Solving Ethereum Rollup Fragmentation with Austin King of Omni Labs

Transcript

you're probably not going to be able to launch a super successful General evm L2 anymore uh I think the market is learning that Omni exists to empower developers to build across all rollups accessing all the users that is why we built it when we are dealing in a world where transaction fees are so cheap using something like you know a sovereign rollup it's going to be fractions of a penny uh you can make it so that you don't force the users to pay for gas I am bearish on the idea of ethereum just becoming a DA layer [Music] welcome to the interop today I'm here

with Austin King he's co-founder of omni and CEO of omni Labs we're going to be diving deep into Omni today understanding their architecture and how they allow ethereum El juice to scale and better interoperate before we get started make sure to hit the like button to get notified when new episodes drop and remember that none of what we discussed here on the Europe is investment advice and if you enjoyed this content please consider staking with us we're validating on a handful of networks and of course if you're heading to Brussels for ECC I really encourage you to to come to nebula Summit it's happening on July 12th and 13th it's going to be a massive Gathering of Cosmos uh Builders Founders and investors and I'd really love it if you come we're going to have

many any of the podcast guests speaking at the conference and of course investor speed dating will also be happening on day one so if you're raising funds and want to meet some of the uh top tier investors in the space definitely register for that as well you can get tickets and register for investor speed dating at neb. Builders and I have a special treat for the interrupt listeners have a 50% discount if you use the code interop at checkout so go ahead and get your tickets and of course if you're a student NE Summit is free so just register for a student ticket and we'll be happy to have you Austin I'm really bummed out you guys can't come uh this year um maybe we'll have to you know do another event with you guys at some point but uh yeah thanks for coming

on the podcast yeah 100% appreciate you having me so let's get started here like um maybe to to start things off like what were you doing before um working nine and how did you want to start focusing your time on building blockchain infrastructure to improve scaling ethereum mhm yeah so I mean my original journey into crypto started about like s seven and a half years ago but you know we can skip over a number of those years I think the best way to actually understand Omni is to understand the stuff that happened right before Omni so we built a defi protocol on L1 um this was a couple years ago at this point and this was just when the ethereum

ecosystem started to wrap its head around like okay we are heading in the direction of a rollup Centric road map um and so it became clear like all right this is where we're going for anybody like in the research community at that point in time even though it wasn't it was kind of far from prod ready um it was clear that like if ethereum had a future this was going to be it and so we started to think about how we would scale our def5 protocol across these different l2s um it went pretty well on L1 we got 50 mil TBL in like two days um and so we knew that a majority of the like activity uh would be on rollups in the future so we wanted to position our protocol uh to thrive during that transition as we started to get deeper into it though

uh we were talking with a number of our colleagues and basically anybody who's doing well on L1 was thinking about this problem just like strategically how do they position themselves well uh for a future where there's a number of l2s on ethereum and so at a certain point we realized that instead of just building the solution for ourselves uh and specifically for our own protocol that if we built it in a generalized way we could serve other builders in the ecosystem and it's just pretty strong signal to us that all of our most success u f colleagues were thinking about the same problem so we decided to Pivot fulltime into building Omni as a generalized infro layer for the ethereum community instead for our own defi specific protocol when executing this transition from a defi protocol to now building infra uh what was the biggest

lift or you know what was the biggest challenge in going from like an application to core infrastructure that now caters you know developers and and other types of users yeah um I mean structurally there's many things that changed so previously we only wrote smart contracts and we still write smart contracts today but we also had to build an entire independent Network around that so there's a lot of go code uh that runs Omni as well so we had to expand the team with people who had uh proficiencies in different areas of been smart contract development um and a lot of the uh like go to market strategy really came down far more to uh yeah like we shifted to being a developer facing product instead of a capital

employer uh focused product so the Target customer shifted as well so there were there were impacts on both engineering and on the go to market side as well yeah yeah of course I want to talk a little bit about you know the current state of the cycle there's this uh there's this great post um I should have it here in front of me but uh but basically the thesis of this post is that we the Cycles go from like infrastructure Focus to to application focus and the technology basically moves in this kind of like infrastructure uh in uh application kind of pull back and forth right and and that this goes back to you know the dawn of innovation and the dawn of Technology uh and I think in

crypto this mirrors a little bit you know what what we typically expect where bare markets are infrastructure focused and like bull markets tend to attract more applications like defi like nfts like social fi things like that typically we've seen these in uh bull markets what state of the cycle are we in in your view because I it feels to me like we are in a in a bull cycle but still I think the majority of the deals that we see as a fund are infrastructure we see very few applications and the few applications that there are have a very hard time raising and uh really kind of gaining attraction and gaining attraction from um from users is there

too much infrastructure are we going to be able to you know leverage all this block space or or at least um you know will all this block block space be used at some point in the future yeah yeah I mean I'm totally with you we really have had about two and a half three years of heavy heavy infrastructure Focus um think you know the like where the apps at has been a common question um it's like okay everybody just keeps raising for infrastructure projects uh nobody's raising for applications I think we're at a point though in the industry and I think this is being reflected some like things we seeing out in the landscape like you're probably not going to be able to launch a super successful General evbm L2 anymore uh I think the

