**A** (0:00):
Welcome back to the rollup. I'm Andy and today we have a second episode with Sunny Agrawal, the founder of Osmosis. We talk about incentivizations, about the recent NFT community launches of sloths as well as mad scientists. We talk about some of the design choices of Osmosis and the modular thesis, and we dive deeper into some of the technical understandings for the current modular ecosystem. Sunny's got a very interesting way of presenting himself and loves to be polarizing and very fun to talk to. We really enjoyed this conversation and hope you will as well. Welcome back to the rollup second episode with Sunny from Osmosis. Here with Rob. We just finished module March and yeah, a lot to talk about. We actually got yelled at pretty bad by the Osmosis community for not including Osmosis in the map, so I apologize for that.
**B** (0:49):
It's okay. Don't worry. We still love the roll up.
**A** (0:53):
Yeah, nice, man.
**C** (0:54):
Leads to another. Now we do a full podcast episode in recreation.
**B** (1:02):
That's how that got scheduled. Okay, that makes sense.
**A** (1:06):
Literally.
**C** (1:08):
I think it's very, very. I don't think anyone disputes the fact that Osmosis is modular. So we can come out and clear the air now and we'll get a little bit deeper into the modular monolithic, the different ideologies, the different tech stacks that are prevalent in modular and monolithic a little bit later. But I think Andy has something that he wants to talk about first.
**A** (1:31):
I mean, yeah, what's going on in the Cosmos community? NFT land. I think that this modular sentiment has been so bad, basically, because honestly it's just a reflection of the Celestia price. And so Celestia has just been down only for a couple months. And then we get this random NFT mint on Stargaze, which actually has this capability before we get into the sloths. The sloths were minted on Stargaze, but you could. There was this interesting back end tech that was able to be enabled, I think through Skip, which made this Mint super easy. Maybe we can just start there. How did that happen?
**B** (2:09):
So I'm going to reveal some. I did not actually buy a sloth, so I did not actually participate in the Mint. But. But I believe what Skip did was like, you know, Skip has built a product called Skip API that kind of does like, you can do like trans. It allows you to like sequence like a bunch of steps doing a bunch of different Cosmos chains all in like one step, right? You make one transaction and it kind of like it goes like, boop. Does Something here. So what was happening was you can with TIA on Celestia Chain in a single transaction. It made it so it would like send that TIA to Osmosis, swap to Stars on Osmosis, send the stars to Stargaze and buy the NFT using those stars. So it was doing these multiple steps all in like so you could basically you're minting an NFT on Stargaze just by making a single Celestia transaction, which was like pretty cool.
**A** (3:08):
I know Algoric and Dean are working on this orchestration idea. It's a similar concept and it was so easy. It was a Friday night, I was just resting. I think I went out the following night for a little party. And that Friday night I had just seen some tweet and I'm sitting here with a buddy who's more of a crypto curious guy and I'm like, yo, we gotta mint these. So we minted a couple and for him all he had to do was send some, some TIA from Coinbase to Celestia, click a button and then click Mint. And for him it was like, it was a two step process, took about five minutes and it was super easy.
**B** (3:42):
Yeah, exactly. I know the stuff that Gore doing is that they're actually like making the dev UX around this even better. Like you know, you will be able like a developer who wants to do a cross chain interaction, they'll be able to just like write these as like each step is just like a single line of code where it's like okay, send from here to here. The next step is like swap and this next step is it. And then it'll like, you can like, it'll make it very easy to like handle the fallbacks like in a line of code like oh, if fail like do this like it. So they're making it. What Orchestration Gork team is doing is making it very easy for a developer to like write these cross chain applications. I'm sure like the Stargaze team had to work with the Skip team to make that UX work properly. The orchestration will make it so any developer can just start doing this quickly and easily.
**A** (4:30):
Yeah. So it seems like the sloths which are now up like 400x from Mint, they flip bad kids Today on the day that we are recording this, which I know has ruffled some feathers, but it seems like they're kind of branding as like this like modular community, which I think that the modular community really needed. I think that Celestia price action was just not great in the Last couple months and I think that people are needed some sentiment boost and yeah, I think it's pretty good timing.
**B** (4:57):
Yeah. Yeah. You know I will say that it flipped on floor price but the supply is lower. So it hasn't flipped on market cap yet. But yes, it's definitely, it's really just good to see like, you know, I feel like we've had Stargaze is this NFT platform and like it. It's basically had one success so far. Right. Which was like bad kids. Right. And like you had. There are other, there are honestly some like really cool other NFTs on like Stargaze that I feel like don't get enough love. Like there's this one I really like called Pixel Wizards which because it's like it an idea that I had for like a while I was telling someone about it. Like I want to like create this like NFT that you link to a coin price and then the NFT like changes emotion based off of how the performance of the coin is doing that day. And I, I always had this idea like just a little fun side project. It was like, oh, you should check out this Pixel wizard thing. It does something similar where. And I was like very cool nft but I don't know, fat kids got all the love and like all, you know, all the attention. But it's like, you know, now we actually with like the sloth. And then there's this project on osmosis called Mad Scientist. Like I feel like you need to have more than just one NFT class collection. Right. You. We're starting to. I think we're going to see like NFT season now in Cosmos.
**A** (6:14):
Like what's the story behind Mad Scientists here? Because I'm a little late to the party here.
