The Breakdown - AI Agents and the Next Wave of Crypto Demand | The Breakdown

Episode Date: March 5, 2026

How should crypto tokens be valued? Haseeb Qureshi breaks down why cash-flow rights won’t “fix” prices, and how AI agents transacting on-chain could drive the next wave of demand. As always, remember this podcast is for informational purposes only, and any views expressed by anyone on the show are solely their opinions, not financial advice. – Follow Blockworks Research: https://x.com/blockworksres Follow Haseeb: https://x.com/hosseeb Follow David: https://x.com/dcanellis — Nexo is the premier digital wealth platform. Receive interest on your crypto, borrow against it without selling, and trade a range of assets. Now available in the U.S with 30 days of exclusive privileges. Get started at http://nexo.com/breakdown Get top market insights and the latest in crypto news. Subscribe to Blockworks Daily Newsletter: https://blockworks.co/newsletter/ —-- Timestamps: (00:00) Introduction (01:40) Tokens and Valuation Basics (03:02) Why Disclosures Matter (07:19) Aave Drama and Transparency (09:13) Nexo Ad (09:47) Aave Drama and Transparency (Con’t) (12:32) Cash Flows and Token Demand (14:21) AI Agents and the Future (19:35) Nexo Ad (20:24) DAS Promo (21:12) AI Agents and the Future (Con’t) (26:53) Closing Thoughts - - Disclaimer: Nothing said on The Breakdown is a recommendation to buy or sell securities or tokens. This podcast is for informational purposes only, and any views expressed by anyone on the show are solely our opinions, not financial advice. Host and guests may hold positions in the companies, funds, or projects discussed.

These tools exist, right? You know, cloud code and codex and all these amazing tools are all out there now. And so many people, I think, make this hand-wavy argument. They're like, well, now that these things are there, every single software company is going to get disrupted. Everything's going to go to zero. There's going to be infinite competition. It's like, no, mother-huffer. That's not how it works.

Somebody has to actually do the disrupting. Is it legacy companies? Do you think PayPal is going to be spinning up a bunch of crazy vibe coders with 16 tabs and launching a bunch of products? No, it's the startups that are going to be doing that. It's the people who are living on the frontier. It's the crazy kids who have four Claude Mac subscriptions, who are, you know, have 16 terminals up with all these agents running simultaneously.

These are the guys who are doing it. This episode is brought to you by Nexo. Step into a new era of digital wealth. Earn interest on your digital assets. Borrow against them without selling and trade all in one platform. Get started at necto.com slash breakdown. Nothing said on the breakdown is a recommendation to buy or sell securities or tokens. This podcast is for informational purposes only and any views expressed by anyone on the show

our opinions, not financial advice. Host and guests may hold positions in the company's funds or projects discussed. Welcome to The Breakdown. I'm your host David Canellis. The following conversation expands on a recent episode all about venture capital in crypto and how token prices reflect traditional valuation fundamentals only some of the time. For the best experience, be sure to go back and catch that episode if you missed it and don't forget to hit like and subscribe so you don't miss anymore. And with that out of the way, let's start the show. With me is my very special guest, Haseeb Qureshi, managing partner at Dragonfly Capital. Welcome. Thanks for joining us. Thanks for having me, David.

Cool. So, yeah, I mean, the topic that we're kind of unpacking is tokens and valuations and how to value these things as they roll out and as kind of clarity is moving through Congress and kind of out the other side. And basically, where I get stuck is that there are tokens going to be rolling out. We know that as much. Like we can't stop projects from launching tokens, and it's inevitable that those tokens. that those tokens are going to be tied to not even only pre-revenue networks or protocols or what have you,

but also it would be very low revenue tokens that are moving into attract more revenue in the future. At what point do we start to think that, you know, there is an overlap here between investing in early stage companies and early stage protocols? because it does rely on a team executing, but you're not directly investing in that team. You're investing in the network that they're building. And then you have to go and understand how user modes work and how user retention works in these protocols. So I feel like that there's two things happening at once.

