Future of Crypto, Blockchain & Web3 w/ Chris Dixon

Major AI companies are closing off access to content and data (Reddit API, Twitter API) because they recognize the value of training data and want to prevent competitors from accessing it.

“what you're seeing now like Reddit just shut down their their API Twitter's doing that you have more and more artists taking content offline because they're just they see it as just fodder for training data”

Blockchains combine the best aspects of open protocols (incentives to build at the edges, no single company capture) and Web 2 platforms (advanced functionality, composability, network governance) to create a third way of building internet services.

“blockchains are they provide a new way to create networks um a new way to create internet services and specifically they they're kind of The Best of Both Worlds of the prior two ways to build networks they combine a lot of the societal benefits of protocol Network so you have an incentive to build on the edges if you build a business or you build an audience as a Creator you own that business you own that audience you have full control of it um the economics are much more favorable to the network participants so there's a whole ser and I have a series of chapters that go [8:16] through the economics and the what's called composability which is the ability to build services on top of other services the way you do with open source software”

Regulatory solutions to platform consolidation are implausible because the regulatory apparatus in both the US and Europe appears to want consolidation rather than preventing it.

“the regulatory the regulatory apparatus in both the US and in Europe seems to actually want that right to want the concentration I mean they certainly are that seems to be their their their they're behaving as if they do even though they don't say that”

Inflation and easy monetary policy have priced young people out of the housing market and intergenerationally transferred wealth from young to old, representing a breakdown of property rights and economic fairness.

“you know there's a there's a there's sort of a structural impediment U and and by the way maybe that means that but bi's idea actually happens in Europe first and look there have been some precursors to it there was the pirate party um you know which actually became quite successful I think in Iceland um and um you know there's there's you know there's more and more um sort of polit I say there the European political landscape is Shifting very quickly right now so um yeah maybe that maybe that uh maybe that happens um and then yeah look I you know the other thing is just um I the thing I I would think about a lot on this is just like you know look a lot of politic like most people's politics for most of recorded history right are you know kind of people's offline lives you know kind of by definition like why do I care about politics well it's like you know I've got my job in the real world I've got my possessions in the real world I've got my money in the real world I've got my f you know my friends are in the real world I've got my you know it's it's like it's it's like you know my ability to speak and politically organized is in the real world and then you know young people now to the extent that they're living more and more their lives online it does make sense that you know sort of Internet politics are going to become more salient and possibly dominant”

Network effects are the primary unit of power on the internet, and the crucial distinction is whether those network effects accrue to a community (as with open protocols like email and the web) or to a single company (as with Web 2 platforms like Facebook and Twitter).

“the idea that sort of the P power on the internet comes through Network effects this is sort of the unit of power you build a service like a social network or the worldwide web or email and these have Network effects and network in in the sort of offline World typically you know Moes as Warren Buffett calls them come through economies of scale and scope and more traditional kind of economic modes on the Internet it's most the mo the main Moe that matters is Network effects”

The 'Read' era (Web 1, 1990s) was dominated by open protocols like HTTP and SMTP where network effects accrued to communities rather than companies, allowing millions of profitable businesses to be built on top without competing for the network's core.

“and so what you do is if you're an entrepreneur or an investor as we have been for the last you know 20 plus years is you try to either build or invest in networks right and and up until blockchains we had two primary ways to build networks we had one which was the the original approach of the web [3:40] and email which is via a protocol right so you a protocol is a set of Standards adopted by a community and the important feature is that it has a network effect email gets more valuable as more people use it but that Network effect doesn't ACR to a single company right it acrs to the community”

In 1996, Netscape had 90% browser share but only 55 million users total (first 55 million people on the internet), with 50% on dialup, making economics nonviable for applications that need millions of users and complex software infrastructure—the economics didn't work yet despite the technology being correct.

“if you go back to 96 I think we had 90% browser share Mark and we had 50 million users so there was like 55 million people on the internet so that's the first thing the market wasn't that big and then the second thing was you know half of those were on dialup so there was limited kinds of applications you could run and [35:40] then the other thing was like building a website was really complex the the software infrastructure the tools were really really primitive um and you know not really designed for anything but serving a page and and so the amount you know the number of Engineers you'd have to feel to get like a mediocre working application deployed to 55 million people was just like it didn't work the [36:10] math didn't work on the economics for most things it worked for a few things like eBay and Amazon and so forth but it certainly would have worked for you know Facebook or uh Twitter or these kinds of things”

Adoption of cryptocurrency (especially stable coins) is higher in countries with volatile currencies (e.g., ~18% of Argentinians own stable coins) than in the US, showing crypto solves real problems for vulnerable populations.

