2: Michael Dempsey - The Craft of Investing in the Future

[00:00] Michael Downsing. [00:01] If you're not on camera, but you are live. Yes. It's great to be here with you. This is so weird. I'm sorry. I can't. We're like staring at each other. Okay. [00:08] You are someone who is deeply interested in the future and in change. [00:13] And this obviously... [00:14] It shows up in science and engineering, but also in markets, in how culture changes and how art changes. [00:21] What is technology? [00:22] and what boundary do you draw around it? [00:25] Oh, Jesus. [00:26] This is proof that you didn't send any questions ahead of time. I sent a few, but your answer rises. Technology is, I feel like this is the time where someone would be like, [00:37] have some really long like Rick Rubin style like pliffy thing. [00:41] I don't know. I think technology is just like... [00:43] the [00:44] like pushing the limits on... [00:47] the possible and science is creating the impossible. [00:50] That's my cliffy version of that. [00:52] do you think your [00:54] the way you think about technology and practical purposes and investing has evolved [00:58] materially in, or that any of us do, has evolved materially in the last 20 years. [01:04] investing has drastically evolved in the last 20 years. I think in my... [01:09] 13, 12 really, your career of investing. [01:14] it has changed also pretty materially. But it's always been rooted in... [01:20] How do I take a prescriptive view of the world and express that view through [01:24] deploying money. What's changed is like [01:28] The gradients of that, the timeframes of it, the optimism of it.

[01:32] The... [01:33] risk parameters, maybe like the views of like... [01:37] terminal upside of things, which I think is also what has changed in... [01:41] traditional markets. People view like [01:44] compounding is like compounders like most obvious straight ever now if you look at [01:47] Most of the major hedge funds, all they own is like the same 20 names. [01:51] And they all have the same kind of view of like asymmetric upside and these businesses that will just keep going forever. [01:56] and then I feel like Adventure, that was the viewpoint and [01:59] maybe 19, 20, 21. And a lot of people lost money that way because they held the things on the way up because they thought they would keep going and they all crashed. [02:07] And the initial like being early to that framework was like software eats the world and running that. So [02:12] Yeah, a bunch of stuff has changed in there. We... [02:15] And again, this is probably... [02:17] product of being relatively young. I've sort of thought about technology as a [02:21] I watched it go from being a thing I was really interested in to sort of like [02:25] eating the world to your point or on software. [02:27] There's also... [02:28] ways to push that, e.g. markets are a form of technology. There are even things, technology, things like crypto, let's say, that are part technology in the sort of science and engineering sense and part cultural change, these types of things. [02:41] Which is sort of the root of my question, like, [02:43] You've obviously done things that most people would squarely put in the... [02:47] technology bucket, particularly before they were in Vogue, e.g., [02:51] robotics, AI, ML, bio, these types of things. [02:55] Did you relate to it in that way of like technology is coming for these things or is this? [02:59] I'm just a technology investor in the sense that I invest in the way the future can change in a

[03:03] mechanistic way. [03:05] I would say it's like, what is the... [03:08] what is the future that you believe in? And then what are the things that impact that future? And [03:15] and what is canonically thought of as technology, and then sometimes it is... [03:18] waves of like [03:20] Scientific or technological change pushes cultural change, which pushes scientific change, which pushes cultural change. And so you can look at like... [03:28] And like we're obsessed right now with like self-directed medicine and biohacking as an example. That is a... [03:34] structural societal change driven by scientific progress and technological progress. And we can talk through the nuances of that if you want, but [03:43] crypto was underpinned, I think, largely by societal change and then technological change kind of like [03:50] worked its way around this problem and subset of beliefs and then societal change continues to [03:57] actually be the thing that [03:59] pushes the industry forward, not technology in my opinion, which is [04:03] why I think crypto is so weird to invest in because you're betting on [04:07] progress [04:09] Out of like... [04:10] figuring out how to express the financial and or like create value from the societal values, which is like usually not how it works. Usually it's like technology pushes the behavior, which then you get this like flywheel going. [04:23] Yeah, I think that's why a lot of traditional investors can't grok crypto, don't believe in it, don't invest in it, because they're like... [04:30] It's just a bunch of people who... [04:33] have these beliefs and like...

[04:35] Why should I believe any of those beliefs if I can't see the tangible thing, which is... [04:39] the technological change. That's also why I think a lot of people now are grasping onto like AI and crypto because they're like, oh, I get the AI thing. [04:46] And like I can at least very simplistically draw a line into... [04:51] why [04:52] you know, on-chain agents matter. Again, I think it's [04:56] Kind of like a bullshit framing, but like... [04:59] I get why some people think that's important. [05:02] in the concept of [05:04] sort of like technological determinism [05:07] Do you view... [05:08] every instance of that on a kind of purely case-by-case basis or are you kind of bucketing them in the context of oh if it's related to crypto and finance and markets [05:16] There's way more of this cultural... [05:18] component of it, shifting technology and vice versa, versus if it's [05:22] I don't know. At some point, the cars will be able to drive themselves. And so it's determined and you don't need to worry about the cultural stuff. [05:29] Mm. [05:32] Can you rephrase that a little bit? [05:35] There's one view that says... [05:37] You should never have to worry about anything other than inventing the future. Yeah. In all contexts. Yeah. [05:42] Okay. From a science and engineering standpoint. [05:45] You just made the point that crypto in some ways [05:49] Whether or not it was determined, it certainly rose out of more cultural shift and cultural change. [05:53] You've written about this in the past, about how various, whether nuclear bomb or Bitcoin, things have emerged out of really a cultural situation in the United States. [06:02] in that context, [06:03] Do you view every incremental new technology or opportunity or problem

[06:08] Is it a... [06:09] Are you evaluating on a case-by-case basis, or are you drawing broad frameworks that say, [06:13] In biology, for example, or biotech, [06:17] it's we don't really need to care that much about the cultural stuff until a certain threshold in crypto. It's everything. Yeah. Self-driving and so on. Yes. Uh, [06:25] I think it's, [06:26] categorically [06:28] You have to think about it like case by case, but I think that the. [06:31] the [06:32] complexity over the past... [06:33] three years of our investing. So like, [06:36] 2016 to 2021 was [06:39] Will these technologies work and will they [06:42] reach some level of commercial viability. [06:45] Now it's [06:47] They work, kind of, and they're proliferating and they become like platforms. They intersect. So you go from vertical to horizontal. [06:55] And that changes. [06:57] how you then have to think about each of these individual things. And so like, [07:01] in the [07:02] AI case, it was like you back companies, but doing fundamental AI research to try and get [07:08] the broad concept of AI to work. [07:10] And now it's like, okay, if AI works, what are all of the cascading effects of that? [07:15] And an example of that is like, [07:17] The first order version of this is like these fast following, like highly skew amorphic companies. [07:22] The idea is [07:24] You take a thing that everyone knows, [07:25] You have this new technology, you plug it in, you get like 30% to like 5x ROI on the thing that you did. [07:31] It's really boring. That's not how we invest. We are not good at backing like one of five companies that all look the same. [07:37] and we don't think we can make money that way. Other people can.

[07:40] Instead, we try and think about, okay... [07:43] if you reach some sort of like commoditization curve or like development curve, [07:46] you then actually want to figure out what are like the native or non-skeomorphic ways in which companies are built. [07:51] And that is what is much more dictated by... [07:55] societal structures. So like [07:57] An example of that could be [07:59] And if you were to think today how most people think about like AI-assisted therapy, they think that you're going to talk to an AI bot and it's going to be you and your AI therapist and you'll chat them and then they'll chat back to you and... [08:08] It'll be async, it'll be very cheap, and it'll be better for everyone's mental health, and blah, blah, blah. [08:12] The like native version of that is actually probably quite different. What that looks like is unclear, but maybe it's like your AI, you know, you have something that understands like your mental state and you tap it into your group chat every now and then, or you wake up and it's like a co-star like experience that you have every day. [08:26] or there's a bunch of different non-Skeuomorphic versions of it. [08:29] We hopefully will back that, not... [08:32] Talkspace, BetterHelp, whatever, et cetera, with AI, where we're just like, you know, I open up my thing, I chat. [08:37] Do you need to see it? [08:39] Do you need to see the live new technology kind of like [08:42] Mi. [08:44] regular people or meet the market first before you can start to draw out those conclusions? No. [08:49] No. [08:50] I think if you do that, you probably are too late in your thinking. Okay. Okay. So this gets to an idea that [08:55] There's two quotes, uh... [08:57] I want to give you one of which [08:59] you've been talking a lot about lately. [09:01] And they all sort of sit under a broad idea you've written a lot about, which is inflation points. And this notion of how do we think about the future in this cascading way. [09:10] The first quote is Matt Kohler. Yeah.

[09:12] It sounds like people bring this up to you a lot. And it's our job is not to see the future. It's to see the present very clearly. [09:18] And the second quote is from you. [09:20] Granted, I'll admit or I'll add the context that it is primarily in the context of driving signal in a financial markets context rather than explicitly venture capital. [09:29] But you say the complexity of market dynamics continues to outpace our ability to model and predict them. [09:34] Within this, though, the role of an investor is still perhaps to see what others don't, understand what others can't. [09:41] and act when others won't. [09:43] Yeah. [09:44] So how do you reckon these two things? Was Matt Kohler totally wrong? [09:47] Matt Kohler... [09:48] was not wrong because he's a really, really fucking good investor. Matt Kohler of Benchmark. Benchmark, yeah. Amazing kind of legendary investor, legendary time there. Yeah. Yeah. [09:57] I think there's two things, right? There's where... [10:00] What is the moment in time when you're saying the thing? And the funny part about that is that's from like a Quora post and then... [10:05] There's a little bit more to it that he says, but everyone pulls that quote and then shoves it in my face. I think they... [10:10] The thing that I've started... [10:12] Coming up in venture, I would talk to everyone about how I wanted to do investing. [10:17] this highly prescriptive thesis-driven research-centric view, [10:21] And everyone was like, [10:23] It's all about what the founders show you. You know, like you got to understand the future that the founders are. So you just react to that. They're thinking about this all the time. [10:30] I think that's maybe true when there's like four firms investing at seed stage and like you all kind of share deals and that was. [10:37] 2013-14 when it was first round and Floodgate and a few others. [10:41] Now, like every person in the world has a venture fund. And I also think like,

[10:46] There's so much money. [10:48] There's no consensus. Venture capital is, in my opinion, the lowest consensus financial asset class in human history, I think. And sorry, highest consensus, lowest conviction. [11:00] Oh, very different. [11:02] Very different, yeah. Highest consensus, lowest conviction. [11:04] And so what that means is that conviction pushes people into consensus quite quickly. And part of this might be that. [11:10] more mature financial markets, kind of implicitly, if you have consensus, you just can't make money. [11:15] Theoretically, but it's just not... [11:17] You must, by definition, be somewhat high conviction and somewhat anti-consensus. Yes. Maybe aside from... [11:25] investing in facebook and google but even then that wasn't completely consensus at some point right like yeah like what what some people out of the conversation with someone who manages a really large endowment and [11:34] He was saying there are investors in some of these large venture firms and hedge funds that just own, you know, [11:39] the big AI companies and the big public companies. [11:42] And he was saying... [11:43] Sometimes the most obvious trade is the one that like people who manage money and charge, you know, two and 20 on it don't want to make. [11:48] And so like they just... [11:50] There's value in just buying that. [11:51] which I think is true but [11:53] To my usual point, [11:55] Venture just got way more competitive. It got way more [11:59] of a high dispersion of people, of types of people, of returns. [12:03] And [12:04] I think it's up until 2016, 40% of all people in venture capital had a Stanford or Harvard MBA. [12:09] And now there's, I don't know, everyone has fun. And so I think if you want to have alpha... [12:14] And if you want to do it from a small fund like ours with a...

