Read Write Own: A New Era

Steph Smith: People always ask, like, what problems do blockchain solve?

Chris Dixon: I think I've been pretty good at predicting what happens on the internet. I've been pretty bad at predicting exactly when it happens.

Steph Smith: The prize when you're building an internet service is always the network.

Chris Dixon: I felt like the worldview that I have that sees a very positive vision for how blockchains and tokens and these other technologies can be used was not widely understood and not properly represented.

Steph Smith: What ends up mattering is the products.

Chris Dixon: It's a worldview of the internet.

Steph Smith: Hello, everyone. Welcome back to the A16Z podcast. This is your host, Steph Smith. And today we have a very special episode. We will be hearing from Chris Dixon as he just released his new book, *Read, Write, Own*. Chris has been a participant, a creator, a funder, a founder, all within the internet era. And in his book, he breaks down the cycles of crypto specifically. And what's the why now? Plus, he addresses the important distinction between the computer and the casino. All right, I won't hold you up any longer. Hope you enjoy this very important episode of Chris Dixon and Robert Hackett. And don't forget to check out Chris's book, *Read, Write, Own*, which is available now.

Sonal: Welcome to Web3 with A16Z, our show about building the next generation of the internet from the team at A16Z Crypto. I'm Sonal, your host, and this show is for anyone: whether artist, developer, company leader, entrepreneur, media, or policymaker. Today, we're excited to launch a limited series on the new book just out, *Read, Write, Own: Building the Next Internet* by Chris Dixon. It explores the power of blockchains to reshape the internet back into an open network for fostering creativity and entrepreneurship, which affects us all. The book also goes beyond crypto and blockchains to cover intersecting technologies such as AI, social networks, marketplaces, virtual worlds, and more. This new series is produced and hosted by Robert Hackett, beginning with this episode with author Chris Dixon and continuing with other guests on other topics. You can find the book at readwriteown.com. And as a reminder, none of the following should be taken as investment, legal, business, or tax advice. Please see a16z.com slash disclosures for more important information, including a link to a list of our investments.

Chris Dixon: I got into the internet for its open democratic kind of anyone could go and build a website. I think that's a serious risk at this point. I think blockchains are the only kind of credible way to turn that. I go through that in detail in the book. That's what the book's about. That is what kind of motivates me at this point.

Robert Hackett: Welcome to the Web3 with A16Z Crypto podcast. I'm Robert Hackett, an editor here at A16Z Crypto. And I'm here with Chris Dixon, founding partner of A16Z Crypto and author of the new book, *Read, Write, Own: Building the Next Era of the Internet*. I had the privilege of editing you, Chris, throughout the book writing process. And I'm thrilled now to have the opportunity to talk to you about what went on behind the scenes, the big themes of the book, the challenges you experienced, and also to branch out and talk about the industry at large, as well as what we can expect from the crypto industry in the future.

Chris Dixon: I see this book as a culmination of my entire career. I started my first internet company in 2004, so 20 years ago now. Before that, I was an engineer and working on the internet, so in some way involved, not as deeply involved. I really spent time trying to be a student of the internet and understanding how it really works, not just how the technology works, but how it works from the perspective of economics and governance and power, how you gain power on the internet and how that power is used. Specifically, what that means on the internet is understanding networks.

Robert Hackett: I remember when you came up with the timeline originally. You had a timeline, you had an outline of when you wanted to write this book. I didn't tell you this at the time, but I was super skeptical. I did not think that you were going to hit the deadlines that you were setting for yourself. It was very aggressive. But you cleared your calendar and you pumped out a first draft in like a matter of a couple months, which I was stunned by.

Chris Dixon: In retrospect, the calendar was naive. It only worked because one, I had a lot of help from people like you and others on the team. And two, I went back and spent a lot more time than I expected. I got on this schedule of working, getting up really early for long periods of time and working for four-hour chunks every morning, and a whole bunch of other things to get it done. This book is the best possible condensation of what I've learned over 20 years. You take all that 20 years, all those meetings, the thousands of meetings, and all the other stuff. I actually cut out like 200 pages of material and could write another book if people are interested. But the key parts are distilled and organized in a way that's meant to maximize the ability to transfer that learning to somebody else. When you realize that, you think, how could I be spending my time as someone interested in the world in any better way than benefiting from that knowledge from other people? That motivated me. I still think books are maybe the greatest technology. The idea that you can stare at squiggles on a page and have essentially a hallucinatory experience, that you can commune with people that have been dead for a thousand years in a very intimate way and get into their head.

Robert Hackett: Coming from a tech investor.

Chris Dixon: It's interesting as a side note that books have been almost completely unaffected by the internet, the book industry. For better or for worse.

Robert Hackett: I would say for the most part, that's good. It hasn't been hurt the way a lot of other industries have and hasn't had to do some kind of crazy machinations. For the most part, it's a pretty healthy business.

Chris Dixon: On the flip side, when you actually look at the book sales figures, which I hadn't really looked at until I wrote a book, it's shockingly small. If you take out romance novels and things like this, the best-selling kind of nonfiction ideas book, which I'd put this in this category, I believe last year had sold like 400,000 copies, which is like a niche website or something in terms of user base.

Robert Hackett: You've already answered one of my big questions. I wanted to back up and ask, why write a book at all? You're a tech guy.

