Bringing Web3 to Billions through Interoperability - TOKEN2049 Singapore 2023

all right all right great to see everyone today my name is Jacob Koh I'm from superscript we're an early stage fund based out of Singapore and New York uh investing into crypto and web 3 and we were seated by temasek but enough about me we're here today with a really stacked and exciting panel to talk about interoperability of course we've got Beyond Wagner the lead contributor to polka dot and CEO of parity Technologies Sunny the co-founder of Osmosis one of the largest tvl dexes on built on the cosmos SDK tarun the CEO of Gauntlet which does crypto simulation and jorgos the co-founder of axela now the title of this panel is bringing web 3 to billions through interoperability that is a mouthful do we guys do we really believe that interoperability is the thing that's going to onboard your mom's sister grandmother Uncle into web3 and maybe I'll go first the answer is no no I think what will bring web 3 to billions of people is amazing developers like sunny here building incredible applications that people find Value in where interoperability comes in is that we can enable those developers to build seamless user experiences into their applications I think it's consensus among probably everyone here that the future is going to be one of a hundred at least a hundred chains but all have substantial economic activity on top of them and in that future High throughput applications like osmosis here are gonna live on their own chain and the only way to make this future sustainable is by having seamless interoperability across chains so that the user an avalanche you know can buy an asset on osmosis without having to think about switching wallets or all of the UI complexities we see today right so while interoperability is not what's going to bring us to you know billions of users it is a necessary component and this is a core infrastructure on top of which is like all those application developers are going to build I could add to that a little bit um a lot of this what he just said resonates with me I think it's somewhat important to remember where this whole topic interoperability came from right like that I came came out like around 2 16 17 when uh Cosmos polycodile right came forward and said hey that's a different way to do to give developers tools to build useful applications in the world right beyond what was standard back then which was the ethereum layer one chain right Evie and um interoperability the need of our interoperability really came from the realization that like ethereum or the evm on one hand doesn't facilitate it doesn't give developers a tool with which they can build um optimized solutions for certain problems they encounter on one right so the VM was just very opinionated in what you can build with it and how you can bid on it there was one the other ones are some other limitations right for example wait we all share this one execution environment right and that means we are all subject to this free market and when crypto kitties jump through the roof the Demand right like every other application deployed on that execution environment is subject to the negative outcomes of um seeing so much demand in another application in the in essence by the femark spiking right so another applications users couldn't engage with it anymore because suddenly transaction cost a hundred dollars and so I think broad broad um you know a large part of our broad ecosystem right like came to the conclusion hey there's another way doing that right they're more optimized there isn't just putting everything on Smart contract layer but that meant you have um disconnected execution environments and when you end up there right like what is the whole premise of blockchain Technology right it's you have this ubiquitous computer where everything can seamlessly you know call another function interoperate and so forth right but as soon as you leave the evm as lean as you leave the one state machine you end up having that problem that these things need to be talking to each other and that's why the whole topic came up in the first so it's more for me it's more like a technical limitation that we encountered to serve a Glo a global skate such a computer got it um that makes sense so speaking of this interconnected systems and and you know there are different ways in which entrepreneur interoperability works right there are different Stacks we've got three of them represented here would you maybe we'll start with you sonny can you maybe frame for us the cosmos interoperability stack and we'll go through XLR and polka dot as well yeah um the cosmos interoperability back we've mostly focused we've built this one protocol that we called IBC it stands for inter blockchain communication and the goal for us was we took a lot of inspiration from the design of like the TCP IP stack in the internet and very much focus on like okay be very like specification based uh make a set of standards that will have many implementations on it and so you know we we built the first implementation on the cosmos SDK but there are other implementations of the IBC standard now as well including on substrate which is the development environment for uh primarily used in the polka dot ecosystem there's people building IBC implementations on evm on a bunch of different stacks and the idea here is by B one of the goals of IBC was to be this like as credibly neutral non-rent extractive system you know you always hear this claim that people are like oh imagine like tcpip had a token and imagine we could have invested in it back then it's like what they don't realize is the reason tcpip1 was because it was non-rent extractive it didn't have a token it was this credibly neutral thing that everyone was able to standardize around and that's the goal of IBC is to build that credibly neutral stack that anyone can Implement into their uh system got it and and you know the approach of Cosmos effectively you have the cosmos Hub which sits in the middle and the transactions are meant to sort of go through that right no actually so that's actually a common misconception that the way uh Cosmos IBC works is the causeless Hub is a little bit of a