# 20VC: Trillion or Bust: The New VC Game | Coreweave Sell-Off Creates Panic | Town at $1B, Anduril at $90B: Revenue Multiples Out of Control | General Atlantic Eyes IPO & Modal Raises $750M at $15.75B 20VC · 2026-10-22 · 61 min · https://thetwentyminutevc.libsyn.com/20vc-trillion-or-bust-the-new-vc-game-coreweave-sell-off-creates-panic-town-at-1b-anduril-at-90b-revenue-multiples-out-of-control-general-atlantic-eyes-ipo-modal-raises-750m-at-1575b ## Transcript Speaker 0 You have to have billion-dollar positions in everything in big-boy venture today, I think. It’s billion-dollar positions or nothing. There’s no point investing otherwise with large funds. The entry price—$50 million, $100 million seed rounds—you can’t make money. Speaker 1 If the end-user demand is there, then everything is going to work out. Speaker 0 CoreWeave is B-tier compared to NVIDIA and friends. Oracle is B-tier. Speaker 1 Compared to NVIDIA, we’re all fucking B-tier assets. How about that? Speaker 0 I actually think code review is the single most important category in software today. Speaker 1 Coding is the mother lode of LLM value. Speaker 0 We’re just taking revenue multiples too far today. Bending Spoons is like the Berkshire version of software. Harry Stebbings This is 20VC with me, Harry Stebbings. It's my favorite show of the week. Rory O'Driscoll, Jason Lemkin, the biggest news in tech. So what are we talking about today? Well, are the AI cracks starting to show? CoreWeave is hit pretty hard. Next up we have Anthropic. The talk of IPO town is it could be priced at 1.5 trillion, not the 2 trillion originally suggested. Then we move down to the world of private markets where we have Modal, which closes a $15.75 billion valuation fundraise. Angel tends at 90 billion. And then welcome to the agent land. We have Town that raised $108 million at a $1 billion valuation. This and so much more in the episode today. But before we dive into the show today, Base44 is one of those products I used. And I just thought straight away, man, I wish I'd had this when I started 20VC with $50 and no contacts in venture. I wanted to see what it could do. So I built a guest research hub for the show. Upcoming guests, company notes, previous interviews, sponsor contacts, follow-ups, all in one amazing location. I described it in plain English. A few minutes later, I had the product open in front of me. Database, login, back-end and hosting already handled and just all in one. It made it so easy. And that's what I loved. I didn't write a long spec. I didn't stitch five different tools together. I built the thing and I started using it. For any founder listening, that's the point. The faster you turn an idea into something real, the faster you launch. Apps, dashboards, internal systems, AI agents, whatever you need to build, Base44 gets you there. Go to base44.com/20vc. That's base44.com/20vc. While Base44 helps you build, AlphaSense helps you find answers really, really fast. We used AlphaSense on an investment that helped us close an $8 million deal. $8 million, baby, that's a lot of money. That's why I'm genuinely excited to have them as a partner on 20VC. AlphaSense combines AI with one of the world's deepest libraries of market intelligence, including expert interviews, broker research, earnings calls, company filings, and real-time news. Every answer is grounded in this incredibly trusted evidence and fully traceable to the original source, which is so important. So you can make really high conviction decisions with confidence. But the best part, they're building super analyst, an always-on AI analyst. So instead of starting your day with another search, you'll start with work you already done, your coverage monitored, the important developments surfaced, and your investment brief already waiting for you. See for yourself. Head to alphasense.com/20vc. That's alpha-sense.com/20vc. While AlphaSense finds insights, Intercom helps you serve customers. As AI agents become more common in customer experience, teams often end up juggling multiple siloed tools for every job. Well, Fin was built to change that. It's a single unified agent that works across your entire customer experience, from service to sales to success and beyond. Fin is the agent making perfect customer experiences possible for thousands of customers. It's powered by custom models, trained on years of real customer interactions, so it understands the nuance and complexity of customer service better than any other agent. That means faster resolutions, more consistent support, and just better experiences for every customer. It's also designed to be fully self-manageable, so you can easily improve and adapt it as your business evolves. No third parties required. Leading companies like Gamma, Asana, DoorDash, and crypto.com already use and love Fin to deliver better customer experiences. So for a limited time, you can get $500 a month in Fin credits for your first three months. Learn more at fin.ai/20vc. Speaker 1 You have now arrived at your destination. Harry Stebbings Boys, I just tweeted something that has got a rather large response from the founder community. I think venture is broken today, in the way that I’m meeting more and more companies that fit this profile. I just met a company today that’s at $20 million in ARR, growing incredibly fast—600% or 700%—and profitable. Brilliant business with a great team. Crickets. VCs do not want it. Is venture broken today? Speaker 0 Well, I’m sure Rory has a lot of thoughts. Let me add one thing first. We were talking about this company, which will remain nameless. I invested in a vaguely similar company doing similar things for similarly sized companies. They raised about a $3 million seed round, and they’re planning for that to be it—even with this type of growth, even with it all being agentic, even with it being all the things we’re supposed to be excited about: AI, voice, and resolution agents. It does something that’s actually pretty unique, but they’re literally planning to never raise again because you get the 15%–18%-a-month venture update, but no one wants to meet. I’ll put in a little more money, but it may even be like the old days, where your early investors may need to carry you if you’re not hypergrowth and checking the right boxes. The three-box check. Your investors may have to carry you. It’s like the old days. Harry Stebbings Jason, you’re right. You said earlier that the specificity box has really narrowed in terms of what’s interesting to investors. Speaker 0 Yeah. Harry Stebbings I’m seeing that more and more, with this concentration of capital going to fewer and fewer names and everything else being wholly unfundable. That’s more than I’ve ever seen in venture before, which is surprising for a sheep-like market like venture. Speaker 1 But what you’re saying is that, if you think about the role of venture financing innovation, there are lots of companies that look on the surface to have very good numbers but can’t raise money. That’s what you mean, correct? Harry Stebbings Correct. Speaker 1 I think that’s probably true. Let me give you the contrast, though. At the same time, I don’t think it’s just going to 10 or 20 companies in total, because there are a whole bunch of projects that probably wouldn’t get funded at all in another era, but are getting venture money: data centers in space, quantum computing, you know. So we’re very risk-on for stuff that has ultra-high upside. I think the stuff that’s struggling to get funded is, as you say, software for specific markets where there’s some inarticulate fear that AI will make it all go away. Is that a fair next-level summary? Harry Stebbings Yeah, I think so. I think you’re absolutely right. It’s this power law on steroids requirement where it’s like, “Unless you can pay me the trillion-dollar market cap, I’m not interested.” Speaker 0 To answer your question a little more tactically, Harry, I think you’re right that you roll with a high-powered crowd. If you want Sequoia and Benchmark and one of the better partners at Andreessen to fund you, the box is very specific, right? It’s Cursor or better, right? OpenRouter’s okay, but let’s see what we can do better next time. For companies like you described, still growing at numbers that approach the iconic outlier metrics, here’s where I think the world is today: you’ve got to meet every VC. You’ve got to meet 100. It’s like the old days. You’ll find someone who takes that bet. It won’t be one of the five or seven Benchmark partners and all the