**B** (6:19):
Yeah. So Mad Scientist is really fun because so the story behind it is okay there's this project called Backbone Labs. They you know, came from the Terra ecosystem and they kind of what they are really interested in like NFT fi. Which is something I've always been pretty interested in. And that's how I kind of got in. Got in touch with them originally. But they have done a very, I think very, very cool idea for how to build like an LSD ecosystem. So what they do is. So okay, most LSDs, right, they like usually take like a 10 like take rate off of the staking rewards. Right. That's like that somehow just become the industry standard. That's what stride takes, that's what lido takes, etc. And on something like stride, like you you know, the stride token gets 10% of the staking rewards from all of the LSDs that Stride does, which is, you know, it's, it's a pretty good investment probably. It's probably like the, one of the best ways of just investing like OSMO and Stride are probably like the two best ways of investing the Cosmos ecosystem as a whole, because one gets all the volume, the other gets the staking rewards. But you kind of like end up, you know, as Stride grows and grows. They have like over, like you know, over 10 LSDs at this point, right. I think like the community, like they have to choose where do they put liquidity incentives and like, where do they like, you know, focus their limited marketing and attention, right? What Backbone did was something cool where they said, okay, for every LSD they launch, they're also going to launch a NFT collection. And what's going to happen is the, that NFT collection is going to get like 60% of the protocol revenue. And then Backbone's native token, it's called SOL S O U L, it gets 40%. So what's happening is, but now what you do is you have this like really passionate like set of community members. These like NFT community, NFT communities tend to be, you know, much more, have a lot more vitality and stuff. And so they're going to go and like promote that LSD within each ecosystem. So, you know, they have one on, you know, Migaloo, called Migaloo Whales. They have one on Kujira called the Kojirans, they have one on Terra. But so they basically built this one on Osmosis called the Mad Scientist. And so, you know, there's an LSD that they have for Osmo, it's called B Osmo. And mad scientists are going to get, you know, 6% of the staking rewards from, from B Osmo. And so this is like, it's cool because now you actually have an NFT that actually has like real yield on it and then on top, you know, and then kind of what happened was by doing this like mint on Osmosis, you know, Osmosis just is usually, you know, it's, it's the most active community in all of Cosmos, right? We have the most users, like daily users, most transactions, like all that, right? And so their like Osmosis Mad Scientist collection was just like a runaway hit where so they used this, we built, we designed this protocol called Stream Swap a while ago. It's meant to be this like fair auction mechanism. And so we designed it but then we gave it to another team called Omniflix to actually just like build out. And so there have been a bunch of stream swaps already like you know, I think and they're used for token launches on osmosis and I think over 10 at this point. But then this one was just like a runaway hit where like so, so what they did was they created a token called Lab and they sold it via this stream swap and then the lab can be burned in order to mint the nft. So it is sort of just like really streaming. You're auctioning off the NFTs eventually, but via this stream swap mechanism. And I think it did almost like 2.5 million Osmo or something was like, or 2.4 million Osmo participated in the stream swap. So it's like one of the biggest NFT collections, I guess mints in Cosmos now and then since then the mints actually open up next week. So since then the lab token has just started permissionless and it has already been trading on osmosis and it's also like, I think like 3x or something since then. So yeah, it's fun, it's exciting because it's like the first NFT on osmosis, but it's also cool because it's like using a lot of mechanisms I've personally been interested in for a long time, like fungified tokens to represent a collection like the Lab token using stream swap, using this LSD protocol under like real yield nft. So I don't know, I just think it's one of the most interesting NFT experiments right now.
**C** (11:20):
We spoke to Tarun from Gauntlet about the value accrual thesis in the modular stack and just a little bit of prehistory. Right? It used to be before modularity that you had value accruing either at the protocol level or at the application level and it was just like flowing up the stack or down the stack. But then I think the conclusion that Tarun and Andy and I reached was that in the modular world, in the Cosmos world, the modular world, value accrues to the users. And so we can see that very, very clearly in these nft. In NFT fi like, as there's real use cases for these protocols and these products, they earn fees and then that interest and that value tends to accrue back to the users. And I think NFTs are a great tool, instrument, if you will, to, to capture that value in the hands of the users. And so it is really cool to see some of the cost Savings, some of the protocol generated revenue as a result of these products end up flowing to the users and then, and then it's in the user's hands. It's like it, it is just much more peer to peer. It's much more like, you know, it's like the people who are using the thing are the ones that are reaping the rewards and I think that's the way that it should be.
**B** (12:34):
Another thing is actually I think the somewhat of the illiquidity of NFTs makes it almost a better governance item than like fungible tokens because you know like it's almost like a private market thing right where like you actually want NFTs feel a little bit more private markety where like the way of buying and selling them is yeah we have these marketplaces and they're getting more sophisticated but they, they're still more similar to like OTC stuff actually than it is to like an amm. And so you actually want the people like doing governance of protocols to have this like little bit more lock in. So I'm actually curious to see if we'll see more things like this where like projects do like you know, governance is actually done by an NFT series rather than by a fungible token.
**C** (13:26):
It's, it's also more akin to a nation state governance structure. It's more like one person, one vote. And so I think there's some very different voting and governance dynamics that come as a result of that, that type of governance structure.
**A** (13:47):
Yeah, it's like identity as well. It's like your identity, it's your person.
**B** (13:52):
Yeah. I think Dowdao, which is the most popular governance tool in Cosmos, I think it's going to be rolling out NFT based daos pretty soon. So yeah, that's really exciting. So I know the mad scientists are using this other one called Enterprise that was built by actually by tfl. Um, but yeah, so they'll be using that for like the, the government, the governance NFT daos basically. So I think, I think that will grow. There's also a product that we're, we kind of another like one of these, you know, idea we had that we like gave to another team to work on called Fractal, which is for fungifying NFTs. So it's like you know, you'll be able to create a token like a token that tracks the value of like the floor price of bad kids for example. So that way they can be like traded more. I think part of what made labs like mad scientists fun Was like the liquidity of being able to buy and sell lab was very easy. And now there's a game that a lot of people are doing of like, oh, do I keep my lab tokens for the liquidity or do I burn them to mint the NFTs and get the yield from the LSD? And so I think people are kind of debating what they're going to do. And so.