“there's a lot of data that shows for example stable coin adoption is much higher in countries with with um volatile currenc so like Argentina something like last number I saw was 18% of argentinians have own a stable coin um in America 50 million people um have crypto and mostly Bitcoin”

Bitcoin and blockchain enable American economic dynamism and entrepreneurship that is critical to US superpower status, comparable in importance to military dominance and the dollar.

“the core I mean America is very good at entrepreneurship and Innovation and we want to see and I think by the way this benefits VC's interest when I get asked is this in your interest like am I doing this am I arguing for a dynamic entrepreneurial Rich internet out of self-interest to some extent I am I work in the in the Venture Capital business as you guys do like we depend on there being a very Dynamic economy”

Most innovation comes from the edges—from entrepreneurs and creative people at the periphery—not from centrally planned efforts by large corporations or governments, and this principle should guide internet policy design.

“I deeply believe that most Innovation comes from the edges it's not just about sort of you know fighting over how a technology is built right it's fundamentally like do we have an Internet that creates incentives for entrepreneurs for Creative people what is the society we want what are the incentives we want do we want human creativity to still flourish”

Ethereum is undergoing an upgrade (EIP-4844) in three weeks that will provide approximately 100x improvement in L2 (Layer 2) scaling, reducing transaction costs to sub-penny levels and enabling applications currently impossible due to cost.

“there's an ethereum upgrade happening in three weeks 4844 which is going to Pro like basically 100x the Improvement of ethereum l2s which are the main way of scaling ethereum today and you get sort of sub peny transaction costs for most applications so it feels like we're getting very close to something that's very us for a lot that just really widens the aperture of possible applications”

LLMs are currently good enough to defeat password-based identity verification, Turing tests, and email spoofing protections, making current internet identity systems obsolete.

“and everyone's supposed to do their own Turing test on every email they receive to see if it's real which is not going to be sustainable there probably isn't today right I mean these These llms are good enough to fake all of that voices everything right so that whole system is about to break down”

Important policy decisions about AI, crypto, and other emerging technologies should be made through democratic processes with transparent discussion of pros and cons, rather than through courts interpreting 100+ year old laws, which produces outcomes based on judges' interpretation rather than collective democratic will.

“I I think the my argument is the best way to think about policy whether it's AI or crypto is look at the technology expansively look at the pros look at the cons and then through a democratic process ideally you know and this is very idealized because it's not the politics of today but like have a [1:02:44] intelligent discussion discuss the pros and the cons and then come up with a policy solution that maximizes the pros and minimizes the cons that would be the the I think my view the smart way to approach new technologies from a policy perspective”

The distinction between Disneyland (corporate-owned, controlled experience) and cities (open infrastructure enabling entrepreneurship) is a useful metaphor for internet futures: one leads to stale innovation, one enables it.

“you have a great metaphor in the book as you kind of get into you know things like okay how is VR going to work and so forth and you talk about well we can have Disneyland um which is fantastic by the way um or we can have cities uh and maybe talk a little bit about that and you know kind of why like what's good about cities versus Disneyland”

Take rates are the primary mechanism by which platform companies extract monopoly rents, and comparing take rates across different internet architectures reveals the economic advantage of decentralized systems.

“one thing I want to do in the book is you know you hear these words like decentralization I I try to almost avoid that that word in the book and a lot of tech jargon and really dig into the detail so like when we say decentralization that can mean decentralization of software production it can mean decentralization of economics and so when I talk about that's one sort of choice I made in the book was to really sort of say like let's get specific and like how does the money flow and that's the that's the idea with take rates”

Social media platforms are demotion links and making algorithmic changes that keep users in the app longer, not because of user preference but because they need to retain users as organic growth has capped out at ~10% annual growth.

“on social media what's happening is that people are are basically they've capped out their kind of organic growth for the most part they're growing like 10% a year or something to get to to get to juice further growth they need to keep you on the keep you in the app longer and that's why they're demoting links and doing all these other kinds of things”

If companies move up the stack to the service layer after losing OS-layer power, people who like open source should logically support blockchains as the service-layer equivalent.

“but then what happened right is the the corporate the companies responded by saying we're g to move up the stack we're going to move up to the stack to the service layer we're going to monopolize Services data right right and and I would argue like this is one of the things I I wish I could convince the open source folks of if you like open source you should really like blockchains as well because it's essentially open Sour's response to that move up to the service layer”

American technological and cultural dominance depends on three things: military power (declining), the dollar (being weaponized ineffectively), and the internet (currently being forfeited through bad crypto policy), with the internet being the most important soft power for exporting American values.

