Biconomy’s “Supertransactions” are changing the way we use crypto

Transcript

there are decentralized apps and decentralized infrastructure the interim is like centralized apps and decentralized infrastructure and in the future it will be decentralized STS and decentralized infrastructure right and so that that's really how I see you know where things are going my vision here is that web 3ux will ultimately even be better than web 2ux there's no special role that bomy takes in this network Beyond being the developer of this network and the promoter of certain technological advancements like anyone can join anyone can run a node anyone can be a solver be a bundler be a whatever they want that's that's the beauty of Designing um permissionless [Music]

protocols I'm here with Amed alagi who's co-founder of economy and mlav yavor who's VP product hey guys thanks for joining me today hey thanks for having us hi thanks for having us yeah so bonom has had a really interesting run it's a product that was launched in the last cycle and was trying to solve some major problems with user experience specifically providing developer products that would allow developers to abstract away a lot of the complexity of using blockchains and so I wanted to before we go into the technical bits and talk about like the recent announcements with modular execution environment and everything uh talk a little bit about

the origin story of iconomy and what were the problems you guys were trying to solve when you launched a couple years ago yeah uh absolutely so origin story was uh it was around 2019 when uh me and my two co- finalist came together and we actually started off doing um staking um so back in the day tezos was doing dpos or and for well they call it baking uh um which was which was a pretty it was a cool concept back in the day and uh we essentially um did yeah we ran a sing operation initially um and then we sort of realized um that all right great we're validating blocks the

blockchain secure however we're not seeing like a lot of transactions and so uh we went into the the weeds as to why why that was happening and we realized you know one of the biggest problems in the space especially in ethereum back then um was just the ux really sucked um a lot of daps um had problems acquiring users um they had a lot of problems getting users to basically connect wallets to do transactions and so essentially our first product um was called gasless the it was the idea that developers sponsored gas fees um uh for their users um and this could really be for anything it could be for a defi

transaction and nft tment um a swap um anything really and so uh the really cool thing here was that uh the user didn't need to have eth um in their wallet in order to transaction and so um before gas Fe skyrocketed on ethereum WE that that's where we were I'd say very active and we even had OG projects like curve using by economy on ethereum before the gas fee spiked and that's where we sort of really focused on the l2s um afterwards so that was I'd say the original story and the original the the initial product that that we we launched with and so since then like how has the the mission uh you know evolved and you know essentially you we we've

just kind of come through a pretty pretty hard bare Market um you guys have stuck through and kept building and building really valuable products how has time say influenced the mission and the products Direction yeah really good question we so so we've um I'd say as a as a company um especially with our vision um it's I I would say it's all about enabling and empowering and anyone anywhere to navigate web fre seamlessly that that is ultimately our vision um and that really hasn't changed the the products has changed so we we did add to it um so we did actually build uh a bridge initially

called hyphen it was of it was basically using liquidity pools to allow users to move from one chain to another um and that was I would say more as a direct result from applications that we saw on polygon wanting to have like very quick transfers um of of of eth or other assets um from one chain to another because the users were were using those chains and uh I would say we quickly evolved and sort of realized that account exraction just before it it became a thing uh we realized that this was going to be the future for how users would interact on chain um and for how applications would also allow their users to have accounts um and so that's

where I would say we really focused on because it kind of provided a base um for what uh Au would look like B um it would better the security for users interacting on chain and C it would also um enable just a better flow uh on a single chain or even AC cross chains right so we had that foresight of seeing that account structure would help with these different elements um and that's why I'd say around two years ago the product really just focused on the c exraction um and that that sort of allowed us to sort of just build you know regardless of the market because we were solving um specific problems um and so yeah so I say this is like personally

it's my sort of third Market or you know third cycle and for B economy um yeah it it's sort of two bare cycles and this is a second let's say bll I guess we're in theable yeah it's really impressive I mean like uh to see teams um really navigate through the cycles and stay true to to their Vision like um I I guess maybe you know to to to to you as a founder and M love to get your thoughts on this too like as as markets come and go so like Cycles come and go um How do you stay focused and how do you find that like very narrow Focus to stay on course

when you know Cycles come and go narratives come and go Etc I do think that uh for for a lot of companies to stay focused you need to build products from the zero principles this is one of the ways me came about me came about with sort of in a in an intellectual vacuum I would call it where it wasn't like is this a good narrative is this something that people are latching on to because the origins of any c go before even chain abstraction was a term that people were using uh it was just basically out of sheer frustration just what Ahmed said about 2019 and then you know like with the gasless stuff and the people having trouble connecting wallets or stuff like