[12:17] growing but like not the biggest brand. [12:20] you just can't really say like, my job is to just like meet the, you know, ex- [12:26] uber you know engineer who has an idea that everyone's going to understand and go and like win that investment [12:32] unless you're going to pay highest price, which... [12:34] We've seen people blow themselves up doing that or you have a ton of money. [12:38] And we just refuse to do that because we like making our investors money in an outsized way. [12:42] And it's just more fun. [12:44] Like, it's better to be able to, like, look at the world and say, like, what should exist and let founders come in and shatter them or... [12:50] fit within those and then like [12:52] it creates an actual... [12:54] theoretically an actual competitive advantage. [12:57] this might feel like zooming out, but I think it's very connected. Years ago, we had a conversation, I think that you tweeted about that was like, [13:05] Me realizing how much of my conversations were... [13:08] cached basically basically just remixing and replaying ideas that i'd heard before and [13:13] I think I wrote about this. I rewrote about this point in a post recently. And maybe I read it. Maybe that's what popped it up. [13:19] And then on top of that, I think [13:20] Both [13:21] with respect to the answer you just gave, and also sort of anyone who knows you, would throw around... [13:26] words like original or contrarian. [13:29] Do you think it's possible [13:30] Or today, do you think you're overrating or underrating being original? And on top of that, [13:35] How do you relate to the word now, the word contrarian? [13:40] I don't like the word "conteriors" become a meme. That's like that thing of like... Okay, reduce down the meme. [13:45] The notion in your – the program running in your head on some level that is oriented around –

[13:51] I think... [13:54] I would frame it as... [13:57] Original is like... [13:59] I think is still underrated as an investor because I think that, [14:04] Again, like when there's excess capital, originality is if you want to be early. And if you want to compete on the idea of being early... [14:12] and [14:13] You don't maybe have structural advantages, if no small venture fund does, except for [14:18] the burden to exit companies is lower. It's the only structural advantage we have. [14:22] you have to be early and do things that other people just don't get. [14:26] And the beauty of venture is you have to be right one to two times every 30-ish times. [14:32] And so we try to do that as much as possible. [14:35] The contrarian part, I think, is more like... [14:39] Thank you. [14:40] It's kind of like [14:42] How do you exist in a space such that you can extract the most amount of alpha from it? Yeah. [14:50] What is Alpha? [14:51] the ability to have excess returns relative to what the overall market would give you. And in a framing, you could just say to have [15:00] something that [15:01] A level of intelligence that is or a few or something that feels excess and asymmetrically skewed. Do you think thinking about alpha, whether explicitly or implicitly, is important for people who aren't investors? [15:14] Yeah, mostly because I think everyone does that anyways, though, because they think about [15:19] what makes them special, or they think about what

[15:21] makes them unique. [15:23] and [15:24] or what makes them not and i think people struggle more and more culturally with not being unique in some way yeah [15:31] I think we see that across a bunch of things. [15:33] And so I think about kind of [15:36] consistency as maybe the contrarian thing that you could do. [15:40] which is like, [15:42] Can you, there's a famous quote we use a lot. It's not famous actually, but it's a quote we use a lot. [15:47] called the, if you want to understand how something works, study it when it's coming apart. [15:50] And the idea is... [15:53] In these areas, they cycle up and down and hype and consensus, whatever. Yeah. [15:58] and [15:58] to truly have [16:00] contrarian or whatever, some structural advantage [16:04] The thing is to understand the counter-cyclicality and be there the entire time. [16:08] And so you can exploit being there early and being original. [16:11] You can exploit it. [16:13] counter trading the market and thinking something is interesting out of other people. [16:16] And then... [16:17] When it's really interesting, you can exploit not being a part of it and just let the world go and still learn. But you don't have to lose money doing it. You don't have to. [16:25] Spend too much time doing it? [16:26] And then when it cycles back down, inevitably... [16:28] You have... [16:29] compounded all these earnings by just continuing to be there. [16:33] This obviously fits. [16:35] kind of very intuitively to anyone who's spent any time in crypto over the last seven, ten years. [16:40] Are there other areas where you've seen this play out materially recently? AI, for sure. [16:45] What point of the cycle do you think we're in? [16:47] in there. We're definitely at the top part of the state club. So you're less active.

[16:51] would be the... We've made like... We've made... [16:54] one AI investment in the past 18 months. [16:57] Yeah. [16:58] Zooming out again a little bit, we talked a little bit about science and engineering. You've written and talked about post-science projects being the place where you like to play. Maybe that... [17:08] tension of the early part of that [17:10] That curve? [17:12] There's also a quote I really liked that you brought up from Richard Hamming, which is, in science, if you know what you are doing, you should not be doing it. [17:18] In engineering, if you do not know what you are doing, you should not be doing it. Yeah. Which is awesome. Yeah. [17:25] extrapolating this out and especially for people thinking about this concept of alpha, whether they're entrepreneurs or [17:32] employees or [17:33] people who are just interested in the future, [17:36] How do you think about that tension? [17:38] And when something goes from like, yeah, great, Mike thought about the future a ton and he determined that like biotech is going to be a big deal. [17:45] and gene editing, all these things. [17:47] Timing matters. [17:48] Timing certainly matters if you're an investor, but it broadly matters too. [17:52] What does that post-science project space start to look like? [17:56] I think it's changed a lot in the past five years, which is that we used to... [18:01] be very comfortable as a world. [18:03] letting value accrue in non-capitalistic places. And you could deem those places as academia or research organizations. [18:10] And now we are not comfortable with that as a financial system. [18:15] structure. And this is kind of as we've seen technology move. And so [18:18] People now... [18:20] push the limits of science into companies earlier than ever before. They fund them too quickly.

[18:26] I think what that means is that you [18:28] If you want to operate in those spaces, [18:30] My view is you have to have... [18:32] two very distinct frameworks. [18:34] or two very distinct free-mangs maybe? [18:37] The first order one is [18:38] You have to have a very, very clear view of how do you sequence to this terminal future. [18:43] and [18:44] You have to be very comfortable in [18:46] testing those assumptions, but [18:48] being pretty damn right along the way and sequencing those things. [18:51] The famous ones are like the master plans from Tesla and SpaceX. There's a bunch of other companies that have done this as well. [18:56] But you have to be basically... [18:59] super high conviction with super low data for probably your first four years, because if you're pushing a science project, you're not going to have tons of proof points. You're doing a lot of really hard things. You're front loading a lot of really hard work. [19:09] that you are making a very explicit bet that everyone will care about once you do the thing. [19:14] That's one. [19:15] The other is something called the Equal Odds Framework, which Chris from Runway has [19:22] told me and pushed on me many times. [19:24] And Chris builds runway this way, which is... Runway is... [19:27] kind of the premiere or generated video. Yes. [19:31] And so his view is... [19:33] They're building a new camera. They're building a new world model. They're building the future of creativity. [19:39] And [19:40] In a time in which everything moves so quickly and in which their core asset is they think about the world very creatively, [19:48] They are incredibly generative, no pun intended, as a firm. [19:51] you want to get the maximum number of ideas out into the world such that you find the right one, because every idea has equal odds of being the right one.

[19:57] And I think if you operate in that post-science projectee time, [20:02] You're implicitly [20:03] saying that the world is going to change very quickly. It's going to change more quickly than many people appreciate. [20:08] because you're taking something from a science project and pushing it out. [20:11] And... [20:12] you either know perfectly how it's going to change or you know how to maximally take advantage of that change. [20:18] And this is maybe one cut on how to... [20:22] increase your chances of [20:23] making a jump on the schemorphism idea. [20:26] Like, how do you get from radio and television to the new thing? Right. And I think the dangerous part is that, [20:33] You... [20:34] If you're building a company, you get put in the middle, and if you're investing, and if you're... [20:38] working, whatever, all these things. [20:40] you can get put in the middle where you start to worry about these short-term optimizations. [20:44] And so you can... [20:45] Say, [20:46] Well, I don't know. [20:47] The models only allow me to do X, Y, or Z today, so I need to build product around this little constraint and then [20:53] That'll allow more people to, in the next six months, become users. [20:56] And in reality, you should [20:58] Either no, you trust the trend line, you trust the extrapolation, you trust the nonlinear growth of this thing. [21:03] Or you should know that [21:05] In service of that, you should know the exact [21:07] number of ideas that you want to push forward on all these different constraints because you [21:11] or kind of if you're in that framework, you don't know exactly what those constraints are. And so I think where a lot of these companies, the past two years of AI company building at least has been, [21:20] A lot of companies that have grown revenue very, very quickly and then kind of died, like churned out 80 percent of revenue. [21:25] And it's because they made all these short-term optimizations or short-term bets on what was possible.

[21:31] in the next 12 months. [21:32] And I think that is like the death valley for most companies. [21:36] In the context where they are... [21:39] in an area that's starting to work. Yeah. [21:42] And the timing is probably right otherwise. Like, there will be serious winners in this time frame. Sure. But, like, maybe, right? But maybe not. Maybe it's actually... [21:51] two to three years from now where the like native companies come up. Right. [21:54] in the broader inflections points. [21:57] kind of idea you talk about false inflection points. [22:00] Put simply, [22:01] this notion where sort of it seems like the wave is coming for a given technology and then you're too early. So a classic example of this would have been like VR when, [22:10] You've written about this when Facebook bought Oculus. [22:13] It seems pretty intuitive to me that [22:15] there's like a pretty clear science to project and not science project curve for [22:21] certain areas, let's say [22:23] At times, self-driving cars, maybe bio, [22:26] But there are also other areas where it's much more hazy. I'm curious if there are any... [22:31] Maybe to take the VR point right now, as we're at an interesting point post-Vision Pro, and then more broadly, are there any other areas that you think are more likely to be false inflection points today? [22:42] I think I have a [22:43] more... [22:45] I think this is maybe my most smooth brain approach to anything in life. As you know, I really hate that general framing or maybe left-curve, right-curve approach. [22:55] I think that... [22:56] The beauty of what I do is if I believe in a future, I should just continually bet on that future and keep losing money until I don't.

[23:03] And so... [23:05] I'll start with that framing, which is... [23:07] If you fundamentally believe something will happen, [23:10] Thank you. [23:11] As long as you're not wrong by like 40 years, you should just keep doing the thing. [23:16] assuming you make other good decisions around that thing. Right, you can't lose money for 10 years straight. Can't lose money for 10 years straight. But again- That's what crypto is for. Yeah. Also, that's the beauty of what we do as venture capitalists and [23:28] Also, to be clear, as a founder building company, we [23:32] we're totally okay with people failing. It's... [23:35] If you were a founder, this approach would not work at all. [23:39] I think it would not work at all in the way of like, you would start a company, it could fail. But like, [23:46] what do we love more than failed founders like in tech the the best like you've [23:51] You become a founder, you fail. [23:53] You... [23:54] maybe get a better job than if you didn't become a founder of film? Yeah. [23:58] The point I would most agree is there's nothing cooler than... [24:00] Or more compelling when you see somebody who's been... [24:03] sort of running around a problem. Yeah. And just trying to attack it from so many different or I found out recently the ramp guy did Paribus. Yeah. Like, it's just cool. Like, yeah. Oh, clearly, like, maybe to your point about consistency. [24:15] Somebody who's just like, nah, this is, this, this is going to happen. Yeah. Eventually like. [24:20] the tide turns on these people and they're like, Oh God, I can't believe they're still working on that. Like, this is so lame, but like, [24:26] I don't know, that's the way the world is, fine. More explicitly, though... [24:29] to just push back a little, like... [24:31] investing in VR, investing in self-driving cars, I remember in 2015, maybe this is a good example of what you're saying, I remember in 2015, 2016, Uber was happening and the early self-driving car stuff and I lived in LA so I was biased but I was just like,

[24:43] 10 years from now, for sure, like, cars will be gone. Like, we will have solid iron guns. And if you had asked me about it a year ago... [24:50] two years ago, I've been like, [24:51] I could have never been more wrong. It's going to be another 10 years. I wrote an awaymo this week. [24:55] And... [24:57] It's a good reminder that both of these things are kind of true. It kind of was 10 years, but it really... [25:03] Like it took nine and a half years. Yeah. And even then we're not really quite through the woods yet. Yeah. It's like underestimate in a year, overestimate in a decade type things. Or opposite. Yeah. [25:13] Okay, so are you still willing to answer the false inflection point? Sure, yeah. I think the false inflection point stuff, I think... [25:20] VR is... [25:22] doing the thing that crypto is doing, which is like, it's, well, actually now it's different, but it's riding a commoditization curve down on hardware and it's not moving nearly as fast as people thought it would. [25:32] And... [25:33] I think eventually it will get to... [25:35] A [25:36] hardware scale that will drive... [25:39] insane adoption. I think [25:42] now would probably be a good time to start losing money in VR. [25:45] Because... [25:47] You have one company effectively now too, I guess. [25:51] just blowing tens of billions of dollars on the problem a year, which is kind of what Waymo did for a long time. [25:56] And [25:58] I think the problem is that people... [26:00] Thank you. [26:01] People try to do the, like... [26:04] they don't think about value capture enough. And I think that that's also one of the parts of- Value capture in building a company? Yeah. [26:11] I think that's one of the parts about being somewhere for a long time. And there's this chart that we have in, I don't know, we use it a bunch.

[26:18] which is like if you look at [26:20] major [26:21] cambrian explosion moments of like deep mine being acquired cruise being acquired oculus being acquired um [26:26] the foundation of Bitcoin and Ethereum ICOs afterwards. [26:30] Typically, [26:31] financing for companies spikes like 6x in the next 18 months. [26:35] And if you look at the companies in those [26:37] areas, they all failed. Like all of them are bad. [26:41] And I think it's because people are like new to the thing. [26:44] get excited, deploy money, don't think about [26:47] the principles of building these companies over long term, but they know they need to get in. [26:52] And there's bad incentives. Why? [26:54] And. [26:56] Was Oculus Post Science Project, in these examples, like... [26:59] It was not. Sorry, the better question. Would you have argued then Oculus was not post science project or is that am I trying to overlap two different ideas? I think that. [27:08] Oculus was... [27:11] I think that Oculus was probably more of a science project than not. [27:17] But the counterpoint is they built a hardware thing that at the time worked relative to what we expected. [27:23] And it was relatively cost affordable. Like you could buy it. It wasn't crazy. [27:27] The DK2 at least wasn't bad. [27:30] The problem was, is just like, there wasn't any ecosystem around it. So it was, [27:34] The technology was not a science project, but the [27:36] category was [27:38] And what people try to do is then invest to brute force the category creation. [27:42] And... [27:44] Thank you. [27:45] That's just not that valuable investing in VR studios. And maybe to the value capture point, you would make a case that in almost all cases, the actual just...

