E28: Andreessen Horowitz GP Chris Dixon on Crypto, AI, and Web3

26:18When they start off, they're quite the opposite, right? They're out soliciting and doing everything they can taking free photos, inviting you to parties, charging low fees, right? And then you and then And as the network travels up the S curve, you know, then what happens is, so you start off, they're kind of, this is true of Facebook, Twitter, all these things early on, right? You're out there hustling, trying to get people to come on, trying to get media brands to come on, being as friendly as possible, keeping rates as low. And then you start growing. Hopefully you're going up the S curve and then just like everybody's happy because everything's growing.

26:47With a growing pie, everyone's happy. And then you hit the top of the S curve inevitably because you run out of people in the world. and and at that point with networks they almost always start turning the screws and they say how can we continue the same revenue growth when our kind of core organic growth is is flattening and the answer is almost always more fees more you know charge more and it's not always like by the way like like an airbnb it's an explicit fee in other cases like with google search it's just like put more google properties in there put more ads in there you see on amazon now like more you have to scroll down further and further to get past the sponsored links.

27:26Right. And so it's just sort of this steady process of kind of extraction. And but because it takes so long, people kind of seem to continue to get tricked by it, like we're seeing it happen again now in the world. People are going to start building other people's platforms and they think it's great because they're in the track mode and they'll all five years from now be saying, oh, God, I should have listened to those people. But those those boomers actually knew something. Turns out like it's just that this is an inevitable process of these platforms is that they, they switch modes over time.

27:56I remember you guys saw the opening, I post about Sora, which is their video generation based on a single prompt. And it was like watching that tweet. It was like the sound of a hundred startups dying. Thinking about it's so good. It's like, wait a second. And it's just interesting how the platforms end up competing against their partners who are building on the app. Oh yeah. I mean, very few platforms don't, I mean, the, the, probably the biggest counter example is maybe windows like some some of the old os you know apple used to be like this i think they've changed now with the app store um but uh like to window i don't normally say that many great things about microsoft but the the um to their credit um you know they they always had the attitude of like we need to have the gates said this at one point we need to make sure that the developers are making more money than the platform you know in aggregate right he wanted but he didn't want one of them to do it because they'd have too much power.

28:49But like you wanted a whole bunch of fragmented ones to an aggregate. And that was always his model, which was a, which was a right and brilliant model. And, and part of that was keeping it, um, you know, keeping it predictable. I mean, to the fact, the fact that today I'm like gaming is very interesting, right? Because like Microsoft is essentially at this point, enterprise software company and a game company. That's their primary consumer business at this point. And yet the windows platform is extremely open platform and, and their biggest rival exists on its team. And so, you know, there are exceptions, but for the most part, and especially social networks have been very extractive.

29:23And I think it's just the new norm now. Like I think I would expect with all these AI companies, you're just going to see constant kind of platform and application kind of cannibalism as you're describing Antonio. Can we just go back to the person waking up from the coma that you're now telling them that Microsoft is the open company, that they're pro developer. They bought GitHub. GitHub, you can run Linux with it. It's insane, right? To think that like, you know, Zuck did this video on the Apple Vision Pro and he's talking about open paradigms and he's talking about Microsoft. I think the 18-year-old Zuck probably had a very different opinion of Microsoft.

29:57And so it's just amazing to kind of see how these things change, right? Yeah, it is interesting. And, you know, I mean, look, part of it, I think I tend to think everything can be explained through incentives. You know, and if you look at Microsoft today, they, like I mentioned, they really are, I mean, primarily they're an enterprise software company that exists through bundling, right? I mean, so how'd they get Azure so big? They got it so big by bundling it with, you know, their SLAs with office and Active Directory and all their enterprise products, right? And so they they go to an enterprise and they say, we'll just be your partner for everything.

30:32And so there's no, you know, their lock in comes through distribution. They don't need to have lock in through software in that sense. Right. And they kind of gave up on some ways on on Windows outside of that bundle. and so that actually made them in a lot of ways the most aligned with the open source software in an interesting way. Facebook is the open source AI champions that's another thing I wouldn't have predicted five years ago. I think it's great if they keep it up. I was a little surprised when Zuck said that they were open on MetaQuest because I don't think of MetaQuest as very open.

31:06I had a bunch of friends who've been very frustrated trying to build on MetaQuest but it would be nice. Chris is very humble. Let's, let's remember that before Chris went all in on crypto, he was also the earliest investor in Oculus. So I think, uh, you know, you, you've been also interested in AR VR stuff for a while, right? Yeah, no, I mean, I, I, like I've been writing for years that I thought that, so I, I've always had this theory for, I don't know, I've been writing about it for 10 years, blogging about it. Um, that, uh, you know, the last wave was mobile social cloud, right? That was 2008 to 13.

31:38And so sort of at the time, there were all these people saying, which is going to be the real trend? And it turned out the answer was all of the above, right? And moreover, the three trends reinforced each other, right? So phones were the, you know, got us from 500 million to 5 billion devices. Social is the killer app, two and a half hours a day. People spend social media that made phones so popular. And cloud is the backend infrastructure that enabled it all to exist, right? And so I don't know. But the theory I've had for a long time is AI, VR, crypto, right? And VR, like, look, I think VR is nuanced.

32:14Like, people also say spatial, Apple saying spatial, metaverse. You can ask questions like, is it the headset itself? Or is it just the idea that people are going to be spending more times in 3D experiences? um and you know i think i think i think that the the simple almost non-debatable thing is the idea that people are going to spend you know that right now people spend x number of hours on average in three-dimensional experiences versus two-dimensional and that number is going to go up way up right um and then moreover maybe a lot of that's done in different devices and headsets and things so that's metaverse spatial whatever you want to call it xr or something um and then ai is obvious, right?