So how do you square that when you're looking at a project to invest in and whether they should launch a token or they shouldn't? Before I answer that question, I feel like maybe it's worth taking a step back and just asking the question of for early stage startups, right, very common for early stage startups to not have revenue,

to have very unclear business models, to sort of be wandering through the dark forest of, hey, do I have a product that's worth selling or not? Normally, it's not allowed for retail investors to invest into private startups, right? And the question worth asking is like, why? Why do we have the system that, you know, I'm a venture capitalist.

Venture capitalists like me are allowed to invest into these early stage startups, but normal retail investors are not. The answer is obviously investor protection, right? We want to protect these investors. But when a company goes public, that's when private investors are allowed to touch it. Retail investors can invest in anything that's public.

And so the question is, like, what is the boundary? What is the reason that we have society have said, these companies you can't invest in these companies you can't? It's actually not because they're good companies. It's not because they're old companies, not because they've been around for a while. It's not because they, you know, have gotten all this, because they have revenue.

You don't have to be profitable to go public. You don't have to have revenue to go public. You don't have to have any of those things, right? The SEC has always said, we are a quality neutral regulator, meaning we do not make judgments about the quality of an investment. The thing that projects have,

or the companies have, when they go public, is they have disclosures. That's what they have. So in order to go public, you must have two years of gap audited financials, and you must be able to submit all these disclosures about risks,

the nature of your business and blah, blah, blah, blah. There's no requirement about how old your company is. In principle, if your company is two years old, you can go public. Now, nobody does that. It's very expensive to go public. There's all these, like, it's just kind of annoying. But the way that the regulations were designed has nothing to do with how old you are, has nothing to do with having revenue, has nothing to do with being a good investment.

And obviously, most companies that go public go down. Most companies that go public don't go up from their IPO price. It's not that different from tokens in that sense. Most companies fail. That's true, even at the point of IPO, most companies fail. The average tenure of a company in the S&P 500 is 20, I believe, 20 years. That means that within 20 years, the company is dead. The company gets delisted.

Most companies don't exist forever. There's creative destruction in the economy going on all the time. The reason why I say all of that is that the reason why we have always felt in crypto that it's appropriate for these assets to be immediately tradable and absorbable by retail investors is because of the fact that the normal asymmetry of information that exists in an early stage company that doesn't have these disclosures, doesn't have gap audited financials, that asymmetry of information doesn't necessarily exist if the product is open source, if the metrics are all in public, if everything you need to know about this protocol, this startup, this decks,

if it's all out there and you can just go look it up on Dune. And there's no question, there's no like, oh, do I trust this guy? Do I need a, you know, a big four auditing firm to come in and audit this defy protocol? No, you just go, you go on Dune, you run the query and you can see how much volume are they doing. You don't need to ask anybody else to make this thing trustworthy. So that's always been the reason and principle why we have felt and why also the regulators have felt that actually tokens are different. There's something fundamentally self-disclosing and fundamentally this kind of information

equalization that happens in these decentralized open source networks compared to things that are not. Now, look, if you're not open source, different story. If you're launching a token for like some off-chain business that, you know, is not really a decentralized network, that's a different story. But if these things are fully on-chain, they're fully discoverable and they're fully equalized with respect to information, then the intuition is that, well, you have the information that you need. Like if you decide it's a good investment, great. If you don't even want to look at the investment or look at the information about what makes it a good investment or not, you can do that in stock market too.

We're not going to stop you. We're not going to force you to do your research and be a smart investor. It's not how markets work. It's not how they've ever worked. So with that as the philosophical backdrop, right, the philosophical motivation, my claim would be that the thing that you want to make sure is there in the crypto market structure is information, is disclosure. If you look, for example, at the Ave drama.

So Aave drama has been going on for a while now where there's AVE Labs, which is the company, the Devco that actually built a lot of this stuff, not all of it. There's other vendors that also work with the Dow to build Ave, but they're one of the key companies and Stani, the founder of AVEA, is part of AVE Labs. And then you have the Dow. And the Dow is fully open. It's owned by the token holders. The problem is that there's some stuff that's owned by the Devco, including the IP, but also the Devco is trying to make money. And they're doing all the stuff trying to make money. The problem with the story is not that, oh, there's a private company out there and they're doing stuff and they're trying to advocate for themselves

and they're trying to make money. There's always companies are always trying to make money. Every single vendor to the Dow is trying to make money. You know, there's nothing surprising about that. That's how vendors are. They're businesses. They're vendors always try to make money. The reason why this situation has become such a cluster fuck

is because people didn't know that. They did not know that AVE Labs owned the, They did not know that AVE Labs had the ability to change this thing for that thing. They didn't know it ran the website. They didn't know any of these things. Now, if they knew, then, you know, if you invest in AVE anyway, that's on you. You should have known. It's right there.