market is learning that whereas uh six months ago that probably would have gotten some good uh terms for investment so I think there's a number of things here there's like how good is the actual infrastructure can people build new types of applications and then on the capital deployment side it's like is it going to seem rational to VCS to continue allocating to infra projects at this point I think uh we may be overallocated to infra but that was fundamentally necessary if you look at the app layer we didn't a lot of innovation installed so many of the applications that are out there are just like Forks of what uh happened previously you know it's like a lending Fork a DEX Fork um and that's from my perspective because it's an extremely

constrained design space um I think you know we can get into the different like impacts of all this different modular innovation um and what that will Empower at the app player but I do think we are kind of if we're talking about it like a pendulum of infra to app I think we're swinging towards apps now because we just hit the point where we can sew a bunch of these different modular pieces together to create fundamentally new like a fundamentally new design space uh at the application layer yeah what what do you think that design space opens up in terms of applications I think uh we'll be able to build non dgen applications that's my take on it um right now I think the user experience is stuck in a dgen brain space and that's because we make people

buy a volatile asset and then send it to a network before they can do anything so it fundamentally locks them in like before I use anything I have to go some buy something that might go down in price so it's hard to mentally Escape that like all right this is an investing thing I need to like think about how I make money back here um when we are dealing in a world where transaction fees are so cheap uh using something like you know a sovereign rollup it's going to be fractions of a penny uh you can make it so that you don't force the users to pay for gas uh you can actually and especially account abstraction will make this much easier but you can make it so that users won't even need to know that they're using crypto they'll just be able to sign in with email and just

make transactions uh I think the substantial increases in performance throughput and uh capacity for abstraction will open that like new application design space that is less dgen focused I wondered to what extent those applications really need blockchains so you know because as you abstract things away like private Keys users full control over funds Etc and also I think as we move away from applications that are Financial in nature the utility of a blockchain starts to erode a little bit I think like I'm not saying that social fi is doesn't have like interesting use cases or that um you know Ai and crypto don't have interesting use cases mhm but they

feel a little bit less pertinent and and maybe important than the financial ones I mean the the the financial ones I feel are deeply rooted with values and you know political positions as well about like what money should look like and uh people's like Financial freedoms whereas these other applications I think have a little bit less of that um so yeah it starts getting into for me at least starts getting into this kind of weird territory where where it's like okay how much of this is actually useful and how much is just kind of like a fancy use case for a blockchain um similar to you know a lot of the a lot of the U the

hype around blockchains in Enterprise like the sort of mid 2010s right where every Enterprise was trying to build a block you know supply chain like let's you know I I built and you know built a like a failed company around around this very uh this very this very use case and and and Pitch so like yeah what are your thoughts on on that and like you know our our users you know do users actually sort of care that social fly applications exist and and does a does a blockchain social fly application really introduce novel use cases for users that are are that far away from the existing of social network infrastructure or like application space

yeah so I actually totally disagree with you on this uh so let's chat about it the um my perspective comes Downstream of the fact that blockchains as a technology Empower us with property rights on the internet that is the lens at which I look at things um money having having the ability to own money obviously is extremely important but property rights are impactful throughout Society for things far beyond just money um and so I'm thinking much more widely than just social Pi I think this much more is about empowering people with provable ownership in a digital space um because right now we effectively rely on like five companies in order to

attribute ownership over things and you know with that we know it's not ownership because nobody's making money off their like online presence uh and those that are are explicitly like paying a significant amount of that to these five corporations and so that's the lens through which I look at it's not about like having tweets that are directly attributable to you it's about having your entire online presence uh provably owned by every individual out there uh AKA like digital property rights yeah no I I think like that narrative makes sense I just it to me it comes back to this this idea of like in the end do people care right there was this whole moment in like 2018

2019 when privacy kind of came to the Forefront I believe it was when wetsap changed their their their their terms I can't remember exactly what but but basically like a bunch of people started using signal including myself yeah I I saw it as essentially you know folks uh moving over to Signal because like a lot of people were also moving over to Signal not really understanding what was happening there and then you know eventually like nobody used signal anymore except for you know people like us who are sort of you know nerds and privacy conscious but the majority of people don't care and um yeah I mean I get the argument I just U I guess I'm like a bit cynical here I I would argue that yeah the my what I'm saying is not

that people are going to shift Downstream of because something is private I think we're going to be able to create fundamentally new types of like networks and applications out this that provide direct utility the selling point to to end users uh the selling point to end users is not uh hey this is more private or hey control your own data it's uh you know what just like at a much more left curve level like what are they deriving utility from um so yeah I think anybody that goes to Market on like this is the Privacy x i if it's end user facing it's probably not going to go that well and I think we have seen many different laps in companies attempting that and none of them I think have succeeded anywhere near the scale

that uh companies that did not focus on that as the core value proposition have seen in the market yeah I mean I want it to happen man I mean like I'm yeah if you're using signal I imagine but no I want I want it to happen like I I I I I would love uh a user owned alternative to you know what we currently have where users drive not only value creation but also ACR value from their from their data from their usage of a platform uh you know are able to govern the platform uh Etc right like these are really laudable use cases when it comes to crypto um yeah in my opinion it's

actually when we're thinking about it from a consumer like and user facing perspective I don't I don't think anybody's going to want to govern things people don't even like vote in the country that they live in sometimes let alone helping manage some online application that they use I think what is going to be more compelling is the fact that and you see this in open source cultures where uh everybody can contribute freely and the rate of innovation accelerates uh instead of just being like closed down in a propri platform external Innovation slows and so I think it's more about betting on an ecosystem of people being able to collaborate better people who don't trust each other can cooperate better at large over the Internet um and I think that applies specifically at The

