Scaling Ethereum Summit | Connecting our Multichain Future - Jon Kol

foreign [Applause] [Music] speaker for today which is John Cole from hyperlink uh John's going to talk about now with alternate Roll-Ups that we have whether it's Zeke or optimistic or anything else around modularity how do we make sure they all work nicely with each other and that's where multi-chain world comes in we're understanding what's possible what's not possible how do you think about making what should be possible is going to be important so for this last talk for this Summit I'd like to welcome John here with us and uh have him get started hey John thanks for having me I'm going to start by sharing my screen here and we'll get right into it wonderful and we'll get that loaded up SMS we see on our end all right so clear okay so what are we going to talk about today all right Karthik just told you that we have a multi-chain future ahead of us and we're going to need to figure out how to deal with it it means really a lot of things right so this talk today it's titled connecting our multi-chain future so in this talk we're going to cover the current state of blockchains the need that stems from that current state as it pertains to interoperability then we're going to touch about how all of that relates to the scalability of blockchains which of course in particular means the modular blockchain thesis and then we're going to finish the call with this little thing that we're doing here that we call the interchain highway and so that's why you see that little subtitle of the interchain highway so this is uh if you're curious you want to know what it is well this is what stable diffusion thinks it is uh and I kind of like this little impressionist style that I gave it but let's get right into it so the current state of blockchains as I like to say with my dad joke style the state of state of State so we have these amazing internet computers and they're creating loads a valuable state but they're isolated they can't connect everyone here knows this right everyone who is really sad throughout the day and and is going to participate in the scaling hack like this is really not news to you now you realize that the valuable state that exists and is created on one chain by and large is not easily accessed by another even if we're talking about a roll-up and certainly if we're talking about a roll-up on a roll as we will get into uh later today there's been this uh expansion now into a multi-chain environment which is that most often it's meant deploying your app on multiple chains and effectively this has resulted in creating multiple instances of your app on those streets right definitionally if State can't be easily accessed between chains as a developer building on a single chain you really can only access and serve users that exist on those chains and that pertains obviously to every bit of State that exists there this is really sad because what does it mean it means that if you now want to expand to another environment right you were um the very pot let's say you launched as an app on arbitrum and you're like whoa I want to expand to optimism as well for whatever reason maybe those are distinct audiences well your next deployment does not benefit from your existing one so now we're expanding actually creates another form of fragmentation between these blockchain environments it's uh yeah as I said it's it's quite sad Now worse yet we're transitioning into a world where there's not only going to be more chains and deploying your own chain is going to be easier and bear with me today I I know this might not be acceptable nomenclature but I'm going to use the tame the term chain as a blanket term to include what would be an alternative layer one what would be more familiar as a roll-up whether it takes the form of an optimistic roll-up and a ZK role and certainly when we need to have those distinctions we're going to get into them specifically and thirdly I'm going to again treat something like a Cosmos app chain simply as a chain so when I say chain don't get hung up on the notion that it's just a layer one it is a blanket term for any one of those three so again we're moving towards this world becoming whoever more popular through the realization of the modular blockchain thesis that not only is deploying your own chain more common it is also made easier because there are many more of these toolkits but what then right you deploy your own chain your own environment and effectively you've created you've sprouted out a new planet in a solar system a new island in the ocean if it's a rope for say uh and it's an ethereum based role you have one main line into the main economic Hub right into the into the main continent to call it like uh into the main planet in that star system but that's all you get what about the other planets in that star system the other Roll-Ups what about the other Islands on any island chain how do you connect to them yeah are there any what are the trade routes that you can access right and so this that is uh where we start getting into some problems lastly as we analyze kind of the current state we have to talk about just how many infrastructure interactions we have as users in crypto in fact the new user almost always has their first interaction with an infrastructure before they ever reach an application you know this is something that whenever you take a step back you realize this is this is certainly not the way recently a good friend was talking about their experience onboarding a close friend who had never been into crypto but was interested in collecting a poapp that poapp was on gnosis they tell their friend to download the wallet of their choice the friend downloads rainbow finding it to be a very acceptable and easy to use um that Mobile Wallet then they realized off gnosis is not supported by so