Robert Zaremba - Cross Chain DeFi: opportunities and challenges

Transcript

thank you hey everyone I hope you have a good time so um few words about myself uh I'm leading uh I'm a tech lead at Umi uh or ux chain we rebranded we're one of the cosmos layer one chains um the biggest in fact money market native money market in Cosmos uh yeah my background is U for I think already six years in Cosmos before in ethereum uh I'm also part of the your contract standards uh committee work group I used to do lecturing at the University of Geneva but I don't have any time uh uh time anymore um that's my website if you want to like follow

me so I want to talk about like this uh discrepency between different chains we talk about layer twos we talk about chain abstraction account abstraction about layer ones layer twos layer threes and so on um I want to start with this slide uh I think everyone is familiar with uh because like on every chain we want to launch another mcoin another nft sell it uh and this like replicates like okay maybe you can try on eum then maybe you can try on a base and on Solana and an avalan or whatever is there and and see your success so for nfts for like this kind

of images yes like maybe the strategy will work like you don't use any um lose any efficiency uh right but there are many other m coins and then you can ask yourself like you know how much sense does it make that you launch yet another mcoin and then you need to bridge it like you know how do you form this ecosystem um so this causes many problems right like uh now if you think from the financial perspective for mcoins it works you can just lunch it another one right but you know nobody is using mecoin for serious stuff it's it's good to make money but um nobody will trade it like uh in a uh like traditional sense uh so um now like abstracting from

that way yes like you know how this works like okay you you you take the same smart contract yes have a template and then you redeploy it and maybe you just change the catalog of of the images if this represents nft collection if it this is like aex then probably you Fork Unis swap or any other solidity contract and then redeploy it Rebrand it you add maybe one more feature change maybe the governance parameter and voila uh maybe uh maybe you have another successful protocol um and then you know you do it for Layer Two you do it for another layer one um soon there will be layer freeze um all of it then if you if you apply it for the for the DS it obviously

will fragment the liquidity right because your Unis swap pull on on ethereum and then you need to bridge some tokens and and then uh I mean you get the point right it fragments the liquidity and then it causes the liquidity wors because like how do you incentivize people to um to actually move the liquidity so what's very often happening is that okay there are this uh incentives yeld strategies and then you know um all the Deion will basically milk your incentives the the tokens you have uh and as soon as the incentives are finished uh most likely your protocol is that but who much is like you know you made I mean um probably few thousand if you are good a couple hundred thousand

if you are very good maybe some Millions right so you made that job but overall yes like after a few months your protocol will be dead um so it's just like you know phenomena in in uh in the def of today uh next thing is that you know if the produ like let's say a or or Unis fabas which is lunched on few layer tools or and ethereum and uh there is a brand right so they uh they keep the value uh but they lose on uh Financial efficiency right so um if you take for example a AA again um it's very important that the the liquidity on the on the man Market perspective is there

right because the money markets uh they are supposed to provide to to be that you know that the very stable big source of liquidity to help other other protocols and the other protocols should be built on top of of that so uh there is a whole strategy right uh when and why uh the mark money market will have to be replicated or not and how it works so again uh to make it efficient then we have a a pro from a DEX perspective especially yes we have a problem with with I mean it only works when there is arbitrage right if you are on let's say um arbitrum and you need to make some some some trade um uh either yes you you break it down into fut trates until the

the Arbitrage Bots will you know will um equalize the price uh which is not perfect from the user uh user perspective or you lose because the slipage will be uh will be bigger so the the world of today is really like depending on on the on the branding on the people on the trust in the protocol sry teams right and the Arbitrage b stand they will be efficient consequences so reduced liquidity of course right your protocol is less efficient increased complexity because now you either wait or you again you need to apply those Arbitrage Bots um inefficiency in U capital and uh resources allocation and uneven access

to the information right so the the Arbitrage B is a huge money to make but um it's really for sop sophisticated users only so what I want to talk here is that we we should as a community as um as Builders as innovators thinking more on the bridge powered apps so uh rather than like fragmenting and copying the the same solution start thinking especially now when you know ethereum is moving to layer twos and there is not one layer two but many many layer TOS they have to be connected they need to work they need to be uh know uh interoperable so it's not only about the composability right but it's the connectivity this is still missing if

