The Stablecoin Chain Wars: Codex’s Bet on On-Chain FX | Cofounder Haonan Li

Haonan:
[0:00] Most cross-border transactions involve an FX component.

Haonan:
[0:04] And so if you are the cheapest place for FX to happen, that chain will be a natural magnet for all these stablecoin flows. We want to be the cheapest place for T to C swaps. You want to be the cheapest place, fastest place to do this fiat stable swaps as well. This is not what typically happens at a chain company. A chain company classically doesn't want to think about fiat. Why would you think about fiat? I'm at a crypto company. Why would I think about fiat? Now, if you think about the reality of these systems, like how do you make it so that stable coins aren't just useful to like people in crazy countries and like crazy situations? Well, you need to make the friction go down to zero. You need to make it so that moving between a unit of fiat and stables is frictionless on a cost and speed basis. Need to delete the border between fiat and crypto. So that's what we are focused on. We're staffed against that. We wake up every morning, we think about this. How do we delete this border? Make it go to zero.

David:
[0:58] Bankless Nation, I'm with Hounan, the co-founder and CEO of Codex. Codex is a stablecoin-focused Layer 2 on top of Ethereum. Hounan, welcome to Bankless.

Haonan:
[1:07] Hey, David. Good to see you.

David:
[1:09] So you are working in a pretty competitive part of the market. There's Stripes Tempo, there's Circles Arc, there's Plasma, there's Stable, not to mention the deep incumbent Tron. Competitive. Are you intimidated?

Haonan:
[1:23] You know, I think there's a ton of competition if you think about it from the perspective of a VC PowerPoint slide, right? If you had a category that said Staplecoin chain, you would have all these logos. I think what that abstracts away is actually most of the reality and the detail underneath. And the reality is that these different chains are pursuing totally different directions. And I can just say for us, I can't speak for others, when we are going to market against our customers, we don't really hear about any of these other chains. We don't hear about plasma and stable because they've taken this sort of maybe more bare chain type approach. We don't hear about arc or tempo. Perhaps it's perhaps not quite clear exactly what it is they want to do yet. And so we just see all these charts going up and to the right, and we're growing a lot. And customers seem to like what we are doing. And, you know, we just want to focus on the customer.

David:
[2:19] Maybe you could teach me and the listeners a little bit about what makes these chains so different from each other. Because like if you, I think you talk to the average person in crypto, they're like, oh, yes, there's just this whole payment chain sector where we're all trying to disrupt Visa. and we're going to put the stable coins on the very fast blockchain and that's going to become the payments chain. And that's what all of these, all the chains that I just named, that's what they're all going after. Again, Tempo, Arc, Plasma, Stable, Tron has been the payments chain, the stable coin payments chain. Is there more nuance there or like what are the construction differences between all these chains and how they point to different outcomes? Maybe you could educate us a little bit there.

Haonan:
[2:58] Yeah, I think there's multiple layers to it and there are a lot of differences

Haonan:
[3:01] across multiple layers. I think, let me start with saying that it's such a new category that nobody really knows what they are doing quite yet, right? They're all exploring this idea maze, this giant design space, and people have chosen different parts of the maze to start with. So, you know, what is the approach of, say, a plasma or perhaps a stable, right? The approach is the Baruchain playbook.

Haonan:
[3:23] So what is the Baruchain playbook, right? You do a token event and you attract a bunch of flows that come onto your chain. Of course, AbleCop, right? You're giving away free money. Then you try to make it stay because if it leaves, you know, you're screwed. So how do you make it stay? Well, you fork over a bunch of DeFi stuff that exists elsewhere, the same, you know, 10 apps that exist everywhere. And you sprinkle token incentives on top and you hope to God it stays. The problem with this approach historically is that nothing of differentiated or substantial results from this, right? And of the day, you've got the same DeFi apps. And so what ends up happening is people dump the tokens. These are mercenary farmers, right? We are far, far away from the days of like the uni airdrop, right? This is like industrial scale farming type dynamics. People will dump the token, the price will go down, but you still need to keep the money there as much as possible. So you give away more tokens to hit the same APY, people dump the tokens more, the price goes down, and then eventually it all goes to zero, right? Or near zero. So we've seen this movie play out before with a bear chain or a blast. Will it play out this way this time around? It's not clear. Perhaps it won't. It's just a very different approach from maybe our way of processing the problem.

Haonan:
[4:45] When it comes to a tempo or an arc, I think both these guys run into a ton of problems when it comes to neutrality. So imagine that you are, and here I like to say like neutrality is not just like an Ethereum prayer, right? It's not this like religious word that we say to each other at these Ethereum parties,

David:
[5:02] But like it actually means something. Although it might come off like that, yeah.

Haonan:
[5:05] It comes off like that, right? It's almost like the word deluxe. It just means good to a lot of people. Like deluxe cheeseburger is just a good cheeseburger, right? But neutrality means something. What does it mean here? It means if you are a competing fintech from Stripe, do you really want to bet the future of your business on Stripe chain? If you are a competing issuer to Circle, do you really want to bet the future of your business on Circle chain? It's hard for me to imagine. Imagine the CEO of a large fintech going to his board and saying, hey guys, we really fucked up. Stripe's way ahead of us. We're just going to go use their thing. I don't know. I don't, it's not very sort of credible to me. So that would be maybe my first cut at distinguishing all these different approaches.

David:
[5:52] Isn't that the neutrality aspect, isn't that just true for all of these chains? All brand new chains in my mind have on day one, have the most compromised neutrality that they'll ever have. And then ideally they become more decentralized, more equitable, more spread out over time. But ultimately any brand new chain is always just not neutral. It's just by the truth of the fact that it's like brand new. Like there's all of these have VCs. All of these have early insiders.

David:
[6:22] Like ultimately it's the neutrality thing. Neutrality is not really something that you can engineer.

Haonan:
[6:26] There's neutrality and then there's neutrality, right? There's sort of like compared to say a Bitcoin, you know, some would argue Ethereum is not neutral, right? But I think that's somewhat of a pedantic point. I think the neutrality that really matters in the market today, instead of like this kind of pseudo religious thing that people might have on Twitter, the neutrality that really matters in the market today is are you really going to bet the future of your company on a competitor's infrastructure?

