Google Loses $500B in 12 Hours | DoorDash Buys Instacart for $17B | Kalshi Raises $1B

Jason Lemkin 0:00 Agents are an existential risk to DoorDash because their number three competitor is nobody. All these app platforms have a risk of slight disruption from agents.

Harry Stebbings 0:07 Today we have OpenAI's big fundraise, $1.25 trillion price down from $1.4T. Then we have the AI trade hitting a wall. Alphabet, whoa, $500 billion off the market cap overnight. Intel falling 29%. Then we have DoorDash buying Instacart for $17 billion. This and so much more in the show today.

Rory O’Driscoll 0:27 Kalshi is a money deal. If you want revenue velocity right now, nothing gives you revenue velocity like inference. Physical intelligence, company with relatively little revenue is worth $20 billion. Dropbox, company with billions of revenue ain't worth a lot. It just shows revenue multiples don't get you anywhere, right? It's all about the future.

Harry Stebbings 0:43 Ready to go? [upbeat music]

Harry Stebbings 0:55 Boys, we are back. Luckily, there has been a huge amount happen in the last seven days. So let's start with OpenAI. Its valuation got cut from $1.4 trillion to $1.25 trillion, and the round reduced from $30 billion to $20 billion, with only about $11 billion committed. As a result, everyone is going, "Oh no, this is further examples of the AI wave cracking." How do we think about that?

Rory O’Driscoll 1:23 I would dispense with the whole AI wave cracking narrative. I think people are trying to extrapolate their, you know, desires, wishes, whatever prejudice and biases on the story. It's as simple as the other company is g- bigger, growing faster, profitable, and about to go public, so you can buy it liquid at $1.5 to $2 trillion. What would you pay for the smaller company growing less quickly, still losing money, that's illiquid? Answer, less than $1.5 trillion. How much less? Well, that's what they're figuring out, so it's, it's, it's utterly not surprising. It's, it is as simple as that. No projection about AI, positively or negatively, just there's two securities are gonna be sold in the next two months, and which one

Rory O’Driscoll 2:02 do you wanna buy?

Harry Stebbings 2:03 But Jason told me last week that it was the most underappreciated.

Jason Lemkin 2:06 I think it is. It just may, maybe it needs to drop a little more.

Harry Stebbings 2:09 [chuckles]

Jason Lemkin 2:11 Well, don't forget, listen, there's gonna be bumps, right? Um, and, you know, broadly what we call AI stocks fell $5 trillion last week, right? Now, maybe this AI round was the week or two before that. There's, there's obviously a temporal gap, but when we're seeing this level of volatility... And listen, I'm surprised that folks didn't talk more about it. I mean, Anthropic had a down quarter from net new revenue. It was down in, in this last quarter in terms of net new bookings over Q2. Um, there should be some, some nervousness. Um, maybe it would fill up at a trillion. Maybe I'm glad we didn't put in our bid until the end, Harry. Maybe, maybe this, the second tranche will be lower than the first. [chuckles] It makes sense there should be a little bit of wobbling at the moment. The stock market's down. We could argue, maybe Rory will challenge me, we can argue that Anthropic's last quarter wasn't, wasn't, uh, quite as crazy as some folks might thought. There, there should be wobbliness, um, and maybe it doesn't matter. It matters where it'll be in 12 to 18 months. Um, but it, it makes sense there's wobbliness. There's wobbliness in the debt markets. Cohere's debt is p- paying 15%. That's wobbliness. There's, there's w- there's just Broadcom having, I know Rory didn't, kinda, kinda took me to task on this last week, but Broadcom having to guarantee Anthropic's debt is a little wobbly. It's just w- it's just w- wobbly, and, and IPOs are always thin and wobbly. Like they, no matter how many times we say they, they get oversubscribed 20X, so many of them are just a little wobbly going into it. [chuckles]

Rory O’Driscoll 3:28 I think all those things are true. I think it is. There's a lot of... The way I think of it, there's a lot of kinda pushing, shoving. There's a lot of price discovery, risk discovery going on on the supply side of the AI trade. You know, what price do you need to be paid to lend money to Cohere? What price do you need to pay to buy OpenAI in the private markets when you get in the public markets, right? All that, I, I agree with you, all that stuff is going on, and I think, you know, when you go from assumed euphoria regardless to tempered euphoria, uh, there's this kind of price discovery process on the way, and I think... So I think that's fair, Jason. Those things are all true. As I say, my point is merely if the other side of the equation, the demand side, holds up, all these things will be, you know, forgotten in six months. And if there's a problem on the demand side, on the growth side, then all these things will be, you know, straws in the wind of a very significant change. That's my point. It is. So when you zoom out to what matters, if the demand for AI is anything like some of the bulls would say, then let's, let's talk about the Cohere bonds, 'cause everyone's like, "Oh my God, the bonds are gonna go..." If the demand for AI is as strong as people think, then everyone's gonna need every damn computer, GPU, and data center they want, and those bonds will be money good because someone'll buy the damn company. Right? There's just no risk of that, right? If the demand is not there, then the first person who's gonna suffer when the demand is not there is not gonna be Google or Amazon, who have a big-ass balance sheet. It's gonna be the guys like Cohere, and that's where the problem will come, and that's why there's some risk in it. So that's all that's going on.

Harry Stebbings 4:58 You mentioned price discovery. It was a, it was a tough week for Alphabet or Google, um, falling about 12%, which is, you know, is the most significant drop in a decade. Um, Google Cloud doubled to an insane number. It doubled to, like, $130 billion, growing 103%. But search did slow to 12%. Free cash flow was negative, unusual when looking at Google. Is this just another element of price discovery?

Rory O’Driscoll 5:27 Again, it's, on the search side, it's been true for two or three years that there's probably gonna be some element of cannibalization of search, and, you know, it still grew, um, pretty nicely. I think it grew 12% or so, but there was some unsearch. But the, I, I think it was a combination of search angst and general angst, right? And Cap, and CapEx angst. So nothing more than that. But I mean, it's so funny to watch the last two or three years as a... You know, we've gone from Google is dumb as rocks and Satya is a genius to then you fast-forward, oh my God, Google is the most amazing company ever. Poor dumb Satya, he missed the boat. And now this is a switchback to, oh, maybe, you know, maybe the things that we're worried about are, about Google are still true, which is, you know, search cannibalization and... Exactly. But, you know, to be fair, to your point, the growth is there. These guys are saying they can They can sell, compute as fast as they can make it, and that's the number one thing, and that's where the growth's coming from.

Jason Lemkin 6:15 I just think we're in a... Listen, I, I think w- we, we are, uh, um, understandably in a, in just in a nervous period. Yeah, I mean, I... It's a... I, I think it's at 120 billion dollar run rate growing to 100%, right? That's about as good as it gets for the, the folks in the cloud. The cloud group should feel pretty proud of themselves. Um, I just think, um, a- a- a- when any metric comes up short, I mean, Shopify just fell 15% as we're discussing this, right? Because they came up, uh, a smidge short even though they raised guidance for the next quarter. Just everyone's just nervous AF, right? J- they're just nervous AF, and I, I think it, I think it's appropriate. I think that the Anthropic IPO's gonna suffer for it. Um, I think there's evidence of substantial pricing pressure. There's evidence that, you know, as dominant as they are, it might be not quite as dominant as there is, and I just think it's, it's, it's, it's okay. There's gonna be these squiggles going here as most things still hopefully go up and to the right until we can, uh, distribute out

Jason Lemkin 7:08 to our LPs.

Harry Stebbings 7:08 What do you think Anthropic's gonna suffer for? You mean the, the consumer excitement won't be there? You think that they won't have the pop that they could have? Like, what, what does that mean?

Jason Lemkin 7:16 I think folks are just gonna be nervous about a deceleration in, in quarter-over-quarter growth, and I think they're gonna be nervous over pricing pressure. Look, it's hard to process everything out of, uh, out of, uh, out of the S1A, the quarterly amendment, but it, it sure appear... I mean, OpenAI is basically, uh, running around selling somewhat equivalent products for 30% off. We can discuss how much C- CISOs want China-based open weight models there, but they're putting pricing pressure. Jeff is putting pricing pressure, and it... And the latest models are effectively cheaper. It's just, it's just, it's just you have to have so much momentum to overcome a certain amount of pricing pressure and deflation. I just think the numbers suggest some, some nervousness in it, right? The quarter-over-quarter growth, uh, m- m- i- is slowing, and what does that mean? Um, and how much nervousness does it create? Um, we'll

Jason Lemkin 8:02 just see.

Rory O’Driscoll 8:03 And it's worth pointing out the numbers in very precise terms, 'cause the funny thing is even after we point them out, people can still disagree, right? What it means.

Jason Lemkin 8:11 Yeah.

Rory O’Driscoll 8:11 But at least let's get... I mean, rough and tough, in Q1, and I'm using GAAP numbers here, not runway, 'cause GAAP numbers, as I say, are audited, and the CFO will go to prison if they're wrong, so I always assume they're correct, right? Roughly 5 billion in Q1, roughly 11 billion in Q2, roughly 17 billion in Q3. Those are the numbers, objective fact. What it means from Q1 to Q2, you know, you can look at it as an absolute number. They went up 6 bill, and they went up probably a little under that, about 5 billion Q2 to Q3. So on an absolute number basis, you know, th- that's one perspect- Let's look... You can also look at it as a percentage. They, they, they were at, you know, what is it? 5 or 6 billion in Q1, they went to 11. They went up. They, they doubled Q1, so their Q on Q growth rate was, you know, 100%, which is an astronomical year-on-year growth rate. Then they added exactly the same number in Q3, but by definition, 'cause they're starting at 11, right? Now you only a- now you're adding the same absolute number of, you know, ARR, but on double the revenue size, so the growth rate naturally, the Q-on-Q growth rate has, quote, unquote, "plummeted" to a roughly 50% Q-on-Q growth rate, 'cause you started at 11, you add roughly 6. I'm just using rough numbers. It's roughly growing 50% Q on Q, which is still 5X year on year if you compound it. So if you s- if you... So if you buy into kind of percentage-based growth, you could say, "Hey, they grew 100..." I mean, they grew 100% in Q1, Q on Q. They grew 50% Q2, Q on Q. Q2 to Q3, right? So if they stick at 50%, that's still 5X year-on-year growth rate. That's the positive spin, right? The n- the, the negative spin is, "Oh, they were, were growing at 100%, and only a quarter later, they're growing at 50%," right? It's... Those are equally true statements of the same facts. And as Jason said, another way to think of it is just really at the margin, they added a little high 36's. I can't remember which, Q1 to Q2, and the absolute number, right, that they added Q2 to Q3 was slightly less than that on a GAAP basis. So the most negative interpretation is, which I'm not saying I agree with, by the way, to be clear, is that you're adding a li- a, a, a, a, uh, you're adding roughly 20 billion a quarter of revenue, in which case you, you know, you're, you're slowing out pretty quickly, right? And literally, as I said, you can throw out those numbers, and I'm sorry that was a wall of numbers, but, you know, you can squint one way and say, you know, they're still growing at 50% Q on Q, which is 5X year on year. That's freaking amazing. You can squint another way and say, "Oh my God, it's deaccelerating from 100% Q on Q to 50% Q on Q. Does it go to 25 next quarter?" All the money is in that bet. That's the only discussion to have.

