Security of L1 vs L2s - DeFi Security Summit 2025

Transcript

Uh thank you everybody. Uh please come in and uh we've got an amazing panel uh tonight. Uh so please uh welcome Usman, head of protocol security from Plasma. Um I was told that Usman uh didn't know what he's getting into. So that's going to be fun. Um Yan, co-founder of Circuit, one of the uh uh newer Zik rollups. Um, Vlad, head of protocol security from Metrolabs, the wizard of ZK Sync as I know him, uh, for many, many years, and one and only Lumi, um, who's apparently now at Arbitum.

>> It's an honor to be on stage with David Geta. So, [laughter] >> so uh, there is uh, this tradition uh, on my panels that I moderate that, you know, I try to make things spicy. So, I do apologize in advance, guys. You know, nothing personal. I know most of you uh really well. Uh we just want to make it entertaining for you guys and maybe uh in a way we're going to learn a thing or two. Uh so um I'll just start with a super brief intro uh of myself like I've been for many years building D5 protocol at make a DAO and then I moved on to look at the infrastructure and uh a

couple of years ago I founded L2 beats which is like a community watchdog uh for L2s and we're trying to make these L tools honest essentially we think that we all should be honest uh we should all um expose pose uh risks to users and we're at the DeFi uh security conference. So I think you should be interested uh in risks especially pertaining to DeFi. So I guess my first question um to you guys um and I start with u Lumi um like why DI users should even care about the risks uh of underlying infrastructure like you know why why

this is important to them like DI protocols are complicated as they are you know like we've seen recently with balancer uh drama um why should we even care about like L1 security versus L2 security that's sort of you know hopefully will uh set the stage uh for further discussion >> I I think Cloudflare uh downage is a good example of this right so cloudflare's down I'm working on a presentation nothing works right the same is true for web 3 but the problem is if your protocol gets hacked if the underlying layer gets hacked or you have things like reorg you might actually just lose all of your money right so I I think on a very simplistic level it's

you as a user should never have to worry about where your assets are and if you might get rugged at any point of time. So it's it's a simple answer but I think overall uh this is why we really need to pay attention to security properties of both centralized decentralized L1s as well as some of the other config like considerations for L2s that inherit some verifiability from L1 like like security um uh committees. So if uh you kind of think about such basic properties uh of chains as uh safety, livveness, reorg protection, censorship resistance, uh that type of stuff. Uh flat, what do you think are like the most important uh uh for uh D5

users? Well, I think everything is important. It um it's not like you cannot, you know, you're saying that reora not important or livveness is not important or any other properties of the blockchain. I think that you know we want to achieve the perfect security for layer 1 users. That's what we do on Ethereum and you always want to when you put money on chain you want to know that your money will be there and will be available for you at any moment of time. That's why we moving from web two security web two traditional finance to web three. um to tell a bit more about this I would say there's events in history when you start using the normal traditional banks and

at some point you just lost your access right and some people decide that you should know should now not be able to withdraw your funds and that happened in many countries Argentina is a good example of such incident but also for me I am from Ukraine and then I know that some people had a limit of withdrawals on your credit card and it is not on control of user. So we should protect this value for financial freedom for everyone that is using blockchain and if you just disallow one of these properties for the blockchain then you will lose something important. So with reworks you made some actions on chain and then at some point your action was declined which is not good. you can made

it right and then it appears to be declined right but then for livveness you want to access your funds at any moment of time and also safety is the maybe the most important part of all of these properties because you want to know that your funds will not be stolen without your proper access. So I think everything is important just a matter of in you know how you achieve it. So just to set the context for everyone um um Lumi uh essentially right now works for the biggest uh with respect to TVL um and one with the longest history rollup on Ethereum arbitum right uh and platform for ZK

syncing so these are two let's say leading uh L2s and and Yan I mean you're working at zirkuit right which um maybe could be characterized as upcoming ZK rollup, right? Like a new kit on the space, but still L2. So um like from the point of view as a founder, why have you chosen being an L2 uh on Ethereum as opposed to launch um your chain as an independent L1? >> Sure. I think uh I think um one of the big properties that is overlooked when we talk about L2s is not these core principles of blockchains that we've talked about like livveness or safety which are all important and all have a place um but actually diversity and um