Chris Dixon: It's a good question. I got this question from a lot of my tech friends. Why write a book and why go to a traditional publisher? Why not just write blog posts? I used to blog. I blogged for years. I wrote many hundreds of blog posts, and I'm a big fan of that format. But I do think books are special in a couple of ways. Even though the best-selling books in this style maybe sell a few hundred thousand copies, I think they have an ability for the ideas to penetrate culture at large in a very powerful way that is much harder for blogs and podcasts to do. One way to think of a book is it's really a physical delivery mechanism for a set of memes that, if successful, will penetrate the culture. Concepts like Black Swan or Tipping Point are now mainstream concepts. The number of people that know those concepts and think about them is far greater than the number of people that actually read those books end to end. With a book, maybe 1% read it all the way through, 5% to 10% maybe read the intro and skim around, and then 90% or 95% who understand the ideas didn't actually read the book. Books have this ability to penetrate more broadly. That's the breadth. And then there's depth. People meet who say a book really changed the way they think about something, changed their career, changed their life. It's not that something written on the internet couldn't do that. I have heard people say certain blog posts were really important to them. But it goes to this quality of the gravitas of a book, the length of a book, the fact that it's the right length to fully present a worldview. A blog post can present an idea. A book can present a worldview. This book is a worldview. I'm sure a lot of people will disagree with it, but it's a worldview formed over 20 years. It's a self-consistent worldview, a fully worked out system that's consistent with the facts in the world. I think it's a theory that a lot of people will find novel because it's a different way to look at things than if you just sort of are a casual internet user and haven't seen how the sausage is made behind the scenes. There's a meme out there in the tech world that why write a book when you can write a blog post? Why write a blog post when you can write a tweet? I generally agree with that. But in this case, I really felt strongly that you cannot transmit this worldview on podcasts and blog posts. Every time I tried, I managed to transmit this piece of information, but then there were all these questions about these other things. I finally realized the problem was that you couldn't take a piece of it without the whole thing. You just needed the whole worldview.

Robert Hackett: It's interesting to hear you describe it as a worldview. People might think about a book about crypto or blockchains as a tech book, as a business book. But really, this book is kind of a philosophy book.

Chris Dixon: When I say worldview, it's not a worldview of how to eat and live your life. It's a worldview of the internet. It's a complete theory that's a different perspective than most people will have of the internet, which I argue in the book has become in some ways this global mind. If you think of it as a mind-body dualism, this is the digital mind, this global hive mind that we all plug into and that influences the way we think about things, our politics, and our economics. It's a distinct worldview of the internet. I don't want to overstate it. It's not a full life philosophy. But I do think it could only have been presented in a book form. I finally came to the realization that I wanted to present that worldview. There are specific things happening in the world. For example, there's a big policy debate about whether tokens should be illegal. I obviously have strong views on this. I felt like the worldview that I have that sees a very positive vision for how blockchains and tokens and these other technologies can be used was not widely understood and not properly represented in that discussion. I wanted to present that through a channel that would be accessible and potentially reach a broad mainstream audience. This is very much a book written for smart high school students. There's no knowledge presupposed outside of basic common knowledge and a willingness to be curious and interested. It goes from first principles. I studiously avoid technical words. I have a few, but I define them. It tries to present it in a narrative form that goes through the history of the internet and how these things developed and give a lot of examples. There are 36 pages of end notes.

Robert Hackett: Chris, one of your superpowers is this ability to take a deep concept, some hard idea, and distill it into a catchy, sticky phrase. You've got so many of these that you've blogged about forever: "come for the tool, stay for the network," "the next big thing starts out looking like a toy," "can't be evil." What is your process for coming up with these taglines? How do you know when you have a winner?

Chris Dixon: I think it's less the tagline and more that what I'm good at is taking lots of information and trying to distill it into simple ideas. It takes me a lot of time. I would often get stuck in the book and spend literally like three hours exercising other things, like how can I simplify, simplify, simplify? What are the core basic principles here? The catchy phrases just end up being like "come for the tool, stay for the network." I'll explain that idea. The prize when you're building an internet service is always the network. The internet is a network of networks. There's the base layer physical network from the internet, and then what we do on top is we build networks. The original networks were the worldwide web and email. Later on, there were Facebook, Twitter, YouTube. These are all networks. Networks are what create value, business value, profits on the internet because networks are very defensible. They have moats that Warren Buffett would talk about. Once you build a large audience like on YouTube, it's very hard for another video site no matter how much technology they have. They could have the most beautiful thing in the world and the fastest and everything else, but it doesn't matter because the network, the audience, is somewhere else. "Come for the tool, stay for the network" is something I started to describe many years ago. Every entrepreneur, including me, when you're an internet entrepreneur, you try to create a network. It's very hard to create a network because the network effects cut both ways. A network effect is this idea that as a network grows, it becomes more valuable. Let's take a dating network like Tinder. When you get to the point where there are tons of people on there who want to date each other, you have a very defensible business. It's very hard for somebody to compete. But when you start off and there are like three people on it, no one wants to come there. There's this age-old dilemma among internet entrepreneurs of how do you get over that bootstrap phase? My partner Andrew Chen wrote a book about it, *The Cold Start Problem*. It's the chicken and the egg. How do you get over that? There are also what I call tools. Networks are multiplayer, meaning they only really matter when there are many people using them. Tools are generally single player. A word processor or a spreadsheet, you can share it with somebody, but you can get a lot of value out of it just using it by yourself. The nice thing about tools is that you don't have that network effect problem as an entrepreneur. The case study in the book is YouTube. YouTube did the strategy "come for the tool, stay for the network." They started in 2005. Originally, they were a dating site. They pivoted to being a video site. Their pitch was that they didn't have a network. They didn't have an audience. No one went to the YouTube site. At the time, everyone had their own website. I have my website, cdixon.org, and they would have their own audience, and they would want to put video there. YouTube said, "Hey, you can come to our site, upload the video. It's all free. We'll subsidize it. And then you can take this little piece of HTML and embed it in your site." They acted as a tool. You didn't need a big audience on YouTube to have that be valuable. People started doing that and being very popular. They had a very nice tool. It was very seamless, and they paid for the bandwidth. Subsidies are a huge part of how the current networks were built. I talk about it in the book. That's why they had to all raise billions of dollars as they were subsidizing all the early users. They built that up, and then at some point they had millions of websites using it. They said, "Hey, by the way, whenever you put it on your site, we put it on our site too, because you may get some more viewers." People were like, "Great." And then at some point they had this website with tons of great videos, and people were just like, "Why am I going to Chris's site when I can just go to YouTube?" That's how YouTube started.

