{"id":"PETPmAAZQfk","title":"Across’s take on CAKE: Defining a Settlement Layer for X-chain Intents","description":"Speaker: Hart Lambur, Across\n\nWe Came, We Saw, We Abstracted!\n\nNEAR, in partnership with Frontier Research and Agoric, hosted Chain Abstraction Day, a massive event that attracted over 1,200 attendees. \n\nThis gathering brought together the brightest minds in blockchain, including industry leaders like Sreeram Kannan from EigenLayer and Adrian Brink from Anoma. \n\nA cornerstone of the event was unveiling the CAKE framework (Chain Abstraction Key Elements). This framework provides a foundational understanding of chain abstraction and paves the way for further exploration. Discussions dived into cutting-edge topics like threshold signatures, multi-chain settlements, and abstracted security. \n\nRecap of NEAR at Eth Denver: https://bit.ly/NEARatEthDenver\n\nJoin NEAR's community:\nWebsite: https://near.org/\nReddit:   / nearprotocol  \nDiscord: https://near.chat/\nMedium:   / nearprotocol  \nBlog: https://near.org/blog/\nTwitter:   / nearprotocol  \nGitHub: https://github.com/near  https://github.com/nearprotocol\nDev Docs: https://docs.near.org/\n#Blockchain #NEAR #nearprotocol","channel_name":"NEAR Protocol","published_at":"2024-04-11T00:00:00.000Z","thumbnail_url":"https://img.youtube.com/vi/PETPmAAZQfk/hqdefault.jpg","transcript_text":"uh awesome guys all right so if you want to hear about uh intense uh something that I think everyone's going to be talking about a lot today uh come listen and um since the frontier guys are big on cake uh we're going to be talking about a cross's take on cake cuz in many ways we are similar um to them in their way they think about in all right so what do we want uh what we want is pretty easy right we want fast cheap and secure interop that is so seamless you don't even know that you're talking with blockchains um you just want it to work and uh there's this sort of concept that people like talking about that interop should be like tcpip it should be like this bass layer where like messages just flow between blockchains easy no problem uh so simple so so like water you really know what's happening um and I don't think I agree with this um because in this model of uh interop like tcpip we're sort of distilling the problem where um bridging is just like sending a message the formatting here changed a little bit but um it's just like we're going to send a message from blockchain A to blockchain B and it's going to work but we want to send this message uh quickly cheaply and securely and this is like basically impossible to do we might do it in 5 years with super fast CK maybe that happens maybe we get super fast super cheap ZK interop Tech but otherwise sending a message from A to B is not a simple thing to do um and let's again to do it securely too security there's trade-offs here if you want to make something secure it's usually going to cost more or take some time and so you just kind of think through these things you're like this is going to be a really hard thing to do and so my answer is just stop sending messages that's that's the problem um we're trying to do something that isn't very easy to do and that we should just avoid and um the observation here this actually follows something onid said in his talk earlier too um um tcpip is moving bytes we're just moving data around this network um but blockchains are mostly and I'll come back to this mostly they're mostly moving value and there's a lot you can do when you're moving value and this has to do with like fun ability and an made the same uh presentation one of his slides or same point in one of his slides um he he estimated like 95 to 99% of transactions that we care about uh are about moving value and not build BS and if that's true maybe we can solve some of these interon problems and solve some of the ux issues by focusing on moving value and um the reason is that value is fungible uh if uh if I'm moving money I don't really care about how the money gets there I just care about the money getting there um it it's it's it's fungible and this is where my mental model for my intense work comes from is you are moving value don't care how it gets there you just want to specify the output uh don't care about the path don't care about the execution just want the most value at the end of the day possible and if we're sending value and it's fungible we can do fun things um where we don't actually need to send a message to fill a user on an inup transaction so stop sending a message from blockchain A to blockchain B and instead have a network of solvers uh we are the solver layer of the cake right now so instead of a network of solvers that are observing what's happening on blockchain a so user deposit transactions and are then filling a user on blockchain B with their own capital and uh suddenly we get rid of this messaging problem at least a real time messaging problem and so this is my take on why crossan intents are the the the future of interop here with this intent concept solvers can fill users with their own capital and they can fill them quickly and cheaply and uh then the solvers get repaid later and it works it sort of solves our dilemma that we started with so um going a little more detail because I it was pretty high level about just how the solar thing works I look at and I always have looked at kind of the across system as something of a cake too um and follows very nicely with what the frontier guys are talking about we got three layers we got the intent layer where the user is specifying what they want to do and this might come from a front end like the across bridge across. two it might come from a bridge aggregator it's going to come from RFQ systems increasingly I think it's going to come from applications themselves that are controlling user orderflow um we then