Jason Lemkin 10:45 Look, the way... Listen, Anthropic obviously isn't a startup at this point, right? It's not even a scale-up. But i- if I as a startup saw someone that did instead of 6 billion, 6 million last quarter, and only did 5.4 billion this quarter, I would not invest.

Rory O’Driscoll 10:59 But hang on, let's be clear.

Jason Lemkin 11:00 I, I would not invest.

Rory O’Driscoll 11:01 Can I just... I, I'm just gonna put out... I'm gonna be very clear. It's not that they did 6 million and left out, that they added newly ARR. I mean-

Jason Lemkin 11:07 Yeah, they added, but they, but their bookings-

Rory O’Driscoll 11:09 So it's the first-

Jason Lemkin 11:09 I think the booking-

Rory O’Driscoll 11:10 No

Jason Lemkin 11:10 ... bookings de- de- decelerated.

Rory O’Driscoll 11:12 Yes, the rate of bookings growth de- deaccelerated. The absolute number of bookings went up. I mean, they sold more shit in Q3 than Q2 by about 5 billion.

Harry Stebbings 11:19 Uh, what's actually interesting though is given the bookings decelerated, one thing is like, oh, reaching TAM levels in enterprise, they're 0.1% of Fortune 500 spend. And so actually, there is so much more budget for the to co- for them to consume, and the bookings still decelerated.

Jason Lemkin 11:37 There's not enough data in here for... Or, or maybe I'm not smart enough to parse it. That's why to me, the answer is simply pricing pressure. It's that simple. The d- the d- the demand is there. The market should, if you look at OpenRouter and others, the demand does not appear to have declined. I don't see any metrics that say p- penetration of the Fortune 500 has reached 100%. Um, but you've just got, you've just got a lot of pricing pressure everywhere. Um, so whether it's a combination of folks buying from other vendors 'cause it's cheaper or implicit discounts that they've added in Haiku 5.5 and others, um, and that, that, may- maybe we've reached the point where, where, where Anthropic Ability to command a price premium has hit a local maximum. That's one potential thesis. It is Open-- I mean, I, I'm sure if we saw OpenAI's numbers, they'd look much worse if they're literally offering a 30% discount on Astra, right? With a, essentially what appears to be a higher cost basis [laughs] . And they're offering a 30% discount, so lucky they're not, they're going public second, right? But it appears to be working. It appears to be working. 30% off is a big deal. It approaches the 40% that we target for open weight models. And, um, this is, this is my thesis is that pricing pressure hit them in Q3. It's just, I just think that's the, the only thing I can point to, 'cause everything else is going pretty darn up and to the right.

Rory O’Driscoll 12:51 I mean, yeah, probably true. You're, you're right. I mean, it's gotta be something like that, Jason. I mean, it is just worth stepping back and reminding ourselves again on a GAAP basis, $4.5 billion in GAAP revenue last year, full year, exiting a run rate of $9 billion, and they're probably gonna do $56 billion in GAAP revenue this year. So year on year, their growth rate is 12X, right? That's just an objective fact at this point. But then the interesting thing to your point is you look at the same kind of growth rates, and you come up with roughly $120 billion of GAAP revenue next year. So they're probably gonna go from 12X growth rate to 2X growth rate in a year, right? So to your point, pretty sustained slowdown to still magnificent numbers. You know, it really is, it really is glass half full, glass half empty. You squint one way and you say, "Oh my God, this is a company doing $56 billion in revenues that can still roughly double next year. That's freaking amazing." And you squint the other way and you go, "Oh my God, it went from 12X growth to 2X growth. Is it gonna go to 50% growth or five, 10% growth in, you know, the following year?"

Harry Stebbings 13:50 Do you remember when, do you remember when we had Stewart Butterfield in Slack and zero to 10 in 18 months was, like, the gold standard? Like, the gold standard.

Rory O’Driscoll 13:57 Yeah, yeah.

Jason Lemkin 13:58 You know what's just a little bit interesting? I know it's just an anecdote, and people will make fun of me, but I'm trying to launch this app, Sastra Connect. Hopefully it goes out in the last, next couple weeks, but I've been working on it for a while, um, this AI matching recruiting app. And when I switched it myself to Haiku 5.5 and Jev, um, total cost went down about 70%, 72%. Now, there's some costs that are gonna keep going up. We're all gonna run five agents a day, 10 agents a day. We're gonna run them on multiple platforms. I just upgraded to Muse paid yesterday, right? So there are, there are trends going the other way, right? As we see in the Anthropic quarter. But, but, but, but again, 70% for... If, if a lot of mature applications can see that kind of, um, deflation from im- from improved quality of tokens and pricing pressure, uh, that's a lot. You know? E- even though ev- everyone's gonna be coding at 72 hours a day with 20 agents, uh, s- a 72% decrease in effective price is not, is material. And it's faster. It's faster and 72%, uh, cheaper and, and better. So that, that hadn't happened yet, at least in my personal ex- agentic experience since the beginning of 2015. On an, on an as-used basis, nothing got cheaper. No matter what anybody said on evals and token price, nothing got cheaper until Haiku 5.5 plus Jev. Then I saw a 72% decline for B2B workflow running constantly. Um-

Rory O’Driscoll 15:18 For the same task.

Jason Lemkin 15:19 Yeah, for the same task, and q- higher quality. So, like, it's just one, and people can make fun of me, but, um, it, it is interesting. For me, it is interesting for that actually to happen in the real world for the first time. It just makes this hard to predict for me. It just makes it hard to predict how it's all gonna... I mean, 12 months ago, none of these models barely worked. Um, today, they're epic. Um, in black October, they've become weapons, um, to destroy Binance and everybody else. It's hard to know in 12 months how these, how these sub-metrics are gonna play out. I think it's almost unpredictable.

Harry Stebbings 15:50 But I think that exact statement, Jason, is why we have the unpredictability or the binary nature in venture markets. It's why we have the volatility in public markets, which I think is just, like, a lack of certainty around what the five-year looks like. I think the whole world does not know what five years out looks like in a way that we generally have done for the last 30, 40 years.

Rory O’Driscoll 16:10 We don't know.

Jason Lemkin 16:11 I'll tell you, not to share too many other personal stories, but just comparing to the show, this morning I finally churned off Canva after 12 years. After 12 years.

Rory O’Driscoll 16:19 I saw your note.

Jason Lemkin 16:20 Don't need it.

Rory O’Driscoll 16:22 Oof.

Jason Lemkin 16:23 And l- I just think at the start of this year, Canva was still accelerating. You know? There were report, there were some rough reports this week about its current growth, right? They kind of varied. I don't know what's true. Um, one of them said growth was zero. Others said they were preparing layoffs. Who knows what it was? They, it, it at least seems to have decelerated 20%. It's just that predictability, right? It's just, you just don't need it. The agents don't need it. So it's, it's hard to predict one year out, let alone five, so.

Harry Stebbings 16:48 Jason, what happens to Canva? Does it go out?

Jason Lemkin 16:51 I don't want to be a Debbie Downer, right? The Debbie Downer case is obvious. They missed their window, right? The Debbie Downer-

Harry Stebbings 16:57 And then, but then what? They miss their window, and they're just like a private company that grows slower.

Jason Lemkin 17:00 I don't know. I mean, it seems to be working for Automattic, o- outside of a few, uh, CEO turnovers and [chuckles] , and temper tantrums [laughs] .

Rory O’Driscoll 17:07 I, I want to be clear. Was that humor?

Jason Lemkin 17:09 I, I-

Harry Stebbings 17:09 [laughs]

Rory O’Driscoll 17:10 Okay, got it.

Jason Lemkin 17:10 The Automattic one?

Rory O’Driscoll 17:12 Yeah.

Harry Stebbings 17:12 Yeah.

Jason Lemkin 17:12 I don't think it's humor. What I mean is I think Automattic will stay private forever. I don't think it will e- it has raised its billion, right, as 40% of the web. I don't think Automattic will become an agentic company, and I think Matt is fine with it, and I think it will stay private f- I genuinely think it'll stay private forever. I don't know what the, what the VCs do with it, right? But I don't think it'll ever go public, and I don't think it'll ever get acquired, Automattic.

Rory O’Driscoll 17:33 The only comment I'll make is software companies in particular, you know, if they're not moving forward, you know, it's like the shark. You're moving, you, you, you, you're going down, right? And over an exte- I mean, it might take an extended period of time, but you know, you, you're... And I remember, I remember there used to be mainframe companies in the '90s and early 2000s that were still, you know, living off their 20-year-old contracts, but the asset just becomes very terminal valued. Yes, you can stay private and kind of continually cut costs and live off your install base or, you know, do your own little bending spoons on your own, but it's just, it's a tough place to be, and it's hard to get value from it as an investor. So I, I, it will be interesting to see what all, many of these companies decide to do, right? You know, can you, can you clamber aboard, even if it's a little late, the AI train and get enough traction to maybe not be the winner but grow 5%, 10%, 20%? I don't know But I think if you don't do that, if you decide to just not play that game, I think over the medium term, you're just heading to quiet obsolescence. And no-

Harry Stebbings 18:32 I'm so picking a fight here 'cause Cliff, Cliff always messages me. I mean, Cliff loves the show, and he always messages me-

Rory O’Driscoll 18:38 He's the best

Harry Stebbings 18:38 ... like, "You guy- you guys are harsh on me," and I'm like, "You know, Rory and Jason. I never say anything. I'm just-" [laughs]

Rory O’Driscoll 18:44 I didn't say anything harsh.