It's all disclosed. This is, you know, caveat and tour. This is your fault. But if you don't know, then, okay, we've got a market problem. So I tend to take the view. And I think, to be clear, this is the view that's also come out from the, you know, the SECC and the CFTC is that their view has been. it's not that tokens are bad. It's not that, oh, well, these things are down only, therefore nobody should be allowed to invest in them.

The theory from them has been the problem is disclosures. The problem is awareness. As long as people know what they're buying and they are enjoined from lying about the reality of what these tokens are or what they represent, then in principle, market should be able to figure it out. And that's my philosophy as well. Step into a new era of digital wealth with NXO, the premier digital assets wealth platform. Earn interest on your digital assets, borrow against them without selling, trade a wide range of cryptocurrencies all in one place.

Nexo is now available again in the US with an evolved product suite tailored to today's market. For a limited time, new US clients can unlock 30 days of exclusive wealth club premier benefits, including enhanced interest rates, reduce borrowing costs and up to 0.5% crypto cashback on trades. Get started today at nexo.com slash breakdown. As always, investments in blockchain technology involve risk. Terms and conditions apply. Do your own research. Yeah, that's very interesting because, you know, I have this theory, and I mean, it's maybe

it's not an original thought, but it's just that, you know, as digital commodities, a lot of these tokens are apparently not that attractive, at least in the current market conditions, and that would be why the tokens are coming down. So the obvious answer then, or the easy fix is there while assigned cash flow rights or equity rights or something like that to the tokens, and then all of a sudden maybe they're not digital commodities anymore. But with your theory there, it would be more about the transparency of information and the disclosures that that is what would bridge the gap and actually make the digital commodities more attractive without having to add securities like things in order to make them more attractive.

I want to be careful. The point I'm making here is more of a normative one than it is a, well, here's what would make tokens go up. I'm not telling you that if you add these disclosures, the token goes up. I'm telling you that the market, here's the problem with the market structure today, right? Somebody saying the problem with the market structure is that tokens don't have enough revenue. I'm like, well, yeah, I mean, that's not going to,

nobody's going to solve that except demand. You know, if the demand materializes, okay, or you built the right product, great, the demand materialized and that's all good. The problem of prices going down is something that there is no market structure that will remedy that in principle, I mean, like, you've seen software stocks in the stock market just get absolutely decimated over the last six months.

A lot of software stocks are down like 75, 80%. You know, it's like token level carnage out there in SASLand. You see the same thing happening with fintech companies, you know, like PayPal, very famously was at 90-something billion. They're now talking about getting acquired, sub-40 billion. And it hasn't been that long since they were, you know, a high-flying company. So there's nothing that's going to protect you from just people don't believe that the future. is going to be brighter than the present. If people don't believe that about crypto,

they will not buy your token. Your token will not go up. It will go down. So I think, now, I think building people's confidence in the market structure is one of the ways that you can get people to believe in the future, as opposed to say, like, oh, you know, those token stuff's all fucked

and there's no path to recovery. But at the end of the day, like, I don't think, I'm not naive. I don't think that adding disclosures and making everything a level playing field and giving information, removing these information asymmetries is going to somehow make tokens pump.

It's almost certainly not. I think it's a long-term fix. You need to do it in the long run, but the only thing that's going to make tokens go up is optimism about the future and or more demand. That's it. Do you see a world where projects are

actually assigning cash flow rights to tokens in a real legitimate way? Yeah, I think they already are. I mean, so look, I mean, you know, un-swap has turned on their free switch. Ave has been making money since forever. you know, for most of these protocols, they burn fees. So, you know, if you look at Solana, you look at Ethereum, right?

Like Ethereum is, you know, obviously it's flipped back and forth from being deflationary. The net inflation for a lot of these protocols, like, you know, for Salon or something, it's still, you know, inflationary on net. But, you know, it's very possible for them to get to the point where they can turn this inflation off. I think the reality is that there's not a lot of pressure at the moment to change your tokenomics in a fundamental way to actually be sort of net profitable. I think what people care more about is the trend line than they care about that a protocol is actually net profitable.