have a solver layer and the solver layer is competing to fill the intent specified at the uh at the user layer and then lastly we have a settlement layer and what's actually going to happen is the Sol the user funds are going to get escrowed in the settlement layer the solver is going to fill the intent and then at the at the end after we verify the intent was fulfilled uh the solver gets paid back from the sell layer that's how I think the future of interop is going to work and now we'll talk about like why so the advantage here intense can be really really fast um this is maybe impossible for you guys to see on your screen um but this is AC cross's median fill time from L1 to L2 from L2 to L1 from L2 to L2 and this is Stargate and this is ccip and our medium fill times on L2 to L2 are 7 Seconds very go 12 seconds for L2 and L1 23 for L1 and ler and we are uh one to two orders of magnitude faster than these other systems and how is that how is that possible um it's possible because we have these solvers competing on speed and they can actually take finality risk they can trust centralized sequencers which is why the L2 times are so fast here they can trust centralized sequencers and just fill users as soon as they they get a preon back um and even on L1 they can make probalistic decisions where there hasn't been a a three block block reorg in a very long time so two blocks good enough I can start with clal users um and this is something that you can't do if you don't have these offchain actors taking finality risk you can't have a protocol uh chain link cctp absolutely cannot take finality risk it's not worth it so this is why intents are faster um intents can also be really secure and the idea here is that this architecture we have separates the NRL problem into two parts there's the Urgent part where users want their fast fill and there's the complex part where we like communicate between blockchains and understand whether the intent was fulfilled and what's going on there and by separating these two we can we got a lot more options for how we do cross chain messaging the Urgent part we talked about uh the solver just fills the user with their own capital and then the complex part the uh the the messaging part we can verify using any number of cross chain messaging techniques we can use the canonical bridges that protocols WR themselves even 7day Sol Bridges possible just be a really long repayment time we could write ZK proofs um for proof Generation Um and because because maybe we only need to batch aggregate payments so we can do one CK proof for many many um intent verifications or we can do optimistic systems and that's actually how AC cross Works uh we optimistically verify batches of intents um and repay them every hour or two um and so we are able to solve for uh inop uh security without worrying about doing it quickly which opens up a lot more options for helping to do this securely and cheap the third Advantage um is just cost and um the reason why this intent architecture can be so damn cheap and again you're not going to really see these numbers here but it crosses L2 uh L2 to L2 L2 to L1 L1 to L2 across this uh cost here this is all in gas and transfer costs um over the last 90 days um across can be really cheap because of two things we do um instead of verifying each intent and having costs associated messaging costs associated with verifying each intent we batch them together in mundles which we can do because the solvers are already making the loan um so these batches allow us to verify hundreds or thousands or tens of thousands of intents in one transaction and then further we can do off of verification which is slow but it's really cheap and quite secure add those two things together together and we have really cheap costs the tradeoff is that uh solers are making short-term loans so making loans of capital we have a cost of capital to the university gets but um if we do the math this tradeoff makes a ton of sense so here's my theoretical example and I'll kind of walk you through it Pretend We're bridging $50 million of volume a day and we're going to do do this over 25,000 transactions um so okay so I got a $5,000 outage transaction size 25,000 transactions what does this cost me let's assume that we we have a default cost the user has to make a deposit and then on the destination chain there has to be a fill so there is going to be some gas costs that's unavoidable of the the user deposit transaction and the destination chain fill transaction but what are our verification costs on the a across side with this intent Bridge we need to make loans for about an hour and if you assume a 10% annualized interest rate um that $50 million of loans for an hour is $500 small amount of money it's kind of tiny and then we've got to send a certain number of conical messages in these batches to verify these repayments and I'm estimating these conical messages at something relatively High put a higher number if you want I'm going to say these are expensive all one messages but we're spending like under three grand a day in our verification process and that's advertised over 25,000 transactions if I send messages for each of these transactions like something like what Stargate does I've got to send a message 25,000 times and even if these are like L2 based messaging is really cheap I'm I'm I'm being generous here and saying this cost an average of two bucks it's a lot more than that it's going to cost you 50 Grand and so this Tre off here where the cost of the loan is just not much relative to the advantage we have from this batching and we get faster cheap uh faster speeds and better security so um when you talk about scaling this across is this settlement layer that could scale to transfer a billion dollars a day but we're actually going to only send like 150 canonical messages