Jason Lemkin 18:45 You know, you know what? You know what the surprising thing from all-- where I've gotten surprised, guys, going to this point? Listen, obviously, if you're doing what we're doing, you have 21 agents running, and they're running agentically, and they're using Higgsfield and Opus and the other things. It makes sense you'd turn off Canva, right? Because your agents don't log into it, right? And it, and it, it's really not useful for our agents. The surprising thing from all this is that the real world isn't buying much of this crap anymore, right? If I would've thought that a Canva, a Squarespace, which just launched 20% layoffs, a Wix, which is gonna be bailed out by Base44, others, I would've thought, like Shopify, they'd be doing fine because most of the world isn't running 21 agents and then, and then, and Higgsfield and Musefine and everything. We're not all tech nerds. But it isn't really playing out that way. It isn't... And, and obviously some of this is because, you know, ChatGPT is so good for images, right? We're, w- there are, there are things that are so good. Groq is great for images. But I'm just, I'm surprised that a lot of these core businesses are being so, so severely impacted. Maybe it's the limit of my knowledge, and it, and it wor- it wor- it worries

Jason Lemkin 19:44 me.

Harry Stebbings 19:44 Instinct, Instinct's amazing for images. Yeah. Maybe.

Jason Lemkin 19:48 Instinct's great, but I mean, you s- you s- I mean, uh, in your news, you know, Elon just had to cut back GroqBot use because it's too expensive. He can't afford it.

Harry Stebbings 19:55 [laughs]

Jason Lemkin 19:56 I think I underestimated what the costs of Instinct were-

Rory O’Driscoll 19:59 Yes

Jason Lemkin 19:59 ... because I didn't understand how persistent the VMs can be. Yeah, I, I think Instinct could cost as much as 20 bucks a month per user to service Instinct the way it's supposed to work today.

Harry Stebbings 20:07 Way more.

Jason Lemkin 20:08 Yeah, well, that's not-

Harry Stebbings 20:09 Way more

Jason Lemkin 20:09 ... more than 20?

Harry Stebbings 20:11 Yeah. Oh my God.

Jason Lemkin 20:12 Well, it's, it makes... It's, then it's, it's m- it's a moronic business model at $40 or $50 per user per month. I got it wrong because the lovable Replit thing, right, which is a highly tuned on-off VM, is about four to five bucks a month, okay? They have hu- they have large teams tuning that so that you can have a little bit of AI, you can have a little... But the, the VMs spoo- spool up and down very carefully. They've gotten it down to four to five bucks a month. So I assume that Muse and friends were, like, Muse could even be cheaper, I assumed, right? And Instinct might be twice that because they don't have the resources. But it's much higher. It's much higher to have these persistent agents running. It's much... You think it's more than 20 bucks a month? How, how... I know you wanted to do it at 30, Harry, but how does this math pencil out at $30 or $40 a month per user? All these tickets we're gonna buy?

Harry Stebbings 20:53 Well, I admit that that seems relatively paradoxical to believe that it's more than 20 bucks a month and to want to do it. It's a very good argument from you there. Uh, but my thing for the power users, it's very significant, and I think for people that use it once or twice a day with limited automations, it's very, very cheap. And so I think your blended is probably much cheaper. But, like, for me, I, I will be terrible as an economic profile for them. I mean, the worst. [laughs]

Rory O’Driscoll 21:15 Look, I, I, I think real, the real truth is, just like every digital product, it always takes off at a point in time when the economics don't quite work. I remember when Facebook was ramping early on, and people were like, "Oh my God, the server bill." It was, you know, using scads of servers, and it was maybe gonna have CapEx of $100 million, and people were like, "That's the end of the world." And the bet you take over and over again is digital goods get, you know, reduce in price 50, 60% year on year. So if people like the shit and they get hooked on it, over two or three years, the math gets better.

Jason Lemkin 21:47 YouTube.

Rory O’Driscoll 21:47 Right? And that's really what's going on here.

Jason Lemkin 21:49 YouTube.

Rory O’Driscoll 21:50 Right, one of the best businesses in the world.

Jason Lemkin 21:50 May- maybe, or maybe-

Rory O’Driscoll 21:51 Yeah

Jason Lemkin 21:51 ... or maybe we're gonna see these waves where for all but a handful of folks, the subsidies just can't last. We've also seen this in AI. The subsidies just can't last. It's too expensive, right?

Rory O’Driscoll 22:01 And they won't.

Jason Lemkin 22:01 Um, and I don't know whether Instinct is on that side, right, or, or not, right? I mean, Cursor barely pulled it out. Cur- the, the Cursor V2 would've died, right, reselling Anthropic tokens, right?

Rory O’Driscoll 22:12 If you, if you're going to regard Cursor as a failure, then, yeah, you can win the argument. I mean, I'd take the 60 billion and call it a win myself.

Jason Lemkin 22:18 I said the V2 version of Cursor-

Rory O’Driscoll 22:20 Yeah

Jason Lemkin 22:20 ... would've died, just selling Anthropic tokens at a loss perpetually. If that's all they were able to do, they would've died. Great teams figure it out. I get the venture thing. I wouldn't mind being Sarah Guo and investing in Instinct at 50 or 100. I, like, I wouldn't bother me to. I wouldn't be too bent out of shape about the business model if I invested at 50 post. Don't get me wrong, okay? Wouldn't bother me a lick. [laughs] But I, I just don't know which side... I mean, A- Amjad from Replit said yesterday, like, the era of these free subsidies for the agents are coming to an end. If Elon can't afford it, uh, Muse can afford it for now, but what i- what is, what is Muse, what's the scale of the Muse subsidy, right? This is the second fastest-growing product that they've ever had. If they're, if, if, if it's 10 or 15 bucks a month for them, uh, who knows? Maybe Amjad's right. Maybe, maybe, maybe it will come before they can raise that six billion at 30.

Rory O’Driscoll 23:07 Yeah, well, as a side, j- just for context here, if it's 100 bu- uh, 10 bucks a month from users and they're subsidizing, that's, call it 100 bucks a year, right? That's, uh, for every million users, that's 100 million bucks. This is a guy who's spending $70 billion on CapEx. He's in, right? He ain't gonna blink 'cause, you know, the compute bill goes up by a couple of billion dollars, right? And I, so I, I do hear-

Jason Lemkin 23:29 Yeah, Meta can afford it. Meta can afford it, right?

Rory O’Driscoll 23:31 Look, just, I, I wanna go back to your comment 'cause it is inter- You... 'Cause and once... Look, I, I tend to be the boring guy who says that the numbers have to work in the end, and yeah, gross margin, Cursor, blah, blah, blah. But when things are moving so quickly, you can't wait until the economics work perfectly to launch a product, 'cause someone else armed with venture dollars will launch that product 12 months earlier when the p- when the economics don't quite work, grab all the mindshare, and then they'll basically lean into the cost curve and, and hope that the improvement economics bails them out, right? And that's the bet you're taking. And then to take the Cursor example, I mean, they're doing some version of, hey, look, if it, if, if you don't get a big acquisition for 60 billion, well, if my costs, if, if what you said earlier, if I can get my own model and my costs come down, I make out like a bandit. If I'm still stuck on Anthropic in 2027, I'll probably get fucked over, right? But it's still a better bet than waiting until 2027 and saying, "Oh, now that there are lots of open source alternatives for f- state-of-the-art coding models, I should launch a coding company." 'Cause it doesn't matter, it's done, right? The, the truth is the capital markets and the venture markets are just so aggressive that waiting until everything makes sense is not an option. So what that means is kind of by the reciprocal property, almost everything that's new and digitable will look crazy for a while in terms of economics, and the only question is, can the economics get better before the money runs out?

Jason Lemkin 24:50 Look, no question. Uh, and we can move on. But it ju- it's interesting that Elon had to cap it.

Rory O’Driscoll 24:55 It is.

Jason Lemkin 24:56 They have resources. They had a pretty good IPO. He has re- he's going long. He has bet his company on AI and subsidizing this thing. He subsidizes everything round trip, right? Uh, even with his personal fortune, he subsidizes everything. This is a man that would not want to cap Grok bot if he did not have to.

Rory O’Driscoll 25:12 He's also the man who's getting so many dollars per unit of compute from Anthropic that he's like, "Yeah, I could do some free Grok bots, and these people could make cute cartoons and send pictures of sexy a- sexy anime pi- girls to each other. On the other hand, I could get, like, 50 bucks an hour from Anthropic. Screw it, I'll do that."

Jason Lemkin 25:28 Well, they're not doing that. They're building applications, they're building code. Gr- Grok, Grok is f-

Rory O’Driscoll 25:32 I know

Jason Lemkin 25:33 ... fairly technical-

Rory O’Driscoll 25:33 I'm just picking a little fight

Jason Lemkin 25:34 ... un- unlike the Instinct and Muses, right?

Harry Stebbings 25:36 Okay, boys, we had a, we had another pretty big transaction which went down that in a normal world would have dominated news. Um, i- in today's news, it kind of lasts through a couple of hours, but DoorDash buys Instacart for $17 billion. Under half its 2021 private price. Uh, D1 still owns 9.7%. Sequoia owns 8.1%. Um, before we talk about agents, which was a disclosure, and the thing of note here, an, an inevitable ending to a clear second-place provider to DoorDash.

Rory O’Driscoll 26:11 I mean, inevitable is a bit strong, but yes, they clearly tried to do it before. It clearly makes sense. There's also probably an element of do it now while you can if you're DoorDash. So yeah, all makes sense.

