The Exchanges

Q Do you think these deals are priming the pump, so to speak?

A I think all of these deals now are priming the pump. I mean, you basically announce the deal. They're 10 or 20% funded, and then you have to go raise capital to, to, to fund the rest of it. And so, you know, everyone announces these deals in gigawatts, not dollars anymore. And I think most people don't know how many dollars a gigawatt is. And so the rough math is, you know, a gigawatt is forty billion dollars to, to build out. Jensen says it's 50 or 60 if you use, uh, the next generation Vero Rubin chips. So let's say it's somewhere between 40 and sixty billion. So, a hundred gigawatts of power build out, which is what people are talking about now, that's eight, that would be AI's eight trillion dollar question. And then, two 50 gigawatts of power is AI's 20 trillion dollar question. So, we've totally upped the ante, and the magnitude is just, it's just much, much bigger. But of course, that's not funded, um, and so I think the funding for these deals is, is gonna be an important thing that has to play out.

AI assessment note: “I think all of these deals now are priming the pump.”

Q There's so many things I want to unpack within those. I do want to go to what did you not predict or foresee that did play out that you were surprised by?

A I think there were two big misses last year. I think the first big miss was These like big talent acquisitions. I mean, I think that if you'd asked me the probability a year ago that, you know, if you're a 25 year old recent grad from an elite university who is perceived to be an AI expert, you can get a fifty hundred million dollar pay package right now. And if you are a brand name that everyone recognizes your name, you can get a billion dollar pay package right now for a single individual. I totally did not see that coming. And I think that you asked me a year ago to predict that I would have said you were crazy. So sometimes I do think the beauty of AI is like reality is stranger than fiction, and a lot of crazy things happen.

AI assessment note: “I think the first big miss was These like big talent acquisitions.”

Q Do you think these deals are priming the pump, so to speak?

A I think all of these deals now are priming the pump. I mean, you basically announce the deal. They're 10 or 20% funded, and then you have to go raise capital to, to, to fund the rest of it. And so, you know, everyone announces these deals in gigawatts, not dollars anymore. And I think most people don't know how many dollars a gigawatt is. And so the rough math is, you know, a gigawatt is forty billion dollars to, to build out. Jensen says it's 50 or 60 if you use, uh, the next generation Vero Rubin chips. So let's say it's somewhere between 40 and sixty billion. So, a hundred gigawatts of power build out, which is what people are talking about now, that's eight, that would be AI's eight trillion dollar question. And then, two 50 gigawatts of power is AI's 20 trillion dollar question. So, we've totally upped the ante, and the magnitude is just, it's just much, much bigger. But of course, that's not funded, um, and so I think the funding for these deals is, is gonna be an important thing that has to play out.

AI assessment note: “I think all of these deals now are priming the pump.”

Q I completely agree from the employer side. On the flip side, um, when you think about, like, advice to them, if, if you were advising your younger sibling on choosing their first job, I, I saw on LinkedIn you said, follow the smartest people a year ahead of you. That moniker of advice may not be relevant anymore. What advice would you give to them?

A Well, this is like the biggest learning, because I've met with two or 300 young people a year, I have a very big data set, and I think I've probably spent more time than anybody at Sequoia on this specific, you know, thing. And my biggest lesson is that the way that young people choose their career is this, what I call the memetic algorithm. And the memetic algorithm is, yeah, what did the people one year ahead of me in school that I thought were the best, what did they go do? And it's a recursive algorithm, right? So it's like, what did the people a year ahead of me do, but those people chose based on the people a year ahead of them did, and those people chose based on the year ahead of them did. Now, one reaction to that would be negative of like, oh, that's so mimetic, they should think for themselves. I actually don't have that perspective. I'm fine with it. I think it's like a reasonably good algorithm. When I graduated from college, Palantir was the hottest company to go work for. All the really smart people went to go to work for Palantir. Going to work for Palantir would have been a great life decision at that stage. Uh, before that, you know, in the early 2010, Google and the big tech companies were the hot place to go work. And I think, you know, those companies were all 10 X's over the, uh, over the 20 tens. Some of them even, I think, 25 X's. So The, the, it was a g…

AI assessment note: “I don't think the memetic algorithm is inherently broken, and I respect it”

Q Tell me, dude, what's your biggest miss, and what did you not see that you should have seen with the benefit of hindsight?