rest, but someone will take it at a reasonable but still elevated valuation because it’s the best deal they’ve seen this quarter. Someone will take it. It’s the old days where you’ve got to meet 100, and there are growth investors whose names are not Index, Andreessen, and Bessemer, and they will take that bet because it’s a good bet. To Rory’s point, it could lead to a Listen Labs or better in some cases. You just better meet everyone on the planet now if you have top-decile numbers. Top-decile numbers used to mean an automatic term sheet for years. You’d walk into Rory’s office. Rory knew every deal since the dawn of time. He’d know exactly if you were in the top decile. If he believed in you and trusted you, you’d walk out with a term sheet from Scale. The world just doesn’t work that way today. Speaker 1 You might still walk out with a term sheet from Scale. But what you’re saying is there’s a small number of extraordinary potential startups that the very best VCs are focused on, almost to the detriment of anything else. That’s your point, Harry, correct? Harry Stebbings No, I’m saying the whole venture landscape has realized that certain categories are the most interesting and exciting, and everything else is uninteresting in a way it hasn’t been before. Speaker 0 Rory, back me up here. If you’re raising a $5 billion-plus fund, you have to have billion-dollar positions in everything. You even need a $5 billion position here or there, right? So what? A $2 billion Listen Labs just isn’t good enough, because the odds are pretty tough unless you have 51% ownership, which most founders are not cool with. It’s pretty hard to build a billion-dollar position. Speaker 0 If I were working at one of those mega-funds and we were playing our game, that would be my final question to the team: “How are we going to get at least a billion-dollar position out of this?” I’ll take 1% of Anthropic. I think that still gets me to a billion-dollar position, even with dilution, right? The Eric Schmidt story. But you’ve got to have billion-dollar positions in everything in big-boy venture today, I think. And I don’t mean “boy” with a gender attribution. Big folk. You’ve got to have billion-dollar positions or nothing. There’s no point in investing otherwise with large funds. Speaker 1 Well, I still go back to the statistics we cited last week, which is that there are 10 to 20 decacorns per decade in the entire US tech industry, and we clearly believe there are going to be 90 of them this year. One of those 2 facts is going to be at variance with the other, and I’m willing to bet on 50 years of data, albeit lifted up for growth. This isn’t just predicting based on the past, because you allow some kind of growth over the years for bigger exits. I don’t think you can confine yourself to saying, “I’ll only do deals where I can get a billion dollars to work.” What’s the minimum? A $25 billion outcome. You want some of those, but the whole industry isn’t going to be able to say, “We’re all going to do those,” and have more than 5 funds, because there just isn’t enough. Speaker 0 Maybe 5 is enough. Harry Stebbings Jason, I think Rory’s not going along with our thread here. Speaker 0 Listen, I haven’t sat in an Andreessen or Index partner meeting, but I’m pretty sure the math is something like, “How can we make at least a billion dollars off this?” It’s not even enough. Everyone has the 4 different scenarios when they write up the investment memo. A billion-dollar position—not an exit, a billion-dollar position—is just the B outcome, right? This is 10% probability analysis: $25 billion or higher, 25%, you know, this percent; we only make a billion on the deal, right? We’re raising these $5 billion to $10 billion funds every 18 months. We have to have so many billion-dollar position gains to even be playing this game. It’s just math. It’s just the maths. Speaker 1 To take the other side of the argument and talk about seed, which is your world, Jason, I’m assuming you’d agree that, from a multiple perspective, it’s going to be far easier for the best seed investor to generate a 5X net fund than it’s going to be for someone writing billion-dollar checks into things. Speaker 0 No, it’s harder. It’s much harder. Speaker 1 Okay. Speaker 0 Because unless you really do the Roger Ehrenberg thing and find something in Antarctica, right? The entry prices are so high, and ownership is typically lower unless the seed fund is very large. And all these VCs are cutting my pro rata, Rory. How am I going to make it up on the back end with my SPVs? I can’t make it work. A 5X is so much easier. I’ll do it at $1 billion in town, and I’ll raise it to $5 billion next week, and I’ll take the rest of the year off. At $50 million or $100 million seed prices, you can’t make money. Harry Stebbings Just to add to that, not only can you not make money, you can do even worse because it’s a multi-stage product for those big players. They’re the ones who lead the rounds, and they crush your allocation. So not only does the price go way higher, but your allocation goes way lower. You have $300K to $500K at best. I’ve got this in a company that’s now going to $18 billion, and it still doesn’t move the fucking needle, Rory. Speaker 1 Because what? You only got $300K into the deal? Harry Stebbings Because you’re $300K in, and it’s at a $100 million price. Speaker 1 Yeah, no, that doesn’t move the needle. I’d agree with that. Harry Stebbings And so not only do you have a way higher price, but you have a way lower check going in. Speaker 1 Look, over time, in a rational capital world, the return multiple from early-stage investing should be higher than growth, just logically, over an extended period of time. Otherwise, why would you do it? Harry Stebbings I agree, but wouldn’t you say that, for the first time ever, we’re actually going to see that inverted? Speaker 1 I totally agree. I think the next 5 years are clearly going to be inverted. Harry Stebbings Yeah. Speaker 1 It’s clearly going to be. I mean, someone said it. It was a very good line: “The aggregate market cap of SpaceX, Anthropic, and OpenAI went from around $400 billion 3 years ago to something like $4 or $5 trillion today.” Sorry, trillion. Adam, change billion into a trillion. $400 billion to $5 trillion, right? So you’ve got $4 trillion-plus of value created in the late-stage business. Nothing else matters. If your theory is there are going to be 10 more of them next year, then you’re right, that’s all you do. If your theory is that was amazing, but it’s probably not going to be a runaway business, then you make different bets. Harry Stebbings We’ll go back to schedule, which is: are we starting to see a crack in the AI trade? CoreWeave has had a tough week. I admit, I’m a holder. It has not been great. Rory, can you please provide some context as to what’s going on? What do we need to know as background? Speaker 1 Yes. CoreWeave had a tough week last week, I think down 15% or 20%. Bond prices went up, implying lower default risk. But zooming out, when you think of the whole AI trade, these neoclouds are always going to be riskier than anything else in the system because they’re underwriting high growth with large CapEx and lots of debt. You try to thread the needle. If you grow too slowly, you don’t make your customers happy. If you grow too quickly and demand slows even slightly, then you’re screwed. I’m doing this from memory, but I think CoreWeave has something like a $7 billion runway, $10 billion or $20 billion of CapEx, and $35 billion or $40 billion of debt. That’s the way you have to do it, because you’ve got to build these data centers and fill them full of compute. So it’s just very tight. It’s not like a software company where you can hunker down and cover your nut. There’s just no way out here. They’re always going to be risky. Any time someone wakes up and says, “Oh, AI is growing a little less quickly,” all these things are going to go down. Any time anyone says, “AI is growing amazingly,” they’re all going to go up. Last week was an AI-go-down week, probably because of the Anthropic numbers, which in retrospect was a bit silly. But there’s no doubt last week was a things-are-slowing-down week, and when that happens, CoreWeave jumps down. Harry Stebbings Was this a gross overreaction, then? Speaker 1 If the end-user demand is there, then everything is going to work out, including CoreWeave. I think their biggest customers are people like Microsoft and OpenAI. Provided the demand is there, they want the data