“if you think about America and why it's important in the world uh you know we often refer to ourselves as you know a superpower the superpower and that that superpower is derived really from you know a few areas one is military power which I I think is kind of [1:04:49] declining for us and that you know the there's disputes spread out all over the world I think as a relative um you know Superior strength as the technology of warfare is changing our lead is not what it once was the second is the dollar itself and we've seen we've tried to use that in lie of military power in the Russia Ukraine conflict and it it's been very ineffective and then the third is the internet where we're exporting our values our culture um and making the rules of the internet and that was [1:05:20] enabled Because the Internet largely got built in the United stes States and you know the government didn't undermine it in fact it um promoted it and that that that remains a huge source of power for the country”

The most compelling pitch for blockchain-based platforms to creators is simply: 'Do you want all the money instead of none of the money?'—this is a more direct and powerful motivation than governance arguments.

“the fact that you can this is also what makes me bullish on these networks succeeding is if you can go to these creative people and say hey do you want all the money instead of none of the money like that's a that's a pretty that's a pretty powerful proposition um”

Current Web 2 has consolidated power in a small number of companies through network effects, and without blockchain-based alternatives, we face a future of five companies controlling the entire internet.

“so in that context so what are blockchains [7:46] blockchains are they basically my core argument in the second half of the book is they provide a new way to create networks um a new way to create internet services and specifically they they're kind of The Best of Both Worlds of the prior two ways to build networks”

Each decrease in transaction cost and increase in blockchain performance unlocks new concentric circles of application categories—high-cost apps like trading/lending are profitable even at $5, but games and social media require sub-penny costs.

“so I sort of think of it as like each each decrease in cost and increase in performance sort of it's like concentric circles get unlocked of different application categories right um and then at the same time we have you know like in our portfolio you know something on the order of many dozens of high quality application um companies and entrepreneurs building things on you know that take advantage of that infrastructure”

Regulatory ambiguity around whether crypto tokens are securities has incentivized bad actors (since they face lawsuit risk similar to their criminal alternatives) while discouraging good entrepreneurs (Stanford graduates) from entering the space, contributing to scandals like FTX.

“that ambiguity um we've argued for a long time for many years has had the effect of encouraging Bad actors because Bad [1:00:40] actors like ambiguity because they otherwise would be stealing money and hacking and doing whatever and so having some risk of a lawsuit is actually appealing to them versus their other career options um and meanwhile good actors if you're a topend student out of Stanford do you want to go into an area with gray with legal gray area or where you just stick to AI or something right and so what we've seen for a long time is that that ambiguity has created a disincentive for good entrepreneurs an incentive for Bad actors and I think has in many ways contributed to a lot of the [1:01:11] things that happened like for example with FTX and other um scandals”

Open source software and regulatory clarity are the best ways for countries to create new innovation hubs (new Silicon Valleys), more than free office space or direct funding.

“people have come to us forever saying they want to create a new Silicon Valley um and and our you know and they they say things like we'll provide free office space we'll provide funding right these are not what people need or want um what they want is regulatory Clarity like that is the way like getting I believe getting ahead on regulation whether it's again AI crypto whatever the emerging area might be is [1:08:28] the best way for a country that you need other things you need good universities and sort of Entrepreneurship and things like that but that that is the smartest way to get ahead of um and create a new Silicon Valley”

AI and blockchain/Web3 together can solve important problems: proving provenance and authenticity of digital content/video, creating decentralized identity systems (solving the broken CAPTCHA problem), designing incentive systems for content creators, and enabling decentralized GPU and data networks.

“one maybe I'll start with the simplest which is proving that something came from an actual human as Ben just alluded to proving that a sort of the provenance of a video or piece of media a very natural way to do that to me to me the only correct way to do to to sort of say this this video is attested to be true [1:10:31] by Mark andreon or this video is attested to be true by the New York Times the natural place to store that attestation is on a blockchain a blockchain is a community-owned database a public resource um and is made to have sort of an you know as a key feature has an immutable audit Trail”

Musicians can sell digital collectibles, backstage passes, and other adjacent value on blockchains while giving away the music itself for free, creating new revenue streams without relying on streaming platforms' restrictive terms.