David:
[6:50] Are you really going to see that control to them?

Haonan:
[6:52] And does that really sell at the executive level? And I think what we are discovering, what we are seeing is that it doesn't really sell. Nobody wants to use a competitor's substrate. I think another way of saying this is like, let's take us, for example, Codex. We don't run an issuer. And we certainly don't run a fintech. We are not going to be competing against our customers.

David:
[7:15] So isn't that so that same aspect is probably also true of plasma and stable and you're just saying it's not true of arc and stripe because stripe is stripe or tempo because stripe tempo has stripe and then arc has circle they're they're the issuer but and then but then what your critique is of plasma and stable is like well they're kind of doing the bear chain thing as well which is separate critique, but at least they have the neutrality going for them is kind of like a summary.

Haonan:
[7:44] I think that's fair. I think like if you dig in, like they're mostly focused on tether, right? So is there a really issue of neutrality there? It's, you know,

Haonan:
[7:51] somewhat debatable, but I think, you know, the main thrust is correct. Yeah.

David:
[7:55] Well, you're highlighting the differences that you're highlighting. It's kind of like to go to market, but ultimately are these all the same things at their end game? Like, do they all converge on the payments chain, landscape. And that part is actually the common denominator that there aren't any nuances there. They are actually the same kind of outcome that everyone's going for. Is that true?

Haonan:
[8:14] It's not clear to me what the end state here will be necessarily. I think certainly in this first sort of leg of the race, people are taking totally different approaches. And They're building different products, really, right? Because if you think about it, you're like, okay, I'm going to make a change for stablecoins. Great. What does that mean? What matters? What doesn't? Like, here's one thing that could matter, right? You could say, well, I think it's really important that people can pay for their gas and all kinds of stablecoins. So I'm going to enshrine a DEX. And, you know, this way, I think this is super, super important. And obviously, if I enshrine a DEX, then I've got to really make sure it's robust. And how do I figure out this, like, the effects mechanics of that and how do I make sure that doesn't break? You could do that. You could say that, you know what? I think the most important thing is to build a neobank. So I don't keep too close with what Stable and others are doing, but last I heard they were doing something like this, right? What is a neobank? It's like a payment app with Stable coins in the back. And you say that's the most important thing to build. And so you're going to pick up these little flows one by one and you put it on your chain. Do that. Or you could do what we're doing, which we think is much, much smarter.

David:
[9:23] What's that thing? Yeah.

Haonan:
[9:26] So I think what we want to focus on is wholesale FX flows.

David:
[9:32] Okay.

Haonan:
[9:34] Why is that? Well, because most cross-border transactions involve an FX component. And so, if you are the cheapest place for FX to happen, that chain will be a natural magnet for all these stablecoin flows. And so, that's what we're focused on, right? We want to be the cheapest place for T to C swaps. We want to be the cheapest place- T to C? What's that? USDT to USDC. The stablecoin people don't want to keep repeating USDC. Yeah, I understood.

David:
[10:02] Okay.

Haonan:
[10:04] Those cheap swaps, you want to be the cheapest place, fastest place to do this fiat stable swaps as well. And so this is not what typically happens at a chain company, right? A chain company classically doesn't want to think about fiat. Why would you think about fiat? I'm at a crypto company. Why would I think about fiat? I want to think about shared sequencing schemes. And I'm going to rotate the sequencer and all these- Ivory Tower stuff.

Haonan:
[10:29] Yeah. Much more interesting from an engineering perspective, certainly, right? But if you think about the reality of these systems, like how do you make it so that stable coins aren't just useful to like people in crazy countries and like crazy situations? Well, you need to make the friction go down to zero. You need to make it so that moving between a unit of fiat and stables is frictionless on a cost and speed basis. You need to delete the border between fiat and crypto. And that means like understanding the fiat infrastructure, right? Which again is very contrarian. So that's what we are focused on. We're staffed against that. We wake up every morning, we think about this. How do we delete this border? Make it go to zero. I think the

David:
[11:09] Conversation we're drifting into is the golden goose conversation. That's what I'll call it. And each one of these chains kind of has like what their perceived value is. And I think the most basic way to articulate this value is like kind of similar to like the Solana idea of just like a ton of volume at very low fees. And stable coins clearly is the killer app after Bitcoin, right? After like non-sovereign store of value. And then we get stable coins. Maybe DeFi is like our second killer app, then stable coins. And so stablecoin transfers, payments, remittances, the same words that we have used in crypto for a decade now. It's like, now the time is actually here. Let's put all the flows on chain. And then what does Tempo and ARK and Plasma and Stable and Codex want to do? They just want to transfer a ton of stablecoins and charge half a BIP on a trillion dollars of flows. And I think that's like the level one, like bull case articulation, getting a little bit more nuanced than that. It sounds like what you're saying is like, yes, and there's a whole FX market

David:
[12:16] that is extremely large that we are also, we want to go for. Maybe you could just like color in the FX golden goose. It's like, why is that so big? Why is that the logical conclusion for you guys? Just like help me imagine that a little bit better.

Haonan:
[12:31] The ERC-20 sign of this is the simplest thing. Like, you could throw R&D at it and make it faster, better, whatever, sure. But that's not really the problem, right? The problem is the friction going from fiat to staples. So what you'll hear a lot, VCs, observers, like Twitter thread people, right, is this concept of stablecoin sandwich. It's like you start off in fiat in country A, you turn it into stable coins, you send, you know, it's an on-chain send, then it comes out in fiat the other side and this is great, right? This is the story. The reality is across most countries today, it doesn't make any sense whatsoever because they're too expensive. The fiat to crypto conversions are too expensive.

David:
[13:13] And so- The on-ramp cost, right?

Haonan:
[13:15] That's right.