centers. It will all true up. Harry Stebbings Isn’t that a generational buying opportunity, though? It’s down 25%. Of course demand isn’t cracking. We all see that every single day. Speaker 1 The problem is, if it were that easy, we’d all be doing it, Harry. The real question is how demand compares relative to what these guys are spending. Can they fill everything they have now? Of course, they can fairly quickly. But if you’re continually investing for next year, as I said, if your CapEx budget is $20 billion, it’s not enough for demand to go up. Demand has to go up by a lot every year. The beating heart of the whole thing, and we’re seeing people share it, is that you add up OpenAI and Anthropic’s numbers together. That gets rid of the shift in market share, and the growth rate of that overall is what’s going on in the AI economy. That’s the growth rate you have to look at. The question is, how do CoreWeave’s CapEx plans look relative to that? Speaker 0 The simple lesson is that markets are going to be very sensitive, as Rory said. Any bump with Anthropic, any bump with OpenAI, any bump anywhere, of course CoreWeave and Oracle should be hypersensitive, because they have massive debt loads and massive risk. But they’re also the B-tier players. CoreWeave is B-tier compared to Nvidia and friends. Oracle is B-tier, right? You can make money in the B-tier, but why would you want to hold the B-tier? CoreWeave exists because Nvidia and friends need it to exist for now. It has no inherent reason to exist. It isn’t producing a singular set of silicon, GPU, and software that no one else can do today. Oracle just can’t sell enough enterprise software. People don’t want to install enough of this crap on their servers, so they’re stuck renting out their excess capacity, buying GPUs, and doing Stargate. These are B-tier assets, and B-tier assets should trade at a premium. More importantly, they should be more sensitive to shocks in the system. The A-tier assets should, in theory, be less sensitive. Speaker 1 The only comment I’d make is, I think B-tier, A-tier, CoreWeave versus Nvidia—I understand what you’re saying, but it’s almost like you’re comparing apples and oranges. Nvidia’s core IP, yeah, it’s the dominant provider of the whole ecosystem. Look, compared to Nvidia, we’re all fucking B-tier assets. How about that? Speaker 0 Yeah, but why buy the B-tier? Harry just asked, “Listen, why buy the B-tier?” If you just look, it’s like IPOs. Speaker 0 Why buy? Braze is getting acquired for—whatever—$5.8 billion. It’s down 40% from its IPO. Why buy any IPOs that aren’t A-tier? There’s just nothing in the public markets. You can make money on the B-tier if you get your nice seed entry at an $8 million post, right? It’s the start of this conversation: it’s A-tier or bust. The discount’s too high. What the hell am I gonna do with my CoreWeaves? Harry Stebbings Yeah, Rory. Speaker 0 In 5 or 6 years, I’m not going long on it. I want to hold—I want—I’m not passing CoreWeave to my great-grandchildren. I’m not putting it into their 529s that I’m giving to my great-grandchildren. There’s no way. Speaker 1 This is the “Jason stock picking is dead” argument. This is the public equivalent of the private comment: only the biggest companies matter. Again, the great thing about this is I can prove you wrong more easily in the public markets. When you go back over a year, Nvidia is up 18% and Nebius is up 50%. SOXX—I mean, that’s… Micron’s up 350%. The truth is, you buy the second-tier assets because you think they’re gonna pop. I hear you, but no. Speaker 0 Well, let’s see if you put 20% of your personal portfolio into those Microns and Shmicrons. There’s a lot of risk there, right? Speaker 1 Agreed. I wouldn’t dare. Harry Stebbings So, Rory, you know we have Jason’s IC. We’re gonna have Rory Recommends. Speaker 1 No, we’re not. Harry Stebbings Yeah. Speaker 1 No, we’re not gonna do it. Genuinely, we’re not gonna do it, because I’m not gonna sit here and have a… One thing I’ve learned about doing this now is that pulling stuff out of your ass because you’ve done a little bit of work is just not right. No, I’m not gonna do it. Harry Stebbings Isn’t that venture? Like, that’s your job. Speaker 1 Yeah. So, no. Harry Stebbings Okay, okay. I’ll leave it there. It’s okay. We can talk about what maybe caused a slight crack in that, which is Anthropic’s numbers. Anthropic announced that they’d turned profitable on an adjusted basis. Growth slows, and the IPO price is potentially rumored to be going down from $2 trillion to $1.5 trillion. Q3 revenue was $16.8 billion, up 45% on Q2. 85% gross margin, which we’ll talk about separately. What should we take from these numbers? Speaker 1 I think they were good. Look, again, it’s— Harry Stebbings You should tell your voice that, Rory. Speaker 1 Sorry, long day. Growth is clearly slowing fairly quickly, but from an unsustainably high number. It had 10x’d for the last 3 years. It’s highly unlikely to 10x, you know, from what was the run rate coming into the year: $10 billion. If it has 2 more 10x’s, then what is it, $100 billion? You’d be at a trillion. Growth was going at a crazy rate. This is all about how quickly the growth is going to slow, right? Honestly, I thought people might be sitting there going, if you add linearly—if you did $5 billion in Q1, $11 billion in Q2, $16 billion in Q3—you kind of lean in and you’re at $20 billion, $21 billion in Q4 if you keep adding $5 billion a quarter. I’m not saying they will, but, right? Then you start next year, which would be when you go public—the future year. If you’re at a $20 billion run rate, adding $5 billion a quarter, then you do $105 billion, $110 billion. Now you’re saying it’s 12, 13 times forward revenue for the most exciting company in the world. It doesn’t feel crazy. It doesn’t take a lot of squinting to make that work. Speaker 0 I’d add just one thing. For what it’s worth, tactfully, Rory, I think you’d agree: Chamath’s tweet was just, “Raise whatever you need at $1.5 trillion and move on,” right? If you can raise it at $2 trillion but it’s gonna stress your capital needs, it’s just a valuation. In absolute terms, everything’s great, right? This is unprecedented on all levels: great team, great application. What I suspect will be a reminder is that it’s gonna be very choppy after the IPO, because, going to Dev’s point when he was on the show, developers are not loyal. The amount of change in market share since this show started between Anthropic, OpenAI, open weights, and Jev—which didn’t even exist, what, 45 days ago—is huge. This is a winner, but there’s only so much predictability in the incremental growth and in the market share. There’s only so much ARR in the ARR here, and depending on how effusive the markets are, it may create stress. It wouldn’t surprise me if at the end of next year Anthropic’s market share is down 10% from where it is today. It could still be great on an absolute basis, but down 10% wouldn’t be a shocker at all—not just in tokens, but in revenue. Easily could happen. Look at, despite whether OpenAI’s at 70 or 50, it’s a lot. Speaker 1 When you say the share is down, you mean market share? Speaker 0 Yeah, market share could be down, right? So when you’re building your models—not just DCF models but your top-line models—you could be materially off from your good case, your best case. It should be expected. It should be expected. Speaker 1 To state the banal, it’s a little scarier than the average recurring-revenue business, right? That’s what people are wrestling with. How do I think about it? If you have continued market-share shifts between the key players, then the next shoe to drop is typically, “Well, I wanna keep my market share, so I’ll look at price.” Everyone’s trying to work out, from a game theory perspective, what that looks like as a P&L going forward. Which is why, going back to the boring thing I keep saying, provided the end-user demand at the top is there, everybody can win. We all love businesses where everybody can win, right? If that starts to slow down and price competition becomes literally the thing you gotta do to take it from the other guy to keep a growth rate going, then it could get pretty tough. Harry Stebbings Do you think Sam was super strategic raising his supposed round at $1.4 trillion, in terms of some form of anchoring and messaging to the market? Speaker 1 I think when you need money, you take money. I don’t know if there’s a big, sophisticated “Oh, $1.4 trillion as messaging versus $1.5 trillion.” I