Robert Hackett: Maybe I'm interpreting this too liberally, but I actually see this book as a tool, like a "come for the tool, stay for the network," where this book is the tool and the network is this computing movement that's going on, recruiting new people to that idea, to that software.

Chris Dixon: I think that's one way to look at it. The book is a standalone thing that you can read and understand. If somebody finds it interesting, then the next thing afterwards would be to join. There are all sorts of communities that are interested in these topics, and that would be the next way to get involved.

Robert Hackett: In startups and in life, timing is everything. Tell me about the timing of this book. Why this book? Why right now?

Chris Dixon: Good question. I've been in this space for basically 10 years. I led our investment in Coinbase in 2013. I wasn't exclusively doing crypto. I was doing other things. I had some AI investments, some drones, some other random things. I was doing kind of frontier investing, just new emerging areas. But then, I guess five years ago or so, I switched to full-time on blockchains and crypto. Through that time, there have been all sorts of ups and downs, many cycles, positive and negative developments. We had probably the roughest period around FTX in 2022, the most recent downturn. It was frustrating because there was a lot of really interesting stuff getting built. I talk about in the book this distinction between the computer and the casino, the two cultures that have coalesced around blockchains. The computer is what I see myself as part of and what the book is about: people trying to use these technologies for productive new use cases. The casino is people who co-opt aspects of the technology and use it in ways that are really focused on gambling and speculation. FTX was one example of that. They were even worse than that. They committed fraud. But there were a lot of others who just promoted speculation, and it still happens today. I see it as a struggle between those two movements because the casino stuff really hurts consumers. People who had money at FTX lost money. One of the really frustrating things that's happened on the policy side is I would argue that a lot of the actions of policymakers have actually encouraged the casino and discouraged the computer. It really helped the bad actors and hurt the good actors.

Robert Hackett: How? Say more about that.

Chris Dixon: A simple example I talk about in the book is something called meme coins. Meme coins like Dogecoin are tokens that are created that have explicitly no use case. They are the emblem. I'm not anti-Dogecoin. Technology develops in unexpected ways, and maybe Dogecoin will develop in ways that are quite interesting. But this idea of a token which does nothing, has no purpose except to be traded and bet on, I'm not a fan of that concept. One of the great ironies of the current policymaker approach is that meme coins are perfectly legal. But if you take a meme coin and you decide to add utility to it by working on it, you then are at risk of tripping up securities laws. As long as it's completely worthless and does nothing, it's fine. There's no ban on speculation. Creating tokens is not illegal. Creating tokens and then working on them and trying to make them better is what is in disputed territory.

Robert Hackett: It does seem a bit backward.

Chris Dixon: There's a reason behind it. Securities laws are built to avoid asymmetric information. With Dogecoin, there's no asymmetric information. Everyone knows it's just a totally meaningless, useless coin. You're just betting on a number going up or down. There's no asymmetric information. Everyone's on a level playing field. It may be a dumb playing field, but it's a level dumb playing field. Once you start having a team behind something building something, that team might have asymmetric information. The guidance from the SEC in 2019, the last time they gave guidance, said you need to be sufficiently decentralized. It's okay to have people working on making something better. For example, Ethereum has a bunch of teams making it better, but there's not one team doing it. There's a whole bunch of teams. The information is sufficiently decentralized such that there is still a level playing field. There are ways to build things, to have tokens where you build blockchains in a way that is decentralized such that it's not a security. It's just a very complex, nuanced thing. It's very simple to go and create something that is unambiguously clear from that. If you go to a bunch of these other websites who aren't really crypto companies, like Robinhood, they removed a lot of very useful blockchain tokens, but they support Dogecoin. We'll probably see a Dogecoin ETF at some point because it's legal.

Robert Hackett: Some big ideas books have this problem where they kind of just come across as scattered blog posts. What's nice about this is it actually has a through line.

Chris Dixon: It's interesting because I feel like the best part in a lot of ways is the end of the book, which I struggled with. The end of the book is really the payoff because of all the work you do before. There are seven sections that are seven really interesting application areas of how you apply blockchains to finance, to artificial intelligence, to games and the metaverse, to media and storytelling, and a bunch of other areas. As you're thinking about how you write this, don't you want to have the juiciest parts at the beginning? But the reality is you really need the background to have the payoff. You have to do the work and understand what are the alternative ways that these things could be built. That goes through a little bit of the history and how we got there. Then I do a deep dive. People always ask, what problems do blockchains solve?

Chris Dixon: So I kind of do a deep dive on why blockchains are a better way to build networks than these other methods. Then I very specifically go through the benefits: the economic benefits, the governance benefits, the software development benefits. People use words like decentralization. I don't use that word that much, but it's about decentralization.

