二级市场为何正在吞噬IPO｜All-In 流动性二级市场圆桌

Everybody wants access to these private markets. Joining us right now to discuss all of this is Kelly Rodriques. He's a Forge CEO. We see a world where the private market opens up and is accessible to any US and global investor. There's 19 companies in the private market AI basket. These companies have grown on average 300%. Please join us in welcoming Gavin Baker, managing partner and CIO of Atreides. The ROI on AI has empirically, factually, unambiguously been positive. Investing is the search for truth. We welcome in Brad Gerstner. It's good to be back with you. You have a program called Invest America. I think we have a historic moment right now to get everybody into the game of capitalism.

Do we have a few slides from Brad to kick this off?

Let's start like old times. I was backstage and said, “Gavin, do you know we're talking about secondaries?” He's like, “What do you mean?” So let's just set this up for everybody. The room's full of people who are allocators. People are looking for distributions.

This is the secondary market over the course of the last decade. This is the amount of money going into venture capital each year and the amount of money coming out of venture capital each year. The red line represents the net effect of that. So, Chamath, we're in about 5 years, right, where a lot more is going in than is coming out. But the secondary market is at record volume.

I call these companies quasi-public companies. These are later-stage companies, and there's buying and selling going on every day. Look at that, Jason. Relative to the 2021 peak—we thought that was crazy at the end of 2021—we're double that now in terms of secondary transactions.

This is the amount of employee secondary activity. This is people buying into Anduril, Anthropic, and SpaceX. This now represents 31% of all primary venture activity in 2025: buying into these secondaries. Secondaries are now competing with IPOs and acquisitions as the principal way these companies are exiting.

I thought that was a decent setup to start the conversation this morning, just to level-set how important secondaries have become. The final point is that secondaries over the last couple of years were trading at a discount to the market. If we wanted to sell shares in one of our companies to buyers out there, they were willing to give us 80 cents on the dollar in order for us to get liquid so that we could send DPI back to our LPs. Today, it's at 106%—a premium in the market as a whole.

This doesn't include some of the wild west of SPVs that have been unwound recently. People are charging 10% loading fees, double carry, and a lot of gray-market, off-market stuff.

This is also having a profound impact, Gavin, on employees at these companies. I want to hear about that because you've seen it up close and personal with SpaceX, and they have a very orderly process here. So why don't we start there? What impact is this having on the employees, Gavin? And then, on the market, how orderly is this? Who are the buyers?

Are the buyers the suckers at the table? Are these family offices and high-net-worth individuals who keep hearing us talk about Anthropic, SpaceX, or Anduril and say, “I have to own the name,” without being discerning? So, Gavin, maybe you can start by talking about the impact on the SpaceX employees you saw firsthand.

Well, maybe broadening it beyond SpaceX, I do think that if companies are going to stay private longer, this is absolutely necessary. There are a lot of people who are very wealthy on paper but actually cash-poor. If you're making tremendous sacrifices because you work for a company you really believe in and you're contributing a lot to that company, it's hard if you can't buy a nice house for your family. It's hard if you can't afford to do nice things.

Especially in year 7, 8, or 9 of working at the company, when you tell your spouse, “We're worth $10 million on paper, $30 million on paper, and you don't own your home.”

Yeah, or year 15. So I think this is necessary and important. Whether it's good or bad, I think it's very clear that companies are going to stay private for longer.

What's the reason to stay private longer, truly?

I don't think there's actually a good reason to stay private longer.

Hear, hear.

I completely agree with you, too.

Why has it happened? Let's just call it what it is: founders don't want to be under a microscope. They want to build, enjoy life, and have it easier than being in the public market under a microscope.

I think there's a perception that life as a private company is easier, that you have more freedom, and that you can think long-term. I don't agree with this. I always think about Mark Zuckerberg's commentary that, had he been public—Facebook, I won't call it a near-death experience, but long ago, it's difficult to believe—in 2010, Facebook did not believe in apps. They believed in something called HTML5.

HTML5.

It was the cataclysmic debate, and it was me versus Bret Taylor. I was like, “Apps. I want to go build a phone.” Bret was like, “HTML5.” Zuck picked Bret and spent the next 3 years unwinding that decision.