A One big financial miss is Datadog, and I worked on this before joining Sequoia, but I remember, you know, Datadog was this amazing company. The numbers were incredible. It was proper. Like, it was one of these businesses where you just, like, your mouth waters looking at the financials of that business, and, um, and I remember we lost, uh, and, and, and we lost to Dragoneer, and I never confirmed this with Dragoneer, but the story behind it really stuck with me, which was that Dragoneer had this list of 20 companies, and they only worked on those 20 companies, and they had been spending years and years and years courting Datadog, and, like, this was their number one priority, and they knew it was their number one priority for a few years, and this was probably six years ago now that this happened, but, It's a principle that has really shaped how I pursue new investments, which is, um, if it's not, I want to really focus my time, and I've actually adapted this to, if it's not one of the top five opportunities, that's where I really want to be spending 80% of my time, and then I want to spend the next 20% of my time on the next 15, and then after that, just, like, really trying to focus your time, and so that actually shaped who I became as an investor, and I learned a lot from that.

AI assessment note: “One big financial miss is Datadog, and I worked on this before joining Sequoia”

Q Can I ask you what specifically makes you say you can see the fragility?

A Well, the circular deals, I think the circular deals dynamic is probably, when I think about why did, why did this AI bubble narrative go from contrarian a year ago to consensus today? I think the main thing driving the consensus is these circular deals and the big tech company dynamics. Let me, let me unpack that. A year ago, hyperscalers were holding up the AI ecosystem and everybody felt very comfortable with that because everyone knew that these were very robust businesses. Microsoft and Amazon specifically were driving the vast majority Of the AI CapEx growth. And they were explicitly saying, hey, we're going to buy out your generator capacity for five years. We're going to sign a 20 year lease on this data center and we'll back it up with our credit. So they were basically putting themselves in front of all the risk. And the way I thought about it a year ago and wrote about it a year ago is like they're almost grabbing the hot, demand hot potato and saying like, it's, it's ours. Don't worry about it. We got this covered. A year later, Microsoft and Amazon have really stepped back. And this started And again, the information has done a really nice job reporting on this. They started in the beginning of the year. There was this big public announcement or leak or whatever you want to call it, where Microsoft walked away from two data centers and it sent a message to the mark…

AI assessment note: “the circular deals dynamic is probably, when I think about why did”

Q How does the data center change? You mentioned it changes in real estate moving from, you know, a 100,000 to 300,000. What would one expect to see in that changing landscape?

A Here's what's happening. And I think Basically, no, a lot of people are not paying attention to this right now. Amazon in the last six months has announced fifty billion dollars of new data centers, right? This is, when I talk about AI's six hundred billion dollar question, like, this is the cost line on the six hundred billion dollar question, like, these data centers are getting built. They have to go hire people to go build these things, right? Like, you're gonna go to some town. There's a company called Cyrus One. There's another company called QTS. These companies are the real estate developer for data center. So Microsoft or Amazon goes to this real estate developer and he says, hey, I want you to build me this data center. And the real estate developer goes to this company called DPR. DPR is the biggest general contractor in building data centers. And the DPR is like, Hey, I need you to build me this data center. DPR goes and finds a subcontractor. The subcontractor needs to go find you a thousand electricians and all these people. I mean, labor is the single biggest cost in data centers. And now you're putting ads on Facebook and you're like electricians needed. Please come to this random town in the middle of Illinois. We need you to come build a data center, right? And then you have all these people who, like, they're getting on planes, they're gonna be put up in hote…

AI assessment note: “just the sheer physicality of what's gonna happen in the next 12 months”

Q and he said, really, there's like three things. There's compute, which we've discussed, there's, uh, algorithms, and then there's data. He said, I take the alternative to a lot of people. I think data is the core bottleneck on model and AI progression today. To what extent do you agree with him? And when you look at compute algorithms and data, which one do you think is the core constraint?

A It's funny. I used to kind of agree with Alex, and I think that my mental model has actually shifted. I kind of think compute models and data have kind of converged, like all the big guys know what they're doing. They're all doing the same thing. Everybody's using scale. Now I love Alex. Alex is fantastic CEO. He's doing a great job, right? Like he is kind of this arms dealer. Everybody has to buy from him. He's providing all these big, big model companies with data. But I think it's really hard to argue today that any of the big model companies has a data advantage. Compute, it's just a commodity that you pay for. So it's hard to argue that any of them has a compute advantage. And then models, they would all argue that they have some secret sauce. But again, if you believe the bitter lesson and you believe that scaling laws are the things that matter, then the secret sauce is like not that material. And so I'll propose my own three things that I think are the three things that matter. And I would summarize it as servers, steel, and power. So I'm just much more interested in the industrial nature of AI and like what is happening is this industrial revolution. And so servers, that's NVIDIA, AMD, Broadcom, like the chip innovation. NVIDIA has an amazing gross margin. There's going to be a ton of competition. The chip wars are just getting started. So there's a lot of interesting …

AI assessment note: “I used to kind of agree with Alex, and I think that my mental model has actually shifted.”