differences in functionality and differences in use cases and differences in target audiences and I think if you see the last couple of years we saw other companies start L1's because they want total control uh you know over things whether or not they're saying that out loud I think that's the underlying reason a new L1 occurs these days because honestly we have battle tested tech. Uh maybe Ethereum is not sort of the beall answer. I think it is. U but you know even people who like Salana and others realize that there are chains with established user bases that they could build on. And I think the only reason you'd want to go away from that is to have this total control. And I think that's really you know interesting and it's it makes sense from a sort of a management point of view if you're an enterprise because you have

stakeholders that you want to report to and you have shareholders who want you to do the right thing. Um, and that's really, you know, good, but then there's a lot of engineering effort, and I don't think a lot of the technical stuff will necessarily improve, uh, or draw users away. Um, if you do an L1, but if you do an L2, you get most of that, almost all of that. You can still change your functionality. If you don't want to be EVM, you don't have to be EVM. If you want to do your data somewhere else, you can do your data somewhere else to some degree. U, but you still get to do a lot of this control. You can still hold a lot of the keys. Whether or not you should is another story, but you certainly can. But you can also on-ramp to the rest of this ecosystem without

you know relying on other weirder bridges. Um you know on the L2 will probably have a canonical bridge itself but also layer zero will probably just work and and that's really easy and that's really nice. So you get a lot of that control and for me it was a pretty much no-brainer that we would do an L2 not an L1 because we actually don't want to be isolated. We want to be connected to everyone else and we wanted to serve this community in a way that is unique you know for us. and um if we started our own one, I don't think we would have had that opportunity. So I I think it's very critical to to stay grounded to the people building and to bring something back to them. Um this is a very interesting point. I mean we'll come back to that like being connected to to

the overall community. So uh Usman, you're the only L1 representative. So it feels like three against one or maybe even four because I probably am hardly impartial here in this discussion. Um, so my understanding is that you don't want to be connected to this community, right? >> You know, I think a lot of the talk around being an L1 or L2 focuses on, you know, L2s have better access to liquidity. L2s are more aligned with the underlying chain. But I think from the perspective of most like retail DeFi users, there's really very little difference. like they go to layer zero front end to bridge their tokens and they don't really know that okay this chain settles to Ethereum or what

settling is or where the data layer is. Um there's this ongoing debate whether value acrru to you know Ethereum from the L2s or if it does or it doesn't. Um, I think it's easy to say that the L2s, you know, give more back to the chain or are closely aligned to it, but from the perspective of most of the users, I think it's very similar when they're on an L1 that has good integrations with the ecosystem and when they're on an L2. And again, most users don't even want to use the canonical bridges, right? Because they don't want to wait, you know, they want their bridge to be done in seconds, not days. Maybe I can add to that a little bit of like the thing that makes blockchain valuable is verifiability or decentralization right

otherwise use AWS use a bank none of this matters right like it's really easy to have financial services the thing that these properties that you mentioned bic of like livveness censorship resistance uh these all come finality all come from the fact that the underlying layer is as maximum decentralized as possible. The second you start putting more throughput on a L1, you're just kind of by definition because of blockchain trillemma or whatever you want to call it, you're diluting the value of the chain that you're using. And I think this is the reason why like users might not care, but they also can use a bank account, right? So if if you want any of the the properties that are important to users,

you need to maximize towards this decentralization. So you get actual finality of transactions and a small validators that can't just reorgg away your transaction. for banks that means you reduce counterparty risk if you have real finality not this sort of single slot finality on an alt L1 right you get censorship resistance which is an important property of blockchains and I think that's one of the reasons why like we need to scale blockchain using like actual proofs in order to have L2s built on top that's how you get the scale into a blockchain while maintaining the number one property that gives you any value in a blockchain >> oh yeah I absolutely agree I mean that's the reason that like we are here, we're all building these things and like you

know it's good that we are thinking about them but at the same time from a product perspective I think users of the like the end apps are not always fully aware of you know what's the difference between this app's running on L2 this app is running on an L1 but but I think actually where you were going to go where I was going to go which is um the end users of those apps might not but those apps themselves might they might have different requirements they might have different needs they might want something different um whether that's as simple as the the virtual machine it's running on or something more advanced or more controlling Um, but that is a reason why you would want to actually figure out what chain you're on um as an app. And you know, there's some really good examples of this if you want to pay

uh for some of your users to be free cuz they're paying in your stable coin. I mean, to me that that's an easier engineering lift if you're doing that on an L2. Like to me, it seems easier to fork a stack and make that modification than if rebuild a brand new L1, but um I know I could be wrong. I'm not I haven't built an L1. I don't know if that's true. Uh, but there are all these nice things that you can do where the users may not care and they may care a little bit about about that cost, but L2s are racing to the bottom on the cost for transactions anyway. So, I I'm not really sure anyone's put off by like subscent transactions these days. Probably there is some people, but I think we're getting so low that these other features will start to dominate