Absolutely. Basically, the idea was that the iPhone had come out and initially there wasn't a big app ecosystem. There was a thought that there was no need for apps—you were just going to use the web browser on your phone—and HTML5 was a way of making websites look mobile-native.

Dynamic.

Yeah, and this seemed like the future to a lot of very smart people, including Google and Facebook. But it was not the future. It was wrong. What Mark Zuckerberg has said, I think several times in public, is that he profoundly believes that, had he been a public company during this internal debate, things would have been different.

The detail was actually that I went to Zuck and said, “I need $1 billion to build this phone.” We were in this moment in 2010 where we could have the third leg of the stool. There's Android and there's the iPhone, and neither had really taken off yet. He said, “We don't have $1 billion.” I said, “But the public markets will give us $1 billion.” He said no, but then we went public a year later.

But that year made all the difference.

Made all the difference. And he said that, had he had the constant pressure-testing from public-market investors, there was a dynamic. I was talking to another CEO here this morning. When you're the CEO of a private company, you are the most special flower to all of your investors. You're as important to your board members, particularly if you're really successful, maybe as their families or parents. Board members think about you a lot.

Once you're public, you're one of thousands of companies. That's its own dynamic. But the consequence of this is that private investors are often selling to management teams. At some level, that can mean telling management teams what they need to hear because you want to be able to keep participating in the rounds.

Once a company is public, you can buy or sell as you wish. This means that investors feel freer to give management teams feedback. Zuckerberg said, “Had I been public, had I been getting rigorous, detailed questions from really smart public-equity investors, I think I would have—”

By the way, the second unwritten story of that, which has never been said, is that he called me. He's like, “Hey, man, what the fuck is going on over there?” And I was like, “Yeah, I know,” because I had just left. Then we wrote a deck, and I walked over to Zuck and said, “Here's the deck of what you need to do.”

Yeah, do these things. Well, this is a key point. I think, Gavin, when you're private, you do not get clean information as the CEO and the management team because people want access. Once you give the truth or ask the hard questions, you might lose access.

100%.

The sycophantic nature of private markets is real. Now, an exceptional CEO—

Elon.

—seeks out negative feedback. But not many CEOs are wired that way. And, by the way, I do think we have to give Brad credit. That was a very good deck. You said it back in 2012.

No, because he did a second one. He had—

When he did the second one, he did the open letter to Zuckerberg at the end of—was that at the end of 2022?

October 2022.

Why don't you call it “Get Serious”? What was the—

“Time to Get Serious.” That was impactful. Those are 2 very impactful ones.

So, you're hearing the bulls on going public, but Kelly, take the red team—the other side—because you're on the other side. You built a private business and sold it to Schwab, so clearly one of the largest financial institutions is going to ram its way into this market. But then you're seeing a lot of pushback.

Anthropic is saying, “Hey, dissolve these SPVs.” OpenAI, I think, was saying today, “Dissolve these SPVs.” Should we dissolve the SPVs? Where are they coming from, and why are you on the right side of history?

Have you had to dissolve any of the ones on your marketplace?

No. No, look, first of all, being a private-company CEO for most of my career and then being a public-company CEO for 3 years, I recognize that the job is incredibly different. It's much less fun. You're not doing—

What do you mean when you say “much less fun”?

Turning into an investment manager primarily as a public company CEO is a very different job than being a visionary, product-first, first-principles business. When you become a public company CEO, everything changes. I would say, in the world we're in now, the kind of capital you can raise—the kind of capital that was represented in the very last discussion—allows you to extend your private life. SpaceX has been a private company for 24 years.

The reality is these SPVs that are now emerging because these companies are getting so big is that a market is trying to happen, and a company like SpaceX has done this extraordinarily well. They've run essentially liquidity programs for almost a decade because there's so much pent-up interest in both being an investor and getting liquidity, for some of the reasons that Gavin was mentioning.

So I think what we see now is the next phase of this. This Schwab deal with Forge basically says to the world, “This is a real asset class. It's more than just secondaries. We're going to put these companies' equity into fund products, into very well-managed, regulated SPV structures,” because they do serve a purpose in the market.