Q chatting to Pat before about Bluntly, you and the venture career that you've had so far, and he mentioned you're being 27 and being the co-head of, you know, the venture division at Co-Two, and you worked with Databricks, UiPath, Snowflake, some of the best businesses of our time, and the first thing I really thought was, how did that and those companies impact and teach you about deal selection?

A I mean, first, just to calibrate, I was the associate on those deals, so I built the models, I did the customer calls, but I was not the lead investor on those deals. What I will say, and just because you picked those companies, I'll give you kind of a through line on those four, which is, I think the lesson from those companies is listen to what people do, not what people say. And I'll tell you why. When I called the customers on Marketa, you call DoorDash, Instacart, Square, those were kind of the big customers. And you ask them, do you guys like Marketa? They would say, it's fantastic, but it's too expensive. We're going to rip it out. We're going to build this ourselves. That's what everyone said when you called them. Now, fast forward five years later, Marketa is a public company. None of those companies have churned. It continues to do super well. UiPath, same thing. You call all the customers. They say, oh, it's a bandaid solution. We're just using it for now. Eventually we're going to fully automate. Again, fast forward five, six years since that investment company went public. Fantastic company. People continue to use it. Even Snowflake and Databricks, same thing. You call customers in 2018 and you ask them, what do you think about Databricks? They'd say, oh, well, Amazon has their own competitor. Google has their own competitor. Snowflake. Oh, there's BigQuery. There'…

AI assessment note: “the lesson from those companies is listen to what people do, not what people say”

Q podcast. Cause they're like, Hey, we're not meeting investors. I'm like, I agree. They're all twacks. Uh, I'm, I'm a media person. I'm just a podcaster. And then whenever it comes something else, I'm like, no, I'm an investor. It's amazing how you can switch hats. Um, you mentioned kind of, they don't want to spend time with investors. What's the craziest thing you've done to win a deal, David?

A Look, I think you probably actually maybe can relate to this. Like in the beginning, it's super hard and nobody wants to give you the time of day. They ignore you. They don't respond to you. And so I think in the beginning it was about how do I just get people's attention? Once I get in front of people, then I think you can make your case, et cetera. But I think getting in front of people is really hard. And so one thing I used to do is I went through this phase where every time I met a founder that I really liked, I would go through their Twitter. I'd figure out something they really liked, let's say some TV show, and then I'd go in cameo and I would find the actress or actor that they like really liked, and I would get that person to record a video for them. And I would send them that video. In hindsight, it's kind of ridiculous. It was kind of a ridiculous thing to do. And yet it was shockingly effective. Like I will tell you. What is that ridiculous?

AI assessment note: “I'd go in cameo and I would find the actress or actor that they like”

Q Okay. You're an early morning riser. Pat always tells me, like, I'm good in the morning. David beats me. Can you just tell me, like, the morning routine and how you organize yourself so efficiently to smash the day early?

A I love the mornings. I just probably, like, naturally, just biologically always like the mornings. I wake up at five a.m. I'm at the gym at five 30. I bike to work every day, and so I bike from the gym up Through Stanford and then up Sand Hill Road. And I'll tell you, that's like such an inspiring experience. Like you bike through Stanford. It's like, this is the history. This place is built around Stanford. All this great technology has come out of here. You kind of bike up Sand Hill Road, the mountains in the background. The sun is coming out and you just feel like, man, like, like this is going to be awesome. There's so much history here. Sand Hill Road is so phenomenal. I'm going to participate in this thing. That's sort of been going for 50 years. Hopefully we're going for many, many more years. And you get to the office and you're super fired up. And I try to meet a founder in the morning, you know, like I want to come in fired up and meet with me with a great founder and talk to them about their business. And it's just wonderful.

AI assessment note: “I wake up at five a.m. I'm at the gym at five 30.”