Innovation layer where we'll see New Primitives come out people start building around them other people start building around these new uh things that people layered on top so I think it's more about facilitating open Innovation across all these different uh platforms and that will lead to better user experiences and new types of applications yeah absolutely well uh let's let let's maybe switch gears a little bit here you know one one one thing we were talking about here before the show was um you know what is ethereum and uh we we were kind of joking about the fact that that uh you ethereum and also Cosmos are are stuck in in this identity crisis right now I think they're they're a little they're slightly different but but they have a lot of

similarities so yeah from your perspective like what's what's what's happening uh in ethereum right now where um you know that that's that is causing this confusion and and maybe this pull in different direction yeah it's actually very funny phrasing it that way because you can look directly at some history is repeating itself in many ways uh with Adam 2.0 we're like okay we're going to be sharing security from The Hub it's like reaking ethereum um and this is Downstream of the fact that people are like what is this core L1 Network here for from perspective the reason that I love the ethereum community is one the uh the community and the like Innovation that happens there I feel it's like very genuine and

like uh just most aligned with my personal values of actually trying to create Sovereign software for people across the world um but I do think ethereum is going through an identity crisis right now where the obvious technical path to scaling ethereum was l2s but that kind of destroyed the existing business model and a lot of people are wondering what happens next and what they should even optimize for with the core L1 Network like should this become a shared sequencer thing should it focus on issuing pre-confirmation to different L2 platforms or should it just be a DA layer I don't think there's uh consensus around what we are trying to optimize it for so I do feel fragmentation in the mind share of the ethereum ecosystem of

where people are trying to take things in a way similar to what happened uh to Cosmos over the past you know like five plus years yeah as you're saying that I'm reminded of like having a conversation with my uh my excellent co-host over at uh at epicenter Mah Roy right around the time this ethereum um rollup Centric roadm post came out and ethereum 2.0 was was rolling out and and he basically you know was sort of predicting this very this very problem that uh proof of stake and the rollup Centric road map would put into question the ethereum business model of acing fees M um and um and it looks like you know that has sort of

come to pass um where where do you think we go from here you know with uh with reaking and then rollups like what is the best outcome for ethereum layer one in this new reality right where like these things are not going away yeah um yeah it is fascinating I think uh if you think about the ultrasound money meme that really through people uh off course personally that's my that's my take that it was I don't know how helpful it was to push that Meme because now it's not ultrasound eth is inflationary again because a majority of the activity happening in the ethereum network happens on l2s now and L1 captures way

less in fees Downstream of that so it was honestly incredible that ethereum was deflationary for a bit like that is astounding that we were able to create a decentralized network at the scale of ethereum and have it be ultrasound money like this thing is literally deflationary um that's not happening anymore though the network now after we have 4844 so like blobs on ethereum uh it's inflationary again so that I think is just whiplashed people a decent amount um and understandably it makes complete sense why people would be confused about that moving forward I am bearish on the uh idea of ethereum just becoming a DA layer I think that is is uh I don't think that has a compelling

business model um and it has so much Tech debt when you compare it to new uh platforms that you know just can use the latest Innovations and simply offer a better product um so I'm quite bearish on that future if I'm thinking in terms of somebody who cares about eth the asset I think a lot of people are biased by that though and they think about eth the asset instead of ethereum the network the reason I care so much about ethereum is because it's the most decentralized uh Network that we have out there um and so it can provide the highest guarantees on a number of security Dimensions that I think is where ethereum really excels compared to some of these other networks and so I'm not a huge fan of some of these proposals that are out there for like uh

reducing the validator Set uh and like putting constraints there so I think that's what ethereum really excels at being this like massively distributed core layer that uh can yeah just achieve the highest levels of decentralization and ideally censorship resistance on top top of that so that's what I would prefer if the ethereum community optimized for um I don't know if that's the best business model but that's me thinking about like what I think the world needs not how do I make eth go up yeah I I think you're right about eth not being a great da layer um yeah I mean it it just it can't compete with the likes of Celestia and and ail and like aan da in terms of uh technical advancement like this this technical

debt I think just puts it as a disadvantage the community the highly decentralized nature of ethereum the highly aligned values of that ecosystem I think are its strongest attributes and I guess the question is like what is that useful for what what what do those these things allow ethereum to become um and I think that one uh one candidate is for ethereum to serve as a ZK a verification layer for most of crypto where it's just like does ZK proof verification very reliably in a credibly um like credibly neutral way

with a decentralized network of validators that will just you know reliably do that um at scale you know possibly with like kind of super proofs the likes of which like electron are building yeah wonder what you think of that yeah I I think it is such a fascinating design space to look at if you think about all the lessons we have learned from the time that we originally uh launched ethereum we understand far more about the different properties here and the different like layers really I think Celestia has done a great job with helping people understand like no it's not blockchains it's like more akin to web 2 honestly where they're going to be you know we're going to call them modular networks but really they're like microservices in the decentralized stack

