Why a Leading Venture Capitalist Is Betting on a Decentralized Internet


Credit: gmast3r/iStock/Getty Images Plus

Chris Dixon, arguably the most prominent venture capitalist focusing on blockchain, is friggin’ tall. Prior to sitting down with him inside the spa-like Silicon Valley offices of Andreessen Horowitz in August, I had talked to him a handful of times over the years, but always on the phone, so I didn’t exactly know about his height.

As far as either of us can remember, Dixon and I initially talked in the mid-2000s, when he was running his first company, SiteAdvisor — a web security startup that was bought by McAfee in 2006. Next, Dixon cofounded Hunch, a wisdom-of-the-crowds recommendation site that eBay scooped up in 2011 for $80 million. Dixon joined VC superpower Andreessen Horowitz in 2013, around the time he made his first blockchain investment, in Ripple. He’s since led the firm into crypto investments that include Coinbase, OpenBazaar, and Mediachain. In June, Andreessen Horowitz unveiled a $300 million crypto fund and brought in blockchain guru and former federal prosecutor Katie Haun to run it.

After Dixon and Margit Wennmachers, the firm’s communications chief, greet me warmly in a hallway at Andreessen Horowitz, in Menlo Park, we make our way into a blah conference room. Dixon sits and stretches out his long frame. He’s wearing a gray polo shirt, gray jeans, red-and-white-striped socks, and sneakers. The following is an edited version of our conversation, beginning right after Wennmachers reminds me that Dixon doesn’t like to talk about his private life.

Kevin Maney: You just don’t like to talk about personal things?

Chris Dixon: There are a lot of crazy people. We have people coming to our house thinking that we keep bitcoin there.

Okay, well, this is a little personal: I’m intrigued that you were a philosophy major.

It was A.I. meets computer science that first got me introduced to philosophy, and reading Gödel, Escher, Bach, which was really influential for me. I wrote my high school thesis on whether machines will someday think, and then I studied philosophy at Columbia: philosophy of language, philosophy of mind. Then I got interested in history and how innovation works, and this has been a constant interest of mine.

Does any of that factor into what you do now?

If you go back and look at a lot of great innovations, they were driven by small groups of slightly crazy people who were motivated by what I would call the interestingness of the problem. It wasn’t practicality, right?

That sounds right.

For example, if you look at a lot of the early technology that bitcoin is based on… these people were not trying to disrupt the Federal Reserve or create computers to take on Facebook and Google. They were interested in these very obscure, specialized problems.

I wrote this blog post that kind of got popular. I wrote “what the smartest people will do on the weekends is what everyone else will do in work in 10 years.” It’s more than a coincidence that so many important technologies began in garages and dorm rooms. It has to do with the fact that there are very few contexts in life in which people are able to focus on things that have a 10-plus-year horizon.

There’s academia and maybe government-funded things, but the vast majority of people are private-sector people — that’s where a lot of talented people are, right? It’s their weekends and nighttimes. They’re motivated by interestingness. It tends to be predictive of things that will matter 10-plus years from now but have no practical importance today.

Satoshi’s paper describing bitcoin doesn’t talk at all about disrupting the world.

I have it on my wall. The white paper actually fits on a poster—it’s actually that short. It’s a very modest kind of opening. So, there are two interpretations of the paper. One is that it’s a submarine strategy of, “oh, this is a little modest thing.” The other interpretation is that it’s a deliberately understated technical discussion that had a secret intent.

Do you lean one way or the other?

It’s complicated. There was a civil war in bitcoin, and I was on the losing side, I guess, the so-called big block side. In theory, it was a debate over literally a single number in the GitHub code, which is called the block size. In practice, it was really a debate between two broader visions of blockchain technology or bitcoin. One was that the purpose of this is to disrupt the Federal Reserve — the cypherpunk, libertarian view. The other side, which I’m on, says that this is a really interesting new architecture where you could build lots of things that could be financially related, like money.

I think bitcoin itself is interesting, but could do much more than that as well. And this side wanted to not only change its parameter, but do other things — like add a more rich programming language. Bitcoin actually has a very basic programming language. So our side lost. But then Ethereum came out [in 2014] and those folks were on my side.

So you lost, but in losing, won.

That became a rallying point for the technology community — Ethereum and all the things that it inspired.

Was that an ah-ha moment for you about blockchain?

