Bankless - Cloudflare Needs 100M TPS from Crypto to Fix the Internet | CEO Matthew Prince

Episode Date: May 25, 2026

The internet is about to be flooded by machines, and Matthew Prince thinks the old business model is not ready. The Cloudflare co-founder and CEO joins Bankless to explain why AI agent traffic could overtake human traffic by 2027, how crawlers are changing the economics of content, why ads and subscriptions may not survive in their current form, and how x402, stablecoins, and pay-per-crawl could create a new payment layer for the web. --- 📣SPOTIFY PREMIUM RSS FEED | USE CODE: SPOTIFY24 https://bankless.cc/spotify-premium --- BANKLESS SPONSOR TOOLS: 🔮POLYMARKET | #1 PREDICTION MARKET https://bankless.cc/polymarket-podcast 🟦 COINBASE ONE | MEMBER MONTH https://bankless.cc/coinbase-one 🧭OKX | TRADE, EARN, PAY to OKX | 120M+ USERS WORLDWIDE https://app.okx.com/join/USBANKLESS 🦊 METAMASK | DOWNLOAD NOW https://go.metamask.io/BL-Pod-Download 🌐BRIX | EMERGING MARKET YIELD https://bankless.cc/brix 🎯THE DEFI REPORT | ONCHAIN INSIGHTS https://thedefireport.io/bankless --- TIMESTAMPS 0:00 Intro 1:18 When AI Bot Traffic Overtakes Humans 4:19 What Cloudflare Actually Does 10:40 Life on the Internet’s “Ice Wall” 13:33 The Internet’s Business Model Problem 17:43 Google, Traffic, and the Attention Economy 24:56 AI Agents and the End of Ads 28:47 Why Faster Answers Can Still Hurt Creators 33:26 AI Companies as the Next YouTube or Netflix 38:18 Creating Scarcity for Content Markets 41:52 Content Independence Day 45:24 Should AI Agents Count as Human Users? 51:37 The Risks of Subscription-Only AI 57:24 A Golden Age of Content Creation? 1:01:46 Pay-Per-Crawl and x402 1:04:34 Why Stablecoins Matter 1:07:56 The Cloudflare Stablecoin Question 1:14:18 The Final Form of the Internet 1:17:40 The Park Record and Local Journalism 1:22:16 The Centralization Risk 1:27:12 Advice for Content Creators --- RESOURCES Matthew Prince https://x.com/eastdakota --- Not financial or tax advice. See our investment disclosures here: https://www.bankless.com/disclosures

The internet for a long time is growing like crazy. And then it kind of stalled out for the last little bit. And the number of like new websites that were being created, a number of new domains kind of plateaued from about 2012, 2011, 2012 until actually quite recently. What's changed is I think that because a lot of these tools have allowed more people to be creators, to be coders, to be developers,

actually for the first time in really the last 15 years, the number of websites that are being created, the number of new things that are happening is actually growing again and growing at rates that look like what the early 2000s growth rates of the internet look like. So I actually think that's kind of optimistic that there's more things that are being created that are out there and are coming online. Bankless Nation, very excited to introduce you to Matthew Prince. He's the co-founder and CEO of Cloudflare. That's a $70 billion company that sits between the Internet and roughly 20% of the world's websites run on top of Cloudflare. I think he's one of the most thoughtful voices on the question of how is the business model

of the internet changing. He's also recently joined X402, got some stable coin stuff to talk about. Matthew, there's so much to discuss today. Welcome to Bankless. Thank you for having me. So we've seen estimates that at a certain year, AI bot traffic is going to exceed all human traffic on the internet. I don't know if that's 2027, 28, 2029, or maybe even this year. when do you think that happens and is the internet ready for it? Yeah, you know, I think that based on what we've seen, we were saying it was 2027. We're now saying it's the first half of 2027 that that will cross.

It's just been, it's been extraordinary to watch the growth of agentic traffic, bot traffic over the last little bit. And for a long time, bots were bad. But you increasingly are saying these agents, which are actually helping humans trying to do things which are really, which are really useful, just driving an enormous amount of traffic. And I think any of us who have used any of the AI tools can see how that happens. If I'm personally going to go, like, think about buying a new digital camera,

I might like go to a site or two, you know, a few pages that are out there. And then I'll make a decision. Someone who's maybe more, more thorough than I might go to five, you know, sites and compare reviews. The agents are, you can just watch it as you ask, you know, a chat GPT or a grok or a or a Claude to do work on your behalf. they're taking, you know, oftentimes a thousand times more in-depth research. And that means there's just a lot more traffic it is that is hitting the internet. In terms of whether the internet's ready, I think that's a little bit of an open question.

I think one of the challenges, you know, from a technical perspective, I think we can manage that. And we can scale through it. We've seen explosive growth in internet traffic before. I harken back to like April of 2020 when COVID hit and all of a sudden overnight. internet traffic doubled over the course of two weeks. And it was extraordinary how much the system still held together. And while it created challenges, it was able to scale and we were able to keep the internet functioning

and help us get through that time. I think this is the same sort of moment where you're going to see a lot more traffic. So I think technically will be okay. I think the big question, though, is what's the business model going to be and who's going to pay for it? Because there's a lot of servers, there's a lot of infrastructure,

there's a lot of stuff which is behind all of that. And if the business model of the internet for the last, you know, 30 years has been ads and subscriptions, promise, agents don't click on ads. And, you know, buying one subscription and then having agents be able to basically pick, pick all of the content back up from that, that's not going to help make sure that the people who are creating that content get compensated for it, the people who are providing the infrastructure to power that content, get actually compensated for the work that they're doing. And so we have to figure out, some new business model.

And that's a lot of what I'm spending my time thinking about these days. I can't wait to discuss that with you. I think that's going to be the bulk of the episode. Before we get there into the new business model that might be required, why ads are breaking down, why subscriptions are breaking down. Talk about Cloudflare for just a second. Because I think our audience may not know exactly what Cloudflare does, even though they're probably a daily active user of Cloudflare.

because if Cloudflare turned off tomorrow big chunks of the web that would just go completely dark. And you guys have been part of the story of how the internet has scaled over the last decade or so. One way I think of Cloudflare is sort of like, you know, that ice wall from Game of Thrones, where it's just like this giant ice wall that kept the white walkers out

and all of the bad bots and, you know, Russia and North Korea and just malicious things. You guys play a role in that. I know that's not all you do. do, though. Talk about how would you explain Cloudflare to, I guess, you know, a normal tech savvy audience member? So, you know, I think it's kind of like, you know, any, any superhero needs an origin story. And so our origin story started pretty simply, which was we saw that all of the world of software was going to the cloud. We saw that all the world of like servers were going to cloud.

And so it seemed obvious to us that sort of networking and the security appliances that were out there before the firewalls of the world, we can go to the cloud as well. And so Cloudflush started with a really simple idea, which is how could you take a firewall and put it in the cloud? And one of the challenges we had was we knew that in order to be successful, we had to then someday sell to big banks and governments and healthcare organizations and things like that. But we were tiny. And so the challenge was we didn't have a big network. We didn't have the data to really stop the most sophisticated threats at the time. And so one of the things that we did was we made a stripped-down version of the service

available to everyone in order to basically help us bootstrap getting the network built and also to help get the data to be able to identify what the threats and challenges were. And the combination of decisions really caused us to then have what were just a series of almost endless problems. The next thing we knew we had every human rights organization in the world that signed up, which meant that then every sort of authoritarian regime in the world was attacking us trying to knock them offline. And so we had to build, you know, not just a firewall, but now we had to do DDoS mitigation. We had hacker kids that were, you know, trying to, you know,

literally steal our domain. So we had to build a registrar. And we had to our own employees that were getting, you know, attacked as they would go online. So we had to build our own VPN service that was there. We even had to build our own developer platform in order to be able to keep up with all to feature requests and be able to build things ourselves. And so the story of Cloudflare has really been, like, start with this simple idea, how do you put a firewall in the cloud, make it available to as broad a set of people as possible, have a series of problems that that then creates and then solve those problems for ourselves. And then it turns out, you know, if you're building, you know, whether it's an AI company today, where 80% of the major AI companies and labs are our customers of

ours, you know, in the crypto space, almost 100% of the crypto space uses Cloudflare today. And then a Across Fortune 500 and Global 2000, you know, we're seeing, you know, closing in 30, 40, 50 percent of those companies that are relying on us as well. And so today, you know, what we really think of ourselves as is what's the next generation cloud? If you could start over and you didn't have sort of all the assumptions that an AWS or a Google or a Microsoft had, if you could sort of think about instead of serving, you know, Ashburn, Virginia or one particular region, if you could serve, if you're the region that you served was the earth. And if you could take everything that we've learned from powering 20% plus of the internet and make it available, like, I think that's

what we, that's what we're doing at Cloudflare. And, and I, and I think our audience is, is really, how do we make the internet better? Our mission is to help build a better internet. And the reason, you know, I'm still so excited to, to come to work every day is, I think that's a really important mission. I can't think of anything more important I could be working on. And especially right now, when so much is changing about the internet. I really love the fact that we have a seat of the table trying to figure out how do we make sure that works for everyone. If you could get a 3% top-off on your entire portfolio today,

would you take it? Because Coinbase is basically offering exactly that. If you've been looking for an excuse to consolidate your assets, this is it. Until May 31st, Coinbase is giving a straight 3% Bitcoin boost on any crypto or cash deposits all month. Coinbase 1 is built to be useful day-to-day. You've got zero trading fees on thousands of crypto assets,

around 3.5% APY on USC, boosted staking and lending rewards, and up to 4% Bitcoin back with a Coinbase 1 card. Plus, you can still claim 20% off the first year of Coinbase 1 annual plans and a $50 Bitcoin bonus when you spend $100 in a new Coinbase 1 card in the first 30 days through May 31st. Get your 3% boost at Coinbase.com slash bankless. Visit coinbase.com slash bankless. That's coinbase.com slash bankless to claim your Bitcoin bonus. Offers valid until May 31st. Terms apply, Coinbase 1 card is offered through Coinbase Inc and Cardless Inc. card issued by First Electronic Bank. Bitcoin back rates are based on cardholder assets on Coinbase.

