120 - Reinventing the Internet | Marc Andreessen & Chris Dixon of a16z

Welcome to Bankless, where we explore the frontier of internet money and internet finance. This is how to get started, how to get better, how to front run the opportunity. This is Ryan Sean Adams, and I'm here with David Hoffman, and we're here to help you become more bankless. Guys, we have a fantastic episode for you today. This is an episode that's gonna break the internet, or maybe fix the internet, David. One of those two things, because we have two big brains on this podcast, Mark Andreessen and Chris Dixon. Mark was here at the genesis of the internet, okay? He was a builder in web one, he turned into an investor in web two and web three, and also Chris Dixon, whom you know from previous Bankless podcasts, the master of Web3 mental models, put them together and we get this fantastic episode. A few things to look out for. Number one, we talk about the original sin of the internet, what exactly that is. Number two, how Web3 actually fixes the internet, or maybe a better term, reinvents the internet. Number three, all the different ways crypto is like the early internet, including the fight for freedom. And of course, Mark knows these things because he was there in the front row. Number four, why the open version of the internet, including crypto, eventually wins. And number five, we end with this, and this is a perfect way to end it Mark giving advice to his 22 year old self. Okay, what a fantastic episode, David. Any thoughts going into this one?

Yeah, we frequently use the line and we brought it up a number of times in the podcast of how crypto is speedrunning the history of money and finance. And so

History is known to rhyme, of course, and history never repeats, but it rhymes. And we ask Mark about the parallels between the rise of the early internet, the rise of web one, and the rise of web three, and how strong those parallels actually are. And of course, this is when the history always rhymes line gets brought up, but Mark says that he doesn't take that statement lightly. It's always easy to say that history rhymes, but Mark says that to the degree that Web3 is playing out in the same exact path that Web One played out is quite salient. All the same players exist, all the same patterns exist. And so having this one particular person who was here for the fight of the open and free web one and is also here for the fight between the open and free web threes, I think is a very unique perspective, especially as he grew up from being a builder of the early internet to an investor in the web three internet. He has a lot of lessons from history to teach us in an industry that so well is mimicking the early path of its own predecessor.

Guys, there's so much more to talk about. In fact, David and I are about to go talk about that. That's contained in the debrief episode. Okay, so if you are a bankless premium subscriber, you get access to the debrief episode where David and I talk about the episode after the episode to unlock that, upgrade your membership, become a bankless premium member. We'll include a link so you can do that in the show notes. Guys, we are going to get right to the conversation with Mark and Chris.

Bankless Nation, we have a special treat for you today. Never before been assembled. We've got two of the biggest brains alive in technology investing. Mark Andreessen, Chris Dixon. Allow me to introduce them to you for a moment, although they need no introduction. Mark Andreessen, he is the Andreessen in Andreessen Horowitz, otherwise known as A16Z in the crypto space. He's also the founder of Netscape. I believe he accomplished this feat at 22 when he built out the mosaic browser. And he is perhaps the only person in the world who navigated both the birth of the internet, going from a web one builder to internet investor, to where we are today. A16Z, of course, is a massive investor in crypto, in web3 technologies. We're going to talk about that, including a $4.5 billion fund, their fourth fund that was just announced. Mark, how are you doing?

I'm doing fantastic. Thanks for having us.

Yeah, it's great to have you on bankless. We've been wanting to have this conversation for a while. And of course, we also have Chris Dixon, who is a repeat bankless guest and one of the key people leading that new crypto fund that I was just talking about. I've called him on Twitter anyway, the Kendrick Lamar of Mental Models. Okay? Because this guy drops influential ideas.

I had to Google Kendrick Lamar, but thank you.

You're very welcome. I know Mark. Everybody else.

Anyway, he's certainly given bankless listeners a fantastic set of mental models through which to understand and explore this whole crazy thing we called crypto or web 3 or whatever we're gonna call it in this episode. Welcome back, Chris. How are you doing?

Good, good. Thanks for having me.

All right, Mark, we gotta set this up because we haven't heard from you yet. We've heard from Dixon on a few things already, but um, you were here for the birth of the original internet, the thing that we now call Web 1.0, though it wasn't so obvious, you know, it was just the first version of this thing. But there's this perception around, I think it bleeds through crypto as well, that there's a problem with the internet of today, that it's somehow broken. Do you think the internet is broken today? And if so, where did we go wrong?