misnormal of a name it all packets are sent point to point between Every Chain so there's no single point of failure in the uh Cosmos Network got it um what about the actual Network urls I think uh we've done a lot of inspiration by the way right from Cosmos and polkadot and without them we couldn't be here so maybe let's go in chronological order and talk about polkadot first sure do you want to frame it for the audience Beyond yeah I would love to um all right um I think there are two parts to interoperability in the cosmos world right one is interoperability with on within that fairly integrated polka dot system and then from polka dot to other consensus system such as uh individual Cosmos chains right because they're all individually consensus systems or ethereum are led to roll ups and so forth so Lana whatever it is so let's speak about internally within polka dot first right I think my mental model for polka dot that's I think the most accurate one it's a multi-core execution system right so um you have um there's spell it in a network right that obviously had um you know skin in the game via the stake and uh it's simply it's basically a sharded system and each shot in that instance execution shot is one core right just like a CPU like if ubiquitous computer and like today this is how it works today each of these uh execution costs can be assigned or is assigned to a parrot chain right okay it's not too dissimilar to what you see in rollups just that I think it sits somewhere in between optimistic roads and ZK Roll-Ups and combines it almost the best of two words because you get um via validity proofs you get kind of like the security and guarantees of zika proofs but you still have the flexibility in terms of what you can develop with is it uh more decentralization of um of optimistic roll up so and now um so that's good right that's why we talk about shared security umbrella all of these chains that rent one of these least one one of these execution cars is secured and is being made interoperable because they polycodile provides to all of these chains that connect to it um what we call what is in other ecosystems Bridges trust minimized risk minimized Bridges got it so when you say shared security you're talking about the relay chain of polka dot which sits in the middle and it secures all of the para chains and basically that's a kind of bridging right it's an interoperability stack essentially exactly it does two things right it provides like hey you don't have to build your own validator network right you get the full Economic Security of the greater got it greater thing but also you're provided with a trade route yeah but the very crucial Step Beyond that is um that we have somewhat um going we are going away from for example this one core assigned to one power chain yep and going to more of a model that's much more keen to what you're used to from cloud computing right where you say like well um a core could actually like if you're a chain right you could acquire a core on the spot market right yeah for the next block or go acquire five cores and thereby you get elastic scale got it so two interoperability approaches somewhat different one point-to-point packet one a relay chain in the middle now how does actually play in all this your house yes so with acceler we're fortunate to you know learn from both polkadot and Cosmos and we were built at very different times right I think when Cosmos and polkadot were introduced there was only ethereum and they were trying to introduce this new ecosystem of change that were I think very legitimately like there was a future where everyone would bid on build on either Cosmos or polkadot or both right but there was nothing else and fast forward a few years when we started building axillary in 2020 uh I think it was the boom of the various layer once uh there was like Avalanche Solana a few of those so we have to think about how do we connect those heterogeneous ecosystems right so xcmp IBC these are optimal protocols if you can build the stack from the ground up but with acceler we did not have this ability because everyone is was building their own stocks everything was heterogeneous and at the time the way everyone was connected those networks because these are through centralized expenses or through centralized Bridges right and how do centralized Bridges work you have a custodian and no one really knows how they work on the back end right you trust the team you have a custodian you send them the asset you they send you the asset on the other side right so there's effectively a guy in the middle that just facilitates this communication but no one knows what this guy does right there's problems with that right like we saw on multi-chain for example multi-chain I'd say the biggest hacks in the space the Ronin hack the harmonic hack were all due to centralization right so the the easy answer to that is how do we decentralize this entity yep but you know and it's actually the same way that we decentralized blockchains right if you remember in the early days there were a bunch of proof of authority blockchains then I think Cosmos were the first to um introduce this notion of a proof of stake blockchain where you could have a permissionless decentralized validator sets so we took inspiration from that and we created actually as a blockchain with the sole purpose of creating of bootstrapping a decentralized validator set and today where we are with acceler is that every time you pass a message from one scene to any other chain right it is secured by 75 validators it's a fully decentralized permissionless validator set and if you look at the overlap of the validators of acceler with any other chain we connect there's very big overlap right so in a sense attacking the actual validator set is similar to attacking like the validator set of any other big thing right including but when we look at polka dot and Cosmos their you know their market caps are in the billions right so it's much harder to attack those networks since they're sitting in the middle of the chains I think axela's market cap is close to 135 mil right so what happens when the tvl that you guys to carry on your Bridge gets really large it's it's arguably