Trading is changing. Not gradually right now. OKX just launched trading buzz directly inside the OKX app. Grid trading, DCA, arbitrage. You set your strategy once and it executes around the clock. No staring at charts all day. No manual entries. No missing moves while you're asleep.

And for the first time, automated trading actually feels simple. But OKX is thinking bigger than just trading. They also launched the agent payments protocol, an open standard that lets AI agents execute full commercial transactions on chain. The Ethereum Foundation, Unoswap, and AWS are already building on it. And now it's live inside the United States. And new users who deposit in trade can get up to $500 in Bitcoin through the bankless link. The link is in the show notes to learn more, not investment advice, not available in New York or Texas.

$100,000 is up for grabs. Metamask and Ondo are giving it away in the Ondo trading competition. Five weeks, three tiers. Whoever earns the best percentage return in each tier wins the biggest cut. Pick from 260 ondo tokenized stocks, funds, and commodities to swap inside Metamask and let your portfolio do the talking. If you think you can outtrade everyone else, this is your chance to prove it from now until June 18th. Click the link in the description below and opt in on Metamask mobile. Trade RWA's on mobile or desktop to compete. Not investment advice.

Can we illuminate what the internet is like to the people who don't, they don't really know what's on the, to use Ryan's metaphor, what's on the other side of the ice wall? There's a lot of just naive users of the internet. And I say that as a good thing, like Cloudflare and like the protective services that you guys provide for the internet means that I don't have to worry about the white walkers on the other side of the wall.

You do, and you see that every single day. Can you maybe just illuminate just how hostile of a place the internet is for the people who don't really have exposure to that? Yeah, what's it like to be on the wall? Yeah, it's like, yeah. I have served this many days, right? My watch is done. Yeah, your watch is not over.

My watch isn't anywhere close to it. I mean, it's pretty amazing to see, you know, the things that we have to deal with. And again, like, I would never imagine these things. When, you know, in February or in February 2020, or 2023, when the, when the Ukraine war broke out, like we had provided our services at no cost to Ukraine. and all of a sudden the entire Russian government was trying to figure out how to take us down in one way or another.

They weren't successful. We helped make sure that Ukraine stayed online and my reward was I'm now personally sanctioned by the Russian government. So I don't get to go to Moscow anytime soon. You know, when right now the conflict is going on in the Middle East with Iran, watching the Iranian hackers

who are very sophisticated, try and find any possible vulnerability and we can track them as they, as they do that is extraordinary. And so, you know, there are in any given second, you know, literally tens of millions of attacks which are hitting our network. And what we act as is that immune system that helps make sure that, you know, you can just put Cloudflare in front of it and not have to think about any of those other challenges. And I think

that's been a big piece of what has really driven our success. And that's in the sort of cyber threat side. I think that the thing which I find interesting is that even as we change and bring in things like agents, like that may not seem like a cyber threat, but it's really a threat to the business model of people who are doing content creation and other things. And so we get pulled into kind of those other other conversations as well because we're sitting in that, in that position. And again, I think we have a responsibility not just to ourselves, but not just even to our customers, but to the internet as a whole because, you know, any given month, six billion people, basically everyone online passes through our network at one point or another. And when we're doing our job right,

You don't see a capture. You don't have to identify bicycles. You don't have to do anything. We just make sure that the white walkers stay out and that the rest of the normal people are able to transact and do so safely. So as the ice wall on the internet, you guys have a unique advantage point. It sits kind of higher above, I guess, the rest of the kingdom.

And maybe let's get back to that thread of the business model because you said your role, your mission really at Cloudflare is to build a better internet. And there's almost an assessment of like, how good is the internet today? Do we have a good internet? I mean, I remember the early dream of the internet in the 1990s, set of open communication protocols. It felt very decentralized. It felt very organic.

It felt very bottom up. And then we hit sort of this period where it turned into kind of a Disneyland, if you will, where everything felt synthetic, you know, big corporations kind of were the, you know, it wasn't they open internet. so much as it was an app on your phone that Google or, you know, some big tech company harnessed. And now we have a world where it seems like these AI agents are starting to break some of the original business model that formed the basis of this. Talk about that. How has, how has the internet evolved? What's the ecosystem like? Who pays for it? Yeah. So, so, so big question

and at risk of falling back into my own or my old sort of professorial ways. Let me kind of take that in various bits. So at the very beginning, I remember we were in a venture capital meeting at one point, and they were like, who's your competition? And I said Facebook, and the VCs all rolled their eyes. But actually there's something that was really true to that. This was like, what, 2009 or so when you guys were? 2020, 2011, 2012, early days.

Because I think that what Facebook, a lot of the value that Facebook provided is it provided all of this safety within kind of their Disneyland. But in exchange, you had to give up a lot of what was unique about whatever it was that you were doing online. And I think if Kleffler hadn't come along,

actually Facebook would be a lot larger in terms of how much was there. Because in the early days of Kleffler, in 2009, 10, 11, 12, there's a period of time where people are like, everything's going to end up on Facebook behind that walled garden, and there's not going to be as much

of kind of the independent internet anymore. I think that we are to Facebook as like Shopify as to Amazon. We provide kind of what the base functionality of keeping you safe, making sure that you can do whatever you want, but then it's up to you in order to really innovate and create and have your own, your own experience. And so you said we sit above. I always think we kind of sit below. We're kind of that deep infrastructure layer that's down, down, you know, buried somewhere, which again, if we're doing our job right, you shouldn't even have to think about it. And you can be as creative as possible. If you then, the look, kind of over the last period of time

that Cloudflare has been around, the internet for a long time is growing like crazy. And then it kind of stalled out for the last little bit. And the number of like new websites that were being created, a number of new domains,

kind of plateaued from about 2012, 2011, 2012 until actually quite recently. What's changed is, I think that because a lot of these tools have allowed more people to be creators, to be, you know, coders, to be developers, actually for the first time in really the last 15 years,

the number of websites that are being created, the number of new things that are happening, is actually growing again and growing at rates that look and resemble like what the early 2000s growth rates of the internet look like. So I actually think that's kind of optimistic, that there's more things that are being created that are out there and are coming online.

I think that your kind of Disneyland analogy is the right one. Because I think those of us who have, But I mean, I've been on the internet for, you know, since the early 90s. And I remember the quirky, weird, people were just putting stuff up for the sake of putting stuff up. And what feels like it changed along the way is that a business model really did develop around it. And the core business model of the internet really was advertising.

And Google was the company that did more to push that than anyone else. You know, we all think of Google as a search engine. But beyond search, they were the ones that. that provided, you know, with AdSense and double-click and all of those things, they provided the infrastructure that actually said, I'm going to take traffic and turn it into revenue. And so the game became, because of Google, how do I get as much traffic as possible? Now, people have taken that a step further.

So Google kind of turns into Facebook, which turns into TikTok, and we're increasingly playing this game where the goal is, how do I get as much traffic as possible? because every bit of traffic I can then put an ad in front of someone and there's some percentage of those ads that I'm going to get paid for. So the more traffic I have, essentially the more that I'm going to get. The problem is traffic has always been a terrible proxy for value. Like if you think about things in life that, you know, you might generate traffic.

Like take it off the internet for a second. You know, like there's a car accident and people rubberneck to look at the car accident and it creates a traffic jam, that's not a good thing. That's not appealing to our sort of better angels. That's kind of the morbid sick part of humanity where we sort of stop to look at the car accidents that are happening.

And so traffic has actually always been a somewhat bad proxy for actually the value that's there. And in fact, a lot of the media space for a long time, I remember sitting with a senior, a founder of a very prominent media firm and talking to this person, and she was bragging about how, they generated content articles as, as inexpensively as possible, and that they would then

then A, B, test headlines to see which ones provoke the most rage, because the more rage they could provoke, the more likely you were to click on it, and the more likely they were able to serve you an ad. Oh, my God. That's just a horrible, like, that's, that's, that's, that, that isn't making society better in any, any possible way. And so what I think that, that feeling that a lot of us who've been on the internet for a long time have had is that the traffic-based, attention-based economy that the internet has had,

that that actually has kind of led us to both kind of a Disney version, but with kind of rage bait attached to it, right? So it's sort of like an unfriendly Disney version. Rage Disney is a terrible. And so I actually make an argument that a lot of what's wrong with the world today, a lot of why you've had this massive rise in populism, a lot of the reason why the world feels so divided is because that's been the business model of media. That's been the business model of the internet is how do we actually provoke rage and cause people to click on things so that we can serve them ads? And so I think that as we think about what a new business model of the internet

should be, we should actually try and say, how can we not reward like just traffic, but how can we reward actually content creation or even more important than that, you know, like knowledge creation? And I think that's, you know, that's a lot of what, I think, as we think about what that future looks like, we should be trying to figure out, you know, the people who actually really create knowledge, those are the ones who should get compensated, not the ones who can write the most, you know, incendiary headline. Well, let's think about what that future looks like. But before we get to kind of bright futures and rosy futures, let's talk about blame for a second.