[27:52] Product quality. [27:54] is going to lead any kind of commercial aspects of it. [27:58] For things like that, maybe. Yeah. I just... I also... But the... [28:01] The new learning and all these things is the, well, some people still don't know this, but like [28:06] The first company is the one... [28:08] First movers captures tons of value. Fast followers, in my opinion, destroy a bunch of value. [28:13] But some will work. [28:15] Native non-Skeu morphics capture next order value. Final movers capture issue ton of value. And so now everyone's like, [28:21] It's the canonical founders fund, how many space investments have they done? It's like SpaceX and then nothing for a decade, right? [28:26] And so... [28:28] I think why that happens is because those businesses, what people realized is they needed to build full stack companies. [28:34] And if you build the entire thing, it turns out you can capture a lot more value and you can brute force a lot of the dynamics around these industries. [28:42] problem is is that everyone looks at those examples and says [28:47] well, yeah, it's like SpaceX, and it's like Anderol, and it's, you know, like Oculus, and [28:51] DeepMind to a lesser extent. [28:53] All of those examples basically were funded by like billionaires who could not care less about the sequencing of milestones over time for the first five years of those businesses. OpenAI, similarly. [29:02] Are there enough of those cases? [29:04] particularly in kind of [29:06] core science project. [29:07] domains [29:09] that [29:10] Could. [29:12] ruffle some feathers on some of these frameworks or broadly the notion of effectively investing in science-driven stuff. Like so many of these [29:20] core science problem

[29:22] companies and ambitious projects. [29:25] They sort of map to your frameworks, but they also... [29:28] To your point, like... [29:30] didn't really work with the traditional venture model. Yeah. Mark Andreessen talked a little bit about this, about like, [29:35] Maybe getting ahead of like, where are there? Why don't we have more Elons? And one notion is like the current venture capital model actually isn't set up in a way where you could fund. [29:43] SpaceX? Yeah. Or maybe even Tesla? Yeah. [29:47] I think it is now. I think it hasn't been for the past decade. [29:50] Aside from a small fund called Compound and others. I think it just... I think now... [29:56] This is the problem with what we do. Like now we actually can't fund those things because everyone's like, oh, it's remotely interesting. And it kind of pattern matches OpenAI. Here's 50 million bucks. [30:04] And it's like when we seeded Runway, it was a $2 million seed round. Like the company is now since raised a lot of money, but like... [30:11] We see it at Wave, which we view as the final mover in autonomous vehicles, is a multi-billion dollar company now. [30:17] It was $2.2 million. [30:18] And so... [30:20] That was because they were like, we're going to do this very constrained problem and we're going to show the world with a very different approach how you get a car to drive itself in a more nimble way. [30:28] Now, any of those businesses... [30:30] people at some big firm who [30:34] doesn't give a shit, would be like, here's 10 million bucks, let me know in 24 months if you do anything with it. [30:38] It's good for the entrepreneurs. [30:40] Uh... [30:42] No, but I'm biased. [30:45] In the context of AI, in this inflection point and cycle standpoint, technology's obviously had a number of – there's – [30:51] Every 18 months, there's some new thing. Again, crypto is an amazing example of this upswing and downswing.

[30:57] Some people might say that the kind of modern AI stuff is a version of that. Others people would say, no, no, no, it's a much more transformative thing in the vein of the dot-com bubble, which was not really a bubble. It was the introduction of kind of everything that was to come. [31:10] What's your cut on kind of where we're sitting right now? You briefly... [31:13] alluded to it earlier, like where we're sitting right now in this AI stuff. [31:17] Not. [31:18] Let's disregard kind of like robotics and some of the other things that it came into. [31:23] You mentioned you're not doing as much. It's super crowded. We're kind of post-wherever. [31:27] that upswing is. [31:30] I think – [31:32] All of my opinions are actually more consensus on this, which is like... [31:35] If you stopped model development today, there's a lot of juice to squeeze. [31:39] I think that there will be continual model development [31:42] I think that we need science breakthroughs to have super intelligence. I don't think we need science breakthroughs to have general intelligence. [31:49] And I view those things as very different. [31:51] Um, [31:52] And I think that like the... [31:54] value comes from. How do you [31:56] build products around these things now. And I think that [31:59] the clearest... [32:01] signal of that is that [32:03] Opening eye, very... [32:04] quickly realized that [32:06] they needed to become a product company far more than they ever thought. [32:09] Do they realize or do they... [32:12] Another... [32:13] The argument would be that they... Stumbled. Yeah, they stumbled... [32:17] And maybe it's not even the right thing. They just... [32:20] The product's growing so much that they can't look away. [32:22] I know people who think they should be totally focused still on being an API company, ultimately.

[32:27] or developer company. Yeah. [32:28] I... [32:31] I think that they... [32:32] did not stumble. I think they noticed [32:36] Well, they knew Anthropic was going to push out Claude, and they front ran them. [32:40] And... [32:41] I think secondarily, they... [32:44] have taken very... [32:45] specific approaches to [32:49] trying to be more thoughtful around like UI, UX and how they've hired people and pulled people from different places. [32:55] Did they stumble into the fact that ChatGPT is like whatever it is now, 65% of revenue? Sure. Like maybe they would have thought it would have been 30, 70, not the other way around. [33:03] But I think as the things progress... [33:06] This has always been kind of the complexity with like, [33:09] a box that enables you to type in anything and get a bunch of different outputs. [33:13] you start to really wonder why if you own the box, you don't own the outputs as well. [33:19] You... [33:21] And if you truly believe, [33:23] And I think the other thing, too, is the incentive is they actually have to do this because... [33:27] If they didn't believe that they were building AGI, [33:30] then they would want the API because then you have products being built around the constraints of this non-AGI thing. Right. [33:35] And so the moment that [33:37] starts getting doubted [33:39] The... [33:41] the kind of like value of the thing starts to get down a little bit. And I think that right now where we are is [33:46] You cannot... [33:47] be in a place if you're building open AI or anthropic [33:50] where anyone doubts that you are doing the most important thing in the world. [33:55] because of the valuations and scale of ambition and keeping talent or something else i think because of the

[34:01] Yeah, the talent, all those things, and the fact that you just need to raise an insane amount of money. And you need to bend the world to your view on the world, which is like you need to – [34:09] get people to be like, okay, fuck, we'll turn nuclear back on, which they literally, like, Microsoft literally did. [34:15] Like, I think, like, [34:16] To do that is just... [34:18] If anyone has any doubts and anytime looks up and is like, [34:22] Well, what if not? Like then the entire thing can start to collapse and like, [34:27] Again, tying it back to crypto, like we see what a death spiral looks like and it can happen very, very, very quickly. [34:32] if you lose faith in the core institution, [34:35] And so – [34:36] I think that that... [34:38] It's like both a very smart decision to want to own product and almost an existential one. [34:44] Speaking of maybe OpenAI, and we talked about Elon earlier, [34:48] We definitely are living in an era, maybe tech's always been this way, of sort of grand heroes. Yeah. Sam Altman, Elon Musk, Mark Zuckerberg, antiheroes, SPF. [34:57] I'm curious, one, how you think about that phenomenon. [35:00] And also how you think about the sort of classic jockey horse. [35:04] talent idea when it comes to [35:06] technology and building companies. [35:08] The premise there simply being... [35:10] Is it more about this? [35:12] The talent behind it, or in some cases, this grand, amazing hero, visionary, etc., Elon Musk type. [35:18] Or is it about the market and the timing and kind of the right business model? [35:23] I think that... [35:25] we have become... [35:27] Way to

[35:31] Obsessed with trying to pattern match humans to each other. [35:36] And [35:36] I think that the tech community... [35:39] is [35:41] Thank you. [35:41] so masturbatory about these people. [35:44] And [35:45] I think they're incredible. And I don't get why. [35:48] We feel the need to... [35:51] Almost... [35:53] Extrapolate. [35:55] what they do all the way down to who they are as humans all the way down to like why they are the [35:59] perfect version of a human. Right. [36:02] And I think the... [36:04] The [36:05] Other side of that are the [36:07] People maybe listening or talking about this around the table would say, [36:12] That's because they're so ambitious and you're not or something like that. [36:16] And they frame it as this ambition framework. [36:21] I just think... [36:22] They're like, they're incredible people who do incredible things. And there's a ton of them. And it's, there's a ton of them. There's a ton of people who do that. [36:30] One counter argument here, at least, would be that [36:33] Almost everything we've chatted about for the last 20, 30 minutes is [36:37] a sort of like silver bullet solved marginally, at least by Elon Musk. Like, [36:41] the [36:42] I'm willing to work on this for 40 years and I don't care about the timing. I'm willing to sort of, I am able to basically like bend the inflection points in my favor. I'm able to bet the timing doesn't matter. Like all of, I'm able to actually raise the capital to like hit the waves. [36:56] in a way that [36:58] One is rare, and two... [37:00] Seems. [37:01] if not necessary, um...

[37:04] very very critical to doing the most ambitious type things is it a coincidence like [37:09] Do we have, and maybe this is my question, [37:12] ignorance, like do we have really good examples of incredibly hard [37:16] Combo science plus commercial plus like narrative. [37:20] juxtapositions that haven't been pulled off by somebody or do we... [37:24] definitionally sort of frame somebody as a hero if they are able to pull that off yeah i think that's the [37:29] So I think [37:31] What Elon does is, like, absurd, right? I don't know if you... [37:34] read his Memphis data center thing recently, but he spun up. Also number one in the world in Diablo, apparently. Yeah, this is crazy. [37:41] Spun up one of the biggest data centers in the world in Memphis in 20% of the time that anyone thought it would be possible. [37:48] Also did it by bringing in a bunch of trucks. There's a light and gasoline on fire to power it. And people were not happy about it, but it's insane. It's incredible. Um, [37:58] There are a million people that in the past five years pattern matched to the Elon version. [38:04] of being a founder to the Sam version of being a founder. [38:10] I actually don't think anyone pattern matches to Zuck, to be honest. I think Zuck is in a category of his own in his aesthetic and... [38:16] the depth at which he talks through problems and the variety of problems he cares about and [38:21] the vision he has that is not... [38:24] actually directly science fiction oriented where Elon's is actually far more science fiction oriented. Elon's directly cosplaying. Yeah. [38:31] which isn't bad. Uh, [38:33] And they lit tons of money on fire and blew up their companies multiple, multiple times. And we see them all the time.

[38:39] It's [38:40] true that you have to be incredible at building narrative if you're going to build large-scale scientific companies. [38:46] It also is true [38:48] that... [38:49] you have to know when to pull the levers off of that and... [38:53] execute or... [38:55] They bridge the gap between narrative and [38:57] operations. [38:59] And like... [39:00] That's the expertise that Elon has. [39:04] Sam's expertise is [39:06] He bends capital markets in a way that people can't seem to get away from. You talked about in the past, too, he's really, really good at betting on the consensus view. [39:15] Yeah. In a sense, like, or sort of accelerating the Kansas SU. Yeah. It's accelerating the thing that's working. Yeah. But again, I think all of these things are... [39:24] This is why I talk about being a founder as an [39:27] maybe not risky or whatever, all these things is... [39:29] we bury the dead quietly. Like, each of these people have done really dumb companies that they raise lots of money for as well. Well, let's go back to Zuck for a second then. Yeah. Both because, one, he hasn't done that. Exactly. And two, [39:40] So you mentioned two things. One, you mentioned not enough people are imitating him. I didn't say not enough people. Yeah, I don't want words in your mouth. But one, I'm curious what that means. And two, you mentioned his aesthetic. [39:51] Which I think most people listening would say, like, his aesthetic has dramatically changed the last three years. I think you were implying actually something a little below the surface that I think has been more consistent. One of the reasons I've been pretty confident in Zuck for a long time. [40:03] But I'm curious what [40:04] One, if you think Zuck today is really that different from Zuck five years ago, and two, what [40:09] What would the model of Zuck if somebody were to imitate it, not that you're necessarily suggesting they do? Yeah, I don't think anyone should.

[40:15] I think Zuck... [40:17] changed because... [40:19] He... [40:21] always was a certain way. He always felt he had to be something. [40:24] for he'd be he always felt he had to be everything for everyone which means you're nothing for no one yeah [40:29] And, [40:30] He then saw a tide turning in which, for the first time, [40:34] ever he wasn't public enemy number one in technology. [40:38] And he said, you know what? What caused that? [40:40] Was it SPFs? [40:41] I think it was... [40:42] uh, [40:44] I think it was a combination of SBF. I think it was a little bit Elon. I think it was a little bit even Bezos. [40:51] And I think that... Also more conversation about Apple, TikTok stuff. Yeah. And I think he just saw a window to be like, I can... [41:00] finally just be me and [41:05] Yep. [41:05] I can try to be something for someone. I think the secondary thing too is... [41:09] He became a dad. [41:11] He got a little older and [41:13] I think for me, like I am, as you know, I'm like far more softer of a person than I used to be. You're far more comfortable as you get older. Yep. [41:22] And I think he probably also went through... [41:24] the [41:25] rise of the whole COVID thing, then getting dragged in front of Congress. [41:29] having [41:30] Politicians hate him. [41:32] the election stuff whatever then came down stock sold off [41:36] He got people trying to tell him how to right size his business. [41:39] And [41:40] then he kind of [41:41] just crushed it and he was like you know what this is like this is what we do and also got kind of days not days like mocking ed but

[41:48] AI was very good for his business. Yeah, for sure. [41:51] But yeah, like that's the funniest part about Duck is everyone talks about how incredibly right sides of business. And Brad Gerstner wrote this letter and Altimeter saved the company. And it's like. [41:59] Six months later, he just kept lighting money on fire in reality labs and nobody cared because he had AI. [42:04] Okay, so that's the juxtaposition. We talked about modeling Elon. People model Elon all the time. What actually are the things... [42:11] either... [42:12] The whole time or most recently with Zuck that you think... [42:15] are being modeled. [42:17] Or are being inspired to maybe even. [42:19] I think Zuck has... [42:21] Zuck and Elon both have the intense ability to go... [42:25] to the metal and up 30,[redacted address] on the way up. That is... [42:31] intentionally [42:32] antagonistic. [42:33] And not antagonistic in a way of, like, necessarily negatively, but he's intentionally trying to paint the maximally ambitious vision of the future. [42:42] effectively never hits timelines. [42:44] uses this way to create this cult personality. [42:47] Zuck is the opposite, right? He's just kind of like... [42:51] Yeah, I think Oculus is great. I think the AI stuff is really important. We've been buying GPUs and we didn't know exactly what to do with them, but we knew they'd be important and that's why we have more than everybody. Could you compare... [43:02] Maybe this is wrong. Listening to this, it sort of [43:05] You could compare Zock to maybe like Larry Page. [43:08] The counter argument here would be [43:10] And obviously the environment shapes the player, but the counter argument would be Zuck has this amazing business model that he can fund his R&D with. And Elon was doing the R&D as the business.