Yeah. But how do you convince Elon specifically to give you access to that when he wants to do it himself and he has a team, and every 6 months he runs it himself? How do you get access to that? What's your pitch to the next Elon?

Here's the pitch. The pitch is, you're going to go from being a private company eventually to a public company. What Schwab represents is 46 million investors and $12 trillion. This will change capital access and the way that you distribute your shares moving from private to public.

How did that work when you pitched them on that?

Well, I'll tell you. We got our first SPVs on SpaceX in 2018 and 2019.

Was he okay with it?

Absolutely. Totally permissioned. And then, as we got closer to the IPO, we said, “Guess what? We've got 30 million retail investors that would like to have a $50,000 slice of SpaceX.”

And he went out publicly and talked about having broad-based distribution—

At the IPO price.

At the IPO price, and Schwab was named one of the IPO allocations.

Beautiful. I do think this is actually a very effective pitch. I think a lot of these CEOs are a little bit ambivalent about it, and I think they understand that maybe the institutions who are investing in these private rounds may represent unions or retirement plans. But I do think they like the idea of democratizing access and, if they're building something that they think is great, giving ordinary Americans an opportunity to participate. I actually think that's a very appealing story to a lot of these CEOs because they're capitalists and they understand the power of equity.

So, Brad, what is the downside, then, because you're part of the go-direct movement now? BG2 Pod, officially 5th bestie. Gavin, officially 6th bestie. You got that? That's Gavin. That's new news. We officially made you 6th bestie today.

But does that mean I'm definitively behind Brad? Because that's the real news.

You're standing behind Brad. You're just giving him that big bear hug right behind him.

Wow. So, are you saying I'm the big spoon?

You're the big spoon now, in the side drawer with the extra spoons. But Brad, it's getting very weird very quickly.

In all seriousness, with great power comes great responsibility. Sometimes the enthusiasm people can have can exceed reality.

Correct. Going direct, you've become more measured. I've noticed as your profile has gone up. You were talking stuff down on CNBC a couple of times, saying, “Hey, I don't think the average American needs to be in some of these companies. There's time.”

I get worried at this point in the market stage, particularly on CNBC, where you're talking to retail investors at home.

Yes.

I was one of those retail guys looking up to everybody on this stage, trusting everybody on this stage.

And when people are telling you to YOLO into, right, double-fee-structure SPVs and all this—

It's time to be careful, to do your work, to be thoughtful. We're in this because we want this to be durable democratization for a long time.

Yeah.

We want to build trust among those who feel left out and left behind in capitalism. We all think that we need to go public sooner. The reason I think it is destabilizing is when you're creating trillions of dollars in private value and 80% of America thinks it's a scam, where they're left out and left behind. That's when they come rushing in, and they could be not so good cards, right?

So all I'm saying, like I said when they asked the question on CNBC last week, is: If you had $100,000 of fresh capital and you were sitting at home, is today the day that you would shove it all into the market? And I said, no. I think about it in sizes, right? We just had 2 of the biggest months in the last 10 years in the public markets. They've been big months.

So, if I had a stack of 100, I may put 30 to work today. I'm never going to pick the bottom. I'm never going to pick the top, but I certainly wouldn't be putting it all to work. And I'd say the same thing about late-stage privates. People who are YOLOing into this stuff then feel really disappointed. They're like, “Hold on a second. I bought the SpaceX IPO and it didn't go up 3x.”

Let me ask you, then.

Yeah.

Do you view this as exit liquidity for you? Would you shape your portfolio and returns and increasingly say, “You know what? I don't know when this guy's going to go public. Let me just pump the stuff out. Let me get the distribution, send it to my LPs, and just call it a day”?

Yes. We are selling into this.

You're selling into this, right?

I have LPs in this room who say, “Listen, we invested in your VC5 or VC6 7 or 8 years ago. If you can go sell a slice of that at 4 or 5x and we get DPI and it's priced really high, then go sell some of it.”

We often don't talk about this in venture land. Half of what we do is in the public markets. Gavin and I get up every morning and think to ourselves, “Should we buy today or should we sell today?” Venture capitalists don't think about the sell part. They think about the buy part.