think it’s as simple as, “We spent a lot of money on compute. We didn’t get our IPO done. We’ve spent all of Masayoshi Son’s money. Better get another $40 billion at a price.” What are the constraints? It’s gotta be up on the last round, otherwise the messaging is shit, and it’s gotta be down below reality to leave room for the upside. There’s only a finite number of digits between one and two. Well, there’s an infinite number of digits, but there’s only a finite number of one-decimal-plus digits between one and two. Harry Stebbings There are a couple of elements beneath this story—or beneath Anthropic this week. One was Replit making OpenAI its default. Jason, I’m intrigued here. Anything from you on that, especially given your Replit power-user status? Speaker 0 I think it was a stupid media story. We use Replit about 14 to 15 hours a day. Basically, Replit has its own router in it, its own model selector. It can be invisible or visible to you, depending on what category you’re in. OpenAI models are just cheaper for roughly the same quality output in Replit, so they just made them the default. TechCrunch miswrote an article saying they were no longer working with OpenAI, which wasn’t true. They just flipped the default from Opus and Fable to Astra. That’s just the default, and it’s what we’re all routing, right? We’re all open-routing one way or the other. If you’re Replit or Lovable or Base44 or v0, you have these extra incentives because your average user just wants high-quality output. They’re very incentivized to get this quality pricing just right for you. Unlike the average idiot preconfiguring OpenRouter themselves and hoping they get it right, these guys are doing the work. It’s more interesting to watch what Replit, Lovable, Base44, and v0 do because they have an incentive to nail the routing. They have a very strong incentive. They can only partially pass it on to the end user, right? So we can squint and see the future based on what these guys are doing. What Amjad’s saying is, “The quality is sufficiently peer-level here, and the price is 30% to 40% better for us today. Why would we use Sonnet and Opus?” And you know what? Next week he may be at a different event, get on stage, and just go into this market-share thing with Anthropic. It may change. They may not have fully thought through how to use Sonnet 5.5 and Haiku 5.5, and they have Kimi and others. But if Kimi actually was so great, Replit would be using it entirely—which is maybe the more interesting learning that was under-discussed, because it will route through them too, and it doesn’t. Amjad didn’t get up on stage and say, “We’ll route everyone through Kimi,” did he? He said, “We’re going from Claude to OpenAI because it’s just as good and materially cheaper.” Harry Stebbings Well, what’s the takeaway from that? That Kimi is not as good as people think? Speaker 0 In practice, he’s not seeing the benefits for his billion dollars in revenue or for his customers. He’s not seeing the benefits, or he would just auto-route it. The customers are very cost-sensitive at Replit and Lovable. The SMBs are very cost-sensitive because they want everything for, like, $25 a month, or they even have free LLMs. Enterprise customers are very cost-sensitive because they don’t want massive variable bills for product teams going off and building prototypes. It’s a super cost-sensitive model, right? Hosting, databases, LLMs. If you slow down and see what they’re doing, they’re each at, like, $1 billion in revenue. It’s a little lab that we should be paying more attention to. It’s a more interesting lab than the idiots spending all the money, misallocating it, and hoping it works. That’s a great business model, but we learn much more from the massive investment these guys are making in optimizing high-quality output for low cost. Maybe others are doing this, each at $1 billion in revenue, but I don’t think that many people are doing it for end users. Speaker 1 I think it’s a great point because, yes, it is a very objective, calculating test of what value you’re leaving on the table by paying top-tier prices for less-than-top-tier intelligence. What do you actually need? With a billion dollars, you’ve got enough time to really think about it. Harry Stebbings Jason, you are the best user and analyzer of these tools. On Anthropic, what happened this week: Claude Code Review is out. We saw Boris Cherny tweet about it. I’d love your thoughts. Which startups are destroyed by this? What did you think of it? Speaker 0 I don’t want to talk out of my rear because it’s changing so much, right? Greptile’s super hot. They just raised whatever this week, right, $75 million. I think Rory’s in CodeRabbit, or I could be wrong. They’re an epic player. All these great folks have benefits. I think, in many cases, they have the benefit of being able to run on LLMs, which is very valuable: having one LLM check another LLM’s work, having OpenAI check Claude or vice versa. Just like there have been since the dawn of time in software, there are a lot of benefits to having pure plays. But it’s also always scary because this is the core of Anthropic. Claude Code is the core of Anthropic’s engine, as we’ve talked about. Making code review, making all of this process, is arguably much more important than building code today. You know how much time we spend reviewing code now because the LLMs are so good at getting you 95% of the way there. You talk to any engineer, and they’re like, “Oh, I’m out at dinner, and I built the feature for Harry. I’m having a barbecue, and my phone is building the feature for Harry.” You did build the feature that Harry asked for, but you’re going to have to spend the whole next three fucking days testing it and fixing it. So, how big a deal is it? I have to imagine it’s a bigger deal than the nothing burger, which was Claude Design or whatever the heck it’s called, right? That didn’t kill Figma, but how much? It’s so hard to predict. It’s scary if you’re working at a competitor. It’s always scary. No matter what you say on Twitter, it’s always scary. Even if it maims you 30% or 40%, you just have to be better. Speaker 1 One of the things we said last week, and I stand by it, is that the best markets for standalone AI apps—and I put coding effectively in that category—are the ones with the highest growth rates and also closest to the sun, closest to the foundation model. In fact, I just did a tweet on that today. It’s the Icarus moment. Icarus had no choice: he wanted to get warm, so he had to fly close to the sun. These big use cases are interesting. Code generation, code review, legal—they’re all very close to the foundation models, and every six months you have the terror moment of, “Oh my God, they’re gonna do it,” right? That’s the downside of being in that real estate. But the upside is that they are marquee, superb use cases for the technology. As a result, CodeRabbit and all these companies are growing faster than we’ve ever seen. If you want to avoid that risk, Harry, then you go back to your dull little 3x year-on-year software companies that you were bullshitting on about earlier, right? The ones no one wants to invest in. You can’t have it both ways. You can’t sit there and go, “Oh, I want infinitely fast growth, utterly undifferentiated from foundation models,” even though the only reason we’re getting the growth rate is because the foundation models have unlocked new capabilities. It just doesn’t make sense. That’s not a thing. I’d like to be a lot of things in my life, but it ain’t gonna happen. Speaker 0 I don’t know if this is true in terms of venture outcomes or revenue, but I actually think code review is the single most important category in software today. By far the most important category. It doesn’t mean it will make the largest fortunes. It doesn’t mean there’ll be a $600 billion code review Cursor. But you cannot vibe-code your own SAP in Salesforce. The rate at which we’re building code is unprecedented, and the code now is pretty darn good. But it’s never production-ready or production-intent. This is what everyone misses. If we step back a minute and define code review more broadly, it is the most important category in building. It is so important. Everyone is building 10, 12, 14 hours a day in Claude Code or Cursor, whatever. But the amount of time it takes to review and process it is still an order of magnitude more than building, and we haven’t figured it out. We’re pretending that this work is complete. It’s not. If I could find the best