DappCon 2023: Why Bridge Aggregation Is A Necessity And How The Future Of Bridging - Philipp Zentner

Transcript

hi all right thanks for making it into this talk it's quite late already in the last day of the conference so appreciate for everyone who who came um we Qui about me I'm philli I'm the founder and CEO of Lei I do have a technical background even though I haven't written a line of code in quite a while now um but I've seen the web grown from the get-go kind of I remember 2002 when I was optimizing websites for Netscape um and we're writing browser switches um so uh and I can tell you 20 years later I'm excited as I was back then that um a lot will change real quick well the company um we are a multi-chain liquidity aggregation protocol um as such uh we have grown quite quickly because the

space is fragmenting and this is what this talk is about it's about fragmenting infrastructure fragmenting liquidity and um how the bridging landscape looks like right now and how it's going to change over the next years um if you want to learn more about this we do a decent job in providing knowledge about inability and and all of that on our blog and newsletter so check out blog. fi or substack um you will certainly learn a lot now if you look back um 2021 um we saw a huge outflow of tvl from ethereum to other chains those chains were less decentralized and therefore faster and cheaper right Solana polygon binance Avalon phantom um

this has been happening over the past two years they all got lots of traction some of them kind of lost the battle some of them were really successful like polygon um but I don't want to bore you with too much history if we look forward um this is what we are seeing happening right now we're seeing a bunch of rollups as a service coming out they're going to make Cosmos app chain future true within within ethereum within the evm ecosystem um and it's going to create vast amount of applications and all of them have to be indexed all of them going to have their own token they need to be connected so there's a huge use case for bridging in general right um at the same time you could say well

shared sequences like espresso going to solve that but there will probably more than so first of all they all need to opt in second of all with shared sequences rollups kind of give away control about meev so they are not ultimately incentivized to do so and third of all um there will be multiple Shar sequences so again then we're going to have fragmentation across that stable coins um we are all used to usdc usdt die um but you know by now everyone that sits on a huge treasury is incentivized to release their own stable coin to give you an example with Ava with the gho token right so go is essentially um essentially aa's way to say wait we don't need to um we

don't need to um collect money from LPS we can simply have our own stable coin bake it up and over collateralize it with our Treasury and the interest we would normally give out to LPS we can keep ourselves it's a great business model um and so is JP Morgan announcing their own stable coin so is PayPal and all of them are going to create not only their own stable coin they're going to create their own bridges for the sake of control and as companies like AXA layer zero warhole with the multi-chain token standards are not established yet um we're going to see a vast amount of new stable coins being established in certain environments certain ecosystems and um and real world asset niches I would say um and yeah as I just told you

about um we going to see more and more multi-chain token standards coming up and this creates another fragmentation there so um those uh those tokens will be minable on chain and they have like a decentralized accounting via the bridges that make that happening um but it is difficult then to um actually trade them with other pairs um so that the the the the fragmentation is going to really it's going to become really messy for the next two three if not even um years to come it's it's incredible and um I I think um account abstraction and all of that and chain abstraction how we want to call it um those are nice approaches but um also

for them to grow up we need a certain middleware that's actually what Lei is working on mainly let's go back to brid which is because for the most of you that are as a i who's technical here all right okay it's 50/50 actually that's nice um so for those who are developing applications um and for the users essentially if you're not into that particular space um you're just you're just going to deal with normal Bridges liquidity networks that connects Stargates hop across um and and and Squid and so on so touching arbitrary data messaging Bridges like layer zero and axel is something if you build native multi-chain applications um but

apart from that um yeah we can ignore these topics for now and just talk a little bit about bridges so if we have two decentralized networks two blockchains they can't talk to each other so they need a man in the middle someone that connects um or relays information from one chain to the other and there are multiple ways to do that um in Ideal World um we could simply read from one another then we would have light client based Bridges so essentially we would Implement a light client of chain B on chain a and the other way the other way around um but because that barely scales well Wormhole is trying to go in that direction it barely scales though um especially in a

future where we have thousands of chains so we need a system in the middle for which it is very easy to watch smart contracts on multiple chains but at that point we're introducing trust boundaries and um now we introducing that topic of trustless and Trust minimization in general we have these three type of bridges so natively verified bridges are the ones I just described we have light clients on different chains that barely scales then we have externally verified often NPC based multichain was one of those uh Bridges um here there are different mechanisms to make it more safe but often we have like a multi sick system or whatsoever in place and it's not considered to be very secure um ESP especially if there are 20 30 50 million

stored and there's a huge incentive um to hack the system and then we have locally verified we try to somehow get some atomicity between um those two actors um but it's difficult it's very limiting so we are losing out on lots of options um and most bridges have like connect and hop have actually moved away from this what we are talking about in general here is the interoperability trilemma so the idea is that all these bridging systems have to pick two out of three trade-offs um they can focus on so the first question is trustless where we are simply asking who is verifying the system and what would it cause to corrupt it right um here in the best case we have a system in the middle that

