Synthetix: A New Hope with Kain Warwick

Welcome to Bankless, where we explore the frontier of internet money and internet finance. And today on Bankless, we are exploring the frontier of the multi-chain Ethereum roadmap. The roll-up centric roadmap has always been about producing many, many new layer twos, each of their own flavor, construction, and type, in pursuit of allowing diversity of chains to all exist under the same ecosystem that we call Ethereum. Synthetics, an application, has been leading the charge into exploring what this means for the application layer of Ethereum. And today on the show, we're bringing in Kane Warwick, actually inside of my apartment. This is an in-person conversation. So you'll be able to pick up on the uh speed and lack of latency there is between me and Kane, I would say. He was in New York for Mainnet. And so we wanted to do a show about his recent blog post that he wrote called Synthetics a New Hope, about the next stage for synthetics in this multi-layer two ecosystem and how he is thinking about some of the complexities and challenges that the Ethereum Rollup centric roadmap presents for the application. Like I said, since he was in town for Mainnet, I was like, Kane, just come on over and we'll do the conversation in person. So that was a pr that was pretty fun. It's always nice to have in-person conversations. I don't do them as much, even though I have the setup ready to go. Anytime there is a conference in New York, however, I will try and get one done. So in this episode, you're going to watch us navigate the complexities, the challenges of Ethereum's roll-up centric roadmap as it relates to applications that desire to exist across multiple layer twos, across multiple chains. Depending on the layer two application, the application construction itself, different challenges can arise in various levels of complexity. Kane and I walk through the three different levels of cross-chain complexity that an application can have that is determined by what that application is trying to do. For example, just mere independent instances like Uniswap deployments across chains is the easiest way to navigate the multi-chain world. Doesn't really matter if different Uniswap instances are deployed across chain. But as soon as there's unified governance, all of a sudden there is complexity because we need to start to weave these chains together. And then there's the next phase, the synthetics phase, which is not just unified governance, but unified liquidity and unified state.

Synthetics has always been at the frontier of navigating the hard problems in Ethereum.

After Unipig, the Uniswap demo on Optimism's Optimistic Rollup, the first optimistic rollup that had a live in-production demo of an application, we called it Unipig back then, uh Synthetics was the number two.

And so they have always been pushing the frontier of what is possible in the world of crypto. And now Kane is leading the charge into doing the same thing once again with synthetics and the multi-chain ecosystem. So if you are curious about the super chain idea for optimism or just simply the shared state of layer twos, they re-recomposing all the fractured composability of layer twos, this conversation is for you. And it's also just a fun conversation. Kane is a great conversationalist, and he's, of course, wicked smart. So let's go ahead and get right into that conversation with Kane Warwick of Synthetics. But first, a moment to talk about some of these fantastic sponsors that make this show possible, especially Bigeless Nation. We are here in person in my apartment here in Brooklyn with Kane Warwick, the co founder of Infinex, semi benevolent dictator of synthetics, house collector, and possibly, possibly the best hair in Ethereum. Kane, welcome to my apartment.

Yeah, thanks for having me. Um I was hanging out downstairs for a little while. I couldn't uh figure out what channel to get you on. So

Yeah, th just like w the problem that we're going to discuss in this episode, how do you reach me? Telegram, Discord, email, Twitter, text.

we need some kind of aggregator.

Yeah, we need one single shared way of getting in contact with each other. You're in town in New York here for uh Mainnet. How's that been going?

Yeah, it's been good. Yeah. Um so uh mainnet and um ETH New York slash pragma, I guess, uh, you know, the Ethereum, um, hackathon slash conference. Um so yeah, it's been it's been pretty good. Vibes are pretty good.

Lovely, lovely. Uh and y do you go in and out of Australia or do you do a circuit before you go back home?

Um I this time I did because uh I went from Togan 2049, landed uh in Sydney and then 24 hours later hopped back on a plane to come over here. Um I had to pick up my kids on the way through, so uh couldn't go directly from Singapore to New York, unfortunately.

Oh man, the life of uh juggling, uh being a dad and going to conferences and also being an app founder, I cannot imagine.

