The Exchanges

Q You mentioned the industrial revolution there. The one thing that strikes me as very different now to then is simply the speed of adoption. How do you think about the speed of adoption? How does that maybe change or alter how you think about the relationship and symbiosis?

A Well, look, there's a great paper that I read recently. It was referred to me about when The electric motor was invented. And before you had these factories that were based around steam or water power, and they were all based around central drive trains. So electric motor is invented. It takes about 40 years for the electric motor to be adopted into factories. And initially, people didn't really know how to get leverage from it, because they were just retrofitting old factories. You had to rip the factories down. Totally changed how they were organized to take advantage of this idea of modular power, right, and distributed motors throughout a factory environment. It's going to be faster. Maybe it's 20 years, right, for AI, or 15 years for AI to be adopted, but it's also not tomorrow. I mean, I think there's a famous Bill Gates line that's something to be effective. People tend to overestimate how quickly things change year to year and underestimate how much they change over decades, and I think that's kind of where we are yet again.

AI assessment note: “It's going to be faster. Maybe it's 20 years, right, for AI”

Q What's the backstory there? How do you split your time? What are the kind of challenges and opportunities of being both a VC and a founder at the same time?

A So I think, I mean, the story goes back a bunch of years, which is when I had my first startup called Drapio, um, in New York, um, many years ago, you know, I started doing a bunch of angel investing. And so I was, you know, one of the first investors in MakerBot and the first investor in Birchbox and, you know, the first investor in Venmo and a bunch of awesome companies. I found it really interesting because I thought that, you know, the experience of being a founder and actually building stuff actually I think makes you both a much better investor and also makes your advice useful. You actually have a relevant perspective to share. So I was at Facebook a long time after they acquired my company. I loved it. I have incredible respect for the company. Um, and when I left, you know, I knew I wanted to do more investing. But I also, like, wasn't quite done. I wanted to start another company. And so I had the opportunity to start an awesome company with a really close friend of mine, Andrew Cortina, who was the founder of Venmo. He also was finishing up at PayPal about the same time. That's where I spend the majority of my, like, hours, um, in the day. But then, you know, at the same time, you know, the, the, I've been at LP in the first few slow funds, which were usually just, you know, they were just a small group of friends investing together, you know, led originally by Dave …

AI assessment note: “That's where I spend the majority of my, like, hours, um, in the day.”

Q No, I agree. Peter Fenton and Jeremy Levine of Basma both said we're in a consumer downturn, so that was a cheery episode. But tell me, I want to finish today on your most recent publicly announced investment, and why did you say yes and get so excited, Sam?

A I'm going to go with, and I'm not actually sure this is true. Oh, no, you know what it would be? The most recently publicly announced investment is a company called Fetcher.ai, which is effectively AAI, or Artificial Artificial Intelligence, for recruiting. So if you, you know, need to fill a role, they will effectively do a lot of the top of funnel, where they filter through candidates for you, and like, basically start, write the email introductions for you, so that you kind of get at the bottom people who are engaged in the funnel that are in the right job and the right role. I think it's great. I think it's a great use of machines and people working together in a specific way to kind of augment and optimize the process.

AI assessment note: “The most recently publicly announced investment is a company called Fetcher.ai”

Q You mentioned the industrial revolution there. The one thing that strikes me as very different now to then is simply the speed of adoption. How do you think about the speed of adoption? How does that maybe change or alter how you think about the relationship and symbiosis?

A Well, look, there's a great paper that I read recently. It was referred to me about when The electric motor was invented. And before you had these factories that were based around steam or water power, and they were all based around central drive trains. So electric motor is invented. It takes about 40 years for the electric motor to be adopted into factories. And initially, people didn't really know how to get leverage from it, because they were just retrofitting old factories. You had to rip the factories down. Totally changed how they were organized to take advantage of this idea of modular power, right, and distributed motors throughout a factory environment. It's going to be faster. Maybe it's 20 years, right, for AI, or 15 years for AI to be adopted, but it's also not tomorrow. I mean, I think there's a famous Bill Gates line that's something to be effective. People tend to overestimate how quickly things change year to year and underestimate how much they change over decades, and I think that's kind of where we are yet again.

AI assessment note: “It's going to be faster. Maybe it's 20 years, right, for AI”

Q kind of privacy and China argument, I do want to stay on the theme that we're on. I had Adam Goldberg, a light Speed on the show. He said that the underlying infrastructure can't support mass market dApps. I think we saw that with kind of crypto kitties. How would you respond to this, and where does it leave your thinking and standing as to maybe the nearer term potential?

