, the guy who's built the whole model. The entire Suno model is from the ground up. Without missing a beat, Martin goes, “Wouldn't think twice about it.” It goes to your point that this is a consumer product. The experience, the interface—think Spotify, right? That's what we offer. How we get there is obfuscated from the user. The user couldn't care less. Whatever gets you there.

Did you ever predict the speed of that?

No, definitely not.

Because I remember when Slack, 1 to 10 in 18 months, was the gold standard. I mean, Suno's multiple hundreds of millions—I mean, half a billion or whatever it is now. It's nuts.

No. No. It goes back to Uber. You know, when Eric is asked, “How did you know? How did you know?” Eric goes, “I didn't.” He said, “The company I saw before, the company I saw afterwards—we underwrote those in the exact same way.” Anyone in my seat who says, “I knew,” is just full of shit.

I absolutely love that. What's been the most controversial deal that you've done internally?

What can become controversial is the what or the where. So certainly, Coupang—I said to Eric, “Korea,” and he said, “Do you even know if it's north or south?” But the magic there is that I'm based in Harvard Square. So people go, “How do you get to Korea?” How do you get to all sorts of places? And the answer is Harvard Square.

So Bom drops out of HBS after his first year and comes to see me. Another controversial company was probably Shield AI, in terms of what it does. I would say the whole partnership didn't necessarily love defense drones, and early on it was, “Is this only defense?” I love it because it's certainly taken us to some very controversial geographies and controversial whats.

Are prediction marketplaces just legitimizing gambling?

It has to be. I mean, if you look at Kalshi and Polymarket, what's the difference between those and DraftKings and Betway? They seem very similar to me. But, by the way, this will be controversial: TVPI versus DPI. The one looks like a prediction market, right, and the one's real. I could say the same—TVPI looks like a prediction market, right?

But candidly, when they're doing $2 billion in ARR, who gives a shit? The one thing that you worry about is a change in the Trump administration and what that does to regulation around them.

Yeah, yeah.

That's a different game.

Yeah, yeah.

What do you know now that you wish you'd known when you started Founder Collective?

For the most part, frameworks have saved us. It's also the place where, if I look at some of the deals that we didn't do, we used valuation as shorthand to say no. Terrible mistakes.

Clavio—I loved Andrew, loved Ed. It came to me first through Hugo van Vuuren, who also sent us Suno, and I didn't do it because of the framework. The framework allowed me to easily say no. So we'll miss a lot. We'll make plenty of mistakes. I think I've freed myself, like you, a little more in that area. Just go, “They're extraordinary.” But the frameworks have saved us as well.

Penultimate one: biggest advice on a happy marriage and relationship?

Kindness and being present. Being present with each other. At dinner time, no phones are allowed anywhere near the dining room table, and I don't take my phone to my bedroom. My phone is never alongside my bed.

Here's the rub: I don't need my phone to be distracted. Here distracts me perfectly. The question is how to be present and involved, look you in the eye, and make you feel, with my body language, that I'm hearing you and that I'm invested in you.

I think of the same things we think of in founders: a happy life, your kids, your wife, your siblings, your parents. And this is the lifelong goal. I have not got this nailed. But how do you show them you're present, you're there, and they matter to you? That's the quest.

Final one. What are you most excited for in the next 10 years? You know, I look at my own life. My mother and I walk marathons. She's got MS. I think there'll be amazing discoveries for chronic conditions that we always just assumed would be forever. That could change millions of lives. I think that's super exciting. What are you most excited for?

I mean, you're leading the witness in a few ways here. But I would say that each wave brings things that we couldn't imagine. I look back to driverless cars, and there was a promise that they were 5 years away. It turns out it was 20 years from now. I saw Waymo driving around London. I think it's coming here soon. We're not quite there.

And yet we're back in the “slow, slow, slow, and then it feels like overnight” phase. Of course, if you were involved—and this is, again, the intoxicating part of what we do—you know before the world knows or the world cares. But you know that it took a long time.

I think we're on the threshold of a lot of really interesting stuff. I think that you and I could be buying the very last driven cars. I think that in 5–10 years' time, our kids will not need to drive. And I think with AI, we're on the threshold of a lot of that.

There's a lot of doomsaying. There always is. But in terms of discovery, in terms of what we know about the world, in terms of health, you look at chemo and the number of friends of mine who have been treated and who have passed away. You look at chemo and go, “That is prehistoric.”

I think that with AI, with the amount of compute going on in healthcare and other realms, there are solutions coming through. Not fast enough, but it's so exciting, what we're involved in.

It's very exciting for me, too, to hear you say that because I don't actually have a driver's license. And so you could assuage me or relieve me of that necessity.

Yeah. Harry, you live in the most walkable, wonderful city. You don't need a driver's license.

Oh my God, dude. I never, ever need to drive. Thank you so much for doing this. Thank you so much for 11 years of friendship. Honestly, it means so much to me, and you've always been so kind to me.

You've gone from strength to strength, and that's my wish for you: keep going from strength to strength. You've been a great, great voice in this environment, a great voice in the world.

Thank you so much, dude.