**A** (0:00):
This is epicenter episode 400 with guest Sunny Agrawal. Welcome to Epicenter, the podcast where we interview crypto founders, builders and thought leaders. I'm Friedrice Ans, and I'm here with Zubin Kutica. Today we're speaking with Sunny Agrabal, who all of you know and love as one of our Epicenter co hosts. But today we are talking to him about his project Osmosis, which is an AMM thing on Cosmos. We'll get into the details in a bit, but before that, let me tell you about our sponsor this week. Solana is the next generation blockchain. With lightning fast blocks and fees less than a cent per transaction. Scalability is perhaps the single biggest challenge preventing crypto from becoming the backbone of the world's financial system. Go to solana.com epicenter to learn more. Cool, Sunny, it's super funny to actually have you on the other side of this for once.
**B** (1:12):
It's weird. Feels really weird.
**A** (1:15):
Osmosis is a pretty recent project, right? So it's not been around. It's not been around for a long time. So tell us about how Osmosis came into being and about the team and what it does in a nutshell.
**B** (1:33):
Yeah, sure. So it's about the project launched about a month ago, but maybe worth talking a little bit about the history of the project. I think. I was on this, my second time being a guest on Epicenter. I was on as well maybe a year and a half ago, me and Ethan Buchman, we talked about Cosmos and I think it was like shortly after the launch of the Cosmos Hub. And yeah, so basically last year I left the core Tendermint team and I sort of went, but, you know, still wanted to keep working on the Cosmos project. And so I went and, you know, I had a validator that I had been running for a while called Sika, and we started sort of running our validators and we were just experimenting with all sorts of different things. We realized, you know, running Validators is not the most fun thing in the world. I mean, it's okay, but, you know, we want to build stuff and we're trying to figure. And you know, this is right around when like, Defi Summer was happening on Ethereum. And we're like, well, all this Defi stuff is going super cool. Let's like, how do we bring it to Cosmos? And so we started experimenting with like, different kinds of things that we want to build. We got, you know, so me and my co founder, Dave Oja, we were like, you know, experimenting with a lot of like privacy stuff. We got really interested in front running resistance and so we started building that out, spent most of fall building out a front running prevention scheme and realized it's a cool feature without a product. And that's why we're like, all right, let's actually launch a DEX called Osmosis. So it's the first sort of DEX on the Cosmos ecosystem that's like IDC enabled and that's sort of like, you know, in parallel while we were working on this Osmosis thing, we were also contributing back to the Cosmos core development and helping with a lot of the work that was going on there. And Osmosis sort of got a chance to be like the first chance that IDC inter blockchain communication protocol was used at scale. And so this is sort of the first real time that like the Cosmos vision has sort of come alive. Oh, and then one thing I also to mention about the team is we kind of actually merged with another team called Chain Appsis. So they built their the Kepler Wallet, which is sort of the metamask of Cosmos. It's the most popular wallet on Cosmos. And so our Sika Validator company kind of merged with the Wallet company and we started building this Osmosis project together. And so we've been doing that for about six months and until we launched last month.
**A** (4:08):
Super interesting. I'm fascinated to get into the weeds here because AMMS and MEV resistance seem to not go together very well. But we'll get to that. So let's talk about Cosmos first. Right, so basically Osmosis runs on Cosmos. Is it its own zone or does it run on Cosmos Hub? And maybe for the listeners who are probably mostly familiar with Ethereum, can you kind of put these into a larger context?
**B** (4:39):
So Osmosis is a chain in Cosmos. So yes, it is its own independent chain. It has its own validators, it has its own staking token as its own security model. But what it means that it's part of Cosmos is that it enables this IBC protocol and this IBC protocol is what allows it to talk to the other chains in the Cosmos ecosystem. So whether that means the Cosmos Hub, you have the Akash network chain, you have region chain. Currently we have about eight chains connected right now, although I think another one just launched yesterday. So it'll be nine, hopefully by end of day today. So yeah, we have, we have, we have more and more chains that are connected. But yes, it's built using the Cosmos SDK and like the Tendermint software stack which is Sort of lot of the stuff that I was contributing to back when I worked at the Tendermint team. But yeah, it's its own chain now built using this stack.
**A** (5:36):
So the IBC just. So the IBC is the Internet of blockchains. What exactly does it mean? Does it mean that the standalone chains kind of trust each other or that there's shared security layer or what unites them?
**B** (5:53):
Yeah, so IBC stands for Inter Blockchain Communication. What it is is it is a generalized bridge protocol. So you know, at this point I think everyone's familiar with bridges and like, you know, you have a way of moving tokens between two different chains. But all of these bridges right now are very bespoke and for every like chain wise pair you usually have to have some trusted set that maintains the bridge and all this kind of stuff. With ibc, the idea is, let's build this native generalized bridge protocol where you have two chains. Let's say you have chain A and chain B. Chain A will have a light client for the consensus of chain B on, on chain A itself. And so it's part of the state machine itself and it's doing light client verification. And so this way your security of your bridge is equivalent to the security of the counterparty chain. And so this allows people to sort of, you know, basically send tokens between these chains with no additional trust assumptions. They have to have trust in the two chains because, you know, if, you know, if you're holding tokens on chain B and you want to move them to chain A, you obviously have to have some trust in both those chains, but you don't have any additional trust assumptions on top of that. And what's nice is because we use the Tenderman consensus protocol, which is this fast finality chain where it comes to consensus on every block and once a block is committed, it will not be reverted. This allows this cross chain communication to happen very quickly. You don't have to wait for minutes or hours or days for some sort of finality guarantees. You can say, okay, I want to send a token from chain A. You know, one block happens and then chain B receives it and, and you, you get it. And also from like the developer standpoint to create these bridges, it, it's very simple because when you want to open a connection with two IDC enabled chains, it literally just requires making two transactions on both chains. So you know, you don't have to get validators to run new software. You don't have to run, you know, all this new stuff. It's Just like transaction, transaction, you have a connection and now these things can start talking to each other.