32:55I mean, we all know that. And then I think kind of crypto blockchains is the other one. And I think the crypto blockchains is the most interesting to work on as a VC. Doesn't mean it's more important than AI, for example, or VR, but it's the most kind of disruptive and aligned with new startup innovation as opposed to things that were valuable accrued to the incumbents. Now that you mentioned AI, can I ask my AI question, Eric? Please. so you have you have a chapter at the end of the book with which is like applications of crypto and one of the more more interesting ones i think is the ai one and the the problem there is um and i think you frame it perfectly i i actually um today i actually had an early review copy believe it or not i use it i actually do ocr on it and i paste it in our slacks or everyone at the company would read it uh i don't know if that's piracy it would be speaking chris i'll help you the whole book or just that chapter no no no no just just the chapter on that and then and then i've also bought actually the hardcover and the kindle version so i think i made up thank you all right you're super because i had to i had to reread it when i was here on paris and i didn't bring with me anyhow yep you have this great setup in which you're talking about the google one boxing phenomenon right like the way the web has been paid for for 20 years has been google indexing the web they upstream you on ad revenue but they give you distribution and there's some equilibrium there and that's that's how the web has been paid for 20 years and that's just kind of going away and that people are you know i always use the zoomer in my office as like the canary in the cage for the ones are glowing that guy doesn't use google he uses perplexity or chat gvc or whatever and so like they're using the content but they don't they don't drive the actual traffic will make a problem there and so how do you actually get the content owner paid and you sketch out this very elegant sort of blockchain based solution for it which i thought was kind of brilliant i don't want to yeah yeah the intersection is fascinating yeah yeah so that i so i call it um sort of a way to describe it as a sort of an existing covenant, I call it between distribution and content on the internet.

34:53And so what I mean by distribution, think of a social networks and search engines, like the place you go to find stuff, right? And then content is, you know, news sites, media sites, websites that get indexed and linked to by those sites, right? And what's happened over the last 30 years is there's evolved kind of this economic relation, symbiotic relationship between those two parties between the distribution of the content right and the basic deal is um the the content sites if i'm a news site i say google you're allowed to index me um you're allowed to crawl my content you're allowed to show snippets in the search results but in return i expect you to send me some amount of traffic back right some some some percentage of the people that search for things and see my results will click through and i'll get traffic back.

35:40Right. And that's basically this sort of equilibrium state that, that drives the entire internet economy. I mean, it's very important. Right. Um, and, um, as you describe in a world where you can now get the answer. Um, so you say, I want to, in a world where, you know, you say, uh, what are 10 restaurants to go to New York tonight? Where should I travel for vacation? And instead of getting 10 links that you click through many of them, you just simply get like a perfect answer from an AI bot, whether that's today or in two years, that's clearly the future. Right. And maybe by the way, there's a link like perplexity, but like, if, again, if you get the answer, do you really need to click through?

36:18Right. So in that world where you just get the answer, like what's the new covenant, like how do those, cause that content came from those websites. Right. And, and those, those bots learn from them. Um, and maybe the answer is there is no covenant there. And you, but you mentioned one boxing, one boxing is the word for when in the past is sort of in some ways there's a canary in the coal mine right it's uh what they google would do over the last decade is sometimes um for certain types of content they would simply show the answer so for like music lyrics um for some programming questions like i was on the board of stack overflow for a long time a programmer q a site and that was always their biggest concern was we get one box so instead of you type in a programming question instead of getting a thing with the link and you click through and go to the website, you just get the answer, right?

37:09Because if you don't go to the website, then they don't get to like some percentage, don't convert and blah, blah, blah. And the whole business model breaks. And so, you know, so, so what's the new covenant in that world, right? And so one answer is maybe there is no covenant and maybe all the content sites are screwed and the internet turns out to be five sites, you know, five AI bots that just give you the answer. And maybe that's okay. One One question that arises then is what happened? How did those bots get trained on new content? You know, new, new genre of music comes along a new piece of news comes along, whatever, you know, news news category comes along the world changes.

37:44Um, and today, you know, they rely on the content sites updating. And if those content sites don't exist anymore because their economics are nuked, like how do they get that? And maybe they get it like the answer, the kind of dystopian answer, which is happening today is they get it through offshore content farms. So you've got teams of people in the Philippines typing in, you know, you've got some new cool programming language in San Francisco, and you've got teams of people in the Philippines doing Q &A to train the AI bots on the new programming language, right? And, you know, so that's one outcome of the world where you have, so we have five big AI systems, content farms, training them with new content.

38:26that seems kind of depressing to me, right? We lose all of this rich, vibrant ecosystem of the internet. So another possibility is we develop a new covenant, right? We say, what should the economic relationship be between you're an illustrator and you design some beautiful thing which gets input into an AI system as part of its training, that AI system in turn gives somebody some valuable generative art. and maybe your art contributed one ten thousandth of that in inspiring that AI system, should there be some economic model where money trickles back to you based on that contribution? Right. And yeah, and as you described, Antonio, a blockchain is a natural way to do that.

39:11A blockchain is a system for creating digital economic covenants. Or another way to look at it is, it's a good way for these millions of websites to get together and collectively bargain against these five big systems and sort of come together and say, Hey, you can use our content, but under these terms. Right. Um, and so, you know, yeah, I mean, so essentially, um,

39:42and I guess I would finally say that I don't see a lot of people talking about these issues right now, like, I think I feel like with the AI stuff, there's a lot of excitement around the kind of initial applications of the technology. And I don't see a lot of thinking around the second order and third order kind of consequences around the economics of the internet, um, and the economics of creativity in general. Um, and I think we should be thinking about that. Cause I think I don't know, my own take is a far more realistic risk with AI is not that it takes over the world and creates bioweapons, but that it just, further concentrates power on the internet and further diminishes the role of creative people.