now my friend has to onboard their friend to metamask on a desktop they have to move the eth that was purchased in the first wallet they didn't want to expose the private keys so he tries to onboard them to that and now they're already 30 minutes into it this person being on board it still has not reached this final application collecting that Co-op that they wanted to they realized they have to enter the RPC for gnosis and very cumbersome they they can't find the RPC address so they go into the noses docs and they get their long story short they spend about 90 minutes for him to he has to explain to his friend and that is only after all of that work dealing with so many parts of infrastructure did the new onboardy and the crypto get actually exposed to uh to the ACT no one here probably thinks that that's acceptable right there's no way that uh that we can work with that so this isn't the end state right like what is right you know uh how do we get there well how do we get to this point where like we're serving people a lot better we know that whenever we do get this demand we have really big scaling woes and obviously this event uh that we're all embarking on in scaling ethereum is deeply related to that we're going to try and find ways to do this more um more acceptably and so these internet computers of ours if there's one thing we know about like we know how they work right we know that they work by making digital resources more scarce than they are otherwise and so we know that whenever demand increases unless the supply of these resources changes we're going to have a rough situation uh we're gonna run into this situation then I'm sure everyone on the uh on the line today has run into it you just realize just how high uh that gas cost is and how do we solve that so we're going to talk about this today and to solve this we're going to talk about the combination of interoperability um modular blockchains and a few other things that will come up so what does inacceptable like multi-chain future look like well it has to be more uh has to be more acceptable than we need to find a solution to that demand problem we don't want the gas to be too damn high all the time right should we ever talk about a future where demand doesn't increase no we shouldn't talk about it because everyone here is very passionate about the industry and future where demand doesn't increase is just not that interesting because it means we can all go home we did not fulfill our goals of crypto being a solution to Global problems so hard step we're not even going to talk about it the only way to cater this demand is by creating more more supply for and that Supply can take many forms uh there might there will be a spectrum of of chains you know on one end of the spectrum I don't think certainly I don't and I'm sure most people here don't feel this way but one end of that Spectrum can look like very heterogeneous environments in very monolithic chains it might take the form of move chains like sui and Aptos and certainly the VM will be a part of it maybe the Solana the c-level VM will be a part of it and on the other edge of the spectrum we have a completely homogeneous environment it's all evm based and then we just have many many Roll-Ups stemming from this uh the singular Hub that Hub can be the cosmos uh the cosmos Hub with dsdk and IBC it could be ethereum with many ethereum based Roll-Ups personally I'm inclined to think they will end up somewhere closer to the hub version but not completely homogeneous so this is really what we talk about when it comes to the modular blockchain thesis a very large amount of Roll-Ups app chains and whatever you want to Define them stemming from from these hubs beyond that what does an acceptable future look like all of we're going to have all these chains where they're going to have to be connected applications are going to have to be connected how do we prevent an experience like the one in that not so great story I just shared about a person trying to redeem trying to onboard in the crypto for the first time to redeem a damn pull-up uh apps are going to have to be connected but if apps are connected it's going to mean that chains are going to have to be connected how can we get there and before we get that let's talk a little bit about the errors in current history of interoperability so before before we got to where we are today which is we have some form of acceptable interoperability majority of it is very user-centric it's still users bridging assets it's still users dealing with infrastructure we had the first instantiation that was what we call here defined rule set interoperability one thing to note you see a timeline here the timeline is a bit misleading notice there are no end dates for any of these this is because all of these errors are playing out simultaneously so maybe the better title for this is not errors of interoperability rather modes of interoperability because these are all distinct in their form of operation but again all of them playing out at the same time and it's not clear that these will remain independent and not uh coagulate in some form or merge in some form as we move into the future so early on around 2015 2016 we have this notion of defined rule set interoperability this is most commonly observed in uh in Cosmos and IBC in polka dot and xcmp really what it is it's interoperably within a set structure and it's the same notion of horizontal um horizontal scaling where instead of saying like hey I'm going to have one computer that can process everything I'm going to have many little computers and they can have some logic about how they can share things between them and they will handle all the press the processing so the dominant form of expression here are certainly uh what you know uh app chains is the right way to Some people prefer news app chains I