because of this batchel um and if you had a non- inent based design you'd be sending something like 10 million messages which maybe that works but that sounds expensive to me um and this is my my reasoning why like this tcip model for interop I don't think is the right one I think the mental model is going to be more akin to financial model where you are aggregating these funable transactions and batch setling up so our vision and I think of over apologize um across at the intent layer we want to help build efficient RFQ and intent agregation systems uh one of the things that ankit's uh cake Frontier research piece talks about is solver price competition versus solver speed competition um I want to research with this ecosystem RQ systems that do both how do we how do we uh auction off intents to solve for both speed and cost um at the silver layer I really want to establish a standard for cross-chaining tents something that can get as many solvers and as big an ecosystem as possible providing offchain liquidity for this intent concept and I think a well-designed standard here is purely POS sum for our ecosystem and at the bottom layer this is where I think AC cross ultimately um really lies where AC cross is this settlement layer that escrow's user funds and releases them to solvers um when these intents and verified and our goal is to define a to build this cheap secure and super efficient settlement system for these cross un intents and so looks like a cake um which is the theme of the day and that is my talk guys so uh thanks","transcript_synced_at":"2026-11-03T06:04:03.249Z","segments":[{"start_seconds":9,"text":"uh awesome guys all right so if you want to hear about uh intense uh something that I think","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=9s"},{"start_seconds":17,"text":"everyone's going to be talking about a lot today uh come listen and um since the frontier guys are big on cake uh","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=17s"},{"start_seconds":25,"text":"we're going to be talking about a cross's take on cake cuz in many ways we are similar um to them in their way they","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=25s"},{"start_seconds":32,"text":"think about in all right so what do we want uh what","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=32s"},{"start_seconds":39,"text":"we want is pretty easy right we want fast cheap and secure interop that is so seamless you don't even know that you're","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=39s"},{"start_seconds":48,"text":"talking with blockchains um you just want it to work and uh there's this sort of concept","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=48s"},{"start_seconds":56,"text":"that people like talking about that interop should be like tcpip it should be like this bass layer where like","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=56s"},{"start_seconds":63,"text":"messages just flow between blockchains easy no problem uh so simple so so like","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=63s"},{"start_seconds":70,"text":"water you really know what's happening um and I don't think I agree with this um because in this model of uh interop","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=70s"},{"start_seconds":79,"text":"like tcpip we're sort of distilling the problem where um bridging is just like sending a message the formatting here","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=79s"},{"start_seconds":86,"text":"changed a little bit but um it's just like we're going to send a message from blockchain A to blockchain B and it's going to work but we want to send this","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=86s"},{"start_seconds":95,"text":"message uh quickly cheaply and securely and this is like basically impossible to","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=95s"},{"start_seconds":102,"text":"do we might do it in 5 years with super fast CK maybe that happens maybe we get super fast super cheap ZK interop Tech","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=102s"},{"start_seconds":110,"text":"but otherwise sending a message from A to B is not a simple thing to","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=110s"},{"start_seconds":116,"text":"do um and let's again to do it securely too security there's trade-offs here if you want to make","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=116s"},{"start_seconds":124,"text":"something secure it's usually going to cost more or take some time and so you just kind of think through these things you're like this is going to be a really hard thing to do and so my answer is","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=124s"},{"start_seconds":134,"text":"just stop sending messages that's that's the problem um we're trying to do something that isn't very easy to do and","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=134s"},{"start_seconds":141,"text":"that we should just avoid and um the observation here this actually follows something onid said in","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=141s"},{"start_seconds":148,"text":"his talk earlier too um um tcpip is moving bytes we're just moving data around this network um but blockchains","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=148s"},{"start_seconds":157,"text":"are mostly and I'll come back to this mostly they're mostly moving value and there's a lot you can do when you're","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=157s"},{"start_seconds":164,"text":"moving value and this has to do with like fun ability and an made the same uh presentation one of his slides or same","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=164s"},{"start_seconds":171,"text":"point in one of his slides um he he estimated like 95 to 99% of