40:25And it's, it's a very much more mundane set of risks, but I think more realistic and ones I worry about more. And that I think now is the time to think about it. Like if you go back by analogy and think about the internet, like, look, a lot of people that first covenant I described, a lot of people weren't happy. Like news corp spent years fighting Google and suing them and eventually reached a settlement because they didn't like that covenant, right? They didn't like the fact that Google had so much power. Um, the time to have really thought about that for the content sites would have been the nineties, um, would have been early on or maybe the early two thousands, right.

40:59When this, when these systems were developing and these, and these systems were getting, these content distributors were gaining, gaining all their power. Yeah. I'm sorry. What could have been done back then? Yeah. It's a good question. I mean, you know, I give the example of the book, NASPERS just said, screw it. We're becoming a internet investing company. They were a South African newspaper chain. That's now I think worth$90 billion or something. They became one of the savviest internet investors. So one thing is opt out or play the other game. Um, no, I think, I mean, people had solutions back then, like robots.txt, like you could, you could opt out.

41:37Right. But of course the problem with opting out is that if you're the one person to out of the million, like it's not going to do anything. Someone will just step and replace you. I think you could have come up with, you know, I mean, like you didn't have blockchains back then, but you would have needed to come up with ways to collectively bargain essentially, right? Or to have the group come together, create industry standards, consortia. I think blockchains are a better native kind of digital way to do it, but, but you would have had to have some kind of system set up. Yeah. I don't know. It's a good question.

42:09I mean, you didn't have, you didn't, I don't think you had all the sort of technological pieces you have today to do things to counter it. So the trippiest thing that I think about, Chris, is like agentic AI on the blockchain, meaning AI that's kind of at the metal level of given rough goals. And then it uses the AI to actually how to reach those goals. Because on the blockchain, like one of the reasons why I got kind of crypto pilled is because like, if you consider the world as becoming natively digital and natively sort of ritualized, then the blockchain is a natural economy for that, right?

42:35Like, you know, I made the joke trick recently. I was, don't ask why I was shopping for diamond rings and I was comparing the four C's and all this bullshit and the rarity score. And it's like, these are like NFTs. There's artificial scarcity. There's a rarity score. And a little bit of this is about money laundering and tax evasion, right? This is basically physical NFTs of what diamonds are. And so it's like, so if goods are not becoming little hunks of carbon mines in South Africa and that they're becoming this other thing, then the blockchain version of it is exactly the economy. That's how I got crypto.

43:03Then you add an AI, which is obviously living in a virtual plane. And if I were to bet there being some sort of AGI Terminator style apocalypse, it's probably going to happen on the blockchain in the sense of there's going to be some runaway AI. That's what we need. We need another negative blockchain. No, no, no. It's okay. But that's way more likely than any of Yudd's crazy virus scenarios or whatever, right? I don't know if you ever think about that, or maybe it's just - Well, no. I mean, what is, what is code running on Ethereum? It's autonomous code, right? It's code that it's, it's, I mean, the, the way I sort of describe blockchains in my book is they invert the power between.

43:43So in a traditional computer, you have hardware and you have software and basically the hardware is in charge of the software in the sense that the human or the company that owns the hardware can ultimately change the software. Right. And that's why ultimately, like when I'm running, you know, I'm using Facebook services. In the end, I'm subject to the humans behind Facebook and their decisions, right? It's how they use my data and how they charge me and the rules of the API and everything else, right? And so the kind of core innovation of a blockchain in my mind is that it inverts that relationship, that power relationship between hardware and software.

44:16And that when you write code on the Ethereum blockchain, that code will continue to run. The promise the blockchain makes is that code will continue to run as designed, even if the hardware operators who run the code, the Ethereum validators, try to subvert it or undermine it. Right. Um, and, and to your point, like, you know, a, a consequence of that is you could imagine a world where somebody is writing blockchain based code that is fully autonomous and immutable and, and, you know, has no kill switch. And, um, this is by the way, a debate. Um, so the EU has a, has this thing called MECO, which is their regulatory proposal for crypto.

44:56And they're requiring a backdoor in every smart contract. Um, and I think part of the logic is sort of what you're describing, which is. Now, of course, the problem with that is the backdoor more likely will be used by the next hand being retreated. And like it undermines the whole kind of value of a blockchain, which is not to trust an individual. Once you had a backdoor and you had a sketchy person, you've just undermined the whole value of it. Right. And so, you know. Anyway, so I think, yeah, I think there's there's some sense in which that's a. That is an interesting kind of long term possibility.

45:30do that and every background to every smart contract when i start pooping on the europeans now that i'm actually physically in europe and i like flip my narrative to the other direction but um in any case it's obviously not gonna happen but um cool um okay and so is the idea chris we talked about how you know web 2 giants had these you know different kind of moats that web 3 projects don't necessarily have but that web 3 projects will be as big or bigger because they'll be able to price discriminate more effectively or they'll have more liquidity?

46:05How do you respond to why will Web3 projects be as big if they don't have some of the same advantages of Web2 companies? I mean, I guess my simple mental model for tech, the whole tech world, is it just keeps getting 10x bigger every 10 years or something. You go back to the 2000s and you tell even the most... Mark Andreessen is probably the most optimistic technology person I know. And if you told him in 2007, if you asked him how big someday could Facebook be, I bet you he would have said 10 billion, 20,$30 billion. I just saw yesterday's 1.2 trillion or something, I think, meta. Nobody would have thought any of these tech companies could be as big as they're, even the most optimistic people, right?

