Writer Talks: Chris Dixon on find out how to reshape the digital panorama

On this version of Author Talks, McKinsey World Publishing’s Raju Narisetti chats with Chris Dixon, normal accomplice at Andreessen Horowitz, about his new e book, Read Write Own: Building the Next Era of the Internet (Random Home, January 2024). Dixon chronicles the evolution of the web and argues {that a} decentralized framework may promote innovation and development. An edited model of the dialog follows.

Why did you select Learn Write Personal because the title to your e book?

The rationale the e book known as Learn Write Personal is as a result of the web is essentially a participatory medium, a community that was designed to be owned and managed by the customers and the communities that construct software program on it and use it.

I see the web as having gone via a number of eras, with every period deepening the extent of participation. Within the first fashionable period, when the web began to go mainstream within the ‘90s, most individuals had been consuming data—you’d go to Google or another search engine, sort one thing, and browse one thing.

Then within the 2000s, the web grew to become far more of a two-way medium, the place there was the rise of social media, the place folks couldn’t solely eat data but additionally publish data. That’s the place democratized publishing and the primary period of consuming data started.

So the primary 12 months you possibly can learn issues; the second 12 months you possibly can additionally learn and write; and the emerging new era of the internet is one wherein customers can go even deeper into taking part within the digital companies they use—by truly proudly owning elements of these companies.

That’s where blockchains come in. Bitcoin was invented in 2008, so it’s been round for 15 years. We’ve seen, all through that interval, plenty of proof that [users are excited about] proudly owning the digital companies they use.

Take bitcoin for instance. Bitcoin is controversial. Some folks adore it. Some folks hate it. However, essentially, bitcoin is a monetary community that’s owned and managed by the customers. This concept has excited lots of people.

Since then, we’ve seen other digital services created with the identical structure—most notably, Ethereum, after which an entire bunch of different platforms that take that type of structure—and develop these concepts of a community that’s user-owned into many different completely different sorts of use instances.

That’s actually what my e book’s about. It goes considerably into the historical past of the folks, the early phases of the web, which I used to be a part of, and I am going deep into the historical past of that. However then, plenty of the e book is about find out how to take all this power folks have for this new structure, for being house owners on the web, and find out how to harness that, channel that power into productive, optimistic use instances that profit the web as an entire.

What did we get mistaken once we first envisioned how the web would work?

One miracle of the web is that it got here out of, in fact, governments and academia and was partly designed to withstand nuclear warfare and for an entire bunch of different causes, so it was in-built a totally decentralized means.

It wasn’t owned by an organization in the way in which TV channels and newspapers had been owned. It was an open, permissionless community: anybody may construct an internet site; anybody may construct a start-up.

And also you actually owned that—you actually owned what you constructed. If you happen to put up an internet site—you realize, you’re Mark Zuckerberg constructing Fb and Fb.com or Jeff Bezos constructing Amazon—you constructed that, and also you owned it. You didn’t pay a tax or charges or ask permission from some middleman. It was a really decentralized community.

This led to a golden interval of innovation via the ‘90s and 2000s, as a result of it supplied incentive for folks to construct on high. What occurred within the 2000s, with the rise of social media after which iPhones, is that plenty of what was the decentralized community grew to become de facto centralized.

In the present day, the highest 5 big tech companies now account for 50 % of the Nasdaq market cap, up from 25 % a decade in the past, in nearly all classes of significant web companies: social media, search, e-commerce. You may have a really excessive focus of market share among the many high two to 5 gamers.

What this implies is that customers and start-ups and inventive folks must undergo these giant companies, these intermediaries, to succeed in their audiences. And these intermediaries have very excessive “take charges.” The take fee is the share of income they take that passes via the community.

They exert their management in quite a lot of other ways. They mainly have found out find out how to arrange the system so that each one the income and management go to those firms.

What I actually concentrate on is how that impacts innovation and the way you now not have the identical type of incentives you had within the ‘90s, the place entrepreneurs and creators may construct companies and go on to their audiences.

Consequently, we’re probably organising a state of affairs the place we considerably stifle innovation and creativity. I’m very professional AI. I feel typically it’ll have many optimistic results on the world, however it’s a very centralizing know-how.