that is the same like with me it was the times changed and you know having a gasless thing wasn't enough anymore because people were interacting with five six seven 8 nine chains and I think we were doing that as well and so I mean the focus is just like what would what a web 2 company would do like what a web zero company would do if you're an ice cream shop you know like if uh you see your customers wanting you know lemon flavored ice cream you you you buy a lemon flavored ice cream and you sell it right so and I think like just focus is not thinking not thinking too far not going too far into your head just looking at the market if being a user yourself that's like critical so

actually using the products chains and the things that you think uh you know your company is building for and then just seeing where the friction points are and solving them and that's like that can keep you focused for 50 years if need be right yeah I I would add on to it to you know I I I don't want to sort of sound um like you know this person that says you know narratives are the only thing um but but of course they're not just the only thing I mean they do matter we all know that they matter and so you know in a very Ideal World whatever problem that someone is solving it does actually it it also has some form of narrative Market fit right like in in web 3 I would say it's definitely the

job of the the founder or Founders or even the leadership team to ensure that whatever they're doing and building um either has either they're building the narrative from the the ground up um or they're actually um a part of a narrative ative that that's also big and allows people to um to S get excited about the project and and and so I would say just this is more of a workr phenomenon to be honest and it it actually makes building a bit harder because Founders are somewhat either implicitly or explicitly pressurized to try and find like you know try and latch onto narratives and so it's a very very I'd

say difficult or interesting dance to to deal with um for for sure yeah token prices apply pressure I think yeah you know and uh in in in a in an industry where a lot of the focus is on short-term Financial outcomes is specifically like on the retail side and and I think also you know professional investors have this some some professional investors have this uh short-term um Financial outcome focus it it can be hard to uh not want to latch onto the latest narrative and and so I think it's it's a huge achievement when you see teams that are able to stay true to their mission and and have that

zeroth principal approach uh as you mentioned ml um narrative is also I would say not every narrative is the same I think that the part of the wisdom is uh you know um churning out through all of them and seeing which ones are worth latching onto which ones are not worth latching on to you know that's also something for they probably comes with with a bit of experience in the industry of course yeah one one other thing I would say just to to um on top of your point SE regarding the financial gains and sort of just the the the Outlook people have um even if we just look at this industry and I I tweeted about this as well and uh like it's very clear the speculation is just way there's just way more

speculation than actual substance um and even if you look at the top 20 protocols that are actually making revenue and money all of them have a trading aspect to it right so it's um your speculation get speculation to to some degree and so that is what is driving the industry um it kind of that that that along with with stable coins are probably the biggest pmfs that that we've seen so far yeah so let's um let's talk about this modular execution environment or the me framework that uh was announced uh I believe in November or December just a couple weeks ago so like what is me and how are you guys thinking about

it in terms of um you know bringing developers or providing developers an easier framework for executing transactions across like multiple different chains yeah so uh maybe on on a bit of a like let's call it a technical aspect I can I can take that one and then ahed can maybe go more on the business part of the things um so I would say it's uh what we realized you know as mentioned working from those zeroed principles is essentially um the transaction model that blockchains uh use is not sufficient for the environment that we

find ourselves in and I think we found it necessary to actually um remove the transaction processing aspect itself and execution itself from the blockchain and this like really coincides with the modular thesis how like call it testia or somebody they removed like data availability from the blockchains onto their separate layer uh there there are other things that are moving onto their separate layers and I think we really will see this modular thesis play out in in many different verticals on blockchain and um one of the verticals that we realized uh was really really necessary was execution and that modular execution is required um so what does it mean that execution is modular it means that monolithic execution Frameworks

would be one which basically executes one transaction on one chain with one signature model uh and and it just basically very tightly integrated with the blockchain on top of which the transaction is being executed u a modular execution environment in comparison to a monolitic execution environment essentially says no like the blockchain itself is just like the settlement layer for this transaction but the way the transaction is signed processed and executed can be done in many different ways so what do I mean by that um let's call it you have a transaction that needs to span multiple blockchains right so you want to um swap usdc for usdt on optimism and then you want to bridge it from optimism to base and then you want to deposit it into a

lending protocol right um so these would be three transactions on three separate chains um they are luckily all of them are EVMS but you know they do have their own intricacies you know having to set chain ID in the transaction all of this U so basically sort of enshrined into those blockchains how they would process the transaction now in a modular execution environment you would essentially encode a thing that we call a super transaction which contains all of those transactions within itself and then you sign the super transaction itself and the modular execution environment makes sure that all three blockchains step of the transaction so essentially how uh how we look at it it's multi-chain multimodel which means that

you can sign things that are not only transactions right so down the line you will be able to sign things like intents you will be able to sign things like you know certain offchain actions anything that can basically be proven onchain that it has been executed will be able to be a part of uh the modu execution environment so that's problem number one that we solved through this model called super transaction problem number two is a problem of horizontal scaling so you know the modular blockchain thesis and the f chain thesis says that we're going to have hundreds if not thousands of blockchains uh in the future and I basically we already have a 100 plus definitely we're in the hundreds already um so I don't see it as being a