like referring to them as app chains that's the cosmos uh parlance and the polka dot World known as parachains some distinct differences between polka dot and uh Cosmos and IBC there but this is the common way that these are expressed and and each as we talk about each one of these modes we'll probably spend some time thinking about talking about how exactly it's working today and so today there are dozens of Cosmos chains IBC transaction volume is quite high now not every chain has a very active IBC channel to communicate amongst the other the other chains that it can communicate with right like if you look at the network graph of IBC connect the chains there's there's a lot of concentration there but there's still certainly a lot of activity similarly there are many um many more polka dot uh pair chains today more than you would expect it would have existed certainly eight years ago and this category of defined rules that ends are really while limiting while you are limited to interoperability only within that set structure and no strong form of communication with chains that don't fit that structure it's certainly an acceptable mode of doing things and one that's going to continue to see more activity as time progressed there was more active participation in crypto more types of users and we went into what uh I like to call acid acid-centric interoperability this was interoperably that focused on giving users the ability to move assets between between different blockchains where the structure of the underlying chain was far less important in fact it really is not it's only important for the operator of that interoperability service for the user's perspective it is entirely unimportant bridges are the most common expression of this right if you in some of these Bridges take the form of the canonical or native bridges that we see for Roll-Ups but certainly Bridges like Wormhole like synapse these are the dominant form of expression for this mode of interoperability and truly there's been billions that have been transacted via Bridges again in the form of asset Centric interoperability there's also been some issues um I want to say that as of today more at least five of the 10 largest security incidents in the history of uh of the industry all relate to the category of interoperability now is it because this category is inherently unsafe because this mode of asset Centric interoperability of the nearly unsafe that's debatable I'm certainly biased but I would argue that that's not the case and it's just that software with super high stakes and in adversarial environment as the permissionless internet computers that we work with it is it is very dangerous and so incidents are going to happen and finally there is this last mode of interoperability it's very very new there's not that much to say about it that isn't speculative and that is app Centric interoperability now if there's one thing that's very common between the two modes that we observe in the beginning at least in their operation until this point in time is where does the burden of interoperability lie there is an operator that provides the service but who performs the action in both of these it's the user it's the user having to move around an asset and you know if you wanna think about this in non-crypto terms having to move your funds from ethereum to optimism having to move your funds from ethereum to osmosis when you want to transact there that has a little bit of the feel of you know I really want to purchase something on Amazon today so how about I load my balance and I transferred from my bank account say with Bank of America and I load up some bucks into Amazon I get some Bezos bucks as my balance and that is quite strange right you probably wouldn't want to do that there are not that many instances where uh common internet products that we use ask us to behave in that way but this is where we are in crypto so the key difference between these two modes in apps app Centric interoperability is this notion of separating or moving the place of the burden of interoperability primarily it's about moving it from the user's hands to the application so that now we we bake in a layer of abstraction such that apps that are built in this way are now handling that transfer on your behalf similar to like we don't know the logic of the credit card or like the payment Network when we get charged on an e-commerce site like Amazon but it moves the funds on our behalf and that seems fine it certainly contributes to a simpler to a simpler user experience the dominant form of expression here is interchain application but what really is an endogene application simply put it's an application that should be through some means connected by some interoperability tool something that can communicate easily between the different internet computer computers that it exists on we believe that that'll be the dominant form of expression but it's still really early it's still not clear that this will be the case what does it even mean to be connected by some interoperability platform like what does that mean for us like from a blockchain perspective so for that now we're going to step in to talk a little bit about flavors of interoperability and by and large they come in two main ones you either get native verification or you get external verification what does native means native means you are leveraging the consensus mechanism of the two or more chains that are connecting and most importantly it means there are no more trust assumptions introduced into your action Beyond this idea of I'm already trusting you know ethereum which I'm used to dealing with and let's say in our example now you want to go to osmosis I'm trusting osmosis and I that's all I'm okay with and so really native verification is the ideal method because it only inherits the security risks of the