ERC-4337: Adoption Analysis by Tom Teman | Devcon SEA

[Music] hey everyone my name is Tom Tean from the ethereum foundation I'm a product manager for the econstruction team so we're going to be a little less technical in this talk and we're going to talk a little bit more product-wise what's been going on the past you know almost two years since we launched the entry point contract to mainnet so um this happened back in March 2023 and since then we also deployed entry point 0.67 that you have talked to with some minor with some bug fixes but in general like you said there's not going to be any more any any ABI breaking changes and the one of the important things from deploying the entry point contract is that uh uh a concern that you have when you're talking about uh adoption is fragmentation and if we go back to the early days like when we had you know just a couple of EA wallet web 3js was very beneficial in a sense you feel like okay each wallet wants to bring their own thing to the table but on the other hand you do want to make sure that as a wallet provider I don't want each application to have to develop a different interface just for me the it creates a problem for the users for the daps even for the wallet themselves so having this very popular smart contract wallet prevents fragmentation so since a con abstraction brings so much to the table in terms of the functionality it created this effect that preent that you know kind of coalesced everyone around this standard and that is a very valuable side benefit that we've seen um so keeping in line with the etherum foundation and uh infinite Garden kind of theme it wasn't just enough to you know plant that seed of releasing the entry point contract we also have to continue watering this garden and making sure that you know when you're a developer you need to make sure that there's developer tools ready in place and things like block indexers so when people start writing code they can actually you know look at transactions that are user Ops and and read them and understand them we had to make sure there's a couple of bundlers and walls and sdks ready when we launch so people can start building on that and in addition besides just making sure that Developers have what they need and can build on top we also want to start exploring account abstraction so we gave out grants to teams to look into interesting ways to sign Keys uh look into complimentary standards like model or accounts stuff like rip 7212 which is um allowing uh you know using your key to sign transaction but that key can be uh you know your fingerprint on your phone or face ID things which are very popular that are very expensive to do with EOS well you can't even do them with EAS actually but when you do it with smart contract wallets it's still expensive because that method of signing the key the the the cryptographic function is not pre-compiled in the node so that's very expensive so we want to have that and through rollups we can experiment with that education and obviously also work on the user op meol we had to make sure that eventually will come and as we all know that has been launched we'll talk on that about that a little bit later and so ethereum is not the only one that wants native account that wants account abstraction we have all the l2s and rollups which we very much love because they're helping us reach the scale that we need but we do again we do not want fragmentation we want them all rallying around the same standards y have touched on this in this talk about the fact that as a user if I have a wallet that works on this rollup doesn't work on that one and then ethereum it's a whole different thing that is also something that we want to avoid so to do that we introduced rip which is the rollup Improvement proposal that allows all the different rollups to kind of rally around the same standards um in addition it's very much worth mentioning that to when you want to add so those rip proposals are very nice but if they don't get implemented somewhere doesn't really help us and adding things together by it's not that easy for good reason you know ethereum as you as as we know doesn't doesn't move that fast for good reason but RS want to experiment want to start adding things and they each we don't want them each foring gu by their own because then that creates a lot of work for them they don't move as fast and again we're starting to create this fragmentation so the EF is working officially with nethermind to build rollup G which is uh basically the there's going to be a core Fork of G that all the rollups are going to work on together and each one can also for that adding their own very specific things that they need but they're going to have this core very much rollup um um friendly uh you know uh client they can all use and save on you know development time have that security benefits of all everybody's using the same thing Etc so we also wanted to make sure that all the you know different uh rollups and L twos are also aligned and we're moving in the same direction and having the same standards So speaking of Standards there has been just like an abundance of standard that came out some of them a few of them with our kind of like a I wouldn't say like blessing our support but a lot of them are just teams in the ecosystem that realize that they need to have H standards to move forward you know in a way that is much better and faster and more secure so I'm not going to go over all of them there's quite a lot and I just the mere fact that I can say that I probably missed a few just kind of shows how many there are but model of smart accounts is a one of the very interesting standards that have Arisen we have two 6900 and 7579 always good to have some sort of like a competition always breeds more better results uh and I'm going to touch on that a little bit later but things like 7702 and 7212 which I mentioned which is the uh using uh uh web aan to or more specifically secp 256 R1 signatures for uh signing transactions and a lot of other ones ways of EIP 5792 is also very interesting which is uh you know the communication between adap