farfetched idea that we're going to be over a thousand uh very very soon uh of like meaningful lab chains and and rollups and things of that sort um botomy has been like a leader in the 4237 space and we do have experience how hard it is to run a bundler just for like 10 chains uh or 30 which was our like uh support here but like 10 major chains just running um bunders for them is is extremely operationally difficult um so this is why modular execution environments are essentially collaborative which means that you encode this transaction object and the entire object doesn't have to be executed by a single entity you know you

you can have um bundler on one chain execute the part of the transaction for that chain a bundler another chain execute the part of the transaction on that other chain and then an intent solver in between those two bundlers actually moving liquidity around from one bundler to another and we do have a full-on permission as cryptographic model which enables those three entities uh to trustless coordinate and to commit cryptographically to the user itself uh to to execute uh these actions so I would say these are the two two main things that that we are solving is basically this new data model called the super transaction which which is able to encode things across multiple blockchains that it removes the execution from from the blockchain onto a separate layer and then we have this

collaborative execution which enables multiple types of actors like St bunders orac C down the line uh to collaborate on on executing these super transactions and yeah and on top of that we have a very userfriendly SDK that that basically enables uh developers to interact with this entire system um as if they're interacting with like uh VM or or like a normal blockchain transaction type so that it's not confusing for for examp existing developers right and so like I I think it could be useful here to dive into ERC 4337 a little bit uh for those who are not familiar with um with this uh ethereum standard so what does that

particular standard allow and um how is me uh applying that standard so one of the things that that ERC 4337 did is it introduced an idea of a smart account basically this is a smart contract that that is acting as if it's a user account this means that you get certain very cool benefits uh so this is like primary things where as as as like Ahmed said like they had this biggest pain Point Way Way Back which was gasas and I do still think that the biggest use case of uh ERC 4337 is actually gas sponsorships and then they'll gasless transactions uh but ERC

4337 is Essen Al you know much broader than that uh you're able to uh you know do like account recovery uh you're able to do multisig you're able to do a resource locking where you basically certain amount of funds are escro um so you're able to do a lot of like very programmable things with your wallet however ERC 4327 is also an execution standard uh where it basically defines certain roles these are called bunders and pay masters uh bundlers essentially post your transaction on chain they actually interact with the chain they extensions to the nodes of the blockchains themselves there's a additional method on the RPC node which which enables uh

these ERC 4237 uh methods uh um and then uh basically pay Master is is a separate role in this ecosystem which can either sponsor gas or allow you to pay for gas in uh erc20 tokens um like like usdc usdt D or or you know whatever else um so yeah these are these are like the core principles of ERC 4327 and they're good like the especially the smart contract account uh thesis is really good and I think it's going to become even more powerful Now with uh being in shrined into the evm and ethereum protocols with ERC 772 where the users will be able to upgrade their eoa wallets to become uh smart project

accounts um but the what we realized with with uh building me is that um it was built for a single chain environment so it was basically a project that like bony definitely spearheaded along with uh some some other players in the industry but it was uh you know done by ethereum Foundation they had a lot of influence there and it was done to solve problems of ethereum at the point when it was developed when ethereum was still primarily like a single chain environment with a few chains here and there you know some maybe a few rollups uh Etc it was never basically designed to solve this problem of of the environment which has 50 60 70 100 plus chains you know um so yeah basically

it's a great idea it's a great starting off point B economy built some of the best infrastructure there but it just times have changed blockchains have changed and uh so ERC 4337 had to evolve into me for it to to to be applicable in this new envir yeah Ju Just to also add on to that as well I mean ultimately ERC 437 is not a protocol change um which people might confuse for it's a it's essentially an ERC which you could think of it as um uh a layer just on top of the actual chain itself which um uh developers could just use there's no hard fork or anything at the at the protocol or the blockchain itself

ultimately what it does is it it it basically separates the signature um uh very verification from execu and that's ultimately what er4 37 is um I would say that that's in way from first principles what it optimized for and and that basically enabled all the you know different features and bells and whistles that that I've talked about um and so for us even if we sort of look at um ERC 437 today and how we're evolving it into model execution environments um is that you know we are are essentially I would say like me is essentially built on top of um account

struction um the the major difference that we're sort of enabling is that um although the the contracts um uh and everything's the same what is different is essentially the bundlers themselves so on our side the me node is basically we you could think of it as I could be to of a bundler um it's just a a better way because uh like M said ERC 437 does not cater for um you know a multichain environment whereas NE does and so this is where um essentially the spe2 of a bundler um as we describe it as a am node um would basically um have just sort of more and and better features