**A** (8:14):
That sounds like it's potentially dangerous. We're going off on a tangent here. But what prevents you from kind of scamming this mechanism?
**B** (8:22):
Yeah, so what happened? So if you actually go right now, there's a site called Map of Zones and what it does is it has a overview of the Cosmos network right now with the list of all the IBC enabled chains. And, and you'll actually notice a lot of them are, there's like on the thing there's probably hundreds of chains, but most of them are actually just test nets like on, you know, weird chains that people just spun up on their computer for an hour to decide to make an IBC connection. And to ibc it doesn't know the difference. Right? Like it doesn't care. It doesn't know that something's a main net with like economic value or if it's a test net that's running on my computer. So basically the idea of the IBC Cosmos vision is a lot of this is pushed onto the users where you design protocols such that the, if a user trusts chain A and they trust chain B, they can like move their A tokens onto chain B and do whatever they want. But any other user that also has a tokens on chain A, if they, as long as they don't move the tokens onto chain B, they're not going to be affected because obviously, you know, it keeps track of how many tokens were ever transferred over a connection and it will never let more token more than that number of tokens come back. So it's sort of up to users to decide which chains do they trust and want to interact with.
**A** (9:50):
So it's kind of a web of trust architecture, is that correct? Where basically the nodes across which you can actually send tokens correspond to how trusted the connection is by connection is by users.
**B** (10:06):
Yeah, it's up to individual users. Right. Like I, I think that the region network chain is trustworthy and you know, I want to interact with it. The question is, would you be willing to hold tokens natively on their chain in the first place? If so, then you should also be okay with holding IBC tokens transferred to that chain as well.
**A** (10:26):
Okay, I see.
**B** (10:27):
But you might not be okay with that. You might not trust that chain and so you're not going to. And you know this happens already today. Right? Like, all right, like let's say, let's say a chain that's like somewhere it's A little bit sketchy, but it is legit. Like, you know, something like Tron. Right. It's like, all right, maybe. I think it's, I think it's okay to hold tokens on Tron because I think it's secure enough. But there's obviously reasonable argument for why many people be like, no, I'm never going to hold tokens on Tron because it's a centralized system or whatever it is. And today we already do that. Right. Like, I can hold Tron tokens and you don't have to. It's the same model for IBC as well. It's up to users to decide what they trust or don't.
**A** (11:05):
Okay, and did you, did you just say that there's currently eight IBC enabled chains in the Cosmos Network?
**B** (11:13):
There is eight when Cosmos, when Osmosis launched. I believe at the moment there's 10. So in the last month, two more.
**A** (11:24):
Okay, and can you give us an idea of what kind of chains they are, what they do, what the ecosystem looks like?
**B** (11:32):
Yeah, so, I mean, there's a wide variety of different chains right now. So there's a project called Akash Network where they're trying to build like a sort of a decentralized cloud similar to like, you know, build a decentralized version of like aws, essentially a compute market of sorts. And so, you know, they're, they're one of the big ones. You have chains like the crypto.com chain which is like this like large exchange. Their chain is actually built on the Cosmos SDK and has IBC enabled. So it's interesting that a number of different big exchanges actually use the Causes SDK for their chains. But crypto.com was the first one that actually enabled IBC and is using it, you know, actively. You have things like Region Network which are this a Carbon Credits marketplace. There's, there's a Sentinel, which is this like decentralized VPN service. So you have, you have, you have all these, a couple of different chains. And you know, it's interesting that in Cosmos right now it's sort of like a very, it feels like Ethereum. If you go back in 2017, there was like, you know, I feel like most of the things that are interesting things that are happening on Ethereum right now are like mostly Defi stuff. But I think like, but back then there used to be like a cool combination of like DeFi and Web3 and all these like different things. I think you see something very similar happening with Cosmos right now, and especially a lot of These products that were sort of priced off of Main Net Ethereum because of, you know, Defi just like takes up all the bandwidth on main Net Ethereum. I think a lot of these projects are starting to like migrate onto their own chains, which is like, you know, so some of these like for example Sentinel, they used to be on Ethereum and they actually migrated onto a Cosmos chain for this reason. Now let's talk about Solana. We all know that scalability is one of the most important issues facing the blockchain industry today. The Solana blockchain has been engineered from the ground up, optimizing for performance and scalability. The network supports thousands of transactions a second with 600 millisecond block times and over 500 different validators. It's not a sharded blockchain, but a single blockchain hyper optimized for performance. And that makes it easy to maintain composability between the apps on Solana. So they work together seamlessly. The Solana ecosystem is growing at a rapid pace and it's a great place to build your project and get involved with the community. So go to solana.com epicenter to learn more. What was the Ammo landscape look like on Cosmos right now? Are you guys the first AMM? Are there other AMMs coming about and does it look like, you know, uniswap constant factor market makers or is there new ideas being pursued in Cosmos? Yeah, so we're not the first AMM built using the Cosmos SDK. There was, you know, so what? So for example, there's a project called Secret Network that they built this like SGX based smart contracting system and they have a AMM on there called Secret Swap. So that was running. The biggest one is Thorchain. Right? So Thorchain is this like I absolutely love the thorchain team and project and they were sort of the first people to build like an AMM using the big major scale AMM using the Cosmos SDK. And they don't have IBC enabled yet, but I know they will be working on that very soon. But they're kind of focused on a different market where they're trying to connect to Bitcoin and Ethereum and a bunch of Bitcoin esque chains. So they're kind of going for a different target market than we are right now. But yeah, we were the first ones to have sort of IBC enabled and being able to provide liquidity to all these tail end assets that are being built on the Cosmos SDK right now. And then there's more AMMs in development right now. So the Cosmos Hub actually is building their own AMM called Gravity Decks. I have all these reasons why I think that, like it's not, you know, I was, as a core developer, I was heavily against putting a Dex on the hub, which is what, you know, when we started Osmosis, we actually