Panel: Account Abstraction, why is it important?14th Oct 2023 | Sunway University | ETHKL 2023

okay uh before we start let's do some warmup how many of you are developers or there are not as many as I thought how many of you are just random people trying to learn about account attraction okay there's way too many than I expected uh who don't don't even know why is account attraction okay wait um I don't know if I should feel happy or sad about that but hopefully you guys can learn something um new about U during this panel discussion so yeah without further Ado let's get started okay cool yeah um I'm wean the cat there so I mean yeah the cat there sorry I'm the moderator for today so today's topic is going to be about account attraction obviously and mainly the discussion will be about um why is it so important even though um Amazon just now has already give a very wonderful speech on um a bunch of uh under uh industry insights on um the current AA State but uh uh nonetheless we're still going to discuss more about Construction in depth with three of these speakers here so I'll just introduce it the speaker one by one so on the left already um we have Emerson from long has Ventures clap clap then in the middle we have Brandon from herera then lastly we have Franchesco from consensus metamask okay so um let's give them some time to introduce themselves even better and U uh while you introduce yourself I also want you to tell us more about why do you think account instraction is so important so just give a little bit background about yourself and yeah why account exraction is important so that's our R you Emerson yeah sure so uh my name is Emerson all right all right that's cool uh so I I help manage the uh portfolio at lash Ventures um supporting uh projects like lit protocol safe um and particle Network on growing uh their user base and uh developing Partnerships within our portfolio um so yeah and why why I think uh AA is important is is essentially because um it's unrealistic to uh expect that everybody will be onboarded with the seat phrase um and uh we need to create a a new safer Paradigm for new users new users to enter crypto in a safe sustainable way yeah cool hi I'm Brendan um from Singapore and uh yeah doing develop relations at hadera um yeah uh why account exraction is important I think there are few things or use cases that it enables which um similar to your point let uh I guess people people who are not like crypto DG Advanced users like the average person can actually start accessing web3 Technologies and like AA is one of the ways that or one of the technologies that enable that um to happen hey what's up guys uh yeah Francisco and part of the Deval team at metamask and yeah I'm quite involved in a since the start even before 437 so it's not it's not uh yeah it's not something new I think we standardized you know the the logic with 437 but uh before even with safe you know like we we had that uh yeah like I don't know if how many of you guys are in 437 Mafia but we started a group end the days and and today is is quite well known for like a lot of like research and good discussion points so if you want to be involved yeah also like be part of that group and uh we also received the EF grant for for our a Meetup we did the one also like in Singapore where we had Alo vitalic speaking about the history of a some of you guys were also there it's it's great to see and uh yeah why I'm I mean I'm I mean not bullish just about the a I think it's just we need we will have different as you mentioned like different Target personas and I a be believer that you know light wallets uh will play a big role uh and it's kind of a natural evolving of needs of uh of of users right people don't want to just uh do uh you know like uh specific recovery they want to have you know limited they want to have Frozen accounts you know they want to have different set of features that I think like with a we can enable that and uh I think we need to nail down a little bit like how is more like around the infrastructure with running bundlers and bunch of like specific discussion but uh that's definitely a good topic to talk about and uh and Advocate about for for all the users cool so I think in the general sense is that we all know that account fraction is important that's like um the conclusion that we have about account ction but looking at the previous let's say Tonka standards or different ercs that we have on ethereum itself uh what's so special about this ERC 437 like what are the new possibilities that account exraction enable or more specifically ERC 437 enabled that we couldn't get previously yeah I think it's partially is actually a developer mind share um understanding what is a what required in order to build an e constraction wallet um and that unlocks the modularity I was talking about um in in the previous in the previous talk um essentially for developers to create their own specialized wallets um and uh loop it that loop back into into a specific standard uh that everybody like abides by yeah yeah um I'm going to have a slightly contradictory you know opinion on this um yeah cuz I don't work uh with ethereum right I I work on hyera and um it is evm compatible so you have Json RPC you have the same type of smart contracts op code level equivalent Etc but the accounts and transactions model is quite different so um so I take like a more I would say different or I guess more lower level view of account struction than say just focusing on 43 7 specifically or like even on Smart wallets so for me what account abstraction