easier to financially attack than Cosmos and polka dot great question right so there's two different attack vectors with those systems one is the financial attacks and the other one is corrupting the validator so it's compromising the the secret keys right and compromising the secret Keys which has been the main cause of hacks last year we counter with decentralization right so on that side we have similar level of security as those other protocols uh now on the financial incentive side the economic security uh today yes the liquid market cap of acceler is around like 1 135 140 I haven't checked the precise numbers the fully diluted valuation is around 400 million right so it's a little bit higher than that and all that stake can actually be staked right so it's almost everything mistake right now on the network and that it's actually a great topic because you have sunny as well we have a collaboration through which we're introducing Mass security right so the acceler network will soon also be secured by the economic security of Osmosis right and this is another like 400 million dollar Network and we can introduce security from other Cosmos stands right maybe even the the cosmos Hub at some point so we do have a roadmap to add to the economic security there's also some other interesting ideas we've been discussing with sunny There's an opportunity with eigen layer to also use ethereum stake so you can basically restake it earn some yield and secure the network and hopefully we can do that with Bitcoin as well and get this 500 billion dollars worth of support on the network got it awesome so three different interoperability approaches and stacks Cosmos polka dot axilla they all have a blockchain in the middle essentially helping to secure and validate the network of messages and values and transactions passing through but you know you guys are the ogs right polka dot and and uh Cosmos you've gotten your white paper route in 2016. your main Nets went live 20 20 21 respectively uh but there's a new narrative that's sort of Shifting around here right we're seeing the dominance of ethereum in many ways many Roll-Ups op stack you know there's arbitrum there's ZK sync all of these Roll-Ups and they are all EX sorry settling on ethereum layer one so in many ways if layer 1 has kind of like almost become like a relay chain itself tarun how did we get here yeah I mean I think an interesting question right at least historically if you look at 2016 there was certainly this view that hey maybe you have to make the protocol itself sort of shard in different ways maybe you have to actually have some homogeneity over control of like the interfaces between these different things and I think ethereum basically just took a longer route to do this because a it first had to transition to proof of stake which you know happened last year and that was a quite big deal but also B I think ethereum doesn't have the bootstrapping problem it has more of the like how do we ensure that our existing applications are able to get all the support all of their needs and what that led to is sort of this very heterogeneous execution environment heterogeneous execution guarantee system um which in a lot of ways you know if you look back at it it's sort of converging to what polka dot is I think once restaking is used to secure ethereum Roll-Ups you will effectively have created something that looks very similar to the polka dot model except there's no auctions right um but I think the the real reason we kind of ended up here is that the ethereum scaling roadmap took a long uh time to kind of fully flesh out but I also got to learn from the mistakes and successes of other networks and you kind of see that in each of the different Roll-Ups like they each take a different view to towards that but I think one other thing that I would say that is sort of interesting philosophically is ethereum doesn't really have a first class principle of owning a notion of interoperability it doesn't really have some notion of like hey arbitrum optimism and ZK sync need to conform to some particular standard now the irony is it will move that in that direction so some of the the ways of optimizing handling multiple execution environments whether they're sort of things like dank sharding will allow single blobs to be used in multiple Roll-Ups and then and the moment you do that you're implicitly creating a communication mechanism so you could argue that polkadon Cosmos took the first principles approach to interoperability like yeah it's a core tenet it's a philosophical uh stratagem that has to be built in everywhere whereas for ethereum it was much more the opposite it was how do we have this emergent interoperability and those are philosophically different things that sort have seemed to converge to to a very small set of outcomes I remember a couple of years ago when you know in the bull market uh people were talking about Bitcoin and ethereum and then they started talking about Cosmos and polkadot it's the internet of blockchains right and that was such a great narrative you guys went out there and built it it's almost like you were too you know too early relative to what we're seeing with the adoption curves um but that is a challenge right because right now we're seeing like a billion different uh startups start up to make it really easy to build rollups and ethereum right uh roll kit Sovereign lab stacker Saga Eclipse op stack slash Dimension fashion Suave Eclipse conduit Caldera that's not even half of them right so how do you guys compete polka dot Cosmos right um I think there's a lot of people building infrastructure uh a lot of block space and I think how we compete is actually just focused on building the products that people want to use um and I think like you know I think Cosmos has done a good job at experimenting on the types of products you know the largest stable coin in the world was built on Cosmos doesn't exist anymore but you know we had this culture of experimentation which is good um at the end of the day today most uh at the end of the most of this stuff is mostly used for like building trading infrastructure and perps are the like primary volume drivers of trading infrastructure and if you look