[43:18] Yeah. [43:19] Could you even do the... [43:22] If you didn't have this amazing cash cow core business internet software advertising thing, like... [43:28] Does that strategy even work? [43:31] Uh... [43:33] Yeah, I think that's a good question. [43:35] I think it doesn't matter anymore because there's plenty of people who will fund it. [43:38] People, Mark Andreessen, many others have asked, why don't we have more Elon's? Why don't we have more Zucks? Maybe is a good addendum to that. It's not cool. Zucks being Zucks. Think about his life. [43:48] Well... [43:49] Like his life is awesome unequivocally. The thing about his arc [43:53] Oh, sure. And maybe it sucks being Elon, but like, that doesn't feel like a... [43:58] One question is, frankly, is just, is that the right question? Like, do we actually have... [44:03] And then two... [44:05] Yeah, it's okay. So why don't we have more? Let's make a slightly longer list. Add in Sam, add in even Patrick Coulson or a number of these guys. [44:15] Is your notion broadly that these people are Tom Brady, that they're actually so exceptional even amongst the exceptional, or is your notion that there's some other lagging reason – [44:25] due to how financial markets work, venture capital works, or how [44:28] Good we are at... [44:29] spreading this across the whole globe versus just in certain markets? Yeah. Is it a funding problem? I mean, all these things are power laws, right? So like a subset of people matter for building companies. Most of them don't. And that's fine. I think that the... [44:45] Other thing that I believe more and more is that at each –

[44:49] technological [44:51] shift and we can say that's [44:53] early computing into web. And we'll say like web one and two kind of can blend together in certain ways. So the founder profiles were different. [45:01] you have like a reshuffling of the cards of what matters. [45:04] And. [45:05] you had intense operational excellence, you had [45:08] you had [45:09] adults basically in that very very first version and then you had a [45:13] intensibility to [45:15] ship, move quick, move fast, break things is a simple kind of canonical way to think about that next... [45:20] thing [45:21] AI could be viewed as the next version of that. You reshuffle once again. [45:25] and the reshuffling [45:27] might mean a new type of framing for what these people look like. My view that is [45:32] that that probably actually looks like a far more of a push on [45:36] How do people most creatively think about the world and the people who can spike creatively the most will be the ones who capture the most value? [45:42] where I would argue... [45:45] Elon thinks about the world very ambitiously. I actually don't know if he thinks about it that creatively. Back to the cosplaying science fiction thing. [45:52] I don't think we're going to call him non-creative. Yeah. Yeah. But somehow it's okay. Um, [45:58] And so you have like [46:00] Yeah, you have novel primitives that come out and people have to... [46:04] create those primitives. I think that might mean [46:08] kind of why I'm against this idea of pattern matching to these existing founders who have [46:12] been building for the last decade plus. I just don't know if that will... [46:17] if building in this next wave will necessitate the same things.

[46:21] Because you don't have capital constraints in the same way [46:24] You have way more leverage. [46:27] You have this... [46:28] overhanging thing of artificial general intelligence, ASI, that you might be able to pull on, and it also might destroy [46:34] the 90% down of uncreative people [46:38] And so if you think about the power law of the next generation of humans, like I personally think creativity is one of the is. [46:46] Maybe the core thing. He's creative in the sense... [46:49] being able to jump to the non-Stymorphic? Is it sort of totally orthogonal to anything that exists? Steve Jobs is wildly creative in the context of the personal computer and what it could be. [47:00] He was also part, frankly, of bringing it to reality. But how do you... [47:05] Are there even a set of words or ideas that you would associate with creative community a ton of things, especially in... [47:09] Maybe like – [47:12] Like immaculate conception. Just this idea of... [47:16] Not. [47:17] I think it's either being intersectional or translational between multiple areas and bringing together a novel idea out of that. [47:24] Or it is... [47:26] Yeah, truly. I mean, again, the memes, like I hate first principles as like a meme, but that idea of like... Is this an aid or are there things people do that make them more creative in this sense specifically? I mean, maybe to the first line, it's sort of having multiple inspirations that are not... [47:42] Overlapping. [47:43] Yeah. [47:44] Um... [47:46] Even in the founders, perhaps, that you work with, the runway guys, as an example. Yeah, I think so. I think it's people who are uniquely obsessed about a few things. And that obsession of a few things is actually the...

[47:58] the next step versus... [48:00] The [48:01] One person, like the person who's obsessed with AGI. It's the old Navalism a little bit of like, he can be number one in the world, but he can be top 5% in new things. Some of these things kind of hold true. Yeah. Yeah, I think that's probably... [48:15] That's probably the framing. I think Cristobal and Alejandro Anastasis at Runway are all like that. They have... [48:21] wildly divergent and [48:23] really interesting [48:24] kind of core interests that they're obsessive about [48:26] I think like Alex at Wave is like that. [48:28] And I also think that that's... [48:30] Sometimes... [48:31] where like [48:33] Anastasi says this thing where he says, [48:35] pre-training is science, post-training is art, which is like... [48:38] Yeah, the science of throwing a bunch of data into something is pretty well known. And then like... [48:43] how you make the model once it's a model kind of become this next order thing. [48:47] is an art, and I think that [48:50] in a world in which if you just assume there's just an AGI model or a few AGI models, [48:55] That statement actually might be far more profound than people think. [48:58] I'm forgetting his name, one of the main open AI guys, Karpathy. Under Karpathy talked about hallucination being the feature, not the bug. Yeah, yeah. I think it ties into this a little bit, especially if things get really wildly complex. Yep. [49:10] What is venture capital for? [49:13] being the riskiest dollars in... [49:17] the riskiest capitalistic dollars in the world. [49:20] and like pushing, allowing people to take risks that no one else would underwrite. [49:24] It's just a risk engine, basically. [49:26] Yeah, with the idea that... [49:28] In aggregate, the thing works to create way more value than

[49:32] It consumes. But it's sort of less about serving your LPs or whatever. Like, in a very structural level, it is sort of a thing that allows... [49:40] a level of risk that probably wouldn't be able to be taken otherwise to be taken. Yeah. [49:45] I think so. [49:46] You've talked about. [49:47] particularly in the job of being a VC, you've talked about [49:50] this maybe slightly contrarian view of, [49:53] A lot of VCs, especially as they get older, they talk about pattern matching and wisdom and experience and all these things. You talked about sort of fighting decay. Mm-hmm. [49:59] Can you expand on that a little? [50:01] I think all venture firms and all investors decay over time. [50:04] And maybe all people do. [50:07] The... [50:08] way in which [50:09] You... [50:11] Everyone wants to believe they compound over time. [50:14] We all just naturally get better as we get more and more things in. [50:17] And, [50:18] It's like a beautiful idea. And... [50:21] We say that and then we talk about how everything takes work. Love takes work. Friendship takes work. [50:27] And [50:29] For some reason... [50:30] VCs are like... [50:32] No, it actually just gets better and better and easier and easier. And, [50:35] even though there's more money coming in, there's more competition, whatever, even though I work fewer hours, [50:39] I'm going to be better... [50:41] without changing any of these inputs because the inputs just feed themselves to get better. [50:45] And I think that's just... [50:47] fundamentally makes no sense. And I think as you do venture, [50:51] You have a few things. [50:52] You have the [50:54] total lack of feedback loops for a long time. And that wears on you a lot. [50:57] You... [50:58] send out many millions of dollars, sometimes billions, [51:01] And you just kind of stare into the abyss and hope you know what you're doing.

[51:05] And, [51:06] Year four, you're like, [51:07] "Oh fuck, I really don't know if I know what I'm doing." [51:09] Maybe you're eight. You're like, oh, I definitely know what I'm doing. Maybe you're 12. You're like, oh no, I really don't know what I'm doing. And I thought I did. And I'm 12 years in and [51:17] What am I going to do with my life otherwise? So you have that. And I think that creates some... [51:22] just natural decay and [51:24] headspace and you have to fight against that. You have to actively think about what are the inputs that matter, [51:29] you know, how do I, what are the small signals I can take that allow me to know if I'm doing better and then kind of push and push and push. [51:35] then you just have the responsibilities to pile up. You... [51:37] companies don't die as quickly as you fund them. And so as you do the job for longer, you have more companies that you have to spend time with and that you have to help in more founder relationships. [51:47] and [51:48] Again, in friendship, in relationships, in anything. [51:52] We all know that there's like kind of some level of maximum capacity to be the best person to a subset of people. [51:58] There's all these numbers around it, whatever. [52:00] But in venture, you, for some reason, can just keep adding portfolio companies and it doesn't matter. And so that pulls on your time. [52:07] And then third is like, [52:09] the world meaningfully changes and the world changes from a demographic perspective. [52:14] you can become [52:16] I'm 34 now and I talked to a 21-year-old and [52:19] Their internet is different than mine. Right. And I try to understand it and... [52:24] But it's different. [52:25] And there are 65 year olds doing this job who don't get my internet. And so [52:30] And the cohort of people building changes. And so your network has to be refreshed. [52:33] And people think that their network is going to continually work and –

[52:37] I don't think it really works outside of maybe Founders Fund. [52:41] And [52:42] All these things are just, you need to just continually figure out what is the [52:46] What are the things that need to be worked on or things need to be cut? That's the other thing that VCs don't like to talk about is they don't like to [52:53] even speak of this idea that at some point you might go to a founder and say, hey, this was great. [52:58] If you ever need anything, call me, but... [53:00] I'm not your guy anymore. And I think that's like really... [53:04] Something that, for whatever reason, we're so... [53:07] precious about these relationships and they are precious and do mean things but [53:12] I don't know, I've had that conversation with some founders before and none of them have ever... [53:16] been like. [53:18] how could you do this to me? They're like, yeah, I get it. I'm either a massive company now and I'm doing my thing and I love that you helped us get in business or... [53:25] I understand I'm just thrushing, I'm going to die soon, and like I... [53:29] I don't really know what to do. And so, um, I'm going to call you randomly, but like, [53:34] I'm in my own world right now. [53:36] And then the last thing probably is like venture firms don't like to scale. [53:39] the rise of the solo capitalists. People think that you can do this job. [53:43] Is there any other... [53:45] Asset class. [53:47] in which people do it as a side project besides real estate. [53:51] You would probably know better than me. I mean, yeah, outside of personal, like commercial side projects are pretty rare. It was people give you money. Right. You charge them for it and you do it as a side project. [54:04] Yeah, something was standardized somewhere along the line commercially in Silicon Valley that sort of said, I'm going to pay for network access. Yeah. And so I think that's more or less how it happened. Yeah. Yeah.

[54:14] And so. [54:15] I just think all of those things means you just have to work. And I think life needs to be worked on. [54:21] I don't know, venture investing generally is part of that. [54:24] You were part of a... [54:25] Ultimately, it was sort of like a lineage with Compound. Compound, first fund you raised, but kind of came out of a fund before that. [54:34] and you're building a thing that [54:36] has a number of people beyond you [54:39] How do you think about [54:40] I don't want to get into necessarily every aspect of your life and we'll get more personal, but [54:44] legacy for this place and this thing you're doing in your building, what it... [54:49] Maybe even in the context of the answer you just gave, like I'm curious how you think about what, [54:53] what you want this to be and what you want the legacy to be. [54:56] I hope... [54:58] Anyone who works at Compound never works anywhere else again. [55:01] And if they want to, that's cool. And that is a... [55:05] hope on both sides. It means I hope that they're great enough that they deserve to continue to [55:10] own more and more of the fund. Everyone who works here is compensated with upside in the fund. [55:16] and two means that they want to do this job do it here whatever. [55:20] And [55:21] The only thing that I probably... [55:23] I would say I'm a dictator on is... [55:26] thesis-driven research-centric investing is what we do. [55:30] It's all these things we've talked about of... [55:32] building futures we believe in, helping people fit within them, etc. [55:35] And all I care about is being really great at it, not... [55:40] The biggest... [55:41] And... [55:42] I don't think about anything else besides that.