next-generation true code review, bro, the best team, I’d make 10 bets because we need it. They may all get killed by Anthropic. We’ll see, but it’s much more important than code generation. This is back to Amjad’s point: I can rotate between Astra and Fable and whatever. They’re all great, right? Maybe Kimi if it catches up. But how am I gonna get it into production? Speaker 1 Agreed. Big picture, it’s pretty clear that we’re making way more software than we’ve ever made, probably orders of magnitude more. Just look at the Anthropic thing they did on employment recently, that paper they did on unemployment. Given that, the whole software development life cycle is gonna change, right? How that gets deployed successfully is probably going to be very different in an agentic world. I agree. I think there are a whole bunch of compelling related opportunities, not just around code review, but end-to-end memory management and testing. Sandbox is where you and I have an investment. I said it last week, and I’ll say it again: coding is the mother lode of LLM value. It’s where it all happens, and therefore there are gonna be a whole bunch of ancillary opportunities around that. They might not be $25 billion, so you guys might pass, but if they’re a $10 million outcome, I’ll take my humble check and be very glad. Speaker 0 Yeah, maybe I shouldn’t call it code review because, as Rory pointed out, it might be construed too narrowly. But build-to-prod is the biggest category. Speaker 1 Absolutely. Speaker 0 Build-to-prod. How do you go from “It’s built, it works, it works in dev, it looks great” to “How the hell do I get that thing into prod?” That’s where the big money’s gonna be made now. The money’s already been made in frameworks for build, and they’re great. Let Sam and Dario and friends duke it out every day. We all benefit. But how do I get this crap into production? Harry Stebbings I want to move, actually, to one of the most interesting rounds. I thought Modal closed at a $15.75 billion valuation, with run rate surpassing $1 billion. It’s a $750 million Series D led by Accel. I actually interviewed Eric the other day. He was amazing in the studio. How do we think about this round and what it means for the wider inference market? Speaker 0 I like it better than CoreWeave. I like adding value to the universe rather than just paying for some outsourced racks of GPUs. I like this one better, even though, to Rory’s point, the competition appears infinite and limitless, right? You better have some cojones in venture growth to do it. But, yeah, I like at least adding value to the universe. Speaker 1 The interesting thing about Modal is they were always perceived as strong by developers, but I think they were at, like, $60 million or under $100 million in the middle or late last year. That level of acceleration is just amazing. It’s stunning when you see it. That’s the first big-picture takeaway: inference is exploding here, right? They’re very developer-friendly, and they have a whole bunch of products. The other thing I was thinking about is, how big could this market be? I think you look at the four or five key inference providers and add them all up, and maybe they’re plus or minus $5 billion or $6 billion in run-rate revenue, which is a lot. But they’re all serving all the open-weight models, while right now—and it’s not like-for-like—the sum of OpenAI and Anthropic’s revenues is probably north of $100 billion, $110 billion. By dollars, it’s less than 5%. So it points to the opportunity here. Going back to Jason’s Replit guy, think about Amjad at Replit sitting there going, “Do I want to give money to Claude Code and give them a 70% gross margin for this token, or do I want to buy it on Modal and give them a 30% gross margin for this token?” Well, if the tokens are roughly the same, you know what you want to do. So, yeah, it does point to the upside there. Harry Stebbings Completely agree. It does. I’m constantly shocked by the infinitely expanding size of the inference market. Speaker 1 Absolutely. No, it’s stunning. Harry Stebbings It’s the one thing I know today that I wish I’d known 12 months ago, because I would’ve done things slightly differently 12 months ago had I been more expansive in my mindset here. Speaker 0 Well, also, you had to be smarter than me to understand how much of this revenue could be captured outside of OpenAI and Anthropic. A year ago, I wasn’t smart enough to really think through this. I didn’t think enough of it would be captured outside of them, so I didn’t think it was that interesting. But are these inference providers going to be worth $300 billion, $400 billion, $500 billion? Speaker 1 I always think it’s a race between 2 things. You can put your MBA hat on and say, going back to Jason’s comments on CoreWeave, all the neoclouds are saying, “Oh my God, let me get this straight. We do all the building work, we buy all the compute, we borrow all the capital, and then these bozos come in, buy from us, sell it to developers, and make a shit ton. Why don’t we do that?” So you put your MBA hat on, and you can imagine a lot more competition here, right? And then you say to yourself, “I doubt you get those kinds of valuations.” On the other hand, when you have this kind of growth rate, nothing else matters. It covers a multitude of sins. My bet is the margins have crept up this year, right? Harry Stebbings They have. Fireworks is in the mid-30s now, and they’ll be hitting 50 pretty soon. Speaker 0 Even mid-30s, if it stays there, the public markets can tolerate that. It’s just the multiple you’ll ultimately get on all of it, right? There are plenty of folks like that. So it’s okay wherever it ends up, right? I mean, it’s important for Accel, probably nailing it at $16 billion, but I think it works either way. Harry Stebbings Do you think the 3 hyperscalers of inference are Baseten, Fireworks, and Modal? Speaker 1 Probably, in some simplistic “who’s the winner” framing, you’re probably right. Those were the 3 biggest. But maybe the statement is this: I don’t think it’s like the hyperscalers, where no one else mattered. Harry Stebbings Jason, this is a big funding round, but it’s too easy for you in the investment committee to decide that you’re probably going to do Modal at $15.7 billion. Speaker 0 Yeah. Harry Stebbings And so if we call an investment committee, there was another round that was announced: Town at a billion dollars. I think it was a $108 million Series B. Are we going to do this from our growth fund, Jason? Speaker 0 Well, look, Harry, as you know, I regret passing on Modal and letting the Accel guys sneak in there. They’re aggressive, and they’re behind in many ways on AI investment. I mean, they did the deal over the weekend. I asked for 1 business day, and I didn’t get it on Modal. Modal was, by all intents, cheap. Yeah, I mean, $3 billion to $1 billion in 5 months. At the end of the year, we end up at $1.4 billion. Next year, $5 billion. Even if the margins never hit the 40s, we make money on this deal. There’s just no question, unless the markets crash. Town, I’m going to pass on, Harry. If you and Rory and the rest of the team want to do it, I’m fine with it. A billion today is the new $100 million, right? A billion is just the entry point for a classic B round. In the old days, when the 3 of us met, Series B investors used to say, “As long as it’s $100 million or less, it didn’t matter,” because the classic Series B wanted a billion-dollar-plus IPO. They wanted their 10x, so $100 million, $80 million—they didn’t care, right? So a billion’s the new $100 million. Town has no data. It has sort of interesting revenue. It has a smart team. This is just a bet. This is just an old classic Series A, Series B with a bow tie at a billion. If this is one of the 10 you want to make in the next 12 months, just do it. I just consider it a chip, and I don’t need to be at the rest of the meeting. There’s just not enough here for me to get excited either way. A bunch of guys from Plaid. I mean, it’s a great one, but we wouldn’t even do Plaid today. I don’t even see a $25 billion exit for Plaid, so I wouldn’t even bring it to the IC. It’s too early to predict where it’s going to go. It’s too early. It’s just a classic bet. Harry Stebbings So we’re not doing it, Jason? Speaker 0 Well, the good news is the team only predicts it needs about $500 million of inference over the next 12 months, so that’s much better than the Instinct team, which needs $5 billion. In a sense, I like it for the fund because