because it could accommodate a multichain environment um so we're still I'd say incorporating the core elements of ERC 437 where we're using smart accounts using the entry point um and uh at least yeah be being on the same terms as sort of what uh that the standards say on the onchain side got on the offchain um aspect that that's where what change comes out okay and so one thing to note here is that yeah ERC 4337 is is not a protocol change it's it's effectively like an an onchain standard but ERC 772 which enshrines uh ERC 4337 allowing EAS to be converted into smart account that would

be a protocol change right so ERC 772 doesn't enshrine 42 G7 that's a that's a that's a like a common misconception okay uh so ERC 772 essentially allows users to install a smart contract on their uh eoa address it can but it doesn't have to be a ERC 4337 compliant uh smart contract so ERC 43 that's that's I think a very important distinction for people to make it's a very tightly defined standard it's not just the smart contract account it's also the specification for the bundler specification for the pay Master specification for the additional routes

on on the nodes Etc so it's a very very uh highly defined standard and this is why people have been building on top of it because it's kind of you know you have a very uh defined way to build on top of it uh but 7702 I think is is even a much much bigger thing is because it will allow multiple differing standards including 4337 but also things like uh modular execution environments like the bonom me uh to be installed on top of your EA so so I think that's a very important distinction to make me is not 4337 on steroids it's actually not even compiled with ERC 4337 it's a new standard uh which is built on top of the same smart contract accounts that uh ERC 4337 was built on top of but it's a

whole new generation of execution from ERC 4337 and there is also one more like when we are skewing numbers there is one more very important DRC it's called DRC 7579 so the smart accounts that are being used are not just um Regular smart accounts they are modular smart accounts which means that users can install various types of modules on their accounts which do account recovery uh Pass key signature signing for transactions uh what I say resource locks estros uh you know certain Escape hatches like a lot of mechanisms can be installed on top of your uh um account and the cool thing about modular execution environments and bonom me is

that it's a fully industry compliance 7579 module which any user interacting with any 7579 compliant wallet can install on top of their wallet so you don't have to use bomy uh contracts to to use the bonom infrastructure and and that's I think a very important thing is like we we really want me to become the execution environment for everyone you know regardless of which uh smart uh smart contract account provider they are using Okay interesting and okay thanks for that distinction now the the super transaction I'd like to dial in on that a little bit um so what are super transactions what kinds of things can be

included in super transactions and like I guess one of the complexities here is dealing with fallbacks and how the ux uh gracefully handles fallback so if a trans If part of that super transaction doesn't execute for some reason um so yeah could you provide like an overview here and you know understanding a little bit how that works from the user perspective when there are problems so crush atomicity is an unsolved issue uh I would say nobody has soled it uh so what we do essentially on this front is is um I think we have gone the furthest uh down this route of of uh cross chain execution in in an atomic manner uh is basically we provide

probabilistic uh you know like fix up transactions that if C certain things fail you can revert uh things but what we we we go like call it defense in depth strategy basically first you have the transaction you assume optimistically that it's going to succeed secondly if it doesn't succeed you have repeat transactions with within the super transaction that says okay try executing it now but with slightly different parameters you know so give yourself a bit of wiggle room uh to actually make sure that this transaction goes through if it doesn't go through have a certain you know cleanup transaction which you know moves the state back to a certain state that it was Prior now none of this is as good as a proper transaction revert but it's as

far as technology will allow at this point however we are already in active conversations with several companies doing like block building uh you know these things with like lower level meev transaction ordering stuff to see uh and then there are good indications that something like this will be possible to see if if a cross chain Atomic uh transaction bundle can be encoded within a super transaction so it's a very very interesting uh problem yeah and and just to add on to that I mean I I think if we were just to like really um think just about how the current trans transaction model works right um I think we mentioned it slightly before in

the form of um when a user does something it's typically um on the same chain it's one type of transaction um and it's always going to be a user signature for for that transaction um and if they want to do more and interact with different contracts then they would have to sign multiple times go on different interfaces um for us how we see what what we really Define as super transaction it's essentially a data model right and this data model basically defines very specifically what um the user is going to be signing and this could be um a normal transaction it could be an intent it could be um an

offchain um Oracle um trigger as well really depending on what the use cases and what the developer wants to enable um it will allow for this I would say less complex multi- um uh chain uh uh coordination of of actions to be done in just one signature for the user and so for us um that that's I would say the exciting part about it is because the user just signs once for you know really for a multi- instruction um uh uh let's say use case um then that basically enables the user to just interact with one interface have no popups um and just