A Well, look, I spend most of the time looking at investments in the crypto space in, you know, third generation protocols. There's no question. Ethereum cannot support anything of scale. I also think we have to be careful about dApps and, like, even their role in the systems. I am much more interested in us sorting out the database layer first, right? Which is how do I have a system that works at scale where we can all agree that the records are the records and no one entity can undermine the records. So before you get into smart contracts and dApps and this kind of ecosystem people want to build, I think we have to start off by saying like, is there a globally secure decentralized database I can trust and that works at scale? You know, in a lot of ways, when you think about what a cryptocurrency is, it's kind of the simplest database, right? It's just a series of keys.

AI assessment note: “There's no question. Ethereum cannot support anything of scale.”

Q Can I ask, do you think it's a winner-take-all market, or do you think it's kind of geo-fragmented?

A No, I don't think it's a winner-take-all market. I mean, I also think this is a secondary issue. It's going to end up being effectively, well, you know which scooter I'm going to use? It's literally the one that's closest to me. If there's a scooter five feet from me and a scooter that's 10 feet from me, I'm taking the five foot one, and I'm totally happy to have eight apps in order to, like, not walk an extra walk, which I think also makes it difficult from an economic perspective. People argue, they say, well, the value will come in recharging scale and stations and positioning, and I'm sure there are some advantages there, but I think it's going to be highly fragmented unless the cities regulate it in a way that they only allow a few competitors. Which is starting to happen, but then it's going to be highly controlled in terms of the profits that the scooter companies will be able to take versus the government saying, look, this is great. You're using our sidewalks and our capacity, and we all need revenue, so thank you very much for increasing our budget.

AI assessment note: “No, I don't think it's a winner-take-all market.”

Q think it's always super interesting. Uh, our, Tendency to paint a picture of evil machines taking our jobs and what that means for us, but it was actually your own partner who wrote a phenomenal piece on the declining productivity in males and never seen before in terms of kind of lack of productivity. How do you think about this and almost to what extent human apathy is to blame?

A So it's really interesting set of conversations and Cortina's writing on it is quite good and maybe you can link to it in the podcast. I think here's the way I think about it. There are two elements. One is in the West, If we want to maintain and keep having jobs, we're going to have to increase our productivity, right? And we're going to have to use technology to do it rapidly, because the reality is that globalization, the way you fight globalization, is not by trying to say no one else should get to play, or we should cut off the global environment, as some might believe. But in my mind, it's saying, look, let's get better at what we do, so that we are actually comparatively better off the technology and tools and productivity we have at completing a given task or doing that type of work. So I think we have to be very, very focused. I mean, I would argue that technology and productivity is the only path to having jobs. So I think that's one part of the equation. Now, the apathy thing that Cortina is talking about is something we talk a lot about, which is, you know, you ask the average person, they say, I work a four hour work week or, you know, in technology or banking or we're pretending they, they might say I work an eight hour work week. But then if you sit down and actually measure what they're doing, say, okay, are you like, how much time are you actually engaged? The …

AI assessment note: “Now, the apathy thing that Cortina is talking about is something we talk a lot about”

Q start today, Sam, back to the schedule before we move too far off it. We spoke in our last conversation two years ago about the man versus machine relationship. Time's now gone on, as we said, two years, so I have to ask, what have you, and what have we learned and observed over the last two years with regards to that human versus kind of computer hybrid system relationship?

A So I just don't think it's computer versus human, and I think that's what you and I talked a lot about in our first podcast, This was the time when, like, Facebook M and things like that. People were very excited about this idea that you could do a bunch of tasks, and you'd have this magical black box AI that would learn from the humans and the data set, and then voila, you wouldn't need people anymore, right? I mean, how wrong has that proven to be as a model? To my own horn, as I was trying to push years ago, and we still push, um, you know, the real model, I think, is you have to think about AI and people as symbiotic, and you have to think about the real transition in the industrial revolution of knowledge work we're going Through is going from, We have a new piece of technology. We need to change how work is organized and focus humans on the most human tasks and kind of productionalize factories of machines around them to get them as much leverage as we possibly can. So it's very similar to how the industrial revolution happened. Steam alone actually doesn't get you massive increases in productivity. There's an incredible study about electric engines, and when the electric motor was invented, how factories were retooled. It didn't have this immediate improvement. What you had to do was reorganize the factory floor. Change how work was done, change what parts of work people…

AI assessment note: “the real model, I think, is you have to think about AI and people as symbiotic”

Q think it's always super interesting. Uh, our, Tendency to paint a picture of evil machines taking our jobs and what that means for us, but it was actually your own partner who wrote a phenomenal piece on the declining productivity in males and never seen before in terms of kind of lack of productivity. How do you think about this and almost to what extent human apathy is to blame?