and the wallet is when it was in eoa was fairly simple but once you start having a lot more functionality in terms of like batching and pay masters we need to start making a bit more sense of the API between those two entities so there are efforts to also standardize that and just like I talked about web three back then this becomes the new kind of thing that you want to have um this has been talked a lot in the previous two talks so I just wanted to kind of give it a slide to reference this but we are not stopping at 4337 this is what has been going up until now but we are growing into more uh into into the future which is going to be native account abstraction a couple of things that are really accelerating the growth in the ecosystem that we've been seeing and at the top of the list is M accounts and I'm going to go forward just a second and come back to this this is if you look at the Smart accounts that have been deployed by provider and you see that little Lego piece I put next to the one the list on the right there those are all model or accounts okay if we're talking about H Alchemy and zero Dev P economy those are all using ERC cc900 579 model accounts are being super beneficial to the growth of the ecosystem for the and the and top is safe well because that safe but um they've been around for so long but it's it's showing you the benefit of it because other teams don't have to build everything from scratch they can build on the existing Security on the existing adoption and and everything around that and just add the models that they want they have a specific way to do account recovery they can take Zer deps kernel just take it as is and build a model that does recovery in the specific way that they want to add something or some unique interaction with their pay master or anything like that but is it's driving growth in a very very meaningful way um there's other things here that have been very very useful if it's zero knowledge that we're seeing a lot of Novel approaches ZK email has been given some talks they've doing some really cool stuff um pay masters are not just for uh sponsoring the users gas fees we've seen some Innovative uses where you can have cross-chain asset transfers there's magic spend Plus+ there's also Magic by zero Dev basically allowing you to have that single balance anybody who's developed a wallet and has to show users their assets across different chains even if though it's the same usdc it's a terrible user experience this allows you to actually allow they're doing things like where you put some money and lock it up and then you can use the cross chains with pay masters it's really cool but it gives the users an experience which is very very friendly interesting point about pay masters so a year ago 98% of the user Ops that we saw were using pay masters like gas fees were like like sponsoring gas was a big thing for for applications we even saw cases where applications would allow users to connect with metamask or any other EA wallet as if it's just an eoa account but behind the scene they would deploy a smart contract wallet use the EA is the signing key only so they can actually give the user the that gas sponsorship because it just improves ux so much we actually saw it go down to 88% I believe that the reason to do that is just because we're seeing you know more functionality that is you know that has become important enough so it's it's not the only reason to have a user up if it's transaction batching on Dees if it's simpler authentication methods or account recovery those those have become a lot more sophisticated and a good enough reason to actually use Aon abstraction not just gas uh uh sponsorship um as you have mentioned in this talk as well we've launched the user off mle that be's mentioning and one Thank The Ether spot team for taking the lead on on on promoting this as well as candid and cilas for also taking a huge role in this H and um also fast lane which will thanks to their work we can also have it on polygon cuz they've really helped solve some front running problem there that existed so this is been this this is always been on the road map and we really care about censorship resistance decentralization and we love user experience that is good but we don't want to compromise on that so it's a very important Milestone that happened two days ago so we're very excited about that let's talk some numbers so these are numbers that you sometimes see everywhere you know close to 100 million user Ops over 20 million uh users and I want to focus a little bit on this well you can't really see the laser but the monthly active users now if you look at the numbers on the left hand side here on the Y AIS see 3 million monthly active users really well well no not really I mean I want to keep it like real and not just like throw numbers that are not necessar neily what they mean so I want to thank zero I want to thanks Kofi uh he got a grant from us to build bundle bear he's a wizard data and he's here to clear up some misinformation put more nicely and um he worked he gave a talk at duncon on the one day before Devcon and he was gracious enough to allow me to also use this data in this in this talk so if you filter out you know things like spam and Cil attacks and frequency trading and me and personally I don't necessarily think that these things don't mean activity okay the whole idea behind blockchain was that we can use machines to do things doesn't have to be a human person doing the activity to consider it real but we do want to understand what is the human interaction with applications not just things that are happening automated we still see very nice growth but the numbers are more around 600k monthly active users which is still impressive uh but these are the numbers that we're talking about in terms of adoption if we go and look at daps and nonfilter it looks like if you look at the one on the bottom there piggy box seems to be a very exciting application uh so is this the killer dab we've all been waiting for well no not really well piggy box is something that is about just like a lottery box that you want to