really do um uh things like we're used to right in in the web two world where you just easily um signing away things and and being able to to do things um very easily and quickly um and uh yeah I mean there are I'd say like we could delve a bit more into like what other features like super transactions can enable um because this kind of comes in part with the collaborative execution that M talks about but um yeah for us it's this data model which um which does a lot of things yeah like please uh I'd love you for you to expand on some of the things that this enables yeah um so what when we think about um uh like a super transaction

sort of as a as a data model um if we were to sort of think about like a use case and and this is a demo that we actually have that's live um a user could basically um and this will come by default what we call like or what the industry and chain of straction calls like UniFi balances um so the idea is that you will have this one smart account um and this smart account is represented on different chains um and you might have um balances across all these accounts right and and they're basically on different chains and so when you want to let's say deploy um some money on um a lending market and you find the highest API um you

basically want to send 50 usdc and you don't really care about which chain it's on um and assuming that you don't have 50 usdc on that chain then what would happen is the super transaction would specifically encode um and optimize for any Bridge action to actually happen um to move the 50 USD to move the required funds in this case 50 usdc to basically Supply to that um Lending um protocol and so for the user it's one signature um but the me node in the background is basically doing all the actions that was encoded um in order for um the user's intent to actually succeed right and

this and and this is where we call like this is why I like to call hybrid execution so the idea that um this this is still using bridges for example right like you might hear in chain abstraction let's remove bridges right um and it's just going to be int all the way down and I don't fully like I'd say believe that I think it's going to be a hybrid it's going to be the case of sometimes you're going to use intents sometimes you still going to use Bridges and sometimes you might use a mix right and so the cool thing with super transactions we we we also um support have support for intents and the whole idea of resource locks which you might have heard about um with one balance and socket and a few others who who are also

I would say just fully going down this approach but you but even with resource locks you still need the user to actually lock the funds and then be able to do a transaction right from from from that instance um and so for us we we do believe we're going to be in a world where it might be the case there is an intent and then a and then a normal transaction happening um and so for us we we see super transaction really blending in well on its own with Bridges or um aggregators of bridges or it could also blend um part and part with with intents as well again this really depends on the use case and so this is why we see um super transactions are I'd say very flexible that and that's just

one aspect but there's also um the other aspect as well where in a super transaction given that there are you know um like essentially a super transaction this data model right it's it's all these leaves in a Merkle tree um and how does it work with like collaborative execution if I'm a bundler on chain a and your Bund Lo chain B um we would both sign on within that super transaction give a cryptographic commitment that we will be executing these transactions right and so this will be essentially recorded on chain this cryptographic um um commitment from these different nodes that this super transaction will be executed um you know

on my side I'll be doing a part of that super transaction and you'll be doing the other part of it and so this I would say is a really interesting and novel um thing that really no one has and that we're building um and and this you know we we definitely think we like to call it like horizontal scaling in the concept of how do we enable um uh you know different actors to come about and and actually you know solve for user intent in a trust his manner yeah understood and and so I wonder if you could maybe explain like some of the um complexities of

building this type of execution technology on EVMS versus other chains and I'm thinking specifically of Cosmos because I have a pretty strong foothold there you know I know projects like anoma that are building intense systems and agoric that have like a very different approach and allow things like async await on chain for instance um you know H how how how do you sort of compare the evm landscape to these other ecosystems and are there plans for B economy to also support uh nonm ecosystems so I I maybe can uh give give a bit of insight on top of that so I do

think uh that there is um a need for an unifying layer on top of multiple virtual machines and uh super transactions have been built from the ground up to support multiple VM standards so there is nothing within a super transaction that says it has to be executed on evm blockchain uh the current implementation uses 4337 style uh smart accounts it uses 4337 style user Ops it uses 7579 modules Etc but that's just the implementation of the me standard for the evm right so that every chain will have its own implementation of the me standard uh as far as uh projects like

anoma or or agoro we we know both of them we've we also had conversations with them uh it's a different approach I would say what we are doing is proof-based so like in comparison to agoric which does these things like orchestration asyn weight stuff uh they do it through uh using one chain Within the cosmos ecosystem as being the orchestrating chain uh the me is a network which is not a blockchain and so I do think that there is a lot of value there uh because we are using pure proof-based execution where you sign a cryptographic commitment as Ahmed mentioned uh as a as a bundler as an intent solver you commit yourself and you post a stake that you

will actually execute what you said you will execute and then this stake is essentially uh used as collateral to keep you accountable for doing the things that you promis to do uh this has a lot of uh very interesting benefits uh one of the biggest ones being that that really turns up the scalability up to 100 uh because you no longer have the overhead of maintaining consensus you don't longer have the overhead of uh you know needing to So like um me could split into five separate smaller networks accidentally and nothing bad would happen like literally the transactions would keep on going you would either get a quote or not get a quote you are communic ating with one node if that node knows about the entire