David:
[13:16] Right, because the only people that have truly figured out the on-ramp to solve the on-ramp problem are exchanges because A, they have the ability to lock up your money and B, like if you do like a chargeback, right? If you like do a transfer and then you like yoink it from Coinbase, it's kind of like a loss leader. So they just kind of like lock up the money and they have some sort of assurances, but then they also have like other money generating valid products on the actual exchange to like, it's kind of a customer acquisition tool. The pure on-ramp startups that have just been in on-ramp have been a terrible business because they have to deal with chargebacks. They can't charge ancillary fees. They can't upsell their customers. And so the only people that have to really solve the on-ramp problem are exchanges because of their unique positioning. And I think that's kind of what you're saying is just like this fiat to stablecoin pathway is a hard problem to solve. Are we talking the same language?

Haonan:
[14:12] That's right. I think it's an incredibly hard problem to solve and it's thorny, right? It's not just like pure research task.

Haonan:
[14:18] This is like, you got to go figure out the fiat infrastructure. You're going to have open bank accounts. You're going to have to get licensing. You're going to have to deal with all this gunk that nobody wants to deal with. We deal with that gunk. For us, it seems to me that instead of optimizing the sends, which already work quite great, what you really need to do is deal with the actual problem. So we talked

David:
[14:37] About the chain fees and you said, okay, yes, all chains collect the chain fees. That's the normal business model. Is what you're saying is dealing with a gunk the actual way that you guys are trying to like, is that your golden goose?

Haonan:
[14:49] That's right. That's right. Because, you know, people get to charge fees for solving customers' problems, right? And so if you are in the business of saying, look, guys, like I'm going to make your URC 20 cent, not 10 cents, but now like nine cents. And the dude sends on average a million dollars per transaction. He's going to look at you and go like, OK, cool. Like, all right. That doesn't do anything for me. Right. This is this is, I think, what not enough people do, which is like, go speak to the actual customer. What does the customer want? What the customer wants is for us to deal with this gunk. And so we're going to focus

David:
[15:25] On this gunk. Isn't that what Circle does? Circle has banking partners, right? And so I would imagine transferring my money to Circle and then receiving Sablecoins would be trivially easy compared to any other payment chain. How is this different?

Haonan:
[15:38] There are all sorts of problems with Circle's banking infrastructure. I think to begin with, there are volume limitations, certainly much imposed by Circle themselves. Even then, even if you strip out the volume limitations, much of these transactions happening on a T plus two basis because the banking infrastructure is not adjacent. It's very ironic, right? Like the sort of 10,000 foot level view of stablecoins, the, I don't know, Jason Kalkanis' view of how stablecoins works is that everything is instant and everything is fantastic and everything just works. But you dig into the actual details and none of it works like that. A lot of this stuff is still happening on a T plus two basis, even to mint a stablecoin, which is crazy.

David:
[16:24] Yeah, but doesn't that also apply to whatever you're doing?

Haonan:
[16:27] No. So I think what we are focused on is eliminating all these frictions, right? We want to make it so that folks can access stable coins on a wholesale basis on an instant basis, sub 10 minutes, and on an extremely, extremely cheap basis. That's kind of what we're focused on. And we want to also have the FX infrastructure so that we can shift FX from being off-chain to being on-chain. On-chain FX is this kind of holy grail that people have been trying to pursue for a while. And our diagnosis is that we have the pieces today to actually make it happen.

David:
[17:00] What are those pieces? I would imagine what you're saying is like, okay, so you have the banking infrastructure in the United States, but if you want on-chain FX, don't you need to get the banking infrastructure like fucking everywhere across the whole world? Seems hard.

Haonan:
[17:13] Yeah. I think there's some secret sauce there, David, so maybe I won't dive into all the details. Okay.

David:
[17:19] That means I'm pressing on, that means I'm going in the right direction.

Haonan:
[17:23] You're a good interviewer, David. But I think, you know, the high level intuition is that the classic problem in onto-net-fx historically has been that non-US dollar stables have no liquidity. So you can't move, you can't really move volumes for them. And that's problem one. Problem two is that you have no demand for it. Like who the hell wants to move money through this on-chain mechanism if the spreads are wider, if it's costlier, if there's technical risk, like why the hell would you want to do it? And so for us, we have solutions on both sides. Yeah.

David:
[18:00] Okay. And I suppose this is where we wrap back around to the neutrality conversation, because not only do you want to be neutral for all of the reasons that you talked about earlier, but you also need to be neutral because there's going to be a ton of issuers of not just dollar stable coins, but I would imagine if you're going to do on-chain FX, you need,

David:
[18:18] all of the fiat currencies that have volumes in the FX market. I don't know what those are, but I would imagine there's at least six of them.

Haonan:
[18:25] Yeah, there's a number of them. I think there's a couple of defining characteristics they have. some are run by great people, high quality people. Some of their infrastructure tax are not bad. None of them are product markets.

David:
[18:35] The euro, Japanese yen. And what are the other obvious currencies that you need to get onto Codex?

Haonan:
[18:42] We're really interested in exotics. The more exotic, the more interested we're in.

David:
[18:45] Oh, interesting. Okay, so like long tail developing country currencies?

Haonan:
[18:50] That's right. That's right. We're really interested in that.

David:
[18:53] Because those are the ones that want to get out the most, the most, is that right? As in like if I have Argentine peso, So like, please connect to me to some other fiat currencies on the blockchain. Is that kind of the idea?

Haonan:
[19:04] Well, we're seeing a couple of things. I think to give away some alpha here, I think one thing that we are seeing is that you're seeing this shift in how commerce works globally. So in trad rails, almost all FX transactions go through the dollar. So, you know, you might be going from, I don't know, Malaysian ringgit to Singapore dollar. And actually you go first through the US dollar. What you're starting to see on chain is that that is breaking down.

Haonan:
[19:34] And folks are swapping directly between currencies, which is a radical restructuring of how markets work and how commerce works globally.

David:
[19:43] So if the United States dollar is currency A, then people are swapping from currency B to currency C without going through currency A.

Haonan:
[19:51] That's right. That's right. And that actually is all sorts of interesting. National security as well as like,

David:
[19:56] You know, government implications. It's the opposite of what Scott Bessent wants.