Yeah, it's it's pretty hectic. Um but yeah, my prints are my kids are pretty good travelers now, so

Okay, so as I alluded to, there is a problem arising in the world of Ethereum. The roll-up centric roadmap, as great as it is for producing natural emergent solutions across the different modules that make up a chain, it also fragments composability. And this is like kind of how I alluded to with like how does how do you reach me with all the different apps that we use? The illusion here is that there's so many different networks. How many networks currently is uh synthetics deployed on?

really two. Mainnet and Optimism are the other two.

Okay, why not more?

Uh good question. I think, you know, we obviously were the first uh um project to deploy to Optimism. Um we worked really closely with the team. Um, you know, I did try to warn people about this and say, like, hey guys, why don't we coordinate around Optimism rather than having you know 100 roll-ups, but that ship has sailed now, I think. So um, you know, uh we are in an environment where we're seeing, you know, even L1s convert into L2s, which I think is great, right? I think people um, you know, who previously thought it was a good idea to launch an L1, realizing that, you know, being uh sort of uh connected to the Ethereum ecosystem is a more viable pathway. That's amazing. Um so you know, in the end of the day, I think we'll kind of

Accept that that's just the direction that things are going. But for us, the challenge has been cross-chain communication. So with synthetics, because we're a liquidity protocol, we need to be able to ensure that on whichever network a user is interacting, that information about those interactions are passing through to the other networks. And that just creates cross chain communication demands that are much higher than like bridging a token, for example, where it's sort of a one shot thing and then the tokens on the other network and you can kind of let it do what it needs to do over there.

I would imagine with the demands that Synthetics has, uh the growing number of chains produces probably like an exponential growing amount of complexity for managing those chains.

Yeah, if you do it the right way.

Yeah.

Right. So, you know, there's uh like unfortunately, we're not Uniswap, right? Where Uniswap can just deploy their immutable code onto a new network. Um, there isn't even the complexity of like moving fees around, right? Fees sit on that network, they go to LPs. Um, you know, you flick a fee switch on, and all of a sudden the complexity goes up significantly. Right. Because now it's like, okay, well, where do the fees go?

Right.

You need token holders on that network, you need token holders on mainnet, you need token holders anywhere. Do we burn the fees? Do we transfer them? Do we convert them? What about if it's on Avalanche? How do you deal with that? You know, so as soon as you start to have any contemplation of like cross chain communication, um, the complexity grows significantly. So if you can just deploy your code and let it sit there and interact only on that one network, it's much easier. Unfortunately, for synthetics, it's just not possible.

Right. And you you so what you're starting to allude to is that there are different types of apps out there that have different cross-chain needs.

Yes.

Maybe there's a few categories of this in your blog post, which is going to be the basis of this conversation. You kind of start to progress through the different kinds of apps that have different levels of demands for complexity when it comes to cross-chain. And I think synthetics would be at like the highest order of a cross-chain, a many multi-layer two world is the most complex for synthetics than it would be for like Uniswap. Uniswap is deployed on like 13 different chains, and that's just a property of the nature of what Uniswap is. Like you alluded to, there's this one thing called a fee switch that if we turn that on, all of a sudden Uniswap goes from like, well, it's pretty simple to deploy Uniswap across chain to like, okay, now we are elevating in complexity. Could you kind of walk us through the phases of complexity that an app, the properties of an app, might be? Like what properties are simple to deploy across any chain, like the roll-up centric roadmap of Ethereum, 10,000 rollups. That doesn't matter for some apps. What kind of categories of apps are those? And then what things make things more complex?

Yeah. So I think on the spectrum, you know, the simplest thing is a Uniswap style deployment. It's immutable code. So you don't need to govern it. You don't need to upgrade it. You just deploy the code and it sits there. There's no need for fees or anything like that to go outside of the protocol. So there's sort of endogenous to the protocol. LPs put liquidity in, then LPs earn the fees. That's it. And so each instance is really distinct. As soon as you start adding things like governance, right? If there's any upgradability, then it's a question of where does the governance come from? Is it governance just on that network or is it governance that is anywhere? Because most of the time the governance comes from token holders and token holders are spread across all these different roll ups.

Right.