two chains that are connected now why why refer to it as the ideal that kind of makes it sounds like it's not easily achievable well sure as it's not sorry for my language it is not uh easily attainable unfortunately it is by far the most expensive and most I mean really most resource costly uh method of verification in fact in systems that were not designed from the ground up to be able to replicate the consensus of external chains within them uh it yeah it is just incredibly expensive and for this reason it works the best and perhaps without external tools really the only place where it works well are within these defined rule set systems so insofar as IBC is concerned IBC gives you a form of native verification now should you be super discouraged not yet uh there are things that will help us reduce the costs of native verification that will allow native verification to be more feasible for environments that weren't built for Native verification and drivability um most promising along this front is uh certainly zero knowledge proofs as a way to replicate uh the consensus of the chain that you are connecting with the chain that you are or sorry to replicate the consensus of the two chains amongst each other in such a way that now having a interesting State transition is much more easily done without having to introduce a new trust assumption but unfortunately uh it's gonna take some time right there are great advancements you've seen stuff with uh with ZK bridge that recent paper and there are great investments in the field of Hardware acceleration for provers and so I'm quite optimistic that within the next few years we'll start seeing the inklings of zero knowledge proofs and increasing the prevalence of native verification interoperability something that is especially useful for the modular blockchain thesis and the future of ethereum scalability the other mode is external verification well you know easy words to say right if if the uh previous mode did not introduce any new trust assumption and this is the mode up you know aside from it or opposite from it well you guessed it it does introduce a new trust assumption so external verification means that in between the two chains that you're connecting now you bring in a third party and you rely on that third party if you spend a lot of time uh looking at or trying to solve the Oracle problem or just dealing with the Oracle problem you are intimately familiar with this in fact I don't feel uh incorrect in saying that external verification interoperability really is just another manifestation of the Oracle problem and solutions to it often take the form of either fully fledged Oracle Solutions or narrowly scope or narrowly defined Oracle systems external verification takes many forms there's many different configurations the easiest and most common looks like a multi-sig if you were unlucky enough to be involved in the Ronin Bridge incident you might be more intimately familiar with this configuration but many times it just takes on a multi-sick meaning that whatever state we're replicating between the chains the ultimate Arbiter of that is decided by a key it used we used to have actually interoperability systems that did not even use a multi-stick and were a single key thankfully we've Advanced far past that and that is no longer the case but that is a very very popular uh mode simply because it simplifies uh the operation but it obviously introduces predictable and obvious attack vectors another configuration is proof of stake where you can kind of enhance that multisig with some type of economic risk meaning where these third parties that are being introduced as external verifiers now have economic risk now they have a real reason to not lie about what they're doing but importantly unlike the proof of a stick that you're familiar with from ethereum where ultimately a validator cannot change your balance a validator can lie about a certain State transition but they act they cannot take away your balance um in Oracle systems you are totally beholden to what the external verifier says is the state of the world if the external verifier says that the sky is red the sky is red for all you know on that destination and this is why proof of stake systems um when it comes to interoperability they are completely different they are in many of those ways they are similar but when it and where it's most important when it comes to the key risks to the system they are quite different than what you're used to when it comes to proof of stake simply for blockchain consensus and another configuration uh one that I'd like to Hope is growing in popularity is the optimistic instruction again different from optimistic Roll-Ups but similar enough to where you can use it as a mental model you bake some amount of time a certain challenge window between when some State transition is initiated on chain a and that being replicated on chain B and within that time Watchers can observe the transition if they find anything malicious they can submit a fraud proof and prevent it and last there are configurations that will involve different forms of advanced cryptography these can involve threshold signatures multi-party compute and we unfortunately won't have enough time to get into those and there's who knows what other configurations can hold but I can tell you that there's going to be more of the so we've done a half decent job of covering the history of interoperability but now I told you we're going to talk a little bit about the modular blockchain thesis and how it relates to scaling uh crypto so look we all know with the modular box of these is we're going to have thousands of Roll-Ups and they're gonna fix scalability right and they have to obviously but what does it really mean what does it really mean for us to have all the state on ethereum and now we want to access it