Q Is there an alternative argument that it allows for the continued concentrating of power, the continued oligopolies becoming even stronger, that we must consider as well?

A I think there is, right? I think that's a super valid perspective. I think the perspective, what you would, what you would argue is these companies are too powerful. They're basically erecting barriers to entry, right? Now, in order to be an AI cloud, you need to be willing to light a bunch of money on fire, right? And so I think that that is a pretty significant barrier to entry for new entrants. And by the way, I think what you heard from the commentary this week is that's not an accident, right? That's pretty explicitly what they're trying to do is say, Hey, we cannot afford to let anybody else, you know, attack our golden goose. This cloud business today is a two hundred and fifty billion dollar business. So the cloud business today is the same size as the SaaS sector. That's the business that Azure, Google and, uh, and, uh, and AWS control is the same size as the entire SaaS sector. So of course they're going to do everything they can to protect it.

AI assessment note: “I think there is, right? I think that's a super valid perspective.”

Q and he said, really, there's like three things. There's compute, which we've discussed, there's, uh, algorithms, and then there's data. He said, I take the alternative to a lot of people. I think data is the core bottleneck on model and AI progression today. To what extent do you agree with him? And when you look at compute algorithms and data, which one do you think is the core constraint?

A It's funny. I used to kind of agree with Alex, and I think that my mental model has actually shifted. I kind of think compute models and data have kind of converged, like all the big guys know what they're doing. They're all doing the same thing. Everybody's using scale. Now I love Alex. Alex is fantastic CEO. He's doing a great job, right? Like he is kind of this arms dealer. Everybody has to buy from him. He's providing all these big, big model companies with data. But I think it's really hard to argue today that any of the big model companies has a data advantage. Compute, it's just a commodity that you pay for. So it's hard to argue that any of them has a compute advantage. And then models, they would all argue that they have some secret sauce. But again, if you believe the bitter lesson and you believe that scaling laws are the things that matter, then the secret sauce is like not that material. And so I'll propose my own three things that I think are the three things that matter. And I would summarize it as servers, steel, and power. So I'm just much more interested in the industrial nature of AI and like what is happening is this industrial revolution. And so servers, that's NVIDIA, AMD, Broadcom, like the chip innovation. NVIDIA has an amazing gross margin. There's going to be a ton of competition. The chip wars are just getting started. So there's a lot of interesting …

AI assessment note: “I kind of think compute models and data have kind of converged”

Q Now, the year of the data center sounds wonderful. We had An amazing discussion last year. What did you predict last year, David, that happened and we are seeing in action now?

A I think there's really, so we talked about last year, this concept of steel servers and power. And I think if you remember, you know, rewind to summer, 20, 24, the big conversation at that time was compute models and data. That's what everybody was talking about. And I sort of had this view that everyone was underestimating the physicality of these data centers. I'm on the front lines. I'm talking to people every day. And you, you know, you talk to people, they're flying electricians to Texas and they're trying to buy out generator capacity and, you know, generators are sold out until 2030 and, and so how do you get in line and how do you do that? And so I sort of had this sense that people were thinking very abstractly sort of in a, in a bits perspective about AI, but they should be thinking in an Adam's perspective about AI. And I think that prediction came true in two ways. Uh, the first way is the best trade of 2025 was the AI power trade. A lot of Wall Street people made a lot of money betting on the fact that power was going to be the constraint and we're going to move away. You know, you hear Sam Altman now talking about gigawatts every day. He's not talking about dollars anymore, right? So we're moving away from dollars and we're moving toward gigawatts, and I think that transition has fully happened in the last year. The second way I think it was right, and I saw, you …

AI assessment note: “that prediction came true in two ways. Uh, the first way is the best trade”

Q Why do you think defense is the next AI? Sorry.

A So I think that, you know, it's funny because I started investing. I, as we were talking about a year after the transformer paper in, in 2018. And, um, you know, I think that it's sort of defense reminds me in some ways of like a few years after the transformer paper, which is to say people who are really paying attention understand that I, that defense is, is going to change. And the transformer moment was the, was the Ukraine war. It was a very odd, you know, before that you had to be a visionary and, and to Palmer's credit and, and Peter Thiel's credit and people like this, like they were visionaries before the transformer paper, you're a visionary. And, you know, Ilya, Andre Karpathy, these people are visionaries. After the transformer paper, you're an early adopter, right? And I think our job as investors is to be early adopters for the most part, especially in the growth business, to be early adopters. And so you see that, you see the change that happened in Ukraine. And, um, and I think it was very obvious that like, you know, warfare, you see these pictures, these tanks, you know, and these like long chains of tanks from Russia. And it's like, wow, like defense technology is 50 years old and technology has moved so much in 50 years. And yet, like, the way that we do war just hasn't changed, and that's because, um, you know, we've been in this period of golden era for th…

AI assessment note: “the Transformer paper moment was the, was the Ukraine war”

Q The second thing I do before we move to the second, do you think those scaled pay packages are justified?