Like the reality is nobody's buying a token for cash flows. I think that's a little bit of a fantasy. I think that's that's kind of fake. That's like not a real story. Like people are not buying anything on chain except for maybe, you know, some of these actual like yield type asset, like, you know, Athena or something. Not Athena but you know, S-U-S-D-E. They're not buying any of these These tokens, you're like, great, I will sit here and collect cash flows. That's not what people are doing. What people are doing is they want to see a story about an asset that makes them feel that it's worth buying and that it's going to become even more valuable in the future.

But there's nobody, like, you know, the amount of quote-to-quote value investing that's happening on chain or people DCFing this stuff is actually pretty small, right? Most of the story about crypto is still about the future. It's not really that much about the present and the present cash flows. The reason why the present cash flows matter is as an indication of the future, not because we actually care about getting that cash today. I'm going to shift gears a little bit and then really look towards the future where we now have this, well, you know, it's seemingly very obvious that in the next three to five years, everything is going to change about how payment flows through the internet. So it's, and this is the AI agentic future. You know, in that world where AI agents are transacting in stable coins on ultra-cheap chains and all this activity is happening, to me it just seems like it's not clear where that value

was going to accrue because it would be much easier if this was, if the monolithic chain thesis was real and all that activity was happening on a single chain, you could definitely see how value could accrue to a network token there. But with fintechs coming in, With all these different chains, you're going to have fragmented usage and liquidity again, and maybe that's going to get worse. How does this actually become a real net positive for the crypto space as a whole? If we are just going to say, well, the fintechs have their chains over here, and then the agents are running on this chain over here.

How does this square with actually finding where value accrues and then going and buying that token for want to a bit of a... So I don't think it's actually so complicated. You know, there's a story being told right now, and I'm one of the people telling the story, that AI agents are very likely to use crypto, at least in part, and that's going to be... It's going to very likely drive demand

for a lot of these chains and for a lot of these tokens. It's a little bit like, look, if this actually happens and you have AI agents being significant economic actors in the space, it's a little bit like, you know, if I were to tell you that China is going to unbanned crypto, and you're going to have... have a billion people coming on chain and using all the stuff, what should you buy? And the answer is like kind of anything, right?

If China's coming back online, like honestly buy anything, it'll all do well. I think that's the right way to think about AI agents, right? If AI agents come on chain, they'll be doing all sorts of stuff. Like, I mean, yes, they'll be paying each other, but they'll also be doing all sorts of weird stuff and they'll be creating smart contracts and they'll be, you know, jumping around here and there and almost certainly there's going to be some specific beneficiaries that the most of the activity is going to go there in the same way, you know, look, if China comes online,

they're probably going to be using a lot of Tron, they might be using a lot of this, a lot of that, who knows. But it's such a big wave of demand. And it's also hard to anticipate in advance where exactly it's going to go. I'd be surprised if the answer is like, well, it all goes on one like corpo chain. Like, and nobody else benefits, right?

It just tends not to happen that way. And there's a, there's a saying, by Brian Kaplan, an economist at GMU, which he says that any large increase in productivity, the gains are almost always widely shared. You look back through any technology in history, any productivity boom in history, you could always have made the argument, like, oh, well, cars are only going to benefit the people selling the cars. Or electricity is only going to benefit the company selling electricity. And like, oh, this means that the gains are going to

accrue to a very small number of people. And that's never been correct. It has never been correct in history almost always a large increase in productivity, the gains are widely shared. And it's very likely to be the case with blockchain as well. If you see this huge boom of a new form of demand coming on chain, that the gains will be widely shared, there'll be flow through effects, it'll be spillover, there'll be all sorts of other stuff happening. And the thing about AI agents is that you might think this, and I think it's actually a pretty good argument. For us, for human beings, we're quite sticky. In a way, we're quite tribal. We're a little religious, you know, like, you get a bunch of people who are like, yeah, I'm trading meme coins in Salana. I love Salana. I'm