network fine if it doesn't know about the entire network fine if somebody commits um if somebody doesn't commit or or goes offline or stops you know generating proofs that's also fine like if there is one bundler for one chain and that bundler breaks down the rest of the network keeps on working like as if nothing happened right uh so it's an extremely resilient cryptographic design it's an extremely scalable cryptographic design uh and so I would say that that is one of the main benefits of our approach in terms of technological implementation instead of actually using a blockchain to do this orchestration and we also support these things like

scheduling uh async A8 these are all parts of the modular execution environment we're actually working on something called the composability stack um it's using um it's using a set of predefined smart contracts on all the blockchains that integrate uh me and it enables like seamless frontend only uh building of call it like pseudo smart contracts where you basically say the output of a transaction on chain a will be used as the input of a function call on chain B right or like you call something on chain a then wait for like 20 minutes then call something on chain B and use the output of that as the input on chain a you know and then you

can build these also very interesting uh uh code promise based like asyn weit smart contracts which could theoretically enable certain things like um multi-chain uh atomicity in a way where you can like revert a transaction after it has been partially executed because the funds that were used for execution were held in a sort of asro model and we are already in discussions with several large uh players who are looking to integrate something like this uh and yeah so there is both approaches have their own merits I am very much biased towards this chainless approach uh because I do think that even uh a lot of other systems and a lot of other

teams have independently come to conclusions that using proof-based generation to for execution is is a much better way to to do things and it's sort of in a way how an optimistic rollup RS right you know it sequences transactions does things but at the end of the day it's all like proof sold the way down you just prove things cryptographically and if things go wrong you reverse them and this seems to be the way that the entire industry is moving uh towards and yeah so it Rose naturally out of technical discussions when when we were designing me that this should not be a blockchain it should be just like a normal peer-to-peer network with uh with with the all the settlement done on actual chains that you the user is interacting with got it yeah underst and

like can you um get into a little bit of the uh the the the gasless transaction part of this and uh what are the some of the challenges there and how how do you optimize for gas efficiency across you know multiple chains how difficult is this to do and how you achieving it it's actually more difficult that people would imagine uh so it was a big part of Designing the protocol was how do you make sure that you enable multi-chain gas abstraction in a sense where you pay for gas on one chain and you execute transactions on multiple blockchains and that this can also support sponsorships and that it can also support token uh gas payments so let's say like one of the most common use cases now that we have with people

is just like a user pays in usdc on optimism but executes a transaction across five chains you know or something like that um so there are caveats you know like 437 does gas uh sponsorships through an atomic model where you know within the same transaction gas is sponsored and the transaction is executed so if it reverts both the gas and the transaction revert together the user you know doesn't pay for anything um actually for the for the call for the call gas that there are certain there are multiple Fields within the RC 4237 standard for how how do you pay for gas but this all breaks down in a multi-chain environment because like one part of the transaction could fail the node could potentially grift you just

take your gas and don't execute the transactions so there was this entire mechanism designing of the staking slashing uh functionality where the node always commits to a certain cryptographic thing where it has to execute all of the transactions on all of the blockchains they are encoded within a Merkel tree type structure where there's a single signature that's describing multiple actions then there are slashing contracts on on multiple chains which uh the protocol can trigger to actually penalize the node for not executing things that they were supposed to execute so a seemingly innocuous problem like uh sponsoring gas on multiple chains actually has far-reaching implications on how you design your protocol uh um one of the

things also it's very interesting like gas uh refunds it's not always possible to accurately estimate gas ahead of time and so you don't want to overcharge your users ever and we don't so so the the cool thing is like NE never overcharges you for gas you always pay exactly the amount of gas that the blockchain actually used plus like a small premium for the Me Network itself um and so for to do that we had to figure out how to do uh cryptographically secure multi-chain gas refunds glad to say we solved that problem as well as uh multiple other problems in that domain so yeah it was um it was it was a tough nut to crack but uh we did we did manage

to solve it like I would call it like four five months ago and uh it's been turning on productively uh ever since very cool so one of the things I wanted to talk about as well is um is the tradeoffs between I guess user experience and decentralization this this feels like a recurring uh topic amongst you know all types of blockchain applications whether it's like at the infrastructure layer at the application layer if we look like very far into the future where you know if we anticipate that most users interact with chains through some kind of abstraction

layer how do you feel that about that potentially departing from um some of the actual benefits of blockchains and that they you know users should be interacting with censorship resistant decentralized uh platforms now I think we can look at you know the Visa Network as an example like every time you interact with your credit card you're essentially trusting one company to execute a whole bunch of transactions across you know potentially several different you know banking networks Etc and um I think that's probably what the future of crypto looks like but with crypto you know we we sort of are attached to this idea that applications

