Coatue's Spencer Peterson on Cursor hitting $1B ARR and what makes a 'transcendent' AI company

again an excellent product uh all these things happening at once and then you sit here and you wake up today and say wow cursor's got an incredible business incredible product velocity an incredible team and now their own model which they shared yesterday is their number two most used model on the platform in composer.

So um you know there's a lot of good companies uh in San Francisco. It's kind of a nice thing like in early 22 everyone was very worried that all of our 21 investments and everything was just going to you know kind of wipe out. There's actually a lot of good companies in San Francisco. It's a great thing.

There's [laughter] there's a decent narrative violation. Right. Right. There's a lot of good companies. Um, there's a few there's a few great companies also, but there are very few of these companies that, you know, the the the word of the day or the word I try to always come back to is transcendent.

And um, you kind of know it when you see it or you know it when you're around it. And that's what we think uh, Michael and the team are doing at Curser. Uh, funny story. Uh I Curser is uh I'm incredibly inspired by by their growth and and user love and and everything they've accomplished over the last few years.

But I had a painful moment when they when they uh when they first kind of exploded onto the scene because I realized that Aman, one of the co-founders, had been following me since like 2020 or something.

I [laughter] was like I I I I I had thisve I've had this like you know lingering fear over the last few years of like you know getting followed by somebody not reaching out not investing and uh certainly certainly would have been smart a few years ago to um to Well listen it's not too late there's still we think there's still a lot of upside you know so we'll we'll see fingers crossed one of the frameworks that I've heard kind of bandied about around what's going on like a lot of people maybe want this to be more of a winner take all market than it is.

Um and yet when and uh and yet maybe because of Jevans paradox or just the nature of like we it's very hard to understand how much code does humanity actually want to write like it's potentially a ton.

And so you're in this like entirely new market that's a blue ocean and it's massive and it's just growing growing growing. And so there's opportunity and I think you're in a unique position where uh you you you're you're in a number of names that do overlap somewhat and there's a lot of opportunity.

I'm wondering if you think that's the right framework or or or is there more of a like a winner take all monopoly thesis with wi with this particular market or if there's other uh historical anecdotes that you go to to kind of understand how this might play out.

Yeah, I'm I'm not sure what the best I mean I'll get I'll get to the anecdotes. It's hard to find the best analog and analoges are always difficult especially for this. We're we're investors in OpenAI, Anthropic and Cursor. Yeah.

Um we're also investors in Glean, Harvey, Open Evidence um a number of companies across the kind of AI stack.

um our view right if you just take a step back right people talk about the the bottoms up tam sizing on developers to your point people want to write a lot of code 30 million developers in the world right a fraction of a percent of the global population today and so what what's that number look like in 10 years 20 years what should that number be I mean that's not really an answerable question but probably much higher and then if you take a step back tops down you've got 5 trillion of global IT spend and about a third of that is labor today.

So hey, Open AI is putting up historic growth numbers, Anthropic putting up historic growth numbers, right? All that's well kind of directionally known. Um we see all these companies growing really really healthy. Um and doing it in a in a way that we think is pretty sustainable. Yeah.

Did you have a reaction to uh Satcha Nadella on Dar Cash X Dylan Patel? Um there there's this interview and he kind of talks about uh it feels like he was kind of sending the signal uh I'm going to steamroll you if you try and build an XL agent.

I'm I'm defending the castle that is Excel but also on a platform for a bunch of other things. Excel is Excel co-pilot. Yeah, exactly. And I think that I I think that I I would bet on him being able to defend the castle that is Excel.

But I've but I was going back and forth with with uh with Jordy about uh it feels a lot harder for Microsoft to put up a fight in AI for lawyers because they just don't have a practice there.

And yes, lawyers use word documents, but uh it is a it it just feels like a slightly different uh go to market, slightly different product.

And so I'm wondering on how you think about the edges of like where the where the hyperscalers or the neo hyperscalers the foundation labs might steamroll versus not uh because cursor is an interesting example of a company that feels like almost steamrollable and yet hasn't been right hasn't been right again if you look back at the past if I told you hey claw code is going to be this amazing product that's this good and you can use it you'd think oh my god right cursor good like I'm short in cursor at the end of the That's what a lot of people.