Jason Lemkin 26:20 Listen, some of this is not my strength, but the fact that the Postmates acquisition worked was a surprise for me, right? I assume you buy -- When Uber Eats bought Postmates, I assume you get a little bit of a brand and a list, and you basically shut the thing down and that I didn't really believe these deals would be accretive, right? Um,

Jason Lemkin 26:36 but they appear to be, right? So if Instacart's got 20% of the market and DoorDash has 60 of, of, of the delivery market, and you can get through antitrust, which you probably shouldn't if you think about the numbers that I just said, and you can get to 80% market share, this is how you dominate networks, right? You dominate ev- every part of it by having 80% market share. And I'm surprised that you can combine these networks and they don't end, one plus one doesn't end up equaling 1.1 a couple years later. Like, that's, that's my experience. But it doesn't appear to be true. Does-- These deals, if well-architected, appear to work. And so, um, it's, it should be accretive. It should make sense. And, um, I think also, like, uh, you, you noted it's half its '21 private price or whatever, right? I think like the Braze deal at Salesforce, you may say what's the similarity? A lot of times folks have been circling each other for years, right? Uh, the, this, th- Tony said they've been talking for years, right, uh, with the s- So it's just there are deals that are-- that don't make sense e- even if they're cheaper, and then there are deals where, like, Salesforce bought Braze. Salesforce was losing to Braze in marketing for years, and the- when it got super cheap, when it got to 3X ARR, now they had to play a, pay a big premium. They just buy the assets they want when they become cheap, and it may be that simple. It may be... And, and so the, um, if they're accretive, buy 'em. Bu- this, this may be just simple things that the corp dev is supposed to do.

Harry Stebbings 27:52 One that I found fascinating was actually relating to a conversation we've had before, and it was agentic usage of these platforms and DoorDash saying that AI agents have sent only several hundred orders against three million deliveries a day. Were you shocked by how surprisingly low that was, or do you think it's symbolic of the fact that people will always want to choose food and have that discovery process?

Jason Lemkin 28:16 I think in a sense it was a dumb comment for what it's worth. But it reminds me a little bit when Brian Chesky, like a year and a half ago, said there was no need for AI, uh, for, for, for, um, you know, for, for, for, for, for rentals and housing, or when Parker Conrad, another generational CEO, said there's no need for AI in B2B 18 months ago. It, yeah, today, today only several hundred orders, but, um, but these, these, these platforms like, like all, all these app platforms have a risk of slight disruption from agents.

Harry Stebbings 28:45 Well, I think it's incredibly-

Jason Lemkin 28:46 They all have a risk of slight disruption

Harry Stebbings 28:47 ... i- wh- which is why I think it's so self-serving. I would be pushing this.

Jason Lemkin 28:50 Yeah, self-serving. Self-serving.

Harry Stebbings 28:51 Whether, whether you're Uber, and when I had Mark, the president of Uber, on, he very clearly was most worried about agentic usage of these platforms because it would disintermediate Uber, and you could get-

Jason Lemkin 29:01 Yeah

Harry Stebbings 29:01 ... free now, you could get a Lyft, you get price optimized, time optimized, and who gives a shit about the platform provider? You have it for Vinted, you have it for Uber, you have it for DoorDash. I think it's, like, very specifically well done because it made headlines and news in a way that actually he wants. No, humans are here, and we will continue to be here.

Rory O’Driscoll 29:18 Maybe the better way to say it is let's assume they're telling the truth. That's the objective fact about what's going on now. And the million-dollar question is, do you see that behavior happen more and more, and then how does DoorDash respond to it? I mean, doing versions of the same thing, which, I mean, I could totally see for some of your standard food orders just it's not that much easier for the record. I mean, telling your agent to order your dinner, it's just as easy as to just go click on the DoorDash app, but-

Jason Lemkin 29:41 Well, but there's a thing. DoorDash has two competitor, big competitors. It has Uber Eats, right? Um, and it has the restaurant.

Harry Stebbings 29:48 Yes.

Jason Lemkin 29:48 The restaurants don't want to use DoorDash or Uber. So most restaurants have self-pickup. At least they have to-go pickup where they can avoid most of the fees, right? Or self-delivery. An agent can figure that out for you. "Hey, Harry, did you know that Lemkin's Curry down the street, actually it will deliver for $2, and you don't need to go through, uh, DoorDash or whatever they have in London?" I did not know. Thank you, Muse. Thank you, Instinct. I'll take-- That, that'll save me $14 on my order? What's the pub right behind your office that I really like, Harry?

Harry Stebbings 30:15 The Grazing Goat.

Jason Lemkin 30:17 Love that place. Uh, five out of five, okay? Uh, if, maybe they could just drop it off at the, next time we do the pod for free. I don't have to pay 20% and plus tip, plus this, plus... You think, think I'm kidding? I think it's, I think there are, uh, is, agents are an existential risk to DoorDash because their number three competitor is, is nobody.

Harry Stebbings 30:33 Can I, can I put out a cultural difference between us?

Jason Lemkin 30:36 Yeah.

Harry Stebbings 30:36 You guys pay 15 bucks a burger, and it gets delivered at 50, and there's, like, federal charge, state charge, delivery charge, service charge. It's ridicul- in London, it's kind of like 15 quid is 15 quid. It, it's fascinating to see the difference in, like, what you pay and what you get.

Rory O’Driscoll 30:52 'Cause, and that's why these businesses are wildly profitable. They figured out exactly how to extract every dollar of rent from the person who wants to sit on their couch, press a button, and get their meal an hour later, or 10 minutes later, whatever. Yeah, it works.

Jason Lemkin 31:03 It's been a brutal business since the old GrubHub days. It's just evolved with the internet, and agents will evolve it again. It's been brutal. It's been brutal. The idea is take as much margin out of the customers or the restaurants as humanly possible, right? It's brutal.

Harry Stebbings 31:14 I'm, I'm with Bill Ackman. I think Uber's a banger of a buy right now, though. I mean, it's so cheap.

Rory O’Driscoll 31:20 I broadly agree, and I haven't spent a lot of time on this, but the two worries you have are way more disruption than agentic disruption, and what you're implicitly saying is border fine.

Harry Stebbings 31:28 Uh, yeah. Agentic disruption is bizarre in taxi hailing. You're gonna go to-

Rory O’Driscoll 31:32 Agreed

Harry Stebbings 31:32 ... your WhatsApp and say, "I would like to get-"

Rory O’Driscoll 31:34 Agreed. I don't see it. I agree. I don't think that's-

Harry Stebbings 31:36 It is, it's a, it's, it's a worse user experience. I would like to get in an Uber. I am at this position on the map.

Rory O’Driscoll 31:41 Yeah.

Harry Stebbings 31:42 And the other issue is people love food. They love looking at it. They love pictures. They love scrolling. They love discovery. The only thing it makes sense for is, like, automation and scheduling, like every day drop it off at Scale's office in, you know, San Francisco.

Harry Stebbings 31:56 That's it. Okay, boys, we have two big rounds in Kalshi and Cohere. Kalshi, a billion at $40 billion, one of Jason's favorites, a low dilution round there, uh, with a two and a half percent dilution-

Jason Lemkin 32:06 I love the low dilution rounds

Harry Stebbings 32:07 ... but still raising a-

Jason Lemkin 32:08 Yeah

Harry Stebbings 32:08 ... billion dollars and a $40 billion price a day after Utah charges, uh, founders of Kalshi. Uh, and then we have Cohere. I think it's the largest private raise for a Canadian company at this point, um, with actually the Canadian government, I believe, uh, well, Canada itself, investing up to $1.14 billion. Which one's more interesting?

Rory O’Driscoll 32:28 Kalshi is a money deal, right? Let's be really clear here. The thing is working. It's growing like a weed. Deep in their heart, everyone knows that we started off thinking we were doing prediction markets, which had some social utility, and we've long since converted into sports gambling markets to, you know, feed people's addiction for gambling, right? So we're all over the doing good. We're all over the higher purpose. Um, you know, we're pushing through a whole bunch of local regulation, which is why Utah is suing them, 'cause sports betting is regulated on a state-by-state basis, and these guys are plowing through on a national basis. So there's nothing to see in this other than VCs will put a lot of money into shit that's working. That's the takeaway, right? And, you know, at f- you know, if, if, if it's at $40 billion, it's had momentum. It raised like three or four months ago at $22 billion. So this is, you know, one of those classic momentum plays, right? I mean, contrast, Cohere's very different. I mean, it's almost not a momentum play. It's a, "We need our sovereign AI play in Canada 'cause you are nasty people, and you've threatened to invade us multiple times, and we've got this product," and it's, you know, one of the original four or five LLMs, to be fair. I mean, back in 2023, you had Cohere, obviously OpenAI, Anthropic, and then all the other ones like the one that Microsoft bought, Adept, that Amazon bought, all the others folded. So give, I mean, give Cohere huge credit. They are one of the survivors of that first group of e- of LLMs before Neo Labs was e- even a thing, and now they're, uh, y- they didn't release revenue numbers, but you can see how you could make a sovereign AI story. It's almost the exact opposite. In the case of Kalshi, the government is saying, "I want to stop you now and put you in prison." In the case of Cohere, the government's saying, "I'm gonna make this work no matter what, even if the economics don't work." So they're almost like the yin and yang of investments in terms of their relationship with the government.

Harry Stebbings 34:11 Does that work? Can the government just do that? At a point, it doesn't work.

Jason Lemkin 34:15 Do, do what? Do what? Do what?

Harry Stebbings 34:17 Well, hey, we, we wanna make you work, and so we're gonna inject you with $1.164 billion. Uh, totally. You can do it once-

Jason Lemkin 34:23 Isn't that what, isn't that what our governments do all the time? They subsidize everything.

Harry Stebbings 34:27 Y- yeah, you can do it once. You can do it twice, but ultimately, if enterprise adoption isn't there, it doesn't work.

Jason Lemkin 34:33 Sometimes it's worth just being in the game. Some- it's like, it's like-

Harry Stebbings 34:36 Yeah

Jason Lemkin 34:36 ... putting a few chips into Anthropic in the D round. Sometimes it's worth just being in the game, and sometimes it bounces, bounces your way, and you've always got the liq pref.