you think from the perspective like again in ethereum that you know ethereum is the settlement layer it doesn't bring what we really need the interoperability between the protocols because interoperability is not about token transfer it's really about uh it's remote procedural calls so um for that yes so okay token transfer is the basic one but okay we need to have the the message passing we to have a gas abstraction right because let's say if if the money market will be let's say on on arbitrum right you don't want to again replicate it to the optimism um uh you just want to like quickly and efficiently talk to one and another without having a user to Ray on a gas and underlying um

complexities uh interchain tokens and another thing uh usdc how many usdc you have you have Axel usdc a usdc um a usdc on on on polygon as AEL usdc on um on arbitrum on uh on osmosis and so on so far right so um then you have those olders 19 yes this is at least it helps um unification around it I mean we those tokens have to be different because they they carry at different risk so but the concept of interchain tokens is again is how we can leverage bridge to unify it yes somehow um and finally interchain authentication uh that comes with interchain accounts and

chain abstraction I did a lot of work for example with AEL and a few other Bridges and there are usually problems because when you have multihop Bridge you lose the proof who actually is the original sender because the the bridge will all usually reprodu the sender from the last leg so you need to think also about you know like how do we do the authentication when when there are multiple legs on your Bridge uh chain abstraction um uh internet computer is doing like a great job with that started it then near uh that's like they want to put it as a team of the of this year I think I have a slide about it vo so it's more about that okay you have your own

wallet and from that wallet you control all the other accounts on other chains and so the concept is that use MPC uh I'm not sure if uh internet computer was like going deep into that but they have a lot of documentation a nice video check it out uh but very roughly that um the uh the subnet or here in case of the of near yes um a chart yes will um derive keys from your let account here to uh construct transaction to sign a transaction for other chains um that's uh from uh from IC uh IC but you know

the idea is like you know we use threshold uh cryptography and and NPC to basically it out so going back to Bridges right like this is from um from Axel [Music] um the applications will be on the chains multiple chains um then you have many gateways many bridges right and then uh how we how we can organize it yes so all the bridges I I think that they they need to have the uh application layer like in the tcpip right so it's not only a communic but application layer Axel allows for it war will allow for it IBC will allow for it and I will talk it in

a second polygon polygon is doing like something similar but in a different way so they introduce that a layer or interoperability layer right where all the messages again messages not the tokens right can can can can transfer in a trust trustless way here in this case yes um they require ZK proof so that AG layer in this slide is interoperability layer will verify it so you have a a zero knowledge proof that the execution was correct um which is usually better than just consensus verification um uh why it's important again because you can focus on developing on a single chain and

leverage a fast interoperability layer to do it and the other it works the uh disadvantage in case of poly or cdk in my opinion is that no it's it only works with evm if you have some custom evm with custom op codes because you do more of the Privacy stuff um then no sorry it won't work for you um the here's another view how how you can look at it so we have some public chains cdk chains private chains again all of it then the proofs can be aggregated merch into that you know concept of interoperability ler again I don't want to talk about about polygon but about the concept the concept Works um a

cosmos chain ZK IBC IBC is perfect for developing application on top of the bridge so um who here is familiar with IBC okay few people um so IBC stands for an inter blockchain communication protocol so it really start like from bottom bottom up approach here so you have the um the networking layer you have the channels connections ch panels same as you you know you you you connect um your web applications web services right you uh you have a server the servers the servers and the client establish a connection once the connection is established you can have a channel there multiple channels and

multiplexing and then uh in terms of IBC these channels are important because each channel is binded to a application to a specific protocol they can then Multiplex the packets between um between the same clients uh crucial is that it's open it's open-ended so you can have any application on top of that yeah so you don't need to to to change probably should be better if I would put her another slide like you know um uh like in O stack um that you don't need to change the channels the concept of like know how the trust is established here it doesn't require you for that yeah think about like uh from um uh from the cdk polygon perspective if you would if if if there is a limitation somewhere

there uh like the whole infrastructure doesn't work yes so we will yet to see yes how how we will be if we will be able to like enable more uh aggregation again if you change the proof system there might be some limitations IBC maybe is not the most efficient one to today it's really focusing on the light CLI verification but it's open-ended you can do like any any type of the verification is peer-to-peer right so you as a user again from a chain is you can have kind of like a governments or um or [Music] or uh risk assessment uh to see like you know with which protocols with which chains you want to inter oper interoperate and that's crucial this is really the