A So it's really interesting set of conversations and Cortina's writing on it is quite good and maybe you can link to it in the podcast. I think here's the way I think about it. There are two elements. One is in the West, If we want to maintain and keep having jobs, we're going to have to increase our productivity, right? And we're going to have to use technology to do it rapidly, because the reality is that globalization, the way you fight globalization, is not by trying to say no one else should get to play, or we should cut off the global environment, as some might believe. But in my mind, it's saying, look, let's get better at what we do, so that we are actually comparatively better off the technology and tools and productivity we have at completing a given task or doing that type of work. So I think we have to be very, very focused. I mean, I would argue that technology and productivity is the only path to having jobs. So I think that's one part of the equation. Now, the apathy thing that Cortina is talking about is something we talk a lot about, which is, you know, you ask the average person, they say, I work a four hour work week or, you know, in technology or banking or we're pretending they, they might say I work an eight hour work week. But then if you sit down and actually measure what they're doing, say, okay, are you like, how much time are you actually engaged? The …

AI assessment note: “Now, the apathy thing that Cortina is talking about is something we talk a lot”

Q kind of privacy and China argument, I do want to stay on the theme that we're on. I had Adam Goldberg, a light Speed on the show. He said that the underlying infrastructure can't support mass market dApps. I think we saw that with kind of crypto kitties. How would you respond to this, and where does it leave your thinking and standing as to maybe the nearer term potential?

A Well, look, I spend most of the time looking at investments in the crypto space in, you know, third generation protocols. There's no question. Ethereum cannot support anything of scale. I also think we have to be careful about dApps and, like, even their role in the systems. I am much more interested in us sorting out the database layer first, right? Which is how do I have a system that works at scale where we can all agree that the records are the records and no one entity can undermine the records. So before you get into smart contracts and dApps and this kind of ecosystem people want to build, I think we have to start off by saying like, is there a globally secure decentralized database I can trust and that works at scale? You know, in a lot of ways, when you think about what a cryptocurrency is, it's kind of the simplest database, right? It's just a series of keys.

AI assessment note: “There's no question. Ethereum cannot support anything of scale.”

Q though, the hard thing for me as an investor today is like, how on earth do you price it? And so, we had Avichal at Electric on the show, and he agreed, suggesting the way that dApps, and kind of the majority of blockchain-based investments, pricing is fundamentally challenging. For you as an investor evaluating it today, Help me out here, Sam. How do you think about and approach this?

A Well, I think I'll give you a few angles. First, we, I generally don't invest in dApps. Again, like we're much more focused on protocols. There are very few exceptions we would make to that. In terms of pricing, I mean, there's a few ways to price anything. One is you price things based on what the market will bear, and that's kind of like fundamentally how, you know, the price of Bitcoin is set by supply and demand. Pretty simple. The second thing that I always go back to is this If this works, is there a line of sight to how it's worth 10 times as much? And if you stretch your imagination and this really works, where can you end up? Again, I don't think it's that different than the other type of venture math from that perspective. The difference is, is that a lot of people believe that if you get this stuff right, it is so fundamentally important, right? But the, the hundred X case, it might actually be a thousand X case in some cases, you know, unless of course the things are priced so highly that you're back to a hundred X case. So, you know, we, we treat it like any other venture investor. I mean, you have to think about competitive dynamics. You have to think about the open source nature of it, and what's really defensible versus not. But it's fun, and I would argue that in a lot of times, it is for sure the wild west, and lots of money will be lost, just as I think a lot…

AI assessment note: “In terms of pricing, I mean, there's a few ways to price anything.”

Q Sam, you mentioned global booms there, so before we move on to the element of new cycles and fake and real, I do have to ask, everyone in Europe is chasing the lime and bird competitors that exist here. However, you've said before, you don't think this is a good business. Can I ask, why is that, Sam?

A Look, I really enjoy the experience of electric vehicles. I ride my boosted board all over town. I've loved the scooters as a consumer using them, not necessarily as a consumer being near them when other people are using them. To me, the argument's quite simple. It's kind of the fool me once, shame on you, fool me twice, shame on me. I don't think the cities are going to let the scooter companies be profitable or be meaningfully profitable. And the reason is really simple. It's like, it's because of Uber. Uber kind of spoiled it for them, which is Uber and Lyft Kind of that ride sharing. I would argue really haven't gotten an incredible deal out of global cities. They were able to build consumer love and demand and incredible service kind of around the sides of most cities. Cities don't really benefit from them financially, despite the fact that cities are supporting the roads, cities are supporting the traffic, the cities are supporting everything that they kind of sit on top of. I think that was kind of a one-time opportunity from a profit perspective. Does that mean there won't be scooters? No, they're going to be Scooters, and I think they'll be pretty good, but I think it's going to be much more controlled. The governments are not, local governments are not going to let them get away with the level of growth and power that I think the ride sharing companies have got away w…

AI assessment note: “I don't think the cities are going to let the scooter companies be profitable”

Q schedule itself, I had Danny Reimer at Index on the show recently, and he said the biggest mistake that VCs make is they underestimate The size of niche markets and just how big a market can be. You mentioned there about kind of the backable versus non-backable and kind of the market sizing element. How do you think about market sizing today? And is that a dominant thought for you?