By the way, some of this is affected by talking to policymakers. For example, I remember one in particular I was talking to last year who said, "I believe much of the other stuff you're saying, Chris, but I don't think you can really build software without a CEO." And I was like, well, then I started talking about how 95 to 99% of the software in the world is open source software, which has no CEO. Linux is by far the most successful operating system. That was a really good conversation, and I think the person I was speaking to found that eye-opening. It's surprising that most of the world doesn't know that's how most software is built, without a CEO, that collaborative bottoms-up way to build software.

I feel like the open source software movement is such a perfect precedent for what's happening in crypto right now. Crypto is very much, in a lot of ways, the cultural and intellectual heir to the open source software movement. Open source software is about keeping the software layer open. Blockchains are about keeping the service layer open. I go through this in the book.

What basically happened in the last 20 years is that much of the software business became commoditized through open source software. It no longer became a good business to sell web server software or operating systems because you could get great ones for free. So a lot of these companies, including Microsoft, now fashion themselves as services companies. Microsoft invented the software business. Services are software that's live and running, storing information, instantiated and active.

One way to view blockchains is that they take a lot of the ideas of open source software and elevate it to the layer of services. A blockchain is an open source service. Something like Ethereum is this computer in the sky that anyone can access. Nobody owns it. It's running all the time. It has its own internal economic model that keeps it running. To run software, you have to host it and pay for bandwidth, so you need an economic model to have self-sustaining open services. Users and developers pay to access it, and that money flows back into the hosting costs. It's very much like the mainframe computing model, time sharing, pay as you go back in the 60s. This open source, bottoms-up, collaborative model got together with mainframes and created this new kind of open community-owned mainframe.

That lets you create services on top which inherit those properties. You can create a game on top of Ethereum which inherits all of those properties I just described—being open and community-owned. That has all these downstream consequences which I go through in the book.

Robert Hackett: What do you read?

Chris Dixon: I read everything. Fiction, nonfiction. I try to read almost anything that has a four or above on Goodreads. A lot of it is recommendations from friends. I always have a stack of books that I'm going to read. Some of it is edification, being a person who's well-read, so I try to be broad. Sometimes I just go deep in certain topics.

My main theory there is in contrast to social media and the internet. Sadly, I don't think we're in a great spot on the internet today for a whole bunch of reasons. I think there's a trap you can get into where you're just checking Twitter and other things. I try very hard not to do that. I think of it similar to people who are into dieting and different food programs. To me, books are protein and social media is generally sugar. I do it sometimes, it's a work thing. It is important to stay in touch with some things going on. But I think it's very important to have limits around that. I think there's a thing it does to your brain where you get into these cycles and you need to avoid that.

I found the 50-page rule on books is a great rule. It's pretty doable if you plan your day right. Obviously everyone's schedules and demands are different, so I don't want to presume everyone can do that. But for me, I make it a priority. Some days it feels like a pain, but it's only 50 pages. Then what often happens is I start and it's a good book and I'm like, okay, I'm going to read more.

I also have a rule: 50 pages by coincidence. I have to read at least 50 pages of a book before I quit it. That's a good rule. I do quit books, and I think it's important to quit books because one of the blockers you get into with reading is you get stuck on a book you don't love and don't get through it.

Robert Hackett: You also have a habit of reading very out of sync, out of the way the book is telling you to read it.

Chris Dixon: Yeah, it varies. Especially with nonfiction, not with fiction. With fiction, you have to read it in order. Sometimes with nonfiction, yeah. If your North Star with reading is just how much time do you spend reading books, that's what you're trying to optimize, as opposed to social media or TV. If that's what you're trying to optimize, you want to avoid the blockers. The blockers are things where you get bored and you stop reading. One way to do that is to always make sure what you're doing is interesting. That means it's okay to quit a book after 50 pages. And it's okay to jump around with a nonfiction book. First thing you look at the table of contents and jump in. People should do it with my book. I hope they do. You jump to a few sections that look most interesting. Hopefully those are well written and you like them. Then you're sort of, "Oh, wow, maybe I should understand the stuff before. Maybe that's well written and interesting too." In the best case, that leads you down the path of reading the whole book.

Robert Hackett: You mentioned Andrew Chen before in his book, The Cold Start Problem. I know he's talked about how he will lock his phone in a crate, in a Faraday cage, so that he doesn't have access to it so he can concentrate. Do you do anything like that to carve out that 50-page period?

Chris Dixon: I definitely... I don't do the thing where you lock your phone in a thing, because I would probably just figure out a way to get it out. I do definitely put down the phone. I really think it's this trainable brain response thing, this attention span sort of thing that people have now. For me too, I have that same feeling. You put the phone down and you have this urge to look at it, but it goes away. If you push it, it goes away. It's just like exercise. You first sit down to exercise and the first 10 minutes are probably the hardest because you're lazy and don't feel like doing it. But once you're into it, it often takes a life of its own. I think it's very similar to these other kinds of habits. In many ways, the initial action is the most important thing, and then it becomes momentum-driven after that.

I think of your brain as an LLM. It's a machine learning algorithm, which we probably are. The most important thing you can do to feed that and improve it is the inputs. You should be very deliberate about the inputs. It's very hard to have bad inputs and then try all these things—I'm going to meditate, I'm going to do this, I'm going to unwind. But the problem is you put all these bad inputs in. You can only do so much to the machine if you've got the wrong input. To me, the good inputs are books and conversations with people. Those are the two things where I learn the most, where I get the most direct primary information. I'm very big on primary sources, not secondary sources. I'm very skeptical of many secondary sources. I always like to go as much as I can to find the source, get a lot of inputs from primary sources, and make my own determination.