Because it's always, you know, we can assign some blame here. And I'm not sure if you blame individual actors or the system itself, maybe it's some, more the latter. I think I read a quote from you. You talked about everything wrong with the world today is Google's fault. And I feel like you're being somewhat tongue in cheek with that. A little bit.

Google has been a massive force of good in the world. Like you just think about the fact that we can all carry around in our pocket access to all the world's information. Had Google not existed or something like Google hadn't existed, like the internet wouldn't be nearly as big, nearly as robust. It wouldn't be nearly

is accessible. Like a lot of the business models of, you know, the old, like, you know, those are you old enough to remember, like CD-ROMs, it was how do I sell you this really expensive thing to access knowledge? Google actually figured out a business model that made it so that knowledge was available and it made sense to make knowledge available for as many people as possible. And I think that's, that's incredibly good. But at the same time, again, the core problem was what they were rewarding was traffic.

And traffic is just a poor proxy for very. value or for quality. And so what I think it would be better as we think about the future is, how do, yes, we create incentives for things to be as open as possible, but also how do we create incentives for them and then the people who are, who are, you know, creating that content to create content, which is actually furthering human knowledge as opposed to just, you know, provoking outrage. Yeah, I feel what you're saying. I mean, it's almost the problem, the classic problem of over optimization, right? If you over optimize in just basically any direction, you get a set of outcomes that feels almost dystopian, right? And so we've been pursuing this ad-based, eyeballs-based

business model for like two decades. And so this is the state. Almost basically 30 years. Google will be 28 years old in September. God, I can't believe it. I still think of Google as like kind of a new company, but it's really not, right? It's been operating this business model for 30 years. So I mean, and maybe no one's to blame. I know Mark Andreessen famously talks about, you know, missing payments. was the original sin of the internet. I think Ben Thompson is... Not including encryption also.

Yeah. Oh, my God. Oh, you're speaking our language, encryption. Yeah, we forgot that one too. And Ben Thompson says, well, look, it was the only way that the model could have grown, right? It was an inevitability. There's no one really to blame.

There's never been a micropayment system that worked. That's right. At scale other than advertising. And advertising is pretty amazing because, like, I'm a pretty, like, Target rich advertiser. If I see an ad, you know, it's actually super, super, super valuable for that to be shown to me. Whereas some, you know, kid in Botswana, you know, they're maybe getting, you know, a tiny fraction showing an ad to them, but it's still a tiny fraction. And the marginal cost of serving that kid in Botswana is so low that, again, the business model of Google,

and the reason why, again, Google gets a huge amount of credit for this, the business model allowed that kid and botswana to have access to the same information that I have access to. And that's that's a sort of per se good. And I think that as we think about this, and one of the problems, by the way, with both micropayments is, yeah, for me, for you guys, not such a big deal. If it's a fraction of a penny every time I access to website, kitten by Botswana actually kind of is a big deal. And so I think there's always going to be tradeoffs as we think about these things.

And we may solve one set of problems to create, you know, a very different set of problems going forward. And again, I think we should just be cognizant of that and I'm certain that, you know, that, you know, 30 years from now someone will be getting like, oh, all the stuff that, you know, Cloudflare did really screwed up the internet.

So, yeah. We do seem to be approaching toward like the end game of the ads based business model of the internet. It's really, it's mature. It's kind of done, it's thing. And this is happening at the same time that now we have agents. Can you talk about

the, just the addition of the agents variable into just the internet itself, the economics of the internet, and how that's impacting things. You called it. I think, Matthew, you used these terms, which I really appreciated. You said the AI agents are almost strip mining the web. That's right. I think that's, so first of all, you're absolutely right. The ads economy has been declining. And that's been true because of, you know, people got, you know, privacy blockers, ad blockers that were, that were out there. You know, that was changing the equation. I don't think that the internet has always explained well enough that the, the,

that the tradeoff was like we'll take the, you know, we'll take the energy and the infrastructure to give you this content. But in exchange, we get to show you ads, right? And that was, that was the quid pro quo that was there. And again, I think that I'm sitting here in Europe, like a lot of people just don't get that. Like, they don't understand that that was the trade. It was just like, well, I should be able to get this stuff for free, but you can't show me any ads that are out there. That doesn't work. And it's, and it's been declining. But the change that's been so dramatic has really been in just the last, in the last two years. And that change has really been driven by OAA AI. You sort of saw this steady decline in, in sort of ad revenue, in what are called

CPMs, which is sort of the rate that you get for an advertisement that was out there. Those were sort of declining over, over time. But in the last little bit, what's happened is you've just seen a, just sharp, almost step function drop in the, in the value that people are being able to get from, from, from, these ads that they're delivering. And what's changed is, like, if you just look at Google, like if you type something into Google, like when did bank lists start as a podcast? In the past, Google would show you 10 blue links,

and you'd have to click on one of them in order to go find the answer to your question. What happens now is Google presents an AI overview at the top of that page. And it's going to give you that answer. And what's increasingly happens, those answers are getting better and better. And the way Google gets them,

they go out and they spider the web, they effectively strip mine the internet, they pull it back, and then they give you the answer. Now, from a user's perspective, from an end user's perspective, that's great because I don't want to have to go spulunking through the web. Like a search engine, the search isn't over when I get the 10 blue links. That's just the beginning of the search. That's a treasure map.

They don't have to go click through in order to find what I'm looking for. What Google is turning into and what chat GPT and Claude and XAI, all of these companies are, they're not search engines. their answer engines. And again, from a customer's perspective, we've all had that experience of like, wow, this has saved me a huge amount of time and given me answers to things that maybe didn't even exist out there before in a way that a search engine never could. But if you're a publisher, if you're a content creator, if people aren't clicking on the links, if they're not going to the sites, then no one is seeing the ads and there's nothing which is compensating you. And so all of

these different companies, all of these answer engines are taking content, but they're not sending traffic back anymore. And what that then means is that because they're not sending human traffic, eyeball traffic back, that there's no one to show an ad to. And since agents don't click on ads, then all of these companies are just massively losing the revenue that they're getting from their ad-based business models. And that's going to start to hurt. It's already hurting content creation, which is out there, either by content creators putting walls up where neither you nor your agents can get to that content, or in many cases, you know, publication shuddering because they can't actually have a business model to support themselves anymore. Can we just double down

on why this is bad? Because me as a user, I get my questions answered faster. We get more truth faster. Things are more free for more people. Now I understand that, you know, independent publications might have a hard time. Maybe, maybe they have to shudder. But maybe that's just an evolution of the internet. Now there's more information. There's more free for more people. Can you just articulate just the downsides of what happens if we let this happen

kind of unfettered? Yeah, so first of all, like I'm not sure it means it's more free for more people. Most of these AI companies, their business model is actually, again, there's a subscription that you pay, and that gives you a certain amount of access to it. And again, for us, that's fine.

Like, I can pay 20 bucks a month or 100 bucks a month or, you know, $1,000 a month if I want. But I think actually the people who are in the global south, people who don't have the ability to pay for those subscriptions, I actually think that the internet's getting smaller for them, not bigger. And I don't think these AI tools are actually solving a lot of their problems. I think it's actually, they feel like those things that, you know,

the wealthy people have on the other side of the wall, but they don't. I think that's a problem that we're going to have to wrestle with. And almost a solution that we come up with is not necessarily going to solve. I think back to the other question, though, is, like, you know, you said maybe the independent publications are going to have to shutter. I'm not sure that every publication isn't going to have to shutter. Like, if over time, and we've watched where the difficulty of getting traffic back to your site, as compared to sort of the Google of old, you know, today, it's 20 times harder with Google than it used to be. But that's the good news. With Open AI, it's 1,500 times harder. With Anthropic,

It's 60,000 times harder. And so if you're doing the work, if you're a journalist that's doing the work of creating things or you're an academic that's doing the work of creating things, and these AI companies are coming in, they're taking all the work that you do, but they're not giving you anything back. They're not compensating you in any way.

Like over time, I don't think it's just the independent publications. I think that anyone who's doing content creation has to be crazy to continue to do content creation. Why would you do it if there's no business model behind it? Like, you know, yeah, sure, maybe at the edges they're going to people who do it just because they love it. But that's a really small universe. People have to eat.

And so journalists have to get paid somehow. And if the way that we're going to be consuming information is through these AI systems, that's going to change the business model. What I think people don't totally understand is, like, the internet has gone through a number of different platform shifts. Like we went from browsing the web on our laptop or desktop on a browser to then social and basically consuming information largely through social.