So, again, I just think it's it's, you know, we get to sit in a unique spot where we see all these things happening and go, "Wow, we're just so early in terms of the pie growing, the tide rising with all boats.

" Um, I think I messed up that somehow, but the uh uh to get back to your question around the the hyperscalers and who wins where, um, it's it's it's hard to say exactly. I think that's kind of the big question right now.

If you look back at like the beginning of the year, I think there was an open question around are are these API businesses going to work? Is there long-term margin in the API business, right? Is something like cursor just going to get run over?

Is something even like anthropic in the API layer going to get run over and squeezed? I think now people are kind of going, well, you know, these businesses look really good. They've scaled really nicely despite all these things, to your point, happening at once.

Maybe now the question is what's the the architecture of that market look like with the hyperscalers over the next 10 years. Um it's it's hard to know.

Um what what I would say is if you take a step back and you look at the cloud AI revenues for the uh for the three hyperscalers I mean they've been inordinate beneficiaries of this trend right open anthropic cursor uh like Harvey all these companies are great.

the hyperscalers have done really really well and the move to cloud and the move from onrem to cloud that's being pulled forward by all this is really really compelling for them. So, like they don't need to worry about winning in our view, right? Like lawyers in beating Harvey.

Like that's that's almost um that's just further down the economic ladder in my mind for them. But, you know, we'll see. Um the uh and then the other point, oh, you asked about Excel. Yeah, Excel is fascinating, right? Uh the ultimate like dominant software product over many decades.

We've been looking for kind of the cursor for Excel uh for a while. We've tested a few products. We've got an internal team here that builds a bunch of things that we also You guys have built your fair share of models over the years, I would hope.

Yeah, we we we do a lot with with data science and and now increasingly with AI. We haven't quite found something that works yet, but there are a few folks building some interesting companies that we're not invested in them yet, so I'm not going to say what they are, but it's an interesting opportunity.

could um I I think that more likely than not Microsoft still ends up owning a lot of that. I mean the closed source thing is very important like there is no VS code in that category.

Uh people talk about this with uh cursor for the for biotech like what's the lab notebook and I was pushing on like I was trying to go I I was trying to go deep with an investor who uh kind of came at me with a thesis of of cursor for bio and I was saying okay well if if you're going to invest in the cursor of bio what's the VS code of bio what's the open source what's the open source standard that you can fork that every bio lab uses and they're like oh well actually like the labs use closed source stuff and those enterprises do not just want to give up the data.

So, you're going to be fighting for the walled garden and all this other stuff. So, uh it does feel like the like the cursor for X model. I mean, it was really hot at YC what two batches ago? Everyone was doing cursor for X. Uh and it feels like uh it was wishful thinking in some markets.

Yeah, I I think that's exactly right. Um, and what's interesting is for biotech in in particular, right, and we've heard Sam make more public comments recently, right, about their effort with the sciences and how much time they're spending there. Yeah.

Um, I just think that's where the value ends up acrewing is something like an open. Anthropic probably throws their hat in the ring at some point. There's uh Isomorphic Labs coming out of Google. There's the Chai Discovery business which is a compelling company.

There's a number of folks who are kind of building these pseudo lab companies that work with with biotech.

I think that's probably what ends up being more interesting because to your point there's not a there's not like a VS code for biotech today and there are just very few markets where there even is a VS code analog right part of the magic of cursor is they identified an incredible market where the models worked really well early which was coding.

Right? This has been the frontier of model capability for a while. You had the ubiquity of VS Code and of the terminal and all these things and then you just built a really nice product around it and have continued to just push the frontier of what's possible and now you layer in their own model with composer.

Uh so yeah, we're excited about it, but it's hard to find markets with the same anatomy and structure and kind of opportunity. to your point, people are kind of squinting and like kind of forcing ideas a little bit. Totally. Totally. Totally. Yeah.

Jordy, uh I'm sure uh I'm sure you guys just try to back uh visionary founders that don't come to you uh for help unless they unless they really need it hopefully. But how how do you think uh private companies how much attention should private companies in AI be be paying to the public markets right now?

because we're seeing so much growth, revenue growth in the private markets and at the same time uh you know we were just you know coreweave's down uh 45% in the past month.

Uh there's a lot of jitters in in the market right now and it's hard not to pay attention uh to some of these signals even if even if a company's own revenue growth is is skyrocketing. So I'm curious how how you think the best companies are going to kind of manage through the next 12 months.