A I think they're symbolic of this sort of desperation in the ecosystem where it's like, we need to eke out progress. We need to prove that all these investments are worth it. And I think there's this logic that gets really abused in the venture world and in the tech world, which is like, hey, if I increase the probability of making a trillion dollars by one percent, that's worth Ton of money, right? That's worth ten billion dollars. And sure, that's true, but it's very easy to overestimate the one percent. Is it one percent? Is it a 100th of one percent? Is it a 1000th of one percent? Is it a 10,000th of one percent? Our brains were very bad at reasoning about that scale of number. And so I think to the extent that you believe that hiring this very impressive researcher increases the probability you win by one percent, I totally can see why you would justify a billion dollar pay package for an individual. That said, I think we are psychologically biased to overestimate what that percent contribution is. And it may be the case that there's these broader macro variables, which we'll talk about, I'm sure later in this discussion, there's these broader macro variables that are actually driving progress in AI that are, uh, that are not a single individual can change.

AI assessment note: “I think we are psychologically biased to overestimate what that percent contribution is.”

Q I spoke to Sonia on your team beforehand, and she gave me a fantastic question. She said, if this is a game theoretic bubble, is there a coordinating mechanism for the spending to stop and the bubble to pop?

A You know, I love game theory. So, I mean, my, my basic framework on AI, and this is actually kind of how I write all these pieces, is there's like 10 players around this big chessboard, and they're extremely powerful, and each of their moves affects the other people's moves, so it's kind of recursive. And, and so you sort of have to think first order, second order, third order, how to, how does my move affect other people's moves? And these are very sophisticated players doing this. And so, one, what my, the simple answer to your question is, it's, it's not coordinated. That's the beauty of the invisible hand. That's the beauty of people's incentives. These are big companies that are acting out these incentives. And so I think until the incentives change, the behavior is not going to change. And so there is no coordinating mechanism. I do think that's one of the surprise. It's always the surprising fact of capitalism. Like everyone wants to believe that everything is kind of coordinated. It's easier for our brains to grok everything being coordinated, but I actually think it's, it's pretty uncoordinated and incentive driven.

AI assessment note: “the simple answer to your question is, it's, it's not coordinated.”

Q I am like, mom, you should have done better. Bad parenting. Uh, you encouraged me to do English, really? Um, come on. Um, yeah, war and peace doesn't quite make it, does it, when you're getting paid three and a half billion by Zuck. Uh, what was the second?

A I think the second one, you know, one thing we talked about in the podcast last year, I predicted that Meta was going to do really well. And I think that prediction was clearly false in a 12 month time horizon. Um, I thought that the vertical integration that Meta had was going to be an advantage. And I think that Meta, you know, these hundred million packages are coming in large part from Meta because they haven't performed as well as they thought they were going to. The reason I thought Meta would do well is that it was vertically integrated and founder run. And I, I sort of continue to believe that in the fullness of time, it is possible, and I think the dramatic actions that Zuck is taking represent this, it is possible that I will be proven right in a longer time horizon, which is to say that Zuck's gonna fix the problem, it's amazing what founders can do, he's so focused on this, he's spending all of his time on it, but I think if you look back a year ago at the prediction that Meta would do well, I think you would say, Wrong.

AI assessment note: “I think the second one... I predicted that Meta was going to do really well.”

Q So David, when we play out your question there of the winners and the losers, just so I understand that, who do you think the winners and the losers will be when we look back on this last 12 to 18 months?