**A** (15:10):
And it's the same thing with the sloth, Sonny. It's like there's this new airdrop meta of like, oh, bad kids were included in a couple of these token launches. Now the speculation has gone kind of red for the sloths and for other NFTs, you know, and implements this game theory of like buying and selling and holding and when is the snapshot going to be taken? And wallets and different wallets and.
**B** (15:32):
Yeah, yep, yep, yep.
**A** (15:35):
So yeah, I'm curious what you think about that though. Is that meta something that is from a protocol token distribution, marketing expense perspective, is distributing to NFT communities something that you think is a wise strategy?
**B** (15:50):
I think so. I do think that NFT community shouldn't be based off of the expectation of airdrops. I think that just, you know, that's not like bad kids never was like done as like, oh, the expectation is that there would be airdrops. Right. It was like, it was just like. Honestly, I think out of all a lot of energy class series, I actually just. It sounds really like cheesy, but I actually did like the art. I think it was like, I think it was fun because what you could do is you can like look for the one that looks like you and you like buy that. I think that's like, that's what kind of what made crypto punks pretty cool as well. Because you like look for the one that looks like you. And so I think that it's. And then it was nice that like a community formed and then people are like, oh, this is a good community onboarding tool. Let's airdrop to it. Right. I don't think communities form like real communities form around the expectation of airdrops.
**A** (16:48):
Yeah, I agree. I mean we've been like memeing a lot of like modular stake tia, et cetera. But I think that we're going to, I personally will stop trying to, to do too much of that because it's just this, this community who just wants free money versus like the understanding and the underlying idea. But that's what's, that's like what's going on in the modular Thesis right now or the modular environment is like you've got this double edged sword and we talked about this with Ajit from Justin, you know Ajit Tripathi, the personality. We talked a lot about this. He's like modular appeals to like these like OG research, like Giga brain esque like in the governance form people and then you've got like this other side of the sword which is this like staking idea where all these people are staking for airdrops and there's like this like governance and it's just kind of like this like meme culture as well. And I, I think there's something to be said about like onboarding both ends of the spectrum in terms of like growing a community.
**B** (17:45):
Yeah, exactly. So you know I think, I think like I said, I think NFT communities tend to be more passionate than like fungible tokens because you know I, Yeah, so I think they are a great community onboarding tool and like a, A like a set of missionaries that can promote. You think that like I said that's what I think Backbone did well, which like oh, you have an NFT per chain that will then go. And that community will go promote your NFT or promote your LSD within that ecosystem.
**A** (18:16):
Yeah, yeah, yeah. I mean it's going to be interesting to watch play out. Another really interesting kind of event that's been going on is Solana's been pretty congested. So we've got this general purpose Solana single state machine, the kind of the epitome of, of monolithic has gone against sharding throughout. Their entire roadmap has stuck steady to this upcoming Fire Dancer upgrade. But then we, we get this Kamino airdrop, this wormhole airdrop, we get this ore farming thing happening and all of a sudden we've got a lot of failed transactions and yeah, I mean what's your thoughts on the current situation there? Is Solana roll ups going to come? Could they learn something from the Cosmos model or the Ethereum model?
**B** (19:03):
So to be fair I think most of the current issues on Solana's stuff has to do with their P to P stack more so than the chain itself being congested. This in my opinion has always been a bit of a. I don't know. I think the Solana's P2P stack has actually been something that I've been, I've always been a little bit worried about. Just the way that they don't really have a mempool in the same way that most blockchains do. You kind of just like broadcast your transactions to validators and you, you're, you know, every validator is almost effectively being ddosed constantly because everyone is trying to submit transactions to them like extremely fast. So I believe that the current issue actually was mostly due to a bug almost like in the P2P stack implementation where it's just not designed for these high P2P storms. But that being said, I do think that eventually Solana will, I just don't think one blockchain will ever meet the demands of everything. Right. And part of the what's annoying about building on like a, a generalized blockchain is that like you know, you have someone else's another application gets super popular and it makes your application like unusable or like, or way too expensive. Right. The nice thing about like osmosis is like hey, if DYDX suddenly starts getting like a shit ton of like transactions and like you know, that doesn't affect trading on osmosis. So no matter what, I do think that like you will have to like major applications will always need app chains. I think that that's, that's the thing. I think generalized purpose blockchains are useful for prototyping or like you know, initial go to market, like launch quickly. But and I, I think the DYDX is the ultimate like, you know, it's such, such a good example of like evolution of, of applications. Right. DYDX was originally a smart contract on Ethereum. Then they kind of turned into this roll up stack. But then eventually like okay, we need full control of our stack and so they kind of build their own app chain. But I think Solana is doing a good job at being this like bootstrap. I think Solana has definitely done a really good job at building a smart contracting environment and ecosystem and eventually it will just, they'll have to start building app chain now. That'll be cool. If we start, Will we start seeing. Actually we do have an SVM app app chain, right. Pith Pithnet is I guess the first Solana VM based app chain. Pith decided to fork Solana and they like they're running an SVM basically. So yeah, maybe we'll see more and more SVM based app chains. I wouldn't be surprised.
**A** (21:49):
Eclipse is coming. Eclipse is like Ethereum roll up using SVM a little different but yeah, but it's still not like an app chain settling to Ethereum. No. Yeah, it's going to be general purpose.
**B** (21:57):
Yeah.
**A** (21:58):
I was talking to the Monad guys because we had to roll up. We're like sovereignty own your block space. Don't have to depend on, you know, Billy Bob Joe who's, who's your neighbor who just his lemonade stand just got a bunch of business and now your chocolate bars are just can't sell them. Like we're trying to get, we're trying to get out of this mental model and get into like a sovereign block space model. Talking to the Monad guys about this, but they're like, well, you know, we think that we're going to be able to scale regardless. And I was like, no, we kind of need some Monad L2s or app chains as, as well. And then they end up asking me and I'll pose this to you as well. Do you have a mental model for app categories where customizability or sovereignty is like really needed and really good and also maybe like not so great or bad even when it comes to types of applications.