Yeah, I mean look, the fortunate thing for the the best companies is almost all of them have a war chest in terms of capital on the balance sheet today.

And I think the capital demands of these businesses is a bit exaggerated relative to what what what it is would be maybe my personal view um relative to like what's on X or what's in the media around these AI companies. Yeah. um [snorts] in terms of like paying attention to the public markets, right?

I think it's I saw um like Subu tweeted I think a few days ago something around right like the the market's been above its 50-day moving average for like over a 100 straight days like something like 130 straight days or something like this.

So, you know that the market's been good for a while uh when you see a stat like that. Um, I think founders are aware of that and are pretty aware that it's a good time to raise capital and fortify a war chest, right, which we've had a really busy Q2 and Q3 and now early Q4.

Um, I think that's kind of the way to think about public markets is, hey, when's a good time to raise capital when my when my cost of equity is lower? Yeah. Uh, but outside of that, like I just think tunnel vision is everything.

Um, and again just to to bring it back to Kerser, like this is a very focused company, very heads down company, doesn't make a lot of appearances, right? Like just stays in their lane and focuses. And I think that that's kind of the ultimate uh value, right? You don't want to be spread too thin.

And look, there's a lot of folks like myself running around uh San Francisco now looking to give founders money and tell them when the when the value of their equity is is pretty good, right? So we we help give them that signal. Yeah.

There's also this uh interesting I it's hard to measure corrections in the venture ecosystem because they're much less quantitative than just looking at oh okay you could you know map the yeah correction often looks like wow that company hasn't raised in three years. Yeah. Yeah.

that or you know there's some hiccup in in LP fundraising or something but uh there there's me there there's many corrections in the public markets that I can think of that were like hiccups in the private markets but for the most part it was everyone just kind of being like oh there's some crazy stuff going on okay like uh let's cancel a couple meetings and okay we're back on two two months later like the coid9 like selloff was like a massive correction in the public markets and for most for most startups it was like Yeah, you needed to understand what could your company continue or were you in like hospitality or something that was going to be heavily affected.

But for a lot of companies, it was just like, oh, your your your March raise turned into a June raise over Zoom and it was fine. Yeah. I mean, I I I think that's spot on, right?

Like there is a [snorts] level where um right because we do have a public hedge fund and so what's interesting is when you have conversations like the one you guys were having prior to this, right? Um, you know, as a hedge fund, you're always behaving in the market, right? There's always a choice to be making, right?

It's very unusual to be in 100% cash or something like this. Yeah. Right. Is is as a private fund, as a growth fund, we can just choose to not invest for a while, right? We can just wait a bit.

Y um and for us, like the the convenient thing is uh you know, we only make I think this year we've maybe added about half a dozen new logos to the portfolio. So for us, we're only making three to five core big investments a year on the growth side right now is kind of our cadence.

So it helps us kind of work through whatever that cycle is and just focus on the asset, focus on the entrepreneur, focus on what we view as the 10year trend and story for why this company is going to be so durable and so powerful over time. Yeah. Um but yeah, like you don't see it, right?

I think the other thing that's maybe like less discussed is we've had a bunch of companies and I've seen companies outside our portfolio too that were kind of your 2017 era SAS businesses that have really seen a nice bounce back and a nice acceleration over the past few years with AI really like reinvention is the wrong term.

It's just kind of like some product extension in the CEOs and the management teams and the the teams are just re-energized and excited because there's this new capability that you can integrate across your entire platform. Totally.

And if you if you were some if you were some, you know, flavor of a source of record, your ability to now use that data um and create economic value around it is just inordinately better. So um yeah, I mean look, the the market's been up for a while. It's definitely interesting. Yeah.

Uh but for us on the private side, it's a little bit easier because we just get to focus on the companies. Yeah. How do you think people are do you think people are reading too much into Sarah Frier's like private phone calls these days or uh [laughter] I mean it it feels like uh you know Open has like a billion users.

Uh deceleration might be sort of like expected like you know people don't accelerate on on DAUs. Um, do you have any thoughts on like what's going on there? I've got a big term cuz I I knew this would come up and so I thought of this before. This was my one big term here. Okay. Anxiety displacement, guys.