I think it's true for everyone in any kind of information intellectual profession, but particularly in investing, which is what I do. It's very important to develop a model of the world that is both true and hopefully differentiated. That's what you're doing in investing. You're taking that model and applying it to financial investments, trying to find things that are undervalued or avoid things that are overvalued because the world model that most people have is off and yours is better. If you think about what you want to do, you need to be very cognizant of how you're training that world model. The easiest thing to control are the inputs.

Robert Hackett: If we're talking about high-quality inputs, do you have a favorite book or a favorite author?

Chris Dixon: I kind of grew up reading. I spent years and years, both as a kid and then college, even after college, reading computer books, programming, and then got into artificial intelligence, what was called artificial intelligence back then. It was very different. I got into a bunch of philosophy books via Daniel Dennett and Douglas Hofstadter and a bunch of these bridge writers. Douglas Hofstadter, I would cite as the biggest intellectual influence on my own upgrade. Gödel, Escher, Bach is just an incredible book. It opened up a door to another world, which was proper philosophy and philosophy of mind and language. I spent years reading that.

I've read all the popular science books. I really love the Steven Johnson books, the early Malcolm Gladwell books. I've read all the popular physics books. Eventually, after my popular science phase when I was young, it became history. The thing I always go back to is history. I read a lot of history books, all eras. My favorite era of history is probably the industrial Victorian era. In a lot of ways, you could argue that modern civilization was built between 1870 and 1940. If you go to New York and look at all the skyscrapers, most of them were built in that era, and the bridges and everything else. Also electricity, airplanes, cars, the whole modern world. You had this era of 100 sort of Elon Musks, these bold entrepreneurs who would build big physical world projects. That's always an area I love.

I spent a lot of time reading computer history, tech history. I think an underappreciated area is the origins of computing. I read an article that was published in the Atlantic Magazine a couple of years ago on that. My thesis was that a lot of computing came from logic and philosophy intellectually. That era when you had people like Turing, Claude Shannon, John von Neumann, Kurt Gödel, and all these other thinkers. It was a very small group of people. What was really interesting to me was that if you'd surveyed the world in 1905 and tried to find the most useless, unpragmatic area in the world, it would be formal logic. But then it turned out later to be a lot of the things that led to the design of computers.

Claude Shannon has this amazing paper where he was the first to have the insight that you could do a mapping of all of formal logic and Boolean logic onto circuits. Once you could map AND, OR, and all these other things—which are what the most base level transistor circuits are—you could then do math and everything else. He knew that because people like Bertrand Russell and Whitehead famously in the Principia Mathematica had written out a thousand pages of how to construct all of mathematics based on simple Boolean operations. They had built all this machinery for a very esoteric philosophical purpose. But Claude Shannon's insight was that you could take all that machinery and just move it over onto electronic circuitry and have computers.

Anyway, I'm going down a rabbit hole. The point is that I think there's a lot of really interesting history throughout that whole period. I really like primary sources because I find that, as an example, people often talk about George Boole—called Boolean because George Boole wrote this book called The Laws of Thought. He's almost always talked about as a mathematician. If you go read the book, it's not a mathematics book. It's a philosophy book. I read the book. The whole thing is about Aristotle and how he's extending Aristotle because Aristotle was the prior major thing in philosophy. I find this again and again. I like the primary sources because I think it's a very different view of history.

If the goal is both as a human and as a professional to build your world model, you should be constantly thinking about what those inputs are. And by the way, why history? One, I'm just interested in history. I think it's important. But two, when you're working in areas that are extremely complex systems—the internet is 5 billion people interacting in a very complex way, with money flowing and culture and politics and power—you can try to come up with little principles to predict what will happen. But I've always found that the best guide is history. History is a series of examples of people interacting in very complex adaptive systems and how that played out. They give you these very interesting rules of thumb. I talk about history in the book, but I'm not actually interested in history per se. I'm interested in history in as much as it helps us. I'm interested in the future.

Your friend, Sep Kamvar, turned me on to this quote from Neil Postman. I'm going to butcher the exact wording, but he basically said history is the best way to learn any subject, any discipline. Something like you shouldn't teach history, but everything should be taught as history.

Robert Hackett: Something like that. It's a great quote, and I agree with you. I think it's a shame when it's not taught that way. These things generally have fantastic and very interesting stories, and they're very human stories. I always find the humanity behind these stories interesting, and all the struggles.

I want to ask about the human element, but first I also want to pose something counter to this idea of the halcyon days of all these technologists and pioneers coming up with incredible new technology. We're living through a period like that right now, and I feel like a lot of people look at the past with rose-colored glasses and don't see that there are weirdos coming up with awesome stuff today.

Chris Dixon: Oh, I think this is... I say weirdos with love and respect. I think this is maybe the most exciting period of technology. In some ways, you could argue that the next 20 years is the culmination of all of the developments in computers and computer science over the last 80 some years. Between AI, blockchains—which play a very important role—self-driving cars, virtual reality, all of this. You have all these new trends, and you have existing past trends that continue to play out. The internet is still early. It's 30 years into what will probably be one of the most important technologies in centuries. The mobile phone is still playing out. People are still figuring out how to use smartphones and all the different applications, SaaS, FinTech, databases. You have all of these past movements that still have a long way to go, and then you have these really powerful new things layering on top.