started from the idea of like, hey, we're going to build this as a Dex on the hub. And as we thought through it, we realized, well, there's all these different reasons why building something on the hub like this doesn't make sense. But we were outvoted in on chain governance. And so there's a Dex on the hub now called Gravity Dex and they launched like two days ago actually. So, yeah, I think the, you know, AMMs are obviously one of the most fundamental building blocks of Defi. And I think, you know, the way that you're going to see DEFI explode in Cosmos is you need to see multiple AMMs sort of all competing with each other and trying new things. And so I think the Gravity Dex people are working on certain things. What we have right now is actually very similar to Balancer. I mean, admittedly, I'll say it's basically a clone of Balancer. We took the Balancer code base and reimplemented it in the Cosmos SDK. We didn't take the code base, we took their white paper and reimplementing it, reimplemented it in the Cosmos SDK. And that's sort of our basis. And then on top of that, we're going to start to build all these like cool innovations that we have coming up soon. Why not build an AMM on the Hub? What is your philosophy there? Yeah, I mean, so I think that the Cosmos Hub is really the core feature and use case of the Cosmos Hub is shared security where there's a number of, you know, example osmosis. You know, we are, we want to secure way more assets and, but you know, we just don't have the security that the Cosmos Hub does. Cosmos Atoms are this like multi billion dollar market cap coin and that, you know, I think what it should be doing is taking that security and offering it and leasing it out to other chains that want it, similar to what Polkadot Relay Chain does. And I think that's really what the Cosmos Hub was meant for. But I think by putting a Dex on the hub, it kind of ruins the credible neutrality of this like base layer L1. Like the cosmos Hub really is supposed to be this base layer L1 and by putting a Dex natively on it. And I've just noticed, like, a massive shift of resources within the Cosmos Hub development team going towards this Dex, as opposed to, like, shared, where it should be going, which is shared security. You know, imagine Ethereum had a native Dex that was, like, built by the Ethereum foundation and, like, instantiated in the chain. It's like, would Uniswap have been built on Ethereum? Would. I just don't. I think, like, you look at. You look at, like, Binance Smart Chain, right? The only people building on Binance Smart Chain are people who are being basically paid by Binance. And if you don't have that, like, credible neutrality at the base layer, it's very hard to, like, build an organic system on top of your L1, which is okay, because, you know, the Cosmos Hub is not this, like, the basis of the Cosmos ecosystem. The Cosmos Ecosystem is sort of a much larger vision than just the Cosmos Hub, and they are sort of distinct. But, you know, as someone who is also very, you know, active in the Cosmos Hub side of things as well, I would have liked to see it go on a slightly different path. But I think it could still happen. I think. I'm willing to bet that within a few months, Gravity Decks will move on to its own chain. What do you think will cause that? Just a realization that this shared security model is probably where the real value lies. And this gravity, the Cosmos Hub should not be competing on becoming a DEFI application. And I think that people will start to realize that, as I think it'll look bad for the Cosmos Hub if it tries to become this DEFI application, then isn't able to compete at being the best DEFI application. Because the problem is it wasn't designed to do that. Like, you know, every. All the tokenomics we've designed and everything that went into the Cosmos Hub went in with a very different purpose. And so it'll be interesting, I think, eventually, you know, hopefully they'll see that.
**A** (19:59):
You said earlier that osmosis kind of launched as a clone of Balancer. You kind of position yourselves as an experimentation center for amms, though. Can you talk about that for a little bit?
**B** (20:15):
Yeah, sure. So there's what I call the Cosmos mindset. And the Cosmos mindset is this idea that, like, we believe in localized systems. We believe in systems where, like, you know, spread a thousand flowers and see what blooms. Right. And that's sort of the vision of the cosmos. Like, okay, let's have all these chains. Some of them will fail. We'll See what ends up happening. And I think Balancer, is that the closest thing to that mindset, but for AMM design, where I think it's the most like, generalized MM that exists currently on Ethereum with like as much flexibility and like, you know, you can change the weight and create the pool. You can have like different multi asset pools. You can have, you know, you can have two different pools for the same asset pair. And is that the most efficient? Probably not, but I think it's the most in line with the Cosmos mindset. And so that's why we were most intrigued by the Balancer design and we decided to implement that. But as we go on, I think what's going to. So one of the key visions that we started with when we started Osmosis was I wrote a blog post almost over a year ago now, back in maybe like February 2020, where I said, it's called Daoifying Uniswap. And I basically said, hey, this AMM thing is really cool, but it's way too simple right now. This like X times Y equals K. Cool, but I think it needs to become more powerful and just more smarter algorithms. And since then Curve came out and Balancer came out and all these like, a lot of AMM innovations happened obviously in the last one year, but I think it could go even further. I think, like, and we're starting to see that now as well, right? Like now you're seeing all the V2s and the Curve V2 and Unisoft V3, and you're seeing like it's still going forward. And essentially what we wanted, this whole idea of being the AMM laboratory, what that meant was we wanted to build a place where it's very easy for people to design custom AMMs and have them like. And then Osmosis provides like all the standardized interfaces and UIs and everything. But in the back end, the AMMs could be working completely differently. And a lot of this actually came from like, watching Zubin design a lot of the AMMs for options that he's been working on for like, you know, many months now. And so, you know, I saw that when I saw him like designing that. That also, like, helped me realize that, yeah, look, the design space of AMMs is still so much left to explore because I think as we have like new types of financial instruments, each financial instrument will like, need its own specialized AMM essentially. And I think these current ones are great for spot assets and like very specific spot assets at that. But I think, yeah, I think there's so Much more room to go. And when I say amm, I'm actually, you know, I don't think, I don't think of them in terms of like this X times Y or like this like two side amm. I think like I have this meme that we. Or we have this meme internally which is like AMMS ITE but bonding curves though, because I think bonding curves are the original AMM essentially. And I think there's a lot of cool design space to be had there when you explore with