the narrative over the next 2 or 3 years because I don't know if there's a demand right now for 130 rollups that L2B tracks, right? There there will be. I think there absolutely will be but right now there's there's diversification needs >> also from a user experience point of view or user point of view the nice thing of blockchain is it has value right so like you have this decentralization trustless access you have these properties and the user doesn't really have to be aware of any of this what this really translates to is cheaper um like cheaper swap fees because there's more liquidity because those market makers don't have counterparty risk if they're using a decentralized or less counterparty risk.

The more decentralized the protocol is, you get higher rewards rates again for the same reason that you don't have counterparty risk. From a user experience point of view, like you the user doesn't have to notice necessarily, but they get the benefits of having some of the properties that blockchains offer if you design it correctly. >> Um, so maybe let's make it uh like super concrete and focus on specific properties. Uh who in the audience think that uh it's important that the underlying infrastructure provides censorship resistance?

Uh the vast majority I would say. So who thinks that L2s have the edge here and they provide a better uh sensorship resistance than L1's independent L1's? uh few who thinks that uh independent L once with uh decentralized uh validator sets provide or potentially provide a better uh censorship resistance. Okay, so that's a little bit more uh I take that the rest are not sure. So let's uh look at some uh concrete data from the recent uh balancer hack. I'm I'm sure that most of you might have heard about it. Um the balancer hack in impacted a few of the chains some of

them being L1 and some of them being L2 and it's kind of interesting to see how they reacted. So the Bara chain uh which is L1 they stopped the chain they reorged and they hard forked um as a result uh of the incident. I can only imagine that uh to do that in a very short time uh they have some uh private signal or discord channel so that they can actually coordinate. So uh while you may think they are decentralized you know it's just a a kind of a small group that can easy coordinate. Polygon uh which is an L2 but with a very independent validator set not posting data to L1. So they have at least half

an hour to to like detach uh from from Ethereum in a way. Um they uh their validators started censoring the attacker right after the incident. Sonic Independent L1 uh they did an onchain freeze. Uh and three other chains were affected. Uh you might have heard about them. Base, Optimism, and Arbitum. And uh even though they have centralized sequences uh which is very often cited as uh like one of the reasons why people say that the L1's are more uh decentralized uh they actually did nothing you know they kind of behaved as if they were Ethereum. So I'm kind of like curious like uh what your chain

would do uh if there was such a uh incident like you know how would uh what do you think is a proper like response because you know clearly uh the I think we are all divided in a way right on the one hand you know we'd like attackers to be stopped on the other hand we wouldn't like to be censored just because of what or where we are from right uh we are from different countries and I'm pretty sure that if validators don't like you know my country of origin and whatn not I I certainly wouldn't like to be uh censored I certainly would like to pres uh preserve my property rights of anything that I actually hold on this blockchain so the question the direct

question to you like what do you think would be plasma's response to similar accident >> yeah I mean in the wake of the balancer incident I mean we also have you know so on plasma we have you know a DeFi ecosystem and Balancer is one of our partners that we run and we have like internally our we have alerts set up on Balancer and many of our other partners that go to us and I'm sure Balancer of course has their own alerting that goes to them. So we probably would have found out about this rather quickly like if you know if it happened on balancer v3 which it didn't. Um and so this has been subject to a lot of discussion internally and obviously the decision is not up to only me but and I don't even mean to avoid your question but really I

don't know at this point what um action we would take in the future. I think what barer chain did and say what optimism others did both you know you can look at them and say they were reasonable. Um, and I don't even necessarily think it undermines your own chain's claims to decentralization if you manage to get the validators to coordinate. It might like that is an example of social consensus and it doesn't mean that say the foundation has the whole validator set in their hands, right? which you know I'm not saying I don't know much about bar chain specifically or the validator set but also Suie which has you know at least probably a pretty good decentralized set in the wake of their last hack I believe

they managed to hard fork and censor their attacker um I don't know it's a tough question to say right >> I mean we know what arbitum did arbitum out of the three uh is actually a stage one uh rollup meaning for uh those that don't know the details of this stage classification. Uh we can circumvent potential uh sensoring of arbitrum by forcing transactions via L1 and uh assuming that Ethereum doesn't sensor which is um realistically impossible. Uh we would get our transaction through. Um so um Lumi, do you think uh this is a

proper response? I mean this is the the systems that we should be building >> honestly like it's one of those things that as a user if you lose money that's a horrible situation for you right like you relied on a protocol bouncer in this case that you assume that should be safe right you you're not really auditing yourself so you have to trust someone in there it's super hard for the user cuz that's not a good eventuality for anybody involved other than the attacker I guess is having a nice day that there. But this is really difficult, right?