Haonan:
[20:00] That's right. That's right. And if you zoom out here, like what's happening? What's happening is that you have this, stable coins so far have been a dollar phenomenon, right? They've created this wonderful structural buyer of U.S. treasuries. It's sort of this kind of modern euro dollar, right? And it's been really great for U.S. government. And I think the Genius Act was pretty smart, right? The whole point is, let's try to move more of these offshore flows onto stable

Haonan:
[20:26] coins, because again, there are structural buyer US treasuries. I think other countries see that. And while they're like, interesting, I'm not sure if it's necessarily in my interest that stable coins be only a dollar phenomenon. And I think you're seeing various pieces move around on the chessboard now.

David:
[20:44] My imagination is going, okay, so you said exotic currencies, you didn't really.

Haonan:
[20:48] Give me too much

David:
[20:49] Clay to go on. But I think if we're also understanding the concept of neutrality and large FX flows, and maybe an alternative interpretation of the word exotic, is, if we're truly thinking of cypherpunk neutrality here, it's like dollar to Chinese yuan and Russian ruble. The three economic superpowers on the same chain, which offends all three of them equally, and then the Ethereum cypherpunk is like, oh yeah, that's what neutrality is. You don't have to comment on that. But I do want to actually talk about neutrality and the Ethereum context. So Codex is the only payment chain, stablecoin chain, that is an Ethereum layer too. And right now, the broad crypto audience might actually say, oh, that actually makes you guys less neutral. That makes you guys more opinionated, actually, because you're an Ethereum layer two, not a layer one. And between the pendulum of valuations that seems to swing back and forth between layer ones and layer twos, the premium, the layer two premium has given way to the layer one premium. And so we are we're currently in the stance of like, oh, there's a layer one premium out there. And the layer two premium is not really here. And so talk to me about like, why build as an Ethereum layer too? What does that give you guys? And why make that trade? Why make that trade off? I think there's a.

Haonan:
[22:12] Logical case and there's an emotional case. If I'm candid about it, I'll start with the logical case.

David:
[22:17] First of all, this L1.

Haonan:
[22:18] L2 premium thing is kind of dumb because it swings and really what drives it is like product market fit, right? Who has product market fit? That's what matters. If it's in the form of an L1 and L2, it is very much a secondary consideration. Not so long ago, you had chains swapping from being L1 to being L2s that can swing any day. So we don't think like that, right? We're not gonna make choices that don't make sense just to try to pump

David:
[22:43] Up a valuation. Satisfy Twitter, yeah.

Haonan:
[22:46] Yeah. I think then the second observation to make is that most of the stable coins are on Ethereum. This is kind of like maybe an underappreciated fact, right? Somewhere around 60% of all stable coins are on Ethereum.

Haonan:
[22:59] Now, they don't move quite as much on Ethereum, but we can help there.

David:
[23:02] They're like store of value, slow DeFi stable coins, whereas the Tron stable coins are like payment, high churn stable coins.

Haonan:
[23:09] That's right. I think there's also higher value transactions on Ethereum. You know, if you're trying to move $10 million, you probably are using something that's not Tron. I think the observation is that Ethereum is actually the default choice today among stablecoin users, is one way of saying it, right? It's 60% of all the flow, 60% of all the outstanding, rather. And Ethereum, in other words, is the dominant where stablecoins are. And I think our mission at Codex is to make sure that Ethereum continues to be the dominant place for stablecoins. I think what Ethereum needs to do here is it needs to keep up. It needs to run a tight feedback loop between go-to-market and product and research and ruthlessly focused just on stable coins. This is very hard to execute at the L1 level. That's where we come in, right? We wake up every day, we think about stable coins. We go to bed, we're thinking about stable coins. We're in the bathroom, we're thinking about stable coins. We just do this one thing. And so maybe the broader point here is that I think we need to rethink how L1s, how Ethereum L1 works with L2s. And, you know, we have a model in mind that I think is a win-win.

David:
[24:17] Okay, so I'm hearing three different arguments as to like why Ethereum. There's like the technical argument. I don't think you really touched on that, but I'm sure it's there. There's the economic argument, which is all the stable coins are on Ethereum, all the high value transfers are on Ethereum. That's the place to be. And then there's like the ideological argument, which I think you call the emotional argument of just like, yeah,

David:
[24:34] we kind of believe in Ethereum. Maybe we can kind of like actually run through all three of those again, the technical argument. So like technically, why is a layer two construction, what does that benefit you guys as a stable coin focus chain?

Haonan:
[24:45] I think the main trade-off there that people need to make is, are you more interested in stability and speed to market, i.e. B&L2, right? Or are you more interested in incremental, experimental, technical properties at the chain level? I think that is the fundamental trade-off. Yeah. Our view is speed of market matters the most, and our view is stability matters the most. You know, I think when they find bugs in GET nowadays, they are bugs at the go level. That's how robust that code base is. This is one of the most hard... I don't know what that means.

Haonan:
[25:23] It's just like, you know, the... Deep in the basement. The Ethereum, yeah, the Ethereum stack is just so robust, right? Everybody's tried to attack it so many times that when they find bugs nowadays, it's like at the programming language level, right? So it's the bug at a programming language, which not sort of at the sort of theorem software level. Long story short, like what that means is you're working off this incredibly hardened and robust stack. Your other choice, of course, is to not work off that stack and to do something experimental that's probably going to take a long time for R&D and might have all sorts of issues in it. So I think that perhaps is the technical argument there. I think there's also strong sort of guarantees that an Ethereum L2 can offer that no LL1 can offer. For instance, for L2s, users always have the right to exit, back to L1, in a way where if you're stuck on StripeChain or CircleChain, there's nowhere to go.

David:
[26:16] Yeah. I would suppose users, when I talk about the exit hatch in L2s, I'm talking to my listeners, right? So yeah, users, you can use L2 because you have the exit hatch. For you, and I think for broadly, we might also want to talk about countries as well. Because like if we're putting global effects on chain and like russia wants to put their bonds on codex i don't know if you want me to even say these words but like if russia wants to put their bonds on codex be like hey country of russia if we do something bad you guys have an exit hatch the whole entire country has an exit hatch and then and then that's actually part of the consideration there yeah.