Yeah, I call this the original sin. So there was an original sin of the internet, and it's actually sort of this ironic twist. So I'll take you all the way back. Once upon a time, it was actually illegal to have money on the internet. Like it was quite literally illegal to do business online or to transact online. And this is prior to 1993. And the reason it was illegal is because at that time the US government was paying for the internet backbone, and they had something called the AUP, the acceptable use policy. And it basically said that it's a research project. It was a federally funded research program through the National Science Foundation. And so, you know, taxpayer money. And so they quite plausibly at the time said, look, this is not for commercial use. You can't like set up a store. You can't do, you know, stock trading or anything. You can't transact money. And so that began the sort of internet ethos, you know, as a very kind of non commercial thing. And there were a lot of great aspects to the non commercial ethos at the time. There was, you know, this is kind of, you know, the rise of open source, you know, the so called free software movement, and then the open source movement, you know, kind of went hand in hand with this. And then just the general internet ethos of sharing and openness and open protocols and so forth kind of emerged out of that. So

So there was a lot to like. And in fact, you know, when I got involved, when I did my work, which was originally also funded by the National Science Foundation, you just assumed that everything was open source. And then when we started Netscape, the company,

you know, the overwhelming reaction we got out of the gate was well, that's just it's impossible to imagine. There can't be an internet startup. There can't be an internet company that would ever make money because everybody knows there's no money on the internet. You can't do business online. And so the big reaction to Netscape was just like, well, that's impossible. Like that won't work. Um

And so, you know, and then at Netscape, we kind of set out to, you know, basically turn the internet into a commercial medium. Like we wanted to make it real for people in their normal daily lives. And a big part of that was to make it real for people in their economic lives. And that worked, right? Ultimately, the rise of e-commerce and so forth. But

the sort of, you know, original sin was this sort of lack of money, right? Which is the big virtue of the internet is anybody can plug in, anybody can like send email back and forth, anybody can, you know, upload whatever data they want, anybody can communicate with anybody else. That's all great. But, you know, there's no money. And by the way, the fact that there was no money on the internet actually was because of a sort of a deeper underlying, you know, I would say flaw or again, sort of sin, which was basically zero trust.

Right. The internet was built as an untrusted network. And again, you know, and again, like that worked really well for what it was, but there was no trust. There was no trust, there was no money. As a consequence of there being no trust and no money, there's sort of none of the other aspects that you kind of expect from a modern market economy. There was no concept of, you know, economic incentives were just simply absent. And the lack of economic incentives has caused all kinds of problems, you know, including like the entire problem of spam, which has been dogging the internet for, you know, 30 years, is basically a consequence of not having correct economic incentives built in for who sends messages, who receives messages. And then, you know, we can spend hours talking about all the ways in which the internet kind of, you know, in a lot of ways went sideways over the years because you just you couldn't have money. You know, the fact the internet, the sort of consumer internet economy got built on advertising, right? Which has all these kind of external, you know, kind of uh, you know, kind of questions, issues now. Again, it's a consequence. Like if you can't charge people directly for things on the internet, then you know, advertising is the only viable business model, you know, basically indirect payments, you know, with somebody paying behind the scenes. So this was just a fundamental issue. We spent actually a lot of time, like we knew this at the time, and we actually spent a lot of effort in 1993, 94, 95, 96 trying to fix this. And I could take take you through the sort of disaster stories of how we tried to fix it at the time, but none of the fixes took. And, you know, in retrospect, we just we didn't have the blockchain. Like this was, you know, way prior to Satoshi. This is way prior to the Bitcoin white paper in 2009. You know, this is way prior to the conceptual breakthrough of the blockchain as sort of was first incarnated with Bitcoin. And so, in retrospect, like we now know we couldn't do it at the time because we didn't have the foundational technology needed to actually do this. But now we have that technology, and so now we believe we now have a chance to kind of go back and fix those original problems.