[55:46] Focus. [55:47] I think also on some level that [55:49] If anything is going to enable it to exist for a long time, it's probably that. [55:52] Yeah, hopefully. Zooming out from B.C., [55:56] It feels like we're living in a time of future shock. You said, I think it was on a podcast recently, time is collapsing. [56:03] I think so. Maybe I shouldn't be crediting you. I thought that was pretty incredible. [56:09] I think you maybe used it in the context of sort of the rate at which companies are accelerating beyond the science project thing or whatever it might be. But I also think it sort of feels incredibly apt for the entire moment we're in. Yes. [56:21] I think this is also a quote from you. This dynamic feels like a moment in time, but as time goes on, I increasingly believe it to be a more permanent societal shift for volatility and, [56:30] and uncertainty are constants. And that was, I think, in the context of financial markets, but also feels pretty apt [56:36] And then finally there's this... [56:38] Weaver Y blog post that people probably have seen [56:41] relating human progress and time. Yes, yeah. [56:45] which is an audio podcast. It's basically like time is sort of, [56:48] It feels like it's on this linear curve, and then it basically just starts going straight up, and we're the little stick figure staring up. Yes. [56:55] Is this just... [56:56] our lives now? Is this where we're going? Do you expect it to continue to ratchet at that clip? Or is there some degree of [57:05] every [57:06] Group of people is always going to think things are changing radically. We're staring at this AI problem. It's like, are we just as likely in two or three or five years to be like, oh, wow, we got a little overeager?

[57:16] It's possible we... [57:17] feel that way, or we say that, but I don't think that we will feel... [57:22] at any point in our lifetime that the world isn't changing really quickly [57:27] I don't think like I've gone back and I've tried to do the whole reading old articles and papers and all these things about how people felt in different moments in time. And, [57:35] People have felt... [57:37] negative on things over and over again and fearful of things over and over again. [57:42] I don't think that they've felt... [57:44] the vol, like the volatility of things. [57:48] as much as I think we will. And I also think... [57:51] There are... [57:52] kind of like first time moments in our lives that is going to happen at a rate that's kind of crazy and might happen over and over again. So [58:00] When I say that, [58:01] The examples I sometimes give is very confident we'll have another pandemic in our lifetime. Like very confident. [58:08] how quickly it's solved, all these other things. [58:10] depends on when it happens. [58:12] uh, [58:13] I think that [58:14] Humans are volatility machines and... [58:16] Thank you. [58:17] it's beautiful that we're really volatile it's like the best part um [58:20] But [58:22] I just don't think that if you keep introducing all of these [58:25] things that push what it means to be human. [58:29] And... [58:29] that allow us more... [58:32] rocket fuel to figure that out, you're not going to have some really long-tail outcomes that people thought were [58:38] not possible happen far sooner than they think. So something that we believe is [58:43] model any long tail outcome in the world and it's probably like...

[58:46] an order of magnitude more likely that it happens. [58:48] And that's [58:49] I think we'll have, I don't know, war relatively soon. I think there's a bunch of things. [58:54] There's a... [58:56] idea that's been [58:57] talked about across a number of different contexts. [59:00] One is the vulnerable world hypothesis. Peter Thiel has been on this bit for a very long time. [59:04] about the sort of fear of the one world government. And really, as I understand it, that's actually a response to this notion that I think you were talking about, which is that as technological... [59:13] change increases. [59:15] The... [59:16] rate at which people around the world will have access to technology. Effectively, the N of people with sort of world-changing technology goes up. [59:25] There's a few... [59:26] pieces to pick apart here. One, I think that [59:28] get to the Teal stuff a little bit, I'm curious for you on is [59:31] Do you think ultimately technology is more likely to be a consolidating force? [59:37] towards sort of like infinite leverage for an individual with kind of absolute power, absolute leverage, or the other end of this, which is that like, [59:44] It's the ultimate kind of democratizing... [59:47] individualistic. [59:49] Force. [59:51] Or maybe that's the wrong question. Maybe it's about the tension and the volatility. Yeah, I think maybe... [59:57] like we [59:58] talk about this idea of technology pushes societal change and it kind of oscillates. [1:00:02] I think it's probably that. [1:00:04] really like... [1:00:05] viewing it. [1:00:07] in such black and white lenses, because I think that [1:00:10] It just... [1:00:12] creates zealots on either side and [1:00:14] I just don't really think that's how the world will work.

[1:00:18] So I don't. [1:00:19] I don't know. I just know that it will, there will be periods of time in which people feel like it's over a lot more than it probably will be. And I also think [1:00:27] Bye. [1:00:28] the core... [1:00:29] thing that humans are good at is [1:00:31] we solve the problems that we create for ourselves. [1:00:33] Yeah, I always like to say... [1:00:36] we are simultaneously like we hate change. Humans hate change more than anything. And then we're also unbelievably good to do that thing to change. [1:00:45] In speaking more kind of, [1:00:47] less about how people feel about volatility and change, being more kind of absolutely about actual risk. Vulnerable world have others you've written about, this sort of notion that... [1:00:55] You're picking balls out of a vase, white balls, gray balls, good, bad, or good, [1:01:00] gray, whatever, [1:01:01] You pull a certain ball and it's like the end. It's the bad pandemic or it's the nukes destroy everyone or whatever. [1:01:08] I'm not sure how interesting it is to explicitly opine on that, but when you think about [1:01:12] that type of risk when you think about ethics and morality [1:01:16] you've talked about this a little bit in the past, but how much of that is a calculus broadly? And then two, especially in the context of sequencing and inflection points, [1:01:25] How much are you and your companies thinking about this to the extent that [1:01:28] even if it's not necessarily existential for humanity. Biotech, for example, [1:01:33] "They certainly could have." [1:01:34] Push. [1:01:35] what it means to be a human and the rate and degree to which that happens like [1:01:40] Is this a... [1:01:42] something you're feeling on a moral and ethical level? Is it something you're just... [1:01:46] trying to be wise about? How do you relate to this type of thing? It could be paralyzing for a lot of people, I think, and it probably will be. I think it will be. I think that

[1:01:53] We've done a really good job of rate limiting [1:01:57] people in [1:01:59] ability to act as independent agents negatively and like [1:02:03] We've done that by... [1:02:05] recognizing things that could be obviously bad and trying to safeguard them. [1:02:08] A simple example of this is [1:02:11] DNA synthesis, for example. You could [1:02:14] There's a few major providers you can [1:02:16] order these strands, you can put them together and create a new myoweapon. [1:02:20] Um... [1:02:21] inject yourself go out into the world kill a bunch of people [1:02:24] people try to [1:02:25] coordinate those things such that [1:02:27] If I order a bunch of strands from a bunch of different companies and it looks like a sequence of something that would... [1:02:33] be a bioweapon, they're like, hey, FYI, someone's been doing this. [1:02:36] It's very bad. It's not nearly good enough. [1:02:38] But those things will decentralize over time, and then it won't matter. Back to the earlier point about sort of... [1:02:43] One broad shape of technology is just that everyone gets nukes. Yeah. Pick the thing. Yeah. But can we do that in a way that's slow enough? Yeah. And yeah. And I think. [1:02:52] The problem is this accelerationist viewpoint, which is, [1:02:57] It almost doesn't matter because someone will keep pushing forward. [1:03:01] And you just want to make sure that... [1:03:04] the people pushing... [1:03:05] for the most, at least have [1:03:08] some sort of awareness of risk. I think this is what, sorry, [1:03:11] I actually think this is quite interesting. You want to make sure that the people pushing forward is most like, [1:03:16] This is the whole... [1:03:18] challenge, right? On some level, it's like, oh, if we had the philosopher king, [1:03:21] Yeah, if we're a feed. [1:03:23] Or do you think people are overrating that part of it? Just trust Sam Altman. He seems to have a pretty good idea. He's got the god machine and the money in the Mithyran.

[1:03:35] I think it's why it's important that there's more than one Sam, for sure. [1:03:38] But I think the... [1:03:40] alternative. The most tangible example is [1:03:43] the stopping AI progress thing, [1:03:45] is so dumb because... [1:03:47] like China won't stop AI progress, right? And we don't want that to happen. [1:03:52] And that's a very specific view, but... [1:03:55] I just don't think you can stop. [1:03:57] this runaway freight train that is science and technology at this point, all you can do is hope to contain it and we might pull the black ball. What hope to contain it is about ultimately sequencing. [1:04:08] It's about... [1:04:09] I actually don't. And leadership. [1:04:10] I think sequencing is... [1:04:13] is wishful thinking in the sense that [1:04:16] you are seeing this now maybe play out at OpenAI, which is... [1:04:20] The right sequencing would be to continue to give lots of compute to the safety alignment teams that have all left the company. [1:04:27] But the actual sequencing is build the thing with enough thought on safety such that... [1:04:33] It's okay, and you have plausible deniability maybe, [1:04:36] And maybe they are doing it right, but a bunch of people who were on that team seem to not think so. [1:04:40] And... [1:04:42] So I don't think it's sequencing. I think it's just awareness and... [1:04:45] broader awareness, which is [1:04:46] Everyone in the world at least knows... [1:04:48] that there is... [1:04:49] some long tail risk that [1:04:52] Sam or Dario or Zuck or whoever, [1:04:56] Could create. [1:04:57] an evil version of the God Machine. [1:05:00] And so they're all at least watching and they're all at least paying attention. [1:05:03] It's just a big tragedy in the comments.

[1:05:06] That would be a pessimistic view. Yeah, I mean, it's possible, I guess, but... [1:05:10] I think the... [1:05:11] The alternative is just... [1:05:12] You just don't get progress then if you try to push it. [1:05:16] Would you say you're broadly optimistic? [1:05:18] As a person? [1:05:20] around this stuff. [1:05:22] Yeah, I think so. I think that [1:05:25] Because you have to be. [1:05:27] No, I think that [1:05:30] So one of my first jobs was in this building, and I took the path every day. [1:05:35] And one day I'm writing the [1:05:37] elevator down and with this guy who works in the building i'm like what do you do and he says [1:05:41] Oh, I work on the safety department for the path. Um, [1:05:44] I help. [1:05:45] build the structures and the tunnels that if a train [1:05:49] blows up, the tunnels close off so the whole thing doesn't flood. [1:05:52] And I look at him and I'm like, what's the... [1:05:54] Like that. [1:05:56] Can that happen? Ethan's... [1:05:57] I'm 21 years old. It's my first job. [1:05:59] He goes, do you really want to know? And I go, I guess so. [1:06:03] He just looks at me and goes, you know, [1:06:05] If anyone really wanted to do it, they probably could. [1:06:07] And when you do this job for long enough, you just kind of realize when it's your time, it's your time. [1:06:11] I'm sorry. [1:06:12] And... [1:06:13] I think that that [1:06:15] That has stuck with me forever. And I kind of think that [1:06:19] People are just generally net good and [1:06:23] I do think that [1:06:25] for some reason, the only thing I'm uncomfortable with right now in technology, where a few years I wasn't, [1:06:30] Years ago, I thought it was incredible that we had these leaders that were willing to, in a world shutdown... [1:06:35] basically scale up like bezos and say hey we got you we're going to deliver stuff to your house we're going to complete you whatever

[1:06:40] And I thought it's incredible. We have these people who really care about humanity, not just margins. [1:06:44] And there's a jaded view on that, sure. [1:06:47] I think now in technology, though, there's definitely a feeling of... [1:06:50] People care more about power than ever before. And that really bothers me. [1:06:56] We could spend a lot of time talking about it. [1:06:58] apocalypse scenarios, maybe in a slightly more zoomed in sense. [1:07:03] You would know far better than me. It seems that the biotech stuff is starting to finally... [1:07:07] Get here. [1:07:09] Obviously, this gets into moral and ethical areas that are interesting. [1:07:13] Maybe one, like, [1:07:14] How soon do you think we're really going to face some real questions around like what it means to be human or superhuman? And then two... [1:07:21] Do you think that there's anything that people should be [1:07:24] particularly open to, prepared for, or do you think that we need to be prepared to, like... [1:07:29] update our model of what a human is? Or am I getting a little silly? I think it's a little... [1:07:36] Early for that. Okay. I think the best thing is that we will look up. I fundamentally believe a few things on this that I don't think people... [1:07:44] Update enough. [1:07:45] I think... [1:07:47] We, and definitely people younger than us, [1:07:49] 100%, 120 plus life expectancy. [1:07:53] I think we will live much longer than we anticipate. [1:07:55] Think. [1:07:57] A large portion of the things that kill us today will not kill us. [1:08:00] 10 to 15 years from now. [1:08:02] I think that [1:08:04] There are... [1:08:05] outside shot to the things that matter the most to us as people, which is [1:08:10] Maybe like making families as an example.

[1:08:13] And the biological lock will be... [1:08:16] defeated in the next 15 years? [1:08:18] And I think if you line up all those things, [1:08:22] you have a very different view of what life should be. [1:08:25] and what life should be, both from a [1:08:28] how do I want to live my life? And, [1:08:31] What do I need to last in my life? [1:08:34] And so I think those things are interesting. I think the second thing I would say that's more applicable. [1:08:41] is [1:08:42] For people who have loved ones that are slightly older than them, [1:08:45] there is probably a meaningfully [1:08:48] different way in which you should think about how you want them to approach their healthcare and their biology. [1:08:53] over the next five to seven years. [1:08:55] Because a lot of what kills us, outside of getting hit by a bus or whatever, is... [1:09:01] you could just say simplistically is like time. [1:09:04] And I think that [1:09:05] we are getting pretty good if you get the early wave of things to not have those things kill us. [1:09:10] um and so the early way of things me catch something early yes [1:09:13] And I think that [1:09:15] In general, when we think about how us as millennials or our parent, anyone, [1:09:20] how much we spend in America as a percentage of our income on – [1:09:24] are making sure we don't die. [1:09:25] It is meaningfully lower than it should be. [1:09:28] and [1:09:29] I was talking to a friend the other day whose dad has had a heart attack. One of our other friends' dad [1:09:34] randomly passed away last week. [1:09:36] And... [1:09:37] The main thing I was telling him as he was asking me all these questions about how to help his parents is I was like, you should think about what percentage of your net worth you'd be willing to spend.