should be decentralized sensorship resistance Etc so you know how does abstraction you know sit within that framework and how can we make sure that these systems actually maintain those properties yeah I I ju Just very quickly um so there is one thing I remember that um someone told me um way back in the day when I first got into this space and um I'll never forget what he said but he said like what whatever can be decentralized will be decentralized um So eventually things will you know there is always Al there are many incentives to decentralize as well right um Financial incentive um the fact that you could sort of you know see control to to

um to a bigger portion of of the community to to come in and be a part of the network or part of um whatever you're decentralizing so there are different incentives to that that's number one number two I would say is that web 3 ux like my vision here is that web 3 ux will ultimately even be better than web 2 ux why purely because of the fact that uh users are in control um there could be privacy at some point as well they'll be able to set you know permissions on their own um uh and and ultimately they they are in control of their data their assets um and will be able to move you know their assets freely and very easily

unlike you know how it isn't in the web2 world and so I would say like we we will ultimately get to that point I do feel in the interim there's a lot of you know like there's this phase of all right there are decentralized apps and decentralized infrastructure the interim is like centralized apps and decentralized infrastructure and in the future it will be decentralized apps and decentralized infastructure right and so that that's really how I see you know where things are going um and again I do think it's ultimately up to the apps that will probably decide how much they want to um you know I'd say h how much centralization they they'd want to have and this will be dependent on the

business the use case um and and the user base as well so um but ultimately I go back to the same point web 3ux will be better than web 2ux and it will just require just that bit more time and I'd say on the me side itself on every metric that matters NE is decentralized censorship resistant permissionless trustless so by colomy couldn't prevent somebody from using NE even if we wanted to uh so essentially uh these are immutable smart contracts which are deployed by us but not controlled by us they don't have admin keys they don't have upgrad proxies these are proper immutable smart

contracts deployed on chain uh on top of all of the chains that me is supported on and the me nodes essentially simply interact with those contracts and they cannot execute anything that wasn't explicitly permitted by the user they will get slashed if they don't execute the things that they have been committed to and an me node uh does not need a contract with bonomi to join the network or a permission from bomy to join the network the entire process is fully fully trustless uh and permissionless and if all of that fails if if all of these protections are not enough there's always an escape hatch mechanism uh which is already built into the

contracts themselves where the user can always access their own funds by not going through me but actually posting a transaction directly on chain so we did we hold decentralization and like not even decentralization call it censorship resistant and permissionless we hold it in very very high regard and I invite everyone to take a look at our GitHub and me contracts and the implementation uh they will there's no other conclusion that to come to than that b economy has no control over these contracts that we have no no no special there's no special role that b economy uh takes in this network Beyond being the developer of this network and the

promoter of uh certain technological advancements uh but yeah like anyone can join anyone can run a node anyone can be a solver be a bundler be a whatever they want so yeah that's uh that's that's the beauty of Designing U permiss protocols yeah I think that's super important and uh glad to see that you guys are really sticking to to this approach of building decentralized censorship resistant uh systems with Escape hatches is that people can always be certain that they're able to withdraw their funds or execute their transactions so low if they if they need to or if they want to so before we wrap up here I also want to

talk a little bit about this other uh part of the product that uh is still in the works which is the the delegated authentic authorization Network or Dan um so what are you guys working on here and and um like I think maybe taking a step back how like this AI narrative and this AI agent narrative sort of taking over right now and a lot of people are talking about it it's like all over Twitter um how important do you think AI agents will actually be to crypto in the next you know two to two to three years yeah um a lot to unpack um so first I would say uh or I think there are two questions um I'll start off with

the first one um so Dan right now um so how did Dan start right uh we basically had looks into the AI agent space in general I'd say um yeah around like N9 months ago or so um purely because of the fact that it was an emerging space that that we wanted to just invest a bit more time in to understand and Dan was more a it was more research project projects with um uh with uh the the company um that we partnered with to um look at the problems that you know AI agents would have and one of them was um basically guardrails around what AI agents can and

can't do um and so you know you don't want an AI agent to basically um you know just like use all your funds so it would have certain permissions and stuff like that um what we realized was Dan was like extremely early um and this around sort of this was summer like you know um June or summer of 2024 um it it was extremely early in its time and what we decided to do instead was actually build our own agent just so that we're able to understand um how these things work and so we're actually building a b2c application uh called nomi nomi is an AI assistant portfolio manager um

where you know it allows you to use natural language to um the agent to do transactions on your behalf and of course naturally this would use an me under the hood as well um what once once it's already um so I yeah I'd say Dan has kind of transitioned into that um as for the AI agent part um it's I'm really glad you ask this question because it's definitely something that's been on our minds you know quite a bit especially since we've looked into this Market a bit a bit earlier than than others and so um we we are definitely of the belief that more and more transactions or there will be a very Bas significant portion