46:50so i look and and by the way meanwhile like excel windows quicken they're all still going strong like so i think the way to think about tech is like the the stuff that works kind of just keeps working and then you just keep layering on these 10x bigger things um and you know someday that will stop but i i don't see it stopping anytime soon if anything i feel like we're heading into like maybe the most exciting growth period yet in 80 years of computing with all these different things happening. And so I just think the next thing will be 10x bigger. And so if even if part of that is you have deflationary economics, which is what we're describing here, like Farcaster would not capture nearly as much of the money flowing through the network as a Web2 social network.

47:34But if the world is 10x bigger, and it's the social substrate, and some portion of the money is flowing through and we're you know people ask me a lot this a lot as an investor like what's the business rationale if we're early investors in the future social economic substrate you know and the tech world gets 10x bigger and even if the economics are significantly deflationary which i think they will be um it could still be a very big venture venture investment right so that's kind of how i think about it right it's the same way same logic by the way people have used on the internet investing forever.

48:08I mean, Craigslist, you know, deflated classifieds, right? TripAdvisor deflated travel agents, right? But they end up being interesting businesses because it's a very big market. And then you have this effect where like, at first you take away the old business, but then it turns out because it's the internet, there's all this other cool new stuff you can do. And you end up growing it into a whole new interesting business as it was Uber. You didn't just take away taxis, right? You created a new behavior of people, you know, and grew the market. Right. So I don't know, that's just kind of how I think about the business rationale side of it.

48:42Can I ask, so one question that I have, and I've often had, I'm going to just ask the questions I've been accepted with crypto and hopefully let's go, let's go deep. Let's do it. I do, you know, enough podcasts where we, uh, we want to make this the 201 version because our audience is a little bit, uh, more sophisticated. Yeah. Yeah. So one question I've always had is why it's taking so long. I mean, present company excluded, of course, with Podcaster. Why is it taking so long for WebTweet and consumer to take off? Because if you were, you know, rewind 10 years, you know, every random startup guy was running around South Park with like a new color filter app that was like Instagram.

49:17It was always a consumer facing phenomenon. Like that's where the most wild catty sort of entrepreneurs were focused. And if you go to crypto and I mean, and this validates the fat protocol or value accruing to the protocol rather than the app later thesis. you've got the guy creating the you know gasless bridge transactions on an elf and like you know this slew of crypto terms and it's some of them actually drive huge liquidity some of them are clients and but it's like it's like 400 daily transacting wallets right and it's like what why isn't it that you don't see the huge focus like and then you know back in the old days of facebook back when i was there it's like the consumer base was this big and the problem was trying to chase and create a business around it and here we've created so much checkbook on business infrastructure and financialize it but it's like where's the consumer side and can it be that blockchain kind of start on the other side of it it's kind of trying to grow towards the consumer side or what do you think it's a good question i think that like i think like as a participant in it i can so i think like i think infrastructure ends up being very very important um and i will say like that what happened with ai recently was a good reminder of that which is like i like i started an AI company in 2008 that I sold to eBay in 2011 called Hunch.

50:31And if you asked me in 2012 or something why it didn't work as well as I hoped, I would have given you a list of answers like the UX, the design, the this or that. We should have started with a vertical. And then I looked at what happened in AI over the next decade. And I think the answer is very simple, which is GPUs. like is just that like the the the systems just got 100x better with the improvements in infrastructure it's not just gpus it was you know the transformers paper and they you know had more data to index and all the other kinds of things but it's but fundamentally like i i think with the iphone like i think an interesting question is had steve jobs not done the iphone like so smartphones right you had a 15 year history of people trying to create smartphones before the iPhone.

51:20Um, and, and there was a whole wave of smartphone entrepreneurs. I was an investor in some of these things or, or involved with some of these things. Um, had friends doing it early two thousands. There was a whole wave of, of a smartphone video game companies, believe it or not, these are on feature phones. Um, the whole thing is you had to go to New Jersey and convince Verizon to put you on the quote deck on the, on the homepage. Um, you know, why, why did the iPhone finally work in 2007. Like now obviously Steve jobs, the genius, but like, had he not done that, would we still be using feature funds today?

51:57Like I would argue that he may have accelerated it a year or two, but fundamentally, if you look at the improvement curves of all of the key components of a smartphone, that it basically hit a tipping point, you know, between 2007 and nine to, you know, touchscreen capacitive screens, the, the modems, the, you know, the core processors, like all the different kinds of key components. Um, and so probably had he not done it, someone else would have done it. Um, essentially that saying that you sort of infrastructure determinism, right. That, that it's similar to the web two stuff I talk about in the book and all the different social sharing and video on the internet, like how much of that was the genius of entrepreneurs, like the founders of YouTube and how much of it was just broadband penetration hit, uh, whatever 40 % or whatever it was in 2005.

52:47And it's just, once you had that, you were going to have video on the internet. And by the way, video on internet, like real player was a public company in 2007 or something. I mean, it was a whole thing, sorry, 1997. It was a whole thing for 12 years on the internet and didn't really work until finally broadband hit some point. Right. I don't know. So I've just over time come to really believe that these kind of core infrastructure things matter. And so speaking of blockchains, right? So you had, you had Bitcoin 2000 launching in 2009. And then, you know, as you described, like kind of 2013, you had people running around trying to create consumer apps that meant creating things on Bitcoin.