is is like you've got a few different um ingredients that you need to enable it and so the ingredients are things like um can you remove the separation between a private key and an account or it or its address right can an account um after it's been created can it then change its private key can that private key be multisig can one of the components of that private key in the multi uh multi- scenario can that be a smart contract right and like when you when you meet all of these requirements oh yeah one more as well which is can an account uh and the signer account and the payer account on a transaction can they be separate accounts for that same transaction right so when you have all of these ingredients you put them together then you get like a lower level of what I think is being discussed as account abstraction um so you get other use cases in addition to to uh what 4337 talks about a lot of them overlap and I think the most useful ones are smart wallets but yeah that they could be more so there like a more lower level view of it yeah I think uh I mean with smart accounts where were there before right but I think with the with the EIP uh itro standardization around what we're talking about he brought like you know special words that we can discuss with and special like architectural designs that we can we can debate of and I think that's that's what is exciting right so I think with this this is like the first uh kind of like process to have you know different design conversation and also like he brought up like uh a standardization around different you know like infrastructure tooling you know and that's where like kind of things came out like you know hey we have pimo SDK economy SDK I think it's uh uh it's it's super exciting what happened right but it's also a process that started before but I think with EIP we standardize like you know the language that we're we we're talking about it's basically what's what happening a little bit with also like nft standardization from ethero mag magician couple of years ago that everybody was building their own you know Marketplace without metadata standard and then you know they came up some standards and and now you know Marketplace can can talk with each other they can exchange specific information so it's I think this one first step but we still have a lot of like different challenges yeah definitely I think if you look at the history of account ruction it didn't start with I mean now everyone knows abouton attraction because of USC 47 and maybe a bunch of other chains as well but I actually started way earlier in 2016 if I'm mistaken but it's only now is that it's getting traction so yeah it's definitely going to get very excited we're going to get more application being built on top of account Traction in the next few years but yeah I I want to know for each of you personally we have talked about so many use cases right bad transactions you know uh session keys and stuff like that but for each of you personally what's your number one use case like what's what what's the application that you're most excited about yeah among all the others depends I think on the use case like on the target audience I think like you know for gamers out there uh having uh uh you know having sponsored transactions uh it could be a very like interesting use case um I think for like you know end user users that are more like targeted like to to like light wallets like limited accounts I think like social recovery like an easier process I think that's that's definitely needed and the full onboarding without the uh secret recovery phrase and uh and yeah and I think it's like whatever we're doing like we just need to always like uh um be kind of mindful of this kind of like five principle of those principle how we want to build things and it's always like about the end users having the end users in control of whatever they're doing and I think uh and they think this is the most important things so I'm I'm still like you know interested in Social recovery but you know it's a use case that we need to do it all are right um yeah so I have a very specific uh like uh concept like answer to this because we actually were not in not in herera but my previous role in in root stock um one thing that we discovered through a lot of user research is is that um it is very difficult for an average person to deal with a wallet that has multiple currencies um you know like basically the average use and this is this is specific to um the company was based in Argentina so it's specific to Latin America several countries that we research so we don't the users didn't want to care about you know multiple currencies but if you want to transfer tokens around like say usdc and you have to pay for gas right that's that's a big stumbling block right um being able to use your account without uh and and transact on a crypt Network without even realizing that you're on a crypto Network that's another you know uh need to have to yeah to to basically get this um massive adoption kind of use case so the I the idea is like create a wallet that enables that enables that kind of thing right so um One account account um like oper the wallet operator can pay for the the transaction fees in gas um in the native currency and then and then have and then maybe deduct that in you know in transaction fees in the same currency of the user or perhaps subsidize it you know so so that kind of thing is is like what I'd say um I guess it's possible to account achieve without account obstruction but with account obstruction it it