that's how I think about it um you know you're talking about this like recursively you know super proof I think you said um but just like a proof that Aggregates into other proof and then just get submitted to a Bas Network that's a pretty elegant architecture um but like we there's so much Social momentum and liquidity uh and users already on the existing ethereum Network that if we could relaunch Everything clean slate today might it make sense for the core L1 Network that is maximizing for decentralization to just verify ZK proofs yeah that's honestly a pretty elegant architecture uh but we're dealing with this world where there's so much momentum already uh that yeah we'll see like what we can

make happen here um but if we think about it from entirely removed engineering perspective that would be a pretty cool architecture yeah I mean we I totally forgot that also ethereum has liquidity also one of its very uh like strong attributes compared to other ecosystems let's um yeah let's let's talk about um Omni here so what's the vision and maybe talking a little bit about the architecture you know in this rollup Centric World in which we live what purpose does omni serve and you know looking at the way Ro you know that rollup ecosystem looks like today what's the desirable outcome that Omni enables yep yeah so from Simply put Omni exists

to empower developers to build across all rollups accessing all the users that is why we built it um and really it's an extremely developer focused product because we started building this for ourselves and we ended up committing to building in a generalized form so other people could get the same utility of just make your application available across all these different rollups um that is what we do at Omni that's what we care about um and the goal really is we provide that utility to developers so they can focus on building the best appli running their preferred stack um but the what we need to do as an industry is make the users not think about like users don't care if like it's windows or Linux running a website uh on

the web and the same will apply in crypto and so the goal is Empower developers to the highest degree uh while giving them access to all the users um while making users not think about the nuances of this infrastructure stack that's very much our design perspective at Omni yeah let's dive into the architecture a little bit here um I was reading your docs earlier and uh you guys have this kind of interesting approach to using Comet bft and some of the cosmos stack combine with you know um parts of EDM infrastructure and allowing mostly ethereum uh rollups I think at least in this current iteration uh to participate in this this network

of of rollups uh can you talk talk about the architecture and how Omni is constructed to enable this mhm yeah so there are two core uh Primitives in the Omni Network that facilitate this uh output of empowering developers to build across all these different rollups one is a messaging layer so that rollups can pass messages to one another and another is a compute layer and so this allows people to run like turn complete applications um that is the like very zoomed out to Big AR texture pieces there um we can dive into each of them now so the messaging layer um is effectively a way that uh rollups can

speak to one another you know if we're thinking in the cosmos comparison we've had IBC for a while the difficulty with rollups today though as it stands is uh most of them are servers that is simply the reality uh and so it's extremely difficult to let one server talk securely to another server with like so so that end users can be confident in the uh like message that gets passed there um so Omni specifically is uh a way that messages can be passed securely to these from one Ro up to another actually secured by eth the asset itself um the other layer here is the compute layer which is an evm uh itself we specifically built it as an evm so that people in the ethereum community

wouldn't have to go learn a different language and as they were expanding their applications across all these different rollups uh they would just have access to the same tools in same programming language that they've used historically so those are the two core Primitives that give the output of like allowing developers to build across multiple rollups okay interesting and so this interruptibility feature can you describe because I mean coming from Cosmos I think we're all very familiar with sort IBC and how that works U did you guys build your own inability solution here or using something off the shelf yep we built our own okay can you describe how it works and maybe like how similar it is IBC or different yeah so I think the best way to understand the structural differences

here with IBC it's really nice because you have local validator sets uh anytime one chain wants to talk to another so if a message is getting passed from chain a to chain B that message is going to be able to be signed off by the validator set of chain a and so it gives you uh like proof that all the participating nodes uh agreed that this is going to be part of the state transition that happened over there uh we don't have that in the ethereum world and so uh what we did instead is we kind of built like a global validator set in the sense that this is an independent network of people uh who are secur like their nodes are secured with restak eth and they are testing to State updates uh from optimism Arbitron base Etc uh and that is kind of where the cryptoeconomic

security is derived from so instead of localized validator sets attesting the State updates it's a global validator set that is like secured by eth the asset itself okay so this this Global validor set is the the Omni validator set so I mean validators are running Comet bft um you you mentioned like base arbitrum um does omni connect directly to these uh these ecosystems or does omni allow basically sort of roll up as a service type uh like a rollup as a service where anyone can build their own roll up I

mean sort of you know Y what kind of users are we targeting here yep so and this I think is one of the most interesting uh the biggest distinction with Omni compared to things like polygon AG layer the optimism super chain um all these different kind of like localized Solutions it specifically links into it's like an overlay Network it links into each of these rups and it doesn't need uh for them to give up control over some part of their infrastructure stack the reason we built it this way is because it's the pragmatic path um we don't want to like build a solution that you know solves fragmentation but is it only solves fragmentation if you build uh a rollup on our platform there's a number of I