53:24That was very hard to do. It's still very hard to do. There's a bunch of reasons why, including the 10 minute block times and high transaction costs and everything else. So then, you know, then you have 2015, you have Ethereum, which is finally programmable, but slow confirmation times, proof of work, which, you know, led to lower transaction finality, high gas fees, right? And there's been a whole kind of story of now we have L2s and we have a whole bunch of other improvements. And I would argue probably the infrastructure is still not there yet. I think you need to have sub, in many cases, penny fees, you know, on these chains to have true consumer apps.

54:09If you look at what, you know, Dan has done, he's managed to do that with Farcaster, but he's done it by doing a lot of kind of technical gymnastics, right? because you can't store everything today on a high quality blockchain like aetherium even an l2 without incurring fees that are too high for that application so i do think a lot of it is that i think we were probably just a little bit naive or definitely naive at some point 10 years ago about the infrastructure being able to support these things um i think part of it is the casino stuff I mentioned it just like a lot of people got distracted.

54:48You know, when it's you can go off and look, I mean, what, what, what is one of the applications you can build when it costs$5 a transaction? It's trading, right? So you, so you had, you know, 2021, like defo the 20, 2020 is like DeFi summer. Um, and the success of all of these, you know, kind of trading applications. and that was a lot of it was due to the fact that that's one of the only applications where you can charge somebody five or ten dollars you know if they're going to make a thousand dollar trade or whatever it might be um so you know i think one version of history which i kind of believe is that you'll just sort of see as you'll see these um there was as you see these declines in gas fees you'll see just sort of new classes of applications open up yeah maybe just to like build on what Chris said.

55:38And from my perspective, it feels like basically the technology was much lower level than we originally got excited about. Like when I joined Coinbase in 2014, you know, Chris brought this up. The two things that I was excited about was it was only a Bitcoin company at that point is Coinbase had an API. So I was actually modeling it to like, oh, maybe they can build like the Facebook app platform, but with money. And they had commerce, which was like, okay, maybe all the people are going to take, you know, internet payments with cryptocurrencies. All that ended up being too early and wrong. But I think the challenge was, it took us a really long time.

56:12We're just getting to a place where these L2s are still expensive, but they're significantly cheaper than any of the L1. And that's just in the last year. And I've been in crypto 10 years this year. And I think that the second thing to what Chris points out is the core original innovation with Bitcoin is money. And so when you have the ability to create new money, different types of money, value that can be scarce on the internet very easily, you can naturally attract a bunch of people who aren't actually building the hard things, the consumer applications that take a lot of work to actually go figure out, especially on a new stack.

56:52And you just get the pure memetic component of things in terms of tradable tokens. And so So where I'm optimistic, having been in that space for 10 years, is I think Coinbase and that generation of companies was critical in terms of like wiring up the world to crypto and making the ability to move between kind of like our analog to digital system in terms of value really easy. And that's never been easier. There are now all these kind of tools that you can basically embed an on-ramp into your product and consumers are using a normal payment method. But we're only finally at the place where the infrastructure, and this is someone who's been spending three years building like an application to what chris said is like i have to do all these these gymnastics you know cryptographically verified but in a way that is um you know kind of optimizing around cost it just feels like we're in the 56k uh modem era and and we're just starting to potentially you know eke into broadband which i think obviously for the internet there is in a much shorter time frame in the scheme of things Well, maybe not.

57:55You had the whole dot-com boom that was predicated basically on 56K stuff. And then you had this huge build-out of infrastructure, dark fiber, Google, all that kind of stuff. And then Web2 came as a result of broadband just actually finally arriving. And then now you can build YouTube, which you couldn't have done in the dot-com era. But people still had high-flying video startups. I mean, I think Mark Cuban basically had a like a media company startup that he sold, bought the Dallas Mavericks and never ended up launching something even remotely important close to YouTube. Chris, what would you need to see in order for you to think, hey, this Web3 vision is actually happening to the way that I want it and not just sort of a projection of the future?

58:37And the flip side of that question is, what would you need to see to believe, to be disappointed and say, hey, maybe this open vision for the future of the web might not happen the way that I want it to happen? Yeah, that's a good question. On the first, I would say, I mean, like, I think what we want to see is to Antonio's point is we want to see applications that end users are using, right? That are consumer applications that are used by, you know, right now we're in tens of millions. actually people people think it's lower than that but it's if you look at all the data um from the wallet providers for example like metamask and phantom and stuff there's tens of millions but i think we want to see um hundreds of millions and sort of and and in use cases that aren't sort of financial that are social that are kind of across the you know games media um so you know i think that in some sense it's relatively easy to measure the success right is you want to see the time an average person is spending, you know, some, some portion of that starts to go to these new applications and that grows over time.

59:49And I think, look, and I think just based on the conversation, you know, the discussion earlier, I think if you believe in the sort of infrastructure maximalist view that I, that I believe in, that, that we're, you know, that should be in the next, in the near future. I think that could happen now or it could happen in a couple of years. On the negative side, I guess if I'm the last person, I don't build these things. I write a book and I invest. If at some point I'm the only one talking about this stuff and there aren't people like Dan and Antonio building things, at some point I'll just be some crazy guy in the street babbling about the old days or something.