becomes like um I guess to your point like a much more standardized way of doing these um these things yeah I've a really idealistic response that may be futuristic from like five to 10 years later but uh one one use case I'm very excited about is around um Revenue distribution with a lot of different um entities or or users um in a case where say you're releasing a film and uh the the revenue you you receive from a film is just movie tickets right uh and what if what if you can automatically distribute these revenue for movie tickets across all the different entities that are involved in that movie um this like in traditional world this takes like 12 18 24 months to distribute but if everybody is onboarded easily with a smart account on chain and this is automatically distributed to all the uh contributors using a very simple to use wallet um that is automatically deployed for you uh that that makes a massive difference to the people involved in that film um immediately cont like getting getting more immediate rewards for their contribution I just want to say uh I really want to add also on your point on sponsor transactions that uh this is obviously like a technical Improvement but in the same time also we are enabling new kind of business model I don't think is even in the future like we will have a uh you know like who will pay for those transaction I think you know dep dep owner will be the one sponsoring the transaction for the users and then also like new Revenue stream will kind of like come out but I think you know I heard for the first time I think was during e Tokyo that safe was doing like sponsor transaction testing with nosis or something like that that was really cool right because then that owner can actually sponsor this transaction and having you know users full uh just one one step I think further and but I want to say also like it's all about also like changing this business model side mindset I think the the same will be also like who will run bundlers how is the revenue model on on that I mean right now with consensus we have in fur so we are powering rpcs but in the same time I think same will happen on the bundler level and I think that's also a new business model coming out and that's also exciting because we're thinking about technology but in the same time we also change like different like business model narrative yeah cool so definitely account attraction enables a lot new application of new business model and uh ultimately it's when we talk about contraction we also talk about smart accounts which they basically smart contracts right so um it definitely improves the user experience on chain that's the at Le at least one of the goals but at the same time because we are trying to convert on ethereum um from EA to Smart contracts and we know that smart contracts are very vulnerable to attacks so how can we make smart contracts safe like very safe and at the same time convince the end users or the next wave of web tree users to use Smart accounts instead of um the traditional account model EAS yeah yeah so uh I think there are two two ways this is being really addressed today um one of uh what what safe is doing is is is one one having a registry and second uh having a security Council to audit the modules um so what whatever module that this could could could potentially be integrated in a smart account they will go through a security Council of very highly qualified professional Auditors of uh smart contract wallets U go through that and then have an onchain verification that this module has been audited and this will go to a the the sa registry um and the registry will show that okay this this module is safe this account is safe uh this wallet is safe um and uh yeah essentially build a onchain record of um modules that can be safe to use uh yeah I was going to say that whenever there is a there's like a new technology a new set of features then the surface area for attack factors is is higher right and there's always a chance and almost a certainty that someone is going to use that and foot gun um so it's going to happen right but what I would say is like compared to when smart contracts just general evm smart contracts came into play the first time like when there's like a feedback loop right and the the question is how short is the duration between when an attack bacteria is discovered and or exploited and when it's fixed and best practices come Out Come U you know arise out of that and I think like when smart contracts first started there wasn't a established you know security kind of mindset um there weren't big players or established companies with with track records and security auditing in the web 3 space etc etc right but now there is and so that gives me hope that you know that uh this this time around you know you won't see such big hacks or if there are big hacks um you know or security vulnerabilities discovered they will be addressed much more efficiently than they than we have seen it in the past in this industry yeah I don't so I want to say that for example the I think there will be a big tradeoff in general right like I was checking like specific reports they were telling like transaction in theory are more expensive on on Smart accounts and uh and basically what what this means also like I think there would be a trade-off on specific like use case being done directly through smart accounts versus like uh with versus other like methodologies it's the same like discussion that we have with like okay all our data need to be on Chain versus not right and uh I think uh yeah