It’s a know-how that rewards firms with deep swimming pools of capital and knowledge and, left unchecked, will proceed to speed up this development towards centralization. Blockchains and tokens and the issues I talk about are a pure counter-lever to that.

They’re decentralizing applied sciences that shift energy again out to the perimeters—energy, management, and cash—and grow to be much more essential as new applied sciences like AI come on-line. Self-driving vehicles and VR [virtual reality] and all the opposite main applied sciences exterior of blockchains which can be on the horizon are typically centralizing applied sciences. It’s essential to have these countervailing forces.

Isn’t it true that plenty of the current headlines round blockchain are about fraud?

There are clearly frauds, like FTX [or Futures Exchange, a defunct cryptocurrency exchange platform]. I do suppose that is true of plenty of emerging technology. You may have dangerous actors who are available and make the most of it. Notably, not one of the points you examine or which have ever occurred have truly been on the blockchain degree. For instance, you hear quite a bit about crypto hacks and issues like this, however Bitcoin and Ethereum and all the main blockchains have by no means been hacked, not as soon as.

There are folks and organizations that use a few of the language and applicable a few of the language of blockchains, like FTX, and commit fraud. With the best policy and regulatory approach, nonetheless, they are often dealt with.

By way of the optimistic use instances, I’ll offer you a few examples. I feel NFTs [nonfungible tokens] are very fascinating. Lots of people heard about them in 2021 or 2022 and consider them as folks speculating, shopping for artwork, and different kinds of issues.

What’s really powerful about NFTs is they’re a means for artistic folks to promote issues and work together straight with their audiences. For instance, let’s say you’re a musician. In the present day, Spotify, by its personal stats, has about eight million musicians on the platform. Of that, solely 14,000 make $50,000 a 12 months or extra, which is roughly the common American wage.

So 99 % plus don’t make a residing via the platform. Why is that? As a result of Spotify takes 30 %. The labels take one other 60 % plus. You ask anybody who does streaming, they get a pair hundred {dollars} for 1,000,000 streams.

That is true all through the entire fashionable web companies. You set stuff on Fb, they usually make some huge cash. They make tens of billions of {dollars} a 12 months on promoting. They’ve what’s known as a 100% take fee. They take all the cash.

TikTok takes all the cash. YouTube is the one giant community, giant web service, that truly shares with the customers. The share is 45 %. For instance, now we have a couple of investments which can be platforms that may let musicians promote digital collectibles, NFTs, on to their audiences. So it’s like promoting album covers, artwork, digital merchandise, and issues like this.

The notable characteristic is that artists take 97.5 % of the cash or one thing near that. Basically, due to the structure of blockchains, the artists actually personal the issues they’re promoting in the identical means they’d personal an offline good. Why do artists, musicians, promote merchandise after they go on tour and make some huge cash that means? It’s as a result of there’s no middleman. In the identical means, after they promote NFTs to their viewers, there’s no middleman.

On this downturn, folks suppose NFTs are useless. There have been many billions of {dollars} of NFTs bought by creators to their audiences. And the overwhelming majority of that cash went on to the creators. So it’s a significant earnings supply for artistic folks on the web at present, though NFTs are at a comparatively early stage of their growth.

I’m typically very excited. I’ve a bunch of sections within the e book about creators—a full part on what I name collaborative storytelling and quite a lot of sections on how you should utilize blockchains to battle deep fakes and different new sorts of points that come up with AI.

One other space is cost. This has been talked about for a very long time. However with the rise of what are known as stablecoins, we’re seeing this actually occur at scale.

In 2022, there have been $6.8 trillion in stablecoin transactions. These transactions occurred on platforms like Ethereum, the place persons are sending {dollars}, USDC [or USD Coin, a digital stablecoin], and different kinds of dollar-backed belongings. A variety of this occurs within the creating world, which is why audiences within the US won’t be as acquainted with it. It’s a very fascinating type of rising use case.

Why can’t regulation assist clear up for the web’s lack of an efficient repute system?