interoperability peer-to-peer you reduce the risk and and you can do it uh so there are many now um projects trying to leverage IBC to um to use the same concept to use the same um API about the packets channels ports and uh uh to apply for other stuff so for example Union recently they they they they launched they they they connecting all Cosmos chains with all the evm chains um all the API is ABC compatible so um finally is going like open that you don't have one specific implementation between the cosmos chains but you already have implementation for other

chains as well um notably there using zero knowledge uh validation for evm chains um few more slides here about you know whenever doing that it's super important to think about the security there's always risk um so uh the most of the bridges like Wormhole before now they have the new version right of the Wormhole connect um they were rather focused on the multi signature yeah we have already seen like many um vulnerabilities with that MPC is kind of interesting that's what I mentioned if you remember about the IC and near but it's still yes that you don't have full verification validators could collude to

forge a signature um for you uh and there won't be slashed yes for to to to to basically then relay to another chain so it's not the best one but it's okay consensus verification and state verification this is what IBC and what I mentioned before about Union uh provides the zero knowledge validation full validation uh is the is the last one so you you you verify the state you verify that operation is correct based on the latest state of the chain so you fully reduce the the the risk here um so what we should really think and and and develop whenever you're making decision building the applications with Axel with um with Wormhole um with Union or other bridges

that will come or IBC in general as a concept yes we need to really be here don't use that um and finally I want to close here with few thoughts about like a defi and what we are doing um so defi V1 um it really fragments the liquidity um that's you know you have pull one pull two second pull those pools they don't talk to another one to another V2 is you have like a protocol on liquidity it helps because then you have like a protocol who can manage the different po who can do some assessment yeah human is Les less involved into that uh but none of them really solv the

uh liquidity fragmentation problem and in my opinion like maybe it should be V3 or maybe V4 I don't know how people would like to call it but for a few years we already talk about UniFi liquidity and in my opinion this is like building the apps with bridge in mind from the first perspective rather than copy pasting uh a few solutions that already let's say leverage that um inter fluid taking super fluid taking super fluid is for example from osmosis you can deposit into the pool some tokens those tokens will be locked bonded uh and automatically let's say staked

into uh other I mean if those are native tokens so super fluid is for Native tokens so um tokens that are like layer one tokens right that you can stake for your proof of stake uh rewards and then basically you you boost to the um your API uh met is also like experiment we are doing uh that's kind of like a bridge token that you can unify different um and Accel usdc from you know chain a chain B chain C and and try to find a protocol with that we released it a month ago we still like need to think how to Market it uh but it's more like a an idea and uh okay I cannot go back I I went back

um uh yeah so I want to like here put as an example this me token because um check it up I I really want to promote like you know Visions or missions for anyone who wants to build with bridge in mind so uh with that yes I hope that this will be the right solution for for for a defi I mean everyone in defi wants to have um better speed better quality and lower costs right just shipping your protocol from one chain to another is okay it's like a short term there will be always better better chains but if you fragment it just copy paste uh it won't help you right you can always move

your protocol from other chain but you still we need to have the interoperability in mind so if you start with interoperability and then you find that okay actually uh for whatever reason Optimus doesn't work for me because uh they don't have to fraud Cru uh I want to go for sui or salana because it's more sexy now uh but if the whole stack the whole API with all the connections will still work because you with bridge in mind you will keep you will improve the speed you don't break the composability or interoperability you improve the quality maybe because the other chain will provide you better um better Solutions and you reduce the the cast um thank you that's all thank

you do we have like two minutes for one question does one one question be the thank you really great presentation one quick question what's your view of on intent networks such as a cross dbridge dln to abstract the need of even having a bridge and offsetting this one trade on one chain to another through solver networks I'm not sure I get a question you ask about the intent yeah like what's your personal view of intent networks like dln dbridge

across instead of even needing to have a bridge you offset this to solvers yeah good question so um intent could be used like kind of a resolver so uh I mean we have like know this chain abstraction is which I mentioned before that you know you abstract how you can control from one wallet another account uh I think that could be a good optimization for like a Bridges or that uh uh from a user perspective right it will our from the interoperability perspective yes it will it will solve it yes because we will have some optimizers there will be many bridges for sure so um from a developer perspective probably

if there is like a good um if there is a good product that will deliver Bridge base intent and resolve all the roting for me keeping the interoperability yes that's the future thank you so much thank you m