Haonan:
[26:57] I i think just just on the sanction stuff like We have a great legal and compliance team here. My mandate to them is that I want no orange jumpsuit risk, please.

David:
[27:07] Yes.

Haonan:
[27:08] Not interested in orange jumpsuit risk.

David:
[27:11] Right.

Haonan:
[27:11] Yeah.

David:
[27:12] Okay. So, okay, with Tempo, Tempo announced itself as a layer one. When they did that, were you like, nice, that's good for me. That's good for Kodak. So what was your reaction to that?

Haonan:
[27:25] I think like the Tempo saga is really interesting because it's just such a small world. You know, my friend Liam is pretty involved there. And one of the reasons I think that Liam is interested in stablecoins in the first place is because of a conversation I had with him three years ago. And so it's really interesting that, you know, that three years later, you know, he's probably convinced Stripe to do this. It's very interesting.

David:
[27:48] What was the nature of that conversation?

Haonan:
[27:50] Well, you know, early on, like stablecoins was quite a contrarian view, right? Because at the time, like, it's payments. Dude, like why are you doing payments? That fucking sucks, right? Like if you think about these chain foundations and I was at one, right? You think about who has the most clout, it's the DeFi guy. That's the bulk of the product market fit. If there's a payments guy around, this is not an interesting guy, right? This is not something. You're talking about like

David:
[28:17] Back in 2021, 2022, back then?

Haonan:
[28:19] That's right. Payments was sort of like the graveyard of, you know, graveyard of crypto companies. But, you know, I think at the time I was just looking at this information from first principles and I'm like, actually, I think stable coins might be the realest product market fit in crypto, right? and Tron in a weird way may have actually some of the realest product market fit in crypto. And this was not a popular view in Ethereum circles at the time. But I think those are some of these early conversations that gave rise to a lot of the folks working on the system today.

David:
[28:52] Yeah, yeah. So like Codex as a name has like arisen alongside of the rise of the mindshare of payment stablecoin chains. But you guys got started forever ago. Talk about, maybe we can just interlude. We're still in the technical argument of like technical argument,

David:
[29:07] economic argument, ideological argument. But let's put a pin in that conversation. When did Codex just come around? When did you guys found it?

Haonan:
[29:14] I started Codex when I left Optimism to start Codex. I just had this observation that it seemed like nobody was really working on the realest part of crypto in a way. And so I went off in search of something that hits three boxes for me. The first box is that it genuinely moves space forward in a way that my grandmother can understand. Second, that it is a massive economic opportunity. And third, that I had some sort of unique edge to work on. And so Codex is what came out of that. At the time when we were working, when we first started thinking about this, this was a deeply contrarian view, right? Yeah.

David:
[29:52] Nobody wanted to work on this.

Haonan:
[29:55] Nobody wanted to work on this. And I think, but if you just stare at the data, It was pretty obvious. And now, you know, sort of that idea has gone from contrarian and sort of maybe fringe to like mainstream, but that's been good for us. I think, you know, we have maybe more time in the idea maze and then others. And I think we can see the map perhaps clearer than others.

David:
[30:16] Interesting. Talk about why it was so contrarian. What about codecs are the contrarian elements? Again, we might have to teleport ourselves back to 2021, 2022, but like, whoa, why the C word?

Haonan:
[30:27] I think the first piece of consensus at the time was that, well, rollups ought to be general. Right? We can go into the history of this more. I think there's all sorts of interesting counterfactual histories that could have happened. But the consensus at the time was, well, roll-ups should be general. And there's no need to focus on any particular use case. And let's run these foundations. I launched Optimism Token, right? So I clapped my hands together one day and $15 billion appeared from thin air. And then the next day we're like, well, how do we give it away? We don't know how to give it away.

Haonan:
[30:58] And so, you know, you spin up all these foundations and all these people are in these foundations and you're trying to figure out, well, how do I, how do I allocate this? And it's very clear, you don't really know how to give it away, right? Because you're spread so thin across all these different use cases. How are you going to be opinionated in smart ways across all these use cases? And so I was there sitting and observing all this and it just seemed to me that really there was room to focus, right? Focus down on one thing. And then the question becomes, which thing? And based off my readings of data, it seemed like stable coins was the way to go.

David:
[31:34] Wasn't that kind of the thesis of the.

Haonan:
[31:36] Super chain though?

David:
[31:36] Or is still the thesis of the super chain, which is like the cluster itself is very generalized, but the Optimism Organization project doesn't really know where to be opinionated. So they kind of leave that up to the market. And what the market is, is all of the individual chains of the super chain. How is what you believed, what you saw different from what that thesis is?

Haonan:
[31:58] I think even for the super chain thesis, I think the super chain thesis sort of maybe revolves, centers around this idea of interop, right? Right. So maybe I'll give my take on this. Carl will probably have a sharper and more correct take than me or Carl and Zhang. But my understanding of this was that, look, you have this problem because you first launched, you first launched, we rollups at a time of experimental technology, right? And you, for a time it was just one or two teams who could do it. And the moment kind of passed where not all the activity migrated to a single well too. And in fact, the tech got more and more commoditized and it actually became productized into RAS, meaning that anybody could go launch a rollup. At which point you have this problem, which is, well, if anybody can launch a rollup, What is, what exactly is the value cruel for a company like Optimism? And so then the thinking is, well, look, let's, let's stitch them together, right? We're going to stitch them together. It's going to have these wonderful properties. It's going to be great. That's kind of maybe, from my view, the kind of heart of the super chain thesis. Like for us today, how much does interop matter? It's not clear to me how much interop really matters to me individually at Codex. And so maybe that's where those theses diverge a little bit.