one sided AMMs for creating financial derivative assets and things like that. So I'm a big user of Osmosis and I'm just wondering what's your roadmap? How is osmosis going to change the world and change defi and when are you going to also have a stablecoin on osmosis? The first part for the roadmap we have a lot of things we want to work on. But so we basically we've prioritized three main big pillars that we're going to focus on for the next six months and these are sort of the three things we want to start with. So one is better AMM designs and, and so whether that includes things like concentrated liquidity. One of the ones that we're really working on is batching because I think, because that kind of feeds back into a lot of MEV stuff that we're doing where we want to make sure that like okay, if you have two trades coming in different directions, they could be batched against each other. And where it comes from really how we think about it is that you have, it'll make actually maybe more sense if I come back to the batching in a second. The second piece that we're working on is a something called threshold decryption. So that is our sort of grand project to like solve front running which is it's a method of encrypting all transactions in the mempool and they only to a threshold key of all the validators. And once they, the validators commit and finalize a block in the order of terms of ordering in a block, then and only then do the all the transactions get decrypted and executed. And so this basically makes it impossible to front run because you can't read what's in anyone's transactions. And this like no more sandwich attacks or anything like that. And you know, this is like I mentioned, this is what we did a lot of fall where we were exploring all the different techniques of doing mempool encryption. We had like sgxs which has its own host of issues for many, many reasons. You have time lock encryption which has its own set of issues for like around UX and security and stuff. And then you have threshold decryption. Threshold encryption also has its own issues of course. But we spent a lot of time looking at the trade offs and we came to the conclusion that okay, this threshold encryption model is definitely the way to go. And so that's what we're building right now and building it into the tendermint. We have our own fork of Tendermint core where we're adding this in natively. And so yeah, so that, so that that will help solve the front running and then, so then going back to the batching. Why we think batching is important here is if you have a system where no one can read the transactions in the order, the transactions that are about to go in the block, you know, the order of transactional block are really just random and they're kind of meaningless. And so it seems. But that what order they end up being executed in means that everyone gets a slightly different pricing, which seems arbitrary to us. Like why should this random ordering affect what pricing people get? And so that's why we want to create a batching system to make and so to create what we consider a fair mechanism. And we have, you know, I think people often use the word fair but then they don't have good definition. They don't define what they mean by fair. And so we, what we've actually done is we've come up with a definition of fair which means that you take all permutations of all orders of transactions in a block and everyone will be getting a different pricing. Let's give everyone the average price across all permutations of orderings within a block. And that's like the execution clearing price that everyone would get. And I feel that's like that basically removes the arbitrariness of transaction ordering within a batch.
**A** (27:21):
I have so many thoughts. So in a way what you're doing is basically you're amalgamating, you know, from, from an Ethereum point of view. Shutter Network and Cowswap said. Correct.
**B** (27:36):
Yeah, yeah. So we, we've been working on this thing for a little, for about like nine months now. Yeah, so shut up. Shutter network school. The problem with doing it on a. So the reason why it's beneficial to do this on your own chain instead of on Ethereum is you just have so much more flexibility on like what you can do on your own chain. So for example Shutter network has to be done as this sort of like overlay system on top of Ethereum, but with your own chain. We have the ability to bake it in natively into consensus protocol and make it such that decryption of transactions has to happen atomically and in lockstep with consensus itself. And there's no. You can never get into a situation where anything is committed but undecryptable, and you can't. Or the other way around. If something is decrypted but wasn't committed, we have all these ways of creating slashing conditions for that. So you essentially get the same security as your consensus protocol itself.
**A** (28:29):
Okay, and then the second part with the batch auctions. Yeah. So basically, from my own experience working with Gnosis Protocol V2 and Cowswap, basically finding optimization conditions that make sense is super difficult. I mean, so basically, even if you look at the average prices, it depends on what you measure the average prices in, Right. So depending on what your basis vectors, the average price and the fair price is going to change, right?
**B** (29:02):
No. So can you explain that a bit more?
**A** (29:05):
Yeah. So basically, if you look at. At the average prices of different assets and you basically you say you go for. For the average, it depends on what you measure these prices in. Right. Basically what your unit is. Your unit, unit of account. Depending on what your unit of account is, the average is going to vary.
**B** (29:28):
Yeah. So you actually have to take it. So we're dealing with four types of orders right now, which is out given, ends in given, outs, and then the same for the other direction, right? Out given, in. So a give me the out given in for asset A. Given in given out for asset A. Then give me the outgoing in for asset B and then the in given out for asset B. And basically we need to create an algorithm that gives us the. Takes all four of these transaction types and gives every individual the average clearing price. And I would call that. I don't know, maybe you're right, maybe the word fair is not the right word to be using here. But I think it's the one that removes the bias of ordering. And I think that's what we're defining as is the. What is the. Because we. So, you know, epicenter hat, you know, we just did the episode with the Gnosis protocol and you know, there they had a sort of a different definition of what the utility function for your batching system is, where they have this concept of surplus value and you want to maximize the surplus value that's given to traders that's based off of where they think the slippage back, what prices they're willing to accept versus what clearing price they get. And I guess we just have a very different definition of what is our batching mechanism trying to optimize over. And what we're trying to optimize over is removing the bias ability of ordering because you know, we kind of remove 99% of the MEV via our front running resistance and all we're left with is the arbitrariness of the ordering. And we're just trying to remove that arbitrariness.