And then to mobile. and through each of those, the ad-based business model has stayed the same. AI is yet another platform shift. The frame that you're going to view content through is going to be through your helpful AI agent. But what's different about it is for the first time, it's going to pull the information back to you

as opposed to pushing you to the information. And as it pulls that information back to you, it might be that the AI companies are going to be able to create great ad businesses themselves, but the people who are the original, original content creators, the people who are doing the work, going out there, you know, reporting on the news, doing the basic research and science, all of those don't have a business if they're not selling ads and they're not selling subscriptions. And so in this new platform that is AI, I think that

no matter what, people are going to be reading from the AI system, which means they're not going to go back to the original source, which means that it's much harder for whoever's creating that original content to get paid. Now, the good news is, at the end of the day, the AI companies need the journalists out there creating content as much as anything else. They need original content getting fed into the system. So they have an incentive to actually create a business. And across almost all of the AI companies that we've talked to, every one of them has said,

like, we understand that we're happy to pay, but it can't be a position where we pay and no one else does, this has to be something which is industry-wide and that everyone supports it. And so I think that's one of the places where we're in an interesting position where we can help set what those standards are, apply them across the entire industry, and then find a way that we can do something where, you know, over time, the AI companies are actually compensating the creators for it. My hunch is that if you look at the AI companies, you know, take something that's not an AI company, take YouTube. When YouTube first started, it was like a science project. It was how do I build the technology to be able to deliver a stream of video more

reliably and less expensively than anyone else? And that was their incredible innovation. Over time, everyone else figured that out. So that became kind of a commodity. And so then YouTube pivoted, and then their big innovation was discoverability, the search that Google had and could find videos better than anyone else. And then everyone caught up with that. And so today, the most modern incarnation of YouTube is one where the creators create for YouTube because YouTube pays better than anyone else. And it's the number one platform in terms of compensation. And if you talk to Mr. Beast or any of the big, you know, creators that are out there, like YouTube actually compensates us better than anyone else. I think AI is going to go through that exact same

challenge. Right now it seems like a science project where everyone's chasing AI. And I'm like, okay, what happens when you get to AI? And it's like, well, you know, if Google gets there tomorrow, then Open AI go there a few days later and then Anthropical will get there a few days after that and X will get there a few days after that and then an AGI won't be enough it'll have to be AGI plus or AGI plus or whatever. There's a real risk that these are just commodities

and that everyone's going to be chasing them. So then how do they differentiate? I think the natural evolution is that over time the AI companies will look a lot more like a YouTube or a Netflix where they're all going to have unique content which is very special to who they are and people are going to compete by saying

Like if I want the kind of ultimate AI doctor, then I want to have the one that has access to science and nature and the latest publications and research and all of the different information that's out there. And it might be that not everybody has access to that, but the ones that do are going to be able to have very unique skills based on that. And that content creators will actually get compensated based on different AI companies where they say,

I want to have a unique specialty or unique niche around. around this space. Very different than what the world looks like today, but it's hard to imagine how it evolves in a way that isn't that. Because otherwise, everyone's just going to be chasing one after another, and the models will keep kind of leapfrogging each other, and it's not clear what the competitive advantage of one versus the other is going to be. That is a hopeful future. It's going to take some steps to get there, and you sort of wonder about the various incentives around those steps and whether the future that you laid out is kind of incentive compatible for all of the players. Right now, I think the way you describe it, that term strip mining, that sort of feels like what's happening to the

open internet. I've made comments to David before. Now I feel like I go to a website and it's like I'm in an advertising slum. So it's not just the website writing quality has deteriorated. It's also I'm hit by an ad on the side, you know, a pop up in my face. It's like more inflation of ads. My only other alternative is something like The New York Times say. And then I'm hit with a subscription paywall that's just, oh my God, I have to take out a credit card, pay for this subscription, and they trap me, and I'm never going to be able to cancel. I forget that I have it. And these are my options right now. And if you pay for all of the subscriptions that you want for all of the websites, then you're, 20,000 a month. Yeah. We've got to have something, like, it's not, this isn't because, you know,

the media companies are getting wildly rich, right? They're not. It's never in a harder time to be a media company. And so, like, there's real cost. Again, people have to eat. And so we have to have some way of doing that. And yes, as fewer humans are going to sites and more people are putting in ad blockers and all these things, then the only reaction has to either be put up the paywall or crank up the ads. And that's, and that's, that isn't, that again is it feels like kind of a decaying system. What I think is I'm optimistic about is that over time, this should definitely be a very, we're about to see a massive change because the platform of AI

is going to just force the rest of the world to change. And to your question of incentives, I think the good news is you have AI companies that really want to differentiate themselves with unique content, which is out there. You have content creators that really want to create that content. They don't want to create yet another story of what happened at 1,500 Pennsylvania Avenue today.

They want to tell unique original stories that are out there. And for the first time in history, I actually think that we have, a real alignment, or at least for the first time since it's like the Medici's, we have a real alignment of like the people who are willing to pay for the content are aligned with the content creators wanting to do real interesting work and interesting knowledge. And so if we can create some sort of system where the AI companies are actually compensating people for creating new knowledge, I think that that may be that we're on a sort of a cusp of a golden age of content

creation. I think you're right. And what you're describing is almost like a Netflix type of model of paying for creators just at scale and much more decentralized. But I guess the point is, the AI companies may not go there willingly. It might be sort of like the capitalists in the industrial revolution, where it really took some union work from labor forces to band together and push for some reforms. And I somewhat wonder if that's what you're doing at Cloudflare right now. So about a year ago. A lot of people on our legal team would be very uncomfortable with that language. Okay. But I think there's something to that where, you know, what we, we sort of stopped and said, okay, at a basic level, what do markets require? Everyone says they require supply and demand.

That's not exactly accurate. What they require is actually limited supply and demand. If you have infinite supply, like wherever we're all sitting, like we're breathing air, there's no market for air because air is plentiful. If I go scuba diving, there's a market for air because there's no air down there. So you actually need limited supply plus demand that's out there. The good news is there's tons of demand. All of the AI companies are trying to figure out how can they get as much data as possible. And they're all trying to figure out how can they find the corner of the internet that nobody has gotten to before. And what they're really trying to do is, if you think about what's the history of why did Open AI start? It was Elon and Sam trying to figure out how do they

make sure that Google doesn't run away with the whole game. You know, why did Anthropics start? Because, you know, Dario spins out from OpenAI. So, again, it's a derivative of how do you compete with Google. XAI, again, same, same thing. Everyone is trying to catch Google. And Google has significantly more data than anyone else that's out there. Every page that Open AI sees on the Internet, Google sees four.

For every page that Open AI or that Microsoft sees, Google sees five. For every page that Anthropics sees, Google sees six. for every page that XIA sees, Google sees 22. And those pages are incredibly valuable. It's local news, medical health, all of these things. And the question is, why does Google have access to all that? It's because they've been doing this for 27 years. They've done all the deals with all the weird corners of the Internet.

They found their way behind every paywall that's out there. They've done all of these things that are there. And so I actually think that there's a, when we talk to the AI companies, all of them are saying, like, hey, you know, I'm a little, uncomfortable if Google gets everything for free, why should we have to pay for it? So one answer is we've got to bring Google down to the same level as everyone else. The other answer is, well, maybe what you pay for is not, you know, every piece of content that's out there.

Maybe you just pay for catching Google, because that's going to take forever for anyone to create on their own. And so that's actually been an incredibly fruitful sort of line of thinking. And as we've had conversations with all of the AI companies, they're like, actually, that's really interesting. If you could help us catch Google and get as much information as Google, we'd be happy to compensate for that. And then in the process, we can actually gather that up and say, okay, now we'll turn around and give that to the various publishers. At the same time, I think that in order to create that scarcity, you need some sort of tool to be able to keep the scrapers from getting your stuff.

And so that's exactly where Cloudflare comes in again and says, we can help make sure that people can't have your content unless you let them have your content. And that's been a big piece of, again, creating that scarcity, which, again, I think is helping now drive demand demand. It's a supply and demand market, and it's a complicated coordination game. And so can you describe maybe there almost seems to be, maybe you wouldn't call it a war, or maybe the legal folks in this episode wouldn't call it a war. But that's sort of... They're fine with war. They're just a little union.

They're fine with war. Let's use the term war. It turns out that you have all kinds of antitrust problems if you're if you're a corn man. people together in order to compete on price. But that's a whole other story, though. Okay, okay. So Content Independence Day. That was July 1st, 2025.

That was last year. We'll use the terms of war. There was, of course, a revolutionary war, which was kind of the creators in the colonies escaping their British overlords. And this was a blog post that Cloudflare authored that was about blocking AI crawlers by default across your customer base.

So essentially... A little asterisk there, but we... Yeah, yeah. Come back to that. Okay, so essentially, as I understood this, Cloudflare has about 20% of the world's websites of the internet, something like this. And you set a Cloudflare put a rule in place such that creators could opt in,

website, publishers, they could opt into this to say, hey, I want to block AI bots from just, going through all of my website's content. At some level, this is publishers, this is creators, you know, kind of pushing back and say, you haven't asked for any of this. You just crawl our website and then you put it in chat GPT. You're not compensating us.

I want to opt out of that. I want to turn that off. That's how I understood content independence day. There's a little bit of confusion here because people then assume that we apply that across all 20%. And that's actually not right. There's a huge percentage of the internet that deeply wants to be,

and in fact is working super hard to be in all of these AI systems. So just take Cloudflare. So Cloudflare, we published a bunch of knowledge-based articles. Those knowledge-based articles are like how you configure our system, how you use our developer tools.

We want every AI company in the world to have that stuff. We do a whole bunch of things to make it as easy as possible for AI companies to consume that stuff because our business model isn't ads or subscriptions. Our business model is how do we get more people to use our tools, right? So we want to make that as easy as possible.