1:00:32So, um, at the moment though, I think there's, we have critical mass of great entrepreneurs. And so, you know, I, I, my, like my experience with like, like with, I, so I was very, probably the most active I was. So in investing, um, in the, or one of the most active periods I had investing was 2009, 10, kind of right at the early iPhone era invested in, like, I co-founded a fund with some friends called founder collective and we invested in the first fund was just like you know uber and venmo and stack overflow and trade desk and coupang and it was a great era um and there you know there were maybe i mean there were a lot of good entrepreneurs then but i mean we knew them all like it wasn't like it wasn't 10 000 it was like a couple hundred probably um you don't need like i don't think you need that many it's much more about the quality and having you know tens maybe hundreds of really good teams to to to really kick off a movement and then of course what happens right is you have a few successes and then like if there's one superpower of silicon valley it's the sand monster reconstructing itself you know, go like the new thing pops up and suddenly people run over there.

1:01:54Right. It's the, I call it the kid's soccer thing, you know, where professional soccer, everyone sits in their position and kid's soccer. You've got 22 people running after the ball. Like this is both the strength and weakness of Silicon Valley. Right. Um, and the one, it's a weakness. It's sort of like, okay, now this, this month we're all into this. The strength is it's really, really fucking good at like, if something's working, throwing more fuel on the fire. Right. And so, So I sort of think of it as we have the Dans and Antonios and another 50, probably a couple hundred really kind of top end entrepreneurs.

1:02:29And they're kind of the pioneers and then a few of them strike gold and then you'll have all the reinforcements come. So just one comment on that and it'll lead to a question. I think the thought of crypto flaming out, I don't think, I mean, anything's possible. I don't think it's particularly likely. I think what might be more likely is the fact that crypto doesn't, in some sense, sells out. Let me explain what I mean by that. Part of the reason why I also got crypto pilled is that like the blockchain is one of these like industry changing shifts, right? That just remakes things like, you know, one of these DC tourism is that no incumbent ever gets unseated, right?

1:03:04Like Microsoft never lost a desktop. Google never lost search. Just those platforms become less important or this mother platform that becomes more important. So the same thing with like blockchain appearing and all the same fan companies, not just those five, but broadly speaking, fame, like that we've been talking about for the past 10 years, kind of either go away or are less important than what's going on. Watch in, which part of my traitorous little heart kind of likes the fact that the companies that I've worked for and been involved with are like suddenly pushed out of the way you think it's, it recreates a lot of the excitement years ago, but like, what is, and like to me, that alone is, is reason to get into crypto.

1:03:36But what if in fact, right? Like some of the tendencies you see in crypto that re-centralize things in weird ways or like specific chains, or I don't know if you saw this day, you won't believe it, but one of the hot new features that Forecaster shipped was this thing called Frames, which is super interesting, and I did a post on it. It could be a whole separate podcast on just that. They're launching an L3 that just is for Frames. Now we've gone from app-specific chains to feature-specific chains. I look forward to the L4. Who's launching a Frame chain? Do we want to put companies on blast in the podcast?

1:04:12No, no, no. I'm just like how fast the space is moving right now, which is exciting. It's like, I didn't even see that. It's my own thing. I didn't know. I mean, this is like a college humor video. Like it was an hour ago. Maybe it's already failed. I don't know. But what they literally tweeted it. And that's it. That's everyone. I don't know. But no, no, it's for real. Like they actually tweeted like an L3 around a frame specific chain. I'm hoping for an L4 just for the specific button on the frame. Like we need, we need specific button. And I'm half joking, but you see how like a lot of the blockchain, you can get re-engineered, re-engineered in such a way that it actually contradicts some of the original founding principles.

1:04:42And it just becomes another way to share a state, which again, it might still be very effective and it might still be something worth investing and working in, but it may not be like the dawn of the new era that maybe some of us hoped that it might become as, as reality kind of sets in and the need to make a buck kind of sets in. So you're saying specifically that the architectures end up being centralized.

1:05:06Centralized are so specific that like, where did composability go? Where did interoperability go? Where did a lot of these good features go? No, it's definitely like, I think that's possible. I think another failure case is just, I don't like I said, like I with you, Antonio, I don't think crypto, I don't think it's very likely it's quote fails at all. I think the more likely kind of quote failure is these degenerate cases that you're describing, like it's, it sort of sells out like that. or I think more likely is just more and more kind of trading and financialization, which to me is another form of kind of a failure case.

1:05:43I guess maybe in the minority here, but to me, Bitcoin becoming a store of value and not a payment system was a failure case. That was not how I want to see it play out. And so, you know, so. Man, the Bitcoin rage we're going to get in the comments are going to be out of control. They already hate me, so I don't know. But here's the kind of counter to that is that, you know, if it is store value, but we have stable coins and we get, you know, lower costs, then the market is solving for it. Oh, no, and I agree with you. And I think in the end, it all worked out. I mean, I mean, in that sense that like those people moved to Ethereum and we created stable coins.

1:06:20And so I think the market is solving for it. So I'm not saying that, but, but I do think in that narrow case, like that community, just sort of number go up community. And like, I would, you know, if Ethereum ever became number go up community, which is very much not today, that would be depressing to me. Right. That's, that's, I'm trying to, to play out the negative scenarios here. And I mean, look, just look at your original post you wrote about Coinbase when you invested in terms of like the headspace that you were in, in terms of what you were excited about Bitcoin and kind of, I think some of your early crypto investments were very payment oriented.

1:06:51Right. Yeah. Yeah. Yeah. Yeah. And so, I mean, like that's the, in the first paragraph of Satoshi's white paper, it's small, casual payments. I know sometimes, sometimes in the, they forget the first paragraph of the scriptures. I don't know. The heretical sect. So I like, I mean, also this going back to the book, I mean, part of what I thought about the book Antonio is that like, I just, yeah, I assume crypto will come back. It is coming back. It always comes back. Like people, I don't know, people love, people love crypto. They love, they, they, you know, it's been 15 years. The skeptics say it's a bubble.