So, if we're going to stay private for longer and we're going to have trillion-dollar private companies, and Databricks at $200 billion, you have to think about: Is today a day we should be selling some and returning it to our investors?

Doesn't it create, though, as Jason said, these very complicated personality dynamics where maybe you get shut out of a new company, maybe you get shut out of an incremental round, and there's bad blood because you're a credible investor and there's this signaling risk? Whereas in the private markets, if you and Gavin decide to sell, nobody knows.

Well, no. In the private market, nobody knows.

Exactly. In the public markets, they don't know until our 13F comes out. Okay.

But in the private market, it's always a conversation between me and the founder to say, “Listen, we're going to sell 30% of our position.” They never like it, Chamath. They're always like, “We wish you wouldn't do that.” They don't want it known, et cetera. But my job as a fiduciary to the LPs of this is to do that.

It does feel, Gavin, like we have crossed over from early-stage venture to a point at which there is a 3rd way. Either your company had M&A—and we saw in the presentation yesterday that during the wrath of Lina there was no M&A and they just froze the market—or IPOs. We did have some freezing of that market for certain periods. But this 3rd way is now fantastic.

I can tell you, as one of the earliest of the early, we are now selling pari passu into every chance we get because our average investment is at $10 million to $20 million valuations. When they hit $500 million, I tell the founder, “You're going to start selling at $500 million. I'm going to sell right alongside you so that I can invest in the next you coming into the market.” Everybody's fine with it.

But I can tell you, 6 or 7 years ago, when I did this with a company, they begged me not to participate. When they hit peak ZIRP in 2021, they begged me, “Jay Cal, you have to be loyal to us. You can't sell pari passu.” And I said, “You guys are clearing $40 million of the $110 million round. I'm just asking to be next to you, same amount.”

Can I ask Kelly a question? How do you systematize this so that it's like an exchange? If we just want to hit the bid, we can do it. What I don't like about the secondary markets is, I ask my CFO, he calls 5 guys, then my fund CFO calls 4. It's like ticket brokers. We get a bunch of bids, none of it makes any sense, and I'm already dealing with, as Brad said, the agita from the CEO.

It's got to be easier than this.

Yes. Look, 10 years ago, we said there needs to be infrastructure to pull this off. This can't just be a big shadow market. We're sort of at this tipping point now, where we spent the last 3 years building this brand-new platform so that a company could plug into it the same way they could list on an exchange and say, “We're going to offer liquidity.”

Furthermore, if you're a VC and you're on that cap table for 10 years and you want to offer LP liquidity, you can do it in—

To be specific.

What do you mean? You're like, “We would be plugging into Schwab's 30 million humans that are buying stuff on?”

There's a platform. We brought a platform with about 3 million investors, and now we're going to add 46 million investors to it.

Yeah. But wait, hold on a second. Aren't those accredited investors? Do they need to be? Because we just had the chair of the SEC on—

So today, if you're trading individual shares, whether it's in an SPV or directly on a cap table, you're accredited. However, there are products coming to market—we can talk about this in detail later—that have 60 companies, including SpaceX, as listed products for unaccredited investors with $500 minimums. And that capital for those funds will be the underlying—

Closed-end funds.

These are interval funds.

Interval funds. He's got one out now.

I think Naval just did USVC [?] as one of these.

Now, the closed-end funds are a very different bet because you're betting on FOMO. If you look at the underlying value of some of the assets in those closed-end funds, they have no bearing on the reality of what those underlying shares are actually worth. So price discovery is another key component of this structural shift.

But to answer your question specifically, if a VC's LPs want to recycle or want to get liquid, then a platform like this will allow them to recycle that capital and put it back into the next-vintage fund if they want.

I have a question for you based on this. When these returns come out, the mean return in venture is going to look incredible. The median return is still going to be—walk us through how people will sort through that and the reality of what's going to happen in the next year.

I think there are 2 very important things. One, I observe that if you were a venture firm and you don't have material exposure to one of these trillion-dollar-plus companies that you had many chances to buy into, not only are your returns not going to be good, but you're not going to have DPI on a relative basis—you're not going to have DPI.