A I've had a very simple framework for this. It's actually, I think, probably the first thing I ever published in AI, in AI's two hundred billion dollar question way back when. 20, 23. And, um, the framework is this consumers of compute benefit from a bubble because if we overproduce compute prices go down, your cogs goes down and your gross margin goes up. So I've had the view that you want to invest in consumers of compute producers of compute. Imagine you're producing any commodity asset. If other people produce a lot of that commodity asset, it doesn't matter. It has nothing to do with you. You might be running the best operation possible. You might be an amazing business person. But if everybody else starts producing the same commodity asset, prices go down, and so it's very hard to control your destiny in commodity businesses. By the way, this is why commodity businesses tend to trade cyclically and tend to trade at lower multiples than non-commodity businesses. So I think if you're a producer of compute, you're fundamentally in a commodity business, just like an oil company is in a commodity business, and that is going to trade a different way, and that is going to have more cyclicality than, than if you're in a non-commodity business, Consuming the commodity, consuming the energy, and producing intelligence on top of that, and so I think if you're consuming this raw resou…

AI assessment note: “consumers of compute benefit from a bubble because if we overproduce compute prices go down”

Q mentioned earlier in the conversation, and we mentioned that the concentration of value of mag seven, a lot of that's predicated around the belief that it will impact GDP GDP meaningfully, and we touched on it earlier. Massa said that he thinks that we'll see five percent GDP impact. How do you think about and respond to the magnitude of which we will see AI impact GDP and productivity levels?

A So I think Massa makes an interesting point here, and I actually agree with him fundamentally that AI is going to affect five percent of GDP. Probably where I disagree with Massa, so I think he used the nine trillion dollars. I think that's the number he used. It's going to Disrupt nine trillion of GDP. And then he says his next assumption is it's gonna, that's gonna, there's gonna be a 50% profit margin, and then it's gonna be four trillion dollars of economic profit. And I think, so I agree with him, it's gonna affect five percent of GDP, maybe more in the fullness of time. Um, but I think this comes back to the point we were discussing earlier, where people over estimate the monopolistic nature of businesses, and that we're living in this sort of unique Gilded Age monopolistic era, and that that Is, is not the steady state of business. And I was reading, I found this McKinsey report recently, which said that if you look at total global GDP, One percent of global GDP is economic profit above the cost of capital, which I think is surprising. And I think, again, confirms this intuition that I think some people that, that I think is important, which is for the most part, GDP accrues to regular people, working people who get wages and salaries. And, um, it is very hard to sustain an economic profit above your cost of capital. And again, to, to, to moralize for a second, like that…

AI assessment note: “I actually agree with him fundamentally that AI is going to affect five percent of GDP.”

Q So you don't think in a market like Profound that Sequoia and the subsequent Quick Round has helped them significantly get great talent, get great customers, and get subsequent funding, which has then widened the moat between them and the plethora of other people? I, I'm sorry, I, I love you, but I respectfully disagree.

A I think that there are flywheel dynamics for sure in venture. And so I'm not saying that having, I think having screw in your cap table makes your company more successful. So I'm not saying that having a brand name, great VC, who's going to work really hard on your cap table doesn't change the probabilities. I just think it changes the probabilities less meaningfully than people think on average. And I think that, you know, you use profound as an example, because I was in the pitch when they came to the IC, the business was ripping. It was an amazing business. They had tons of customers lining up at their door to buy the product, and so, yeah, we're lucky to be in business with them, and we're grateful to be in business with them, and I hope that we can shape the journey in some way, and if there's five engineers that join, and Sequoia help, can, you know, help, having Sequoia involved to help them join, phenomenal, and by the way, I think that's the number one way that companies do benefit from having Sequoia on the cap table, is that, is talent and recruiting, and we can talk more about that, and I'm fascinated by recruiting, and recruiting dynamics, so I do think Sequoia helps with that, It especially helps with folks who are more mimetic, where I think the, the, the branding really helps. That said, I, I just resist the idea that like, oh, you know, I think this is just som…

AI assessment note: “I just think it changes the probabilities less meaningfully than people think on average.”

Q Like, I have to fight and fight and fight. And I'm like, yeah, your biceps are bulging, Doug. I totally believe that you have to fight for the very deal. It's all good. Um, you mentioned a couple of companies that you work with. The, Common critique posed to consumers of compute is margins, margin structure, unhealthy margins. Do margins matter today in this entry point of AI or not?