1:07:28I, I went back, I would challenge listeners to find a financial bubble in history that lasted 15 years, um, and went through four cycles like this, you know, up and down. Um, it's just, is not the case. Like there's something deeper going on here, which is, I think that I think fundamentally the idea of kind of user owned and operated internet systems, whether they be Bitcoin is a, you know, user owned and operated financial system or Ethereum is a user owned and operated computing system. Like that just resonates with people. People like it in an era when they don't trust institutions as much and you can have trust code, not people.

1:08:09Like it just resonates now. You know, part of why I wrote the book is I think there's sort of a right way for that to play out and the wrong way for that to play out. And the right way is what I try to guide towards. And, you know, and that sort of you couple that energy with useful applications that start to supplant a lot of the big applications that the Web 2 giants have created. And that, that to me is the, the, the great outcome, the ideal outcome. And then anything else when it's just sort of marginalized as a financial thing or meme coins and things like that is the failure case. I mean, bubbles are, if anything, a bull signal, right?

1:08:49Like every financial instrument in history started with a bubble, whether it's the South Sea bubble on equities or credit derivatives and the little credit crunch I was marginally involved in. There's always the bubble. If anything, that's like... But people compare it to like tulips and Beanie Babies. And if you go look at it, I was just looking at Beanie Babies as like three years, tulips. They actually did debate whether that really even happened. Apparently it was much... Yeah. It was like a very narrow phenomenon. and it was very short-lived. And anyway, so, but. Well, yeah, and I think that the other thing that people are easy to dismiss is despite the kind of asset price cycles that crypto has gone through, if you just look, are there more people working professionally building on permissionless open blockchains today?

1:09:34You're like, the number keeps going up, right? Yeah, no, it keeps going up significantly, yeah. And so like the technology is not going away as much as the haters want to kind of focus on the negative elements, which they exist. Yes. But and are they exacerbated? Because basically this technology allows you to create scarcity and effectively money or different forms of it. Yeah. And it's global. But I think the like I've been in crypto 10 years. I genuinely like I have not been more excited about the possibilities of what you can actually go and build. like you know we we got l2s last year like for those in the audience that don't pay attention like bitcoin ethereum slower and expensive um and then you have these kind of like their l1s layer twos obviously uh kind of like one step up if you think about from a layer standpoint they're cheaper and faster they're only going to get cheaper and faster this year and then we have all this like new wallet infrastructure so instead of having someone to go download this app It's just going to kind of magically happen within experiences.

1:10:36And so I think you're finally going to be in a place over the next year or two where great consumer app builders, the people who would go and build an Instagram can come build something that kind of, and I've been using this term like the, you know, Web2, Web3 mullet, where it feels completely Web2 oriented in the front. But the power of the ownership component, to Chris's book, from crypto and Web3 can actually happen on the backend side of things. And so I'm really optimistic that the 10 years I've been in crypto, the last six to nine months, from just the pure availability of infrastructure and tooling, has had a huge, huge change.

1:11:15I agree. I think it's, I think it's, I think you're either now at, or a year or two from now at the kind of the, I believe we're at that point, the kind of the tipping point that AI hit and smartphones hit and things I was describing before. Yeah. Looking forward and gearing towards wrapping up, if we're having this conversation a couple years from now, what are the biggest questions that you yourself, Chris, have around how are certain things going to play out or where are you most unsure of what's going to happen? Well, yeah, a big thing we didn't discuss is the policy stuff. I mean, I think that's like, I kind of feel like if it weren't for policy, so policy, basically, let me frame that, that I think the world has changed in that, that software existed in a libertarian sandbox for its entire history.

1:12:11And that ended probably 2016. seen. Um, once people realize how impactful social media was and that, you know, we sort of overlook this thing that's ends up being really impactful. And so I think that, and obviously there's been a lot of regulatory, um, actions and discussion around, uh, crypto and it's starting to happen around AI. And I think it's just the new normal. I think this will happen with every new tech sector. Um, I think it's a matter of time before, you know, Oh, VR is, VR is killing our kids, you know, et cetera, like you name it, new area of tech is going to have a policy implications.

1:12:48And so my, my thinking is if, if it weren't for that, I just have a hundred percent certainty that you sort of let crypto go off and do its thing. And it's just evolutionary thing. It's going to grow into something big and useful and interesting in the same way that open source did. It was chaotic, like crypto is in the same kind of lineage as open source and that it came from the edges, it's, it's sort of driven by hackers, not by institutions. Um, as a result, it ends up being somewhat chaotic and, you know, goes off the rail sometimes, but then comes back on and eventually finds its way. Right.

1:13:25And so I, I'm sure that's what would happen. Modulo the policy thing. The difference now is you have these things happen and then you have regulatory actions taken, which could, could prevent that kind of natural course of evolution. um and you know and that's of course also part of why i wrote my book is that i wanted to i didn't feel like for that crowd that the positive side of blockchains had been properly um presented um and so you know it was and i think it's like my general view of policy is there's sort of two ways to do it and i think it's going to play with ai too by the way like one is let's call it reactive and the other is proactive.

1:14:05So reactive is stuff happens, people build things, and then there's court cases and that takes years to, so open AI. So there's copyright questions around AI. New York times sues open AI that will go to court. That will take three years that will get appealed. And six years from now, we may have a judge interpreting text from a hundred or 200 years ago and trying to decide whether an AI system is copying or is inspired by a piece of art. Okay. which is an interesting, deep philosophical question, which I would argue is not answered by, probably by a piece of text from 200 years ago. And the other way to approach it would be to say, let's look at the technology expansively.