mean most basically every single other uh network is doing this uh we did not do that with Omni because the reality is is it doesn't solve the problem if you're building an application you want to access all the users and liquidity that are out there you don't want to access some subset uh specifically you don't want to access some subset of a new network you want to link into the existing users in liquidity and so we designed Omni so that we do not Omni does not force rollups to give up control over their stack at all uh and it empowers developers to link into the largest uh rollups that are out there today as opposed to something like agay which is quite cool but uh it still is like a subset of the ethereum world right so you you could build an

application on Omni and tap into liquidity from base and arbitrum and like any rollup on ethereum um and this this kind of I think is is different from the way we've thought about building applications so far is to you know have outposts so you build an application you'll have you'll have your application deployed on every one of these networks whereas here a a developer build their application deploys it on Omni and then can tap into liquidity from all of these different networks um is that is that a is that the sort of accurate way to think about how Omni flips uh that model yeah so uh

a simple way to conceptualize this is uh Omni works really really well for applications that want to accept collateral across all these different rollups um but not have fragmented state or liquidity um so let's think just concretely here about what a per protocol could look like anytime you're trading pers you're going to deposit collateral um if you were using Omni the outposts like the deployments on these different rollups uh would just be like deposit and withdrawal uh for collateral those collateral updates would be sent to the Omni evm though and that is where you could uh have this Global Market where you're accepting liquidity deposits from all these different

rollups and so that is how we like think about this in the sense that like you can accept and withdraw collateral across all these different rollups um but you want to keep Global State somewhere um and so that happens on the Omni ABM for your application and that makes it so that you don't have to deal with like fragmented redeployments of your application Le like this Outpost type pattern okay interesting so in my mind when I was first researching Omni I I kind of thought it was similar to I me I guess it is maybe in its architecture but I think in the outcome it's it's a little bit different I thought it was similar to Anisha where you on Anisha you have this kind of like shared settlement layer and sequencer and inability layer that allows you to build

and deploy your own custom application uh on this rollup as a service uh infrastructure whereas Omni feels more like an interoperability solution um or like an inability layer across different rollup ecosystems that allows developers to build applications that tap into that existing liquidity and like build kind of like cross applications in the ethereum ecosystem is that like a fair compare like a fair way to to um compare those two I would say the initial comparison is semi accurate but the like interop like cross we don't think about it as cross-chain applications I think that is necessarily implies that people have to think about the chains and right

I I think that's the wrong design uh perspective because users don't want to think the chains um if we're actually going to expand the users in the space they can't think about the chains uh that has to be abstracted so uh Anisha I think is honestly an interesting example of a project that that we talked about it's like okay everything we've learned over the past nine years or so since we've launched ethereum if we could rebuild things from scratch what would it look like I think Anisha is kind of an example of a project pursuing this path with Omni though uh it's a more like all right we got what we have have you know things might not be perfect in the ethereum ecosystem uh but all the users and liquidity are in the ethereum ecosystem so how do we provide

developers with a way to build across all these different rollups uh so that in the future they can like any user that enters the space it doesn't matter whether it's on like polygon base Etc that can be abstracted from the user and they hear about an app they can just go use it out of the box so Omni is like part playing nicely with the ethereum ecosystem and making it so that the user experience can be abstracted while still empowering developers with access to all those users Anisha is trying to build it on its own and I think it's it's a fascinating experiment um but yeah there's two different like go to market strategies here effectively uh if you think about it yeah speaking of uh playing lice with ethereum um I

think currently Omni only supports evm rollups mhm um in the future you know are you guys planning on also enabling nonm rollups potentially you know Cosmos app chains or you know salonic VM uh l2s to also tap into this network yep 100% And so this is something actually that we uh were very intentional about making sure the network would be compatible with different virtual machines and because we saw that uh sure the first rollups were probably going to be evm but like people are going to play around at this level if we can couple uh this like consensus from the execution side people are going to play around with

execution and see what different types of virtual machines we can bring into play and so you know back then one of the ones I was more top of Mind was uh starkware and think like okay this is an interesting early experiment in terms of different virtual machine architecture um but there's going to be more and so Omni is intentionally fully decoupled from uh a virtual machine uh it's pretty lightweight in order to uh kind of rewrite the interface that links into each of these different rollups if it's in a different language um but initially rolling out the network it's just like it's a large scale network uh reduce complexity where you can start with just DM yeah H how much of that ability to work with other

VMS uh is attributable to the fact that you're using ABCI ++ um of course you know maybe just taking a step back here ABCI Plus+ is the infrastructure layer that sits on top of comet bft that allows Comet bft to talk to an application up until now it has been used primarily in Cosmos chance to talk to the cosmos SDK but with ABCI Plus+ 2.0 is now more modular ready and its ability to talk to other Frameworks so end of end of parenthesis um how much of you know that ability to tap into other um VMS you know comes from this ability this this use of of of