Harry Stebbings 34:44 Jason, the D round was $16 billion, so I'd, I'd take the D round. Yeah.

Jason Lemkin 34:49 You know, I don't, actually, I, in all seriousness, I don't know how, like, Canadian go- like, fed- like, national governments work. When you talk to folks that run at the top companies with large corp dev portfolios, public companies, really, they just want it not to go down in the aggregate. They love the gains, right? But you only get so many credits as a public company for, for the gains from your portfolio, and the losses really suck, right, because they swing your EPS in really weird ways, and you, you know, you have to put asterisks and daggers next to it. So as long as they don't lose too much money on, on, on the investment, I don't know that they, I don't know that they care.

Rory O’Driscoll 35:20 You're right. The motivation for anyone doing this isn't purely financial gain if you're the government, right? Agreed. But Harry asked a separate question is, can it work? And I think the answer there is, yeah, you can keep it g- you know, governments can make things happen for a fair while, but what they can't do is make the private sector use the product. So, you know, you see it with Mistral too. T- to the extent that there's a willingness to spend Canadian government tax dollars on a Canadian government AI solution, or maybe if Canada goes on with its kind of, you know, affiliating with the EU, this may be, the- these guys in Mistral may be the you're not America LLM options, and there'll probably be some spend in that. But do you think it'll be a compelling business that will go public at 3X this round? Not as obvious to me at all.

Jason Lemkin 36:03 Well, in today's world, today's day and age, Cohere not disclosing revenue is hardly a flex.

Rory O’Driscoll 36:07 No.

Jason Lemkin 36:07 [laughs] I mean, it's, it's a pretty bad sign. [laughs]

Rory O’Driscoll 36:10 You know, I, I, I love, I love it. It's like ARR grew more than 165%, and you're like

Harry Stebbings 36:15 To

Rory O’Driscoll 36:16 Since inception?

Harry Stebbings 36:17 Yeah, no. But who knows?

Rory O’Driscoll 36:19 Okay.

Jason Lemkin 36:19 You know, one thing I was thinking on the Kalshi thing before we leave. So Sequoia led the last round, right? At a billion. They put in a billion-

Rory O’Driscoll 36:26 Yep

Jason Lemkin 36:26 ... right after felony charges were against the founders, right? I get that it's just a state. And, and then we have Binance, right? 6.6 billion taken by apparently North Korea in a matter of hours. I had to wonder between Sequoia and Binance, if the ball had bounced just a little differently, would FTX still be around today, and would it still be a winner? Same VC, same risks. Um, FTX got... Um, listen, you, you obviously can't take money from one sister company and put it in another, but I think Binance barely survived this 6.6 billion. They don't have a transparent balance sheet. We don't know how much money they had. They cert- What if it was 13 billion? Where would they have repaid it? I, I think they were a, I think they were one step away from, from an FTX-type situation. Um, luckily, the, he got pardoned, CZ, right? But I just wondered if, if... It made me think of, since you have the same investor and the felony, we've just all changed the way we take risks because it's game on, right? Um, I, it just made me, made me think about a parallel universe where FTX was a, was a shining portfolio company today.

Harry Stebbings 37:24 Do you know what? I mean, the fascinating thing is FTX's $500 million position from Anthropic would today be worth-

Jason Lemkin 37:29 Well, that we know. Yeah

Harry Stebbings 37:30 ... would be worth 39 billion at the $1.5 trillion. Uh, so timing is important is the takeaway, kids.

Rory O’Driscoll 37:36 The most compelling factor about the Binance thing was, um, was that they could pony up, uh, $6 billion to replace the money that was taken. Like, what a good business where, you know, you... I mean, as- assuming that you're not doing an FTX and swiping it from someone else. As, as I understand it, and I did not spend a lot of time on this one, is that they, Binance were able to make good the crypto loss, right?

Jason Lemkin 37:57 Yes.

Rory O’Driscoll 37:57 Which implicitly means Binance makes enough money to be able to make good the crypto loss, right? Which means it's an amazing freaking business.

Jason Lemkin 38:04 I think so. I just think, and again, I'm not a crypto guy, but when CZ said they could cover it but it would be tough, suggested to me they didn't have 13 billion. Um, and-

Rory O’Driscoll 38:12 Yeah, no

Jason Lemkin 38:13 ... and what I really thought was ridiculous is how many folks said, said that their balance sheet was all the, the, the, the crypto they hold for, for [laughs] as a trustee, as a custodian. That's not true. That's what, that's why Sam's in jail and CZ is out, right?

Rory O’Driscoll 38:25 Agreed. Yeah.

Jason Lemkin 38:25 Uh, I don't think they had 13 billion. I think they got out b- in, in Black October.

Rory O’Driscoll 38:31 Yeah, you can't take my crypto dollars and give them to s- uh, Harry, and say, "Sorry, your stuff was stolen."

Rory O’Driscoll 38:35 Agreed. Okay.

Jason Lemkin 38:36 Yeah, their balance sheet. It, I mean, just think how much still happened in Black October, right? And then we have this, the one that I think is craziest, I don't mean to get off track, is this, I know it sounds m- minor, but it's not, this Ardex thing that, that breached WorkOS and a bunch of others. This is just since we did the last one. Th- this one is insane, this Ardex thing.

Harry Stebbings 38:52 To, to... I saw that, um, Michael Grennan post about this, uh, 16 hours ago, but tell everyone what this is.

Rory O’Driscoll 38:57 Tell us more.

Jason Lemkin 38:57 Ardex, we should be scared about. Ardex is a, is initially well regarded. It's only, like, a couple months old. A Chinese open source pen test solution to penetrate test applications. We all need pen testing. This, we always needed it before. We certainly need it in an agentic area where everyone from Binance to the rest of the world's being breached, right? We're all being breached. All the refrigerators in Europe are going down because they've been breached. It's, it's, this is, Black October was much blacker than we made it look. And so some, some really smart developers just a couple months ago in China built an open source thing. Great. So, so it's free, and you could hook it up to your DeepSeek and your other LLMs, and it will just try to hack you, right? And this is what you're supposed to do. You're supposed to hack yourself. Um, and, um, but it became very easy to turn it around and just attack other folks with agents. It became [laughs] it was flipped around, and it's become one of the, the greatest threat. And a lot of folks started saying, "Well, you could always, you could always do this," but now you can do it with LLMs at scale. You can just tune it against anybody you want.

Rory O’Driscoll 39:51 So, so, so they originally built it as defensive, and ba- Yes, everyone said, "Oh my God, this is a pen testing tool." Uh, yeah, make- makes total sense, is that, yeah, it's like the, you know, it's like, it's like the, "Guns don't kill people. Bad people with guns kill people." You built a gun, and you hope-

Jason Lemkin 40:04 Yeah

Rory O’Driscoll 40:04 ... it was in good people's hands, and now it's in the bad people's hands, and they're killing people. Makes... No, I mean, it is surprising to-

Jason Lemkin 40:09 It, people, it worked, seemed to work great at first, right? Until you turn the gun around. [laughs]

Rory O’Driscoll 40:12 There's probably... It's probably hard to overstate, you know, the, the, the, the, the, the, the, the cyber dynamics of the next 12 months. I mean, I, I, you're right. It's one of those stories you never cover it on a daily basis, but you're right. At some point, something really weird is gonna happen, and you're gonna go, "Come on, guys, we should've seen that," you know, when, you know, the entire US Fedwire goes down for a day, and we're like, "We can't move money."

Harry Stebbings 40:34 Yeah. And, and, and we joked about the refrigerators in Europe, which for people that don't know, there was a hack which basically meant someone could turn up the, the temperature on refrigerators, and they did, to 50 degrees. Now, it was luckily rectified, and the frozen peas were saved, um, which is of great relief to everyone in Europe. Um, but you could see this being much more mission-critical things like traffic lights or, like, heating facilities or you name it. Um, so it is, it is quite alarming. Um, ding, ding, ding. Mr. Lemkin, we had quite a few rounds, and it's your turn to bring something to the IC this week.

Jason Lemkin 41:09 Goodness.

Harry Stebbings 41:10 You've, you've got a lot.

Jason Lemkin 41:10 Can I take a break for the rest of the year? We've been busy. We've deployed so much. We've gone back for three funds this year. [laughs]

Harry Stebbings 41:15 No, no, no, we're Andreessen. We just raised a big-ass fund, and so we've got plenty more for you to go through. J- Jason, we're a deployment machine.

Jason Lemkin 41:22 Okay, great. [laughs]

Harry Stebbings 41:23 Which company are you going to take to IC this week?

Rory O’Driscoll 41:26 Wow.

Harry Stebbings 41:28 Serval, Cognition, Baseten?

Jason Lemkin 41:31 Well, listen, I mean, uh, you know much, Harry, as you know, much, much is, in many ways, the Windsurf team re-re, uh, you know, um, lamented having to sell out to Google for just a few billion. The Cursor team's gonna, gonna lament selling to, to SpaceX for only 60 billion, right? Um, Cognition at 80 billion, uh, it, it, it, it seems market correct. Um, now, to get to a 10X with dilution, a trillion-dollar of, uh, valuation, I start to get nervous with the latest, the latest round from OpenAI and Anthropic. Uh, I, I do get worried that fundamentally it only owns so much of its own IP, but jeez, I don't know if you saw the last Q3 of Anthropic. Like, the one thing that is working, that never stops working, is coding agents. And Cognition has a bunch of the field open to it. Um, they made a lot of their own luck. I don't know if the original version of Devon when I brought it to committee, uh, several year, you know, a year and a half ago, I, I don't know that that product worked. They lucked out buying the windsurf team. They lucked out buying IP. They lucked out from Open waits. But this is something with the kind of momentum that I think we can't sit out. Um, and we have a little bit of l- l- money left in the fund, a couple billion, and we need to deploy the next one. I say we close out the last fund with two billion in Cognition. We're already up 5X on paper, we can't lose, so we might as well put the last two billion in Cognition. Um, and I say if we get a little recycling before the, the round closes, we toss it in too.