A lot of the internet wants to make that as easy as possible. And for those people who want to make that as easy as possible, we're not only not blocking the AI crawlers. We're actually making it super easy for you to make it consumable, switch into like code mode, do markup instead of HTML by demand because it consumes fewer tokens, lots of different stuff to make that much more accessible for those companies or creators or organizations out there

that want their stuff to be in the LLMs. There's a whole other universe, and it largely is based on, is your business model based on ads? That universe wants the AI companies to have none of their content unless they compensate them in some way. Because it fundamentally, if they have their content, it's taking it, basically giving people the Cliff Notes versions, which makes it less likely that they're going to go and see the ads and click on it. So we want to support either side of that. We're not religious on either of it. What we're religious about is it's your content, and you should have the tools to be able to control who has access to it,

either making it super easy or making it super hard. And then that's kind of the position that we're in. Content Independence Day was saying, we're going to make these tools available for everyone. Regardless of your budget, you can use the free version of Cloudflare service. You should all be able to control who has access to your content. And that was really kind of the line in the sand that we drew last July. Okay, so I'm of split minds about this.

I suppose. So I'm a creator and I'm also a user. So from a creator perspective, I definitely need to be, want to be compensated for my creator output. And some of the podcast model, of course, is ad model. Some of it is subscription model. So we use both of those at bank lists. From a user perspective, I feel like I want my AI agent to be able to scan different websites and be able to bring that information into my knowledge base. And I get frustrated when I go to, you know, my quad or my chat GPT instance. And I'm like, hey, here's a link. And it's like, I can't read this link. And then I have to go and I have to go copy and paste it or clip it in an MD file and paste it and be like, now can you read it? And I feel like part of my life, a fraction of my life right now is I'm just like a

copy and paste agent for some super intelligent LLM. And that feels weird. I feel like an unnecessary middleman. And so what about the argument that no, you know, AI agents are actually acting on behalf of humans, and it's kind of the same thing. So if you block an AI agent from a website, then you're effectively, you can be blocking a human from marshalling that AI website as their hands to go acquire that information. Yeah, I mean, I guess, again, comes back to something even simpler, which is, like, if you're not going to the site with your eyeballs, then you're not seeing ads. If you're not seeing ads, then the website isn't getting paid. And if the website it isn't getting paid, then the person who's creating that content can't eat.

And eventually they're going to stop creating content. And so, like, I get it. Like, I think as a user, we want to make that as easy as possible. But it's the same thing that, like, I'm so sorry if you don't, you know, subscribe to, if you only subscribe to Netflix, but you don't subscribe to Hulu, then you can't watch, you know, I don't know, or HBO, you can't watch Game of Thrones, right? I mean, there's going to be different things that different people have access to, because at the end of the day, the people who are doing the work to create content deserve to be compensated for it,

and we've got to create some system to do it. I think over time that that system looks a lot more like, again, a Netflix or a Spotify, where some amount that you're paying in a subscription to your AI company then gets carved off and actually sent to the content creators. They're creating content that's out there. And in an ideal case,

the AI companies are also feeding information back to content creators saying, hey, here's knowledge we feel like is missing in the world. It'd be great if you went and did research on that and figured it out, and we could compensate you for that, for that creation of knowledge. And, like, that's a business model that works. I went up to Stockholm to meet with Daniel Eck, who's the founder of Spotify. And he told me a bunch of stories, but the one that's really stuck with me, he said, you know, the way Spotify works, like, if somebody searches for, like, Taylor Swift, shake it off,

like, we return results and we are sure we're getting people what they want, right? On the other If someone searches for like, I want a song to a disco beat about how fun it is to dance with your cat, not a lot of content, which is out there for that. And they know that. They know that they kind of get a bad hit on the search. And so what they do is they actually take those searches that people run where they don't have good answers. And they publish the back to music creators that are out there. And this is the moment in time where like anyone listening is going to be like,

I am in the wrong business. There's a guy, a guy who makes 40 million euros a year. just writing songs for unfulfilled Spotify queries. And he's not alone. Like he's like the top commentator guy. But there's a whole bunch of people are making millions of dollars a year just writing songs for people

things that people search for. And you can... Are they good songs? Hey, I mean, people are listening to me. They don't get paid if they don't get listened to. Hey, why it's subjective, David? You don't get to decide. But I also think that what's...

I actually think there's something kind of beautiful about that because like if you think about it, when someone searches for a song, it's like they're trying to summon an emotion. Right. Right. And if they don't find it,

then they don't find it. it and we might not know what those emotions are, but this is a way of actually surfacing that. And you could think about a model in the future where the AI companies will say, hey, listen, we didn't find such a great, you know, like information about what the, you know, the carrying capacity of an unladen swallow is. So if you go out there and like do research on that, we're going to be happy to pay you for that, that information. And then we can figure out and we'll incorporate into the system. And, you know, the way I think about it is, you know, we've never, until now, had a mathematical model

of the sum total of human knowledge. We kind of do now. Like, that's what the LLMs are. Right? And I picture it's like a giant block of Swiss cheese. There's a lot of cheese. That's the knowledge. But there's a lot of holes in the cheese.

And the same mathematical models actually show you where the holes are. And so in the future, I actually think one of the things would be really interesting is maybe we compensate people for filling in the holes in the cheese, right? Which is different

than compensating them for traffic we're actually compensating you for filling in where the gaps in human knowledge are. And if I were running one of the big AI companies, I would launch, I don't know if it's the equivalent of like a Nobel Prize or an Academy Award, but every year across all of the different fields, they should publish, here is the person, here's the content creator, academic, you know, researcher, journalist, whoever it is, here's the content creator who did the most to advance, you know, biology or physics or journalist. or whatever it is and actually reward that,

both with recognition, but then also with maybe compensation in some way. And I think that it's amazing that for the first time we can actually measure who did the most in terms of furthering human knowledge in any given topic, which is out there. And I think that's super exciting and then starts to suggest what that business model is future.

I just want to make one observation that that model that you described seems a lot healthier than the eye, attention economics model that is rotting our brains. Yes, except, and I don't know the solution to this. So I am all ears for ideas. Done, I think there are two potential problems. Potential problem number one is it all depends on a subscription model for the AI companies,

where the AI companies are collecting dollars, and then they're basically putting those into a ASCAP, BMI, Spotify-like pool, that then goes out to the different content creators that are out there. And the challenge, again, is I think in the global South, you know, people don't have the money to pay for this. And so it might be that we're going to create

a much bigger divide in terms of information to haves and have-nots in the future. And again, that's a, that I agree that we solve a lot of the kind of attention economy whole problems, but we might get into something else. And people, there's, I was just at an advertising conference and everyone's like, oh, someone's going to create the ad word for AI. No, I don't think they are. Or maybe they will, but it's not going to be nearly as good

of business because, like, in the future, I am sure that I will be spending over $1,000 a year for my AI agent, like for sure. Compare that with Google today, which gets about $600 per US user that's out there in the case of like meta, it's like $253,300 per US user that's out there. So $1,000 is a way better business in terms of just the amount of revenue that's there. And I think a lot of people, a lot of households are going to spend that. I might spend even more than that to be able to have a helpful agent. If that helpful agent has ads associated with it, if I say, hey, what car should I buy? And it's like, well, you should buy a Ford because Ford paid me to tell you that.

Very quickly, I'm going to say, I'm not going to spend $1,000 for this biased agent. I'm going to go shift to someone who's unbiased. So I actually think that being trustworthy is going to be one of the things that, for those of us who can afford it is going to differentiate different agents that are out there. If you can't afford it, though, it's going to be an ad-based model. But it's going to be kind of the lowest end of the ad-based model and a lot of just fraud and problems and all kinds of things that are out there. And so, again, I think that that's one of the big problems is it's going to differentiate like halves and have-nots in that space. And I think we should be really cognizant of that

and try to think of solutions to it. I think the other problem is there's a real risk that done incorrectly that any of these models can do. just exacerbate who already has a lot of money. So for instance, imagine in a version of the future that for an art, we just set a price for an article, an article costs, you know, a tenth of a cent or something in order to, in order to purchase. And so every time, you know, your agent goes and does that,

a tenth of a cent goes out of your account. And again, we can talk about the mechanics of how that would happen. But, but we do that. The challenge is that if we do that, especially if it's a bid-based system, then the winner of whoever, you know, gets access to the article, is just who has the most money today. And so that means that Google, meta, Apple, Microsoft, Amazon, like those companies could basically block everyone else who's out there out.

And what I really worry about is how do we make sure that as we create these systems, that a new entrant, a new startup, you know, is able to have access in a way that, you know, that they don't have to pay the same amount as someone like Google. And so at some of all, I think we've got to figure out how do you pay different different amounts depending on what your scale is. And the answer might be it's just, it's a percentage of what your revenue is. Like if your revenue is really high, like you put more into the pool. If your revenue, if you're just starting out, you put less into the pool, but it's still

the same percentage that's getting built in. And then if you're successful and you grow, then that will grow with it. But I think we have to be really, really, really careful as we design the market to make sure that we, yeah, I think we'll solve a lot of the problems we've had, but we may create whole new problems going forward. In 2024, emerging markets generated over $115 billion in annual yield for investors, with yields ranging between 10 to 40%. These are some of the highest, most persistent yields on Earth.