and we have all this demand and so we need to be able to access we need to be accessed block space in a more refined more targeted uh perhaps even less generalized mode such that we're not paying you know we're not paying 500 degree on 200 000 gas transactions on the regular uh as was not uncommon during periods of high demand as we saw in 2021 so what's it going to mean in practice right we're going to have all all these hubs all right we're going to have talked a bit about this a little bit earlier we're going to have ethereum as a main roll-up Hub and we're already seeing this materialized earlier in the day you heard uh Kelvin from optimism talk about Roll-Ups and optimism has done amazing work with the op stack coinbase already took it upon themselves to extend the op stack and launch their first you know really launch the first widespread roll-up built with the op stack this is just the first example of this but really do think we're gonna see dozens of these if not hundreds Celestia with its work on data availability Celestia roll-ups will become far and wide there are already several dozen Celestia Roll-Ups being built and those will seek data available either from Celestia you know maybe from eigenlayer Lord knows somewhere points rates we're gonna have lots of roll-ups whether it's all the app chains that already exist in the cosmos World these are not exactly role since they uh don't separate uh execution as you know as much as your standard rollup does but for our purposes certainly okay to think about them the key point to take away from us so we're going to have these hubs in the lowest level hubs will look like ethereum Celestia um you know Cosmos a slightly higher level Hub can look something like optimism in the op stack it can look something like bass it can look like perhaps Eclipse which is creating a toolkit to make Celestia Roll-Ups more widespread all of these hubs from them are going to emanate so many roll-ups but how do those Roll-Ups talk to each other how do they call cross between hubs right we talked about the future that is acceptable to us a scalable future needs to be a many roll-up future but how is that square away with the experience how are we going to avoid having so many infrastructure interactions whenever we want to move between these Roll-Ups and communicate between them and that's where interoperably comes in once again the modular blockchain thesis needs permissionless interoperability if everyone who wants to is going to be able to spin up a roll-up on demand if apps are going to be expressed as individual Roll-Ups if they're going to be generalized Roll-Ups becoming a hotbed for more accessible and cheaper computation how do we communicate between them are we really going to wait the seven day challenge period going back in between them obviously not no one seems to think that but the thing that gets not enough attention is inter roll-up interoperability and across interably across roll-up hubs into that end interoperably must be permissionless but it never is so I promised you will tell you a little bit about hyperlane so hyperlane is the first permissionless interoperability layer unlike other Solutions in the blockchain space it's not another blockchain to solve blockchain problems it's a network of networks a network between blockchains in that interchange Highway that we talked about before it's the path between them so it's our effort to create the safest fastest and most accessible way to Traverse the empty space between blockchains and most importantly as that space gets filled with new smaller chains that look like Roll-Ups look like app chains look like app specific Roll-Ups even anyone has the ability to deploy hyperlane on those and communicate with any other network that already has hyperlane on it so we're running out of time but high level how does it work hyperlane is deployed on any environment any environments that want to be communicated between them so if you if you use the op stack to create your roll up you Deploy hyperlane on it as soon as you're done creating the roller piece uh it now you put a mailbox on it that's what it means to deploy hyperlane then you or anyone can run a relay or the relayer once again totally permissionless role and contracts send messages to the mailbox this contract that does exactly what it sounds like and relayers sit in between different chains and different mailboxes observing the mailboxes and then they submit uh the Merkle tree that can sorry the Merkel root that contains your message and delivers it to its destination this is what we believe the best way to offer permissionless interoperability and truly the only way that interoperability can scale to a world of modular blockchains which we know is the best way to scale ethereum so it took a bit too long ran out of time but that in general is uh is a high level talk about this today if you'd like to learn more you can find me on Twitter at pale Nimbus um or you can reach out to hyperlane.xyz you can find us on our Discord we'd be happy to talk to you all about it and we will be sponsoring the scaling ethereum hackathon as we do with pretty much every youth Global event we are large supporters and big fans of each global uh so we'll be here for you amazing thanks John uh this was super informative and I feel like we got to do a part two here to actually dig into how hyperlink specifically can help here with all these things and uh suspending this has been amazing and uh a pretty awesome start to the summit so we get to uh really see what people can do with optimistic roll up ZK VMS and cross chain applications thank you so much now and I believe Mark too well we'll make it happen even if you don't want to you'll do a part two this is this is good stuff we need to have this information should be be out there and available for everybody to uh to mask