Rory O’Driscoll 42:55 Interesting. Just for the, um, people listening, because it wasn't obvious, that was a f- much better question than you normally ask Harry, because the, normally you say, "Here's a deal, Jason, tell me why you should take it to committee." And honestly, and respectfully, I say that's kind of stupid because all you're just saying is, "Give me a soundbite," right? The reason this was a good question, much more interesting, was because you have three deals in front of us, and it's worth pointing out to the l- listeners what they were. W- three really amazing companies. You were basically asking Jason to pick between Baseten at thirty billion, up 2X from June, a, an inference company; Cognition at seventy to eighty billion, up from forty-eight billion in September; and Serval at three billion, up from one billion in December. So three different companies, very different businesses. One's an inference provider, one's a coding agent, and one's a h- very hot next generation I, um, AI app in the ITSM space, right? So three different businesses, all of whom have had a round and are getting a second, getting a second re-up within this next, the same twentyf- uh, the same twelve-month period. So as I say, because it forces Jason to choose, that's a much more interesting question, if you know what I mean, than just tell me nice things about this company, right?

Harry Stebbings 44:09 Uh, for the, for editors, can we please cut that segment? That's the first time that Rory has actually congratulated me on a framing.

Rory O’Driscoll 44:16 It, it's okay because I think it happened by accident. Um-

Harry Stebbings 44:18 I'm gonna, I'm gonna play that to myself later just to give myself more confidence. [chuckles]

Rory O’Driscoll 44:22 But, but, but actually genuinely, it's like there really was... So super interesting question, and I, I try, kinda-- So w- w- would you like to do an inference provider at thirty billion, the t- a coding agent at eighty billion, or a IT AI app in ITSM at three billion? I mean, good question.

Harry Stebbings 44:40 Uh, uh, no shit. Uh, I, I-

Rory O’Driscoll 44:41 Which would you pick?

Harry Stebbings 44:42 Serval. Are you kidding me?

Rory O’Driscoll 44:44 Okay. So you do, you think your upside in Serval at three billion is better than the upside of Cognition at seventy billion? Yeah.

Harry Stebbings 44:50 Oh my God, the tide is changing. ServiceNow is an old guard. Bill McDermott's amazing, but he's like Slootman. He's a sales-led CEO. We're in a product-led generation. Jake and Tatiana at Serval are incredible. The product's amazing. The enterprise adoption is better than I've ever seen. And the upside, just pure multiple on cash. Now, there's an, there's an addendum and a star which we can go back to. Multiple way better here. The chance of getting a, a 10X here, I think very likely. The chance of getting a 10X for Baseten at thirty billion, God, you've got to squint. The chance to get a 10X at Cognition, to Jason's point, gosh, we're nearly hitting a trillion there. So yeah, I think the multiple is clearly much more evident there. Now, the challenge though that we have with Serval that we don't have with Cognition or Baseten is if we're an AUM play just putting cash out the door, they only raised a hundred and fifty million. And if I'm Andreessen, I want to put five hundred million, a billion out the door. And so doth butter no parsnips, as we say in the UK. I need to do Cognition because I need to move money. But for the multiple, it's the easiest decision you made.

Jason Lemkin 45:58 You know what? The question about Serval, I don't know. I really don't know it at all. I don't think the founders don't know it. The, the, the qu- This is just my question to the I, to, to the, to the IC is does it expand the TAM? Am I bounded for, for real, for real, right? Not for hand-waving. Am I, am I, am I in essence bounded by a s- a s- a su- a, a subset of ServiceNow's TAM today, right? Now, every founder will expand the markets and add concentric circles, but is this really expanding the, the TAM here? Um, I, I generally don't know, but I personally am not excited about disrupting existing players if you don't expand the TAM because I think, I, I think we don't get anywhere at the end of the day.

Rory O’Driscoll 46:34 And I think the argument they'd make would be some version of you don't just, you know, ServiceNow just is a list of things that IT has to do and, you know, Serval is actually a product that does those things for you, so you can command more value from it, right? And you know, the way you'd validate that is looking at their ACV on a like with like basis. In other words, there's a mid-market company that might spend two hundred grand on ServiceNow. Would that same company be willing to spend three hundred or four hundred grand-

Jason Lemkin 47:03 Yeah, that's my question

Rory O’Driscoll 47:04 ... on Serval? And that's what you'd need to see to get TAM expansion. Because yeah, I mean, we think, I, I know my, one of my colleagues had talked to him in the past. This is an amazing company, amazing team, and, and a great market. I mean, it's always worth remembering until recently, ServiceNow was, I think, the second largest market cap SaaS company. You had Salesforce and then ServiceNow. ITSM is a much bigger market than you think because it kind of becomes the, the general corporate to-do list for service functions like legal, accounting. All the back office folks were using ServiceNow to keep track of their tasks. So it's been a way bigger business than I would've guessed when we saw it in, what was it? 2010 or so, right? And, um, so yeah, replacement for that is a big opportunity, right? But I hear you, Jason. It's, it's the can you expand the TAM, and are you just, or are you dealing with a 10-year grind ServiceNow down deal where you do it forever?

Rory O’Driscoll 47:54 Whereas Harry, to take the other side of the argument on Cognition, I mean, the beauty of the coding agents is, especially with Cognition, they're ripping open all that enterpri-- I mean, I'm not... They're coming into large enterprises and basically saying, "We will help you effectively, you know, rewrite your entire code base, enable your team to be agentic coding," and getting huge dollars for that. So in terms of money velocity on the table right now- Cognition has, you know, just an ability to command way more dollars. I think Serval probably has the compounding effect of there's probably 10 or 15 years of grinding down ServiceNow to a bigger outcome. So I don't know. It, it's, it's kind of velocity but capped versus, um, long-term me- great mega trends, but is it gonna take a while?

Jason Lemkin 48:37 Look, and this, it may well resolve tickets that ServiceNow can only track. It may be an epic play. I'm not... I'm just trying to... I think 2027, Serval aside, 2027 will be a year where we find out a lot of vertical AI plays are nifty little startups but didn't in the end expand TAM enough. They didn't expand TAM. They, they're, they're fighting for a slice of ServiceNow, a slice of Salesforce, a slice of Atlassian, but they're not expanding the TAM, and I just, personally, I don't want to invest in any of them because I just view it as tiers in five years.

Rory O’Driscoll 49:05 I, I don't think Serv- I, I don't think that's the case with Serval. I think they are expanding the TAM, and I don't think, I, they, I'm sure they'd push back very vigorously on the vertical description. I mean, I think it's very cross-functional and cross-vertical.

Jason Lemkin 49:15 No, no, I'm just, I, uh, it's an adjacent point I was trying to make, but I hear you.

Rory O’Driscoll 49:18 Okay. Sorry.

Jason Lemkin 49:18 Yeah.

Rory O’Driscoll 49:18 I'm not, I'm not arguing.

Jason Lemkin 49:19 I think in vertical we're gonna see the biggest tiers 'cause a lot of folks are gonna make investments assuming that, especially for like an SMB or an, or an ARVER, you can charge three to four times the size, and we're gonna find out, you know what, you can charge 1X the size. Um, and now there's five or eight times the competitors in the market, and the leader isn't dead, right? The leader isn't... There's still a lo- they still have a pulse. They may not be growing at the rate before, and so you're just slicing up the same old F and TAM 11 times. I just don't wanna personally make those bets 'cause it's exhausting, right? It's, it's exhausting. Um-

Rory O’Driscoll 49:50 And then it's been funny, we've, we... As I said, that's why it's such an interesting... It's a really genuinely interesting question. I'm, like, literally thinking of it as an investor here rather than just blathering on. It's like we've kind of gone between the 80 billion one of

Rory O’Driscoll 50:00 Cognition and Serval and kind of Baseten in the middle. I mean, if you want revenue velocity right now, um, nothing gives you revenue velocity like Inference. But from your little face, Harry, are, are you implying the angst there is just competition? There's five or six great companies in that space and that's a worry?

Harry Stebbings 50:16 I would 100% invest in Baseten too, by the way. Um-

Rory O’Driscoll 50:19 Yeah, but the whole point is-

Harry Stebbings 50:20 I think, I know, but, but it's not-

Rory O’Driscoll 50:21 ... there's no, there is no two. You gotta pick.

Harry Stebbings 50:24 But yeah, but no, I would not choose it, um, compared-

Rory O’Driscoll 50:26 Okay

Harry Stebbings 50:26 ... to Serval or Cognition. And I think Modal's product suite is unbelievably strong. Like, Arif-

Rory O’Driscoll 50:32 Okay

Harry Stebbings 50:32 ... is incredible for that, and, and Lynn and the team at Fireworks, uh, by far I think have the best inference product. Um, and so I, I would choose Mo- you know, Modal and Fireworks, [laughs] and so it's super clear for me on Serval.

Rory O’Driscoll 50:44 Rory. Okay. That was a fun question.

Harry Stebbings 50:45 Rory, did I change your mind? [laughs]

Rory O’Driscoll 50:48 You made me think. No, a genuine, a genuine comment. This is like, this made me think.

Harry Stebbings 50:52 Mm-hmm.

Rory O’Driscoll 50:53 Right? These are three very different interesting dynamic... I mean, it, it, it, it makes you think about, you know, what's your bet? Is it very horizontal inference but lots of competition, risk of commoditization? Is it the last surviving standalone coding agent that's cracked the enterprise but at a very hefty price? Or is it an app that I think will be the w- If there is a next generation ServiceNow, the probability of it being Serval is probably 80, 70% plus, right? And those are very different bets and, you know, I'm, I'm actually literally thinking, they're all, for the record, pretty interesting bets. The one nagging comment, and I don't know if this is correct, is, you know, the good thing about Serval is you can say, this is your point, Harry, is that the old school company is worth, I can't remember, I used to know the number, 40 or 50. Yeah.

Harry Stebbings 51:39 It's worth 110.

Rory O’Driscoll 51:39 It may be more than that now.

Harry Stebbings 51:40 It's worth 110 billion.