uh you want we look we see that all the linked accounts that are using that have just one user op and they're just trying to do like some sort of like cble uh I don't want to say Cil attack but they're kind of like trying to fight that they want to get as many user Ops as possible from different users because then you get some sort of Lottery card and it's not well look it's an interesting experiment I don't want to badmouth any dap out there but it's I don't think it's that interesting in terms of what we're imagining what ethereum can eventually bring to users so if you look at the real daps in a sense I'm using real air quot you see that things like an RPG game an nft Trading Card Game social apps things that are what we more have in mind when we think about the use cases for ethereum and look if we look at the that RPG game you know we're starting to see games that are not blockchain that are adding games but games that are adding blockchain there is even a rumor that GTA 6 might have you know may adopt blockchain technology this not some this is what I read online no Alpha here or anything like that but it would be very exciting if that were to happen so and now thanks to things like user you know the user experience we can provide that might actually be a reality um talking about money um we're seeing growth in the terms of how much bunders are making how much pay how much pay gas is sponsored by pay masters and this is the Dex volume that you were seeing using 4337 accounts and we're seeing a growth because there are obvious benefits if it's easier key management if it's transaction batching if it's the pay masters there is uh uh and dexes right now are the killer application on blockchain so seeing a growth there is very very interesting and but if you look at the Gras prices themselves you look at at the trend we see that taking away that specific jump over there in June or May or whatever that is it's kind of the same but that doesn't mean there isn't growth because as vitalic mentioned in his Keynote gas prices are going down so we're seeing the number of sponsored user opts going up so don't be disillusioned by the fact that the gas is kind of staying the same it's just because the gas cost is going down which is good so we're seeing more and more sponsored user operations I want to talk about a several applications out there that have been using account abstraction H I feel like now that we are able to do stuff that is I've always been when I was talking about ethereum a lot of the time I was talking about the fact that we were waiting for this Perfect Storm of all the Technologies being together like I said maybe the killer D already was available in 2019 but we just couldn't pull it off because we didn't have the infrastructure we didn't have all the supporting standards I feel like we're getting now to that perfect storm and I want to show those case studies to just kind of explain how we got there so picnic is um is a DEX that exists that is based out out of Brazil and they make use of execute to bash transactions and they make a lot the the the ux is much simpler and it's also in a way safer it doesn't protect you completely from meev but it does help in a wayb when you have all the diff you're not signaling what you're going to be doing block by block because you're doing all the transactions in one block um it's the leading safe wallet index volume and they also received a grant from us to build an open source their defi transaction Batcher called Jam you might want to check it out and they're one of the leading you know in terms of volume by cons indexes you know in terms that they're doing pretty well H even the toll balance that they have on their wallets you can see very nice growth so there is the point of this graph is to show that there is adoption of wallets that are using 4337 it's the the usefulness is very much there and people are putting their money where their mouth is and Mana wal is also a very interesting wallet that also receives a grant from the EF uh basically turning people uh people in Japan get a government issued idea that has a chip inside that can sign transaction and they turn it into their Hardware wall thanks to account abstraction they can do this interesting key signing mechanism so they've been working closely with the government to explore other use cases and it's just like the the experience is super simple people they did an experiment where they brought a lot of people elderly people people who are not not only not crypto people but not technical people and there were all able to just easily you know people are used to doing that with their Visa card doing that with your government ID is just the same but behind the scenes you have a smart contract wallet where this is the signing key and you can have all interesting ways of recovery and stuff like that so this these are very interesting stuff that we're starting to see happening ferse you might have known them as Doo they're the official uh um collaboration app for Devcon they're using 4337 behind the scene and I find it very interesting because this is a non-financial use case so that's very fun to see um you get all the benefits of having like a web two experience for you know as a Google doc but everything is on chain you uh you know your identity is under your control the data is under your control all the benefits that we know from web 3 you can enforce access permissions on chain that leads to a lot of interoperability use cases that are very interesting we've even seen some traditional Enterprise adoption over the past two years which is a Visa you know they're the OG pay master in the way and they've been they released their own H token pay master and verifying pay master and released some articles on that Toyota has looked at you know smart contracts for the cars themselves and this is all on our our there's links in our Twitter so you can kind of scroll back and you'll see the links to those articles and Sony releasing sonum L2 you all know they're very much working with us about Native account instruction uh to sum up you know account instruction is