I talk about a distinction in the book I call inside-out versus outside-in technology. Inside-out technologies come from the cathedral. They come from Stanford or Microsoft or Apple. It's the iPhone. It's Windows. I think AI is very much an inside-out technology. It's come from Google and universities and places like OpenAI, which were spin-outs of incumbent technology companies and universities. Then there's outside-in technologies. The historical classic examples are things like Linux and open source. The World Wide Web is an outside-in technology. It's this hacked together thing of open source and collaborative movement. I think blockchains are outside-in.

We're entering a period where there are all of these very important things happening. There's the inside cathedral stuff that gets a lot of attention. What gets less attention and is less understood is the outside-in stuff and all of the weird fringe things, which is where I've spent most of my career and I think is the most interesting. I'm not saying the other stuff isn't valuable. It's just not what I personally want to work on. So yes, I think there are Wozniaks and Linus Torvalds and Tim Berners-Lees all over right now. It'll take a while before we can tell the whole story and understand what it all means. The scope and scale of the internet and technology has increased a hundredfold, probably the number of people working on it and people using it.

Robert Hackett: We're talking about these geniuses who really wrenched the world into their orbit and changed the way things work. With a book like this, it's a big ideas book, and there is a tendency for big ideas books to get too abstract. You mentioned the human element. How do you keep the human element in this book? People care about people fundamentally. How do you retain that to capture that human interest and make it engaging?

Chris Dixon: I tried, whenever... A lot of the structure of the book is, I try not to be overly rigid about this, but I do try very hard to alternate between narrative, telling stories—many of which I know from having been there or been involved in it at the time—and then providing more abstract lessons from those stories. I tried very hard to balance that and to make that as relatable as possible.

But look, it is an ideas book. It's fundamental. It's not narrative fiction or an adventure story. There are a lot of ideas in there, but there's also a lot of history, a lot of stories, and a lot of specifics.

Robert Hackett: You mentioned a few of the startups that you founded and your experiences with them, and how that has tracked the progression of the internet over time.

Chris Dixon: Yeah, I was the founder of two companies. The first one was 2004, called SiteAdvisor, a security company. The really important thing to me was that we were built on the open web. We went and rated websites and tried to provide security warnings for users if they were dealing with somebody doing phishing or spyware. But very importantly, we didn't have to ask for permission to do that. The web was this open community-owned network, and we could just build new pieces on top of it. That allowed us to exist. It allowed us to be part of the solution to the security problems on the internet. We ended up being acquired by McAfee and bundled into their software. Our software became a very popular, big product for them. That was very much in the Web 2 era, but it was piggybacking off of these open systems like the World Wide Web.

With my second startup, it was an AI company, a machine learning company called Hunch that we started in 2008 and was acquired by eBay in 2011. I think we were right about the long-term trend about AI. I think we were too early. But our bigger issue was that we assumed the internet would stay open the way it had for my prior startup. We used all of this data that we'd crawled and analyzed, a little bit like OpenAI has done now from Twitter and other places. This was 15 years before OpenAI. As they say in tech, early is the same as being wrong. We were dependent on a lot of other data sources, and that was right around when a lot of the data sources started pulling back. In the end, we sold the company to eBay, partly because eBay had their own data and didn't need to use outside data.

Robert Hackett: You talk a lot in your early blogging about this platform risk, basically, before that became a real issue.

Chris Dixon: Yeah, I've been blogging. I cite myself in there once, but I had a lot more examples on my blog where I was obsessed with these issues of the internet becoming consolidated. I started blogging about this in 2009. It's been very consistent and very prescient. I didn't think it would be appropriate for the reader, but there were a lot of things in the book that I'd been writing literally since 2009. Some people approach this topic and they see, "Oh, here's another tech investor. The same thing that happened with Web 2 is going to happen with Web 3."

I was thinking about this the other day. Why did I even work on the internet? Why is that my profession?

Chris Dixon: I was very attracted to the ideals of the internet, the 90s ideals of the internet. That it would be this open, democratically owned, permissionless system. You would have open APIs and collaborative development. Anyone in the world could put up a website, build something, get distribution, get users, build a business, support themselves. To me, that was what was special about the internet. What made it different from broadcast TV, radio, newspapers, or other forms of media. It was open, it was democratic.

Robert Hackett: And you started worrying about that when you started blogging around 2009.

Chris Dixon: I think it got really bad. I go through this history in the book a little bit. Around 2012 or 2013, it happened to come along right when what I eventually now believe is the potential antidote to it appeared: new systems built on blockchains. That’s what got me interested in it. I think a lot of people think it's about finance. I have very little interest in finance personally. I do have a section on it in the book, I think it's an important area for blockchains, but it's one of many. Honestly, I had to force myself to write it because I'm just not interested in finance.

Robert Hackett: But yet you're an investor, a venture capitalist.

Chris Dixon: Yeah, but you don't really do spreadsheets in venture capital. It's finance in some sense, but not trading and interest rates and all the stuff on CNBC. That's just not it. So that's how I came to it from that perspective.

Robert Hackett: The attract-extract cycle that you lay out in the book, I think understanding that one concept is like seeing the code in the Matrix. You reference it in the early section called "Corporate Networks," talking about the logic of how centralized network services like Twitter and Facebook evolve. They start off attracting, saying "come join us, we'll do all these wonderful things." Then over time, the incentive switch happens.

Chris Dixon: This is not some cultural thing. If you analyze how network effects work, there is an incentive switch at some point where it's more important for them to extract money from the network than to attract complements to the network. It happens over and over. It's a repeated cycle. People keep falling for it. It's going to happen now again with AI. A new generation will fall for this again. They'll build on top of these other platforms, then suddenly things will switch and they'll learn the lesson the hard way. I'm hoping they'll learn the easy way by reading the book. But by the way, I think it's about to get much worse. AI will make it a centralizing force. It rewards companies with lots of data, money, and compute. That will further favor the incumbents.