A Look, I mean, whenever I'm doing an investment, the first question I ask myself is, how does it become 10 times larger? How does it become a hundred times larger? I think after a hundred times larger, you get into a little bit silly territory in terms of like how big things can be. Every once in a while, something might be a thousand times bigger than when you invest, but Man, that's like a once in a career, twice in a career type moment. I do think you can do a 10 X, hundred X model. I agree with Danny that I think people systematically under appreciate how large the world is, and therefore, especially in a globalized environment, how big niche markets actually can be when you blow them up and scale them up. But I think it's an interesting calculus. I tell a lot of people professionally, do what you love, because if you do what you love, it's just an economics argument. The reality is you'll probably do it best and get the deepest in it. And the reality is if you love some esoteric Comic book thing, or you like reenacting the Revolutionary War, thinking my brother right now, the reality is the market of those things is much bigger than you realize. But market size still matters, and the world is not infinitely large.

AI assessment note: “whenever I'm doing an investment, the first question I ask myself is”

Q to ask you, Sam, you're, you're a father yourself and kind of thinking forward with the AI future and AI driven years to come, where would you maybe like your child to become very cogent? Is it the development of the technical systems themselves or Is it kind of creativity, maybe, which is the highest of high fruit for AI to try and capture? How do you think about that?

A I mean, I want my son to do what he loves, and I think there'll be opportunities to do that. I mean, I'm very optimistic about kind of the future and giving people the ability to focus on and excel at and have pride in their work. So I'm macro very positive on this stuff in terms of whether it's designing systems or philosophy or math or painting or doing a A thousand other things. I think there's no question. There's, there's going to be abundance of opportunity. It'll just be different. I think the real thing that's going to be challenging is that I think it'll be harder to find steady, repeatable, trusted work. That's always there to rely upon, you know, creativity, creating things, coming up with new ideas. There'll be lots of opportunities, but it is kind of the, can you have a job as an accountant and count on exactly a certain amount of work all the time? And that, People will always need exactly what you do, you know, with no creativity. I think those are the jobs that will be threatened, and I'm, I'm hoping that as we, as my son grows up, he'll be less interested in repeatable tasks.

AI assessment note: “I'm hoping that as we, as my son grows up, he'll be less interested in repeatable tasks.”

Q Can I ask, do you think it's a winner-take-all market, or do you think it's kind of geo-fragmented?

A No, I don't think it's a winner-take-all market. I mean, I also think this is a secondary issue. It's going to end up being effectively, well, you know which scooter I'm going to use? It's literally the one that's closest to me. If there's a scooter five feet from me and a scooter that's 10 feet from me, I'm taking the five foot one, and I'm totally happy to have eight apps in order to, like, not walk an extra walk, which I think also makes it difficult from an economic perspective. People argue, they say, well, the value will come in recharging scale and stations and positioning, and I'm sure there are some advantages there, but I think it's going to be highly fragmented unless the cities regulate it in a way that they only allow a few competitors. Which is starting to happen, but then it's going to be highly controlled in terms of the profits that the scooter companies will be able to take versus the government saying, look, this is great. You're using our sidewalks and our capacity, and we all need revenue, so thank you very much for increasing our budget.

AI assessment note: “No, I don't think it's a winner-take-all market.”

Q start today, Sam, back to the schedule before we move too far off it. We spoke in our last conversation two years ago about the man versus machine relationship. Time's now gone on, as we said, two years, so I have to ask, what have you, and what have we learned and observed over the last two years with regards to that human versus kind of computer hybrid system relationship?