**B** (22:49):
I think best example. So one is, I think the biggest thing is do you need to make your validators do more? Like in Celestia you need. They needed to make their validators do all this like data availability stuff, right? And like so okay, they needed to modify the chain Thor chain. They needed to make their validators run these like bridges to Bitcoin, Ethereum and all this stuff. So they needed their own app chain, you know, dydx. They want to make their validators run like the order books in the mempool, right? They need their own validator set. So I think these are the kinds of things where like anytime you need your own validator set, you need your validator set to provide an oracle for something, right? That these are things that you want to use. Like I think an app chain makes a lot of sense. The other is if you want to have like very opinionated block, block production rules where like something like let's say on osmosis we can say like hey, a block can be produced but we're going to like execute these not in the order of transaction inclusion, but we're going to like take a block and we're going to scan through it. Look for all the like liquidity ads first. Then we'll execute swaps and then we'll go through and execute all the liquidity removals, right? And so this is like, okay, by like be. You can't do these like sandwich attacks and stuff by doing that. Or maybe we want to take all the trades and execute them in a single batch trade, right? So if you want to be very opinionated about block execution, block Ordering rules, that's another use case. So what are the kinds of things that want to make use of these? So I gave some examples of making your validators do more is if you need to do oracles or they need to run bridges and stuff like that. I think Dexs are obviously one of the ones that will want to have very opinionated ordering rules. You might have something like Stargaze, right? Maybe right now they don't do that, but in the future they might say, hey, we want our validators to run like IPS pinning services that like help, you know, pin these NFTs. Similar to like, you know, to be a validator on arweave, you have to provide a certain amount of storage capacity to the network. Similar to like, you know, Slushia providing certain amount of data capacity to the network. So it's like Stargaze might be like, hey, we're going to build a NFT like data storage platform, right. So it's like that's run by the browser, so I don't know long. And so those are like things where you need the custom, custom app chain now when you just need your own dedicated block space is just, you know, as soon as you're a big enough application that you, you're, you know, you don't want your business to be disrupted by another like popular product. That, that makes sense, right? Like, yeah. So I think that like there's all sorts of use cases why someone might want to have their own app train.
**C** (25:53):
Yeah. And you, you touched on the opinionation of, of your, of your validator set and, and it kind of, it's another benefit to running your own sequencer. And I don't want to like make an argument or too heavy of an argument to like run your own sequencer because I think we are trying to get to a place where we're running decentralized sequencers or even shared sequencers. But I'm curious as to like once, you know, as we kind of like see the benefits of running your own sequencer, you can have this opinionated way of ordering and sequencing transactions and then you can also internalize a lot of the revenue that you, that you generate from, from the ordering of transactions. Do you think we get to a point where the, the shared sequencers have some sort of like viable use case or are they just not customizable enough to allow chains to order the transactions in the opinionated way that makes the most sense for those chains?
**B** (26:51):
Yeah, so I've like, I think shared sequencers can eventually like add more rules Logic to the sequencing system that like allows you to do that eventually. But like it just gets more and more complicated to the point that it stops becoming. It becomes a blockchain in and of itself. Like honestly, suave kind of got to that point. Like flashbots, they're like, okay, we're adding more and more rules. So I think you're like, oh, okay, we're just putting an EVM inside the SGX at this point. We're making our own kind of blockchain at this point. And so yeah, you can do that. But then it's like, okay, now you're now okay. What is there extraordinary rules within that? So I don't think, I think shared secret source can provide some more rules as they add more logic to it. But I've thought through a lot of cases in osmosis where we don't want that to happen. Like I said, I mostly think in terms of osmosis. What do we need for our product? But with this functionality called threshold decryption, which is like, oh, you want to make it. So the validator side decrypts transactions in a certain situation, but you want to make it. If an Oracle update is delayed, what you want to do is tell your validators, don't decrypt these transactions until you get an Oracle update. Then process the Oracle update first, then decrypt the transactions. It's like encode. This set of rules in a shared sequencer will be very complicated relative to trying to just do it on, you know, your own chain. And like what's the point of using a shared sequencer? I actually don't understand why you wouldn't just have your own decentralized sequencer.
**C** (28:42):
Yeah, I mean, I think the argument is around cross domain mev, cross cross chain arbitrage, types of interoperability. I mean that's the argument.
**A** (28:55):
Atomic execution amongst different roll ups in the same block on the L1.
**B** (28:59):
Well, if you want atomic execution now, you don't even have to start sequencing. You need like atomic rollback and stuff, right? And that's like you're basically building a single blockchain. You're building a multi threaded blockchain. At that point you're not building distinct roll ups anymore. If the execution of your. If another chain needs to roll back a block a transaction and you need to wait on that before determining finality, you're no longer a unique app, you're a parallelized blockchain which, which is fine.
**A** (29:30):
But it's atomic use case. Atomic inclusion could be a better term.
**B** (29:35):
Atomic inclusion is fine, but I'm not sure what. Or there's a. I'm not sure how much I'd get to. Right. Like atomic inclusion. Because if a blockchain has, like my blockchain has rules of. Just because a transaction is included doesn't mean I'm going to process it the way you think. Right? Because like I said, maybe Osmosis is reordering the, the, the execution, like what order things get executed in. So inclusion gives you some guarantees, but it doesn't give you like guarantees of how it's going to execute.