**A** (31:07):
Okay, but basically you're, you're very much in the MEV must be combated camp because I mean there's also the converse camp that says MEV must be extracted. Right. So basically the Flashbots camp, yeah, I.
**B** (31:22):
Am very much in the mev can, must be combated. And it's this like, it's like a moral issue for me. I think, I think this is fundamentally a privacy violation. Like you should not. The fact that you're able to like read other people's intents before they commit them seems like a privacy violation to me and that needs to be combated. There are specific forms of MEV that I think. I actually gave a talk at the Flashbots like MEV conference last week where I kind of categorized the different types of mev and you know, it's pretty good categorization. But when it comes to, you know, you go down this tree of categorization, you get to this thing called transaction ordering manipulations. And within that there's what I call relative ordering and absolute ordering. Relative ordering is when you're trying to position your transaction in relation to someone else's transaction. And that is what I. But to do that you need to be able to read someone else's transaction. That is what I'm trying to stop. Right. I don't, I think that is like where the issue comes. That's what said sandwich attacks and all and like you know, back running and all these kind of things that, that, those are all examples of relative attacks. Now you have like what I call absolute ordering attacks where like you're not trying to position relative to someone else's transaction, you, you're just trying to position yourself in a block. So you want to be the first person in a block, for example, because you want to go run some liquidations or you want to go do like some exist arbitrage based off of some existing state of the network. Right. That is, I think is a fair. That's not what I'm super concerned about, I mean, you know, we've spent some time figuring out solutions or how if we did want to get rid of that, how we could do that. But I'm not concerned that, I'm not convinced that that's what we need to get rid of. And I'm in that kind of thing. I'm open to having like auctions and things like that. What I'm not okay with is having auctions in order to extract value from other users.
**A** (33:23):
That's totally fair. So maybe if I zoom out a little bit, what made you build this on Cosmos with a very, I'll be gracious and say nascent defy scene. I mean it sounds like you have a lot of very advanced opinions and ideas of how to combat these things that are currently happening in a really large scale on Ethereum. But you're building solutions on a network that currently doesn't have that problem set. How does that fit together?
**B** (33:58):
So keep in mind I'm biased because I've been building the software for Cosmos for four years and I do think it's the best software blockchain development stack that exists today. I think that, and part of it is because to do a lot of the things that we're trying to do, you need this like low level native access to your chain. So you know, like I said, threshold decryption, we need this at the, we need, we needed to like change the consensus protocol itself in order to make this happen we needed to like one, we needed to invent new cryptography to make our threshold encryption scheme work and then we needed to like integrate it into the chain which is something you know, we can't do on Ethereum today or our batching systems. Right. We need to like build like we need, we want code that executes automatically at the end of every block. And yes, you can do this on Ethereum by building overlay network. You can build relay networks, you can build shutter, you could build, build all of these things. But it's like, you know, I just don't think that's the right ux. I want this to happen natively in the chain and also as I mentioned as well, very importantly with the same security guarantees as your chain itself. Where I don't want this like threshold decryption to have us different security assumptions than the consensus protocol. And so that's kind of why it's important for it to be on its own chain. Now just because it's on its own chain doesn't mean it not going to be part of the Ethereum ecosystem. I mean, I don't know if people realize Polygon is basically built using the Cosmos SDK. It is like it's built on Tendermint core using the Cosmos SDK. And you know, I don't think anyone would argue that it's not a core piece of the Ethereum ecosystem right now. And so, you know, our osmosis will, you know, currently we focus on just connecting to the IBC chains right now. But like very soon we'll be connecting to the wider Ethereum ecosystem as well.
**A** (35:59):
So basically the bridges are something that is kind of a mystery to me how exactly it works. So basically the bridges within the ibc, they're kind of built in, but there's also the gravity bridge, right, that bridges the IBC networks to Ethereum. How does that work? And can you in principle actually have an AMM that works across this bridge?
**B** (36:28):
The gravity bridge is essentially. So IBC requires building a light client for the other chain in the native in another chain in that chain. Typically what we do is build this into the core code base of the chain itself. So in the Cosmos SDK, it's in the software of the Cosmos SDK to be able to do this. You can also, if you wanted to write this light client system at a smart contract layer. So you know, I think there's a team working on doing this for Solana where they're, you know, Solana is reasonably scalable. Thank you sponsors that they can like, you know, you can actually build an IBC a tendermint light client in a Solana contract and have it be performant enough. But the problem with Ethereum is it's just too expensive to do that like writing a light client, a tendermint like client in Ethereum people are working on it. I mean that people are working on it, but it's just a much harder lift than going the and just too expensive especially you know, six months ago where it would just have not been feasible from a gas perspective. That's why for the Ethereum to Cosmos bridge we needed to go down the more traditional typical bridge route, not using ibc. And so in that what happens is you'll essentially, you have this chain that is the bridge, gravity bridge and you have the validators of that chain running Ethereum nodes and acting as the like, like the, you know, they all decide what when they witness an event happen on Ethereum and then they, they declare their witnessing of the event happening on the cosm on the grav bridge chain and then once they declare that once there's enough of a quorum of them doing that then it triggers, you know, admits the asset that's being bridged over.
**A** (38:21):
So what does that mean for an osmosis Ethereum AMM bridge? Does it mean it's too expensive or does it mean it's theoretically possible but not feasible?