There are exceptions. Great Series A firms may not have this, but their returns are still amazing with great DPI. I am beginning to see venture firms that don't have exposure to one of these companies behave in strange ways because I think they're starting to feel a little bit of franchise risk. Their DPI and their returns are going to go from, “Hey, top quintile, top decile”—

So they're doing unnatural acts, doing unnatural things.

They're writing what I see as call options, like a bunch of these Neolabs. “Well, I need a story. I've done something.” Maybe some of these call options pay off, but I do think they're engaging at some level and maybe—

They're chasing it.

They're chasing gambling terms. Whereas the people who have exposure to this are being a lot more disciplined because they know they're in a great position.

I think another very important dynamic is going to happen in the world of long-only mutual funds and crossover funds. Long-only mutual funds—my former employer, Fidelity, amazing place, love it—Baillie Gifford, Capital Research, Wellington, T. Rowe Price—they all can, per SEC rules, allocate up to 15% of their funds into privates. These are the biggest pools of capital in the world. They dwarf sovereign wealth funds.

But most firms, because they don't want to get in trouble with the SEC, say, “Hey, we're going to cap it at 3%, 5%, or 7%.” It was very public that Baillie Gifford was forced to sell SpaceX last year for regulatory reasons.

What's going to happen as these companies go public? All of these long-only mutual funds are, by and large, finding it hard to participate in private markets right now because they're at the limits of their self-imposed 3%, 5%, or 7%.

3%.

When a company goes public and the lockup expires, it moves out of that bucket.

Nice.

So this is going to be hundreds of billions of dollars of new late-stage demand coming back to the market after being out of the market for a while.

That's a lot of dry powder. There's a lot of dry powder.

The net trade is up. Then the marginal trade is up. Founders are going to be in the catbird seat. People are going to be looking to put money to work.

There's an interesting buzz going around about accreditation rules. We had the head of the SEC on the All-In interview show. We did it. They're going to have a sophisticated-investor test, something I've been talking about for a long time that would really democratize the way Invest America has access.

And then, funds. I've been getting pitched for years on, “Put your fund on blockchain,” or, “Sell your fund into this ETF.” Maybe you could talk a little, Kelly, about the possibilities around venture funds being more tradable, like secondaries are. Is that on your roadmap? Obviously, there's demand for it.

What would that do? I can tell you what that would do for my LPs—Brad's, Chamath's LPs, and previous funds. If you could come in and out of these funds the way you can come in and out of Anthropic, my lord, that could be incredible for folks who have a divorce, a life event, or just want a little more fluidity.

There's been secondary fund trading for a long time. I think blockchain and tokenization make it more efficient. That world will come.

But the question we're asking ourselves now is, if you're an LP in a fund that's holding something as valuable as this, are you really interested in trading your fund position, or do you just want to get out of—

The big winner, that name?

Our view is it's probably the latter. In some cases, funds will come to us and say, “We've got a vintage fund that has 2 companies in it that are 15 years old, and we can't clear that fund.” That's an application of liquidity to the market that we think is coming.

Are you worried at all over this next year about this idea of retail being exit liquidity for these 3 ginormous companies? Is there any risk? How do you bucket the risk? How do you manage the risk? What is the risk if something were to happen? What's the blowback?

I was talking with Brad about this yesterday. We're watching these valuations and these multiples. We had this conversation at dinner last night and said, “Wow, these are extraordinary, and people should come into this market.”

“Extraordinary” is a coded word for—

It's, you know, it's okay. Fine.

It's a bubble. You're saying you think the valuations are high.

I think the retail investor coming into this space needs to look down market and look at interesting opportunities that aren't the things that are on CNBC every day, and have access to them earlier.

We had a bunch of retail investors show up in 2018 and 2019 who wanted to be in SpaceX, and they're thrilled that they got in when the valuation was $30 billion. If the market opens up, that's what we'll be talking about: What do I want to get into now that's not at the very top of the market, getting ready to go public?

Also, Brad and Gavin—we're getting better. Shout-out to Bill Gurley. We're getting better at pricing these IPOs and not leaving money on the table. They're fully valued in most cases when they go public, or in some cases—

They're still mispriced. They're massively mispriced.