So it's like, well, how do we coordinate the governance? You know, do we want someone who is holding uh Ave on

Mainnet to be governing the deployment on Arbitrum?

Right.

Does that even make sense, right? Like, are they the same kind of stakeholder, right? Or is it only the Aave token holders who are kind of staked in the mod in the security module or whatever on Arbitrum should govern the Arbitrum deployment?

Right.

And then there's another, you know, area of complexity, which is fee sharing. So, okay, if there's fees that are being earned on one network, who do they go to? Do they go to all token holders? Do they go to um, you know, uh just token holders on that network? How do we distribute them? Are they paid out as a dividend? Do we do buybacks and burns, etc.? Um, and you know, there's solutions to all of these things, but

they all add complexity. Um, and you know, like the governance side of things, I think, adds the most complexity, right? Um, then there's some other stuff like bridging. Do you want all your tokens on all of these different networks to be fungible?

So, you know, I gave uh an example a long time ago, um, you know, even before this was a real thing of maker, right? So if you have DAI on one network that is mainly backed by USDC, let's say, right? Um, and then you have a different instance of maker that's deployed on, let's say, Arbitrum, right? Um, and maybe the majority of the uh the issuance of DAI is backed by ARB, the ARB token. Now that's crazy, and that's not that's not the case, right? But like in a future state, maybe that might be the case, right? And so you look at those two things and you say, okay, you know, we've got one network where it's all backed by USDC, another network where it's backed by like treasuries, and then a third network about where it's backed by, you know, an illiquid shitcoin, right? Um, you know, let's use SNX, right, as the example, rather than ARB, right? I don't want to trigger um, you know, ARB maxis, right? Um, so you know, we've got a synthetics app chain where you've got die that's being backed by SNX, right? Um and you know, it's illiquid or or whatever. Um

Is that die that's being issued against those, you know, collateralized positions fungible?

Mm-hmm.

Do you want it to be fungible? You know, does it even make sense? Is it, you know, on some fundamental level, is it fungible? Right. And if it

is, how do you then enforce that fungibility? Right. So you need bridges, you need cross chain communication, you need to know if the dye is moved onto a different network and it needs to be liquidated. How does that liquidation happen? And so all of that stuff just adds complexity.

Is this both, uh if I'm understanding this correctly, is this both like a kind of a more higher level philosophical question of is is the die fungible from like the social layer? Do we perceive these things to be equivalent? And then the other problem is like, okay, maybe even if we uh do get to that point where yes, everyone agrees that these things, the risks of one implementation of die on one chain, we do agree that that should be equivalent with the risks of a different die implementation of a different chain. If we can cross that barrier, there's still the technical barrier of getting it to be fungible at the technical level, like the contract address and equivalency. So like there's two problems here.

I think there's kind of three. There is the social problem, which is, you know, if I'm using DAI, what is the perception of a user of DAI? Do they expect fungibility? Like what is my expectation as a user of DAI? Then there's like the financial layer of, okay, if my expectation is they're fungible, are they really financially fungible assets? Right. Like, you know, if one's backed by nonsense and the other one's backed by ETH,

is that even the same? Right. And how do we reconcile that? Is there a way that we can set up the financial incentives such that, you know, one has a really low LTV and one has a high LTV, and the liquidation means that actually they're both as safe. They've just got different parameterization. And then assuming we get to that point, then the next thing is like, how do we even do that?

Right.

Really, just to drive this one point home, the way the way that you describe this in your article is like Alice has trust preferences that uh maybe Alice is very conservative. And so she only wants to use uh, you know, a very low LTV inside of synthetics on mainnet uh with very good strong collateral because these are her trust preferences. Maybe a more adventurous Bob is out on some OP stack fork that just got deployed. And if there's a synthetics instance on that network that for some reason allows for, you know, shitcoin collateral to back SUSD, that's great if these instances are firewalled, kind of like how Uniswap deployments would be. Then like Alice's preferences don't get commingled with Bob's preferences. You know, conservative Alice is safe from DGen Bob. Um, and so long as there's actually like separation of instances. What is why why I think uh the articulation of synthetics is such an ambitious project about shared composability is that the unified version of synthetics would mean that Alice's preferences and Bob's preferences are actually commingled and those conflict. And so that's where we get into like, okay, so is S U S D on you know OP stack shit shit OP stack fork with shit OP with shit uh collateral, sorry for the language. Uh just like, you know, the all the risk, all the risk on one.