1:14:46Let's look at the good things about AI and the bad things about AI, and let's craft a policy proactively that maximizes the good and minimizes the bad, right? And that would probably be done through legislation and not through court cases. um same thing with crypto like a bad thing happens we go to court it takes six years we try to interpret some rule from 100 years ago um or we say here's the bad we've seen the bad ftx etc here's the good that's what i try to show in the book let's craft a policy that maximizes the good and minimizes the bad right um but now unfortunately you know at least with crypto and probably with AI soon, it's gotten mixed in with the broader culture war and tribalism has kicked in and to your point, like the reason I can attract these haters is I'm seeing, you know, as crypto is, as, as taking a side in the culture war, which I'm not, but like, you know, it's seen as because the book is pro blockchain, it must be, um, by some crowd be deemed the enemy and, you know, and attacked um and so that to me that look that's the big overarching question right now um and i think it's and like there's just so many different things at play here um there's not just crypto as i mentioned there's other areas of technology there's these sort of pro-tech anti-tech sentiments more generally there's the international angle there's the eu there's asia there's this there's a meta battle going on between the right and the left as to the power of different branches of government.

1:16:18Um, so there's many, many forces here and many, you know, far beyond my area of expertise, but, but, but this is all mixed in, like you can't separate the two the way you used to in tech. I think you can't just sort of look at it as a technology phenomenon anymore. You have to look at it in this broader political, cultural, et cetera, consequences. Right. Um, and yeah, so that, so that, I mean, that would be the other thing that, that I just think when we look back, I think that was your question. Like we look back 10 years or, you know, I think how that played out. And like, I would argue that the most important thing is this broader kind of, I, in an ideal world, I'm probably being naive.

1:16:58We would, we would sit down and all look at the tech. I am being naive. Look at all the technologies, you know, in their fullness and they're good and they're bad and try to optimize the good and minimize the bad, as opposed to calling each other names and trying to, you know, of fight each other or something, which unfortunately is what happening now. Or blowing up self-driving cars. Speaking of being called names, can I ask a meta question? Chris, as you know, I've also kind of written a book in the tech vibe, and it's also been received in many favorite swathes. And it's always interesting to do media around the book.

1:17:32I also did a book tour when Kazmoky came out. And I'm just curious, as a meta question, like, what's been like, without calling out names, probably, like, what's been like the weirdest question or the weirdest experience? because you do an excellent job of explaining to anybody basically what crypto is about. All the same, what's the weirdest moment, if we can talk about this? Maybe this is when your publicist types up. I don't know. In fact, there's been a lot of, as you'd expect, there's been some negative, some hostile. Oh, really? Well, I mean, I don't know, hostile, but certainly skeptical.

1:18:09I will say what's I mean, this is maybe I'm dodging your question, but I will say something interesting that, so Steven Johnson, you know, the author, if you know him, he's a really great author. And I had a chance to talk to him recently. He said, the weirdest thing about writing a book is that you'll meet people who read your book and they spent eight hours inside your brain. And suddenly it's like a different conversation. So I will say on the media tour, it has been very striking who I can tell in the conversation who read the book and who didn't immediately. and it's just a radical and it and i say this in a good way i mean again maybe i'm dodging your question but like it is such a like i had an interview with um uh the economist i did a podcast and it was i had just done a few other sort of mainstream media podcasts and no one had read the book and this this uh reporter this woman for the economist had read the book and her questions were like it was just like fuck you know the clouds parting and like the angels singing.

1:19:05It was like the most wonderful conversation ever. Cause we were able to actually talk about, it was like having this conversation or something that we were able to actually talk about it. And you know, she was using the proper terminology and just like the whole, I don't know. And like, we were actually having a real conversation, um, which reminded me of the Steven Johnson thing. Like, you know, she just spent eight hours going through the whole thing. Um, it's just a topic where like, this is why I wanted to write a book in the end. Is it like, I would just have, I've had, I mean, God, I've had 5 ,000 hour long conversations probably in the last eight years trying to explain this topic.

1:19:37Right. And so many of those conversations, people just look at me like I'm on Mars. Right. Um, because it's like the argument, like ultimately in the book, the argument is blockchains are a new way to create networks that have the societal benefits of protocol networks and the competitive advantages or the advanced functionality of corporate networks. But to get that, you have to know like that, that all the background on the internet, that the internet is a network of networks, that the thing you do on the internet, if you're an investor or an entrepreneur, like I, I used to BRM is you build networks.

1:20:05Like that's what any, all the internet veterans know that, right? They know that this is sort of how the internet works. Um, but most people don't have that prerequisite knowledge. And so it's just like, you end up with these conversations where it's just like impossible to kind of get through and explain these things because it just can't be compressed into an hour. Um, and so that has been just a really interesting experience is that like, you know, we'll see it's still early in evolution, you know, in the, in the book and sort of the propagation of the book. But this idea that, wow, maybe you can actually, like if you explain it clearly enough and people are willing to sit down for six hours, like you can actually like have a, you know, have an under come to an understanding was really kind of eyeopening and refreshing.

1:20:50Yeah, I don't know what weird, I don't know. I'm struggling with their actual question. I just don't, nothing comes to mind. That's a great place to wrap. Want to be mindful of time. The book is read right on. We highly recommend it. We'll buy some copies for our listeners as well. Chris, thanks so much for joining us today. All right. Thanks, guys. I really appreciate it. Thanks, Chris. Turpentine VC is a podcast from Turpentine, the network behind Moment of Zen and Econ 102. If you liked the episode, please leave a review in the Apple Store or rate us on Spotify.

1:21:26Thank you.