at the everyone who's building order book based perps dexes they're basically all using the causos SDK right whether it's dydx injective Nibiru hyper liquid not cost Escape but tender mint like as our goal is to build the best stack for the apps that actually matter and that's how uh we don't need hundreds of thousands of infrastructure providers we need like 10 apps that will like bring 100 million users each and that's how we get to a billion as an investor I fully agree with you there's two almost too much infrastructure out there but it's great that you guys built it um I think you said it there though because to build a great app if I have to deal with all of you know bridging my funds from this to that within the app that's not going to work so how did you do that with osmosis because osmosis is kind of built interoperability with interoperability in mind right yeah so we do a number of different things right so osmosis believes heavily in like vertical integration of like bit you know we think that like the best products are very vertically integrated you look you know this is how Apple makes some of the best products in the world is by like you know the OS knows like it's very deeply optimized for the trip and that's why we decided to build an app chain um but at the same time we know when to Outsource where like right and so we were like so you could look at something like uh Thor chain right one of the other like one of my favorite Cosmos applications but they decided to build both like they have their own decks and they built their own bridging protocol um and but you know their bridging protocol has gotten hacked uh twice actually right and like you know they've done a really good job of recovering from that um but we saw that and we're like uh that Building Bridges sounds like a lot of work and that's not what we want to focus on and so what we do is we go find Partners like axillar who create like okay specialization is also important and we can like Outsource okay we don't want to build the bridging ourselves we're going to work with like amazing teams to build the parts that so we can specialize and focus on what we we're good at yeah so just cycling back to that Beyond you know how does you know as a league contribute to polka dot how does polkadot compete right build a better product build an irresistible product that's that's what needs to be done I believe um particularly if we talk about like uh these systems that are there to serve at a global scale right yep and so um I Liberty walk you through right like this aspect of like how many transactions can you do globally across the whole system that's counting us back right get to a point where you have 250 cars maybe 500 cores right that can eat come about like a thousand transactions per second that's simplified right that's one thing the other thing you need extremely good composability between these costs between the things that run on these costs and I think um we I think there's consensus uh generally in the ecosystem that uh Atomic composability is the best of composability you can get right ideally you want Atomic composability right that's very difficult right right but I think that's where polkadot really has an incredible opportunity due to its architecture because like what we are all what we're all tending towards to what we're all going to get eventually right are these trust minimized Bridges which are the trade routes right how one can send a message to another and the language we use in Pocket Edition extremely General you can send basically any matches and pack any information in that message instruction but the other key aspect is how do you know this message will be Faithfully executed on the other side and that's what we call enforced trade agreements right and I think that that's what polka does shared secure impeller and architecture really has an incredible Edge yeah where we can build what we call a court which are enforced trade agreements between these different charts yeah um and that has some of the elements and then again I think for us all it's really important to find The Sweet Spot between sweet spot in terms of trade-off between modularity and integration so to Sonny's point right you can talk about it on integration application layer right like what's the business logic but you can also talk about it on a lower down the stack right you've got to talk Celestia versus Cosmos versus polkadon and so forth okay it's starting to get complicated now right like and I think this is the problem with interoperability as well there are so many different ways that you could cut it different Stacks different approaches it's so difficult for devs to integrate um you know yogas you're kind of like Switzerland here right um you know you work with everyone you have your own protocol that can work with multiple uh blockchains but let's say ethereum Cosmos polka dot right if you had to prioritize ordering the integration of them how would you do that okay so first of all I think it's actually quite easy though I think you should have said Mary kill sorry oh okay ethereum polkadot and Cosmos Mary kill which one what would you do oh my God uh you should moderate this Tyrone seriously like I mean obviously America has to be has to be ethereum I'm not gonna pick between the other two uh I love both we always pick that to Cosmos as we talked about maybe I'll talk a little bit from the perspective of a builder right of course motion why we actually use the cosmos SDK to decentralized axillary validator set right so uh we felt from the very early days that the cosmos SDK was the most flexible stock and we needed to do some low level optimizations in terms of threshold cryptography so our acceler validators could collectively sign a message and at the time we felt like the cosmos SDK was the most mature stack right and this is the reason why everyone comes to Cosmos like it's the best technology if you want to build those low latency defy applications that Sani was talking about right like dydx uh better same same network osmosis there's no other option that Cosmos and uh just to bring up something negative as well though I feel like this is this is the only reason today right uh I haven't