Maximize risk and then maximize conservatism. Uh do do it how do we even square these things?

Yeah. And is there, you know, this is the financial layer, right? Like, is there a way that you can parameterze this to allow both of these people to interact with one another and feel comfortable? Right. And or if not, is there a way for you to sort of quarterize the risk and segment and silo the risk so that you know the user gets fungibility, but on the LP side, the LPs are protected from one another, right? And you know, you don't have this like cascading failure or, you know, um, or whatever. Um, uh, and so you know, these are hard problems, right? But then it begs the question of even if you can get to that point, the technical implementation, there is no way to implement this. And this is where synthetics unfortunately has a bad habit of being like on the bleeding edge. Like the vision is a very bad habit. Yeah, very bad habit. Like, you know, it's it's like this kind of uh very aspirational vision of like, what if we could do this? You know, what if this is the future state? And the reality is that the tech is just not there yet, right? Um, maybe this is solved with like, you know, at some point, uh like an opcode, you know, like level of solution, right? Where like we actually realize that cross-chain communication is necessary and you know, we have this reliance on like Ethereum as a you know a state layer, you know, where we're you know storing state there, and somehow we have a way of communicating that back out or something. You know, maybe there's like a more long term solution. Um, in the interim, we've got you know a bunch of solutions which are fairly novel, um, and none of them.

are really yet in a position to be able to actually do this. And, you know, like anything, we do it and then we go, Oh wow, this doesn't scale. Right. Right? Like or here's a solution that like handles the problem today, right? Ethereum mainnet, you know, back in twenty eighteen. Totally fine. Enough bandwidth, right? Like just barely enough bandwidth, right? Back then

Days of one gray gas.

Exactly. And then like 20 minutes later, it's March 2020. Right. And you're like, oh wow, this does not work at all. Yeah. Like it just is not possible for this to function. Right. And so even if we have cross chain communication and we've got these kind of rudimentary solutions or whatever, um, do they scale to something, you know, that that is uh, you know, across a hundred chains? Like at the moment, no. Can they get there? It's a technical problem to solve.

Right. And just to again really just hammer on the definition of the problem statement, um, the many many app ecosystem, if you have uniswap without governance, just raw instances of Uniswap deployed everywhere, that's like problem level one, not too not too much of a problem. Problem two is when, okay, we have central Uniswap governance and decentral Uniswap deployments. How does that co how do we cohere? How does this one app cohere into this one unified system when we're across, when Uniswap itself is across many different states? That's like um, you know, not a crazy problem, but still a hard problem. It's still like some uh thing, some things to suss out. And then the final problem is like what synthetics wants to do, which is shared liquidity, shared liquidity across states. And liquidity is finite. Uh, if you have many chains, we're starting to fracture liquidity. And so the golden solution here is that when you add liquidity to one network, one implementation, you are also increasing the liquidity on a different network. And that's like the hardest level.

Go through, yeah.

Right.

It's the hardest level problem that we've thought of to date. And this is, you know, like it would not surprise me at all if synthetics or other protocols eventually come up with even harder problems that are, you know, harder problems that have to do with cross chain communication. There's some, you know, demand or some functionality that requires something. But at the moment, that's kind of the you know, the most challenging thing is like to treat these distinct networks as one network requires a level of cross chain communication that we just can't quite support yet.

Right. Yeah. And a and of course just to throw a bone to the Solana camp out there, this is why they are like layer one, one chain maximalists.

Yes.

Like they they get to be the composable chain because they only want to have one monolithic chain. Uh so they're listening to this and be like, Yeah, that's that's why Solana

Yeah, don't like don't fragment your liquidity. Right. Of course. And you know, there's something to be said for that, right? Um, but you know, the this definitely a counter argument of like, okay, fine, but at some point, you know, the L1 monolithic thing breaks, right? And then

they just kicked the can down the