we were a little more conservative than our peers. All we need to do is round up some of our SPVs and put in another $500 million to fund the inference. So, in a sense, there’s a positive here, right? I’m glad we passed on Instinct, even though we look like fools, because I don’t know who’s going to fund this bad boy. At least Town, again, only needs $500 million over the next 12 months. So if we’re about protecting our downside, let’s do it, guys. I just don’t know. Harry Stebbings Would you rather do Instinct at $30 billion or Town at a billion? Speaker 0 All I want to know is, maybe we can get back into Modal at $20 billion. We just blew that one over the weekend. The Accel guys are aggressive. Speaker 1 I think we should put this clip in the vaults, and 2 years from now, when you realize that a billion dollars still really is something, we can revisit this discussion. Whenever we’re throwing blithely around numbers like a billion dollars, it doesn’t sound like a lot. There are only 4,000 or 5,000 billionaires in the country. It’s real money, right? And as for $30 billion, it’s pretty damn scarce, but okay. Speaker 0 Well, you know what I was thinking as I was writing my LP report last night? It goes to the next point on Harry’s list: Anduril at $90 billion. What I was thinking about with Town and all these is, listen, Anduril’s clearly a generational company, right? They got in early on next-generation defense. Clearly deep. They have the contracts, they have everything, right? But is this thing worth 40 times revenue, with the margins and everything it has? I think the meta question is: Are a bunch of the rounds we’re doing—and maybe Modal isn’t one of them—so disconnected from public comps that we either need these massive public comps to recorrect and have Anthropic IPO? Or do some of these deals just make no sense, even though the companies are great? Does Anduril make any sense at $90 billion? Does Town at a billion make any sense? I mean, when Braze was the best exit in the whole public GTM category at 6x revenue, Monday.com’s at 2x, HubSpot’s at 2x, and Klaviyo’s at 2.2x revenue, growing 20-something percent, right? I’m not saying they’re the same. The growth is here. But I do wonder if some of the multiples are disconnected from where they’ll land. That’s just venture, right? Speaker 1 I mean, my comment would be this: I think that’s almost certainly true. The multiples are disconnected, and the only thing that kind of makes that plausible is the growth rate. Look, if you’re a mature defense contractor doing $2 billion in revenue, you’re worth $8 billion, best case. If you’re a $2 billion-revenue defense contractor and you’ve doubled for the last couple of years, and you think you’ve got a line to the administration for the next 5 years, or the Pentagon for the next 5 years, then maybe you go 2, 4, 8, 20 and you justify the valuation. So it’s not impossible, but to your point, Jason, it’s leaning a lot into the future. A lot. And all these valuations—you could wake up one day inside amazing companies just priced wrong. Of all the things likely to bail you out, just because you mentioned one other in passing, it’s highly unlikely that the US market, which is trading at an… You’re not going to get huge multiple expansion from here that saves you, I would guess. Harry Stebbings Does it not go back to the point Jason made earlier about the need to move large amounts of money? If you need to move a billion or $2 billion, I agree with you totally on the multiple compression you’re going to get from here. But is there a chance it’s a $200 billion company? Yeah, not impossible. So there’s a 2.2x, 2.5x. And you moved a billion to $3 billion. Speaker 0 I just wonder—and listen, whoever’s doing this tender offer, these are very sophisticated investors, of course. Let’s not make fun of it. But I do wonder if we’re taking revenue multiples too far today. I’ll give you a different example. I was thinking about Profound, Harvey, and Peak, right? These are interesting companies. But my God, when you look at the public comps, they’re trading at 2 times revenue, okay? And you can say, “No, no, these are agentic guys. These are agentic players, high growth.” But when the growth, even if it slows to 70% or 60%, will it trade at a Cloudflare multiple, or will it trade at a better multiple than Braze, which until today was trading at 3 times revenue? Speaker 0 I get the growth side. It’s the job we’re in. But I do wonder if the AI versions of low-multiple companies will settle too early into the multiples of their peers. I just don’t know, right? I think they’re being punished. I think the whole marketing category is the worst public category, without question. They’re not dying. HubSpot and friends are not dying. Klaviyo is not dying. They’re trading at below 3 times revenue. I would have a hard time getting the IC excited about it. Harry Stebbings Jason, speaking about multiples and how public markets think about them—where they’re appreciated and where they’re not—a fascinating one for me was Berkshire doubling down on Bending Spoons. Obviously, it bought $330 million of new shares, causing a massive bump in the stock price. The assets Bending Spoons is buying are traditionally slightly challenged assets, with slowing growth rates, not the darlings. And then the price premium being placed on Bending Spoons, the holding company, in public markets right now is enormous. I’m just intrigued, boys. How did we think about this? Speaker 1 Did not expect it. Speaker 0 I think it makes sense. I think Bending Spoons is like the Berkshire version of software, right? It’s like See’s Candy or Coke. They’re trusting the boys. They’re not flipping these; these are holding companies. Warren Buffett wanted to hold See’s forever, right? Bending Spoons is saying, “Listen, we’re gonna be diligent. We look at 1,000 deals a year, or whatever you have—what is it, Harry? 5,000 deals a year. We’re gonna do five or six, and we’re gonna find enduring software assets that we’re gonna hold for as long as we can hold them.” If you actually step back for a minute, the quality of the assets is improving. Vimeo was kind of a dog that they kept. I mean, Evernote was kind of a dog. Now, they made them work with price increases, right? They’re gonna make money off both those deals. Eventbrite was kind of a dog. Miro and Airtable, I’m gonna argue, are better. I think Berkshire’s saying, “Listen, you’re gonna make these into a bunch of See’s Candies and Coca-Colas. Folks think you can’t make them enduring, but you’re gonna make these endure for a decade or longer.” Maybe I’m wrong, but I think they ultimately probably see some similar attributes that they can make into long-lasting assets. Speaker 1 Yes. Speaker 0 It’s just like See’s Candy. Speaker 1 Get a See’s Candy from the 1960s and a See’s Candy from today, and look at them side by side. They pretty much look like the same thing. I don’t think that’ll be the case for the Evernote product of 2080. Speaker 0 No, but my point is, it may be as close as you get in software: the Bending Spoons model. They’re gonna keep these assets and invest in them. I’m not saying it’s the same as See’s Candy. Speaker 1 I describe it as conceptually the same idea as Berkshire Hathaway, but perhaps in an industry where the idea doesn’t apply. Speaker 0 Or maybe it’s like Jane Street putting all the money into whatever his, his name, Leo Apicella. Berkshire’s like, “This is close to what we’re doing. It’s not our core. We’re gonna let the Italian guys do it, and we’ll take a meaningful position. We think they have something. We think they’re gonna be able to maintain this incredible growth level because they’ll have 1,000 assets to look at in a year, and we’ll be able to maintain them in a way Vista can’t.” Harry Stebbings Guys, you can also guide. We can do General Atlantic’s IPO. We can do Tesla, the new Roadster. Unbelievable videos of cars that literally elevate themselves with thrusters. We can do Benchmark on stage conceding that competition is a bit of a flimsy term in venture these days. Any predilection as to where you wanna go? Speaker 0 You know, I wanted to do the 37signals one, but— Harry Stebbings We can do 37signals. David Heinemeier Hansson, who is an incredible race car driver, has the most extensive library of race cars. 37signals made $370 million in profit since 2020. Bootstrapping ain’t all that bad, hey? Speaker 1 I was gonna say, going back to the conversation at the start, all these software companies—none of the