**A** (30:07):
But I think this is important because it's the fragmentation problem. This is like the reason why the Sloth's Mint did so well was because you were just able to go from Celestia to Stargaze and the fragmentation problem was solved. And I think this is one of the big hills of the modular thesis and that Cosmos, basically the Cosmos Hub has been dealing with and all the app chains for the years and now we're seeing all these Ethereum app chains are going to start to basically deal with this fragmentation problem. I want to touch on Osmos alloyed assets, but before we get there, maybe talk us through how you're approaching this fragmentation problem from the lens of Cosmos as well as broader space and kind of why you think the alloyed assets is the solution or a solution.
**B** (30:57):
Yeah, I mean, just really quickly to mention on the sloth thing, like, you know, these chains didn't need a shared sequence or a shared validator set. It was operating on Celestia, Osmosis and Stargaze. All those have independent validators. I think this like cross chain composability is mostly a higher, you know, the whole world runs on asynchronous programming, right. All of web development is done asynchronously. We don't need web servers running on the same, we don't need different websites running on the same server in order to get like the composability we have on the web, right? We have APIs and all this sort of stuff and the same thing can be for blockchains, right. You don't need the entire world, all applications running on the same blockchain to get like the level of composability that you look for. That's kind of what like, you know, agoic orchestration is trying to like do. Make the UX of it a lot better. So like. Yeah, so you mentioned like fragmentation, liquidity, fragmentation. So, you know, today on Osmosis there's, you know, when we like Bridge, TIA or Atom, right? We use ibc, which is this like extremely native bridge. And we like kind of, it's directly done between Celestia and Osmosis. And so when we label that, we just call that TIA on Osmosis because it's like the natively bridged ibc. But when are we going to get assets from Ethereum? Osmosis doesn't have a direct IBC connection to Ethereum yet today. And so there's different bridges that we can use. Basically two years ago in Osmosis, we had this great bridge bake off where it was this. Four main competitors at the time, it was Axelar, Wormhole, Nomad and Gravity. And they were all kind of competing to become the canonical Ethereum bridge for osmosis. And governance basically decided to vote on Axelar as the canonical bridge for ETH on osmosis. So that means when you bridge ETH from Ethereum via actual at osmosis on our site, it just shows up as E. Right. But there are other bridges out there. We have Wormhole as well. And so what we had to do is we came up with a standard. When you bridge ETH from wormhole to osmosis, it shows us ETH wh. But if you bridge. And these are non fungible assets, right? And so because the UX was focused around Axelar, liquidity incentives are focused around Axelar. That's what got most of the traction and liquidity. But now we're in this situation where it's like, okay, what about bridging ETH from Polygon to Osmosis? Right? That's also not fungible because if you want to get the right way to get ETH from Polygon to Osmosis, you have to bridge it from Polygon back to Mainnet, Ethereum and then to Osmosis. What we did was if you try to bridge Polygon Ethos, we've labeled it as polygon.e. axle. So it's like the source chain, the asset, the bridge. Now the problem is we have like all these different types of Ethan osmosis. It's like, it's kind of like really messy and they're not, none of them are fungible with each other. We could go ahead and like build up stable swaps for all of these, but it's like, okay, where's that liquidity just going to magically come from? This is why we came up with this idea called Alloy to Assets, where we say, hey, the canonical version of ETH on osmosis will actually be a basket of underlying eths. So this is actually what someone like Coinbase does in the back end, right? You go on Coinbase and with eth you actually have the ability to deposit withdraw ETH from a number of different L2s, right? Because what's happening is Coinbase on the back end is they're just like maintaining inventory of ETH on different places and allowing you to deposit withdraw from all of them. So that's what will happen on osmosis, let's say. For the sake of example, let's just say we have two assets in our alloy. It's Eth by Axelar from Main and Ethereum and eth via Axelar from Polygon. Right? And so let's say what governance has the ability and the canonical ETH is the LP share of this basket. Right? It's. But what governance also does is it says it allows us to put caps on each of the asset based off of their risk level. So what we could say is like, hey, eth from Polygon, it can make up a maximum of 5% of this basket because that version is taking on Polygon bridge risk right on top of Axelar bridge risk. Well, the direct one is removing one of the bridge risks. But that 5% capacity that it has allows us that if someone wants to deposit ETH from Polygon, they can just bring it in and add it to the basket, as long as it's not at the cap yet. And it will just mint the alloy eth. So we have these max caps on each asset and we have rate limits where we say, hey, the composition of this asset can't increase by more than this percent in a 24 hour period. So let's say it's 50%, right? So Polygon ETH makes up 2% of the basket right now it can't go up to more than 3% in a 24 hour period. This is to deal with like, you know, bridge hacks or something like that where it can cap the losses and not cause this like run on the banks. It lets everyone at least you know, redeem on like, sense on the like get the same haircut, haircut discount rather than causing this like bank run out of the system. So yeah, so that's kind of what. So, so the idea is that like, okay, we'll for eth, we'll support like, you know, the Axelar bridge from like Ethereum. But maybe we'll also, if governance wants to say, hey, Wormhole has proven itself over the years at this point, so maybe it's okay to like add wormhole as another bridging option. What it also does is it gives competition to Axela. That drives bridging fees down for the user. Right. So now people can choose which one, which bridge is cheaper. Do we want to use Axelar or Wormhole? We still get the same asset on the osmosis side. Right. But now we can also expand, like, deposit, withdraw functionality from all of these different Ethereum L2s.
**C** (37:42):
Very cool. And I think that the name makes a ton of sense because even if you go into the chemistry of an alloy, it combines two elements from the periodic table. And once you combine two of those metals or two of those elements, you have an alloy. And so that's, that's what's happening with eth in this case. You're taking polygon bridged eth, you're taking wormhole bridge eth, you're taking eth from here, eth from there, and you're composing all of those together into a single alloy.
**B** (38:12):
And, you know, a fun one is on osmosis, like I said, because it's canonical, it's just called eth. When you. These assets will also be ibcable to other chains. And so when they're on other chains, they'll be labeled as a L L eth, Alloid Eth. But it's like also like all eth, you know.