as we are deeply diving in we also like need to see the overall uh the overall picture right and the specific like use case will be enabled but in the same time we have multiple ways to achieve achieve those and having different like targets personal Target users being satisfied with okay so um now that we understand you know um I mean security is definitely one of the challenge that account instraction is going to face but what are the other practical challenges that account traction um we still have to solve and how how might we even solve it in the future yeah yeah I think uh challenges wise I mean like uh obviously I'm excited about snaps uh we were talking you know how we building modular snaps that are depending to each other um I think you know like we would always see single uh use case um single snaps use case around the a that would enable specific things but I'm excited you know about like this kind of like modularity that the people can build on top of uh top of things so that's probably like one of the biggest challenge the other challenge I don't see enough like developer tooling around a like support like this is something in 427 Mafia we're discussing a lot um I think they are like very specific a multiple like developer tooling are literally forking other project and building top features so I don't know if we're actually improving there uh but uh but you know it's a progression right and uh so it's it's just very exciting to see like how we are like compared the yeah smart accounts of safe before even the EIP was admitted yeah um so I'll I'll say like the the big thing is that um can or will account abstraction enable greater user adoption like uh so one of the things that I think it's is a standout what you mentioned like right at the beginning all the list of things that um Argent wallet does right right which is for example you've got um social recovery or maybe another decentralized recovery mechanism um not having to manage a seed phrase or price private keys and just being able to use it um maybe with the same level of technical um effort as a username and a password right which is what people are already used to so these sorts of things I think are enabled by account obstruction and and yeah like it's uh it's it's a interesting to see how this will actually play out in terms of like will this drive user adoption will we get more um so-called web 2 users to actually on board to web 3 as a result of this um yeah and then I guess like the second thing I I wanted to say was like just during your talk earlier on I saw uh the bit about rhinestone and I was like saying hey metam snap should do that you know so it's like it's kind of like uh it's it's kind of interesting right like you come to this conference and then you know you just suddenly get a new idea and the space is so nent that it's not like oh that has already been done right cuz if you do something like that in web 2 is like there's almost certainly someone else has already done it but in web 3 it's like more like a open field yeah definitely resonate with that a lot um well although another another challenge that I I foresee as well is uh vendor lock in um when it comes to uh modules as well because um in contrast to EAS where where every EA is is is is is all the same right um the challenge with u with modules is that when you build a module you're building with specific like provider can be safe can be zeroa can be can be bomy and this creates uh potentially a a situation where modules cannot interact inter interoperate between uh providers um so there are specific standards like EIP 699 6900 coming out that ensures um like modular contraction to be interoperable um but I've heard some some issues with that with that standard as well um so yeah I think in order to ensure that these modules are are freely interoperable between providers um we need that standard all right thanks for sharing speakers so we actually have five minutes left time goes by really fast so I like to open this 5 minutes space to everyone uh that has any questions to the speakers here just raise up your hand anyone if not I'm going to ask more oh there there it is hi I'm Dan uh I have a question for Franchesco um since AA is here is there any plan for metamask to uh do something or to improve your or ux and integrate 437 ouch yeah for sure so uh what I mentioned before in the talk was that with snaps now is the first first iteration and the idea also is that Snaps are we allowing you for like an extension on the UI side so basically like you're you're building your own even a snaps in the future and what you can do is uh you know adding buttons and changing design directly on on production metamask I think that's that's cool value proposition and uh yeah and for I don't know if you're talking about mobile or extension there was a couple of release of uh of the um of UI refresh I don't know if you check it out but network network selection uh on mobile now is smoother I know like there is always like big pains around that but uh yeah we're working on that but I think like the uh also like the tldr of like the the UI part is that snaps will enable also like a UI extension so for any like future snaps developer here I think this is also something super exciting because you can actually like extended functionality with directly like some AI uh uh components uh in uh uh directly in extension thank you all right nice any more questions oh there yeah hi um I guess this is a question directed at Francesco as well for snaps so can you share a little bit about why snaps