I would just add to that, the um I mean Mark, Netscape created JavaScript, created cookies, right? SSL too? Was that Netscape? Okay. So the other thing I think the one I think about the history of the internet is I think of it as sort of these core primitives, which you need to have. So you need encryption. You need ideally you have some system of identity, you have some system of

money, right? And you were, if you go back and look at it in the 90s, you were racing Netscape and there was a fight going on. So I think this is often

Sort of revisionist history around it. There was a there were two visions of the internet. There was a so-called information superhighway, which was Disney Comcast and a very centralized internet. And Mark was on the other side, which is sort of the open decentralized internet.

And as I tell you, you'd know better, Mark, you were racing against those, the that other side to add features. And that's why you had to build JavaScript, right? You had to build SSL and these other things. So to be kind of a feature parity.

You know, but because it was moving so quickly, you know, you people didn't really have the time to set back, step back and say, hey, let's rethink this advertising business model. Let's go build internet money, right? And even SSL, right,

was extremely controversial. And Mark can tell his story better than I can. In fact, it was illegal, right? Uh 128-bit Netscape was classified as munitions. And you'd go in front of Congress, just very similar to crypto today, where they would say, Well, who wants to use encryption? Obviously, only criminals would be and terrorists and things would use it. And of course, Mark would say, Well, someday people will put their credit card on the internet and buy things, you know, buy kitchen appliances on websites and things like this. And they they would laugh at you and say, Who does that today? And so why shouldn't we just ban it? And so they actually literally did.

uh

heavily regulated encryption and in fact there was a big battle to try to actually ban it altogether because it was so obviously only for criminals. Right. Right. But you can tell the story better than me, Mark.

Yeah, that's right. I mean, look, so encryption, right? So the original crypto, encryption, right? Being able to actually like have secrets. Like, you know, encryption has been used in like warfare for, you know, hundreds of years, right? All the way back to the American Revolution, they used uh ciphers and so forth. And then, you know, obviously the World War II was won in large part due to breaking the enemy's encryption. So like there was this sort of heritage of encryption in the sort of military intelligence spheres. But encryption was never used in daily life. Like ordinary consumers didn't really have access to any encrypted anything, basically all the way up through up through the early 90s. And we actually negotiated, I believe, the first contract. I think we actually had the first kind of consumer-facing contract for encryption with a company at the time was known as RSA, which had all of the rights for encryption at the time, encryption protocols. And so we actually negotiated that at the time. And then we as a subsequently released the first consumer product that had encryption built in, which was the Netscape Navigator. And then, you know, as Chris says, we had to actually invent a protocol. We had to invent a protocol called SSL, which was the first encryption protocol that was used online and is still used today.

Yeah, so we had to do that all from scratch. And then, yeah, because encryption had this military heritage, there were literally two regulatory regimes for crypto. We could actually, for encryption, we could actually ship strong encryption inside the US. We were actually allowed to do that, which was good, although they kept trying to ban that. For export, the Netscape Navigator was actually classified under the same munitions regulation as Tomahawk cruise missiles at the time. Right. And so you could have a missile, you could have like Sarah Nerve Gas, or you could have like Netscape Navigator. And it was equally, it was sort of under the same regime. And so we were not allowed to export the strong encryption version. So we actually had, you know, at this point, software still came in boxes, right? And so there were literally two different boxes. There was the strong encryption version for US citizens who got the special one. And then, you know, there was the international one. And we and we put you know a big warning label on the international one, like, you know, this is the weak encryption, like this is the bad version.

You know, but then of course what happened was encryption was like, you know, it was it's just math, right? It's just math and numbers and bits. And so people outside the US knew how to do it. And so they started building competitive products. And then we basically had this giant fight with Congress and with the intelligence agencies of the time. And as Chris said, yeah, they tried to ban encryption outright. You know, they still try every now and then. It still comes popping back out again, where they, you know, different governments try to ban encryption, but they keep kind of swinging and missing on that one. Because it's hard to ban encryption because like if you ban encryption, then everything is open for, you know, basically for attack and hijacking by terrorists and criminals. And so that you have the other side of that problem.

They did all the same, you know, it's all the same dark warnings. You know, they'd come in and they'd do the briefings. They'd give you the, you know, we're gonna give you security clearance for the day, and then they open the briefcase and out comes the parade of horribles of like all the awful shit happening in the world. Right. And it's like it's all gonna be your fault. Right from here on out, it's all gonna be your fault. Like anytime anything bad happens, right? Anytime, you know, there's a bombing or anything, it's gonna be like on your head and we're gonna pin it on you, right? And so, you know, they tried all that stuff. They tried all the stuff they're doing right now in Washington with crypto and blockchain. We fought it out then, and we're basically fighting it out again.