**A** (38:30):
Yeah. So maybe a silly question, but what happens if, like the majority of the eths go to zero? Like the microlo's eth goes to zero, wormhole eth goes to zero and gravity bridge eth goes zero.
**C** (38:43):
Or if there's a bridge, if there's a bridge hack, and yeah, like one of those goes to zero or more.
**B** (38:48):
Yeah, let's say one of them goes to zero. What happens is the first rate limit will get hit, probably because people will rush to get out. So some people will get out, whoever can. But as soon as the first rate limit is hit, governance will be able to say, hey, it puts a new Mac. It says it can declare some of these assets as distressed assets, basically, or corrupted assets. And what that means is the percent of that asset cannot increase at all in the composition. So now that means two options happen. One, everyone can redeem assets from the alloy, but they have to take out their proportional share of the corrupted asset. So that means now, okay, everyone took the same haircut, right? Now let's say one of the bridge assets that made up 5% of the alloy got destroyed. Right? Now, okay, yes, this whole thing goes to like, you know, everyone has to take the same 5% haircut on their other on their deposit. The other option is it allows an insurance provider like, you know, maybe it's the bridge team themselves or osmosis governance with an insurance fund where we can say, hey, we're going to like swap out all of the corrupted asset for like an uncorrupted asset. So that way we can like continue, you know, we can, we can restore the peg of Alloy Eth.
**C** (40:11):
What are the knock on effects of proliferating osmosis alloyed assets across the cosmos?
**B** (40:19):
Honestly, to be honest, we haven't thought that far. We've mostly been focused on it from an osmosis perspective. But it is I guess makes osmosis a little bit more central I guess in the asset issuance on a lot of chains. And I hope it just improves the ux, right where it's like. I think actually one of the problems is that by choosing a bridge provider it's weird because right now if you cannot canonicalize one bridge provider, it does two things. One, it gives them sort of a monopoly that they can charge pretty egregious fees sometimes, which is not okay. And two, it also just like you lock off partnerships, right? Like other bridges have different communities that they have access to. They have access to like capital and liquidity and all this kind of stuff. And you're basically, basically like saying like oh sorry, we're only working with Axelar, right? Which is like, you know, that's why, why limit yourself to only working with one partner when you could work with multiple and like bring on. And so I think that was actually so really I think what it'll do is that this will actually bring in more liquidity into, into Cosmos as a whole.
**A** (41:26):
But like, isn't Noble doing this for usdc? Why don't we just let them do it for ETH and all the other assets?
**B** (41:33):
So USDC is a special magical case because it is a centrally issue. USDC is a hub and spoke system like asset. But the hub chain is like Circle, like C Corp, right? And Circle C Corp, like, like mints and Burns on all of these chains. And what was happening was before Circle brought CCTP which allows a before to go between chains. Like the native issuance. You needed to like, you know, work have a Circle account. CCTP made it so a normal user can like burn USDC on Ethereum and mint their USDC on Noble. There's no such thing as CCTP for Eth, right? Like Noble can't just be like, oh, I'm going to mint Eth out of Thin air. And like there would just be a bridge. Right. What makes Noble special in circles in USDC's case is that, you know, they did this like business arrangement with Circle. Right? But that's not. You can't do that with eth. And so actually the first alloy we're doing is with Tether because even though Tether has issued a native Cosmos USDT via Kaaba, they don't have a CCTP right now. And so there's basically like no easy way for a user to go between Tether on Ethereum and Kava Tether. Right. The only way to do the way we did it was two ways. One, we did OTCs or two, you go to like a centralized exchange like Bitfinex. Right? You go to Bitfinex and you say like, hey, we're depositing e Tether and we want to withdraw Kava Tether, but what is Bitfinex doing in the back end is the same thing as what we would be doing with alloyed assets. Right. And so it's like, why don't we just build that into the Dex itself? So yeah. So with now with USDT on. So because the Kava USDT is like using ibc, it probably has less bridge risk. It will probably make up like probably 90% of the alloy, but 10% of the alloy will still be USDT from Ethereum. And we're also working with another bridge to bring USDT from Tron. And so you'll be able to deposit, withdraw from both of these sources.
**A** (43:55):
When it comes to LP in a Dex pool or just using these allied assets, one would assume there's like more risk by using the alloyed assets compared to just using one of the native assets. Is there going to be like, like how you compensate people for taking the risk of using allied assets?
**B** (44:11):
Yeah. So I mean, liquidity incentives will be focused around the allied assets. So that way they're by. With these liquidity incentives are kind of being compensated for like taking on some of that risk. And I would actually say like the risk is a little bit. It's not as clear cut as saying it's higher. What's different here is let's say you have five bridges in this allied asset. Yes. You're taking on the risk of five bridges. Like there's more chance of a corruption, but the result of a corruption is lower is safer. Right. Like it's a diversification thing. Right. Where you're, you're actually like if one bridge gets corrupted, you lose 20% rather than if, oh, the bridge you chose to hold gets corrupted, you lose 100%.
**A** (44:59):
Yeah. But it's like who's the, who's the bank, who's the insurance? It's like on an exchange, it's like you can always count on Binance, you can always count on, you know, whomever if something happens. Right. Whereas it's like, you know, you're diversifying. Sure. But like as we learned from the past, you can't just diversify, you know, three good and two bad and call it a, A plus cbo. And then the entire mortg crisis happens. Like we have to kind of think differently about the risk and at least give some incentives for taking on that risk.
**B** (45:29):
Yeah. So, I mean, I do think that Osmosis should be working on building up an insurance fund. Right. Like a, a portion of all trading revenue does go into the community pool and that should be that. Like we're actually going to make a proposal about this that should be set aside as a community, like as an insurance fund to help same. I mean, that's how Binance plugs these gaps as well. Right. Like they, they have insurance funds built in. So it would be like a similar situation here.