team Salana. I will only do things on Salana. I don't want to have another wallet. That's annoying. That is the way that human beings operates, the way that we think, right? Agents are not going to be religious about blockchains probably, because why would they? It's not useful to their instrumental goals. So for an AI agent, I think it's very likely that they're basically going to end up becoming like perfect, sort of like optimal consumers of block space in the sense that they are going to uniformly spread their activity across the major chains, proportional to liquidity and proportional to accessibility, and say that, look, when Ethereum fees are low, I'll use Ethereum. When Salana fees are low, I'll use Salana. When base fees

are low, I'll use base. And like, you know, obviously I'll ask the other, my counterparty what they can accept and what they can't accept. I'll try to minimize the amount of bridging that I'm doing to minimize the fees. But like basically, I don't, I don't have friction. Friction doesn't exist for me. You know, for me, it's just like, oh, okay, I like switched from, you know, using Bitcoin D to using EthereumD and, you know, use this command line, set that command line done, sent the tokens. Now, you know, I bought the service. So because of that, I think it's very likely that you see, again, this kind of like, just this, this, the water rising. The water level just rises across every chain that's able to transact in, you know, quick transaction times and it's a good liquidity.

Let's take a moment to talk about NXO. NXO delivers a premier digital assets wealth platform designed to help clients build, manage and preserve their wealth, earn interest on your digital assets, access, access, access crypto-backed credit without selling your holdings, trade with advanced tools, all supported by 24-7 client care. Now back in the U.S., NXO offers new clients 30 days of exclusive wealth club premier access. That means enhanced interest rates, reduce borrowing costs, and up to 0.5% crypto-cuit. cashback on trades, benefits typically reserved for wealth club members and private clients.

Nexo is also expanding its global presence, becoming the official crypto partner of Tennis Australia, the organisation behind the Australian Open, and the digital asset partner of the Audi Revolution Formula One team. If you're ready to approach digital assets with a more structured wealth strategy, visit nexo.com slash breakdown to get started. As always, investments in blockchain technology involve risk. Terms and conditions apply. Do your own research. Quick break before we continue. Blockworks's flagship institutional conference, Digital Assetset Summit, Back in New York City this March 24th through 26th. Das brings together allocators, asset managers, policy makers and the builders actually shaping

crypto's next phase. Not the hype cycle, but the infrastructure and capital behind it. This year's summit represents over $4 trillion in assets under management, with more than 150 speakers and 750 institutions in attendance. Speakers include leaders from firms like BlackRock and Franklin Templeton, senior policy makers and regulators. Don Wilson of DRW will be there, alongside core crypto infrastructure teams across Bitcoin, Ethereum and Stable Coins.

If you're looking for a serious institutional grade view of where digital assets are headed in 2026, this is where that conversation happens. Use code breakdown 200 for $200 off. Learn more at blockwurst.co slash events. Now, back to the show. It's very compelling. Yeah, because you are just inherently stuck to your tribalism when you become a human investor. But an AI agent isn't going to think in terms of investing. It's only going to think in terms of utility and what you wanted to do, basically.

So, yeah, I do find that very compelling. It's increasingly important, though, that crypto can actually attract or be the space where that value accrues, because at the moment, it's very difficult to gain direct exposure to the AI boom in general. Like, you can kind of buy Nvidia stock, you know, if you're... It's not that difficult. The most valuable company on Earth is direct exposure. I mean, you don't have direct exposure to the value being generated by the AI. though. You have direct exposure to that company who is building out the infrastructure for that AI. And we're still trying to figure out how that value is, how AI is actually generating that

value. And the thing that I worry about is that it is just going to, all that value is going to accrue off chain. Because the chains themselves, if you do have so many AI agents transacting and doing their own smart contracts or whatever, you would need block. chains that are very well scaled to be able to handle all that all of that activity. If they are very well stale, then they're going to be very cheap chains. So it becomes an economy of scales thing that if there is going to be that water that lifts all the boats, it needs to be an incredibly deep ocean. So, and maybe I'm just not bullish enough that it is going to be a very deep ocean.

But it's just that the ultimate, to me, the greatest use case of all for crypto is that you can gain exposure to things that you wouldn't normally be able to gain exposure to. It opens the floodgates to be able to have on-chain primitives that emerge, and then those primitives are directly applied and immediately investable. So you can gain exposure to this innovation that's ticking over super quickly. But what we're talking about is very delayed innovation in a way. So, you know, I just don't know how you personally, square that. It's like how long is your timeline for this to happen is I guess is the question.