David:
[33:20] So maybe to articulate why Codex was contrarian at the time that it got started

David:
[33:25] was A, stablecoin payments focused. B, app chain, narrow chain construction rather than a generalized chain, just a narrow focus chain. Do you think that's where Ethereum rollups are just going generally speaking? Because like one of the coolest new like roll-ups that the ethereum community is talking about is lighter and lighter is like just as narrow if not more more narrow than what you are doing which is just like it is a perpdex and it's enshrined in the actual like zk circuits that they've built and it's grown immensely in at least tvl i mean there's incentives there but nonetheless like i don't think anyone's arguing with the product market fit of perpdex these these days so like Is there a broad comment to make about the trajectory of Ethereum Layer 2 designs?

Haonan:
[34:12] Yeah, I think we need to wake up as a community. And I think we need to recognize some uncomfortable truths. The first truth that we need to recognize is that general purpose L2s are parasitic to L1. What I mean by this is if you just spin up an L2 and it does everything, you know, it's DeFi, it's NFTs, it's something, something payments, Is something, something AI? What really is the difference between that L2 and L1? You are most definitely splitting the user base, no? And what is this L2 adding to L1 exactly?

Haonan:
[34:47] Now, pre-EIP 4044, you might say, okay, well, no, it's batch submitting onto L1, it's paying all these wonderful gas fees. Sure, okay, but post-EIP 4044, like, that set of economics is de minimis. So, then where is the value cruel? It just doesn't work, right? The system doesn't work. And so, like, what is the future? The future is you need, roll-ups need to be able to work in a way with L1 where it's a win-win. So what is that win-win? The win-win is the roll-ups should focus on things that L1 cannot do well. What are things that L1 can't do well? Well, it's very, very difficult at the L1 level to run a really tight product research and engineering loo because it's very hard to focus, right? It's very hard to focus on specific problems. And so the role of L2 is to get deep on one thing. And so in our case, we get deep on stable coins. And in this way, we can sustain Ethereum's dominance in stable coins. And in this way, we can also eventually return economics back to Ethereum. That, I think, is the win-win.

David:
[35:54] Is the future. Isn't there kind of just like a migrating Overton window of what

David:
[35:59] parasiticness is as it relates to the capacity of the Ethereum layer one? And so I think now in 2025, people look back on the generalized rollups of 2021, Arbitrum, Optimism, Polygon. And they're like, yeah, those are actually parasitic to Ethereum. But at the time there was no choice. Like Ethereum was doing 15 transactions per second, at least we're doing 45 now. And we just needed something like that to keep people at least one click away from Ethereum while also having like, you know, 100 transactions per second plus. And now we retroactively judge, you know, generalized rollups in 2025 of the rollups of 2021. It's like, oh yeah, those are parasitic to Ethereum. And also Ethereum in the meantime has kind of like increasing capacity. Gas fees have gone down. And now we're saying like, okay, the actual right construction of an Ethereum world is actually an app chain where we can like zoom in, go focus, be focused, do the product development cycle loops that Ethereum will never do. But in 2030, in five years, Ethereum is supposedly going to do a 3x layer one capacity increase over the next three years. We're going to ZK the chain, we're going to get to a million transactions per second, according to Justin Drake. And then maybe in 2025, we look back on the app chains and being like, you know what? Well, Hounen said that the reason.

Haonan:
[37:19] Why he's building

David:
[37:20] A layer two is because of all of the stablecoin TVL on the Ethereum layer one.

Haonan:
[37:23] How is that not parasitic?

David:
[37:25] And actually just in the grand scheme of things, Ethereum is just going to become

David:
[37:27] just as monolithic as all the Solana people said that it was. And we're just going to retroactively judge the app chains of layer twos as being parasitic to Ethereum.

Haonan:
[37:36] Which is doing

David:
[37:36] The same thing that Tempo is doing, which is just beating Ethereum to the punch on scalability and product. What say you?

Haonan:
[37:43] I think that's an interesting way of cutting at it. I think there's sort of a premise embedded in there, which is all that matters is throughput. Do you see how easily people can slip into that? Whereas most of the conversation so far, David, has been kind of me talking about all these other things that matter, right? That nobody's actually working on because they aren't just some throughput number that people can sort of wave around on Twitter. So in a world where Ethereum is doing a gazillion transactions a second, fantastic, that's great. The set of products and the set of features that Codex is building is entirely compatible with that and entirely differentiated from that. If we are unable to do that, we deserve to die. And so when it comes to sort of the Overton window shifting over time, I think that's right, right? I think for a while, I think it was possible that Optimism would actually be the only L2. I think for a moment, it was possible. The token launched, right? Mass evaluation. You know, my own view here would be that we should have done a migration airdrop where you selectively target the top DeFi apps on L1, you move it all over. You know, the pejorative here would be to say vampire attack, right? You're kind of vampire attacking L1. And I was an advocate of this, you know, perhaps like controversial approach. But at the time, the consensus was, look, like execution should happen on L2.

Haonan:
[39:04] So let's do it. Let's have execution happen on L2. Let's move it all over. But of course, you know, first of all, hindsight's 20-20 and it's much harder to be in the decision maker's seat than to be an early employee like myself, kind of, you know, just throwing ideas around.

Haonan:
[39:19] And, you know, that moment passed, right? And so then what emerged is that there were multiple roll-ups with some fragment of the Ethereum ecosystem on it, at which point then we're headed down this RAS path that has happened since. And so, yes, the Overton window has shifted over. I think there are all sorts of these kind of interesting counterfactual histories that are interesting to think through as well.

David:
[39:41] So we've already touched on the economic argument. I think we've covered the technical argument. Let's finish up talking about the ideological argument, which we've kind of like been dancing around. But like, why Ethereum?

David:
[39:52] Ideologically speaking, why a layer two on Ethereum?

Haonan:
[39:56] Yeah, I think I used the word emotional because I think there are ideological reasons as well. Let me start emotional. I think emotionally, like I grew up in the Ethereum community, right? I quit my job. I was very lucky to join Optimism pretty early. I lived in a house of like six, seven other people in Vitalik. You know, I'd come downstairs and Vitalik is on the phone of Elon Musk or something. And one day and the next day, Ashton Kutcher is in the kitchen and, you know, all sorts of crazy shit is happening.

David:
[40:22] Oh, I remember that house.