**B** (38:33):
No, it means that you'll be able, very soon we'll be able to bring Ethereum assets over the bridge into Cosmos land and then trade them on any, you know, you can bring them onto osmosis and trade them on Osmosis and you know, they'll be like a multi hop thing but like, you know, it'll be our UX will like handle all of that where it'll be like, you know, we want to make it as simple as possible. Whereas like you know, deposit. Ask Ethereum if you go on the osmosis website today, like, you know, you'll notice we actually don't use the word IBC anywhere because I really do think IBC and bridges and all of this stuff is like lower level things that users should not be the ones worrying about. And so on the osmosis website there's like a deposit button and a withdraw button which is like very similar to what you see on centralized exchanges today. And you know, it'll be the same where it'll have a button where it's like all right, deposit eth and it will just your eth will show up then after a couple of seconds into your osmosis wallet and then you can do whatever trading you want and then you can press withdraw and send it back to Ethereum. One question I had was how do you bootstrap a new chain and make sure there's security because you're not drawing from hub security. So what was your process of making sure that there was security from day one? A strong validator set, sufficient capitalization of OSMO, etc. Yeah, so one of the things that we did was we actually distributed the Genesis OSMO via a quadratic fair drop. What that means is it's basically an airdrop to atom holders. The quadratic airdrop was basically you. Everyone got airdropped based off of the square root of the number of atoms in their, in their address. Obviously we could not announce this before the airdrop happened, so we had to take a snapshot first and then we announced the quadratic nature of it. But yeah, this was basically a way of like making sure the distribution is like even more decentralized than the, you know, the atom distribution. And also because you know, traditionally Whales are going to be the ones that get more of the liquidity mining rewards. So we wanted a way of just giving every, all the atom holders a little bit of more distributed stake. Unlike a lot of DEFI projects on Ethereum, our token is also the staking token along with being the governance token. And in order to be the staking token of the network, you need to be sufficiently decentralized from T equals zero. You can't be like, okay, all the tokens are in the hand of the team and we're going to have this liquidity mining program and then look, five years from now it's going to have a decentralized token distribution. We needed something decentralized from day one. And so that's why we were able, we did this airdrop to make sure that had it. And then you know, a number of. Because it was distributed to atom holders, that means all the Cosmos Hub validators by definition were airdropped OSMO as well. And so that's how we were able to bootstrap the validator set. We did. You know, I know like a lot of new chains have to go look around validator programs and all of that, but you know, because all 125 of the cosmos Hub validators had atoms, you know, I'd say probably the majority of them also became osmosis validators as well. And then how did the chain get enough security? I mean, I don't know, market dynamics I guess. I mean the price of OSMO became enough from the get go to secure the number of assets that were bridged onto it. What is ion? Because that's something else that people have mentioned has been airdropped recently. Yeah, so ions are this, it's the secondary token of osmosis. And basically I could tell a little bit about the story of ions where what happened was I wanted the smallest unit of OSMO to be called ions and it was, you know how like ether has whey or bitcoin has sats. So but then one of my teammates said no because he's like, oh, you're going to make the wallet's job more difficult. Because the standard in Cosmos is you use metric prefixes for denominations of assets. So if you have a thousand it'll be milli osmo, you'll have micro osmo, et cetera. And both wallets sort of expect this. So I'm like, okay, fine. But then when we were doing the test nets, I still just really like the Ion name. And so testnets were looking at amm. I need two tokens to test with. So I just used to put a second token in there called ions just to test my AMM with. And then two nights before the launch I just, I don't know, I was really tired or something, but I just had this like, stroke of inspiration. I'm like, what if there was a second token in the genesis of the chain and I kind of just pulled this like all nighter. I, I don't really even remember everything that happened that night. But like they always, like I woke up in the morning and I was like, all right, I have a Token distribution and let's just add it to the chain. And I don't know what this thing does, but, you know, maybe people like it and do something with it. And what's really cool is there's a huge community that has like formed around ions. Like, it's probably one of the most active chats I've ever seen in my like, life. And it's one of my friends, Dylan from Commonwealth, we had him on actually a couple of years ago. But yeah, he's working on a project called Commonwealth. And you know, he has this whole idea called Token Curated Communities. And I was honestly a little bit skeptical of it until I saw it happening in front of my own eyes. And like I accidentally created one where it's like this ion thing. Just created this like token curated community where it's just. And they, they're having like every other day they have community calls for like, you know, all these like really active governance proposals. And, and what's crazy is Jay, who is like, you know, the founder of Cosmos, who's sort of been stepped away from the Cosmos ecosystem for the last year, he actually has gotten involved with this Ion thing as well. And he's in the chats like every day now talking about it. He even has a bounty up for people to try to reverse engineer the distribution criteria for ions, which is kind of funny. Yeah, so I don't know, it's just a big experiment to see what would happen if it's a meme coin and you know. But can the memes become real? We'll see.
**A** (45:15):
Well, knowing that you're a big proponent of doge, I kind of know what your answer would be. Tell us more about this osmosis community. Because basically community building has been one of the least planable things in the Defi ecosystem. And a lot of the Ethereum defi communities are also, I'd say, peculiar in nature. So how would you describe the osmosis community?