We've got anxiety displacement. That's what's going on. You know, like people talk about anger displacement, right? All these things. Yeah. Look, market's been up for a while. Like that's a that's a that's like a fact. Y um Oh, ChachiBT is less than three years old. Sure. Is a product, right? Yeah.

It's grown really quickly, right? It's it's they've disclosed they're almost 10% of the world is using the product weekly. Yeah. Right. Like that's pretty remarkable, right? The market's been up a while. We had like I still think as a society people haven't really processed the whole lockdown thing.

We had the government shut down for you know a month and a half. Like there's all this stuff going on and there's all this anxiety and I do I think like that's been very displaced because people you know like open a great company. It's a remarkable business.

Um, I there's just it's hard to think of an analog for what they've achieved in less than three years is a is a product and uh I think Sarah and Sam and the entire team do a incredible job. So people are definitely reading too much into uh anything and everything.

Sometimes I think there's a little bit of a malicious spin on some of the stuff they put out also just to be very candid about that.

But uh yeah, I mean it's understandable for there to be a lot of anxiety in society in the market and it's understandable that it gets displaced onto the company a little bit, but uh I don't think that's fair. Yeah, it was a little weird.

I mean, we were reacting to this idea that there's a private call with investors like what type of investor is then leaking like potentially bad news to the press. Uh it seemed like an odd chain of events.

Um, but there's certainly unlimited demand for uh bearish takes about Chat GBT right now because uh they've they've been on such a tear. Every day people log on to CHBT and ask for what's the most bearish thing about Chat GBT. Yep. Yeah. Yeah. Right. Right. That's like that's the that's the tweet prompt.

Like that's your ex maximalizing like prompt right there. Probably like that, right? What what chart can I put together or what can I do? But look, I like we'll see, right? um worked with the company for a long time. Think it's an amazing company. I've been really impressed by what they've done.

Um I continue to use the product every day a ton. I think 5. 1 the new model is really a nice upgrade, too. I enjoy it. So, you know, um yeah, I think it's I think they do a great job and I think over over rotate on it, but we're we're excited for 5. 67, 67, which would be only a few upgrade cycles away.

Rolling my eyes that one. Can't do it. Can't do the stupid 67 memes. It's over. Killed it just now. They're over them. I like that. I I like the new coinage, though. Uh the anxiety displacement. It's good. Let's talk about other growth funds because that's just a fun uh fun topic for for every uh every investor.

Talk about everyone else. Do you think anybody has um have you seen any other any other funds uh kind of uh blink yet or or or hesitate or or kind of slow down at all?

a little bit worried or because it's just like we're seeing every every day we get like half a billion dollars of funding announced on on the show, but that has that's a lag, you know, it has like a these rounds like really got done uh in in late summer, early fall, and they're just kind of getting getting out there now.

Yeah. I mean, I'm sure there are some funds that have slowed down. I'm sure there are funds that are deploying more than us and less than us and those things, right? Um, I'd say generally the the I think the vibe when you're catching up with someone over coffee or lunch is a lot more positive than over Twitter, right?

And there's a lot more of like uh well, how exposed am I to the AI trend, right? Like there are a few growth funds that feel like, you know, have disclosed, right, hey, I'm a little behind this and I'm trying to catch up here and there.

So, if anything, like I've felt more anxiety about that than, oh my gosh, you know, we just put more money and pick your big premier company and, you know, I haven't slept well about it. So, that's not necessarily a a great signal either, right?

That's kind of like an an average signal, but um I'm sure there are some folks slowing down. I'm not sure who they are per se because we have a lot of funds that are definitely deploying a lot more than us in terms of absolute dollars and just absolute number of deals.

So um for us we just again we really just try to focus on our our core uh you know mission which is hey what are going to be companies that matter as public enterprises you know 10 15 20 years from now and in some cases a little bit sooner right um and yeah what do you think about the capital intensive pre-revenue AI companies we had uh fee Lee from uh world labs on I don't know if they're fully pre-revenue, but it feels like uh a lot of the world model projects, the generative 3D worlds, Gaussian splats, what Google DM's working on with Gemma uh or is it Genie?

Genie is the model. Uh it feels like something that I just think will be like the next roadblocks vaguely, but there's nothing that you could do to underwrite this against revenue growth. And so and yet and yet a lot of these projects are like to do the next thing we need $200 million.