The problem, Defy can't access them. Bricks changes this. Built on Mega-Eath, Bricks takes emerging market money markets and solve them carry and turns them into composable primitives you can access straight from your wallet. While defy investors earned 3 to 6% on stable coins and T-bills, institutions have been harvesting 10 to 50% yields, by sovereign monetary policy. Bricks connects these worlds with institutional gray tokenization, local banking rails,

compliance across jurisdictions, and real-time stable coin settlement. Bricks does the heavy lifting so Defi can finally access real collateral and structured products on top of real world yield. Even the best carry trades can be within reach. Bricks brings Defi's promise to the emerging world and brings the emerging market yield to your wallet. Let the yield flow with Bricks. Some exciting news.

We are launching a new podcast to help people figure out the crypto cycle, how to navigate it. The best crypto cycle investor I know, his name is Michael Nato. He runs the Defi report. This is the guy that sent me a sell alert before the 1010 price drop happened. His cycle analysis has been absolutely on point. I've been following him for years. And this year, we started recording weekly podcast episodes. Each one we get into his portfolio, what he's holding, the market structure, entry targets, fair market value of Bitcoin and Ether. And where we are in the cycle, there's new episodes that are released every Wednesday. They're 30 minutes. They're short. They're punchy. I think this cryptocurrency

cycle is harder to navigate than most. So let's do it together. Go subscribe to this podcast. Search the Defi Report wherever you get your podcast, YouTube, Apple, Spotify, or find a link in the show notes. There's a new episode waiting for you now. Yeah, certainly the implementation details need to be fleshed out. But I do kind of want to step back and talk about how great this might be. What you're saying is a little bit reminiscent of what platforms like YouTube or Twitter or Facebook or Instagram did to the internet where it just allowed more people to create more and allowed for more business models,

more archetypes of people, like influencers as a business model, like whatever people think about influencers, like it's a whole entire economy out there. Twitter is the goddamn funniest website that I've ever come across. Like the jokes and memes on there are hilarious. YouTube can answer any of my questions,

even questions I didn't know that I had. So in terms of just like human creativity and content production, all of these platforms just unlocked something massive for humanity. And I'm seeing something kind of similar to what you're saying where we're creating the new economic substrate that we're creating can incent just brand new economies and allow for more aggregate content production more than we've ever seen so far because of this new substrate. Now like so much of this conversation so far has been solving problems. Like the ads model is kind of fraying.

The AIs are going to cause some problems so we need to fix them, which is totally great. But also I kind of just want to emphasize how great this could be in the future. And totally, there's problems along the way that are going to be created that we need to fix. But I also don't want to diminish the massive potential

that we have if we design this thing correctly. Am I being overly optimistic here? No, again, as I said, I think we might be on the cusp of a gold native content creation. Because again, if we're designing content, which is actually furthering human knowledge and the people who do that get rewarded as opposed to or make people laugh or just entertain. Like, those are all good things. Whereas if you've just written the same story as cheaply as possible and put an incendiary headline on top of it, that's a per se bad thing. And we should not be building business models that encourage that. And so I do think that there's something, there's something that's really, really, you know, potentially amazing if we get this right.

And it's also just maybe more straightforward. Like, it's kind of weird that like you go to the website to read the thing and then you get monetized by an ad that's there. Like at some level, just when you go to the website, you pay and that helps cover the, you know, cost of the journalist eating, but also the cost of actually delivering the website to you. It's kind of a more honest business model. Everyone, like, Cloudflare, like one of the things that's been great about our business is super simple. Like we make money because people pay us. Like, you know, and they pay us because we make them faster, more secure, more reliable.

Like those are the things that are there. Like, we've never had kind of that ad-based business model, and it's always felt kind of a little bit confusing, like who's the customer, you know, who's the product. Like, if you go to an ad-based site, you're the product, right? The customer is the advertiser, right? The customer is whoever pays you. That's not how most people think about it.

if you found it a different way where you're actually doing something like a micro payment, now of a sudden you're the customer, which means that the journalist is about serving you as opposed to about serving the advertiser. And so I think that there's a way that going forward this could be really healthy.

And so I don't want to diminish just because there are problems. It doesn't mean it doesn't make the world better. But I also think it's important to say it's not panacea. They're still going to be challenged. We're going to still have to figure out our way through them. And we should be cognizant of those

as we design this. And think about, like, what do we want the future to look like? The way I describe it is, like, I want a future where there are not five AI companies. There are 500,000 AI companies. How do we make that as easy as possible? I want a future where there's not, you know, a few content creators, but there's an explosion where anyone can be a content creator, anyone can share truth and it's inaccessible. And they get compensated for it if it's actually valuable.

And then I want a world where, you know, businesses large and small can compete and thrive on a really fair playing field. And, like, as I think about this, those are the goals that I'm. playing for. And I think that's that it's actually pretty easy to alignment around policymakers, most internet users, you know, even most, most AI companies say that's kind of what we want as well. I love that goal. And I think more people need to listen. That, that to me would be a better internet. Let's talk about how we can actually achieve it and how you guys are working to achieve it using some of the technology that we have today. So a perfect world, of course, let's say I'm in my chat GPD cloud user interface. And I want to go access some content.

from a creator. I'm actually paying for it. Okay? And it goes, happens in the background. It's a micro transaction. I've scanned 20 different articles. I pay one cent each. Yeah, chat GPT adds it to my bill. It's less than that. Okay. So fractions of a cent. My, my macro paying, not micro paying brain can't even comprehend how small these transactions can actually be. You had this line in a back and forth with theology about a year ago that I read. And this was the moment I was like, we got to get Matthew on the podcast. You said this to apology. Utopia's humans get content for free and the robots pay a ton.

I think that's kind of what we're talking about. The content is still free. It's democratized. But it's the robots that are paying and it's embedded in these micro-transactions. Can you talk about all this? Pay per crawl X402. How do we implement this? Yeah.

So, okay. So right now, Cloudflare handles any given second about half a billion requests. So 500 million requests per second flowing through our system. When we look at those requests, we think that somewhere between 1 and 10% of those are requests that you could actually monetize in some way that people would pay for having access to. And again, fractions of pennies for it. But if you imagine that just the traffic to the Internet is going to continue to grow like crazy, those fractions of pennies can add up pretty quickly.

It's sort of the Superman 3 or Office-based business model where you're trying to find fractions of pennies that are out there. But that's, it depends on how old your audience is and that might have gone over a lot of their heads. Spotify does this model, right? Yeah, it's exactly. It's the same, it's the same thing, right?

And so, like, I think that we understand what we have to do. And the good news is that, you know, 4-02 has existed as a, as a sort of response code for a long time, which is the response code that says payment is required. And so we, along with Coinbase and, a handful of others are working on creating what, what does that, what does that foundation look like? When a website says, you need to be paid, how do we make sure that the rails all exist for that that to happen? I think the thing that's, I'll tell you the thing that's hold this back.

Like, again, everything that's happening in crypto is, is amazing. And obviously, this is going to have to be, like, you can't put this on Visa's rails. Like, it doesn't work at, at the scale, the costs are too high, because you've got to be able to do this incredibly inexpensively. Well, doesn't Visa charge like a 30 cent just initiation fee on every transaction? Yeah. And so like that just doesn't, that doesn't work, right? So in order to do this, it's got to be some sort of stable coin solution. What we've struggled with is people like will say to us, oh my gosh, you know, we're so excited.

We can handle two million transactions per second. And and I'm like, it's awesome. Good job. But I think day one, I need 10 million transactions. actions per second. Oh my God. And I think it's going to just go up from there. And so, and so I think that the thing that that is that we're struggling with, and again, we don't, like, I don't want to build any of this stuff, but we don't see in the ecosystem yet anyone who has built something that we think can scale to the levels, where the, the,

transactions amounts are tiny, but the volumes are extraordinary. And so that's what we're trying to make friends with everyone who's in the space, who is doing something interesting. And again, I think we've, we, we, we, we, we, we, we, we, we, we, we, we, we, we, we, but, but, but, but, but that's the place where I just don't think that people are thinking about the scale, I mean, a million transactions per second is more than visa does. And so like, like, like, like, and, and I think you've got to, you, you, you, you, you probably have to have a way of scaling to, you know, know, at least two and maybe three

orders of magnitude bigger than that. And that's, and that's, that's what we're, we're trying to figure out right now is exactly how does that work. Once we got that, again, I'm kind of optimistic. We got willing buyers. We got willing sellers. We've got, you know, I mean, Sam Altman talks about how important microtransactions are going to be. You know, the anthropic guys thinking about, like, I think we've actually got a lot of alignment. But now it's kind of how do we make sure that whatever, whatever systems we implement are going to actually be able to scale to the potential of, you know, the new business model of the internet.

Okay, so let me make sure. I repeat some of this back to make sure we understand. So right now, what we have is a screaming need for microtransactions so that creators can get compensated for all of the AI agents that are harvesting their content and strip mining them. They can be compensated for this. Okay. So we have that. We also have this standard called 402, X402, which is kind of an open cryptocurrency micro transaction standard on top so that AI agents can go pay with some sort of stable coin and get hooked into the system. So we have the ability to do microtransactions. So we have that. Saying that it's a standard, I think it's an evolving standard. I think we, I mean, there's, there's four or two is part of the original specification of,

of web servers. Then Coinbase took that and said, hey, let's actually build then the infrastructure to hook that into, you know, payments, payment rails and systems. They then donated what they, their work and to the Linux Foundation and then brought other people, including us, into help flesh that standard out. I think we're still in the process of figuring that out. But then even if you figure that out, you've got to then have a stable coin infrastructure which sits behind that, or more than one, ideally, that can actually handle the volumes that are going to be out there.