[1:09:45] or of your annual income or something. [1:09:47] preventative care on preventative care to not to to not die or for your parents to to hopefully you know live better [1:09:54] There's always been way more pop science, humor-man type stuff. Is there... [1:09:58] a specific... [1:09:59] or general frame, specific person, publication, et cetera, or general frame that you would recommend people at least go to – [1:10:07] If you're like a level seven, other people are level nine or ten, somebody who's a level one, like what is the sort of basic way to get a little smarter on this type of stuff? Or maybe even be more up to date. Yeah, yeah. [1:10:18] It's really hard. [1:10:20] I think the beauty about all this stuff, and this is why we're so obsessed with biohacking, [1:10:24] is that you can effectively... [1:10:26] go to something like consensus.app, which is this research paper search engine, and you can ask a natural language question and it'll synthesize all the research for you and tell you what's good, what bad. [1:10:35] And you can put in papers into chat GPT Cloud and ask questions. [1:10:39] explain this to me like I'm 12. [1:10:40] You can have an understanding of science in whatever degrees of difficulty you want. [1:10:45] for the first time ever. [1:10:47] anecdotally I've seen a lot of people talking about clodinopinant and chai gbt actually helping them materially with various health things yes yeah it's crazy and I think that [1:10:55] That [1:10:56] So there's that. And then there's if you want to go to level three. [1:11:00] you then start to think about what are the indications in your health care, right? [1:11:05] And you just... [1:11:06] I think that the thing most people don't know is like there's clinical trials going on everywhere. [1:11:09] And once something reaches a certain level, [1:11:12] Often you can just ping the people who are running the trials and be like, hey, I want this drug, and they'll just ship it to you if you pay for cash.

[1:11:16] What about collecting data and information on yourself or your loved ones? [1:11:20] Um... [1:11:21] Are we at a level, aside from the sleep trackers and maybe the glucose, is there anything that you would expect to be... [1:11:28] Maybe it's Apple. [1:11:29] or specific [1:11:31] People get these full body DEXA scans. Yeah. [1:11:33] I think that stuff is great. Like the full body MRI is great. I think all those things are good. If you're over 40, I think realistically, if you're not, you don't need to, [1:11:39] The problem with all these things is like the action, the actionability of it's not great yet. [1:11:43] The thing that terrifies me, I have a heart condition, I have a pacemaker, I've been through this system for a long time. [1:11:50] The thing that terrifies me most outside of [1:11:52] One day my heart failing and I don't make it 220, which would suck. [1:11:55] is like, [1:11:56] the brain stuff and how little we know about it. [1:11:59] I think we're going to start bending that curve in the next 10 years, but like, [1:12:02] outside of that, [1:12:04] I don't know, next, like, [1:12:06] Just... just... just... [1:12:08] If you are of any means, and you can... [1:12:11] you probably are not spending enough money trying to understand your loved ones or your own health. [1:12:16] You mentioned biohacking briefly. [1:12:18] We've chatted about this recently. What would you say to somebody who... [1:12:21] Bradley thinks Brian Johnson is a rich... [1:12:24] loser or is doing something morally concerning or ethically concerning? And maybe beyond that, what would you say to somebody who thinks... [1:12:31] even trying to push the boundary on longevity, is, [1:12:34] isn't what we should be doing. [1:12:36] I don't think it's about pushing the boundary on longevity to the point of, like, we should live forever. I think it's just, there are things that kill us, we should stop them from killing us. And... [1:12:43] eventually our bodies will shut down and die, our heart will fail, our brain will fail, whatever, it'll happen. So I'm not of this belief of...

[1:12:49] you know, I'm, [1:12:51] some tech bro who wants to live forever. I'm very against that. [1:12:54] I think what Brian is doing is the coolest experiment [1:12:57] a single humanist in a long time, which is he's taking a bunch of money, lighting it on fire, [1:13:01] experimenting on himself and saying, [1:13:03] If you want to see inhuman R&D... [1:13:06] enjoy and also doing it in a very data-driven way and also in a way that is not... [1:13:11] highly prescriptive. He's not shouting from mountains being like, "You're an idiot if you don't do this." He's selling supplements. People are upset about this. He is. He sells the Blueprint stuff [1:13:21] But it's very different than the next order stuff he's doing. If he starts pushing like gray market peptides he's injecting himself with or something, then I would have a little bit more of a problem with it. [1:13:30] Because I think that so many people don't understand the risk of – [1:13:34] doing that type of stuff and how little data there is on a lot of it. Some of it's really good. [1:13:39] But I just don't know why. I don't know how someone can be super negative on him. I think that... [1:13:43] Thank you. [1:13:44] There's this act in Montana called the Right to Try Act, which passed last year, and it's one of a few different ones. And the main thing is that if a drug gets to a certain level of clinical trials and you want to do it, [1:13:56] You can. You can do that to yourself. And there's a reason why that exists because there's tons of people in rare disease communities that [1:14:03] see these drugs and they might be dead before they get through the full clinical trial process. And they're like, you know what, fuck it, I would inject this thing into myself. [1:14:09] If it means I have a chance of living... [1:14:11] And there's tons of data that says it's really good. [1:14:14] And this state is basically, and there's a few others, it's basically like you have the right to do that. [1:14:18] And I think that movement is quite big.

[1:14:20] And I also think it's pretty funny that [1:14:23] That we are really touchy about, but we're happy to accelerate a lot of other things in healthcare. And so... [1:14:29] Um... [1:14:30] I'm pro-Brian Johnson. It sounds like one view would be that the biohacking stuff is actually, especially if it's an N much bigger than one, is how you could start to make material progress on some of these areas that have been really bureaucratic or slow. I mean, human challenge trials is the other big one. Like, how do we just get people to... [1:14:44] sign up, but there's all sorts of [1:14:46] ethical questions about that. [1:14:48] Speaking of sort of closer to the near term, [1:14:51] You've talked about like minimum viable intelligence in the AI context. I'm curious if there are any areas, and this has to do a little bit too with... [1:14:58] the skeuomorphism stuff and so on like [1:15:01] Or even the notion like we just stopped with chat GPT or GPT-4 or whatever it is. Are there any areas specifically that you're particularly amped on from a product standpoint? [1:15:10] It's therapy. I'm obsessed. [1:15:12] Huh. Say more. [1:15:14] Today... [1:15:15] therapy and this is theoretical or you're actually doing this [1:15:18] I have tried everything. I've tried all of them. I've tried all the apps. [1:15:21] I am... [1:15:23] obsessed with finding someone who is more creative than me to push the limits on what this would look like in a non-skeemorphic way. [1:15:29] Okay. And I've been spending a lot of time trying to find those people. What do you think that person... [1:15:33] What are you looking for in that person? [1:15:35] Someone who has a unique view on how... [1:15:40] a certain group of people, most likely Gen Z, [1:15:43] will navigate their mental state and [1:15:47] to [1:15:48] Back to the first part of why I got there is...

[1:15:51] Today, therapy is this super synchronous thing that you show up to [1:15:56] for an hour [1:15:58] 45 minutes, whatever. [1:16:00] You have to be in the right state of mind to do it at that time. [1:16:03] You have pretty minimal context with your therapist. I have been seeing a therapist weekly for a long time. Still, though, like last week, she was like, I never knew that about you. I was like, great. [1:16:15] tens of thousands of dollars. There it is. And, [1:16:17] I think that those... [1:16:18] And there's more and more things that will signal what matters, what's going on in our head. [1:16:25] And then lastly, just this idea that [1:16:27] our brains are [1:16:29] more... [1:16:30] chaotic than ever before as whatever quote you had read of mine from before. [1:16:35] And... [1:16:35] That probably means that rumination happens at far different times than humans are available. And so you could theoretically start to spin up like a late night therapy service. That's maybe just an interesting business where you just have people in different time zones service people in New York City or whatever. [1:16:50] when people are at home at 11 p.m. freaking out about something, is do therapy then? [1:16:53] So I think there's all of that. And then... [1:16:56] There's just this idea that there's all these... I think that people think about their mental health and their conversations as this... [1:17:03] almost always on progressive thing and you kind of [1:17:07] tap in and out of you tap in and out people into that and i think about this with my friends [1:17:12] a bunch of... [1:17:13] My group chats, so you're in some of them with me, are... [1:17:17] You pop in for 15 minutes. [1:17:19] You talk about something really quickly, everyone dissolves.

[1:17:23] I think a lot about the product mechanism of how you can share your location for an hour. [1:17:27] On your phone? Yep. [1:17:28] And I kind of wish you can share your group chat for an hour and maybe you pull in. [1:17:32] like your therapist, your AI therapist, and your other friends, and they all interact with it. And I think there's a lot of – there's some version of these things, and it's not – [1:17:40] Probably doesn't end up being called AI therapy. It's probably called something else. [1:17:44] probably has some different mechanism, but I just think... [1:17:47] It's an undefeated trend that people want people to be inside their head. [1:17:52] There's a bunch of components of this. One is way more non-skeomorphic emergent stuff like you decided at the end, which is like social stuff or new paradigms. There's also a view that sort of says like, [1:18:01] if the average 21-year-old just talked to ChatGBT with like maybe one – [1:18:06] aided starting prompt. Yes. This would be 90% there. [1:18:09] Where do you think... [1:18:11] Do you think that's like probably true? Have you been able to use chat, DUPT or Claude in this way in a way that's pretty productive? [1:18:16] Or is it actually about... [1:18:17] No, some new paradigms need to evolve. It needs to be fine-tuned or whatever. Yeah, I haven't found it that productive because I don't enjoy... [1:18:25] the [1:18:27] One, I don't enjoy the inherent like I'm in the chat GPT interface talking to this thing. [1:18:31] and I've done a bunch of custom prompting stuff, I think that [1:18:35] Also, these... [1:18:36] which again, this could change and this maybe is going to be [1:18:39] Sam has a view on this that everyone will have the ability to pick and choose the personalities of their models. [1:18:45] Similar to kind of how friend thought about this, where they're like, your friend is, it dies when your friend dies type thing. Yeah. [1:18:50] I do think it's kind of interesting, this idea that why people like their therapists is because it's their therapist, not because it's a therapist.

[1:18:56] And so there's a lot of... [1:18:58] Ways in which that doesn't work today. [1:19:00] It probably gets 90% of the way there. [1:19:03] And then you use something that gets 100% of the way there and you think that it actually only got 10% of the way there. [1:19:08] I think a couple other things to me just seem like, number one, [1:19:11] the inversion of back to friend and there's been a few others, um, new computer, like it prompting you is very different than you prompting it. Yeah. ChatGPT is like this magic search box thing to anything. It's like, it's hard to, [1:19:20] And then the other thing I think is just like, [1:19:22] How do you dump as much context into this thing as possible? [1:19:25] quickly as possible to your therapist point like most of these products today don't do an amazing job of that yeah [1:19:29] If Facebook's reading, not necessarily want this, but Facebook's reading all my WhatsApp chats and has all that context already, maybe it's better. [1:19:36] As we talk about people having too much power over these. [1:19:39] Another version of digital relationship, one of the things that you and I were both pretty excited about for a long time, maybe going back four or five years, was this notion of [1:19:48] The way that changes in animation and kind of new platforms would lead to stuff like VTubing becoming much bigger and much more popular. [1:19:56] You also talked about just broadly the way that animation and IP allows you to have so much more service area with a character or with somebody. [1:20:03] Obviously, it hasn't gone away. B2 being still quite big. [1:20:06] I think it didn't accelerate maybe the pace that you and I expected, [1:20:10] Despite parasocial relationships and streaming and content creating being bigger than ever, [1:20:14] I'm curious, one, where you think it is today broadly, and then two... [1:20:18] It feels maybe particularly apt given that we now have this wave of AI, like Twitter agents, basically. Yeah. Granted, a lot of them are meme, hawking meme points. But it's a whole new level of, okay, Hatsune Miku, Lil Mikaela, whatever. Those were cool. Those were programmed by a person.

[1:20:33] Now you have... [1:20:34] you could just put up an animated character and let it rip. It's Spider-Man, but he makes the rules. [1:20:42] Do you think that we're, maybe is that the inflection point we needed? [1:20:45] I think it kind of [1:20:47] The main problem with the thesis was, the initial problem with the thesis was that [1:20:53] It ignored... [1:20:54] The fact that on a creation mechanism side, you're always competing against humans who can turn on a cell phone. And just the scale of that is so hard to compete against. [1:21:04] That's changing now. Right. And the VTubing stuff is a version of that, but it still isn't as high fidelity. Like, it just wasn't as good. Yeah. [1:21:09] I think secondarily is it. [1:21:11] Kind of is. [1:21:13] Was that first order... [1:21:15] problem approach that I talked about in the deep tech stuff. It was a perfect future and sequencing to it. [1:21:20] And in reality, actually, you have to build it as an equal odds problem, which is... [1:21:23] Just... [1:21:24] the maximum number of ideas in the world. Like quantity is what will read quality for these things. [1:21:29] And, [1:21:30] When you say quantity in this sense, you literally mean personas. Yeah, IP, whatever, right? [1:21:34] And I think that [1:21:36] there are some places that have distribution advantages where it's not the case. Let me give XRB a counterexample of that, by the way. [1:21:42] I think Pixar... [1:21:43] Like if there had been a Pixar or VTubing, would it have been huge? [1:21:46] Yeah, I think so. So the notion is just that it's actually really hard to combine [1:21:51] Until it gets to the UGC point and you can increase the N. Yeah. Or it's a bunch of AI agents. People doing it first party like Brad was or you guys were trying a little bit with Shadow. Yeah. [1:22:00] you have to really nail the storytelling part and the technology has to be good enough. Yes.