as a concept was like conceived like what made the developers of metamask or consensus decide to build something that is focused on developers yeah yeah this is a good question so um so before like so I started my talk I think uh I don't know like last year with snaps uh basically like mentioning something like I don't know if you check out like our our metamask Fe like repo like back in the days we had something like you know 1,400 like open PRS like literally like features feedbacks was was crazy right and literally like one every like two days people are coming like hey can we have uh natively integrated our you know our uh Tech it's always the same right and uh we ask ourself you know with this way with this level of ask we'll never be able to satisfy like future development and I think like we actually came back on more on a higher picture we said okay how is actually looking in a web three process how we actually reaching out to this permission Innovation setup and the only way to do that uh is satisfying uh thirdparty builders in the community and uh that's basically like where we came out with this kind of snaps concept like to extend fully functionality uh and in the future UI components on top of like the wallet and uh that's basically how we started so 18 months ago we released fully snaps on on flask meaning you can actually develop uh uh develop snaps on flask and and during permission L we release uh those 36 um uh allow list of snaps that you can directly install on on production so it was really because of all the poor requests on GitHub that kind of LED yeah exactly so mostly because so many people have different requests and uh we don't think that in the future the company are able to satisfy the also like the speed of innovation that you know all our community is asking for and that's where actually we enable all this Innovation so that's how we started and uh you know there are also different things you know like we we would love to you know to have more Community owned uh um uh kind of like functionalities I think like I just want to reiterate you know Snaps are buil by party uh Builders companies like startup individual contributors and this is how really like we're enabling and and right now you know I don't know I did my first metamask transaction I know maybe six years ago and was just a normal uh PE top transaction and I remember like now like how we using metamask like in general like is is an ecosystem thing it's kind of like a cultural common and uh and now like we're going on more uh on a more lower layer where actually like we can expand and people can be like Lally like modular on top on top of uh on top of like the wallet itself and uh I mentioned that before in the talk like you will not surprise me if there will be much more standardization around adoption of snaps on different uh uh wallets I think that's that's super cool the only things that we need to keep in mind is we want to do that preserving the rights of the users uh and not just shipping features that are uh let's say compromising that Vision you know like uh I think that would be at at the end of the day day like a a trade-off and user will decide which type of features they want to use like we we are not pushing a specific snaps use case compared other snaps use case we want the users to make this decision right like as you saw like for for the the directory we have transaction Insight snaps and you see like something like 15 different snaps and uh they're built by different companies doing different like security having different security text snaps and I think at the end of the day like users will perceive and we see uh different uh different Innovation yes this is for users but on the developer side like you're open to it's everything is open source also the snaps being being issued at all you can see the repo you can Fork it you can rebuild it but the cool thing is that uh uh it's really like enabling uh Builder every Builder out there to extend in the functionality of what uh what what the wallet does today I can I can talk very briefly about being on the other side of that so um I or my team have submitted I think two different PRS to to the main metamask um this was in maybe 1 and a half years ago and then waited for multiple months with no response you know or or some response saying maybe but we'll look at it kind of thing so then after that the final response that we got was oh hey there's metam snaps why don't you do this in in inside a snap instead so they're like oh okay finally you know so that that's been my experience yeah yeah know this this is a good point and uh you even build the dras snap back in E Singapore it was amazing yeah and uh like the the inability uh category is also super interesting because right now like every two weeks there are either new layer twos or like uh layer ones and you know like the only way that you can actually integrate directly if it's fully non evm it's like through snaps now and before we got bombarded really by by native integration adding new networks and uh and now like there is a way right so I think this is also our responsibility to provide like a full a full standardized uh infrastructure fully audited for developer out there because if we're not standardizing things we cannot build further same with like what we're discussing with EIP EIP gave us like a way to discuss things and and build on top of that so that's that's where really is exciting yeah all right that's great so that's the end of this panel discussion everyone let's give them a round of applause