I feel regulation, with respect to the web and know-how, performs a job. The correct regulation begins with a deep understanding of the know-how, what it’s able to, what the nice issues are, what the dangerous issues are, and the way we prescribe a coverage resolution that minimizes the dangerous and permits for the nice.

You could suppose very rigorously in regards to the second-order penalties of those rules. For instance, there’s plenty of discuss AI regulation at present, and there’s discuss having certification methods that make people certify frontier AI models.

The chance with issues like that is in case you begin to add in actually costly, labor-intensive certification processes, you are likely to favor giant tech incumbents. So GDPR [General Data Protection Regulation], which is a European privateness regulatory framework, was an enormous profit to Fb, Google, and the massive tech firms which have 1000’s of individuals whose job is to adjust to that regulation. It’s very, very exhausting for start-ups to adjust to plenty of these items.

Within the US, you usually must take care of 5 to 10 completely different regulatory entities. There are state-level entities. It’s very complicated. So I feel regulation performs a job, but it surely needs to be accomplished in a considerate means that doesn’t stifle start-up innovation.

If you happen to take a look at the final 50 years of tech innovation, a lot of it has come from start-ups. I additionally suppose that in lots of instances, you possibly can clear up know-how issues with different know-how. For instance, one massive query with AI is how creators are compensated when their materials is used as inputs to coaching AI methods.

One fascinating software of blockchains that we see folks constructing are methods the place creators can come collectively and successfully put their content material collectively as a bunch and collectively discount with these platforms. They be certain that if some AI content material distributor needs to make use of their content material, they should adjust to sure sorts of monetary and different utilization phrases.

One other good instance are deep fakes and bots. As AI will get increasingly subtle, it’s going to be very simple for folks to create subtle bots that simulate folks. They may very well be on social media. They may very well be used for phishing. How do you counter that?

Basically, I feel the most effective resolution is to have a solution to cryptographically assure that any person is an precise individual, that they’re who they are saying they’re, {that a} piece of media got here from who it claims it got here from. This can be a very pure software for blockchains, as a result of blockchains present cryptographic proof that one thing is what it says it’s and the whole clear audit path that exhibits you the provenance of an individual or a bit of content material.

My view is mostly that regulation performs a job, in fact, however plenty of issues might be solved via technology and innovation, together with regulation. Regulation must be considered in a balanced means that protects customers but additionally encourages start-up innovation.

Can blockchain higher mediate the financial relationship between AI and content material creators?

I speak in regards to the historical past of search engines like google and yahoo within the e book. Within the ‘90s, primarily what occurred was Google rose up utilizing truthful use exemptions and copyright regulation—which means it might go and index an internet site after which use snippets of it. It might use that content material in its index however then additionally use snippets on the search outcomes. Mainly, it stated, “This was truthful use as a result of we’re solely utilizing a snippet. We’re not utilizing the complete content material.” At first, some folks protested. However, for essentially the most half, it appeared high-quality. That they had a system the place you possibly can decide out of it.

In the present day, websites like Google have grow to be so highly effective that it’s important to be listed in Google, and you’ll’t actually decide out, virtually since you lose all your enterprise. They mainly set all of the phrases. You go to a search engine, and, no less than at present, in case you give them your content material, you’ve some promise that you simply’ll get site visitors again.

That’s kind of the covenant, as I name it within the e book, that you’ve got between content material and distribution on the web. For instance, I’m the New York Occasions. I let Google take my content material and present snippets of it, however they’ll ship me some quantity of site visitors again. Everybody who runs a content material web site is used to that.

Within the AI world, it’s very possible that they received’t ship site visitors again; they’ll simply offer you, the person, their reply. You’ll simply say, “What’s occurring in XYZ?” And it’ll simply let you know. There received’t be hyperlinks, and also you received’t return [anywhere]. It’ll use your content material to make these methods smarter, however they received’t ship site visitors again.

Now one end result is folks put up paywalls, which is already occurring, by the way in which. Which means folks block their content material from being searched, and artists take down their content material, and all these different kinds of issues, as a result of they don’t need it coaching generative AI algorithms.

I feel this might be a disgrace. You’re going to find yourself with a siloed web. That’s the place we’re headed now if we don’t suppose this via. One other end result is—and we’re already seeing this additionally—AI suppliers going to content material boards for content material as a result of the content material’s getting blocked off by these companies.