I came to it from the perspective of looking at the original promise of the internet, and the greatest thing about the internet was the fact that it was governed by open, decentralized protocols. And that was basically the case up until 2008 or so. Probably the smartphone was the point at which it flipped and the centralized services of Google, Apple, Facebook, and Amazon became so powerful.

Right. The Four Horsemen…

Until then, Mark Zuckerberg or Larry Page could put up a website and if more people came to that website, great. There was no worry that [big internet companies] would say, ‘hey, that website looks good to me, let me take 30 percent of it, or hey, let me change the rules.’ You just couldn’t because the fundamental architecture was decentralized.

Imagine the way a well-functioning city works. I spent a lot of time in New York, so I think of New York. You have the streets that are publicly owned and then you have this interaction between public and private ownership. You can start a restaurant that’s privately owned. But the private ownership of the restaurant depends on the public ownership of the streets because of the foot traffic.

Now that public-private balance is off?

Today the internet is much more like Disneyland. If I’m building a restaurant in Disneyland and Disneyland thinks I’m making too much money, they may raise the rent or change the rules.

That’s what building on Facebook or Google and Apple is right now. We live in this sort of Disneyland internet and I don’t think that’s good for a whole bunch of reasons. I just don’t think it’s as diverse and exciting and interesting and creative. I think it has destroyed the business models of creative people. If we don’t change that, we’re going to pay a significant price in entrepreneurship and foregone opportunities and not allowing the next Mark Zuckerberg, the next Larry Page, to start businesses and let them grow.

You need private and public. I’m not saying it all should be public because then you end up with another form of planned economy. But we’ve lost that balance today and there are only two ways I could imagine putting the genie back in the bottle. One is regulation. Now, my personal bias is not towards the regulatory. I’d rather see these problems solved through free market solutions, by offering better products.

And blockchain will bring those better products?

For me, the most interesting part of blockchain technology is that you can provide much richer and more advanced protocols. They have the best features of Web One in that they’re governed in a decentralized way, and in a way that the rules are fixed and people can build on them and invest in them and know that the rules won’t change. But they have more advanced functionality than protocols of the Web One era.

One way to think of a blockchain is as a community-owned database. In Web One, there were no databases. In computer science terminology, there’s no way to keep state. You just look at Ethereum today, you can store any arbitrary code, any arbitrary names, any arbitrary thing. It’s a very rich kind of database, and so you can build much more powerful services that also have those properties of Web One and Web Two. Some people call it Web Three.

Web Two was don’t be evil. Web Three is can’t be evil. You bake it into the code that you can’t be evil.

Well, the don’t be evil era has led to a lot of stuff that’s kind of evil in its way.

They are incredibly powerful, these networks — Facebook, Twitter, Amazon, Google, and so on. Trump attacked Google for its news rankings. This is going to be more and more like political flash-points. At some point people are going to really ask, who decides who gets de-platformed? Who gets demonetized? Like what are the rules? Did we as a society decide we want 10 product managers at a random company deciding on what’s real and fake news?

I don’t think that’s a good idea and, by the way, I don’t think these companies want to do that, either. And so, isn’t there a better way? That’s why, for example, in the crypto community there’s such an obsession with governance. Governance is: how do we not make that same mistake again? How do we make sure that we have new platforms where, instead of saying don’t be evil, you can’t be evil. It’s baked into the code. That would all sound like a utopian fantasy except once you do that, it also creates lots of economic incentives for entrepreneurs and investors to invest in those platforms.

Is that starting to happen?

I see five interesting entrepreneurial teams a week coming out of universities and tech companies who say they want to build something. The typical story goes like this: “I was a smart computer programmer in my random town in the Midwest. I read a bunch of books about innovation in Silicon Valley. I moved out here to work on that. I got a job at Facebook. I’ve spent the last years doing ad twerking.” And they’re like, “This is not what I signed up for. But on the weekend I was hacking on Ethereum and this just feels cool and exciting. It’s slightly subversive.”

Blockchain is Silicon Valley raising the pirate flag again and doing something interesting and slightly dangerous and subversive.

The PC was interesting and subversive, and then Microsoft co-opted it and built a world power. Then the internet was interesting and subversive, and 20 years later it’s been subsumed by superpowers. Won’t that happen again with blockchain? Or will decentralization solve it?

We’ve got to at least try to do this other thing.

The common way to make a protocol including web protocols is you have a standards body or some foundation define the protocol, and then you have multiple instantiations. So think of the web browser. You have a separate group like the IETF define the protocols and then you have Chrome, Firefox, multiple instantiations of it. That’s how Ethereum works. That’s how all the best protocols work.