ABCI 2.0 yeah so it ABCI 2.0 does not directly facilitate this but the design methodology that led us to using ABCI 2.0 uh did lead us to to do this which is uh basically just like clean abstractions like where can we isolate functionality and have specific components do their thing really well while leaning on other components to do things uh like totally different responsibilities very well and so yeah Omni it introduced a new chain architecture um historically we've seen many different chains try to create a fast finality evbm in historically this there's been a lot of struggles here uh specifically because people try to

rebuild EVMS on top of the cosmos SDK these are very different they have like different state representations uh there's very different constructions in the men pool um yeah it just like historically from a literal engineering perspective has been riddled with uh difficulties and so what we did instead was ABCI 2.0 came out which is this nice clean abstraction between like all right we have a consensus engine uh like we can push the outputs of that uh elsewhere into a different execution client um and then similarly on the kind of evbm side of the world there's the engine API which is kind of like the ABCI 2.0 equivalent uh but for execution clients and so uh what we did is we

built an architecture where these could uh like talk with one another and so that allows us to get the best from the consensus side as well as the best from the execution side and uh it's a really nice clean pattern because we get to get an extremely battle tested yet performant uh consensus engine and then we also can use just literally right out of the box gu uh ebm implementations and so uh it allows us to we kind of touched on this earlier it allows us to lean on the work of other open source contributions in the space and this is something that uh yeah like our implementation of this is open source um it's something that we think is really important for the space to have as a primitive to build so that we stop kind of like trying to rebuild different ways

to get a fast finality evm right okay is is is this ethereum engine API sort of akin to Ether Min uh no so ether Min is an example of rebuilding the evm on top of the cosmos SDK and then you have complet okay simple ether me rebuilt the evm and and you guys would have built an API that allows you to use the existing a evm with ABCI yep exactly the simple shapes way of thinking about it is you have Comet bft uh ether rebuilt an evbm on top of it we run it in parallel and we didn't rebuild it we just use like gu or WTH as now that it's out okay run in parallel similar to the way ethereum so e 2.0 does this actually it's like

actually two chains there's like a consensus side and execution side uh Omni operates that same way right so so you have the consensus client and the execution client that API just allows you to do that and run I guess this this kind of like ties into my next question which is this dual chain architecture which I I think is kind of related here so what's the what's the reasoning behind this design Choice um does it does it offer better performance or security U than like using U something like ether Min right where the VM has been rebuilt um yeah yeah I would say the two most important things here are well three one battle tested we get to use extremely tried and tested uh

software that has been out there for a long time on the ethereum side this is uh execution clients like gu on the uh Cosmos side it's Comet bft um so that's one it gives us a really it reduces surface area for vulnerabilities um so security is improved through that um two performance uh if you rebuild an evm on top of Cosmos SDK it's going to get choked and we have seen this we ran this in our second test night actually uh we were running etherman and it got destroyed once we hit like a few hundred thousand users um and a lot of this comes down from like having to translate State representations from evm world to Cosmos world but also the cosmos SDK M Pool is just not built for evbm scale

load um and so from a performance side uh what it allows us to do is put all the like heavy execution side things on the ethereum client um while still having this like very fast uh high performance consensus engine to get like 1 second finality uh for things um those are two I'm trying to think of what the third thing I was going to say is but uh oh the third thing is uh it just e you know people throw around this idea of evm equivalence um this is literally in ethereum node like it is all the tools that are out there uh that are built for the ethereum ecosystem will work fully out of the box if you try to like rebuild things often there are edge cases where it's like oh well like this

thing doesn't work super well with the cosmos SDK so you get close to like running a full ebm but it's not 100% and 100% is extremely important for like no literally every tool that's out there will just work uh out of the box with this right earlier we talked a little bit about reaking and so far we've talked a lot about the architecture uh so now I'd like to understand how Omni is secured and specifically what's the role of reaking yeah um so the primary way it is secured is through reaking as we were thinking about scaling our own def5 protocol originally like I mentioned we looked at solutions that were on the market for like okay how could we like create in or what infer layers out there could we use that would facilitate this

experience where we could just build across all these rollups and access all those users um a very like unimpressed simply put like a lot of the uh solutions that are out there just are nowhere near what ethereum developers would consider secure and safe um and so around this time we met the Igan layer team this was a couple years ago at this point um but it was very serendipitous in the sense that it kind of clicked it's like all right we're trying to build an infr layer for the ethereum network like it is a chain but it's the first chain to complement ethereum not compete with it um so uh when we met the igen layer team it's like okay naturally if it's a chain that is built to complement ethereum it

makes sense that these infr layers built around the core L1 Network would also be secured by L1 the asset um on top of that you get fantastic security properties because you can scale the cryptoeconomic security um far beyond uh a multisig obviously but definitely far beyond uh just securing it with your own native token and so a lot of this very it like clicked for us very early on I think just in a weirdly serendipitous way one of the most important applications of reaking uh happen to fit very nicely into uh the solution or into the problem that we were trying to solve and so um yeah how does reaking play into Omni uh Omni is a network of nodes that provide infro services around the

core L1 Network and they're secured by eth uh you know restake eth uh the asset uh to derrive crypto Economic Security so that's the primary way uh that it's secured right so Omni is in effect an an AVS on on aular yeah yeah yeah so it pulls in the security uh from reaking contracts I think we've seen uh like symbiotic recently I saw earlier today they just passed a billion dollars in assets supplied to the protocol um and so I think we are going to see this like concept of abss um probably proliferate because there's a number of different reaking protocols at this point proposing uh like security offerings so