is as decentralized as the systems it connects that would be the perfect scenario but if we have a decentralized system then again we're talking about settlement times and whatsoever so bridging takes while we are trading off performance and lots of other things then we have generalizability where we're asking well what can I do with it so axela layer zero Wormhole and hyperlane they are maxing out on that right you can build anything on top of it that's the idea um so they're very generalizable um but often that comes with other trade-offs it's slower it's more expensive if you want to validate arbitrary amount of like arbitrary messages arbitrary data then you have to keep things very loose you needs lots of

additional checks so Gas cast are rising and then we have that topic of extensibility obviously it's much much um um so the question here is like how easy is it to expand to other ecosystems for example leaving evm and going to non evm um like Solana or going to ZK Bridges uh extensibility is also something what ex layer zero and so on are maximizing on um and uh it allows them to interconnect multiple different ecos systems um but it again comes at the coast of scalability and and and other topics so this is what most of these bridges are actually um focusing on um if you take a look at the bridging

ecosystem it's pretty crowded we are counting around 100 projects I've just listed a few of them here um and we haven't even touched based on topics like St con Bridges like ucc's native one uh Native Bridge Bridges so all the ecosystems like arbitrum optimism they kind of want to have control about how liquidity is bridged to their ecosystem and away from it they don't want to rely on third parties uh nft Bridges well with nfts it becomes even more difficult because we are creating secondary markets for something where we want to maybe wanted to get rid of secondary markets in the first place so from an economic perspective difficult also um what happens if like nft Bridges really need to need to work in a in a very

standardized way and no one is taking care of Standards right now it's really difficult we have some organizations like Kasa and so on that they're working on chain agnostic standards but it's it's no one has interest in actually adopting these standards right now because most companies simply want to want to be that winner especially those companies with huge treasuries like axela and layer zero they obviously want to establish their own standards and then be the standard right so um it's going to take a while we have CMP within the poodot ecosystem we have IBC um within the cosmos ecosystem and connect simply released uh um XX erc20 that's actually quite an open standard um I recommend looking into for

everyone who is uh thinking about releasing the token multichain essentially xc20 is a way to um without slippage without uh greater gas cost having to trans as it gives the ability without slippage to transfer tokens and at the same time um it allows to it allows liquidity networks of all kinds to opt into the network and be one of the Rel layers so it's kind of a bridge agnostic system super nice ofs um are the standards by layer zero and then also we see a new type of bridges coming up because message verification on ambs is very expensive we see more and more cross chain Z knowledge proof Solutions

coming up Herodotus and lrange and axium are three to name here um and they simply allow us to get a mathematical proof of a certain State on another chain without having to relay um lots of messages and um they're going to replace a huge part of what AMS are doing um over the next years as well so complex field there's a lot happening and all of that essentially creates that creates the case for aggregation for us for Lei um because for any project especially those coming from Top they have to understand and how do these Bridges differ which trade-offs are they taking what is the right trade-off for my personal use case and for each developer

this is an absolute pain right um also once you have employed a bridge you kind of want to have a fail safety solution you kind of want to have a backup solution if a bridge is out of liquidity you have a problem user flow is broken um if an amb like the zero axela goes down it can go down things can collude at layer zero there is a 51% attack chance with an axel especially as the token drops so there are dangers and you kind of want to have a f you kind of want to have fail safety and also you w't have optionality what I mean by that is uh that your needs might change um Bridge implementation are very deep down in your product um and you know how it is like you want to go to market kind of

fast um it's a super competitive market it's difficult to make the right choices and right abstract enough code that you can interchange products um over time as you changes need especially when you are building a multichain application from the get-go um you don't know how things are going to change um where your customers are going to come from which Partnerships you're going to form so you might need simply um the ability to have different risk profiles which you can determine programmatically um and uh yeah change systems as they needed and this is where we provide a Smart Auto routing this is where we provide maximum flexibility by whitelisting blacklisting certain uh bridging solu Solutions and um this is simply where Lei is well

positioned but what's next for bridging um we see certain Trends uh one Trend we see in general for decentralized exchanges is RFQ um so 1in Fusion is an example uh Unis swap X is an example Unis swap X well they jump the bandwagon of um intent driven systems but it's nothing else than RFQ right now all these RFQ systems eventually can evolve into an INT system but right now intense systems are far away um of from from becoming a reality and we'll probably going to see lots of intense systems focusing on different parts on selling nfds on um on buying and selling options perps

leverage whatsoever so we're going to see lots of intense systems take care of um these things and obviously we're going to see these systems becoming more cross chain native from the get-go um another system in the same range where we also set up a limit order is The Coincidence of once system where we have a much more peer-to-peer approach cow swap is one to mention here um what's also next for bridging is that these multi-chain token standards like this is just starting lots of you have heard of O of um ofs layer zero standards I guess Axel standards is not as established yet warhole will come up with one xc20 was just announced like this whole multi-chain token standard War is just