And again, I think those volumes are beyond, you know, what anyone is anticipating. So that's what you're stuck on right now, or that's what you're, where I've trying to figure out how to engineer that. What solutions have you seen? So does this be in a blockchain? There's obviously what Stripe is doing with tempo. You know, Circle has a very successful stable coin. They have Arc.

You have Coinbase with base. Cloudflare has talked about, at least we've heard rumors, of a stable coin as well. So what's kind of the latest and greatest? How are you thinking about slicing, dicing, this problem? Yeah, I mean, I think we went into this thinking that we were going to build sort of a layer to stable coin

It would live on more than one different kind of layer one chains that were that were out there. And again, we've talked to every one of the people that you just mentioned. And they're all really smart. They've done amazing things. What I think is hard for people to comprehend is just the volumes that the Internet really looks like. And so that's where we kind of come in and go, that's awesome, but you've got to be able to report two or three orders of magnitude more than what you're talking about today. So we are still looking for that. I think the question for us has been, do we wait until we figure out exactly

what's the right long-term solution before we launch that? Or do we sort of try and cobble together something that'll hold up now and then the background kind of create what that future is? And again, like whether it's us creating it or whether it's, you know, one of those companies, there are a bunch of startups that are doing really interesting things as well. Like, if you're out there and you're working on this problem and you're not talking to us already, you should be because we can give you that incredible distribution. We have no pride of ownership here. And what we want is to live up to our mission, which is help build a better internet.

And that means, you know, yeah, we're going to support everything which works out there. But the volumes and scale are so enormous that, again, I think a lot of what, a lot of, Again, a lot of the companies that you mentioned have incredible, like, purposes for what they're doing, but nobody is doing the volumes that we potentially can, you know, help facilitate. And again, it won't just be us. Like, I think done correctly, you know, this goes across the entire internet. And the volumes could be, I mean, just extraordinary. So we've got to make sure that we've got the plumbing right in order to support what the future is going to bring.

Maybe I need help understanding the magnitude of the volumes that you're talking about. but a lot of these blockchains are being built in the most modern era. In 2026, we know how to build blockchains that go as fast as we've ever known. Arc tempo, Mega, Eith, all of these things have completely taken off the brakes in terms of capacity. Yes. And so I'm kind of confused as to actually how what they've built somehow still isn't big enough to do. I know the internet is very big, but I would imagine on. Matthew is saying our size is not sized, David.

I think that's what he said. He's saying our size is size. Yeah. Crypto size. It's not size. Crypto size is not size. Cryptos size is not size. Yeah.

Yeah. These aren't going to be, like the dollar volumes might be small. But the transaction volumes are going to be extraordinary. Yeah. But on day one? Like, on day one,

it's still going to be like when things are just going to get rolled out here, right? That's the challenge with us. Like that's true for most people. Like if you're, if you've got to go out in there and sell one website at a time, it's fine. We can take, and again, we say 20%, the real number's bigger than that of the internet,

and we can turn it on overnight. Like, the volumes are pretty high. Again, right now we do about half a billion, 500 million transactions per second at Cloudflare. Not every one of them is going to be monetizable, but we think it's between 1 and 10%. So that's between 5 million and 50 million transactions per second that are monetizable.

there's no blockchain out there today that supports $5 million. The highest we've seen is two. And so if you are out there, if you're listening to this and you support higher than two, talk to us because that's exactly what we need in order to do it. And again, I think it's inevitable that it's going to happen.

I'm trying to not have to be the person that has to build it. But again, at some level, if we're the only one who has the problem, maybe we do have to figure that out. So if you want to go build like a layer one, one, you know, blockchain that can support 100 million transactions per second, you know, call us because we, we, that's, that's, I think that's, I think that's going to be, that's, that's, that is what is necessary in order for us to help build what is going to be the business model

that will be sustainable for the internet in a world of agents. I guess I'm confused as to why it seems like it's a flip of a switch rather than a slow increase, because as I understand, if we're going to bake stablecoin payments. You're a Cloudflare customer and we turn this on and we say, oh, only these guys over here get it, you don't get it. Like, how, I mean, that's just going to, like, it doesn't, like. Yeah, but don't you have to, like, fund stablecoin wallet. Like, it still has to, like, diffuse out.

There's, there's, all of those things can happen very quickly. Okay. All right. Yeah, I get it. We're a, we're a, we're a fairly substantial, you know, public company that, like, that, that's not going to be our limiting fact. We're not a startup. And so our limiting factor isn't going to be how do we go, you know, convince everyone to do it. The limiting factor is going to be how do we make sure that when we turn it on, it can actually scale as big as you want. And that's, again,

that's a high class problem, but it's, but it is a problem that we think about. And what the worst case is we do it and it starts to scale and it breaks. And so that's what we're trying to make sure is that we've, if this takes off the way we think it will, that it can continue to scale up and support that because agent traffic isn't slowing down. It's going to go just straight up. And again, something has to help compensate content creators and the people who are providing the infrastructure to make that, those agents be able to have access to information. Let's talk about the final form of this then. Let's say we figure out the scaling problem. Okay, we got micro-transactions, we got 402, we have agents paying microsetro-s,

sense for access to content. What else is needed here? What does the world look like? Or is it just like tech is the missing ingredient? I mean, Matthew, there's got to be some coordination challenges here, right? How do you get the AI companies on board? You know, David was talking about users with wallets, all of this stuff. You're already frustrated with your agent can't get to things because Cloudflare blocks it.

Yeah. So like I think the coordinate, I think we're in this unique position where, again, 80% of the AI companies our customers. We know them super well. Huge percent of the internet is using us. Like, I think we're going to help coordinate and catalyze it. Now, I don't think we'll be the only ones providing it. But once you get that kind of key critical mass, I think the rest of the internet can then go off and adopt it. And then it can become a standard that works across everyone. And is that just like, does that just become the business model for creators moving forward? It's, they're still getting paid? I think it's going to be different, you know, I think there's going to be

different parts of it. Like, one of it will be, again, the micropayments. which is what this is a piece of it. I think there's another part that maybe like how you train models, how you find that. Like it's like the training versus, you know, grounding. Like those things might have all slightly different versions of this. And I do think that over time you still want to have, especially for certain content,

you want people to go back to original sources. And there should be some way of, like, there's still going to be an advertising business that exists. It's going to be smaller. It's the same way that, you know, you still sell vinyl albums. It's just a smaller part of your business than it used to be

whereas the Spotify streaming has become a much bigger part of your business. I think this will just change exactly what those allocations are. But I am optimistic. If you think about the music industry almost 23 years ago to the day was dying. Whole music industry was worth

valued about $8 billion, which is a lot of money, but not a lot of money for like the entire music industry. Right. And so It was pretty crazy. And what was happening was people were just stealing the content with Napster and Grokster and all these various things that were out there. And then, you know, Steve Jobs steps on stage, almost 23 years ago to the day,

and announces iTunes 99 cents a song. Not the business model that won, but it was a critical step to get to the business model that won. And so I, and again, the business model eventually won was Spotify, you know, $10 a month, all you can eat. But you had to kind of go through that. So I don't know. what we're going to end up at.

But I know we've got to do something that puts a line in the sand that says, content is valuable, and the people who create it should get compensated for it. And then we've got to create the technical solutions that make it possible to do that in a way which is fair so that everyone has to pay something which is fair and either depending on their size or however that works.

And then once that happens, I think that the business models will evolve over time, but what we're trying to do is say, how do you create the right conditions for a market to develop and then provide the technical infrastructure or partner to provide the technical infrastructure to actually then allow that market to thrive? Do you think this new business model could power journalism

as small as kind of like the local newspaper? I know that you and your wife, you bought the park record, which is your local hometown newspaper. Circulation 8,000, Park City, Utah. Okay. So apply maybe micro payments to the park. I mean, could it work? Yeah, I mean, what's been interesting about the park record has been.

So if you think about what businesses, content creation businesses, you know, thrived over the last 30 years versus which ones withered and died, local news is one of the ones that withered and died. It was incredibly hard because you needed scale in order to drive the economics and advertising. It just didn't work otherwise. And so, like, again, we bought the newspaper, not because. we thought it was a great business, but because we thought local news matters. And what you want to do in order to support a community is actually have sources of information that people can trust. And the only KPI that we measured, it wasn't around financials or anything else. The only KPI I measured was voter turnout. And voter turnout when the election right before we bought the paper was at 23%. The last

election, which was an off-cycle non-presidential election, it was 63%. And the only thing that changed was we actually covered like the election. a lot better and cover candidates and things because people don't like to vote if they don't feel like they know what's going on it. So I'm really proud of that. What's been interesting, though, is I think that this year we may make more at the park record

off AI licensing deals, which don't require micropayments or anything else. Those are like one-off deals with AI companies. Then we do off digital advertising. Why? Because it turns out that local news is exactly what the AI companies want. Like if you go to chat GPT and you say, hey, I'm planning a vacation to Park City, Utah.