VCs care. If you hold onto it yourself and pull out $370 million, I think you’ll be fine. That was a great result. I knew it was good, but— Speaker 0 I think that was wrong, for what it’s worth. I think the point that was made was directionally correct. They say they do 10% profit sharing, right? So if you do $37 million of profit sharing, that’s 10%. Rory can help me with the math. I think that would suggest overall profits were $370 million. I don’t think it can make that much money. I’m guessing it does about $60-ish million a year, and even if $40 million can go to the bottom line— Speaker 1 Yeah, $40 million times six. You’re right: $240 million. Speaker 0 Yeah. It may be in the zone; it may be a little high. But when you look at the fleets of cars and all that, obviously they’ve been making this kind of money. I’ll tell you a couple of thoughts I had. One is, it’s something to reflect on in this era. I don’t think we’re gonna have a lot of one-person, billion-dollar startups, but we could have much smaller teams. The one I thought about for the AI era is, man, Zapier should’ve been this. People throw out Zapier as the one-and-done example in venture. For VCs, it’s great, right? You raise one round of whatever out of YC. I think there was a quiet secondary round from Bessemer, but really one venture round, and you get up to—they’re at $300 million in revenue today. Growth has probably slowed. It’s probably not gonna IPO as its current company, but it’s a great company that did something generational. What if they raised nothing, and the three of us just split $200 million of the $300 million? You could do worse. Speaker 1 Yeah, I think you could do a lot worse. Speaker 0 The General Atlantic guys may make more, but you could do worse, right? I almost wondered if this one-and-done model, which I used to be a fan of, is gonna be a trap in the new age. None-and-done is much better because it’s tough to do dividends right in one-and-done. You can do them right in one-and-done, but it’s tougher. It’s tougher to do dividends with the Benchmark team, right? They’re not really that interested in the dividends, are they? Harry Stebbings Can you not just buy out investors? If you’re Zapier, you just go, “Listen, guys, it’s a 5x multiple. We’ll buy you out at a $1.5 billion price. You want the liquidity?” Speaker 0 You can. It gets expensive to do it from free cash flow. Maybe Rory can come up with some examples. It does happen, but it gets expensive, even if you have a lot of cash flow. It’s not that simple, right? It can be done. Buffer did it. Expensify did it, but they all did it at pretty mediocre outcomes for the investors. I think Redpoint made nothing on Expensify, but probably more than if they’d held through the IPO, right? It’s tough. I like none-and-done better than one-and-done now. Speaker 1 Yeah. Well, the reason it’s tough is that if they own a small percentage of the company, one or two years of cash flow can take care of it. But if you’re trying to buy out an investor that owns 30% of the company, and you wanna buy them out for six times cash flow, by definition you’re tying up two whole years’ cash. So it gets risky if you’re doing it with debt. You’re right, these are hard deals to negotiate. But look, if you needed the million or $2 million to get going, you’d do the deal to start. The big picture is that there are a number of software companies that might not set anyone’s little heart aflutter, but are kicking off cash and are extraordinarily valuable to their entrepreneurs. I think that’s great. It just shows not everything has to be through the lens of venture: “I wanna make $25 billion or die trying.” It’s really good that these guys are making money here. Speaker 0 It is. In my learning, when I compare it to—I just picked Zapier because they were at $300 million and a generational company in their own way, but probably won’t IPO—I used to think this one-and-done model was a nirvana, right? The three of us get together, we wanna start a company. We’re first-time founders. We don’t have the money. We raise a couple of million from 20VC Seed or whatever, and we just keep growing our company. There’s not a lot of pressure. We don’t have to have board members. But the three of us didn’t have the resources to do it. I used to think that was a great outcome. Now, in the AI era, I think it’s a trap because you can’t dividend out the profits as easily. You can’t become a 37signals that easily. They took money from Bezos or others. I’m sure they figured it out, right? But I almost wonder if no venture rounds are the way to go rather than one in some ways. Harry Stebbings Rory, how do you feel about General Atlantic going public? Are we gonna see many more firms IPO? Speaker 1 I think you have to be a critical mass. I was looking at it there. You have to be $800 million, plus or minus $1 billion, in fee income, right? That’s the real question. Step back: there’s a bunch of publicly held, PE-backed companies, like KKR, Carlyle, Apollo, and TPG. Then the question is, can you do a bunch of these from the perspective of growth and venture? The binding constraint is the fee income. Typically, public investors give you a lot of credit for the management fee income and very little credit for the carry, which is really odd in a venture context. But, be that as it may, to get to a market cap that matters, they trade at four or five times revenues at best. Speaker 1 So you're probably looking at having to have $800 million in fee income. At 2%, what's that? 50 times 8. Yeah, you get a couple hundred million dollars, well more in actual AUM. It's doable. Andreessen could probably get there easily from here, but that's kind of what you're looking at. Maybe you'll be able to do it with a lower amount of AUM because venture gets paid slightly more and has carry that's more meaningful. The very largest companies will do it, but not many is the answer. That was purely a financial-metrics answer. Let me make a totally separate comment: it's really hard to do that and not change the culture dramatically and become a more risk-averse institution, right? It may well be that someone like Andreessen could pull it off, but it makes it a very different company. A second, separate, and probably more important criterion is: is that what you want to do as a venture firm? I think it's great for the founder who gets to take a ton of money off the table, but then you become a corporate venture firm, and that's just a different thing. Harry Stebbings I don't know if there's a difference. I'm not being a dick, but I don't know if there's a difference between the scale GA is at now and what it would be like if GA became a public company. I know there'd be a bit more compliance-wise, but they're such behemoths anyway. The joke is, you get in the lift and you're like, “Oh, where do you work?” And the answer is, “Here.” Huh. And you both go, “Oh, wow.” It's the same for both. So interesting. Speaker 0 Is there— you know, I should know this, Rory, from the public comps. When you do an IPO, is there more fee pressure, though? I assume when you're not public, you can play a few more games, right? The LPs all know, but when you're public, everyone's scrutinizing all of this, right? Every single dollar, where it goes. Speaker 1 I don't think it's scrutiny. I think the comment is that you always have to grow assets. You're always launching products. You just— Speaker 0 Well, you always have to grow. You al— Speaker 1 Yeah. You always have to grow. Speaker 0 You can't just pay yourself $80 million or $100 million as the folks running it and kind of check out, right? You've got to grow. Speaker 1 Yeah. I'll even state it. Put it this way, and it's a silly example because they're not big enough, but if old Benchmark had gone public at $450 million every 3 years, Wall Street would've had a shit fit. They'd be like, “Come on, dude.” And now finally they're giving them what they want: “Oh, good, you got the growth fund. Maybe you should do a crossover fund next.” That's the way it would roll. Harry Stebbings Boys, what about Harvard's endowment? It returned 27.1%, bringing it to $68 billion, helped by a SpaceX stake worth about $2.2 billion. They're an awesome team. Their CIO's amazing. Love them dearly. Worked with them. They're great. Anything here? Speaker 1 Yeah, I mean, if you invest in venture, when you get the big outcomes, it's good. They've had a few lean years, but I think you're going to see that. I think one of the other universities that went directly