• D.C.
  • BXL
  • Lagos
  • Dubai
  • Beijing
  • SG
rotating globe
  • D.C.
  • BXL
  • Lagos
Semafor Logo
  • Dubai
  • Beijing
  • SG

Jan 31, 2024, 1:44pm EST

Andreessen Horowitz’s crypto boss isn’t backing down

Chesnot/Getty Images
Title icon

The Scene

Chris Dixon, who runs Andreessen Horowitz’s crypto fund, has become synonymous with the industry that found itself in its darkest, coldest winter yet. Rather than hibernate (or pivot to generative AI), Dixon has written a book meant to educate the public on the fundamental promise of a technology he still believes is the future.

Read Write Own: Building the Next Era of the Internet argues that we are on the cusp of a third stage. Web 1.0 was the “read” era, Web 2.0 was the “write” era. Now, people who use the web will get to own it, too, and share in the gains.

I spoke with Dixon about the book, what keeps him evangelizing for crypto, and how blockchain technology could help make AI more credible. Below is our edited conversation.

Title icon

The View From Chris Dixon

Q: It’s an interesting time to come out with a book on crypto, right in the middle of the crypto winter.

A: There’s definitely some skepticism when talking to people. I started writing the book after the market started coming down, like the beginning of 2022. The markets crashed, obviously. There was FTX. Bad things have happened. I’m very bullish on the technology, but that was frustrating.

I believe that this is the most important technology movement. But when I spoke to people about blockchains, and how they’re new ways to create networks, they would look at me puzzled and say, ‘well, why would I want to create a network?’


Until recently, we had two ways to build networks. And for me, the a-ha moment was, ‘Oh, my God, this is a third way to build networks that has the societal benefits of protocol networks, the early networks, and yet has all of these advanced features of what I call the corporate networks, and therefore can both compete with these modern services, but also bring back these benefits.’

Q: A lot of your colleagues wouldn’t bother explaining a technology that’s not so hot right now. A lot of people have moved on to large language models. Why do you want to go this route?

A: I could go do AI. I just really believe in this. I really got into business because I fell in love with the ideals of the internet. These computers got connected in this open, decentralized way, which was a magical thing.


I started an internet company in 2004. And then I got into social media, and in some cases, invested in some of these companies. I just assumed that we would keep that ethos. Twitter was this open platform people were building on, people were building on Facebook.

Now we are at a point where there’s five companies that control the internet, and I think it’s going to get worse. I’m pro-AI, but I don’t see a version of this where it doesn’t further exacerbate this because it’s a centralizing technology.

It will further entrench these big companies, because it rewards companies with money, compute, and data. And the world hasn’t fully appreciated that we’re really losing this great thing. There may be regulatory approaches to helping with this, and I support those. So I’m not saying that technology is the only solution. But blocking [Adobe’s deal for] Figma is not going to decentralize the internet. The cow has left the barn in some of those regulatory matters. It was the acquisition of Instagram that was the issue, not Figma.

So the other important thing is can we create a new wave of internet services that embody the ideals of the early internet. That’s what my book is about. My argument is that blockchain-enabled networks can be the best of both worlds. They can have the advanced functionality and the funding that people expect from modern digital services, but also return us to the ideals of the early internet.

Q: When you talk about crypto’s image, it’s really the rampant financial speculation that attracted a certain crowd.

A: I liken it to housing. We all believe, generally, that homeownership is a positive thing. You raise your family and it’s something you own. A lot of people’s personal wealth comes through their home. As a byproduct of homeownership, we have real estate, and people that flip real estate, speculators. I’m not anti those people. They play a role in price discovery. Same with the stock market.

To me, it’s a question of priority. There’s so much focus on the crypto trading aspect. I’m not saying all trading is FTX. But I do think it’s part of that culture that led to things like FTX. But the emphasis has been off. They’ve crowded out a lot of the good innovation.

Q: Is the answer regulation?

A: The right way to do regulation is to understand the technology and every technology has pros and cons. You can use a hammer to build a house or to tear down a house. AI has wonderful things that can make us more productive. It can solve scientific problems and can also destroy business models, or the livelihoods of creative people. You’re seeing this happening in crypto. You may see it happen in AI with things like the New York Times lawsuit. We’re seeing it play out in courts and that will take many years and lead to uncertainty. They’ll probably make the correct legal decision but is it the right societal decision?

There’s a role for legislatures and other people to say, ‘here’s the good, here’s the bad, Let’s design policies that maximize the good and minimize the bad.’ We’ve had this reactive approach, and we should take a proactive approach.

One idea, for example, which is in the draft of the [Financial Innovation and Technology for the 21st Century Act] is that tokens have long term, multi-year lockups until they hit certain targets. That’s an example of something which is designed to tamp down speculation.

Today, Dogecoin is perfectly legal under the current thinking of securities laws because there’s no team behind it so no one has asymmetric information. It’s not about speculation and trading. One of the ironies of the current situation is that meme coins are fine. I can create a Chris coin that has absolutely no purpose, no effort behind it, and that’s okay. If I build a game around it, that’s when I get into trouble. That’s an example where you’re taking old laws, and you’re not doing anything to tamp down the speculation.

The policy goal should be to allow a path for productive innovation, but with guardrails to tamp down speculation, make sure that insiders, including VCs, are locked up for a long time, I went to D.C. a lot last year, and they’re always surprised that one of our main things that we push for are long lockups on investors.