of transactions that will happen via AI agents um uh if you look at the current landscape right now they're really awesome and they're showing I wouldn't say pmf I mean some maybe AI xbt has some pmf because a lot of people actually rely on it but they've done a very good job in in the social social Arena and just content Arena right um they're basically like removing the need for KS to some degree or at least they're deplatforming KS um uh to to you know to like if I was a k you know how how will I compete with an AI agent um or will an AI agent represent me right and so I think on on that level what AI agents

have done is just shown um like the Innovative nature of what AI AGS can actually do and so assuming we continue on this trajectory um then they'll be able to do transactions right so we haven't seen the case where AI agents are actually doing transactions at an enormous scale right maybe they're tipping maybe they're like AI xbt asked um sticks for something and it paid them whatever it was right um but have we I I would definitely say that that probably that then when it comes to sort of the execution side of things it will be humans um firstly like getting agents to actually tell them to hey do these

transactions and before we get to a fully autonomous World um and so I think that's what um is keeping us up keeping us up at night um it's like you know should we focus our me um for AI agents if they're the ones going to be doing it instead of humans right um because that is definitely um a world we could potentially go towards yeah interesting I I am I don't know if you read this uh this tweet by h of dragonfly capital and uh he he talks about this and his thesis is that uh they that will revert back to human K's at some point um in you know in like a

two to threee time frame like there will be you know Peak AI agent and then at some point people will start preferring the uh the human K's again um for like a number of reasons that he describes in his post wonder if you think that this is like a long-term thing or or more of a short-term kind of hype cycle yeah no I again I think there will be a lot of things in in hybrid when you know when it comes to these things will uh I I would just say that AI agents in general um are always going to be a tool for humans to just enable us to save a lot of time and effort and if I as a k am able to

um really expand my breach through a super performant AI Agent K then why not I think it's going to be hand inand um there would definitely be kol like human KS that would be way better maybe than AI agents maybe they have speciic focus and I haven't seen the Tweet by hi see but I'd love to check out um but there will be some AI agents that will be better than humans right in terms of the content they they come out with so I think it's it's not going to be this or that it's it's going to be how we um utilize the AI agents for for our advantage to be honest I think it it the the argument for human KS is not going to go down to Quality because I think I agree with AED

100% there's probably going to be AI you know advisers who will be better than most humans I always see it as a scarcity argument there will be like human KS will be like nfts you know it takes uh there's much less of them it's much harder to produce one so I'd say like it takes 20 years to create a human K you know it takes probably a a month to create an AI k um so it might become just like this type of status play thing where it's just you know you you follow it because it's it's like much rarer at a point where like a AI content is I think it's not just for K I think it's for content in general like I think that human generated content is going to incur a certain artistic value in the

future uh just due to the very effect itself that it was created by a human and I think this is one thing where crypto is is well positioned to actually be the verification layer for the fact that was it created by a human or was it not um and yeah so I think there will be this very interesting social Dynamic at play uh at like AI agents becoming Super Hyper intelligent and being able to give great recommendations and everything uh but just becoming so abundant that uh people get like sort of sick of it at some point and then they revert to just uh human uh advice maybe for even just like emotional SL artistic reasons you know um yeah so that's that's that's my like uh hot take for the for the future

of this so what's on the road map and what can users and the community expect from bonomi in the next couple coming months and for 2025 so there is a lot on the road map uh so initially we will be uh we we've already launched the me devet which is basically um developer preview version of the me and the me asdk the MDK will be launched uh now together with all the other things that we have built in the sdks over the years they have been unified they have been uh made vendor neutral they will be launched under the name of abstract JS

uh which will be the single SDK that any developer needs to interact with any blockchain uh primarily evm now for the beginning but and to use me and to use account abstraction and to use 4337 and all of these things through a single uh unified unified interface um we're full steam ahead uh very soon we will release a public road map but we are full steam ahead on uh releasing the uh production version of me uh so all all I can say for now is that that that it's it's not far into the future uh and that people should wait for the public road map but we'll uh we'll we'll publish that very very soon um and then there is the process of decentralization we have uh

things I'm not sure how how much and what's smart to say here but we have things in plan for the token holders uh to certain you know uh because the the the network itself will depend on the staking and slashing which will be powered by the bico token um so that there is there going to be that aspect of it so you know a lot of things on our plate a lot of exciting are development launching shipping I think we have been on a shipping spree for the last like two months we've just released a thing after thing after thing and then I'm hoping we keep that pace going throughout the entirety of 25 great thanks guys thanks so much for coming on and uh look forward to seeing more from economy in the next s

months have a good one cheers bye [Music] he [Music]