**B** (45:45):
Yeah, the osmosis community I mean, I think it's still, like, so early, and it's, like, still trying to find its, like, aesthetic or its own, you know, culture. What it was was Zubin and I about, like, two or three months ago, we made this, like, political compass chart, if you know that. The political compass meme, where it's, like, left and right and then authoritarian, libertarian. And we planted every, like, major crypto community on this, like, chart. It didn't make that much sense, but it was, you know, it was a funny meme. And, like, you know, okay, bitcoin is very libertarian right. Etherium is like, mock, like, left, like, moderate left. And then, like, somewhere, like, not as libertarian as Bitcoin. And, like, Dogecoin was, like, the farthest left. And we had. We had this whole, like, thing, and I think it's on zoom in Twitter. And then someone asked me, like, oh, where does osmosis fit on this? And I'm like, I don't know. We're figuring that out right now as we have this, like, literal conversation is like, this discussions that we're having in these early days is what's setting the community culture. And so what's cool is a lot of the. I mean, we were able to bootstrap a community pretty quickly thanks to this airdrop where, like, you know, the atom, we kind of. A lot of the atom community has come over and become osmosis community as well. And then also just a lot of the communities of all these, not just the atoms, but, like, you know, we connected to all these, like, other chains, right? You have the crypto.com, you have the Akash network, you have the region. All of their communities have also sort of become part of this osmosis community as well, because this is sort of the first time that a lot of these projects have had liquidity on their tokens. And so a lot of their users have, like, their. And by. By being LPs on it, they're earning OSMO tokens thanks to liquidity rewards. And so now, you know, so they're also really excited about being part of this osmosis community as well. And then there's obviously, like. There's a little bit of drama that happens sometimes between the osmosis community and the Cosmos Hub Adam community. And I really think. I don't know if people have read, like, any of, like, Rene Girard's work on, like, mimetic conflict and stuff, where it's like, you fight the hardest with the people who are the most similar to you. And I think that's sort of what's happening a lot right now between some of the infighting between the Adam and Osmo community. But I do think that's sort of a temporary thing that, you know, I think, I think will get resolved over time. And I think the osmosis community is really supposed like we try to be as pro atom as possible because we are. There's another meme I had which is like the Father, the Son and the Holy Spirit is atoms, osmo and ions. I don't know what that means, but it fits really well.
**A** (48:30):
Well, you heard it here first. So sunny. I mean it's been a couple of weeks and the community has kind of blossomed. Where would you say it fits on the four quadrant model that you just talked about? I mean you said that you didn't know before you launched it, but now it's been around for a little while, right?
**B** (48:51):
To be honest, I still don't know yet. I don't think it's something that gets set in four weeks. I think, I mean it took Bitcoin years to find out its culture. It took Ethereum years to find its culture. I think culture is not something that is defined in four weeks. I think culture is this more long term process and I actually legitimately do think it's too early. I think what's happening right now is the community is finding its. I don't like to use the word but leaders or influencers and there's people who didn't even get any tokens in the airdrop but they just have gotten really passionate about the project and have gotten like, are stepping up as these community leaders and are like, you know, hosting governance community calls and like writing governance proposals and stuff. And so I think, you know, I think we're still in this very early stage of it that it's hard to really pinpoint anywhere yet. What's your role in the community? My chief troll. I don't know. I like to spend a lot of time in the community. It's fun because it reminds me of like early. I used to spend a lot of time in like the Telegram chats back like when before the Cosmos Hub launch. And no, I really enjoy doing that and I started spending a lot of time in the Telegram but sometimes it's a little bit too much where I'm like, I need to get away from this thing and go back to like building the actual product. So yeah, I mean my role is I'm the head of the development team and so, you know, we want to, we have A, we have a roadmap that we are really passionate about and we, and our goal is just to execute on that. Roadmap. Roadmap. And then we hope that the community sort of likes this roadmap and works with us on pushing this to the larger crypto ecosystem. But you don't want to set culture. You think that that's impossible to do from a single person. Yeah, I think single people can influence culture. I mean, I think I share my thoughts in the community quite a bit and the kind of things I'm thinking about, I think it's a dialectic process. I think it's like I'm communicating with the community, but the community is communicating back. And obviously I wouldn't say I don't have any influence over that, but I try not to have too big of an influence over it. I want the community to figure out its own path, which is another reason why I try to not be especially on the ion one especially I try to not be too active on there. I'm sort of passively following along and when someone asks questions or you know, someone, someone has something interesting and I want to like, you know, for example, people were talking about Web of Trust stuff in there yesterday and I, you know, if anyone knows me, I've been super passionate about Web of Trust for like number of years now. And so, you know, I started telling them about, you know, brain dumping, about all my ideas about Web of Trust and I talked about circles a lot and I talked about like, you know, all these cool things.
**A** (51:49):
So yeah, you said that OSMO is both the staking and the governance token. So let's talk about governance a little bit. So what does osmosis need? Governance for which parameters are governed?
**B** (52:04):
You mentioned the staking token. One thing I just want to quickly mention really quick, the third thing that was on our roadmap, that's like the priority. So one was the amms, the other was the threshold decryption. The third is we're building something called Super Fluid Staking, which is this idea inspired heavily by like Dan Ellitzer's post from a couple years ago called like Superfluid Collateral. But the idea is we have OSMO is one of the main base pairs of the AMMs, but it's also the staking token. And this has this weird tension on its security model. And so what we're trying to do is what we're going to build is this Super Fluid staking. So you have these normal staking derivatives where what happens is you take a staked Asset, you stake it and then you get back this derivative asset and then you go use it in defi. The problem is this completely undermines this entire like security model of proof of stake where it's like, oh well, I mean you were supposed to lock up these assets and have them be easily slashable. But if you could just like liquidate the asset, you know, it defeats the whole point. What we're going to do is put this in the reverse direction. So we're saying what if you use staking tokens in defi and then take the defi assets and stake them. So you know, the simple example would be something like compound. Like let's say there's a osmo, I can go lend it on compound and get back C OSMO and then I can take my C OSMO and stake it. And as long as there's a way for the protocol to efficiently know the conversion rate of like how much OSMO is underlying this defi position, it should allow, it should, it should allow you to stake it and treat it as that because that's still slashable. And so while we're doing, the one that we're starting with is with the LP shares. So if you have a pool that contains both OSMO and Food Coin, you can then take that LP share and stake it and it will be treated as the underlying value of the OSMO half of it. And so that, that way it now when you don't have to choose between staking and LPing Osmo anymore, you can do both at the same time and be earning both the staking rewards and liquidity rewards. And this will just help the security of the chain. And yeah, so I really do think this like idea of super fluid staking is going to be, it's going to, it's going to be help a lot of the solution for how to like what you're asking earlier on. But like, okay, in cosmos, like don't you have this like security issue? I think this will be part of the solution for a lot of these chains. Okay, so that was a long way to answer that. Now to answer your governance question, what, what are these things going to be used to govern? So I mean a couple of things, right? I mean there's obviously. So with superfluid staking you actually do need to specify which pools can be used for superfluid staking. Because you can't have a system where like, you can't allow people to just like bring in any. You can't create a pool of like OSMO with like Shitcoin that I can mint infinitely. Because you can't. If you stake that, then you can just like drain the OSMO from the pool. So governance has to manually approve all of them. That's one thing. A lot of like the roadmap and software upgrades have to be approved by governance. A lot of like the sort of like new AMMs and stuff. Like I, I really, it really is like, you know, there's a lot, I think there's a lot of parameters in the system that will, that will like have to be improved by governments and most of them really do come in the form of like roadmap esque like software upgrade style things.