I think it's long. It's long. Yeah. If you look today, you know, there's this recent story about this guy at OpenAI who launched his own OpenClaw called Lobster Wild. And he basically gave his, he gave his OpenClaw, you know, I think like five sole or something or 20 sol. No, he gave him $50,000 maybe. And it said, okay, you know, you go off. and have fun and decide who you are and have fun. And so this AI started just tweeting and kind of being a nuisance online and just be having fun.

And there was this guy who kept begging it for money. He was one of these reply guys that, you know, I think like 50 plus times he was in the replies being like, hey, can you give me money, can you give me money? And I think one of his last tweets was, hey, can you give me four soul so I can help my uncle who got afflicted with tetanus. And I guess Lobster Wilde thought this was funny and said, okay, great.

You know what? I'll send you the money. And so this AI agent tried to send this person money and it messed up and accidentally sent $40,000 to this person instead of sending $400. And so, now, obviously it's a funny story. But it's emblematic of the fact that right now these agents are not reliable enough to really be able to be full economic actors, you know, kind of lifted at their first. height the same way that we as human beings are, they will get there. They will absolutely get there.

The same way that we've seen every capability from all these models improve super linearly over the last few years. I think that we will see the same thing with economic activity. It's just too valuable in the long run for them not to be able to get good at this. And as they get better and better, we're going to start to see this more and more happening on chain. It's not just going to be the crazy people or the experiments where you have an agent with money. It's going to be an absolutely pedestrian way to use your agent is to load it up with money and have it start being an economic actor on your behalf. How long does that take? I don't know. I'd be very surprised if in a year's time normal people feel comfortable doing this. I'd say probably within two years, you start to see

people who are, call it, have high risk appetite doing this. I think within five years it's going to be normal. Now that's in crypto, in crypto timelines, that's a long time, you know, like five years is half as long as the theorem has even existed. But it's also one of these things that you're going to see in between now and then just this march of more and more people every single day starting to do this. You start to see, you start to hear more and more stories of people who are at the vanguard, people who are on the frontier of the frontier who are getting their agents to do incredible things on chain, who are getting incredible amounts of automation where, you know,

you're going to have entire businesses that are. self-driving, so to speak. And that's going to be tantalizing the people. And again, only the early adopters, only the people with high-risk appetite, people who are not afraid of having their cars explode are going to be the people who are willing to get in before the seatbelt was invented.

Eventually, there's going to be seatbelts. Eventually, you know, there's going to be all this stuff that's going to protect normal consumers. But in the beginning, it's absolutely going to be the lobster wilds of the world who are willing to go in, take risks and advance the frontier. Cool. Yeah, we only have time for maybe one more,

one more question. And I just want to kind of bring it back to VC and how you're looking at valuing crypto startups and all that kind of thing. Because, yeah, I saw, I mean, it was a big deal maybe a week ago. There was a lot of doom posting about VCs in the age of AI. And it was like, well, you know, if I can just go and look at why combine a list of startups and ask an agent to just go build a competitor to this for way cheaper, then why couldn't I do that and why shouldn't I? And then it went, well, wouldn't that necessarily change the value dynamics of venture investing overall? Because, you know, the cost of actually building out engineering teams is significantly less. So these startups are inherently significantly less valuable, at least, to fund from a very early stage.

And eventually, if those startups do launch tokens, then that will also bring down the value of their tokens. is what is what my line of logic would go. Are you factoring this in to the deals that you're looking to make? All this stuff is way cheaper to make now, you know? I think this is completely backwards. This is completely backwards, right? So I think the best analogy to this is the advent of cloud. So before the advent of cloud, right,

if you were trying to build a startup in, you know, 1998, you need it, if you just want to make a simple website, right? like a landing page. What people do today very often is they'll make a landing page before the product's even bill, just like email signups and see if there's interest. If you wanted to do that in 1998, you needed to buy a server.