here um there's a lot of factors that have contributed to that why we are seeing this amazing growth and it seems like we can finally offer depths you know that we don't have to compromise user experience in security and also with the release of them meole we don't also have to uh compromise on decentralization um that's it questions thank you so much Tom now time for Q&A sessions I think the first one is already there let me read it how do you see smart wallets moving towards cross-chain compat compatibility um so that's an interesting questions um I I'm trying to what I mean exactly by moving towards like how will it happen so I think H you have mentioned in this talk some interesting things that can happen once we have a con abstraction natively and that will allow a lot of cross chain ER capabilities if it's the key there's a lot of challenges when we're talking about cross chain uh to name a few is asset fragmentation is the key fragmentation and those are the things that as we have native con abstraction um we're going to see solutions that actually allow you know for users to have that experience as if they're using one chain but not behind the scenes it's going to be coordinated across those chains using the technology I hope I answer that okay I think so yes the next question how do you think um AA wallets will surpass AOA leaders like metamask Etc and what needs to be done I think it's going to be about the functionality my personal opinion is that and this is kind of touching on this is related to this so one of the challenges that we have is that when you have an AA wallet and we're seeing this kind of issue happening is that each D bring each dep onboards the user to their own smart contract wallet so there's this type of fragmentation my personal belief is that once we have this really killer application something that really takes off on just like email because when email took off signing with Google became the way that you sign into other applications so I believe that once we have that killer application which going to have an AA wallet other applications might say okay sign in with that type of account and that's going to just be something that suppresses metamask because it's going to well if they stay with EA only I'm not talking metamask specifically I'm talking EA wallets okay um because the functionality that the users are going to get through gas abstraction through trans action batching through a lot better Key Management you're not going to be I I don't see how EA can win in the long term like we I'm worried about using an eoa wallet on my dayto day and I don't see how that can continue on if we want to have mainstream adoption so by definition I think the functionality is going to win out and that's got to happen I think do you think in the future AA users know what AO address they are using well in the future how future if we're talking about future native account abstraction there there won't be an eoa address so they will not be using an EA address H for now also with current account abstraction with 4337 they also don't have an EA um so they right now they don't know if they have they don't know the EA because there isn't one like the bundler has Ana they know their their contract address but you know that's not something they're really going to see too much they're it's going to be shared behind the scenes and they're going to have like an ens linked and there's going to be some sort of like userfriendly address that they're going to be sharing with their friends just like people are used to you know URLs and they don't share IPS okay what are some user of use cases you are excited about besides pay masters um personally well if I had to list the two most things I'm excited about are pay masters and H Key Management I think the way user Ops allows you to basically add whatever signature you want it doesn't have to be just an ERC ecda signature I feel like that's that's a huge benefit because take for instance clave wallet uh another great uh 4337 wallet out there they right off the bat the way you log in is just with your uh fingerprint or face ID and that's all you know and I think that's a terrific use case for the average person so yes Key Management that's the second thing besides pay Master Okay um we still have more time for the last one will native um account obstruction be more or less profitable for bondas SL blog Builders compared to year C 4337 I think that's a question that we should have asked yavo Alex actually I'm not I don't want to give just an answer I I I I believe it's I mean yeah it's I think it's a weird yes because like they said when we're talking about like the eof situation the way bundlers are is a little different so I don't know if that's if I can answer that correctly that question no worries let me go to another one um will AA wallets gain adoption mostly non-financial use cases to avoid the risk of fund loss on new tech I think the other way around and we saw that with picnic more and more people are saying that you know they're willing to put their money into a smart contract wallet because they realize that if they lost their key somehow they can recover their account in a in a way which is still self- custodial which is still decentralized so right now if I want to use a wallet an EA wallet it's either a hardware wallet and it's super hard to use or it's something that's on my phone then it's not secure this way I can have something on my phone from a day to day but if something happen I lost my phone I'm worried I can go to my Hardware wallet and recover it so I'm gaining that everything is with a hardware wallet smaller things are with my eoa but with EA it's just like all or nothing so I feel like I should be more concerned about using EO I mean time is up but yeah I think you're still around so if people want to ask question but um thank you so much Tom I think we end of um the morning session with youf Alex and you I think we had a very great insightful three topics about the future the past and the current state of account abstraction thank you so much Tom for coming and it's time for lunch now