**A** (55:42):
I'm not sure whether I understood that correctly but you have per pool governance, right? And not holistic governance. Why did you do that?
**B** (55:53):
Yeah, because I think that the people who best understand the needs of their pool are the LPs themselves. So let's say, you know, you have this pool that's really popular. It's, you know, I don't know, let's say it's an Osmo ETH pool and it has a 0.3% fee, swap fee. And you know, let's say what's interesting right now, what's happening right now in the crypto space or in the AMM ecosystem is there's been this like 0.3% meme that's been like standardized with Uniswap for the past like three years. But I think the entire space is in this process of like re figuring out what the shelling point is going to be for this, for this, for what swap fees are going to look like. And so I think, you know, you want to wait for the LPs of this pool to basically use governance to say hey, let's go ahead. And like, well you could, what you could do is put a new pool and then have people like fork off and like try to add liquidity to this new pool. But that has like all these issues with like, you know, who wants to be the first one to take the leap? And like, you know, you fragment your liquidity. So instead if you have a way for the LPs of a pool to say, hey, let's all collectively agree to change our swap fee and if people don't like it, they can fork off, right? They can, you know, we have like a rage, we'll have like a rage quit feature where like people can like automatically say okay, no, we want to stick with the old parameterization of the pool. So then we'll spin up a New pool pool that maintains the old parameterization. So that, that's kind of why it's a way of not splitting liquidity. Essentially. That's sort of the main reason why you want LP governance.
**A** (57:35):
I see. Now I, now I think I'm wrapping my head around it because when I first heard this it seemed like needlessly complicating things. But when, when you frame it like that, this is, it makes sense. So can you give us an idea of what the current usage of osmosis looks like? How much liquid liquidity iron pools and how much trading is going on?
**B** (58:00):
Yeah, so I think we have about $100 million of TBL right now, at least when I checked last and then the trading volume, it was somewhere between like 3 to 5 million dollars per day. And so yeah, that's kind of, you know, not bad for one month in. We're, we're pretty saturated right now where I think it's like a lot of the people who in these cosmos chains wanted to like provide liquidity are already providing and obviously we, you know, we're going to continue to do that, but I think we're going to see the big flood of liquidity come in once we get the Ethereum bridge up because that's going to be when a lot of the people who, you know, all the assets currently are on Ethereum. But actually there's also, I'll be a caveat that with like, you know, like I said, I think there's more cosmos chains coming, coming all the time. I think, you know, just in the last month since we launched 32 new ones that have added IBC and I think in the next coming months there's gonna be a lot more ones and then the other ones are like, you know, there's a couple of big powerhouses within the cosmos ecosystem like Thor Chain and Terra. I think Terra actually has a huge swath of interesting assets. So because they have like, you know, they have like stablecoin, they have their, you asked question about stable coins. But yeah, the Terra has a stable coin called ust. They have all their anchor assets, they have all their mirror assets. So I think they have a lot of interesting assets that will be able to find liquidity on osmosis. But yeah, once we get Ethereum connected, you know, we're going to be able to bring in like a lot of the stable coins that are on Ethereum. We're, we're going to be able to bring in like you know, eth, all the ERC twenties wrapped Bitcoin. So and what we're really trying to do is like, you know, we want to focus on making sure we have the best trading infrastructure for these things when they come over.
**A** (59:51):
Cool. Super interesting, Sunny. Where can people find out more about the community? Where should they go to join you guys?
**B** (1:00:01):
Yeah, so our Twitter, ismosiszone, our website, you can just go to Osmosis Zone and there, you know, you can click the button to enter the app and like, you know, trade and everything or, you know, that should also have links to our telegram and Discord and things like that. And that's basically the best way to find. Learn more about Osmosis. The telegrams are especially some of the more active. And so if you really just want to. If you want to drink from the fire hose of the Osmosis community, I think that's the. Yeah, that's the place to go.
**A** (1:00:36):
Cool. Thank you. It's been fun, Sunny, having you on the other side of this. We'll have you back on our side next week or so, but thank you so much for coming on.
**B** (1:00:47):
Awesome. Thank you, guys. Thank you for joining us on this week's episode. We release new episodes every week. You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google Home or Alexa device, you can tell it. To listen to the latest episode of the Epicenter podcast, go to Epicenter tv. Subscribe for a full list of places where you can watch and listen. And while you're there, be sure to sign up for the newsletter so you get new episodes in your inbox as they're released. If you want to interact with U.S. guests or other podcast listeners, you can follow us on Twitter and please leave us a review on itunes. It helps people find the show and we're always happy to read them. So thanks so much and we look forward to being back next week.