A So I just don't think it's computer versus human, and I think that's what you and I talked a lot about in our first podcast, This was the time when, like, Facebook M and things like that. People were very excited about this idea that you could do a bunch of tasks, and you'd have this magical black box AI that would learn from the humans and the data set, and then voila, you wouldn't need people anymore, right? I mean, how wrong has that proven to be as a model? To my own horn, as I was trying to push years ago, and we still push, um, you know, the real model, I think, is you have to think about AI and people as symbiotic, and you have to think about the real transition in the industrial revolution of knowledge work we're going Through is going from, We have a new piece of technology. We need to change how work is organized and focus humans on the most human tasks and kind of productionalize factories of machines around them to get them as much leverage as we possibly can. So it's very similar to how the industrial revolution happened. Steam alone actually doesn't get you massive increases in productivity. There's an incredible study about electric engines, and when the electric motor was invented, how factories were retooled. It didn't have this immediate improvement. What you had to do was reorganize the factory floor. Change how work was done, change what parts of work people…

AI assessment note: “real model, I think, is you have to think about AI and people as symbiotic”

Q Can I ask, Sam, what are you most concerned about when you view it going forwards?

A I'm an optimist. So I would argue, I think there's a lot of challenges, but I'm mostly optimistic about where we can get to. The thing I'm most concerned about is that I don't think we're well set up to make these, to have these discussions and make these choices as a society. For instance, I think that things as basic as, you know, the Constitution, et cetera, really are going to have to be reexamined in a lot of ways, and I don't think that we have the social structures to do that very well in the West. Simultaneously, I guess I am worried that the centralized authoritarian path For human civilization is a lot of ways I see the most straightforward, right? And it will be very powerful. I mean, again, I have a lot of respect for the Chinese and I understand what they're doing, but like take their approach to genetic records, right? Which basically says, look, we're going to get samples from everyone and centralize it and build a bunch of research. We could never do that in the West, right? Our privacy frameworks, our values keep us from doing that. That's going to be a really powerful data set. So how are we going to feel about kind of being left out of these revolutions and how are we going to wrestle with that? I Is going to be the challenge of our age. And he said, you know, what I hope my son will focus on. I think, you know, we're going to need more philosophers as much a…

AI assessment note: “The thing I'm most concerned about is that I don't think we're well set up”

Q I do agree with you. I'd love to see more philosophers. I do want to touch on one final element. Much of the blame, so to speak, with a lot of this content has been placed around the distribution channels of today, be it Facebook's Twitters. How do you view their role today, and do you think maybe the blame's misplaced?

A That's a leading question. No, look, I mean, I think that actually, I would argue a technological determinist. I argue that once systems exist, or they're set up in a certain way, what kind of falls out of them, it's a lot of hard work by individuals, and the details might be different, but the macro picture kind of results from what the system is capable of doing. I would argue that, you know, if you read your science Transfiction or talk to people, people like to pretend as though what's happened on the internet is shocking to them and could never have been forecasted or foreseen. I mean, I think that's like, I'll use a Britishism. It's like a load of bollocks. Like the reality is, is that people have been writing about this happening for 20 years, right? People have been discussing it for a super long time. And as a kind of emergent property, what happens when you build a net, when you build a network the way that we've built one. So look, I mean, I have a lot of respect for the teams at Twitter and Facebook. I think that they're in an unwinnable position in a lot of ways because society is simultaneously telling them we want privacy and we want security. And as I said, you can't have both. You have to make some choices. So I think they're kind of like this incredibly centralized focal point of a lot of fear and confusion and anger. And I think they're doing the best they ca…

AI assessment note: “who to blame? I mean, I think, you know, you're blaming the wind.”

Q And you said there about being the right time in the right team, and that's kind of the perfect combination. And I have to ask them, you know, now is the time of the bots. Uh, we're seeing, as you said, mass excitement. So first question is what use case are you most excited by?

A I mean, I think it's like, for me, it's interesting. That's, that's a complicated question. I'll tell you the thing that I love the most is augmenting my memory. I don't have the world's greatest memory. I like to take lists. I try to keep lists together and things like that, but I find it really, really hard to keep track of all The ideas and people and things I want to do and remember them at the right time in the right context. And I think that that's going to be, that's the thing that I get the most excited about is my augmenting my own memory in a really fundamental way. I also get really excited about the fact that I think that I call it the four and a half star problem, which is the internet was promised to be not just a very entertaining thing, but an extremely informative platform. And I actually think that we've gotten into this direction for a whole bunch of very practical reasons where You know, if I ask a very simple question, like I'm in a new city, I'm like, where should I go to dinner? I'm getting very generic and unuseful results. Right. Um, and so I think the other thing that I'm excited about, which I think can be solved is this, how do we actually deliver the right answer to the right person in the right context to a bunch of questions where in theory, there's a lot of information out there, but in practice, I mean, I challenge you to go to a city you don't …

AI assessment note: “the thing that I get the most excited about is my augmenting my own memory”

Q And then let's, let's finish on because we've talked so much about slow today. We've done a brilliant job of promoting that too. Uh, what's the most recent investment for you? This could be angel or it could be slow, but a private public investment that you've said yes to. And why did you say yes?