met too many people there's that go to Cosmos because they're gonna get ecosystem support right like it's too decentralized which is great because the most much more most mature stack but if you go on building Cosmos you're kind of on your own still right and that's something I feel like the ecosystem has to improve on and I feel like that's something that's also like a really nice um thing that we're doing with acceler right uh sanitized a little bit earlier on it uh to grow interoperability we need to get these five ten first applications that are going to build those core Crossing Primitives like crosstain Landing Crossing swaps Crossing marketplaces and the same way like those Primitives led to the device number we're going to see an interoperability somewhere hopefully next year one of those primitive loans and other developers can compose on top right and with axillary we're super focused on making sure like anyone who's building with us has full access to our go to market team we can help with introductions to virtually everyone in the space I guess cross chain really touches everything and we're in touch with most teams in the space we can help bring developers in tasks like with exchanges with other validators market makers really everything that took us from the actual side like one or two years when we first joined the cosmos ecosystem right so um I feel like yeah as a part of the cosmos ecosystem as well like we have to be honest and and like look at you know the things that are not working well I think again decentralization amazing most robust attack but there's more to do like in how like we take care of Builders and like we we help them like grow their ecosystem and projects well you're really living up to your name of being the Switzerland of this panel because that was the most diplomatic answer I've heard all day today uh okay so to ruin a question for you right do we even really need interoperability it is so complicated you know some of the developers that we talked to when we speak to them they say like we just want simple things we want atomicity in our transactions we want our smart contract on our chain to be able to call another smart contract now we've got to integrate multiple Stacks we've got to work out where to go like maybe the answer is a monolithic blockchain maybe the answer is Solana who knows right like what do you think yeah so I mean I think the the positive case for for Solana is that they are a hundred percent only a believer in atomicity and like that's their philosophical identity right there's like literally nothing else that I would say is the right I mean they might get offended some some people might get offended by Museum but it is a very very uh distinct Choice um I think there's always going to be a sense in which there will be multiple environments and the the things to look back at for the history of that are sort of looking at how Linux evolved and looking at how cloud computing evolved cloud computing is is is sort of a an ever-changing kind of compute need type of system in the beginning of cloud computing a lot of it was like well I just need to run a couple redis nodes or I have a web server and I need to scale it up or down depending on how many users I get but then if you look at data centers today the majority of actual energy consumption is actually going towards high performance Computing which was like not true before and that's because they're particular applications that happen to find better compute throughput and utilization via different means so in the same way you will have high bandwidth so like many chains type of behaviors and low latency behaviors that you know you want to have kind of like a single machine that's very Atomic yeah I think at the end of the day though if you look at Linux Linux was probably the most successful open source development environment that made a transition from a purely single core Atomic architecture to a multi-core architecture that 99 of the developers who are you know that you can talk to now who've probably you know learned to code in the last five to ten years probably don't even know that under the covers the kernel is doing so much to actually schedule transactions across multiple cores to figure out where memory should be paged to figure out all these low level details that you don't think about and because of that you have react because of that you have all these high level languages that you know someone can actually like learn over a weekend yep and in some sense that's what interoperability should look like where it's you don't even the people who are building applications don't even actually know yeah um but crypto hasn't even gotten to the point where the users don't even know they do that right like I'll give you a thought experiment this is like something I try to you know there's two kind of thought experiments that I think are worth testing as a way of like how far we are ux wise the first is suppose I have usdc on three chains on Solana on eth and say on optimism and I want to do a transaction that says move all my usdc to East right now that is a pain in the ass you have to sign three approves on three chains you have to go make Bridge transactions and then you have to go wait that whole orchestration effort for the average user is ridiculous and it's sort of how parallel programming was in the 90s which is why it you know Linux 2.6 was such a big advancement will there be an interoperability standard that emerges or an approach that emerges that we'll be able to do what you just said because right now crypto is so tribal whether it's blockchains whether it's interoperability Stacks I mean intraoperability is about collaboration does that even make sense right yeah I mean in in sometimes I suspect it will be somewhat application specific right like the circle case for usdc I use it as an example because Circle does have this sort of centralized bridge where they can actually burn on one chain and mint on another if it has native uscc and they can do that as a single transaction for you yeah and I actually think the wallets especially some of the newer MPC wallets will effectively be doing orchestration for users to do something like that