So that they’re going off and paying people who find themselves offshore in an workplace someplace saying, “Hey, give me a bunch of images that seem like this, and feed it into the algorithm.” I feel each of these outcomes, a siloed web and offshore content material suppliers for AI, are each miserable outcomes.

A a lot better end result could be to assemble a brand new covenant, a brand new relationship between content material suppliers, AI, distribution, and search engines like google and yahoo. What’s the pure means to try this? Get giant teams of individuals to coordinate on the web and create networks. That’s what blockchains are. It’s the know-how for creating networks which can be open and ruled by the individuals.

There are a bunch of entrepreneurs engaged on issues like this, the place you mainly consider them as financial networks, the place content material suppliers can be part of, AI suppliers can be part of. They will come and agree on phrases and collectively resolve on what the optimum relationship is for everyone to have the most effective end result. We didn’t actually have methods like this on the web earlier than, the place you possibly can come collectively and have giant teams of individuals collectively resolve on what sort of financial relationship they’d have.

You possibly can construct an internet site within the previous mannequin. However then you’ve this web site within the center that’s an middleman, that has a special set of incentives and finally ends up slanting the taking part in discipline to its personal benefit. Blockchains are a good way to do that in a really impartial means. One of many key options of a blockchain is which you can make robust commitments about future habits. That is actually, in some methods, the distinctive technological functionality of a blockchain.

So the blockchain can say, “I’m going to mediate this relationship. And there will probably be no charges. I, the blockchain, am not going to cost any of the individuals within the community any charges.” The blockchain could make that dedication and promise it in code in a means that may by no means change. That is what I name: “Can’t be evil, not, don’t be evil.”

Which suggests folks can be part of and commit their financial future to a system constructed on a blockchain. They will belief that it’ll keep the identical sooner or later in a means they’ll’t in a standard web service structure.

Why do you consider blockchain received’t go the way in which of the web?

I began my profession within the ‘90s, and one of many most important causes I selected the web for my profession is that I purchased into the beliefs of the web. It’s truthful to say that plenty of the early beliefs of the web that I purchased into associated to the concept it might be this open system that may democratize data and media.

I do suppose it took a mistaken flip in some unspecified time in the future. That’s why, within the e book, within the first two chapters, I attempted to go fairly deep into diagnosing what occurred, after which into diagnosing what’s mistaken with the brand new wave of networks—what I name company networks—why it led to “disconsolidation.”

Issues went off the rails, considerably, with the event of the web. The 1st step is knowing precisely why they went off the rails, which I attempt to do within the e book. Step two is saying, “What are the methods to place it again on the rails?”

There are solely two sorts of credible approaches to that; one is regulatory approaches. I haven’t actually seen vital momentum there for something significant that may re-decentralize the web, however that’s, in concept, one solution to do it.

The opposite is technological approaches. I feel blockchains are, by far and albeit, the one severe contender for doing that. I spent plenty of the e book explaining precisely why that’s, particularly across the know-how. There’s a essential distinction between blockchains and, as I name them, company networks, like Twitter. Blockchains are open supply. And there merely is not any firm and no proprietor behind them.

Earlier than you become involved in one thing like Ethereum or any of those different blockchains, you possibly can go take a look at the code. It’s all open. It’s all community-owned.

There’s no must depend on trusting another individual. You’ll be able to go and consider Linux or rent somebody in case you’re not technical your self, but it surely’s all open. And that’s a vital distinction. By way of the worldwide questions, as China has proven, for instance, nations can just about overwhelm any know-how in the event that they need to. They will block something they need ultimately, and a few nations have chosen to.

I feel the most effective antidote to that is to create world web companies so compelling that individuals really feel like the professionals outweigh the cons, they usually need to take part in them. However I do suppose there will probably be some danger within the web turning into extra fragmented over time. However kind of like a worldwide monetary system, finally, you need to make it in order that there’s a lot upside to be included within the web economic system and the web—the data and cash and all the opposite issues that one would profit from taking part in—that it’s extra enticing than blocking it.

Watch the complete interview