It points to this fundamental asymmetry as well, which is

With new technologies, it takes a long time for the positive use cases to develop, and especially with things like encryption. And the early adopters will just naturally be sometimes bad actors.

So you take e-commerce, right? So encryption, SSL encryption, right? In 1995, a senator saying it's mostly used by criminals, that might have been correct in 1995. But you had this slope of adoption of legitimate uses of encryption, which was a hockey stick, which of course today we take for granted, right? But it would be very easy for an overzealous regulator in 1995 to have, you know.

Killed that growth at the beginning. And of course, you wouldn't even know 10 years later you'd done that. You know, there's probably a ton of regulations that have already done that anyway. So I think we're at sort of, we could talk about that maybe later, but I think we're at a kind of a moment like that now with

with you know

our crypto Web3, where

you know, you have this chorus of cynics who only point to the bad things and ignore any possibility of good things.

Well, and I would say there's several, and these are the counterarguments that we deploy, that we deployed then and we deploy now. So one is yeah, the good use cases are going to emerge. So this is going to have all kinds of positive use cases. So that's one. The second is like, look, this is gonna be a giant industry, right? This is gonna be like a giant industry. This is gonna create jobs, like this is gonna create economic wealth. Like job creation is itself is an actual national security thing. Like you actually, if you're concerned about national security, you actually want your economy to do well and you want your economy to be the place where new things happen. There's also another national security argument, which is if these things are going to exist, you want them to happen in your country, right? This is a big argument we used at the time, which is like, look, there is going to be commercial software on the market that's going to have strong encryption. There are going to be hundreds of vendors outside the US who do this from all these other countries if we're not allowed to do it. And then you, the intelligence community, is not going to be able to get to those companies the way that you can just come like talk to us because we're American citizens and you know we're patriotic and so forth. And so you might not get everything you want, but at least it's happening on your territory in your government, you know, under in your country as opposed to somewhere else.

And then there's just the broad thing, the really big thing, which is what I keep coming back to, which is like, look, like, you know, a lot of the same people who are very worried about like bad use of crypto are also worried about like privacy, right? They're worried about like, is people's private information gonna be held secure? Is people's health information gonna be held secure? Is the national power grid gonna be held secure? Is the banking system gonna be held secure? And it's the same, you know, evil cryptography that gets used by bad guys for coordinating terrorist attacks that gets used to secure the power grid and all the hospitals, right? You know, it's a tool. Like encryption is a tool. It's gonna get used in various ways. On net, overwhelmingly, it's a positive. Like you'd much rather live in a world that has high security, you know, than a world that has no security. And then, you know, I think the exact same, you know, concept applies to, frankly, it applies to money, right? Money, money also gets used for lots of bad things, right? Criminals and terrorists use money all day long. And yet we don't say like we shouldn't have money because they use it. Because overwhelmingly, like the positive use cases dominate. And so there is adaptation that the world has to have as these technologies become widely accepted. But, you know, basically 20 years later, you look back and you're like, okay, that was overwhelmingly good that that got adopted, and then the world was actually able to adapt. And that's basically what I think is going to happen with crypto.

Mark, I'm interested in gaining your perspective as to how obvious some of these things were all the way back then. Because the parallels between the fight for the freedom of web one versus the fight for the freedom of web three are so strong, as you guys have illustrated just now. And I see like me and Ryan and like a lot of the cypherpunk ethos built into the Web3 world as truly fighting for the open and permissionless Web3 in the same way that you were fighting for the open and permissionless web one of the world, where we had a fork in the road with two possible outcomes, the centralized web with open surveillance, or the decentralized web with permissionlessness and privacy. And then we had the government regulation player kind of tinkering with the incentives of everyone and having their own role to play. And that same exact pattern is following out now, where we have open public permissionless blockchains versus centralized blockchains, and we are trying to tilt towards open, public, permissionless, and private systems. But I'm wondering your perspective on whether I'm being a naive youth or not, whether I kind of just hope and assume that natural humans will always pick the open and permissionless system, and crypto is just going to win ultimately hearts and minds of the people, and just because it's what the people want, and eventually Web3 will win no matter what. Am I being naive here, or is there a counterfactual universe where like the open and public permissionless web one did not win and we have a different future? Like, how hard do we really have to fight for this thing?