**A** (45:54):
Gotcha. And the other, the other kind of matter at hand is like what Rob was getting out. Like, how do we.
**B** (46:01):
One really quick. I'll also add that like the goal here is, like I said, is not to be like, oh, we're going to add like 10 new bridges, like you know, every random. I think the biggest value here is like to support bridging from more source chains. So even if we just, let's say we don't even add Wormhole, we don't add any new bridges at all to Osmosis. Right. At least we can use Axelar to bridge from Arbitrum and Optimism and Polygon and Base. And yes, what we are doing is we're taking on the native bridge risk of those chains. Right. Like Polygon has a native bridge. Arbitrum and Base and Optimism have a native bridge. So yes, we are taking a little bit more bridge risk on. But that's where, like I said, I don't think it's like, oh, these are going to make up 50% of the. Each of these are going to be 20% alloys. Each of these is going to be like just capital. Whatever we need in order to get the deposit with withdraw flow capacity that we want.
**A** (47:04):
Yep.
**B** (47:05):
Even Coinbase. I'm sure every time they like people, let's say a lot of people depositing to Coinbase on Arbitrum, they are rebalancing it. Right. They're probably moving it, withdrawing it and moving it to mainnet, to holding it in whichever, a safer way. So like for example, we're building our own. The alloy that I'm personally the most excited for is the BTC alloy because I think I just foresee a lot of liquidity fragmentation coming in that ecosystem. And so we worked with Bitgo to issue native WBTC on osmosis. And so I think that will actually be, at least at the launch, will probably be one of the biggest components of the alloy on, of the BTC alloy on osmosis. But then from there we can add like some capacity for like oh, SBTC from Stacks and TBTC from Meso and like just enough to get these like deposit withdrawal flows enabled. While majority of it is still like the WBTC that most people trust at this point.
**A** (48:05):
Yep, yeah, that makes sense. The only, the only other consideration, while I think that the aiming to solve the liquidity fragmentation is certainly the goal of the current state of crypto as just Cosmos in general, you kind of get the knock on effects that Rob was saying, which I understand that you guys may not be too focused on that right now, but once you have alloyed assets and users are trying to do different things with them on different chains, there may be a bit of a rough around the edges user experience for users and also the teams may have to kind of implement different, I don't know, ways of managing these assets. So yeah, there's just the other kind of concerns that we're seeing.
**B** (48:53):
Yeah, I mean traditionally what's happened in Cosmos is most of Cosmos takes the lead of osmosis on this kind of stuff. Like osmosis voted to make Axelar the canonical bridge provider and then like actually sort of became the canonical bitch fighter for all of Cosmos. So I think the ecosystem tends to follow the lead of osmosis here. One thing also worth noting is that alloying is also implemented as a swap on osmosis. So even though you're depositing into a basket and minting the token, it's implemented as the UX of going from ETH axle to alloy. ETH is just like it's the same as just doing a normal swap transaction. So it'll work automatically in like Skip API and Squid and all of these integrations will just be pretty like seamless.
**C** (49:45):
How, how does that change with ETH to Cosmos ibc? Do we still need alloyed assets in that case?
**B** (49:59):
I think that a lot of these Cosmos IBC Ethereum IBC implementations like like, you know, union and composable and stuff. They still have code risk, right? Like, when we think about bridges, there's two types of risks. Yes. There's like, you know, economic security and decentralization risk. Like, some of the biggest bridge hacks have been from that. Right? The Ronin hack, the Harmony hack. These are like, like, centralization issues, right? You had, like eight nodes turned out. Four of them are owned by one person. It's like, yeah, it's like, okay, that's one. But the other is like, code risk. The wormhole hack was not a centralization, but that was a code bug. Right? And it's like, I think these new IBC implementations from these, like, honestly, like, unproven dev teams, right? Like, very passionate, but, like, you know, they had never been built up. This is going to be one of their first production products. And so I. I think we want to still. We're not gonna be like, oh, okay, you know, ibc, you know, composable has IBC tomorrow. Let's shift everything to using that, because that could still have code bugs. So I think what will happen is as people become more comfortable in these things, they'll slowly increase the caps allowed for that version of the asset in the alloy.
**A** (51:18):
Oh, okay.
**B** (51:18):
So.
**C** (51:19):
So that's how. That's how it could look in practice, right? Is like gradually the. There is native ETH via IBC in the alloy, and then gradually you kind of like increase the cap until perhaps, like, we treat it as secure, and then we entirely have native eth on. On osmosis.
**B** (51:38):
Yep.
**A** (51:39):
Don't worry, Sunny's not going full Justin sun on us. We still got him, boys. He's safe and sound. You brought up a good point. Kind of to end us off here. Last five, ten minutes or something. Hyper bitcoinization. Bitcoin on Cosmos Babylon Nomic just launched, I think. Yeah. So maybe just. Can you give us the TLDR as to what's been going on there? Where I think our audience probably is a little bit left out there and maybe some of the developments that you're seeing. Osmosis, et cetera.
**B** (52:14):
Yeah, Hyper bitcoinization. I started working on Cosmos to build the app layer for Bitcoin. That was, like, my goal. I was interning at Consensus at the time, and I was really interested in all this Ethereum stuff. Like, wow, you could do all this cool stuff on blockchains, but it's like, what is this eth shit? Coin? I thought we're building for Bitcoin.
**C** (52:41):
What time frame Was this?
**B** (52:42):
Like, 2017.
**C** (52:44):
2017.