A So, I'll talk about one that's pretty recent in public, because I always forget exactly what's public. Um, but, um, the, um, I mean, one I'm really, really excited about is a company called Common, um, in New York, uh, started by a really close friend of mine named Brad Hargraves, who is one of the founders of General Assembly. Um, and they're working on effectively a co-living community where people effectively, after college, it's effective, it's adult dorms, right? Uh, he probably wouldn't want me branding it that way, but that's kind of what it feels like. And it's the idea that the community where you live is like as important as, you know, having a cool apartment or whatever. I just like fun, again, like there's, this goes on my list of, of course, this is the way the world will work. I can't imagine a world in which, you know, everyone's living in these kind of low, these apartments where they don't know their neighbors and there's no sense of community when you can start building dynamic communities and vibrant cities. So, you know, there's, we live is another company that's working on this from the, we work guys. Like there's a few of these, but it just goes on my, this is obviously going to happen and it just requires hard work.

AI assessment note: “one I'm really, really excited about is a company called Common”

Q So, so what do you ask the entrepreneurs then that you've invested in? Now, you know, the kind of tumultuous and difficult experiences that they've been through. Do you know what I mean? What are those questions that you kind of say, honestly, do you have a month's burn left or, Honestly, are you about to lose your VP of sales? Do you know what I mean?

A I just put the word honestly in front of normal questions. That's my value add. No, I think, I think, you know, the reality is startups are a emotional rollercoaster. I think it's like the most manic depressive way to live you possibly can. Like you're never doing well. You're either crushing it or doing terribly. And I think that the question for me a lot is one, you know, do the people, I think the only way to get through that and be successful is to truly be missionary about what you're working on. And so I think for me, like a lot from an investing standpoint, understanding what truly drives people and why they're doing what they're doing is probably the most important thing. And the second is like, are they in an emotional place and a personal place where they're going to be able to weather the downs and also not get too manic at the ups? Right. Um, and, and keep a level head because I think that ultimately like so much of this ends up being a head game. Um, I don't think starting companies in most areas is actually rocket science. It's usually just a lot of hard work. And so helping teams, uh, finding teams and then helping teams that truly have like a missionary stance that they're going to get through it. They have a reason why they're trying to, they really care about getting through it and then have the stability to get through it. I think it's some of the most import…

AI assessment note: “understanding what truly drives people and why they're doing what they're doing”

Q What about an affiliate model through bots? So when you order a, uh, in and out through a bot on Slack, you have an affiliate model.

A I mean, first of all, that's just, that's like a worst form of an advertising model, right? In general, that's like, Taking ads and making it more performance driven, right? This is the slow crumbling of my idea by some, but it's like, it's not like, I think, and the second is like affiliate models, just, I don't know, like there are very few companies that have made them sustainably good businesses, right? You know, I would challenge you to come up with a company that's deeply driven by affiliate fees that works in the longterm. You know, a commerce, it's, it's something I've learned a few times in my life. Like, Commerce is a very, very difficult business in any form. And I don't know, a world in which everyone's just collecting, you know, cents on the dollar, affiliate fees, trying to get you to use them and not someone else to buy something you were going to buy anyway, is, uh, it's a rough business.

AI assessment note: “there are very few companies that have made them sustainably good businesses”

Q The thing that strikes me always with, with bots and this kind of shift in excitement is will the transition to bots and conversational interfaces be like a major point of disruption or do you think it would be more an evolution in kind of the interface paradigm?

A Yeah, I mean that, I wrote an article from IWISE publication called The Information about this, um, a few months ago. I think that is the multi-billion dollar question. Um, The question is, is if all we're doing is taking a bunch of websites and apps, which are annoying to develop and slow and have all sorts of issues and putting a new front end on them, that's not that important. It might change a little bit the power dynamics of who's in front of the system like that. Again, it could be good from a platform perspective for a Facebook, but it doesn't really change that much. The question you have to ask yourself is, are there behaviors or interactions or business models probably most important that couldn't work before that can work now? You know, I think obviously in a bot world, a lot of the display driven, engagement driven advertising that, you know, we're all used to seeing as the way services monetize and exist becomes a heck of a lot less relevant. It's when you move where your application or your, your service is, you know, something people spend a lot of time directly looking at on a full page down to you get to push a message. The business has to change. And the question is, do the services change as a result or not?

AI assessment note: “are there behaviors or interactions or business models probably most important that couldn't work before”

Q No, absolutely. And we mentioned Avichal there. We mentioned a thousand X's. A thousand X's in the past generations have tend to happen and be produced from the valley. Avichal said on the show recently that crypto companies in large Part will re-centralize to Silicon Valley. With your experience and kind of hindsight that you have, would you agree with this, Sam?