“and so you see people doing this with music for example we have a couple of startups who allow musicians to sell digital Collectibles and behind backstage passes and other kinds of things that that complement the [1:25:35] music right and they can then get either give the music away for free or be more lock about it or maybe even sell the music whatever and it's an additional Revenue stream”

It's unfortunate that people on the left have decided to hate blockchains, because in doing so they're opting out of participating in a technology that could serve their values rather than contributing to shape it in the direction they want.

“I would also argue this is why it's kind of a shame that it's that if we're going to get into politics here why you know the it's a shame that that a lot of that some folks F on the left have decided they hate blockchains because they're basically opting out of participating in this world um when they could be contributing to it and and sort of nudging it in the direction they want instead they're just sort of saying it's the bad guy”

In the 'attractive cycle' phase of Web 2 platforms, companies actively recruit users and content creators by offering favorable terms, but over time they 'squeeze' these participants as organic growth caps out and they need to maximize revenue extraction.

“I call it the attractive Str cycle when they start off they are trying to um uh recruit people to into the network whether they're users or compliments like content providers but over time they start to squeeze you and you see this like look in your Google Google search [24:12] today or Amazon searches today Google search most of your searches now are Google Google properties the results are Google Properties or sponsored ads very very different than it was 10 years ago”

There is significant pent-up developer energy that has been suppressed by closed platforms over the past decade, and the right blockchain application could release this energy in a wave of innovation similar to early Twitter's developer excitement.

“there's there's a lot of pent up developer energy um and that sort of we've lived in a decade of very closed platforms yeah and people have forgotten what it's like to be able to really build um on a on a on top of an Internet [28:56] service yep and you know we saw that how the excitement level early Twitter was all these developers building things and we saw that level of excitement and I I my belief is that if you look at the 80 years of software history like that's a long ongoing Trend most vividly embodied in open source software y but also with the early internet you know with that 2008 year of social media and I believe it's sort of this pent up energy that with the right blockchain application we can release energy and there will be this huge wave of kind of developer [29:28] excitement around it um I don't know so I think there's sort of a dam that could that's about to break or could break with the right application”

Most people's political interests are grounded in their offline lives (job, possessions, money, community), but young people increasingly live online, meaning their politics will become centered on internet policy issues (free speech, data rights, economic systems), making AI and blockchain policy increasingly politically salient.

“the thing I I would think about a lot on this is just like you know look a lot of politic like most people's politics for most of recorded history right are you know kind of people's offline lives you know kind of by definition like why do I care about politics well it's like you know I've got my job in the real world I've got my possessions in the real world I've got my money in the real world I've got my f you know my friends are in the real world I've got my you know it's it's like you know my ability to speak and politically organized is in the real world and then you know young people now to the extent that they're living more and more their [1:33:47] lives online it does make sense that you know sort of Internet politics are going to become more salient and possibly dominant in how they think about things and and sort of therefore what what sort of Internet policies get put in place um and and and correspondingly Ai and and web 3 policies are actually going to become more Central to their entire concept conception of politics”

Worldcoin (proof-of-human technology using zero-knowledge proofs) cannot operate in the US market and cannot list on US exchanges due to regulatory hostility, forcing it to operate only through foreign exchanges.

“we have an investment in a company called worldcoin which is like a very kind of topical relevant technology because it it's proof of human they can do a zero knowledge proof that proves you're a human being not an AI bot but the one country that they [1:09:31] cannot list on an exchange in is the United States they only list in foreign exchanges and and they're not going to launch the product in the US either”

No major apps will build for Vision Pro because developers have been 'so screwed over' by Apple's 30% take rate and control that they avoid app store dependence.

“no you see this that no one none of the major apps would build in the in uh for um Vision Pro because they're the Spotify of the world because they'd been so screwed over like they're much harsher 30% and just people don't in business or just don't realize like how owner that is on a business”

A stable coin (USDC) on Ethereum provides a digital dollar already, and there is a pending stable coin bill in Congress that would provide regulatory guardrails; the smart policy is to regulate stablecoins rather than reinvent them with a central bank digital currency.

“there's one idea which is having a digital dollar right and I would argue we have a digital dollar today and that's stable coins like usdc on platforms like [49:09] ethereum and that um this the my argument would be the smart way there's a there's a bill a stable coin bill that's pending in Congress that you know that you do need some regulatory guard rails around that and that bill would provide it but don't go and try to rein event what the technology Community has built already um”

Web 1 applications were 'skoric'—they ported old media behaviors (magazines, newspapers, brochures, encyclopedias) onto the internet rather than exploring the internet as a fundamentally new multi-way medium.