transactions that we care about uh are about moving","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=171s"},{"start_seconds":179,"text":"value and not build BS and if that's true maybe we can solve some of these interon problems and solve some of the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=179s"},{"start_seconds":187,"text":"ux issues by focusing on moving value and um the reason is that value is","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=187s"},{"start_seconds":194,"text":"fungible uh if uh if I'm moving money I don't really care about how the money gets there I just care about the money","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=194s"},{"start_seconds":201,"text":"getting there um it it's it's it's fungible and this is where my mental","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=201s"},{"start_seconds":207,"text":"model for my intense work comes from is you are moving value don't care how it gets there you just want to specify the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=207s"},{"start_seconds":216,"text":"output uh don't care about the path don't care about the execution just want the most value at the end of the day","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=216s"},{"start_seconds":223,"text":"possible and if we're sending value and it's fungible we can do fun things um where we don't actually need to send a","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=223s"},{"start_seconds":232,"text":"message to fill a user on an inup transaction so stop sending a message from blockchain A to blockchain B","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=232s"},{"start_seconds":240,"text":"and instead have a network of solvers uh we are the solver layer of the cake right now so instead of a network of solvers that are observing what's","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=240s"},{"start_seconds":249,"text":"happening on blockchain a so user deposit transactions and are then filling a user on blockchain B with their own","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=249s"},{"start_seconds":257,"text":"capital and uh suddenly we get rid of this messaging problem at least a real time messaging","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=257s"},{"start_seconds":264,"text":"problem and so this is my take on why crossan intents are the the the future of interop here with this intent concept","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=264s"},{"start_seconds":274,"text":"solvers can fill users with their own capital and they can fill them quickly and cheaply and uh then the solvers get","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=274s"},{"start_seconds":281,"text":"repaid later and it works it sort of solves our dilemma that we started with","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=281s"},{"start_seconds":289,"text":"so um going a little more detail because I it was pretty high level about just how the solar thing works I look at and","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=289s"},{"start_seconds":297,"text":"I always have looked at kind of the across system as something of a cake too um and follows very nicely with what the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=297s"},{"start_seconds":305,"text":"frontier guys are talking about we got three layers we got the intent layer where the user is specifying what they want to do and this might come from a front end like the across bridge across.","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=305s"},{"start_seconds":315,"text":"two it might come from a bridge aggregator it's going to come from RFQ systems increasingly I think it's going to come from applications themselves","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=315s"},{"start_seconds":322,"text":"that are controlling user orderflow um we then have a solver layer and the solver layer is competing to fill the intent specified at the uh at","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=322s"},{"start_seconds":332,"text":"the user layer and then lastly we have a settlement layer and what's actually going to happen is the Sol the user funds are going to get escrowed in the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=332s"},{"start_seconds":340,"text":"settlement layer the solver is going to fill the intent and then at the at the end after we verify the intent was","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=340s"},{"start_seconds":348,"text":"fulfilled uh the solver gets paid back from the sell layer that's how I think the future of interop is going to work","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=348s"},{"start_seconds":356,"text":"and now we'll talk about like why so the advantage here intense can be","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=356s"},{"start_seconds":362,"text":"really really fast um this is maybe impossible for you guys to see on your screen um but this is AC cross's median","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=362s"},{"start_seconds":370,"text":"fill time from L1 to L2 from L2 to L1 from L2 to L2 and this is Stargate and this is","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=370s"},{"start_seconds":378,"text":"ccip and our medium fill times on L2 to L2 are 7 Seconds very go 12 seconds for","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=378s"},{"start_seconds":385,"text":"L2 and L1 23 for L1 and ler and we are uh one to two orders of magnitude faster","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=385s"},{"start_seconds":391,"text":"than these other systems and how is that how is that possible um it's","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=391s"},{"start_seconds":399,"text":"possible because we have these solvers competing on speed and they can actually