Haonan:
[40:23] That's right. That's right. And so like, that's how I grew up in this ecosystem. system. And so I think if I'm intellectually honest about it, I'm probably a bit biased that way. I think, you know, ideologically, I think first of all, you don't have to reach for ideology quite yet. I think the empirics back it up, right? Like most of the stable coins are on Ethereum. We just have to sustain that dominance into the future. And if you think about which ecosystems are truly kind of true to their values or neutral that are likely to really be a settlement layer globally for the world. I think Ethereum is obviously I think anybody who's arguing otherwise is, you know, maybe raising money for no one.

David:
[41:08] There was a while I was hosting an episode, like a roundup episode with like a handful of different people on it. And one of them was Bread from MegaEth. And this one very strong Ethereum community discord I was in was giving me shit for hosting Bread from MegaEth. And I'm like, what the hell do you guys want? Like, MegaEth is an Ethereum layer too. They're doing the logical conclusion of like the Ethereum rollup design. And all of a sudden, like hosting bread from mega ETH is antagonistic to Ethereum. So some parts of the Ethereum community have their antibodies, very sensitive, very turned up. And so what would you say, how would you address those people who are just like very tuned to be like, no, you're actually extractive to Ethereum? What is Codex? What is Codex giving to Ethereum? What values is Codex giving to the Ethereum project?

Haonan:
[41:59] Yeah, I think you see this happen quite a bit with communities that become sort of more and more mature, right? I think you see some, these are some times, in 21 or 22, it would be like, these are kind of Bitcoin-y traits, right?

David:
[42:12] Yeah, yeah.

Haonan:
[42:13] Or like any attempt at changing something about the Bitcoin ecosystem is an attack on Bitcoin and let's get the pitchforks out and let's string them up, let's hang the guy, you know, which, which, which. So I think it's a natural human tendency. I understand it. I think what we need to do as a community is discern between true threats and progress. So what are true threats? True threats are Alt-L1s attempting to supplant Ethereum. Tempo is a real threat. And we should not engage in motivated reasoning about it. So one version I've heard about this is, well, it uses the EVM, so therefore it is Ethereum. I mean, come on, right?

Haonan:
[42:58] That's, that's BS. Like sit down and really think about that for a second, right? That's, that's BS. So we need to be able to discern between like true threats and progress.

Haonan:
[43:10] What does progress look like? Progress looks like, well, what is the future here, right? We have this problem where, you know, perhaps general purpose L2s are not all that additive to the Ethereum ecosystem. Well, then how do we proceed? Because we have, we have RAS, right? You can't spin-up row-ups. We have a thriving Ethereum community. And there are all these threats on the horizon. So what should we do? Let's sit down and think about what we should do. Well, we need to move quickly. We need to run a tight product iteration loop. And we need to preserve Ethereum's dominance. And probably the best form of doing that is in a stablecoin-focused L2. And so, I mean, obviously I'm talking my book here, but we are very much not a threat to Ethereum or very much aligned with Ethereum. And, you know, we hope to preserve the future of stable coins on Ethereum.

David:
[43:59] What about the whole like topic of conversation of like beta-based and native roll-ups? What do you think about those in terms of like technical, even stronger technical alignments to the Ethereum layer one? What do you think about that conversation?

Haonan:
[44:10] Yeah, I think those are interesting. I think we're happy to explore those things. I think, again, this falls in a bucket of like interesting technical conversations that customers do not care about.

David:
[44:18] Sure. Yeah.

Haonan:
[44:18] So, like, if you go on one of these chain websites and all they talk about is stuff that sounds good to, like, R&D team, I would look out, right? Yeah.

David:
[44:30] One theory I have, one thesis I have about tempo. You know, tempo is like, you know, what is good for tempo is good for Ethereum. We're EVM. We're going to, our software contributions will flow back to Ethereum. We're going to increase the pie for everyone. And also, we're just, like, stable coins, you know? So, like, don't mind. And so we're not coming for DeFi, we're just stablecoins.

David:
[44:50] And I think anyone who says that on the Tempo side of the aisle is like probably being true and authentic. Like, yeah, they're just focused on stable coins and payments. And I don't think anyone in the Tempo complex is being dishonest if they just say, yeah, we're just thinking about stable coins. I think as Tempo grows in the future where Tempo does get adopted as a payment chain, as a stable coin chain, and like the weight of Stripe actually gets expressed in the dominance of Tempo. Tempo evolves from just a payments chain just because of the fact of that's how layer ones work. It's like, why would you not put Aave on Tempo? There's going to be an Aave on Tempo. Why would you not put any sort of DeFi app on Tempo? And why would, there's no difference between a stablecoin payments focus chain and just like a very high capacity, high capability, generalized EVM chain. And so, yes, everyone in the Tempo orbit is like, yeah, we're focused on stable coins. But Tempo as a construct, Tempo the actual, if you give agency to Tempo the blockchain, Tempo's like, yeah, no, I want to become the dominant layer one,

David:
[46:02] the generalized layer one. What do you think about that thesis?

Haonan:
[46:04] That rings true to me. I think the play here, it's a very smart team over there, right? There's a lot of great people. I think if I were them, what I would be thinking is less salami to slice salami slice the shit out of this. What I mean by this is you just progressively and slowly inch through everything. So you make no moves immediately that are threatening to others. You just slowly and progressively get there. Under cover of maybe some smart rhetoric.

David:
[46:35] My general attitude towards almost everyone in this space is like everyone in crypto is a one of two camps. You are either unbundling, and this is the perspective from Ethereum. So again, Ethereum is the agent here. Ethereum is the organism. Anyone in crypto is either unbundling Ethereum. And this is like the archetype of multi-coin capital. What does multi-coin capital want to do? They just want to unbundle Ethereum. Like one by one, take all of the different values of Ethereum, different value that Ethereum produces and like make a faster alternative version of it somewhere else that they invested in. It's like the archetype of Ethereum unbundler. So Tempo, big Ethereum unbundler. And then there's the Ethereum bundlers, which is like Ethereum is the chosen open source ecosystem. It is this gravity center of open source contributions. And that is like naturally accruing of mindshare. It's credibly neutral. Things just gravitate towards Ethereum.