out and sounds bad but there might be at some point also a huge hack from which we all going to have to learn from because it's a new system so again this can become expensive last year we have lost 2.5 billion in Bridge hacks so let's see what's going to come in the future we haven't seen a bridge hack in a while though like the systems are maturing it's good that's good unfortunately with this whole stable coin War we also going to see more more centralized Bridges from those players who simply want to keep control uh either from a technical perspective or from a regulatory perspective Circle for example has their own centralized um Bridge um it has a disadvantage it they

take things very seriously and and and and slow they have a long finality time which is why circle is partnering up with a bunch of liquidity networks to speed up the minting burning withdrawals Outsourcing risk for a bit um but if you have a huge transaction let's say you want to transfer $10 million from one chain to the other and a certain liquidity pool cannot deal with it then this bridge can also talk to Circle API and actually burn and mint that amount of tokens which is uh nice um but also yeah given that so many large players like JP Morgan and payer going to onboard huge ecosystems behind them they're going to onboard their clients right over time we're going to see lots of bridges and lots of stable con

fragmentation Bridge fragmentation around this um and um th those are the things we need to get used to in the next two years 3 years next three years is all about who wins the multi-chain token standard um we're going to see a huge fragmentation of stable coins in different contexts and for different reasons we have economic reasons um um regulatory reasons uh and sometimes uh yeah it's it's it's a complex topic which deserves its own talk actually of juice systems coinci of ones and people that call themselves intense systems this will come a lot I specifically didn't write intense down um but uh that's there what does it mean for us

well Lei is always catching up on all these things we do a lot of research around this but for now it's hard enough to go outside evm and support non EVN systems like Solana and Cosmos we are actually working with the unisoft foundation on um message aggregation so building a system where you can simply say well I want to send that message across ax layer zero and hyperlane and at least two out of three should be the same and then I'm trusting that message right um so so this kind of stuff multi hops if we think about that huge rollup as a service system Bridges will barely be able to keep up so we going to work on Multi hops where it's easy to go from a certain app roll up to polygon and then from there to salana if the users

desires that way um multihop is going to be difficult because already it is hard to get guess estimations right and then um last but not least aggregating RQ systems as part of a multi-chain liquidity aggregation protocol this is also going to be a challenge because aggregating limit orders is not that easy um you also need to be able to cancel them and all of that so it's kind of difficult um without making it frustrating for the other participants um that's what we are dealing with and if you are interested in that if you're developer um reach out to us um build on top of us use our blog to learn ask questions at me on Telegram and happy to have a chat about anything interoperability anything Bridge related we also make introductions to the bridges um aggregation is not for everything so whenever we know like for

example if you want to bring your token multichain we would give you um options which Bridges make sense and then we can redirect you and introdu you to those teams that makes it much easier for you guys otherwise yeah thanks for coming and um open to questions yeah thank you Philip very insightful are there any questions I see one on the top left here thank you so what's your general

view of intents just a holistic perspective of the current industry I think so the the idea of intense is nothing new right um and it's it's a novel idea it's good it's it's going to happen at some point but the question is here also is is here um What's the timing um we recently and IFI were looking closer into this we're talking to VC's and most of these VCS that do have researchers around this and talking to all these teams they are not very much convinced that this is the right timing to build right first and foremost you're going to build an RFQ system you set up a limit order and someone can fulfill it um so that already fulfills that purpose

of well the user shouldn't care how an order is fulfilled um then also we're probably going to see um solvers stepping up so like really um established solvers they all they that are really good they have like teams from 15 to 30 people or rust developers and they what they already do right now is trying to have the optimal execution by taking liquidity from centralized exchanges decentralized exchanges and market makers and those are going to be able to step up and fight in these intense system or RFQ systems for the right thing but then the question is okay what's next um with

nfts and and and all of that it's just I I think it's far out that's probably the one thing I think it's very far out it also requires aggregation right um so if you want to offer that for leverage and perss you need to aggregate these different Services need to partner up with them and then you you quickly run to boundaries let's talk perss if you if we are talking about perss you need high frequency um trading abilities you need very low latency you need to update prices like twice a second um uh it's highly volatile um so so the question is which intense system can actually

address this Market Niche difficult right and then um with nfts it's different issues and so so just based on the conversation we had is just like it seems like far out and I think the problem this the space has to solve different problems first I think we first going to see things like shared sequences being very established I think first we're going to see um our cues and coincident of one systems be far more established I think first we're going to see order book style uh Solutions far more established before we see intense systems really taking off so I think the timing is a little off maybe it's a much better way to think about this in two

years it's like a counter section was already available 2015 right and now we see it coming and even if we take a look at AA SD today most of them still suck still like there is still so much to do still too much to abstract away because um it's probably just too complex to build it right now that would be my guess does it answer your question any other question okay all right I don't think so

then thank you very much Philip thank you h