A I don't know. I mean, I know Veatch pretty well. I have to listen to this episode. The reality is the world is far more global than it's ever been. You know, my wife runs an application called The Information, and I would argue the news coming out of Asia and China is read very heavily by people in Silicon Valley in a way that wasn't historically true and cared about a lot. I do think Silicon Valley is a powerhouse. I do think that just like New York is for finance or fashion, the Valley still is a self-fulfilling prophecy, where it is the Valley because it's the Valley, because people want to be here, and because basically a lot of great people want to come here, but look, New York is a real center of finance, and the intersection of crypto and finance, stuff's going on there, stuff's going on in Israel that's incredibly exciting, and then, you know, there's a huge amount of a market in Asia for it as well, obviously, so crypto is by far the most globalized The tech revolution or boom we've seen in my mind to date , I think the Valley has a very important role to play, especially from a technical contribution standpoint, but it's not clear to me that this is where the answer is in the way that it has been for other things.

AI assessment note: “it's not clear to me that this is where the answer is”

Q What's the backstory there? How do you split your time? What are the kind of challenges and opportunities of being both a VC and a founder at the same time?

A So I think, I mean, the story goes back a bunch of years, which is when I had my first startup called Drapio, um, in New York, um, many years ago, you know, I started doing a bunch of angel investing. And so I was, you know, one of the first investors in MakerBot and the first investor in Birchbox and, you know, the first investor in Venmo and a bunch of awesome companies. I found it really interesting because I thought that, you know, the experience of being a founder and actually building stuff actually I think makes you both a much better investor and also makes your advice useful. You actually have a relevant perspective to share. So I was at Facebook a long time after they acquired my company. I loved it. I have incredible respect for the company. Um, and when I left, you know, I knew I wanted to do more investing. But I also, like, wasn't quite done. I wanted to start another company. And so I had the opportunity to start an awesome company with a really close friend of mine, Andrew Cortina, who was the founder of Venmo. He also was finishing up at PayPal about the same time. That's where I spend the majority of my, like, hours, um, in the day. But then, you know, at the same time, you know, the, the, I've been at LP in the first few slow funds, which were usually just, you know, they were just a small group of friends investing together, you know, led originally by Dave …

AI assessment note: “That's where I spend the majority of my, like, hours, um, in the day.”

Q by mentioning the security versus freedom and the China argument. You recently tweeted about, I have been stalking you, yes. About the terrifying element of algorithmically generated fake video. Compounding this, you said, for the first time, we need to explicitly choose between security and freedom. So, how do you view the current landscape and perception towards fake news today, and what's that dichotomy and balance between security and freedom?

A I mean, this is, again, a whole, we could talk about this for two hours, but I think this is the challenge of our time, is that we have, in the physical world of the high friction environment, Built societies where we make choices about how free we want to be and how secure we want to be, and that's worked really well. Like in the U.S., we strongly believe in freedom of speech, but there's a limit, right? It's not completely universal. The thing about technology is because you're encoding things in binary, in yeses and nos, and kind of stretching into the limit, the reality is you have to make pretty fundamental choices as a society about where you want to sit on a spectrum, and I think the Chinese have made a sensible choice for themselves, and they continue to kind of do a great job with it. Which is choosing extremely high security at the cost of freedom, saying we're going to centralize everything, we're going to manage everything, et cetera, and that's where we're going to sit. The freedom argument would be, I think, is the crypto story, which is going to be basically saying, look, it's actually really scary in a lot of ways, but it does choose freedom over security. You know, how it relates to fake video and this type of stuff is, look, I think a lot about the story about the Tower of Babel. Babel? I never know which one it is. You know, it's in the Bible. But I think it …

AI assessment note: “you have to make pretty fundamental choices as a society about where you want to sit”

Q What do you think? Do you think they will? Do you think these platforms will accept it?

A I, I don't know what the business model is of bots yet. I mean, our business model for Fin is extremely simple, which is we charge people for it. And, you know, we charge people for it both because of the service we offer is expensive to run. It's, I mean, AAI is not an inexpensive thing to do well because it's not just bits. It's not just computers. It's not just processing. We do use people too, and we use great people, um, in the right components at being like the best humans. You can charge people. I think that works for some services. I don't think that works for a lot of services. Um, and I haven't seen anyone articulate yet a good bot business model. You know, the app store business model ended up not being a very good Right. And so question is, the question is, is like, is this next, what does this next shift really mean? Who are the winners out of it? Who are the losers? And man, if you had a good answer to that, that you were pretty sure was right, you should go out and do some stuff because that, that is an important thesis.