Rory O’Driscoll 51:42 So you can say there's a thing that exists today that's 30 times the value of this, therefore I can have that money. Whereas for Baseten and Cog- for Cognition, you have to say there's no standalone coding agent worth 80 billion. So it's, good news, it's a new market. Bad news is what are you mentally taking over? You know what I mean? And the same is true of Inference.

Harry Stebbings 52:00 No, no-

Rory O’Driscoll 52:00 You know?

Harry Stebbings 52:00 ... bad news is, uh, you're fourth in the market. You've got Codex, you've got Claw code, you've got Cursor, and then you've got them. Geez. Uh, don't get me wrong, I'm, I'm all for independence. Uh, you know, I'm an investor in Factory as well, and I totally buy that, but oof, that versus, like, Serval versus ServiceNow.

Rory O’Driscoll 52:20 Yeah.

Harry Stebbings 52:20 Bit of an easier one.

Rory O’Driscoll 52:21 Hmm. Okay.

Harry Stebbings 52:24 Wild. Uh, Rory, you'll be surprised to hear that my girlfriend is still with me. I know. [laughs]

Rory O’Driscoll 52:29 Hello.

Harry Stebbings 52:30 My mother is-

Rory O’Driscoll 52:30 She's still-

Harry Stebbings 52:30 My mother is shocked too.

Rory O’Driscoll 52:32 Yeah.

Harry Stebbings 52:32 But in the legal world, whoa, mega news. One of the big law firms, the prestigious, uh, OG brands, Weil, uh, shut down very quickly and weirdly because of partner dispute, whatever. AI law firm Crosby, who've raised 85 million bucks from Sequoia and the like, bid to take over its assets and its, like, its brand. Anything to say here? I thought that was fascinating.

Rory O’Driscoll 53:01 I mean, it's two separate stories. There's the Weil Gotshal crashing, which is a, you know, American law... It's not a tech story, it's an American law story. You lose a couple of key partners. Um, it just, you know, these, these firms are, even though they don't have a lot of liabilities other than leases, it's just a bunch of connected rainmakers, and sometimes the tissue, the, the, the-

Harry Stebbings 53:21 Uh, that's a, that's a flim- no offense, Rory, that's a very flimsy way of looking at it.

Rory O’Driscoll 53:24 How do you mean?

Harry Stebbings 53:24 A couple of partners leave does not mean a 90-year prestigious law firm-

Jason Lemkin 53:28 Yeah, but if you take enough of the book of business with you, it means.

Rory O’Driscoll 53:31 It's a small number of rainmakers. Absolutely. It's not like a firm-

Jason Lemkin 53:33 It's worse than if you leave, if you leave a VC firm, the VC firm dies but over 20 years. If the top rainmakers leave, you, you know, there's no, there's nothing to pay out at the end of, at the end of the year.

Rory O’Driscoll 53:42 I mean, if you look through it, I mean, I read all the accounts. I don't know as well. It's, as I say, it's four or five key lawyers, each of whom ran a key part of the business, and it, it just, you know, you, you... There seems to have been a clash on a couple of them. I think the guy who ran all the PE practice, which is probably 20, 30% of a company like this, left, takes his book of business, and yeah, it can go wrong pretty quickly. So that, but it doesn't matter 'cause that's not a tech business. That's just a wow, that's interesting. And then I did a- I did admire Crosby for just, that's just a, "Hey, if there's some, if there's some attention to be paid here, why don't I put in a bid for the business?" That was a bit of a fuck you, but I was, I was impressively, um, chutzpah, I think is the relevant word there for that. That was chutzpah.

Harry Stebbings 54:18 I loved it. Ryan just like, "Yeah, we'll take it." [laughs]

Rory O’Driscoll 54:21 Yeah, if you guys don't wa- yeah, if you don't want this 100-year-old franchise, we'll take it here.

Harry Stebbings 54:26 Nice. Gasp.

Jason Lemkin 54:26 I just, I d- listen, I'm not, I don't mean to be too negative on the Crosby thing, but u- u- ultimately, if you're providing Slightly higher value service is gonna have the same issues. Pe- people wanna keep all the mo- the partners, the rainmakers wanna keep all the money. Now, Crosby has, probably has no rainmakers, right? Um, and it's doing transactional stuff, but I, I don't know that this idea that lawyers will give up a bunch of the profits they could make for shares in a company that hopefully will IPO, I, I don't know how sustainable the business model is.

Rory O’Driscoll 54:55 For the record, I, I don't know if I fully buy the model, but I disagree with you in the sense that I actually think... It's actually the same story. What Weil Gotshal says is there are some parts of law where the partner that you are dealing with as the customer is so important because you value his or her personal advice, right, that when that partner leaves Weil and goes to someone else, you take your business with them, right? And that's like, "I'm doing a complex PE deal. I need this guy. I'm paying him five grand an hour," some absurd sum, "but he's my guy and I trust him and I'm, you know, running KK&R, and that's the guy I want." And then what Crosby is saying is there's a whole bunch of other law where you really don't care the name behind the thing 'cause it's pretty standard and transactional, and that business can be done with Crosby Law. Well, I actually think th- they're actually very nice counterpoint, right? If I'm doing an NDA or a Series B stock purchase agreement, I don't need the five grand an hour guy. I don't need to know the name of the guy doing it, right? It's just, you know, do the usual stuff, check the usual things. So I think-

Jason Lemkin 55:53 Yes, but, but, but even at the low end, incompetence hurts you.

Rory O’Driscoll 55:57 Yeah, but incompet- yeah, okay

Jason Lemkin 55:58 It damages you on-

Rory O’Driscoll 55:58 Okay, but ag- agreed, but I don't-

Jason Lemkin 55:59 Have you ever done a routine transaction where some idiot lawyer cost you a lot of money? I have.

Rory O’Driscoll 56:03 Yeah.

Jason Lemkin 56:04 I have. I've been damaged seven figures by incompetent junior lawyers.

Rory O’Driscoll 56:08 But I think... I'm sorry, I'm just gonna push again 'cause I'm grumpy. Like, I think that that's exactly the Crosby argument, which is it's less likely to have that kind of mistake if it's automated, if you've got AI, like, for the standard stuff. So I, I, again, I, I think that the-

Jason Lemkin 56:21 Okay, well, let me flip it around. Let's imagine I go to Crosby and I make $170,000 a year and I get .01% of the company for the IPO that I'm not even sure-

Rory O’Driscoll 56:29 Yeah

Jason Lemkin 56:29 ... is gonna happen.

Rory O’Driscoll 56:30 Sure.

Jason Lemkin 56:30 Right? And then, uh, I can go to Weil Prime and make 250.

Rory O’Driscoll 56:35 Yeah, but I can answer that-

Jason Lemkin 56:36 And I, and if I'm actually good, I ma- every great partner now makes 10 million bucks, and every good... And the entry-level ticket at Weil was over 200,000 for an associate, for a soul-crushing, horrific job, mind you. Like, I get the Crosby... I just, my only point is, and I'm, I'm not being critical. I haven't talked to them. I'm not convinced this is, is as stable as a lot of VCs might hope it would be. It's too human based.

Rory O’Driscoll 56:58 I agree there are issues with the kind of human plus AI delivery model, and I have questions on that, but I'm just gonna push back on your comment on why would I go to Cro- like, uh, it doesn't... That's easy because for every lawyer who wants to be a rainmaker and work 24/7 and make $10 million, there's also lots of lawyers that's like, "I wanna do 40 hours of good law. I don't wanna be a rainmaker. I don't like that. I'm just a good lawyer, and I wanna go home. Maybe I'm, you know, dealing with kids. Maybe I'm dealing with a parent. Maybe I just wanna take it easy," right? There are lots of people for whom being a lawyer, a well-compensated, nice upper middle class salary, working for someone like Crosby, doing 40 hours a week with control of your own time, is a perfectly good product. So I don't have a pro- they'll be able to get lawyers, right?

Jason Lemkin 57:41 I just don't know if I can see a $10 billion outcome here.

Rory O’Driscoll 57:44 Different question. I'm sorry, I'm just keep- yeah.

Jason Lemkin 57:45 For the... I don't think I can take this to IC. It's a, it's a lot that has to happen-

Rory O’Driscoll 57:49 Agreed

Jason Lemkin 57:49 ... for the corporate company, its corporate affiliate, to be worth $10 billion, 'cause it has to suck so many billions of years of profits that the lawyers could otherwise take home as partners.

Rory O’Driscoll 57:59 Please, that I agree with. I think that there will be a coterie of law firms five years from now that are focused on the Crosby kind of work. Crosby will obviously be one of them. They'll do really well. We like the space. But I agree, the question is, if you look at the billings of Weil Gotshal as an example of American law, you know, if 70% of the dollars come from the high rainmaker, irreplaceable by AI stuff, and all you've got left is the 30%, and then you have to have the value of that to get the business, then you're right, then you're left with the low margin parts of the business. So I do agree. That's the question, is how big can it be? But whether you can build a business there? Absolutely, I think you can, is the question I'd make.

Jason Lemkin 58:36 Yeah, you can build a small business, for sure. It's just for our IC, 10 billion is the, is the lowest possible exit where we're allowed to bring it to IC.

Rory O’Driscoll 58:43 Agreed. As far as I can see, 10 billion is-

Jason Lemkin 58:45 It's 2026, boys.

Rory O’Driscoll 58:46 Yeah.

Harry Stebbings 58:46 10 billion? I thought 10, is 10 billion the round size?

Jason Lemkin 58:49 Look, Rory and I are a little more conservative. We're okay with a $10 billion exit as long as the entry is appropriate. [laughs] We can't make the numbers for the whole fund work, but we can make our own little slice of the fund, we can, we can hit our, our individual 6X goals with those

Jason Lemkin 59:02 exits.

Harry Stebbings 59:02 Guys, what, what have I missed? Creandum, I think probably the most exciting fund that Europe's produced in the last three years, raises 750. Eh. [scoffs] Um, Physical Intelligence, uh, in talks to raise at a $20 billion valuation. They've had rumored acquisition talks in the last six months as well. Um, that's with Avenir and Lux. Dropbox, uh, cuts 9% of staff and CTO leaves.