Robert Hackett: There's another group that has a similar concern as you about decentralization, but they are not on board with crypto or blockchains. The Mastodon folks, the Blue Sky Protocol. Vitalik, the creator of Ethereum, had a post earlier this year called "Make Ethereum Cypherpunk Again." He said there's a large ideological rift where significant parts of the non-blockchain decentralization community see the crypto world as a distraction, not as a kindred spirit and powerful ally. How do you bridge those two forces that should be allied?

Chris Dixon: It's unfortunate because I think they do share values. The blockchain people like me and Vitalik, and the folks who are pro-RSS and things like ActivityPub... Look, I love those things in theory. I've been on the internet for 25 years and I've supported these protocols. The reality is there are two protocol networks that have succeeded in the history of the internet: SMTP and HTTP. Email and the web. Both started in the 1980s and grew before there were centralized corporate networks. The closest one since then to succeeding is RSS, and RSS failed in the end. It still exists, but it's a tiny fraction of the user base of Facebook and Twitter.

We've now had 25 or 30 years of protocol networks being developed. I go through in the book, there are literally hundreds of them, and none have succeeded. After 30 years of failure, you should consider maybe there's something wrong with the approach. Protocol networks are great from a societal benefit point of view. What they're not great at is features and funding. They ask you to do things like go get your own domain name. That's not a mainstream consumer behavior. Ask someone on the street in New York, they have no idea why they would ever buy one. They don't want to pay $8 a year to use an internet service. People are so accustomed to things being free.

And funding. Google has, if you count contractors, 600,000 employees. Facebook has hundreds of thousands. These are vast companies that spend massive amounts of money subsidizing these networks. All of that free stuff you do on YouTube, do you know how much money it costs to host that video? A protocol network can never do that. If you have a new video protocol to compete with YouTube, you'd need probably $10 billion just to match the subsidies. You'd need $100 million a year minimum to support a top-flight development team.

To me, a blockchain is a super-evolved form of these old protocol networks. Blockchain networks have the societal benefits of protocol networks like HTTP and SMTP, but the competitive advantages of corporate networks like Facebook and Twitter. Blockchains have an inherent funding mechanism and software features that let them compete with the modern user experiences of centralized systems. They are further evolved, more likely to lead to real competitors, but share many of the same values as other movements.

Robert Hackett: Fred Wilson, who read and reviewed your book, had a post recently where he said he believes 2024 is going to have a ChatGPT moment for crypto. A breakthrough moment where suddenly everybody realizes, "Wow, it works. It works really well." Do you think that's going to happen in 2024?

Chris Dixon: I don't know if or when it'll happen that soon. AI is a very interesting case study. The first neural network paper was McCulloch and Pitts in 1943. Literally 80 years before ChatGPT. A long history, tons of money went into it. I started a company in 2008. A lot of people would argue it didn't really work until 2023.

Robert Hackett: Did you think you were late in 2008 when you started it, or did you think you were early?

Chris Dixon: I thought it was the right timing, but I was obviously early. The thing people didn't know was that neural networks were not the most effective method in 2008. People thought they were cool in theory because the brain worked that way, but empirically they just didn't have the best results. What was interesting about neural networks is they happened to take advantage of the same hardware that video games did: GPUs. Companies like NVIDIA investing in GPUs for video games ended up having a spillover benefit to neural networks. As neural networks got to some level of scale, they improved very dramatically.

Robert Hackett: This is the perfect encapsulation of your idea that the next big thing starts out looking like a toy.

Chris Dixon: It took 80 years. When we were at my startup, we called it machine learning. AI was a bad word. That's what happens in the pre-ChatGPT moments of technology. You have all these meaningless debates over terminology. You see that in crypto today. People say, "Oh, you shouldn't call it Web3, you should call it this." In the book I just use descriptive words like "blockchains" and try to sidestep the tribal battles. In the end, it doesn't matter. What matters are products. Fred Wilson is exactly right. At some point there will be a ChatGPT moment. There will be a breakthrough. I don't know when exactly. I've been pretty good at predicting what happens on the internet, but pretty bad at predicting exactly when it happens. I put wide error bars around my time predictions these days.

Robert Hackett: The timeline of a book is so long. It comes out months after you finish writing it. How do you future-proof your writing when you're writing about something so fast-moving?

Chris Dixon: It's a balance. The more abstract you go, the more time-proof it is. But if you go too abstract, it's hard to read. You want to use examples, name particular products. I've read a lot of business books. I remember rereading *Crossing the Chasm* recently, which is 30 years old, and he talks about the Palm Pilot. It really hurts the book that the examples look so old. New startups will pop up, some will do better, some good ones will go downward. You'll get some of that wrong and it will date the book. You have to be careful to avoid things that are too timely, but you can't be too abstract.

The AI stuff in particular, when I started writing it was pre-ChatGPT. I thought I might have to change more of it, but I didn't actually change that much. I talk about this idea of needing a new covenant between creators. There's been a long implicit economic covenant between websites that provide content and search engines that consume and share that content. The covenant is: I'll let you use some of the content if you send traffic back. New AI systems might change that covenant because they won't be sending traffic back, they'll just give the answer. So what's the new covenant? That feels like a timely thing that will last for some period of time.

Robert Hackett: You've got seven applications in the back of the book for potential uses of crypto. If you had to pick one, what's your favorite?