**B** (52:45):
Okay, so Eth was definitely like much earlier in its life cycle than it is today, right? Like at the time it was like, you know, ethos around for like two years or something. So it was like. And so I was like, I want to build for Bitcoin. And so I saw Cosmos, Learner Cosmos and I'm like, oh, this is it, right? This is how we build the app layer for Bitcoin. You know, a network of app chains. Bitcoin is the original app chain. It's meant for, you know, just for money and issuance and payments. And it's like the bitcoin asset just. We just need to have the BTC asset be securely bridged off and it'll be the money of every ecosystem of all blockchains. And Cosmos is the way we build that. Blockstream was the one who came up with the concept of sidechains and never really built anything with them. Then it took us a while to build the cosmos stack to where it's possible. And then bitcoin just. There was no demand for it, right? There was no one using it, nothing was happening. And so I was always pretty involved with the community, but it was just like, I don't know, there was a few, we called us the bitcoin progressives who were pushing for this vision, but it was just a very small contingent. And honestly what kept happening was our contingency kept shrinking and shrinking over time as people got demotivated of ever seeing any change. And it would. That contingency was shrinking over time. And then ordinals came and just changed everything, like ordinals then, plus like UD's movement that he created. But the culture, there's a culture shift in bitcoin and I think there's vitality there now, which is really exciting. And with all the BRC twenties and runes and honestly most of the stuff is just like NFT sees NFTs and meme coins. But now you're actually starting to see this have like ripple effects where now it's like, now you actually have people developing like Bitcoin L2s and stuff and like, you know, bit VM based stuff. And I actually think that, you know, for the first time there's no lead maintainer of bitcoin at the moment. Like, and I think like, you know, this is, I think, I think that like we're going to see like the core development of bitcoin itself, like chains. And that does not mean we're going to have smart contracts. And some of my bitcoin I think that's like we shouldn't have that, right? Bitcoin main change remain simple, but at least enabling the few technologies needed to enable secure bridging, right? Like OPCAT, let's do OPCTV like 119. These are the kinds of things that we need in order to enable Bitcoin to be securely bridged and then from there we can. And so basically I think Cosmos is super well positioned for this right now because a little bit of tinfoil hat. I think one of the reasons that good bitcoin bridges weren't really promoted within Ethereum is I don't think the Ethereum mafia wants bitcoin as a competitor to eth as money in the ecosystem. But I think Cosmos actually doesn't have that issue, right? I think like Cosmos we don't have a base money and so bitcoin will be that base money of our ecosystem.
**A** (56:03):
I think that fits well with the outlaw vibe of Cosmos. Cosmos is very bad. Kids like outlaw, like censorship resistance like bitcoin, libertarian vibes.
**B** (56:14):
I mean so Babylon, Babylon was actually my idea I gave to David to do hey, bitcoin restaking. This is the thing. And so I think that's going to bring a lot of attention. Mezzo is Cosmos space. You have Thor chain, the biggest Bitcoin dex is Cosmos based. I think Cosmos is just very well positioned. I think we should rename CosmosM to BitWASM. It's a slight branding thing, but I think that's an important thing to do. But yeah, I think Cosmos is well positioned. And for us osmosis, we're trying to be the place where a lot of these L2s where we're helping them with their go to market and stuff. And osmosis will basically be like the. If you want to buy bitcoin like buy into any of these Bitcoin L2s. Like this is the place where you do it. And with alloyed btc this is the way how you move between the ecosystems. If you want, if you have SBTC on stacks and you want like BTC on botanics, like you do that by going using the osmosis alloys as a sort of stable swap between them.
**C** (57:23):
Just taking a quick step back, it seems like bitcoin kind of skipped a cycle. Bitcoin in 2017 you were working at Tendermint. We had bitcoin cash hard fork like the middle of that year. And then we have. Nothing really happened in Bitcoin in 2020 and 2021. All of the action was defi on Ethereum. Now fast forward to kind of like this cycle, if you will. We have all of this action, all this vitality around bitcoin and scalability and what you can do with it. When you were in at Tendermint, you were part of this like, small contingent. Why did you choose to build, better build bitcoin apps on Cosmos? Why weren't you part of that bitcoin cash regime?
**B** (58:09):
I mean, I was a big blocker during that fork time, but like, I don't know, I gave in and I'm like, oh, we lost the fight and like, it's not worth forking off over. Like, bitcoin is the brand, it's the flag. It's like, you know, bitcoin cash, it's just, just, it's just another altcoin now, right? It's like, yeah, if anything, like, if you wanted to fork off of bitcoin, you might as well just use Ethereum at that point, right? It's like the, you know, Ethereum actually has the better technology. It just doesn't have the bitcoin brand and stuff, right? And the history and all that stuff. So it's like, no, I mean, I used to be involved. One thing is I used to be involved in every community I was involved with. I used to go to bitcoin cash conference. I've spoken at bitcoin cash conferences, I've spoken at Ethereum classic conferences, I've spoken dogecoin conferences. I go to everything. But eventually the bitcoin. You know, it's funny, the problem with bitcoin cash is it once you have a little bit too. I like to look at, I use history of religion as a good analysis into how blockchain things work. And it's like bitcoin is like Catholicism. It's extremely conservative and rigid, but as soon as you have a Protestant movement, as soon as forking off is okay, you have endless fork offs, right? So bitcoin cash then forked into like, you had bitcoin SV that spun out of it, then you had bitcoin ABC that forked out of it. That community just kept splintering and splintering to the point that it became unrecognizable. And I also just don't think that some of the folks that were spearheading the bitcoin cash community, I think they were just not very political about it. They just did a bad job of the way Jihan treated. They were like sore losers, basically. I think that was the problem.
**C** (1:00:10):
He has one of the best quotes of all time. But I won't say it on camera.
**A** (1:00:18):
I will. Your mom if you want to, or something like that. Yeah, I think that's a good stopping point. Boys. Thank you for coming on. Sunny, it's been a pleasure.