“and I and the reason I call it the reier this is not my terminology this is a common way to describe it is that most of the applications built in the 90s were were kind of around democratizing the consumption of information so you know the kind of killer app of the 90s [4:42] was a search engine you type Abraham Lincoln into Google and you get a page about Abraham Lincoln and that was this miraculous thing but a lot of the applications were kind of skoric as we call them which meaning took and this happens with all new forms of media it sort of poured it over old old behaviors from from magazines and brochures and encyclopedias and ported them to the internet the newspapers some of they still are to some extent were formatted exactly like newspapers”

Peter Thiel's framing that 'AI is inherently communist' (centralizing) and 'blockchain is inherently libertarian' (decentralizing) is less politically useful than Dixon's observation that AI is centralizing and blockchains are decentralizing, which is a technological rather than ideological statement.

“my default answer and this is sort of what I argue in the book is is blockchains are decentralizing um AI is centralizing so it's a it's sort of a less political less political spin on the same idea”

In Farcaster (a blockchain-based social network), creators who build an audience own that audience and capture all revenue (subscriptions, sponsorships, ads) directly, unlike Twitter where platforms capture most value.

“we're investors in one called farcaster which is an emerging social network where you know it for the end user it feels like a kind of a tra like a typical social network like Twitter um but a very important difference is if you build an audience there and you want to you know put up advertising or sell them a subscription or do whatever you want you get the money you get all the money right and that's just a dramatic shift that's really like $150 billion dollar a year in current social networking Revenue that you sort of like I think if it like a train track you can reroute it back to the users”

Bitcoin developer David Marcus is building Lightspark, a payment layer on top of Bitcoin where users transact in stable coins but eventually settle in Bitcoin, solving Bitcoin's payment problem while preserving its store-of-value nature.

“people are trying like we're investors in David Marcus's company light spark that's that's building a kind of a layer on top and this is sort of how people are approaching it they're building like sort of these layer two systems that in his case it's a system where you can pay in like stable coins like dollars but then ultimately know that you can settle in Bitcoin”

A strongly centralized government CBDC could enable social credit scores, government control of payments, and weaponization of the banking system (as seen with Canadian truckers), making decentralized blockchains an urgent counterbalance.

“the strong form of the of the Central Bank digital currency is an actual government created digital dollar payment system [49:39] that's tightly controlled and I think that there are people that worry that you know sort of the trucker thing in Canada that you're essentially going to have that as a way you know government to do government assigned credit score government controlled payments it becomes new lever in um the sort of the the operation choke point stuff where people are getting debanked”

Video games (growing from $20B to $180B over internet era) evolved by accepting and leveraging internet's properties (copying, sharing) and monetizing adjacent layers (virtual goods) rather than fighting to control the content itself—a model other media should follow.

“I have a chapter in the book on nfts which I argue so it's a little more detailed but I essentially argue that that if you look at different form forms of media and how they've responded to the internet that video games are by far the most sophisticated um and that they so the internet is very good at the internet likes to copy [1:24:01] things it likes to propagate information and when the video game industry started off on the internet their business model was to sell a video game sort of sell the scarcity and over time they evolve that model and the into the dominant model today which is to give away the video game for free to let people stream it and copy it and let the memes propagate and then to charge for an adjacent layer of value namely virtual Goods right and if you whereas let's contrast that with the music industry they evolved a little bit they now allow [1:24:32] streaming but they still have this sort of scarcity model where if you go and share the song and do other things they'll sue you over it right”

NFTs (non-fungible tokens) are a general-purpose technology for encoding ownership, but have been misunderstood as only about digital avatars ('pfps'), when they can represent any atomic unit of ownership including identities, social networks, collectibles, and more.

“I think that people have overtrained on the idea that nfts are like digital avatars like pfps because that was the I mean one thing people remember is nfts were invented or standardized in 2020 it's literally four years old and even with the kind of downturn there were 8.6 billion in sales of nfts last year so it's actually quite [1:26:05] large for four-year-old technology um but a lot of the original stuff was around these avatars and in fact it's a it's a very general technology it's a an nft is an atomic unit of ownership that you can you can put anything in that bucket”

Media coverage of cryptocurrencies is asymmetric: when prices go up, there are no news stories; when prices go down, there are 'a thousand stories' of Bitcoin's death, creating a distorted negative impression.

“and even more so with crypto I think than with the internet the Press only like there's only ever a news story if the price of uh Bitcoin and ethereum are going down like when they go up like it's not in the news at all but when they're going down there's like a thousand stories”