How would you as a growth investor even square that or or would you just say hey let's just come back when we can actually see some adoption data. Yeah. Um normally we would fall into the second bucket like generally in terms of how we operate. It's different for everyone.

I think all those rounds are ultimately just a representation of the idea size. Totally right. And of the opportunity. And then and then of course also of the entrepreneur and the team. Yeah.

Uh we did make you know one investment very early that that maybe fits this mandate which would be skilled in the robotics space which has been a great company has executed incredibly well and we Luke Metro is over there now. Oh yeah that's right. Yeah. Yeah.

We just know someone who's on the team uh but not not one of the founders that left left and to uh to join. It's skilled right. Skip. So that was that was one where we leaned forward a little bit earlier than we would with with most things.

But and again that was on the back of just being very bullish on robotics on a 1015 year time scale and that team and where they sat in the stack. But generally we tend to wait a little bit to see a company be a bit more flushed out but it's always case specific. That's amazing.

Well thank you so much for coming on the show. Thanks for hanging out. Well we will talk to you soon. Hope you have a great rest of the day and congratulations on the cursor round. Uh we'd love to see you Spencer. Have a good one. Cheers. Goodbye. Uh up next we have Tyler and Cameron Winklevoss.

Um but first let me tell you about Turbo Puffer search every bite serverless vector and full text search built from first principles and object storage fast 10x cheaper and extremely scalable. Um uh the Winklevoss twins are in the reream waiting room I believe. Um if not we have a uh slight delay.

Let's follow up on that. Go back to the timeline. Back to the [music] timeline. Dario Amade predicts that we will get to 90% on SWEBench verified in a year. That was one year ago. It's been one year. Uh the best performing model, Sonnet 45 with parallel compute gets 82% on SWEBench verified.

Close to 90% but not quite there. They put him in the truth zone. RIP to Dario. What do you think, Tyler? Bearish. I mean, I think he was pretty close. It was it's pretty close. Good enough these days. That's That was way more of a bullish take than most people. Yeah. No, it is. It is very impressive. Uh totally.

Uh do you think it's saturated? People are saying uh that Yeah. Yeah. I think a lot of those kinds of benchmarks are generally saturated now. Yeah. Yeah. Yeah. Yeah.

Yeah, I mean it is interesting that like simultaneously you hit the 90% on Sweetbench, did very well there, but on the flip side uh you have Audrey Karpathy saying like it's slop and like you can't actually use it for like the frontier software development that he wants to do.

Uh anyway, I believe we have the Winkle Voss twins in the reream video. [music] Let's bring them into the TVP Ultradome. Tyler and Cameron Winkle Voss, welcome to the show. How you guys doing? What's going on, guys? Hey, very nice background. Uh give us the update. Give us the news. I know we're running late.

Uh, so let's just jump right into it. I assume everyone knows who you are. Cool. Um, so we launched Cippy Punk, a Zcash, uh, DAT yesterday. Uh, it trades under Scythe, Cyph PH. Um, and we're really excited about it. The the mission of the company is privacy and self- sovereignty.

Starting by accumulating Zcash and in time we hope to invest in other technologies that promote privacy and self- sovereignty. Uh this delay having some Wi-Fi issues. Uh sorry uh say more about the structure. This is leap therapeutics. you guys have have rebranded it.

This was an existing public company, but what more can you say on on how this came together? Yeah, so it was an existing biotech company um and we basically took it over um and invested via PIP into it and then um rebranded and changed the ticker today. Awesome. Uh so I guess uh Zcash has had a lot of momentum recently.

At the same time, uh, a lot of the DATs have have, uh, you know, struggled, uh, in in, uh, more recently. Uh, what like talk through kind of like the next, uh, how how you're thinking about, you know, making sure, uh, just kind of navigating this time when the markets are are pretty choppy overall. Sure.

So I think number one the longterm theis we really believe in and I think that when you look at Bitcoin as a store of value or how you store your value Zcash is really how you move your value is is very sound and we're obviously very bullish on that. But in addition, we're the largest investor into the DAT.

Um so we don't have like a lot of fast capital that's looking for a trade. we're just long-term hodlers or in the case of Zcash uh zodlers and we just you know we plan to to hold for for a very long time and I think that's one of the