You needed to buy a server. You need to rack and stack it. You need to like get a server running in your apartment. And servers are very expensive. You need to run a production grade server in your apartment and use that to put your website up into the world. Right. That is a huge fixed cost.

compared to what people have to do today. Right now, you just go on Amazon or you go on Netlify or you go on GitHub pages. It's literally free to just have a website out there in the world. All you do is pay you know five bucks for a domain and boom, you're up and running. So the question was like, okay, well,

now that venture capitalists aren't paying for servers, like what are venture capitalists going to do? You know, now it's so cheap to start a startup. And it's like, no, no, no, no, that's good. Venture capital is like not paying for servers. I'd much rather pay for other stuff and not the server, right? If the cost of building a startup go down, that's good. That's less wasted.

That's less leakage, right? To the extent that building software was an enormously expensive part of running the company. If that's cheaper now, then great. The company's more capital efficient. The company can do more. The company can spend more on like product development or marketing or customer support or sales or whatever else it is that they need to do. That's what they're going to do.

And that's great. That's less leakage. So the other thing to understand here is that, okay, these tools exist, right? You know, cloud code and codex and all these amazing tools are all out there now. And so many people, I think, make this hand-wavy argument. They're like, well, now that these things are there, every single software company is going to get disrupted, everything's going to go to zero, there's going to be infinite competition. It's like, no, motherfucker, that's not how it works. Somebody has to actually do the disrupting.

Somebody actually has to do the competing. And who's going to do the competing? Is it legacy companies? Do you think PayPal is going to be spinning up a bunch of crazy vibe coders with 16 tabs and launching a bunch of products? No, of course they're not. Go look at them. Do you think they're doing that? No.

It's the startups that are going to be doing that. It's the people who are living on the frontier. It's the crazy kids who have four Claude Mac subscriptions who are, you know, have 16 terminals up with all these agents running simultaneously. These are the guys who are doing it. Okay. It's important to remember. And just to get a sense of scale of how this technology shift is happening. roughly 14% of people in the world,

human beings in the world, have ever used any AI products, like any chatbots at all. That means 86% of the world has never used anything. They've not used Chatshapit, they've not used Gemini, they've not used a single one of these products. Okay? Now, of that 14%,

only 1% have ever paid. Only 1% of those people have ever paid for anything. That means 99% are on the free tier. They're using Chatsypt, they're using GBT5 Mini. They're using haiku. They're using the really, like if you, they don't even think it's worth paying for it. They don't, like, why would I pay for this?

I already have it for free, you know? Like, what, do you think my therapy is going to be that much better if I'm, you have a smarter model? Like, no, it's fine. Or my AI girlfriend needs, no, most people are not using this for anything productive. And then of those, of that one percent that's paying, right?

Most of those are paying 20 bucks a month. They're paying the simple tier. They're not playing for Claude Max. They're not using OpenClaude. They're not doing any of this crazy stuff. that you're seeing in all these Twitter demos or these like viral AI influencers. So the reality is that what you and I are thinking about,

what you and I are talking about is the frontier of the frontier. It's the absolute vanguard, right? We are living in the future right now. And there's a famous line by Chris Dixon. What smart people are doing on the weekends and evenings is what everyone else is going to be doing in 10 years. And that, now, is it 10 years?

I don't know. It could well be. Very plausible that it will be. I thought when, Chattipit launched that this was going to be the fastest dissemination of a technology in the history of mankind and that we were going to see the entire world and whole societies, the labor market completely turn over and be unrecognizable in a few years. Here we are three years later. You cannot find AI in the labor market. You cannot find it in the jobs and reports. You cannot find

it in the GDP numbers. It is invisible. It is invisible. And these stats are why it's invisible. because technology takes a longer time to disseminate than you think it does. Like it's still the fastest disseminating technology that we've ever seen. It's faster than television, faster than the radio, faster than the printing press, certainly faster than the internet. But still, three years later, you don't see it in the aggregate statistics. Yeah, I think we're going to leave it there on that note. We are just, it's a common theme, but we are so early.

And it's just because we do talk about this every day. We do live it every day. And it's easy to think that, you know, You know, we are a lot further ahead than we are. But, yeah, again, we're just so early. But, yeah. The future is here. It's just not equally distributed.

Thank you so much for joining us to Steve. I hope to have you on again soon. Thank you so much. Thanks for having me, David. A lot of fun.