[1:22:05] And the other thing is you... [1:22:07] have [1:22:08] materially different budgets than the people who are playing a distribution game in which they're selling the upside. Sure, but isn't that like, I mean, you can basically, maybe like Mid Journey, Runway, Sora, like isn't good enough for video yet. [1:22:21] Yeah, no, I think that will change. My point is, it's just... [1:22:24] you're when you're building a studio, [1:22:27] The premise on all these were they were venture-backed studios, right? And they were venture-backed studios that wanted to... [1:22:32] because they're Avenger-backed, own the entire upside. Right. And when you own the upside... [1:22:36] You can't strike the same types of distribution deals that you can if you sell all the upside, which is what studios do often early on. Right. [1:22:43] And, [1:22:44] The premise was that... [1:22:46] That upside, you could capture it outside of the traditional distribution mechanisms. [1:22:50] It just turns out [1:22:51] You're not competing against Pixar then, but you are competing against... [1:22:56] i don't know emma chamberlain or insert tick tocker x y and z right and in that [1:23:01] and especially the mechanism of TikTok, [1:23:04] explicitly is about quantity, right? And so [1:23:07] you've [1:23:08] just the thesis was just wrong that you could create. Not only do you need the Kalo to be putting out a hundred videos a month, you need a, a, [1:23:16] 20,000 Michaelis to even have a chance. Yeah. [1:23:18] So maybe, so two vectors that can potentially shift this. [1:23:22] alongside video creation being easier. One would be them actually being agents or AI. - Yep. - And the other is, I think a thing that you and I have chatted a lot about in the past too, which is this broad idea that like maybe pseudonymity should be much more of a common kind of like identity frame on the internet.

[1:23:36] I think you did a lot of versions of it. One of the earliest ones I remember was you telling me about Scooby Trooples. It was a crypto founder who had raised money as Scooby Trooples. You have somebody on your team, Smack, [1:23:47] Do you think that... [1:23:48] And just to kind of clarify this for people, like, not anonymous. [1:23:52] Very clear identity that's specific, but it's something other than their real name. [1:23:56] Do you think that this kind of thing becoming more common and growing is more about tools that enable it, whether that be... [1:24:04] 3D high fidelity avatars for calls and things like that, all the way down to just simple identity attestation for social platforms. [1:24:13] Or is it ultimately just about cultural adoption and people getting used to the notion that it's normal to not use your real name online, to use a synonym? [1:24:22] Yeah, I think it's cultural because even at our annual meeting last year, Smack – [1:24:28] He didn't have to, but he gave a slide on why he is pseudonymous and... [1:24:33] And... [1:24:34] half of the LPs... [1:24:35] had never even thought of that idea. And a lot of his framing is... [1:24:39] around wanting [1:24:40] his ideas to stand outside of any sort of persona, right? [1:24:44] which I think most people [1:24:46] don't go that way. They go the maximally jaded view, [1:24:48] of like, especially in crypto, just all the negative components of being... [1:24:53] not yourself. This has been one of my claims the whole time is that like, [1:24:56] The baseline is I don't want to get doxxed because I have some security risk. But all the things above that are actually, like, about expression and about a much wider, like, identity fidelity or identity design space, basically. Yeah, yeah.

[1:25:08] And I think that's really interesting. I just... [1:25:10] I do... [1:25:11] find [1:25:13] There is some... [1:25:14] amount of... [1:25:17] I think humans underrate [1:25:19] vulnerability forever, like in massive ways. And I do think that [1:25:25] The problem with a lot of [1:25:27] the [1:25:28] uh, [1:25:30] internet generally is people aren't vulnerable and so if you can be [1:25:33] pseudonymous and [1:25:35] have some sort of vulnerability tied to it. I think that's almost like the superpower because you're [1:25:39] basically saying, [1:25:40] The thing that might curb my vulnerability would be like my physical appearance, right? And it might amplify it as well, but people might be like, look at this person there, you know, think of the influencers who post these really vulnerable videos. Everyone's like, oh, this is just some hot person who's being annoying. And then you're like an anonymous person who's being, or a student-on-a-most person who doesn't have any physical components theoretically on the Internet, and you're purely standing on your ideas. [1:26:03] Thus. [1:26:04] The words you say and the actions you take mean... [1:26:07] you know, X percent more. But I don't think... [1:26:10] I don't know when that will change, and I don't think it [1:26:12] probably will change for that long that like more and more people will become i think more more people will but i don't think it's gonna be a mainstream thing anytime soon [1:26:19] Yeah, one of my zags historically has been [1:26:21] And you got to this a little bit that like... [1:26:24] maybe this is actually a path to help people express themselves more. Truly. For sure. One of my favorite blog posts. [1:26:30] you ever wrote is, is this one about this Tim Crider quote about authenticity and like being known the, you like that transition? I just pulled it. Tim Crider, who's an amazing, amazing author. If people don't know him says, if we want the rewards of being loved, we have to submit to the mortifying ordeal of being known.

[1:26:49] And then a couple of additions from you that I really liked. We string together fragments of various selves, but rarely do we see the entire self. Because what's the incentive? There's just too much risk in being known. [1:26:59] By being somewhat known, we are effectively minimizing some of the beautiful human volatility. And then finally, and I feel like we've conflated the idea of knowing someone with having an idea of someone. [1:27:09] I think this is especially interesting in the context of [1:27:12] you're someone who's written online for a really long time. You've done a decent amount of expressing yourself, but like, [1:27:17] I think also like people maybe pull little snapshots of you or whatever. We're all dealing with this a little bit in this kind of like social media is partly performative. It's partly trying to be authentic. It's partly trying to connect. Do you... [1:27:29] How do you feel about this? I don't know, five years later since you wrote that? Yeah. [1:27:33] shit [1:27:34] I think I feel as... [1:27:35] strongly about it as I ever have. I think the only complexity that I would be really, I'll be really honest about is that, [1:27:42] As you... [1:27:44] start to build something that has people that rely on you. [1:27:48] you [1:27:49] have to think about your actions a little bit more on the internet. [1:27:53] And I don't know, maybe not in Trump's America cancellation. Yeah, maybe not. I think in that moment in time, particularly on the Internet, just to be clear. [1:28:03] Everywhere. But like the internet is just the maximally scalable version of you, right? Oh, well, so I would wonder about our friend Blake and I have chatted about in the past. Like Blake's always said like... [1:28:13] past 30,000 Twitter followers is just like all diminishing returns. Like, yeah, [1:28:17] in a way that I think vulnerability and authenticity and being known in a

[1:28:21] real-world context. [1:28:23] Obviously, there's still nuance there, but it feels sort of... [1:28:26] structurally different than online. I so I agree. I think actually being [1:28:31] The real world context is... [1:28:33] It's the best. It's amazing. Yes, 100%. [1:28:37] And the best part is there's the famous line from Social Network where it's like, the internet is written in ink. It's not written in pencil or whatever. But that's the internet. [1:28:49] On the Internet, you have something that's theoretically kind of permanent. You have really large-scale... [1:28:54] and [1:28:55] You have people who don't really ever... [1:28:57] think they'll see you in person, so they'll do whatever the fuck they want with your ideas or you. [1:29:03] And when I wrote that, [1:29:05] post, I was... [1:29:08] there was, I don't think there was anyone else working. It was just myself and David, I believe, and, and Tara RCFO, but, um, [1:29:15] It wasn't like a risk of other people's careers in the same way. [1:29:20] Now there's a full... [1:29:22] team and we're going to continue to build that team. And there's people who, um, [1:29:26] you know, hopefully we'll [1:29:27] continued to... [1:29:29] be owners of the firm alongside of me and [1:29:32] but I still have a responsibility as the person who runs the firm to... [1:29:35] make sure I don't do anything too stupid, or say anything too stupid. [1:29:39] And there's also just... [1:29:40] Obviously stupid, but there's also just things that probably... [1:29:44] a public leader of an organization. Maybe this is not the thing in 2024. Anything goes. Yeah. Like there, there are probably things it's a level of, I don't know if maturity is the right word, but yeah,

[1:29:54] There's some responsibility. Censorship. I mean, I think this is the funny part, though, is this is... [1:29:59] Back to the point on tech that I think is funny is, [1:30:02] In that moment in time, that was also peak [1:30:05] which this was not at all what this was about, but it was also peak cancel culture and all these things. And everyone was like, we are a sports team. We're not a family. And people started saying that stuff. And it was like, Toby Shopify is amazing. Brian Armstrong is amazing. No room for politics, no room for any of these things. And then 2024, it was like, [1:30:23] "Hey, so we are Andreessen Horowitz, and we're donating X millions of dollars to these political campaigns with these viewpoints, and this is how we feel about things, and we're going to try and coin a term." And we are X firm doing that. [1:30:36] it's, [1:30:38] It was like the most hypocritical turn I've ever seen in... [1:30:42] tech in the past five years. Well, [1:30:45] I guess they would argue it's not about being known. It's about like these political candidates are going to help our investments. Sure. Like that's great. Like, again, I think that's fine. But like the whole point was not about – that whole talking point though was – [1:31:00] There isn't room for these things that aren't about... I guess their argument would be this is about... [1:31:04] Yeah, helping our bags, which is the Brian thing would have been 2020 was we don't talk politics at Coinbase because it, you know, that hurts our bags. So that's a that's a way to frame it. But I think. [1:31:15] Yeah, I don't know. My broader point in all this is I do think that – [1:31:19] there is a continual tension I feel about how [1:31:22] known or how [1:31:24] purely uncensored can I be? I think luckily,

[1:31:27] And this might be self-selecting because I hire everyone at Compound, but luckily everyone else at Compound is also pretty chaotic on the Internet. [1:31:34] you know, myself relative to Smack. Like, we both have pretty high chaos on Twitter. And conversely, on some level, like... [1:31:40] back to the therapist thing, like you have... [1:31:43] It depends on how personally you want to go, whatever. There is a whole bunch of cool stuff that comes from you having dumped your mind onto the internet over the last [redacted address] that has made you a lot more legible for people who want to go down that rabbit hole, and I think that's pretty cool. [1:31:56] Yeah, I hope so. [1:31:57] Do you... [1:31:58] slightly different. [1:32:00] point, but do you... How do you relate to... [1:32:03] Discipline. [1:32:04] And to what extent does curiosity and discipline have a relationship? This is ironic coming from you. Exactly. [1:32:15] I think I have the easiest job in the world for someone who has high curiosity. [1:32:20] And so... [1:32:21] Because of that, [1:32:22] I get paid to be really curious... [1:32:25] And [1:32:27] I just have to be disciplined enough to know that I'm doing it through... [1:32:30] a long-term lens of [1:32:32] capitalizing on that curiosity. [1:32:35] And that makes this one maybe a little easier. [1:32:37] Yeah, I also think in general I'm just like – [1:32:39] I'm weirdly a person who has like a... [1:32:43] very high agency and ability to kind of just if I need to do the things do them I don't like it right [1:32:48] pulled, but that's not an interesting thing to say. I think it's... [1:32:51] Yeah, I just am very lucky that I get to – that that's kind of – that was what I did at CB Insights. [1:32:56] what I did at the hedge fund, why I used to do adventure...