A I think they matter, and the companies I've invested in typically have reasonably high margins. Um, that said, I think they matter. They're, they're a directional indicator of how much product you've built on top of the foundation models. They are not Absolutely important. I, you know, I remember investing in a company many years ago that had the 30% gross margin, and now it has a 70% gross margin. And so gross margins go up over time. I think one thing as an investor that I guess you viscerally experience is that plenty of companies that get critiqued for having low gross margins end up being super healthy businesses in the long run. You know, Snowflake was one of the big, uh, indicts on Snowflake in the early days was that it had a low gross margin. Obviously it's a, it's a very good business. So I think if you have A real product that delivers a lot of value, and there's reasons why, as you get bigger, the cost is going to go down, and in AI, there's such an obvious reason, which is the cost of compute just keeps coming down every year, so the trend line is very clear. I think you can build a healthy business, and so I would even go to the extreme, and I haven't invested in any of these companies, but I would go to the extreme to say that even some of these companies with zero percent gross margins, I can imagine how they're going to work. Now, the companies I've invested in…

AI assessment note: “I think they matter, and the companies I've invested in typically have reasonably high margins.”

Q very quick successive rounds You know, if we look at, say, a Rillet, or a Profound, or, uh, do you worry about them? I remember Pat Grady once saying to me that his biggest challenge is that when he does the deal, everyone else wants to put in money at double or treble the price, and that really stuck with me. Do you worry about these very quick successive rounds?

A I think we try to find the right balance, and, uh, to be, to be honest, this is a conversation I have with a lot of founders, right? So this is like a, Very active conversation. We're all having these conversations all the time, and we're obviously in a market where capital is very abundant and very available, and so I see the argument for why people want to take the capital. I think one lesson we've learned is more capital does not make a company more successful. Capital is a, is fuel, but capital does not create the engine, and so I think this is a tension. I think this will always be a tension, and I think this is definitely a tension for companies right now where, and we learned this the hard way in 2021, Getting over capitalized has downsides. I think it leads to the biggest downside in my opinion is that it leads to this sort of internal perception of like, we're, we're winners. We're so successful. We're so great. And, um, the only thing that makes you a winner is having tremendous product market fit and having customers who love you. And, um, and so I think that's attention, some founders, and I've seen some founders do a great job of this that I've worked with. They, they, they really act like the money is not in the bank account and they really behave diligently and the team size doesn't grow too fast and all of this stuff. But I think that is the exception. Not the r…

AI assessment note: “we learned this the hard way in 2021, Getting over capitalized has downsides.”

Q features, And you have Canva offering it, and Adobe offering it, and Sketch offering it, and 10 others. Even though it adds more value, if you have a mass of other providers offering the same, all pricing power goes, and you are forced in a race to the bottom, which I think we're already seeing across a lot of spaces. How do you think about that, and am I wrong?

A No, I think you're right in its nuance, right? I think the nuance here would be in businesses with barriers to entry, You have pricing power. In businesses with low barriers to entry, you don't have pricing power. So I actually think that will be sector by sector. There will be some sectors where you're right, things get commoditized, and they just basically get bid down to zero percent gross margin. In industries with structural low gross margins, it's very hard to raise the gross margin. So if cost goes up, you'll raise price a little bit to compensate for it, but you won't raise price that much. In industries with higher barrier to entry, for example, industries that have data moats, industries where if I put all my data on your platform, then I can't really move. I think those people have pricing power, and if you had pricing power before AI, you'll have pricing power after AI. If you didn't have pricing power before AI, you probably won't have pricing power after AI.

AI assessment note: “No, I think you're right in its nuance, right?”

Q So do you think we will see the introduction of a new financial instrument to facilitate the movement of on balance sheet to off balance sheet?

A It's very possible. And I've spoken to a bunch of real estate investors. So something that people don't know is a lot of the capital going to fund data centers comes from real estate developers. Uh, real estate investors and these real estate investors, I, I talk to them and I say, Hey, do you believe that the data center you're financing is going to get used? Like, is there going to be enough demand for this data center? And they say, I don't really care for me. I'm giving Microsoft money. This is a loan to a big tech company. My, the deal that I'm doing is backed up by the credit of Microsoft or the credit of Amazon or the credit of Google. And I'm earning, you know, I could buy Microsoft's bonds at X percent yield and I'm earning X plus two percent yield. And so this is a, you know, go to risk. Adjusted investment. And to me, what that says is, you know, these big tech companies are basically issuing debt against their own balance sheets, but it's happening through these intermediaries, so it's perceived as kind of off balance sheet financing.

AI assessment note: “It's very possible. And I've spoken to a bunch of real estate investors.”

Q How does Facebook being the only one not having cloud as their cash machine change how they behave? Do you think?