AI assessment note: “I, I don't know what the business model is of bots yet.”

Q brilliant brains joining my terrible British accent, and I'm just going to dive straight in. We're doing state of the seed market today. Sam, I'm quoting you. Seed investors need to come to terms with the fact that this is not an 18 month timeout, it's likely much longer, and even the death of systematic thematic seed funds. Can you help me understand, Sam? What did you mean by this?

A I think I said it. No, I mean, like, look, I think, I think, um, look, I've seen a lot of eras of seed investing in my life and career so far. Actually, hilarious. I was just reminded of this by someone, you know, my father, as he used to say, was an incredibly active seed investor in the late nineties in the East Coast. He used to call himself, he'd say, I was the big, he used to say, I'm the largest seed investor in the East Coast. Which is kind of like being the smartest utilities executive, right? And so I, you know, but he, you know, I used to sit in a little ton of meetings when I was a kid in the late nineties and kind of watch the rise in the 2001 crash, did a bunch of angel investing myself through the next decade, you know, ended up starting a seed fund with some friends, you know, in the last decade. So I've seen a few eras of this. And I think what was unique about the last era of seed funds, which really was the first time it got big and institutional, right? Was this era where really the role of a seed fund was to kind of be the first step in a factory production line of startups that you popped out the back end, one to ten billion dollar valuation companies, pretty predictably, right? And, you know, it was a factory system. You found a company, you packaged it up, you knew what the Series A firms were looking for, you knew the Series A firms, you say, okay, we're…

AI assessment note: “role of a seed fund was to kind of be the first step in a factory”

Q And then Jason, we chatted the other day about actually why party rounds and a proliferation of very rich tech execs from the liquid years we've had over the last years is more than ever. Are these two ideas arguing with each other? Are they the same? I'm trying to understand. Do we have more party rounds, less party rounds, more rich execs, less rich execs? Which one is it?

A I mean, my basic point, I believe, is I think that the YC thing in particular was if there's a factory to, if there's, if we're talking about venture capital as a factory, that was like ground zero for the construction of the factory. It literally was like, oh, venture capital and like how you raise money and form these companies is like this opaque playbook. It's hard to figure out. They're like, literally here is playbook and then here is demo day. We bring in a bunch of every investor they bid They bid it up. They justify our seven percent. We're done. It's packaged. It moves on the factory line, right? It is the purest part of the factory you could imagine, right? And, you know, as an investor, you know, we've stayed away from it for a long time, partially because there's no signal in it because everything comes off the factory line looking exactly the same, packaged properly. Well, it's fine, right? But it's packaged the same, you know, et cetera, et cetera. So to me that, you know, that's, that's the factory part is the dead part. Party rounds, rich tech execs. Sure, there are lots of rich tech execs. And like, again, seed investing is fun, right? Like, And you get 50% of your money back when it fails, right? So the nice part about mortality rate being higher is you get your money back faster at 50 cents on the dollar, right? Like, you know, as a tax write-off. So like, s…

AI assessment note: “I think that's more about the capital side, in my mind, than it is the factory part.”

Q And then Jason, we chatted the other day about actually why party rounds and a proliferation of very rich tech execs from the liquid years we've had over the last years is more than ever. Are these two ideas arguing with each other? Are they the same? I'm trying to understand. Do we have more party rounds, less party rounds, more rich execs, less rich execs? Which one is it?

A I mean, my basic point, I believe, is I think that the YC thing in particular was if there's a factory to, if there's, if we're talking about venture capital as a factory, that was like ground zero for the construction of the factory. It literally was like, oh, venture capital and like how you raise money and form these companies is like this opaque playbook. It's hard to figure out. They're like, literally here is playbook and then here is demo day. We bring in a bunch of every investor they bid They bid it up. They justify our seven percent. We're done. It's packaged. It moves on the factory line, right? It is the purest part of the factory you could imagine, right? And, you know, as an investor, you know, we've stayed away from it for a long time, partially because there's no signal in it because everything comes off the factory line looking exactly the same, packaged properly. Well, it's fine, right? But it's packaged the same, you know, et cetera, et cetera. So to me that, you know, that's, that's the factory part is the dead part. Party rounds, rich tech execs. Sure, there are lots of rich tech execs. And like, again, seed investing is fun, right? Like, And you get 50% of your money back when it fails, right? So the nice part about mortality rate being higher is you get your money back faster at 50 cents on the dollar, right? Like, you know, as a tax write-off. So like, s…

AI assessment note: “that's more about the capital side, in my mind, than it is the factory part”