Yes, the the big bull case on the open systems and the internet then or crypto now

The big bull case, the way I describe it is there was this guy, Bill Joy, who was one of the legends in software who developed one of the main versions of Unix and co founder of Sun, which was one of the central companies for the build out of the internet originally. And he had this thing. He had this thing he called Joy's Law after his name. And Joy's Law was no matter how many smart people work for your company, there are a lot more smart people who don't.

And you can apply that also, by the way, for countries. No matter how many smart people are in your country, there are a lot more smart people who are outside your country, right? It's the same thing. And so the the advantage of an open system is that basically everybody can participate. Like you can go basically recruit and draft, right, and build a community and you know, of basically every smart person in the world, every smart person in every company, every smart person in every country, and they can come participate. The economists use the term permissionless innovation, right? Which is like, okay, I want to build something new. I want to build a new app, I want to build a new site, I want to build a new service. And, you know, do I have to fly to the corporate headquarters of, you know, XYZ corporation, you know, and put on a suit and go into a room with 40, you know, other people in suits and like give them a PowerPoint presentation and try to get their approval, or which is permissioned innovation, right? Or I can just like build the thing, right? And so, you know, the internet was classic permissionless innovation, crypto is classic permissionless innovation. So the potential is to draft all the world smart people and then to collectively build something that's just like much bigger and better and you know more comprehensive and more innovative than any individual company could ever build on their own. You know, kind of as Chris said though, the the problem with that is it's like, okay, it's like in the moment, it's like, okay, everything I just described is like hypothetically possible, like it's hypothetically what's going to happen, right? But it's not what, you know, XYZ government bureaucrat today wants to happen. It's not what XYZ, you know, it's not what Jamie Dimon, I'll pick on the people who have picked on us. And so this is all me gonna, I'm I'll return a little fire. I'm not gonna attack anybody, but I'll return a little fire. You know, it's not what Jamie Diamond thinks should happen, right? It's not what Warren Buffett thinks should happen, right? It's not what certain, you know, government officials think should happen. And so you're arguing this hypothetical against people who are like, no, I want to stop it now. And as Chris said, like the problem is the dog that doesn't bark, the problem is like if it's stopped, it's all of the gains are in the out years, right? All of the things that were going to happen are in their out years. And when they don't happen, like nobody ever knows that they were possible. And so you are in this weird situation up front where you're kind of arguing in favor of a hypothetical future that has emerged yet.

My hope, right, would be that like we've seen this movie before, we've seen this happen over and over again. We saw this happen to the internet, we saw it happen to the PC, we saw it happen to the iPhone. Like there have been lots of examples over time of how you have these systems that other people can build on top of, and you get this much bigger and kind of more powerful thing. And so I always kind of, you know, I always kind of I don't know. I'm like they're getting married for the sixth time or something, right? It's like, you know, finally, you know, optimism, you know, is gonna win over experience, right? And you know, finally people are gonna like understand this. And I just kind of find like, no, you know, it's like whatever bureaucrats are in charge of the status quo today, they're still gonna fight it the exact same way the last ones did. You know, we're gonna make the exact same arguments, you know.

Quite honestly, there's a generational component to it, right? Which is like, you know, kids coming up look at these new things as an opportunity for them to make their way in the world. You know, a lot of people when they get older, not everybody, but a lot of people when they get older, they view this as like they want, you know, they have status quo bias that, you know, they don't want anything to change. They have their existing power, you know, they run their existing companies their own way and they don't want things to change. And so there's a there's a generational momentum kind of aspect of it. There's a real movement aspect of it, right? There's a whole aspect of all this stuff that has very little to do with the technology, just has to do with movement characteristics, like being able to have a lot of people who sign up and sort of into an army and actually have, you know, kind of collective power. And so I think we're kind of doomed or whatever, fated to kind of fight these things over and over again. And I don't know, I know what side I want to be on.