take finality risk they can trust","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=399s"},{"start_seconds":408,"text":"centralized sequencers which is why the L2 times are so fast here they can trust centralized sequencers and just fill users as soon as they they get a preon","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=408s"},{"start_seconds":417,"text":"back um and even on L1 they can make probalistic decisions where there hasn't been a a three block block reorg in a","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=417s"},{"start_seconds":425,"text":"very long time so two blocks good enough I can start with clal users um and this is something that you","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=425s"},{"start_seconds":432,"text":"can't do if you don't have these offchain actors taking finality risk you can't have a protocol uh chain link cctp","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=432s"},{"start_seconds":440,"text":"absolutely cannot take finality risk it's not worth it so this is why intents are faster um intents can also be really","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=440s"},{"start_seconds":448,"text":"secure and the idea here is that this architecture we have separates the NRL problem into two parts there's the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=448s"},{"start_seconds":456,"text":"Urgent part where users want their fast fill and there's the complex part where we like communicate between blockchains and understand whether the intent was","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=456s"},{"start_seconds":465,"text":"fulfilled and what's going on there and by separating these two we can we got a lot more options for how we do cross","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=465s"},{"start_seconds":473,"text":"chain messaging the Urgent part we talked about uh the solver just fills the user with their own capital and then the complex part the uh the the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=473s"},{"start_seconds":481,"text":"messaging part we can verify using any number of cross chain messaging techniques we can use the canonical","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=481s"},{"start_seconds":488,"text":"bridges that protocols WR themselves even 7day Sol Bridges possible just be a really long repayment time we could","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=488s"},{"start_seconds":495,"text":"write ZK proofs um for proof Generation Um and because because maybe we only","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=495s"},{"start_seconds":502,"text":"need to batch aggregate payments so we can do one CK proof for many many um intent verifications or we can do","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=502s"},{"start_seconds":510,"text":"optimistic systems and that's actually how AC cross Works uh we optimistically verify batches of intents um and repay","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=510s"},{"start_seconds":519,"text":"them every hour or two um and so we are able to solve for","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=519s"},{"start_seconds":527,"text":"uh inop uh security without worrying about doing it quickly which opens up a lot more","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=527s"},{"start_seconds":535,"text":"options for helping to do this securely and cheap the third Advantage um is just","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=535s"},{"start_seconds":542,"text":"cost and um the reason why this intent architecture can be so damn cheap and again you're not going to really see","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=542s"},{"start_seconds":551,"text":"these numbers here but it crosses L2 uh L2 to L2 L2 to L1 L1 to L2 across","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=551s"},{"start_seconds":559,"text":"this uh cost here this is all in gas and transfer costs um over the last 90 days","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=559s"},{"start_seconds":566,"text":"um across can be really cheap because of two things we do um instead of verifying each intent and having costs associated","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=566s"},{"start_seconds":575,"text":"messaging costs associated with verifying each intent we batch them together in mundles which we can do because the solvers are already making","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=575s"},{"start_seconds":583,"text":"the loan um so these batches allow us to verify hundreds or thousands or tens of thousands of intents in one","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=583s"},{"start_seconds":591,"text":"transaction and then further we can do off of verification which is slow but it's really cheap and quite secure add","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=591s"},{"start_seconds":598,"text":"those two things together together and we have really cheap costs the tradeoff is that uh solers are making short-term","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=598s"},{"start_seconds":606,"text":"loans so making loans of capital we have a cost of capital to the university","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=606s"},{"start_seconds":615,"text":"gets but um if we do the math this tradeoff makes a ton of sense so","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=615s"},{"start_seconds":622,"text":"here's my theoretical example and I'll kind of walk you through it Pretend We're bridging $50 million of volume a day and we're going to do do this over 25,000","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=622s"},{"start_seconds":631,"text":"transactions um so okay so I got a $5,000 outage transaction size 25,000 transactions what does this cost