Chris Dixon: There are two dimensions. What would have the greatest impact on the world, and what do I think is intellectually cool? For example, the finance section. If you could dramatically improve and lower payments throughout the world and lower remittance costs, that could have a massive economic impact on lots of people who pay 7% to send money back home. That's instantly life-changing. If we figure out the right way to let creators use NFTs to build a direct relationship with their audience and sell them digital merchandise, that could have a transformative effect on how much money creators make on the internet.

From an intellectual coolness point of view, I really love the idea of collaborative storytelling. A community of people collaboratively develops a universe. Think of Harry Potter or Marvel movies, but their own original universe. They are rewarded for their contribution proportionate to how much they contribute through tokens. They can continue to develop the world, fork the world, add new characters, evangelize the world. That community can make money by licensing the IP to make movies and cartoons, and some of that money flows back proportionately to the users.

Why do I like that idea so much? It's so different from the typical media enterprise that has a death grip on IP and rides it out for decades. It solves a number of problems. Hollywood has a problem: it costs hundreds of millions of dollars to market new IP, which is why we have all these sequels and superhero movies. On the flip side, you have all these active, creative, enthusiastic fan bases who want to be a bigger part of it. They write very good fan fiction. They want to be participants but feel left out and have no way to make money doing it. This idea brings those two things together.

It harnesses the power of community. I talked a little dismissively about Dogecoin before, but there's something really positive about it: the power of that community. Millions of people united around a token who go out and evangelize it. Bitcoin is the most extreme example. Tens of millions of true believers who spend all their time blogging, writing, tweeting, going on TV. What if you took that energy and put it into a narrative universe? People who are passionate about it, who help create it, who fork it, and who evangelize it. That solves the marketing problem. There's no Bitcoin Incorporated putting up billboards. It's the distributed community.

It also creates an economic model for these users. These people can be anywhere in the world. They don't have to be in LA. They don't have to know people. The same way Wikipedia democratized the left brain, this democratizes the right brain, the creative side. You take folks who are left out of internet economics. There isn't really a model to be a writer on the internet outside of going to LA and joining the traditional media world. This provides an internet-native way for those folks to pursue creative activities.

I think this should be a golden era for creators. You have five billion people with smartphones. You can push a button and instantly reach those people. People are as passionate as ever about music and stories. The only reason it is not a golden period is because five or six giant intermediaries inserted themselves between the creator and the audience and took most of the economics. If we can figure out a way to bypass those companies, we can transform the economics.

Robert Hackett: The take rates chapter is really important for understanding that idea.

Chris Dixon: "Take rate" is the percentage of money that flows through the network taken by the network operator. Facebook and Twitter have pulled off close to 100% take rates, which is unheard of in any other area of the economy. It has destroyed the economics of the internet for most people.

Robert Hackett: What will make this book a success in your view?

Chris Dixon: A couple of things. I really hope it will be the book that people who work in this industry can give to their family and friends to explain why they're excited about it. A lot of people who work on blockchains have told me they have trouble articulating that excitement. I wrote this book to be a concise encapsulation of why I'm spending my life doing this. The initial signs are positive. 30 to 40 folks in the industry read it and seem very excited. A lot of them are buying it for their company.

The next rung would be reaching the people who don't have an opinion. Crypto and blockchains have become somewhat politicized. There are people who have decided anything involving them is evil, and people on the other side who love them. But 80% of the world doesn't have an opinion. I'd like to reach some of those people and say, "This is really promising, and here's why. Maybe you should be open-minded to getting involved, whether as a developer, entrepreneur, or just a user."

Robert Hackett: My favorite review so far has been from Kevin Kelly, the founder of Wired. He said, "This book changed my mind."

Chris Dixon: I know. I've been a huge fan of his my whole career. I've read all his books. I had the privilege of meeting him a few times, and it was always frustrating because he agreed with everything but didn't agree with that. For him to say that and to let us put it on the back of the book... I think if you get a quorum of enough minds changed, that really makes this book a success.

Robert Hackett: If you had to write a second book, what would it be about?

Chris Dixon: I don't have any plans. I did like the process, so if I had the right thing, I would do it again. When I was writing this, on the negative side, it was frustrating how no one seems to understand this topic. On the positive side, how often in your career do you have the opportunity where there's a major tech trend and the delta between how you see it and the popular view is so great? It was the opportunity of a career to try to close that gap. With AI, people in AI think it's great, and the rest of the world thinks it's great. There's some difference in details, but not that much of a gap. Here the gap is so broad between the truth and the perception. It would be hard to imagine that happening again. I hope the space evolves and I could write a sequel. I have some other ideas on software and innovation, the history of software, but nothing planned.

Robert Hackett: Last question. Pitch yourself 15 years into the future. You're thinking about a sequel to *Read Write Own*. What is the word that captures that next internet era?

Chris Dixon: I've always thought of it as thesis, antithesis, synthesis. The read era was the protocol network era. It showed the great societal benefits of that architecture. The corporate era broadened and democratized publishing, extending the ease of use and features of the internet. The own era combines the best of the read and write. If you believe that, there is no act for it. It's the end of history.

Robert Hackett: It's the end of history?

Chris Dixon: If you had to pick something, the next obvious vector would be deep immersion. I know it's a sci-fi cliche, but it's going to happen. People living inside of these metaverses with full three-dimensional immersion. The AI stuff is also very interesting. Does that lead to some other kind of era of the internet? I don't know.

Robert Hackett: Chris, thanks so much for sharing this conversation. Everybody else, this book is on sale. You can buy it wherever you buy your books.