Well, no. We have seen some that have gone down after they go out, so—

Nothing good that anybody wants.

What do you guys think? Are we closer to correctly pricing them?

Gav and I have been doing this for 25 years. There are moments when the public market is undervalued relative to privates and moments when privates are undervalued relative to public markets.

Right now, everything in the world of technology is pretty fully valued. You can't have the parabolic moves we've had and think that everything is cheap. That's not to say that we're not going to go higher, but when you've been punched in the face many times, as all of us have over the last 15 years in technology, we know it's a jagged line up and to the right.

For the retail investor, as long as they have staying power—if you're going to launch a product, as long as the retail investor can stay in that product through the drawdown, they're going to do fine. The problem is most of them YOLO at the top because everybody gets them all jimmied up and excited, and so they're levering up. They're doing 2x-levered, meme-y trades and all this.

There are 14 ETFs launching on the day of the SpaceX IPO that are levered ETFs into SpaceX at, like, whatever, $1.75 trillion. This just tells me that there's a lot of signal. We may not be at the top, but we ain't at the bottom.

We're bouncing along. The top might be fair, you know. You've got to allocate accordingly. That's what active management is about.

If we're not thinking about that, when people are puking into their garbage cans at the start of the Iran war and the market is down, Gavin and I are looking at each other and saying, “Good God, these Anthropic revenues are off the charts. We've got to get more dollars at risk. Shove more onto the table in both Anthropic and public-market stocks.”

But then 75 days later, it's all changed. Right now, the market—

Have you guys ever been in a market cycle where these moves are just so concentrated in time, where you take 1 or 2 years' worth of moves and compress them into 30 or 60 days?

This is nothing relative to 1999 and 2000.

Nothing.

This is nothing relative to that.

Describe, describe. Yeah, just package. Sometimes they wake up.

What was 1999–2000 like? In terms of, if this is a roller coaster, what was that?

Yeah.

And what was that? I mean, this is like a roller coaster that's kind of a gentle sine wave.

1999 was Vegas on a Friday night after way too many drugs.

Okay. It was out-of-control nuts. CMGI had no revenue, and the stock went from $2 to $2,000 over the course of 6 months. They bought Foxboro Stadium, they were on the cover of Time magazine, and they were out of business 2 years later. That is very different from Anthropic, OpenAI, and SpaceX. These are extraordinarily real businesses, so I think the better comparison is 2021.

Right. Where valuations get ahead of themselves, or they're at the top end of the range. We could have a normal, run-of-the-mill consolidation in the public markets, with the semiconductor index down 10% or 20%, which means high beta would be down 30% to 40%. A lot of people who just got in would be panicking, but the people who have been in for 6 months or 3 years would notice that it's just a blip. So I don't think it's at all like that.

Okay. I have a question for the 3 of you. Four. Yeah, final question. Take the top 10 private companies off the table. Forget those—you can't pick those. Give me a sub-$50 billion private company, something in the tens of billions rather than a few hundred billion, that you could buy a secondary in today, that you don't own but would want to own. I'll start with you, Brad. Just go around the horn.

I would take a company in what I call inflection growth, Jason. These are the 1,000 companies that are over $3 billion, but let's call it sub-$50 billion. I think it's the trickiest area of the investing landscape because they're the beneficiaries of high valuations, yet they still have binary risk.

Right.

Right. Like Anthropic, OpenAI, and SpaceX—I don't think these companies have binary risk, but there are a lot in that bucket that do.

It's a hard question. I'll give you one. I'd say Sierra, Bret Taylor's company.

What do they do?

They're building basically Salesforce—

Agent-native?

Got it. Sales, marketing, and customer-service agents that are agent-native. I'll give you the downside and the upside. We also own a company called Parloa in the same space in Europe that I think is really interesting. The downside is that OpenAI and Anthropic say, "We're going to do this," and all of a sudden it eviscerates hundreds of millions of dollars in revenue. The upside on these businesses is that they have already built very sophisticated agentic layers, and Meta, Google, and SpaceX all come along and say, "We want to buy you because we want to accelerate our path into agents."