Q: My understanding with your first crypto fund is that there were scheduled sell-offs of tokens and that actually did return a lot of money. So you’d actually rather not have that be the case?

A: I’m very constrained on talking about that kind of thing. In the crypto funds, we have held 94% of the tokens.

Q: But that 6% could be very lucrative, right?

A: But we are long term. And in all cases have held for multiple years.

Q: We’re on the cusp of a new wave of a lot of value creation. The internet also created a huge amount of wealth, but it went to a small percentage of people. Do you think the “own” phase of the internet, as you call it, is an antidote to that problem?

A: That’s the idea. The value of Bitcoin, unlike the value of Square Cash and PayPal, has gone to the community. The same has been true of Ethereum.

We’re not building steel mills. We’re building networks. When you go to TikTok, they didn’t pay for the content. Instagram didn’t pay for it. The users created it. Why wouldn’t the users get rewarded? It just makes intuitive sense that the people who do the work should make the money.

It’s a really exciting design space. Maybe my favorite idea in all crypto is called collaborative storytelling. The idea is you have kind of Wikipedia-style communities of fans who get together and create narrative universes, like the new Harry Potter. And then they get rewarded with NFTs and tokens, according to how much they contributed. And then those things can become movies or TV shows, and those fans get rewarded financially. So instead of just debating should Obi-Wan have done that, they can actually decide and then GitHub-style fork it.

And what’s so beautiful about this idea is it’s like the extension of Wikipedia. It’s the next level of doing it for creative things. Two, it provides a new business model for creative people at a time when it’s critical, with generative AI dropping the cost of things like illustrations to zero. And three, it solves a problem for Hollywood. Why is Hollywood creating so many sequels? They’ll say because it costs $300 million to market a movie. What if instead, you have this army of fans who created something who go and evangelize it themselves.

One of the arguments is that it’s also a more democratic, egalitarian distribution of the money, which is why I wanted to write the book, because the gap between the reality and the perception is so wide. People think it’s laser-eyed people getting rich and I understand why. People are on Twitter saying dumb stuff, like ‘have fun staying poor.’ But in fact, this technology can be used in a very democratizing way.

Q: Critics would say that’s a great idea, but why do you need blockchain for that?

A: The alternative architecture would be to have a company behind it. The problem of corporate networks is they start off saying ‘let’s help the community.’ And then over time, they just end up changing the rules, extracting. You’re just not building on solid ground. The idea I just described involves ownership. So the users are owning a piece of the property. And a core part of my argument is that this is the breakthrough of blockchains — they enable credible ownership. The rules are baked in. It can’t be evil.

Q: We already have structures, like fractional ownership and co-ops but that’s complicated. Is buying some tokens just simpler?

A: I have a section in the book on nonprofits. There have been three kinds of grand nonprofits, which I think have worked. But the two big nonprofits that needed a capital investment were Mozilla and OpenAI. And both of them ended up pivoting to being for profit companies, or at least have for-profit subsidiaries that drive their business. They had to because they need billions of dollars to credibly compete on the internet. People tried this with social networking to create consortiums that would store the social graph.

You had all these problems with how do you coordinate it. How do you guarantee it’s going to stay the same way?

Q: We’ve gone from an era where everyone was talking about blockchain, but not actually using it, to an era where people are using it, but not talking about it. Have you noticed that?

A: Yes, it’s definitely a time where the sentiment is more negative. But there’s a bunch of stuff happening in high end fashion and places like Louis Vuitton. Jamie Dimon, despite his negative comments, I think has something like 1,500 people working on blockchain stuff. BlackRock has a big effort. Fidelity has a big effort. Visa has a big effort. PayPal has a big effort. There’s more going on than might be obvious from some of the public discussions. But I have been in tech for a long time, and I think it will swing back. Obviously, I’m a true believer, so we’ll see.

Q: There are big technological hurdles right now that need to be overcome to have a distributed compute network capable of training and running AI models. Is that where you could participate in the AI boom?

A: Fifteen years ago, there was the social, mobile, cloud. Social was the killer app that drove the sale of mobile phones. None of that would have worked if you didn’t have cloud backends to scale these systems up. So they are interrelated and I expect the same thing with VR, self-driving cars, AI, and crypto. What I hope is that they will intersect and reinforce. Like with AI, blockchain is a very natural antidote to things like deep fakes. Imagine a system where a video has a cryptographic signature, and Semafor attests that this is a real video that we put up. And that’s held in a community-owned database that keeps an audit trail, i.e. a blockchain.

Q: Can you have a decentralized web that has consumer friendly offerings?

A: The web browser is a good example. Chrome and Safari are pretty nice but you can switch. We found a good balance between the user and the service.

Q: So web3 won’t just be like a bunch of direct democracy DAOs [Decentralized Autonomous Organizations].

A: You’ll have companies like Coinbase.

Q: But those companies won’t be able to get as big and powerful?

A: Yeah, they’ll be like web hosting companies. They can charge you, and they should make money. But they shouldn’t make network effects money, which has a 100% take rate. A normal business take rate is 10% when you compete on scale, quality, and reliability.

Q: That also goes back to my earlier question about who shares in the gains?

A: We have the web 1.0 example, where it worked pretty well. The World Wide Web is still going strong. That’s thanks to those protocols. If that goes away, or it can become so marginalized that it becomes irrelevant, people stop using email because they’re texting and people stop using the web because it’s been so marginalized. It’s going to happen if we don’t do something. Then all this stuff that had really positive effects will have gone away.

Semafor Logo