in a single transaction but my point is that is hidden from the user it's not it's not like sitting in their face that they're like clicking metamask 20 times yeah maybe it's axelot who fill that role for people yeah I mean uh I guess one question we get a lot right with acceler is what's the killer application of cross chain and to me this question makes no sense you know Crossing is just about user experience as Tyrone just explained any application you can build on a single same yeah it's exactly the same applications you build Crossing but it's much harder right because you have this fragmentation so crosstain and what we're focusing on with axillary is building the infrastructure that will make it as easy as possible for developers to build this seamless user experiences and if we're looking to onboard the billion users to the space it's not going to happen in the future where you have to do like all those multiple approvals know that you have your asset on Solana the user shouldn't even know that Solana exists right you should have an application on your phone use your asset there maybe on the back end the assets on Solana the application is on ethereum and you're routing it through osmosis for whatever reason right the users know none of that and towards that future interoperability is the infrastructure that we lead us there yeah so you know right now ethereum if you if you believe the stats out there they've got close to 2 000 developers I think polka dot around 600 Cosmos around 500 right I guess the last couple of questions I had was for yourselves Beyond and signing how do we get more people using the polka dot and Cosmos interoperability Stacks right because you're there but being the Godfathers of bridging would love to see more I'm gonna I'm gonna have a seven so um it's a whistling polka dot right like the parrot chicks how do we make sure they they make use of the power they're given to talk to each other right ixia um you're going to consider that the delivery extend and the ability to even do that is quite new right even power chains are one and a half years old like pair of jeans were only able to deploy one and a half years ago exactly and but one and a half years is like 15 years in crypto land correct but not the interoperability step so today we have roughly across polka that like there are 800 application the projects that are being developed of which ADR change that are connected to kusama polkadot right and they can now use that power to talk to each other they give you three examples to make it a bit more relatable right what this is being used for so we have one example is Oak network is a default protocol works with mangada to have essentially Auto compounding right where users one can use a functionality another one and this is all integrated they can relatively easy actually do that auto compound it another one is killed which is a did Network yep um that partner with Deloitte deliver also verifiable credential around here CML and polymac which is another Network in polka dot that is essentially a compliant fundraising platform right Jesus these credentials from another chance via xcm in order to be compliant right um now how do we do it how do we get more folks right um well there are already tons of people right like developing building art very active development but I think his education make it much easier to understand make these developer tools so much more ergonomic and smooth working like orders of magnitude right but it's all the time and time in question like we it took time to build a system that's as powerful as what we've built and that takes time to build these Primitives and now it's really the time to make it easy and educate the world around and help to understand or to coming from the evm the ethereum development mindset what you can do with them all right Sunny what's the exciting development that you think will get more people building on the cosmos stack yeah um I think with terun said on like you know ethereum's ecosystem was able to learn from like the mistakes of uh Cosmos and polkadot on the interoperability side I think actually we have the opportunity to learn from the evm ecosystem on more of the application development side where you know we have a smart Contracting system called cosmosum but it's way the cosmos SDK because it has less developers on it we actually have some more flexibility to be nimble and like develop things differently than the way the evm works so you know I've personally been spending a lot of my time on like a lot of account abstraction stuff but like baking it more natively into the causes SDK in a way that's like is just not really possible in the evm ecosystem right now or um a lot I think a lot of the stuff on the Privacy side like my my co-founder Deb at our conference two months ago he gave a talk about like hey we can modify the cosmolism compiler that like you can add like one or two lines of code and you can turn your contract into having like private State and like I think taking the how the cosmos SDK and cosmosm as a development stack is going to win is to provide features and things that people can only do on the causes SDK style and this is like for example one of the reasons why dydx chose to build on the causes SDK was they wanted to have more fine-grained control over how transactions get ordered in their blocks right they wanted to figure out you know if you're building a purpose decks you they want to figure out how do you make cancellations free right they need fine-grained control over gas markets on transaction ordering and things like this so how cosmosuke is going to win is uh building functions that people can only do on the cause of SDK and nowhere else got it so I mean just to recap right part of it is education part of it is developers building good apps that people want to use part of it is bridging infrastructure SDK is becoming so simple and easy to use they kind of Fade Into the background you know that's part of our investment thesis as well when we invest in in web 3. um thank you so much guys this has been an amazing panel ladies and gentlemen these guys are the gods of bridging they're in the arena bridging stuff a big hand for them [Applause]