Bitcoin is the one that doesn’t do that. That they don’t have a separate protocol spec. The protocol is the code which is controlled by a few people at one company, or mostly one company. Yeah, look it’s very possible that history will repeat itself. But at least we’ll get 10, 20 years.

Ethereum was a big turning point for you?

It’s a very interesting platform. It’s a full programming language. It looks very much like JavaScript. You can write ten lines of code and you’re already up and going.

Blockchain networks like Ethereum have a radical new business model. It is not to take some piece of the action or sell ads to generate revenue — it’s to issue coins or tokens and have the network hopefully grow in popularity. If the tokens are well designed, they will grow in value along with the network.


That is a really beautiful business model. With every network in the history of technology, there’s always been a second phase where the network has internal battles. [For instance, Facebook vs. news publishers, who accuse Facebook of sucking up all their revenue.]

And it feels like a wasted thing because the network and its ecosystem have so many common interests. Like, why are you fighting each other? One of the many beautiful things about the token model is it potentially resolves that because contributors to the network own the network’s tokens, and they all benefit when the tokens appreciate.

Imagine if Twitter had been built through this sort of blockchain protocol. It could have been the micro-messaging protocol for everything [not just on Twitter itself], and Twitter could have had a model where Twitter and everyone using it just owned a bunch of the tokens. The more valuable everyone made Twitter, the more their tokens would be worth. It gives everyone the right incentives.

You invested in Ripple five years ago. Why launch a crypto fund at Andreessen Horowitz now?

Right, that was the very first investment. And then Coinbase in November 2013. And then a few others along the way. But then it really ramped up post-Ethereum. A whole bunch of new people got inspired. They saw it as something to get behind and it wasn’t this cyberpunk libertarian thing. This is a technology platform. Fundamentally what we’re in the business of doing is following technical people. So we just saw this massive influx of talent.

And so to our fund we said, “We want to be aligned with this new business model, and the right way to be aligned is to buy the tokens not to try to buy equity or invest in some toll-seeking gateway.”

How do blockchain entities fit into a traditional VC world?

On one hand we see it as very much like VC where we’re investing in technologists, trying to build new technology. On the other hand, a venture capital firm has a specific definition by law. And if we are going to be doing what we’re doing in the crypto fund and buying tokens, we could no longer fit into that definition.

Really? Officially one of the most recognized venture capital firms in the world is actually no longer an actual venture capital firm?

No, no. Just the crypto fund. We’re not, according to Dodd Frank, a venture capital fund. And that means we had to go register more like a hedge fund.


I can’t use encrypted messaging. There’s all sorts of SEC rules. But we can buy the tokens.

How did that go over with your limited-partner investors?

I went out just three months ago to all the LPs and described what we were doing. I said it was crazy and risky, and different types of risk.

What are you looking for at this stage? What do you think you’re going to be investing in?

Two very broad categories: infrastructure and applications. Infrastructure is, basically, layer one blockchain. Bitcoin would be in that category. Ethereum. Infinity. Competitors to bitcoin like Zcash, and Monero. It’s very hard to see bitcoin getting displaced just because of the brand momentum and trust level. It’s been around for ten years. The two most viable replacements are probably Zcash and Monero. They’re bitcoin competitors that have privacy features. One of the commonly misunderstood things about bitcoin is people think it’s anonymous. It’s the opposite of anonymous. It’s all public. If you’re a Venezuelan and you’re trying to get around the draconian tyrannical government, bitcoin might actually be a really bad solution.

So that’s the stored-value camp. We spend a lot more time on the other side, which is what we think of as blockchain computers. And those are computing platforms that let you build any arbitrary application in a way that you can give trust guarantees. You’ve never had a computer before where you can give trust guarantees to different parties. That’s really the new feature of this computer. The system as a whole can give users, developers, creators, other parties, guarantees about how it will behave in the future.

Trust guarantees are the game-changer?

That is a brand new thing on earth. If you look at every kind of computer over history, there was a new thing about it. The personal computer was new because it could sit on your desk. This [holds up his mobile phone] computer was the first computer you had on you all the time. It had a camera. It had GPS. And then people came up with Snapchat, and Uber, and Lyft, and Instagram. Notice that each time there was a new set of applications, the applications took advantage of the new capabilities of that computer.

We’re probably not in the Apple II phase of those computers yet. I don’t know exactly when the Apple II period will come. Hopefully soon. One of the good things about these computers is you don’t need to build new hardware. They’re software-based computers, which is kind of a weird thing, but it’s a new thing. They can be built relatively quickly.