David:
[47:32] Even Tempo, even though it's like an alternative layer one with no material technical alignment with Ethereum, it's still the EVM. And that EVM grows in the gravitational center. And actually, it's Ethereum at the center of the orbit. And so I'm interested in growing the Ethereum project. So I'm an Ethereum bundler. I will say that you are an Ethereum bundler because you're building an Ethereum layer too that's focused on bundling up payments and stable coins into Ethereum. And so I actually don't really have a question here, but like, that's how I view the world. It's like, you're either, you're either an unbundler of Ethereum or you're a bundler of Ethereum from the perspective of Ethereum. What do you think about that?

Haonan:
[48:10] Yeah. I think there's like, you're like, Ethereum is kind of like Craigslist, right? Is kind of what you're, talking about here. Yeah. I think that's interesting. Yeah. I think the thing is that the Ethereum bundle, like What is the reason why you might unbundle Ethereum? You might unbundle Ethereum to focus, right? So somebody might focus, you know, Craigslist has rentals and you take out rentals and you just do that and it's Airbnb, right? I think in the case of Ethereum, we have this wonderful thing called L2s that allow you to do that. You can go focus on one thing and yet the whole thing still coheres. So it would be like if Craigslist had an incentive for people to spin off little sections of it while still being economically aligned with Craigslist. And that's what's powerful here, right? This is the model we have to get to. And I know there's all sorts of factions inside the Ethereum community that have different theses on it. I've heard one version of this, which is, look, like, sorry, F the L2s, right? F the L2s, and let's just focus on L1, and let's do that. There's obviously another camp, which is like, don't do that. Like, focus on the L2s. And I think the truth here is somewhere in the middle. Like, we need to be pragmatic about this, right? We got to make sure that it makes sense. Like, L2s need to interact with L1 in a way that's additive.

Haonan:
[49:32] And obviously, L1 needs to, like, there needs to be a reason why you want to be in Ethereum L2, right? So I think, again, I'm biased, but I think the model that we are pursuing is the future.

David:
[49:43] Yeah. Yeah, I think I agree with you there. It's just like, let's split the difference between F the L2s and focus on the L2s. But I'll give a slight bias, so like tie goes to the Ethereum layer 1, so I'm not perfectly splitting it down the middle. I think like in 2021 the ethereum community over index on layer twos and they were like all layer twos are valid by default 10 000 generalized roll-ups they're all they're all valid they're all they're all great that's right and and i want to actually my perspective is like actually by default a layer two is not valid it's not it's not legitimate you can't just you can't just spin up a layer two you have to earn your keep you have to do something net additive and being a generalized rollup is not net additive to the Ethereum space. You have to like earn your place as a valid contribution to like the Ethereum landscape. And if you're not valid, then like you're actually parasitic, right? And so like the way that you are valid is you have tight product development research loops. You do grow the pie and you grow the pie in a way that is additive to the Ethereum side, not in the tempo way where it's like, yeah, we're growing the pie, but we're also eating it too. So that's, that's, that's my kind of fault. It's like, don't, don't, discard the layer twos, but also they're not valid by default. They have to, they have to earn their, their Ethereum logo branding, not, not get it by default.

Haonan:
[51:05] I think that's right. I think that's right. I think another way of thinking about it is if a rollup is adding nothing other than existing, it is fragmenting the user base. That's the truth of it. Right. So it better either address some use case better than L1. It better be attracting new users, better be attracting new flows, creating new functionality, or it really ought not exist.

David:
[51:30] Yeah. There was a previous year, maybe a couple of years ago, where that was like blasphemy. But I hope that this becomes like the status quo, like thought in the Ethereum community. When tempo announced itself and they're the rambling from the ethereum community got really really loud and matt hong had to like answer to the fact as to why it was an ethereum why it was a layer one rather than ethereum layer two he said something to the effect of like not wanting to be, tied to ethereum's development roadmap and like a bunch of other things like the the general answer was like if we were at ethereum layer two we would be constrained by ethereum by like ethereum throughput, the 4844, like blob space, growth targets. When you read that analysis from Matt Huang of like, we wanted to be just technically independent from Ethereum, what was your reaction to that?

Haonan:
[52:19] I think Matt's in a tough spot, right? I think Matt has to balance a ton of different things. I think, you know, while there's Paradigm LPs to balance, right? Why is the GP of this fund starting a port co? And how is his time split between Paradigm and this new port co? He's invested in much of the Ethereum community as well, right? So is he now adversarial towards those investments? I think he's got a tough balancing act and, you know, obviously a very smart guy. So I think he's doing a good job of it. I think, again, it has this R&D fetish, that we were talking about, right? I think it's very easy to slip into like, oh, well, we just, what we really need is just more throughput. That's what we really need, right? We need a gazillion transactions a second. And often that's really not what is actually the bottleneck. And so, again, I'm biased, but I think if you analyze the situation correctly and you spot the true bottlenecks, you would conclude that an Ethereum L2 is the faster way to win.

David:
[53:19] How on end, I'm into it. I hope the best for you guys from my Ethereum side.

David:
[53:25] Are there any other parts of this conversation that I haven't asked about? Any stone I've left unturned?

Haonan:
[53:29] I think it's just a really exciting time for the space. I think we're seeing all these sort of geopolitical forces play out on chain as well. And I think all this will get decided in the course of the next half decade. So I was excited to be on the frontier.

David:
[53:44] Cool. Well, I hope the best for you guys. Rooting for you. Not an investor. Spankless, nice. And no, I just like the Ethereum. Roadmap. I like the Ethereum project. To what degree Codex has a firm foot planted inside of the Ethereum ecosystem. I'm a supporter. So, Halonin, thanks for coming on the show and just educating me about this space.

Haonan:
[54:02] Thanks, David. Great to chat. Talk to you soon.

David:
[54:05] Banklessation. You guys know the deal. Crypto is risky. You can lose what you put in, but nonetheless, we are headed west. This is the frontier. It's not for everyone, but we are glad you're with.