I would add to that that I think of the history of computers as sort of there's really two threads. There's the open and kind of closed. There's open cathedral and the bazaar, if you've read that essay. That was a famous essay around open source software.

And so the cathedral is, you know, Microsoft,

later on Apple. Most of the big tech companies today are cathedral. Bazaar was early PCs are interesting. They had aspects of both. I would argue the early 70s PCs were more bizarre. And in fact, Steve Jobs, Homebrew Computer Club. It was a bunch of, you know, that was effectively like a, you know, an NFT Discord or something of the day. Right. And then I think the big, then Mark was involved in, I think there was this sort of understood history of the internet where there was this kind of battle and the open side won. Open source software is probably the most successful example of that kind of ethos playing out. 99% of the software in the world now that we run is open source software. So I think to me, crypto has picked up the baton there, right? I mean, open source software continues, but this is kind of the major important new kind of fight over openness.

I would say one thing, David, is that the open side has Joy's Law, what Mark just described, as an advantage. The disadvantage in the open side is a two-step process. You first recruit the developers and then you build the apps. Users didn't opt to have Linux in their data center, right? They opted to have a system that didn't crash. The developers went and built that system and a system that had robust libraries and ecosystems and everything else, right? And that's very important for understanding. And so, like I example I used in a blog post a couple years ago was Wikipedia versus Encarta. So Microsoft had an encyclopedia, a CD-ROM encyclopedia thing that had beautiful pictures. It was 1998, 2001. Actually, 2001, I think was Wikipedia. So at the time you have Encarta, beautiful pictures. You go look up an elephant, you see a picture of an elephant, some you know esteemed person wrote a description of the elephant. Wikipedia in the beginning is this chaotic shit show, right? Uh, you know, it's people, whatever, putting whatever they want up. Um, but but it's two steps, right? First, you recruit the contributors, and once you get that set of people that are excited and putting in good content, if you look at the stats, it's just hockey stick past, and Cartoon was shut down. I think it was 2006. It was pointless, right? And so once you hit that, you get that incredible growth.

But I think that's important to understand. And I think that's why so much of what we've been doing in crypto in the last few years has been kind of developer focused. And I think only now are we seeing kind of that second step really kind of mature.

I think there's a great story, Mark. One of my favorite Mark stories is uh this also affects your design choices. So in Mark's world in Netscape,

they chose to make all of the transport protocols of you know HTTP, if people that use that, you guys chose to make it all text based, as opposed to binary, right? And if you go back in time oh I'll let you tell the story, Mark, but

Yeah, yeah, yeah. So any expert protocol design, anybody who had ever worked at a big company who understood how to design networks, because there were networks, there were computer networks for you know 40 years before the internet merged.

Yeah.

They were just all proprietary. And so you'd buy an IBM mainframe and they would sell you an IBM network. They had a technology at the time called token ring and it would connect all your IBM computers together.

You know, it you couldn't connect other computers to it. You couldn't connect your Apple Macintoshes or whatever to it, but you could connect all your IBM computers together. And so then they'd have all these protocol designers, right, who were like, you know, world experts at building network technology. And the one thing that they all knew for sure was you do binary protocols. And the reason you do binary protocols is because you need to, you know, networks were very slow at that point. You need to optimize every single bit, right? And so you want to like hack as much data as possible in a few bits as possible, because that's how you get performance out of what we're

So so not human readable.

Not human readable.

Text is not human readable.

Yeah, that's right. So if you looked at the protocol data going over the wire, right, you would just see a gibberish. So it's just a bunch of random numbers, you know, coming across. And then you would need specialized software to try to decode that and repack that. And this was this whole kind of arcane area of network design. And then the internet people, starting originally with things like SMTP and so forth, the internet people started to do these text-based protocols, and then HTTP was a text-based protocol. And the idea of a text-based protocol was it's human readable, right? And so I can read it without having to look for things, and then I can write it. I can emit it, right? And so one of the things that made the web take off early on was that you could build a web server with four lines of Perl script at the time, right? Because you could use all of the standard text parsing technologies built into languages like Perl and Python, which made it very easy to process and to emit text. And so you could basically have this little text processor that basically generated this protocol, and then you could create a web server to do whatever you want. And at the time, you know, there people ran thousands and thousands of experiments of different kinds of web servers they could build.