we'll be quite interesting to see how they compete with one another um but honestly uh it's healthy uh it's cool to see different people come with different perspectives right I mean how do how do you see this Market kind of evolving right where you know there's there's several reaking protocols now like symbiotics one of them um you know will protocols like Omni abss choose to use the security of multiple reaing protocols and then what are the considerations that a protocol has to taking into account when choosing these protocols quality of collateral quality of like the valid sets Etc yeah um yeah

this is a great question it is my perspective that we're extremely early in the reaking game um if you think about it uh reaking isn't even here in the sense that we don't have slashing if there's no slashing there's no security that is simply the reality um and so yeah we haven't even had V1 really of any of these products shipped and so I know that there's like 20 billion dollar collectively in these contracts um but that's if we're being real that is people farming uh tokens like those tokens aren't actually providing Security today um that's a simple truth about it so I think we're extremely early in this I think it's uh healthy to

see people trying to innovate and on this idea in different ways if we go back really this idea like you know we've talked some about Cosmos in this uh this idea of reaking got branded really well over the past two years in the ethereum ecosystem but it's been around you know it was in pokka dot it's been in the cosmos ecosystem so uh this idea has been around for quite some time so I'm not surprised to see different teams experimenting with different methodologies I can tell you from our design perspective at Omni uh we care deeply about permissionless of this uh there should not be some multisig controlling all of it that's defeats the point in our purp or in our opinion um there should not be forced

rehypothecation uh this is something that I think uh different networks or different reaking offerings are going to uh play around with um and there should be slashing uh there should be like simple slashing interfaces so that networks can design uh the best way uh for them um yeah there's going to be a lot that we'll see uh happen here over the next couple years I think we're quite early in this game yeah I think so too and I mean of course like Bitcoin reaking is also now playing a role uh in like kind of pushing this reaking narrative to a different network with you know arguably

a lot more security and a lot more potential uh stake um so you know I think we will start entering into a world where there's like a Marketplace of you restake providers that compete on on quality of security on fees on liquidity uh on all these sort of different properties um you know when thinking about the re I just had the veilance team on uh last week and they're building like a really interesting uh Marketplace for protocol own liquidity and the thesis here is that you know currently defi is mostly what you would compare to in sort of traditional Finance household thatb um um where institutional debt has not

really entered the market crypto institutional debt and that's simply because we don't really have the protocols for that we don't have sort of protocols that enable a Marketplace of protocol and liquidity um you know in Cosmos we're familiar with this idea of protocol own liquidity because like Cosmos SDK has enabled and entered a lot of deals with different networks to provide protocol own liquidity but but these have been mostly kind of manual right right um you know what what do you think creating more programmatic opportunities for protocol own liquidity to enter into the ABS ecosystem the reaking ecosystem you know what kind of exponential growth

effects do you think that that can have uh you know you said we're still early but like that that that that liquidity boost how much how much growth could that bring to the ecosystem you think yeah I've been chatting with a few different people people about there's interest there's an interesting design space for aligning people who are securing the network with using whatever Network it is um and so if you can align the liquidity the capital that is securing it and somehow facilitate that into the usage of the network that is a super interesting space because there are yeah like I said there's about $20 billion in these contracts right now um I think protocol owned liquidity itself um is an interesting idea uh you know

actually the def5 protocol that we built kind of was an early uh iteration upon this where it still wasn't like protocol owned liquidity it was still renting but it was like a way for uh these networks to create create and establish liquidity on it was kind of like a decentralized market making protocol where the Dow engages with this protocol and can Source eth from these different uh Capital providers I think protocol and liquidity specifically is an interesting possible blurring of responsibility where these networks themselves many of them are far from product Market fit and need to like iterate and innovate uh heavily until they like establish a solid spot in the ecosystem when you think about protocol

and liquidity it's like a hedge fund uh it's like now a DA is controlling some liquidity so it's interesting I remember being extremely excited about this concept A few years ago um but it it is an additional responsibility for innovators to think about uh in these networks so I think the more interesting question is how can we align the interests of people who are securing these networks uh with you know incentivizing them to uh participate in those networks themselves yeah I mean that that that liquidity provisioning you can come with if you Pro if you create programmatic agreements I think you can include conditions for that liquidity provisioning to also come with certain conditions right like like this

this sort of like participating in governance um actively participating and deploying uh capital in you know different D5 pools that may like need liquidity or Market Maring I think there's like a whole bunch of things that this enables that I find really fascinating I think will will continue to develop over the next couple of years um I agree you know we're we're almost at the end here but I did also want to we talked about this a little bit but this is a topic that I like to ask my guests about you know and since you you know omnis sort of deep in the evm ecosystem you know what do you think are the prospects for or um alternative EV alternative VMS that are gaining traction you know the move VM the salana of VM to some extent cm

wasm and you know coming back to what we were talking about earlier with ethereum having a lot of technical that I think the evm also has a lot of technical debt how do you think that will fair for the evm in the next you know like 10 years um yeah yeah um yeah it's a great question I think uh it is highly ambiguous we will see what happens here there's kind of two camps there's one it's like let's optimize the evm and this is with like optimistic parallelization you know creating like database improvements um and this is kind of like say monad uh Mega eth is another one um and so they're kind of taking a bet that the evm is like JavaScript and even though uh it is