Rory O’Driscoll 59:28 Unlike the last group of three, these things are, like, literally a grab bag of shit. Physical Intelligence, company with relatively little revenue, it's worth 20 billion. Dropbox, company with billions of revenue, ain't worth a lot. There. Um, it just shows revenue multiples don't get you anywhere, right? It's all about the future, right? Th- so yeah, these aren't, these are just non-comparable things. I mean, look, on the Physical Intelligence thing, yeah, I think robotics is going through an amazing explosion. We have a small investment in Ge- like, I think these next generation models are amazing when you see the demos. I've done a lot in robotics over 10 plus years. When you see what the new models can do, it is truly amazing, right? You go, "Wow, the level of dexterity, the old models just don't have that," right? And I think it's just a real breakthrough moment. So two other comments, though. I think deployment of robotics is just way longer than deployment of kind of classic LLMs, and people forget that. I saw someone... Yeah, somebody's coming and starting to price now on an hourly API basis, which makes sense. So you pay, you know, if you're replacing, maybe you pay three bucks an hour for access to the model, and then when you... There's only 300,000 robots in the US, so to make the math work at kind of 20 billion, which is where PI is, you gotta have a shit ton more robots. So, uh- The technology's amazing, the business model's got a lot of developing to do, but I think the last sentence you said is the real one. I think all the big foundation models company buy one of these things. But... And I think the reason is, I mean, e- either they build one or buy one. It's gonna sound... I think I've said this before, the really simplistic reason they do it is this, I'm writing my prospectus, there are 81 million knowledge workers in the US, we have an LLM, and we sell to them. It's a great business. There are 87 physical workers in America, not knowledge workers, working in blue collar or kind of, you know, physical world jobs. Now we have a robotics foundation model, we can address them, I've doubled my TAM. So, I mean, it's a bit simplistic, but I think the lure of robotics for the foundation model companies, and the desire to say, "Oh, we do robotics, too," I think will... means that there could well be some acquisitions in the space. It's one of those markets, it g- goes back to something Jason said, it's one of those markets where you can kind of say to yourself, if you just try and be logical about the steps between here and deployment, you probably get caught up in the minutia, and I'm often guilty of that, where, you know, it's gonna take a long time, customers are gonna be slow, BMW's not gonna roll out 100, um, humanoids in a week. But if you say to yourself, "The technology capability is amazing, the teams are amazing, we love our position in generalist," and you kind of go, "This is gonna change the world," somehow you'll make money out of it, and that's probably the investment thesis.

Harry Stebbings 1:01:57 Jason, any final thoughts to you on this?

Jason Lemkin 1:02:00 The only small... This one's really minor, it's almost not worth, uh, bringing up other than it's something I'm tracking. But a, a tin- a tiny amount of anxiety that the Hugging Face has drawn some DOJ scrutiny. I don't think there's been a formal request for information or anything, and, um, it just, as investors it would be b- b- a bummer for us if M&A isn't really back, right? Um, as we discussed, we need $10 billion deals to even bring deals to, to, to the IC. I- if this $12.9 billion deal for Hugging Face isn't gonna sail through antitrust, it w- it would just, it would just be a bummer, it would be a, a... Politics aside, um, it just seemed like we were able to do normal M&A deals, right? Uh, with normal tax benefits and normal this and that. If, if it, if it doesn't turn out to be the case, um, then we've got to go back to doing convoluted team and IP deals or slow down deals, it'd be a bummer. And, and, and look at the other f- I mean, Salesforce is on an epic M&A roll. I mean, they j- they b- they've Hoovered up everybody this year, and, and I don't know that they would do that if the antitrust environment, n- again, no politics aside, if the antitrust environment was where it was 24 months ago. It's good for venture that they're just Hoover-

Harry Stebbings 1:03:04 Yes

Jason Lemkin 1:03:04 ... I mean, who did they just buy this week? A- Accresta, right?

Harry Stebbings 1:03:07 They just bought Accresta for 2.7 billion.

Jason Lemkin 1:03:09 Yeah, bless, bless Benioff and friends, right? We need these deals to... They happen in every two weeks, right? Um, and w- and we n- for, uh, for venture to be, like, a- at least to be de-stressful in my opinion, we need M&A to work like clockwork, otherwise we're just, we're just stressed over these d- goddamn IPOs. So I just worry a little bit about this thing, 'cause we want, we want the Salesforces and the ServiceNows and Micros- uh, we just want them buying companies, uh, you know, e- every, every couple of days for a billion and up. We need it. We need it to get out of our, get, to get to, to, to liquid our portfolios. So this one, if something happened here I'd be bummed, but it's just something to track if it gets anywhere. Salesforce says the opposite, they don't... They're game on for M&A, man.

Harry Stebbings 1:03:50 Dude, I agree. Um-

Jason Lemkin 1:03:51 Bless the team there.

Harry Stebbings 1:03:52 [laughs] Yeah. Good

Harry Stebbings 1:03:55 old Mark.

Jason Lemkin 1:03:55 It's smart, too. You gotta... He's buying his way into the, into the future, into the present. It's... There's some point where it's too much-

Harry Stebbings 1:04:02 I- is he, is he with Cr- is he with Accresta? Like...

Jason Lemkin 1:04:05 I think what, what's hap- if you... I think what's happening if you just look at all the deals that have been bought, this, this, this I saw, I'm dating myself back at Adobe, but ev- right, obviously there's a little bit changing in the world with AI and B2B, right? So basically everybody's getting a punch card, right? They bought a, some defense startup, I don't even know, that, for a couple hundred million bucks, right? So s- w- you know, and, and what you do is you go around the table with all your SVPs and they get their best idea. And whoever was in service or support or whatever, C- V- Accresta was their best idea, and whoever was trying to get, uh, um, these sort of, um, uh, uh, support and sales agents faster, Fin was their best idea, right? And you just keep... It's not that complicated. You go down the best idea, and then someone goes out and figures out if they'll actually sell, 'cause some folks won't sell, and you just make this best idea and then people know their jobs are on the line, and so this... None of these deals are Mark betting the future of the company, they're, they're SVP chips. And I've been in the room when they, when you went around the room, "Rory, what's your best idea? Harry, what's your best idea? Come back next week, I want your top three ideas." And it costs you nothing in a sense, because it's assumed some will fail, right? But basically they're getting around the table and they're saying, "We're gonna spend this much of our market cap and cash to do these deals 'cause we gotta do, we gotta do something."

Rory O’Driscoll 1:05:14 And the interesting question is do you do Accresta, you know, like, you know, a m- a more mature company that's using AI, but probably the original stack was pre-LLM and they've probably transposed their stack once the LLMs come out in the last four or five years. But in return for that they have some scale, they have some kind of enterprise deployments. Or do you buy something new, bright, and shiny where it's all LLM first, it's two guys from YC in '24 and, you know, they've got to $10 million and they're exploding, right? And that is the interesting thing.

Harry Stebbings 1:05:43 Well, well, the interesting thing for me is that you've got a relatively similar price being paid for them versus Listen, but you're gonna have a 10X increase in revenue for Accresta versus Listen. Listen were 20 million, sold for 2 billion. Uh, Accresta are well over 200 million in ARR.

Rory O’Driscoll 1:05:59 And given that Sequoia's in both, it's not like they missed the fact that they're getting a different revenue multiple. So yeah, on the other side of the table, you're right, the, the seller said, "Yeah, I get why I, I want a high revenue multiple for Listen 'cause it's got growth and it's new, and I guess why I'm getting a still good but not as, you know, a stellar revenue model for Accresta because it's a mat- more mature company," and yeah. No, the, you're right, it is interesting. It, it's the buyer deciding which bet do I wanna take. And then thank you for reminding me, Listen, that's a very good example of that. Do I want the ultra modern deal? But the problem is I'm gonna have to pay a scary revenue multiple in terms of I'm tr- I'm tra- when you're trading at 7X, to be fair, it's really hard to spend a lot of money at 40X. You can spend maybe a couple of billion, but it, it gets scary.

Harry Stebbings 1:06:43 I think the Accresta deal's actually pretty good. You're paying 10, you're paying 10X.

Rory O’Driscoll 1:06:46 I agree. No, I, I, it, it, I agree. Exactly like, yeah, I think the Intercom deal. I think they're smart deals for Salesforce. You get critical mass, you move the ball forward. Maybe it's not as bright and shiny as the newest bright and shiny thing, but it, it, you know, your service cloud business is your second largest, I think largest business now, and you just make progress on that. I agree.

Jason Lemkin 1:07:04 What you'll find if you ta- talk to a lot of the big, big players, what you'll find is if they want to do an M&A where they want to put their resources against it, their GTM machine, their platform, they, they'll all tell you this, they want at least 100 million of revenue. They don't even... And they'll wait, and they'll take the meetings, and they'll get excited. And the, what's the one, Listen Labs, right? That's a tech bet, and, and most folks will tell you those are the dumbest bets, right? Either buy a small team and pay less or buy something with nine figures. You only do a Listen Labs deal in times of stress, because they'll all tell you that's their lowest, uh, uh, uh, investment performance. That's what it says. But it's just, it's a reason, if, if you're confused as a founder why deals happen, a lot of them, they, they just decide if it's not 100 million, it doesn't have scale. Like, I can't put my guys on it. I can't put my team on it. I, I've heard this from a lot, a lot of teams. It just is what it is, right? Um, and, um, you know, even with Fin, like, their core, their agentic business had to hit 100 million, right? The overall business was larger. It's the same story, and then all of a sudden it's like, "Okay, now I can see how I can make this material in my $50 billion business." [laughs]

Rory O’Driscoll 1:08:07 It's probably true. It's a shame, but it's probably true.

Jason Lemkin 1:08:09 I, I didn't even think... You know, it is a shame, but the Borg otherwise will just, uh, they can't even assimilate you. You're too small.

Harry Stebbings 1:08:15 All right, boys, it's time to go make some money.

Rory O’Driscoll 1:08:17 It is.

Harry Stebbings 1:08:17 Back to our day jobs. Let's do, let's do this.

Rory O’Driscoll 1:08:20 Let's do that. Okay, Harry, thank you very much.