What types of applications do you think we’ll see on these blockchain computers?

It’s a new computer. It has new capabilities. The new apps should take advantage of those capabilities. If someone comes in and says, “I’m just gonna make a Fortnite on Theory,” I’m like, “That’s not a good idea. It’s going to be slow and it’s not going to take advantage of the new features, right?” It’s got to take advantage of the new features.

There’s all sorts of interesting stuff happening [in apps]. Stuff on the financial side, stuff on the gaming side, specifically around what are called non-fungible tokens [like Cryptokitties], which is kind of a new way to think about virtual goods. There’s a bunch of stuff happening where people are taking real-world assets and putting versions of them on the blockchain.

Paying for a cup of coffee is probably the least interesting application for cryptocurrency. That’s been solved pretty well. Between one billion and two billion people have smartphones but not bank accounts, or even an official identity. You have to have an official identity if you want to have a relationship with the bank. And so, I think the developing world is very interesting.

Any other apps to watch for?

I think virtual goods could be really interesting. I think we’re going to see a lot of experimentation. A lot of people think you could re-architect a lot of core internet services. Filecoin is an interesting project where they’re trying to re-architect storage using blockchain. Airbnb for storage. Anyone can offer storage to the network. Anyone else can consume storage from the network. It’s all mediated by these blockchain protocols.

You brought on Katie Haun to help run the fund. She’s a former federal prosecutor. Why was she the right person?

She is a former federal prosecutor, but she got into crypto five years ago or so. She ran a lot of the government’s crypto task force, and then joined the board of Coinbase with me. That’s how I got to know her over a year ago. She started teaching a course at Stanford on crypto. She advises a bunch of companies. Kept hearing her name from various crypto projects about how she was helpful. Regulatory issues are important in this area. We care a lot about compliance.

You’ve been outspoken about the promise of decentralization. Will someone re-create Facebook on blockchain and it would be better?

My guess is it’s better to fight the next war than refight the last war. Facebook could continue to be Facebook. Google could continue to be Google, but there will be a whole new set of services built on blockchain. The Big Four might even miss those things and still be incredibly huge companies.

VCs like working with company founders. Now we see companies that, in some cases, might not have any founders at all, right? They’re founded with a white paper.

No. The ones we invest in all have founders, at least initially.

We’re announcing an investment in [decentralized cloud startup] Dfinity. There is a founding team. They have founding people there. Over time, eventually, if they’re successful, it will just become this global network and the company will dissolve away. But I think that takes a fair amount of time.

So it feels similar?

I think so. I would actually argue that if you’re in VC today, a lot of our competitors and peers are wondering what to do in this space because even if they don’t believe in it, when 20 percent of the smart engineers that walk through your doors are working in this category, don’t you have to do something?

Are initial coin offerings a positive development? Dangerous? A challenge to traditional venture capital?

ICO has become associated with stuff that happened a few years ago, none of which we participated in, when people were using some of this new technology to get around securities laws. We are not fans of that. And we didn’t participate in that. But to other people, ICOs mean token sales. If that’s what you mean, then we’re all in.

As a lot of money flowed into the space, some bad actors came in as well. People would come up with white papers which were not technically sophisticated and get people to invest in them. There has been bad behavior. What’s unfortunate is the bad behavior has led some people to think that the whole space is full of bad behavior. I would liken it to the 1990s internet where you did have a lot of bad companies and a lot of nonsense, but you also had Google, Amazon. In the end it was a very, very important technology movement.

You see that period of bad behavior fading now?

Most people that we know that are technologists, they want to follow the rules. They just want to know what the rules are. And right now they don’t always know what the rules are. And so, they all seek out very good counsel. We try to help them. But it’s a new area.

Regulations are actually helping?

Our belief, and I think this is supported by emerging thinking from the regulators, is that as these networks launch and as they become truly decentralized, and they are no longer controlled by kind of the creating entity, they will become regulated under commodities laws, not securities laws. Commodities laws allow for the exchange of these goods for purposes of use and other things. There will still be lots of regulations making sure you don’t manipulate markets and other kinds of things. But as commodities, crypto currencies are more freely exchangeable. And we think that’s very important for the networks to be used.

Bitcoin has been declared a commodity. Which means I can sell you a bitcoin and not have to make sure you’re an accredited investor. That would be a pretty clumsy process if I had to do that.