A I think that, you know, the cloud guys are playing defense, Meta's playing offense, right? And I think, so I think that's like the easiest way to think about it. The cloud guys are protecting their existing business. I think Meta can afford to be pretty creative. And also I think Meta Has to play less defense because if they decide it's not worthwhile, they don't have to keep investing. The cloud guys are stuck in more of a prisoner's dilemma. They have to keep investing. If they do not invest, they risk losing market share in one of the greatest businesses of all time. And Meta gets to sort of play for the future and Zuck is still young and he's doing a fantastic job. And I could imagine Zuck playing a pretty important role in the future of AI.

AI assessment note: “the cloud guys are playing defense, Meta's playing offense”

Q features, And you have Canva offering it, and Adobe offering it, and Sketch offering it, and 10 others. Even though it adds more value, if you have a mass of other providers offering the same, all pricing power goes, and you are forced in a race to the bottom, which I think we're already seeing across a lot of spaces. How do you think about that, and am I wrong?

A No, I think you're right in its nuance, right? I think the nuance here would be in businesses with barriers to entry, You have pricing power. In businesses with low barriers to entry, you don't have pricing power. So I actually think that will be sector by sector. There will be some sectors where you're right, things get commoditized, and they just basically get bid down to zero percent gross margin. In industries with structural low gross margins, it's very hard to raise the gross margin. So if cost goes up, you'll raise price a little bit to compensate for it, but you won't raise price that much. In industries with higher barrier to entry, for example, industries that have data moats, industries where if I put all my data on your platform, then I can't really move. I think those people have pricing power, and if you had pricing power before AI, you'll have pricing power after AI. If you didn't have pricing power before AI, you probably won't have pricing power after AI.

AI assessment note: “No, I think you're right in its nuance, right?”

Q How does Facebook being the only one not having cloud as their cash machine change how they behave? Do you think?

A I think that, you know, the cloud guys are playing defense, Meta's playing offense, right? And I think, so I think that's like the easiest way to think about it. The cloud guys are protecting their existing business. I think Meta can afford to be pretty creative. And also I think Meta Has to play less defense because if they decide it's not worthwhile, they don't have to keep investing. The cloud guys are stuck in more of a prisoner's dilemma. They have to keep investing. If they do not invest, they risk losing market share in one of the greatest businesses of all time. And Meta gets to sort of play for the future and Zuck is still young and he's doing a fantastic job. And I could imagine Zuck playing a pretty important role in the future of AI.

AI assessment note: “the cloud guys are playing defense, Meta's playing offense”

Q your time. I think you might have been in Singapore, uh, and you were on with Carrie from KOTU and him. Just continuously. And what I actually thought, before we get to, like, actually how you organize time, is just, like, a young person's game. It's, it's easier when you're younger to do four a.m. every morning and have less sleep. Do you think Venture is a young person's game?

A Look, I think there's advantages to youth and there's advantages to experience. And I think you have to make the most with the cards you've been dealt. And so I'm young. And so I need to play the best young person's game there is. I can't play the Doug Leone game. Doug and I are not the same. Doug is way more experienced than I am. Right. And so I think the value proposition that I offer to founders is quite different. The value proposition that I offer to founders is I'm going to grow with you. You're probably my age, right? Like we're the same age. So like, Over the next 20 years, we're gonna become great together. That's the value proposition that I have to offer to founders. Doug is gonna offer a different value proposition. And so, for me, it's about delivering on what you have to offer and doing the best you can with what you have.

AI assessment note: “there's advantages to youth and there's advantages to experience.”

Q Okay. Unpack that for me. Which parts of the supply chain do you find most interesting and least picked over?

A Well, I think the real estate developers, you know, KKR owns Cyrus one, Blackstone owns QTS, fantastic investments. There's gonna be huge money-making investments for those firms. So those are great investments. There's other companies in this space that are really interesting. So I think that space will be interesting. Real estate developers in real estate, if you just look at real estate, it's a good business, right? And so being in that business is a good business. I think the power area is something that everyone is talking about. And, you know, I, I went, I was in West Virginia, uh, two weeks ago or three weeks ago, maybe. And I visited this company that's building long duration batteries that I had invested in at KOTU. They literally built a billion dollar factory in the last 12 months. So There's a just tremendous industrial movement happening, and I think in 12 months you'll see all these charts coming out of, like, more factories are getting built than ever before, and all this stuff. The industrial revolution is just getting started.

AI assessment note: “I think the real estate developers... I think the power area is something”