What's the best new restaurant? If you don't have the information from the park record, you cannot answer that question. And so actually, more local news is more valuable. Compare that with, like, the New York Times versus the Wall Street Journal. If one of those papers doesn't license their content to you, but the other one does, you basically just take the Wall Street Journal and tell it to rewrite its content

as if it's a New York liberal and you get the New York Times. There's another way of saying that it's unfair to the New York Times, but not radically unfair. Another way of saying that is that the media business for a long time was tell the exact same story to an increasingly narrow tribe and an increasingly narrow audience. I don't think that that has much value in the future. Whereas imagine in future version of the New York Times that reviewed not hotels in New York, but hotel rooms in New York. I can tell you whether 1313 versus 1314 and the Marriott Marquis is better, right? Incredible local news. You know, I can tell you what bodega in New York is the best place to get a bagel, right?

Or whatever it is. Like, I think that's the content, which is actually going to be incredibly valuable for AI companies going forward. You can't be AGI if you don't have that information. Whereas if you don't have, you know, yet another take on what happened to the White House today, like there's plenty of places to get that information. So I think the future of news actually might be much more unique, much more local, much more original. And it's the same reason that if you take the New York Times, great, amazing media organization. And you take Reddit, also amazing media organization.

It turns out that if you look at the corpuses, they have about the same number of tokens in them. You know, New York Times has been publishing a lot longer, Reddit's higher volume. But Reddit got substantially more for the same number of tokens for the same number of tokens for their content than the New York Times. Why? Because if you don't have Reddit, you don't have Reddit. If you don't have the New York Times, go get the FT or the Washington Post or the Boston Globe or the Wall Street Journal and you can kind of cobble it back together. I think that the future media is going to be not that. Like, I'm just going to tell the same story, you know, with my hyper-local audience. It's going to be much more. I'm going to tell a story that nobody else is telling. And by the way,

as an audience member, how much more would you rather actually see that? You don't want to see yet another story of what happened to the White House. You want to see, like, hey, here's a really interesting thing that happened in your hometown or your neighborhood or what was going on. Like, people are craving that. And so if we can create a business model that rewards it, I think that's actually, you know, really pretty amazing. Matthew, I feel like there's one last problem we have to solve here. And you've alluded to it several times throughout this conversation.

and I'm not sure this new business model for media has fully addressed it. And that is the concentration of power issue. The concentration of power in maybe the top tech companies, the top AI labs. I know you have some history here. So you have your JD from Chicago. You also used to teach cyber law. And I know even Cloudflare started as kind of like open source technology. And you care about building a better internet.

and so much of that has been based on the open standards and civil liberties that we kind of built in this web 1.0 era. So from like I guess maybe a balance of power perspective or a civil liberties perspective, how can we fix that problem? How do we solve the problem of too much centralization of power or censorship ability in the hands of a small group of tech oligarchs, let's say? Yeah. So again, I come back to what am I playing for? And I'm playing for a world where you have lots of AI companies, you've got lots of media companies, you've got lots of businesses,

and they can all compete in a way which is, which is, which is, which is, which is fair. That isn't the natural state of things. Like, we already feel like there's a very small number of AI companies that might win the game. And if it turns out that building any of these models cost billions of dollars, like that's going to limit the number of people who are able to do it. You know, I, I, like the black mirror version of this is, like, content is going to be enormously valuable.

But could you imagine Open AI launching their own version of the Associated Press? Would it cost that much, right? A lot of out-of-work journalists right now. And they've got to have unique content that's out there. And so they had bureaus around the world that was reporting on what was ever going on. Then, you know, it would feed that back. And by the way, what you would then see is you'd have one AI company that would be, and you're already seeing this to some extent.

You'd have the liberal one and the conservative one, and they'd all have different opinions. And by the way, that would then lock us into silos because we're basically all going to subscribe to one or a very small number of these things in that world. And so we get kind of focused and just in our own filter bubbles that are out there. And then, by the way, if we want something else to just have nightmares over, like the real challenge going forward. I think we can maybe solve the media challenge. The one I'm the most worried about

is what happens to small businesses that are there. If you think about any small business that you do business with today, the reason you do is business with them tends to be either because they have physical proximity to where you are. And so you can save time because your time is limited. Or because you have some emotional connection

to them. Your agent doesn't care about either of those things. And so your agent is actually going to bypass the local bodega to figure out where they can get eggs the cheapest or that match whatever your preferences are. And who might win that in the future is not a bunch of small businesses. It might be a bunch of very big corporations that are very good at talking to and answering to agents. And so, like, there's a lot of ways this can go wrong. And in fact, I think the natural tendency is going to be, we have very few AI companies, very few content creators, and very few big companies that solve it.

I was talking to a guy who ran small business that ed into it and then ran PayPal for a while. And he said, in the future, I worry there are five companies, literally. One that makes things, one that holds the land,

one that transport things, one that holds the money, and then one AI company. And again, that's an extreme version, but you could see a version of that. And so I think the first most, important thing for all of us to do is articulate what future we want. We want one that doesn't

have five AI companies but 500,000. We want one that has anyone being able to be a content creator and get compensated for creating original knowledge. And we want one where businesses large and small can compete and thrive in this interesting world. And then as we go forward and we think about this, we can ask ourselves, is that what we're doing is the business model we're creating facilitating that or kind of tearing that down? And again, I don't have all of the answers, but I'll say these are the things that I'm thinking about, and these are the things that everyone at Cloudflare is thinking about.

And if you or your listeners are interested in this, like these, I think, are the most interesting and important questions over the next five years, and the entire business model of the Internet is going to change. And if you're not thinking about what that change is going to be, you're not thinking about what the most interesting question in the world is. Matthew, we've answered a number of questions, and I feel like we're also closing this conversation

with many more remaining, right? and it's going to be up to us to shape the path and figure out how this future works out. Your message lately has been to content creators, and I'm sure we have a number of different content creators listening right now. If for someone who's a content creator, let's say, they're involved in kind of the media side of things.

They have a blog, maybe they have a podcast. With the advent of AI, do you have any advice for them? Like, what should their move be? I mean, I think you've got to think about who do you want to have access to your content? and the answer is going to be different depending on what the content creator is, and it's going to be different based on what your business is.

If you're running an ad-supported business, it's going to get really tough for that alone to work if the AI companies can just slurp your content down because they can literally just say, you know, you can type in, you know, what happened on the bankless podcast today, and then they don't have to go listen to you, right? So which if you're embedding ads or something,

and that's really a challenge for content creators. And so I think the first step has got to be like, I'm going to take control of the content. That's why we at Cloudflare, and again, it doesn't have to be us. We're trying to say, let's give the tools to everyone that's out there. We need to make sure that you are controlling

who has access to your content and who doesn't have access to the content. And that doesn't have to be the final step. Like, over time, you want anyone who's compensating you in a way that seems fair to be able to have access to your content. But if they're not compensating you, then why would you give access to them, right?

If your business model is somehow doing that. If you're just doing it because you love to create content, great, that's great. Do whatever you want. We think the key is let's make sure that content creators have the tools to either make it super hard or super easy for their content to be accessible based on what their rules and their preferences are. And again, I think that's where I would start. And once you've done that, then you can kind of figure out what that future is. And again, I don't know that we're going to invent what the business ball is.

It might be some content creator that's out there that says, hey, I've figured out that I can put my content available. and when a bot comes to see it, I'll actually serve them with kind of, you know, a prompt that helps kind of inform things and I can sell access to that prompt. I don't know. There's a lot of different possibilities. We're going to have a ton of experiments that are out there.

I don't think we have all the solutions, but we're sure that there has to be some sort of control over content in order for content creators to get paid. And then we're sure that there needs to be some really, really, really efficient way to deliver those payments at an enormous scale that, again, And I don't think there's any financial system out there that has been built to support yet. What do you think about substacks positioning here?

Because exactly what you said, where there's a business model that make sure the creators get paid. Well, substack is a platform for creators. It's already subscription-based. I can easily see substack just baking in AI communications, agent communications. And then they could easily pivot to being like the most primary modern business model for content creators. internet if they just implemented some of the stuff that you're talking about. What do you think about their position? I mean, substacks is a customer, so I don't, I don't want to speak for them. But I do think that I think that there, you know, there's going to be a lot of opportunities

for people to experiment and figure out what the business model looks like in the future. And, and again, I think we're in a unique position to do that. Someone like Substack is an unique position to do that. You know, the X's and metas of the world are in unique positions to do that. And what I think is we've just got to all kind of recognize the internet of the last 30 years for all of its amazing things and all of its challenges is going to change. What it changes to, I don't exactly know. But I think that's so exciting. And so I would hope that the folks at Substack are trying to figure out how to do that. And again, we're happy to help in any way that we can.

But anyone who's creating content, any platform that's helping like host or or, or surface that content has got to be thinking, like something is going to change here because it can't be that opening AI buys a single subscription to each of their publications, slurps all the information out and then publishes it to everyone. You know, that's kind of exactly what's happening right now.

And that's not sustainable. Matthew, this has been great. Thank you for the mission of building a better internet and we'll hold you to that. We need some help. Helping build a better internet because we can't do it. And so again, And we're always trying to find ways to partner and do more with others.

And so if anything about this sparked an idea, I love to chat. Well, you are dialed into the crypto audience. And I think there was a clear call to action. If anyone knows how to build a super high performance to Cloudflare levels, layer one, you know, contact Cloudflare, contact Matthew. Guys, got to let you know, of course, crypto is risky. So is the entire internet. You could lose what you put in.

But we are headed west. This is the frontier. It's not for everyone. But we're glad you're with us on the big. bankless journey. Thanks a lot.