[1:32:59] Be curious. [1:33:01] And... [1:33:01] The best part is that it's explicitly... [1:33:04] curious about things that not everybody already believes. [1:33:07] I think to be, because that should be inherently valuable. [1:33:11] if you're doing long duration investing like we are, [1:33:15] versus seeing the present clearly, as Kohler says, which you might actually have to be much more defined in your curiosity. [1:33:22] Right. [1:33:22] But the... [1:33:23] Also, to some degree, you've said this a few times, I think you used to write like 10 to 15 articles a week for CB Insights. [1:33:30] curiosity driving that. [1:33:32] Some level that is discipline, but curiosity being the engine there feels a lot easier than... [1:33:37] whatever else. Obligation. Before we started recording, I read a bunch of writing in preparation for this. [1:33:45] you're really prolific. Yeah. Yeah. [1:33:47] And on some level, there's clearly discipline there. [1:33:50] But there's also some, I would have to guess something more. [1:33:53] Yeah, I think it's... [1:33:55] I think it's curiosity, but it's also kind of like competitiveness, which is the moment... [1:34:00] There's a few kinds of writing, and there's... [1:34:03] I have an idea about something and I need to get it out. [1:34:07] And I need to get it out because it's like tormenting me. [1:34:11] there's [1:34:12] I have an idea about something and I'm starting to notice that other people are starting to dance around this thing. [1:34:17] And... [1:34:18] If I don't get it out, someone else will, and that will really bother me. Hmm. [1:34:22] And that, so that comes from like competitiveness versus like curiosity of, [1:34:26] And then there's the third, which is like, I want to organize my thoughts. [1:34:28] And so I think there's, sometimes it's discipline, sometimes it's curiosity, sometimes competitiveness, sometimes it's,

[1:34:35] I don't know, like... [1:34:36] annoyance maybe is the right word like this like I wrote something on power and technology. They're just so wrong. They're so wrong and I need to tell them. Yeah well it's like it's like an or just like it's a feeling that I need to be structured in the world so other people can. [1:34:50] can... [1:34:51] who I know feel it can have a thing to like, [1:34:53] Which is an amazing thing to give someone. I often say one of the best. [1:34:56] types of writing is the type of thing that the person sort of reads nodding their head being like yes yes like this is what I've been saying. [1:35:02] Yes. Which is an amazing, it seems easy to do it, it's really hard. It's really hard, but that's, I think that's something people get wrong about a lot of writing is they want [1:35:09] everyone to read it and actually there's something which is like [1:35:13] great writing is writing something for one person and like [1:35:16] I wrote... [1:35:18] I wrote this thing, I send letters to friends sometimes, and I wrote this thing that you got at the time, something about life. [1:35:26] the point on it was... [1:35:28] about having changes in life and [1:35:31] a random person who's now a very close friend who I text every day DM'd me. I was like, Hey, [1:35:36] My brother is actually going through a divorce right now, and I sent this to him, and it was incredibly meaningful. [1:35:42] And you don't really know me, but we like work in the same industry, um, [1:35:45] And that was like kind of... [1:35:47] Thank you. [1:35:48] That's like kind of just worth it. Like that makes all of the things worth it. A thousand percent. Yeah. You know, especially in a world where you're like, your value is how many, like I got 37 Twitter likes. Yeah. There's an idea I really like from this guy, CT Nguyen, where he talks about sort of like [1:36:00] the way we measure things affects our values on them. When you gamify communication, your value set for communication follows. So he gives the example of like a tweet versus a teacher in a classroom of students. And the teacher could say something that 29 students don't really understand.

[1:36:16] blink and one student's eyes light up and it's everything. And you don't get that. On Twitter, it's just like all a hundred people kind of maybe liked it a little. Yeah. It all happens in the DMs, I guess. I don't know if it was that post, but one of my favorite posts you've ever written was this one on inertia. I often say inertia is the strongest force in the universe. I can't remember if I got that from you or not. I'll read a couple of things to give the readers context. The first bit is... The important context of this post is that it was a private letter [1:36:46] friends that year [1:36:47] And then I... [1:36:49] randomly one day was like fuck it and I just tweeted it out. [1:36:52] Also a good reminder that sometimes the things that are intended for one person or a small group [1:36:57] Great art, many great things are universal through being highly specific. Very well said. [1:37:03] How much can I or should I attempt to impact or change those close to me in their lives? I don't know the answer, but most people would say that it's not your life to live. My counterpoint is that it's part of what being a friend is. I want my friends to try to make me the best version of myself and vice versa. Pushing when you have no incentive and being there, regardless of the outcome, is a clearer way to do that than just idly standing by. And the implicit context here, of course, is... [1:37:28] the notion that most of us [1:37:30] sort of just get flattened out by life, carried by inertia. We don't end up making our decisions. There's this line I love from John W. Gardner. He says, most of us attach, we're like a barnacle. We attach our head to a rock and that's it. And so not only did you get at this really powerful idea, which is that are you succumbing to inertia yourself? Are you doing the things you've chosen or did you just wake up here? But two, what should our relationship to other people do?

[1:37:55] You wrote this five, six years ago? Yeah, something like that. I think this is especially relevant in the context of friendship. I think you end that post with the classic, like, that's the thing about friendship. There's nothing in it for anybody. One of my favorite quotes. Maybe friends in particular are one of the areas this can most be expressed because you have the bandwidth or the lack of rules or obligation to be able to tell someone something that you know they're not going to want to hear that's explicitly not in your interest to tell them because they're going to be mad at you. [1:38:25] One, I'm curious how your thinking on this has evolved at all, and two, [1:38:29] Have you found ways to be able to do that outside of the, like... [1:38:33] Your your buddies who yeah, you're fine pissing off. Yeah, I my my thinking on this has evolved a lot. [1:38:41] I feel like... [1:38:42] I think the following year I wrote something which was in the same letter context. [1:38:46] which said, [1:38:48] Sometimes I feel like life is just... [1:38:50] learning the things that everyone a few years older than you has already learned. [1:38:56] Like something, something to that effect. [1:39:00] And so I will say, [1:39:02] There's two framings. You were 29 when you wrote this, I believe? [1:39:05] Probably, yeah. 28, 28, 29. [1:39:09] I would say... [1:39:10] I don't feel as strongly anymore. [1:39:13] in the sense of I think some people – [1:39:15] Everyone is scaling a mountain. [1:39:18] and [1:39:19] when you [1:39:20] are younger, you think your peak is the same peak as everyone else's. And you also think your peak is the correct peak.

[1:39:27] And... [1:39:29] People start going off to different base camps as you scale the mountain. [1:39:31] And... [1:39:33] Some people judge those base camps. [1:39:35] What I think the reality is is that [1:39:37] you're actually wandering through a desert with all your people. [1:39:40] and you have these [1:39:42] different oasis [1:39:43] type things that everyone has. [1:39:45] And so I think my framing has become a lot less [1:39:49] I don't know, prescriptive. [1:39:51] And so more... [1:39:53] It's, you know, the best people in these types of things are very good at asking questions or they do the... [1:39:58] I wonder if type framings. [1:40:01] And I think in friendship, I was never the I wonder if person. I was like, you're a fucking idiot person. And sometimes I'm still like that. But I think that. [1:40:11] it becomes... [1:40:12] more powerful to... [1:40:14] at least be the person that can push, but can know when to not push. Because if you're always pushing, you're just that annoying pushing person. And sometimes what somebody needs is not... [1:40:24] tried it being fixed. Yeah, exactly. It's the canonical relationship thing, which is, do you want a solution or do you want to just talk about your problem, right? [1:40:33] I do think that, I was talking about this with someone, [1:40:36] We're trying this exact like [1:40:38] problem and question and [1:40:41] She said, what is the... [1:40:43] point or do you really think you need your partner to push you? Like is that really their job? Because I said, you know, in a partner I like someone who is [1:40:52] makes me a better person in some way. [1:40:54] And my pushback to her was...

[1:40:57] sometimes it's really hard to respect someone or to feel that the... [1:41:03] support from someone is as good or as real if they haven't also at some point tried to push you in some way. [1:41:09] And. [1:41:10] I've known some people who... [1:41:12] They say the other thing, which is the only thing that matters is how you make people feel. [1:41:15] And they take that to, in my mind, insane extreme in which they will never say anything bad to anyone or push on anything. It's just purely confirmatory. We love in different ways, I think. Sorry, I'm not saying you and I, like people. Yeah, yeah, yeah. I think you and I are people, to your point maybe, who like one of the ways we love, I think most of our friends would agree, is like by – [1:41:38] offering thoughts. Yes. Um, and again, I think I'm probably behind you a little bit. I think we both made progress on the degree, but also man, people love in different ways and they need love in different ways. Um, [1:41:49] And you can't bring facts to a feelings fight. [1:41:51] No, you can't not. [1:41:54] But I think that's, I would say that's changed quite a bit. [1:41:57] And I think it's probably net... [1:42:00] correct and healthier. [1:42:02] I still do at times... [1:42:05] Wonder you know, what are the [1:42:08] What's like the incubation period that you're willing to let someone continue to wander in the wilderness for? And when do you at least maybe try and jump back in? [1:42:15] is. [1:42:16] Or are there even areas that are worth dying on? [1:42:20] I guess what I mean, one classic thing here is like, should you stage an intervention if you don't like your friend's partner or whatever? That's not something to ultimately die on. Yeah. But are there areas that we're like, I actually care about this person so much and this is so critical –

[1:42:35] at least from my vantage point, and maybe I'm wrong, about the trajectory of their life that, like, [1:42:39] I'm willing to risk it all to try to knock them on. [1:42:42] knock them onto the different path. [1:42:45] Um... [1:42:46] It might be a little dramatic. Yeah, this is very dramatic. [1:42:49] Outside of things that could actually really ruin something, like kill someone's life or whatever... [1:42:54] The relationship thing, I don't think so. You only know 10 to 20% of anyone's relationship. [1:42:59] I think that [1:43:00] a role I often play on boards is, [1:43:03] and have had to play on boards and then with companies is... [1:43:06] There are times in which the broader investor base of a company will kind of tap me to go and – [1:43:12] give the push. [1:43:14] which is hilarious to me. And... [1:43:16] I think that... [1:43:18] I always frame all these things with founders of make the decision that if your company fails, you can sleep at the end of the night. [1:43:24] And so... [1:43:26] I have found that [1:43:27] In some cases, I've had to go and do that, and sometimes the founders have been like, [1:43:33] thank you so much for coming and doing this. [1:43:37] There have been times where the founders have said, fuck you, and... [1:43:40] probably don't like me anymore, because I was the only person who pushed back, despite maybe everyone asking me to go and push back. [1:43:47] But I actually think that is like a far easier dynamic than someone you love or a friend. Yeah. Well, there's ironically the desired outcome is more clear and more aligned. Yeah. Yeah. Well, yeah, that's true. [1:44:01] Though I do think the, I think it's like in order of ease, it's...

[1:44:06] the founder relationship, the romantic relationship slash partner or whatever, and then the friendship. Right, because it's the most blurry. Yeah, but I think the friendship one... [1:44:15] If you have... [1:44:16] you know the friends in which it will, because of who they are, you know the friends in which it will go well or not, and it's just the pain tolerance you're willing to go through, and how... [1:44:24] How extreme do you think their reaction will be? I also just love this area because I think it's one of the areas where you actually see the uniqueness of the love that goes into a friendship. Yes. Which is like I actually – there are cases where I will be willing to say something to you that is explicitly bad for me and might ruin our friendship just because I want your life to be better. Yeah. Which is pretty cool. Again, you've got to be really careful. That's just pretty cool. My last note for your question is – [1:44:48] you... [1:44:49] Clearly from all kinds of sources of inspiration, I don't know if maybe your dad being in hospitality, you clearly love art and music and film. And then it also ties into maybe some things you've talked about, maybe you were alluding to earlier in the conversation around... [1:45:02] This idea of craft. [1:45:05] I'm not sure we have more craft in the world than we did [1:45:08] now than we did have in the past. Maybe we have more. [1:45:12] Is that something that you think about either professionally or personally, and how does it... [1:45:16] effect. [1:45:17] What do you do? [1:45:18] I never thought of if we had more or less. I actually think we'd probably have less, realistically. Okay. [1:45:22] I think Kraft is like the... [1:45:26] thing I like love most in the world outside of people I think the fact that [1:45:33] And you see this in how I... [1:45:35] talk about people, think about people, respect people, which is

[1:45:38] people who are on a [1:45:40] It's kind of the point of decaying, right? [1:45:42] Kraft is fighting against Decay or trying to [1:45:45] continue to be better in an infinite pursuit. [1:45:50] In my work, I think about that. In my relationships, I think about that. And then I think the, I get a lot of energy thinking about that in terms of like mastering [1:45:59] crafts that [1:46:00] bring feelings to people. [1:46:02] and that's hence the food and the art and the i mean i'm terrible at art but the music [1:46:06] So I think... [1:46:08] I don't know if... [1:46:11] That's a really rough statement. I don't know if I could love someone if they didn't have some respect for craft. [1:46:18] like in a very deep way. Like there's some people who, they're fun and, you know, they're the kind of friend that you enjoy spending time with, whatever. But I don't know... [1:46:27] if I get deeply... [1:46:29] connected to someone if they don't [1:46:31] appreciate craft of some kind. What is... How would you... You don't need to define it, but what do you feel when you... [1:46:38] think of that word [1:46:40] What is precisely inside of it? [1:46:42] I think just the idea that you're doing something because you want... [1:46:46] to [1:46:47] Work on it. [1:46:49] And [1:46:51] It's not, it might not be about, sometimes it's about the outcome. You want to be good at it. Right. [1:46:55] Sometimes it's about the process. [1:46:57] but you want to just do the thing, [1:47:00] in an indefinite [1:47:02] pursuit. [1:47:03] And I think that is... [1:47:06] It's kind of like it kind of underpins a lot of stuff like it underpins the consistency. It underpins the conviction, underpins the feelings. A lot of the things we talked about, I think, now that you're bringing this up, which is why you're good at doing podcast stuff.

[1:47:19] Yeah, I think it's the first-order... [1:47:23] integer to the equations to all these things. I think, I hope at least it's [1:47:29] it's a real key to the path of like, [1:47:32] Playing an infinite game the reward for good work is more work. Yeah, yeah, I mean yeah, hopefully [1:47:38] And at a certain point, your craft can change, too. I think that's the thing that we don't... But it's a disposition, to your earlier point about connecting with other people. The disposition towards craft, honoring craft, pursuing it. Yeah. Yeah. [1:47:48] taking it seriously... [1:47:50] Yeah, respecting it. Having a respect for... [1:47:54] the art of craft, I guess, is how I frame it. [1:47:57] Yeah. [1:47:58] This is really fun. [1:48:00] This was fun. Michael Dempsey, my friend. Thank you very much. Thank you.