into SpaceX made out even more like a bandit. I think it's Washington University. I wouldn't swear which university, but someone did really well. So yeah, I think it's going to be a thing. It's interesting when it's concentrated in only a few assets. The good news, if you're an LP—not a manager, but if you're an LP—you probably have some exposure to that. You're going to see a lot of this goodness. Harry Stebbings If you're an LP and you do not have exposure to Anthropic, OpenAI, or SpaceX, you're feeling a lot of FOMO right now. Speaker 1 Of course you are. Yeah. Harry Stebbings I think the venture market's about to get a lot hotter, by the way, because a lot of people are starting to get cash back now from SpaceX. They're seeing Anthropic on the horizon, and they're hoping for OpenAI on the horizon in Q1 or Q2. I think we'll start to see that cash recycle back into the ecosystem that provided it, and LPs will come in with more money into venture. Speaker 0 In 2021, the average college and university did 36.6% from venture and PE; 41% was top quartile, and Cambridge had 62%. Harvard's great, but I'm waiting for a banger 27. I want to get back to the nutty IRRs we saw in the last boom, right? The outcomes here are an order of magnitude larger, maybe even more than an order of magnitude larger. But I want to see 60% IRRs. Good job, Harvard guys, but this is no 2021 yet, right? Yet. Harry Stebbings I feel that $68 billion is pretty impressive, to be fair. Yeah. Speaker 1 Yeah. Harry Stebbings Scale matters. Speaker 1 The $68 billion—you do remember they're the oldest university in America, so they've been piling the money up for quite some time. Speaker 0 Wesleyan had 95.3% returns in 2022 from the 2021 vintage in venture and PE. 95.3% returns, Wesleyan. Harry Stebbings Wow. Any others that I've missed, boys? Speaker 0 I think the Tesla one is the last thing we're going to see before the merger with SpaceX. I think it's the swan song, if you want to know my view. I think Elon kind of said it. He said, “This is the last...” He said, “China's going to take over the cars. This is the last car a human will ever drive.” Both of which are clearly true, right? China is going to dominate global EV and autonomy. No one can compete with their cost proposition in electric or otherwise. No one will drive their own cars in a few years. Literally no one. I think this is—I don't know why he did all of it. Some of it's clearly for fun, but I think it's the last standalone big hurrah before it becomes one big company over the next X months, which was always Elon's goal. Having these companies split up makes no sense for him. They'll all be merged, and this is the last Tesla party as an independent company. Harry Stebbings Is it not the most masterful marketing from Elon? The brilliance of Elon is the Elon premium and his ability to have extremely high valuations because of his presence. Doing these stunts justifies it like no one else can, and gives it life for another 12 months without any recompense. I think it was brilliant. Speaker 0 Well, that and giving a SpaceX share to every American is pretty brilliant too. I thought of it just from a sales perspective. There are so many Elon haters. I think he's the greatest founder of our lifetimes. Politics aside, the greatest founder of our lifetimes. I really don't see how you can argue with it. There's a lot of haters out there. Not just on the left; there's a lot of haters on the right, all across the spectrum. It's pretty hard to hate him when you grow up with SpaceX shares in your account and you watch rockets blasting every day to the Moon and to Mars. It's pretty hard to hate the man when you're a shareholder. It's that much harder to hate someone when you're a shareholder. It's really that much harder, man. It's that much harder. Harry Stebbings I don't know. I hate Grok today. Speaker 0 I know, but— Speaker 1 But you didn't get it for free. The thing is, Elon's going to park one share in every kid's account. These so-called Trump Accounts for children are a very interesting way for super-rich folks to make significant donations to everyone, if they want. The idea here is literally every kid gets one share in their account, which in aggregate—I saw the number—is something like $10 billion. It's real money. It's a super interesting way. You could argue, $200, some compounding—what does that mean? But it's better than a poke in the eye, and it didn't cost you anything. Because it's associated with the share, as Jason said, there is an interesting element of marketing to it. So it's good. Speaker 0 Yeah. Speaker 1 Cool. Speaker 0 It's hard to hate the man who just gave my kid a SpaceX share. You can hate him, but it's hard to hate him as much. It's really hard to hate him as much, right? Michael Dell is going to get some kudos too, but this is the big lift. I think this will just gain momentum, too, right? This is the next version of the 1:1:1 thing, which was nice, but... It's going to remove a lot of the toxicity around the massive wealth creation that's only just begun on a scale we've never seen before, and give it directly back to the people rather than through some weird Gates-whatever foundation. This is a much better way to deflect drama: give it directly to people. The ultra-rich should do this. It's brilliant. Harry Stebbings Boys, the brilliance of Elon is a great one to end on in terms of giving. Speaker 1 Yes, always end on a win. Harry Stebbings But before we leave you today, Base44 is one of those products I used, and I just thought straight away, "Man, I wish I'd had this when I started 20VC with $50 and no contacts in venture." I wanted to see what it could do, so I built a guest research hub for the show. Upcoming guests, company notes, previous interviews, sponsor contacts, follow-ups, all in one amazing location. I described it in plain English. A few minutes later, I had the product open in front of me. Database, login, back-end, and hosting already handled, and just all in one. It made it so easy. And that's what I loved. I didn't write a long spec. I didn't stitch five different tools together. I built the thing, and I started using it. For any founder listening, that's the point. The faster you turn an idea into something real, the faster you launch. Apps, dashboards, internal systems, AI agents, whatever you need to build, Base44 gets you there. Go to Base44.com/20VC. That's Base44.com/20VC. While Base44 helps you build, AlphaSense helps you find answers really, really fast. We used AlphaSense on an investment that helped us close an $8 million deal. $8 million, baby. That's a lot of money. That's why I'm genuinely excited to have them as a partner on 20VC. AlphaSense combines AI with one of the world's deepest libraries of market intelligence, including expert interviews, broker research, earnings calls, company filings, and real-time news. Every answer is grounded in this incredibly trusted evidence and fully traceable to the original source, which is so important. So you can make really high-conviction decisions with confidence. But the best part, they're building super analyst, an always-on AI analyst. So instead of starting your day with another search, you'll start with work you already done, your coverage monitored, the important developments surfaced, and your investment brief already waiting for you. See for yourself. Head to AlphaSense.com/20VC. That's Alpha-Sense.com/20VC. While AlphaSense finds insights, Intercom helps you serve customers. As AI agents become more common in customer experience, teams often end up juggling multiple siloed tools for every job. Well, Fin was built to change that. It's a single unified agent that works across your entire customer experience, from service to sales to success and beyond. Fin is the agent making perfect customer experiences possible for thousands of customers. It's powered by custom models, trained on years of real customer interactions, so it understands the nuance and complexity of customer service better than any other agent. That means faster resolutions, more consistent support, and just better experiences for every customer. It's also designed to be fully self-manageable, so you can easily improve and adapt it as your business evolves. No third parties required. Leading companies like Gamma, Asana, DoorDash, and Crypto.com already use and love Fin to deliver better customer experiences. So for a limited time, you can get $500 a month in Fin credits for your first three months. Learn more at Fin.ai/20VC.