I'll give you the name that I was convinced of today—yesterday—by Thomas Leant [?], which was Revolut.

You know, I had always had the kind of—I owned some Coinbase, I owned some Robinhood, we did all of that stuff. It was fine. I kind of ignored fintech, and Thomas backstage gave me an incredibly compelling pitch for Revolut. I actually went and said, "Okay, show me what the Revolut share price is in these secondary markets." I got curious. Maybe I should pick up some of that. So that would be my answer.

What does Revolut do? Explain it for the audience.

It's a bank. It's a bank. What's interesting is that it's a neobank with a completely next-generation stack. It's kind of what Brad said: that theme of rebuilding it in the modern era and unbundling the incumbent has a lot of legs. In a regulated market, that has a ton of legs. They're doing really well in Europe, they're coming to the United States, and the founder seems to be an absolute star. They have tens of millions of customers and 14 lines of business. They're like a billion—curious like that.

Do you have one that you've bought recently, Gavin?

No. I would just say that 2 names we've been involved in publicly are Arrcus and DriveNets, and they're both in the networking space. As data centers get more specialized and complicated, you're going to have increasingly specialized chips. It's called the disaggregation of inference, prefill, and decode. To make all of these chips work together like a symphony, and have the right chip for the right job at the right time, I do think we need to reinvent networking. Arrcus and DriveNets are coming at it in a very different way.

You've been one of the earliest. I'll give you credit. You framed this on a podcast that I saw: there's an impending supercycle in infrastructure networking and silicon, and you've really been at the front of it. I buy into it completely now, too. It's really good. Any names?

Neura Robotics in Europe.

Neura Robotics is a company name?

Yes, and it's AI-powered logistics robotics.

Love it.

They're not in the main strip of high-value real estate in Silicon Valley. They're in Germany.

A quiet company, big investors, $100 million in revenue, kicking ass.

Love it. Well, I have a couple of theses that I've been looking at. One is what Elon is helping put into space as the price goes down. We did a direct investment on the cap table and an SPV for Vast, which is building space stations, and we think they're going to win.

The other one is what I'll just call Uber 2.0. Gurley and I took a lot of notes on that, Brad, as well. We were able to do Zipline, and we put a small ticket size into Zipline as well, because if you can take the delivery cost down from $15 to $5 and then eventually $2, that's going to drive consumption massively. It's going to happen in the air.

These drones had such a false start that everybody gave up on the entire sector, and now it works. It was just a very simple innovation that Keller told me: the drone stays up in the air and drops a tether with the box in your burrito. If you grab the tether and pull it, it just comes down. You don't have to land this giant robot in your backyard with blades spinning to kill your dog.

I think there's actually a very important point on Zipline. It's amazing and has done great things for the world. My firm, Atreides, is also involved in Zipline, but Zipline started with the hard thing: to make anything autonomous work, you need to get it out into the world and gather real-world data. This is how AI works, and it's hard to get approval to fly things around autonomously in American airspace.

Keller had the idea that they were going to go to African countries and help deliver medicines to small villages. They focused on maternity care, and they have cut the maternal mortality rate in some of these African countries by 90% to 95%. You're in a small village, there's 1 midwife, there's an app, and a woman goes into labor. They press a button, and an hour later, a Zipline drone drops a refrigerated package of modern medicine, blood, and everything needed. They did it for 7 years, and it's had a huge impact on health outcomes in these African countries. Now it's come to America.

This is an incredible story, and I've basically now restructured my firm to do the barbell. I missed the seed investment. I turned him down because I was like, "We don't invest on that continent. We don't have any insight into it, we don't understand it, and hardware is hard." He has the email, whatever.

I've stayed in touch with him, and he said, "Listen, I figured it out." I said, "Hey, I have the syndicate. Let me see if I can correct that mistake. May I invest?" He said, "You're my dream investor. I've wanted you on this whole time, and it's just so important."

We've been friends all this time, and I've had him on the pod 3 times. He said, "When are you going to be on the cap table?" I said, "You know what? Learning from you guys, specifically this late-stage stuff, I'm like, well, I can do that." And here we are.