me","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=631s"},{"start_seconds":639,"text":"let's assume that we we have a default cost the user has to make a deposit and then on the destination chain there has to be a fill so there is going to be","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=639s"},{"start_seconds":648,"text":"some gas costs that's unavoidable of the the user deposit transaction and the destination chain fill transaction but","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=648s"},{"start_seconds":655,"text":"what are our verification costs on the a across side with this intent Bridge we need to make loans for","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=655s"},{"start_seconds":662,"text":"about an hour and if you assume a 10% annualized interest rate um that $50 million of loans for an hour is $500","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=662s"},{"start_seconds":671,"text":"small amount of money it's kind of tiny and then we've got to send a certain number of conical messages in these batches to verify these","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=671s"},{"start_seconds":679,"text":"repayments and I'm estimating these conical messages at something relatively High put a higher number if you want I'm going to say these are expensive all one messages but we're spending like under","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=679s"},{"start_seconds":688,"text":"three grand a day in our verification process and that's advertised over 25,000 transactions if I send messages for each","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=688s"},{"start_seconds":697,"text":"of these transactions like something like what Stargate does I've got to send a message 25,000 times and even if these","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=697s"},{"start_seconds":705,"text":"are like L2 based messaging is really cheap I'm I'm I'm being generous here and saying this cost an average of two bucks it's a lot more than that it's","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=705s"},{"start_seconds":713,"text":"going to cost you 50 Grand and so this Tre off here where the cost of the loan is just not much relative to the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=713s"},{"start_seconds":722,"text":"advantage we have from this batching and we get faster cheap uh faster speeds and better security","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=722s"},{"start_seconds":729,"text":"so um when you talk about scaling this across is this settlement layer that could scale to transfer a billion","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=729s"},{"start_seconds":738,"text":"dollars a day but we're actually going to only send like 150 canonical messages because of this batchel um and if you","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=738s"},{"start_seconds":746,"text":"had a non- inent based design you'd be sending something like 10 million messages which maybe that works but that sounds expensive to","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=746s"},{"start_seconds":755,"text":"me um and this is my my reasoning why like this tcip model for interop I don't think is the right one I think the","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=755s"},{"start_seconds":763,"text":"mental model is going to be more akin to financial model where you are aggregating these funable transactions and batch setling up so our vision and I","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=763s"},{"start_seconds":772,"text":"think of over apologize um across at the intent layer we want to help build efficient RFQ and","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=772s"},{"start_seconds":781,"text":"intent agregation systems uh one of the things that ankit's uh cake Frontier research piece talks about is solver","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=781s"},{"start_seconds":788,"text":"price competition versus solver speed competition um I want to research with this ecosystem RQ systems that do both","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=788s"},{"start_seconds":796,"text":"how do we how do we uh auction off intents to solve for both speed and cost","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=796s"},{"start_seconds":805,"text":"um at the silver layer I really want to establish a standard for cross-chaining tents something that can get as many","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=805s"},{"start_seconds":812,"text":"solvers and as big an ecosystem as possible providing offchain liquidity for this intent concept and I think a","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=812s"},{"start_seconds":820,"text":"well-designed standard here is purely POS sum for our ecosystem and at the bottom layer this","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=820s"},{"start_seconds":827,"text":"is where I think AC cross ultimately um really lies where AC cross is this settlement layer that escrow's user","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=827s"},{"start_seconds":834,"text":"funds and releases them to solvers um when these intents and verified and our goal is to define a to build","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=834s"},{"start_seconds":842,"text":"this cheap secure and super efficient settlement system for these cross un intents and so looks like a cake um","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=842s"},{"start_seconds":851,"text":"which is the theme of the day and that is my talk guys so uh","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=851s"},{"start_seconds":863,"text":"thanks","url":"https://www.youtube.com/watch?v=PETPmAAZQfk&t=863s"}],"youtube_url":"https://youtube.com/watch?v=PETPmAAZQfk","url":"https://near-pedia.xyz/videos/PETPmAAZQfk"}