poorly designed from an objective perspective uh they're going to optimize it and it's just here to stay and so the when are going to be people or teams that optimize the evm there's this other Camp uh of people experimenting with like WM VMS uh you know Cairo from starkware move um and they're taking a bet instead that uh l2s offer an interesting uh opportunity to now that we have decoupled execution from the like consensus layer here in the ethereum ecosystem we can play around with new VMS and so uh I think this is healthy I'm excited to see the experimentation that happens here I honestly don't know all all I do but the

only thing that I would say I'm confident in is that on a 10-year time Horizon I do think we will reconsolidate around a single uh virtual machine architecture um we've kind of seen this in web 2 where there's like an oligopoly of VM architectures uh and so there's only like a couple you need to think about really um the rest are just super longtail and not you used widely um I think just from the perspective of developer tooling when you think about like how support you know when you maybe this will be different in 10 years with chat gbt but uh when you're picking up a new programming language at least when I was in college uh being able to Google a

question and have a stack Overflow post come up and somebody else who ran into that problem it's extremely helpful as a developer when you're trying to learn a new system and so there there's a strong social momentum as well as like open source engineering momentum for these different architectures and so I think we're going to see things expand in a ton of experimentation over the next like one to five years and then on that five to 10 year Horizon I think we're going to contract uh heavily back to either one or just a couple uh virtual machine architectures for crypto systems yeah I think that sort of mirrors the way I think about it and you know in web 2 there is this oligopoly of essentially node and PHP uh and then for other more Enterprise applications maybe

Java and you know I hear you can get paid you can get paid pretty well if you're a Cobalt developer 100,000 a year that is a meme that I've heard since College like hey if you want to learn for Tran and work on some 70y old code like yeah just like for some reason that's a meme all around um but yeah I think like just uh on that note I think uh we will have experimentation with different languages in flexible environments like if it's a wasum VM um but I do some I do think that something like move which is specifically designed by like the like these are top tier programming language experts like they came from Facebook and Facebook just fumbled the bag but now it's this open source framework but it's

extremely high quality something that like a tech Titan would ship um so yeah it'll be super interesting to see how it plays out yeah yeah I I had a I recently had a conversation with um freder Ernst uh of nosis on on epicenter and we were talking about this and I think one thing that kind of came out of that conversation that really shifted my perspective is is AI and and the role that AI will will play in really shaping the way we build applications um you know I also spend a lot of time in stack Overflow and you know trying to like fix any sort of problem or like uh you resolve any issue that I was having while coding and I don't code so much anymore but like I

just built you know we run a fund now right so like my whole life is an Excel and we just I just built like this massively complex um uh tool to track our portfolio and like all the formulas were built in chat GPT and you know I just kept like and these are really complex formulas right I built this thing to be modular and reusable and so like I want I wanted it to be I wanted it to to work like to be constructed a certain way I was really thinking about as like a programmer would and and chat PT built all of it and um you I think that perhaps that tells us something about where you know blockchain development is going and development in particular in general but you know it may very well be

that at some point you know we abstract away most of the development language you give a query and that thing gives you op like it it produces op codes um and you don't even have to think about a language you can just like talk to a VM directly so that that might give us an indication about where things are heading I think you know we'll probably have programming ledes for a long time but there is there is a possibility that that also gets abstracted away uh I know I I don't know to what extent this actually came to into reality but you we interviewed uh eming gunier um sunny and I uh in um in Barcelona at their um Avalanche event like two years ago and they were building this kind of like llm

on a blockchain I don't know that they actually like ended up building this but this was kind of the idea right where you could give a prompt and that prompt would get translated directly into uh op codes and transactions so yeah that's yeah sort of like thinking very far ahead but where where we might be heading yeah it's possible that we start with like op code level stuff like assembly and then and I'm I'm speaking outside of crypto I'm speaking about just software engineering in general it's possible that it started with assembly and we created incremental abstractions so we got to things like JavaScript it's possible we end up at English and like and when you program in the future uh you just speak in your

like native human language um we'll see what happens uh here I think that is definitely possible though um because yeah like I think we're at the stage where you need to be an engineer to be able to talk to chat GPT and kind of like at least string the things together and identify what it's doing wrong um but I imagine we will pass that point in the next few years here um but right now I feel you like I mostly participate the like protocol research and like architecture side on engineering I don't really code anymore but like some weekends when I've experimented with like all right how how good is this thing as like co-pilot it's good it's quite good especially when you're like picking up a new stack yeah yeah I know it's impressive how you know in just a

few short years how good it has gotten at cod in well Austin uh thanks so much for coming on uh working people find out more about for Omni and maybe yeah if you have any sort of announcements to share or any upcoming uh things people should be looking out for yeah sure thing so um core website Omni network uh Twitter Omni fdn like Omni Foundation uh and then uh yeah upcoming test net uh here super excited to have this like final lap taken to test out all these infrastructure pieces um yeah and I'll be in Brussels so if anybody's there uh say hi yeah absolutely it's funny I I've seen a lot of these twitters fdn I thought it

meant fully decentralized Network I didn't realize it meant Foundation um but yeah all right I guess I guess that also works yeah cool well looking forward to seeing you in Brussels cool thank you very much man cheers