Like you go down this extremely slippery slope if you start uh reorginging transactions. You're not just affecting affecting those people that actually got hacked. Like someone might have just bought a Ferrari on your chain. And if if like this happens on the L2, now someone's out of Ferrari and the money or the the cash, right? We can't just fork away specific parts of the state. That's that's kind of one of the properties of an L2. And honestly like like I I think it's up to the op like different blockchain builders to or Dows in our case to decide on a direction but we we are trying to inherit decentralization and verifiability from

Ethereum because that is the majority like I just said is the majority of the value of blockchain. That's what makes it powerful. So I I feel very bad for the people that got affected but this is one of the things that you do when as an L2 that tries to f follow the values of the L1 itself. >> Um I guess same question to you guys with this caveat that both ZK sync and circuit right now are stage zero which in practice means that you may not be able to actually force this transaction. Uh I mean you may try maybe there is a facility but uh there is a operator can actually censor you and uh I know that uh it did happen uh in the past in Zika sync uh because the also there was a

security incident and uh the operator decided to to to censor the attacker. So, um, so I'm I'm kind of like curious, uh, is there an incentive for you guys to actually put the effort and get to stage one, which is what at least part of the Ethereum community, including Vitaling, really wants. Yes, the very short [clears throat] answer is yes, we want to get stage one. I think it's very important for the community. is like for users if they decide to go to the chain and trade on a chain and they see the stage one page is it actually I think it's very good that we have L to bit and users can be aware of these properties they can distinguish different chains and I think it just different values that you get from such

a chain right it can be good or bad values because it really depends on a scale of the incident if you imagine the incident that touch the whole ecosystem and let's say 90% of funds are affected that you definitely want to roll back from the chain and on layer one we have a social consensus that allow you to do that on layer two you don't and so with you know I think it's important to have mechanisms that will allow you to react on incidents regardless of stage one or stage zero it for me it's just a matter of toolings that we have how we coordinate internally we should be prepared for any kind of incidents and with regards to stage one I I think

we not protecting only users but we protecting oursel as well because if you are stage zero rollup you can be targeted to censor someone specifically you can be a target to get a private key of the operator to do some malicious activity what not so this important for both chains and user and I think this is uh yeah we have incentives to to get there >> what do you think about all this Yan >> yeah so I I think uh I'm going to shock the people here but uh first let me start by saying I definitely want to get to stage one but it's going going to be a a bit of a different consideration for us at Zirket. Zirket, if you don't know, was purpose-built to stop hacks like the ones that affected Balancer. We have built into our sequencer an integration with various uh security providers that

detect these hacks occurring on other networks in real time. We actually would have been able to detect Balancer if they were on our chain and stop that transaction from hitting RL2. People would not have lost money in the short term uh on circuit. And we were purpose-built because we were tired of people losing this money. Uh now that doesn't mean we should censor indefinitely or indiscriminately. I think there is um you know a place for a quarantine but even a couple of hours could have saved a lot of people some funds in the balancer incident and um building out that functionality is what we aim to do. So you know I don't want to say we'll we'll we'll uh you know never release things from quarantine but we we we would have stopped this or at least a big part of it. we know that the

security partners we used detected it. So, we would have caught it. That that transaction wouldn't have hit our chain and there would have been no complicated reorgs. It would have just been, you know, a bit of a hiccup for that attacker and everyone else would have been living a pretty good life. Um, including those people who had the money. So, um, you know, I think that's one of the perks of the L2 is that I can choose to, you know, solicit that value and and share that value and say on my chain, you don't have to worry about that and you won't have to hear platitudes of I'm sorry, you lost money in DeFi. Because I think the reality is people don't have the time to audit.

There's a whole wall of auditors behind me when this, you know, panel is over that there's big business here, which is fine, but it's very complicated, hard business. And if you're an institution trying to come on chain, I don't think an elder operator saying, "Oh, sorry, we lost your money." is going to be very happy for them. Like they have, you know, a lot of money invested in it. So, I think, um, you know, playing the role of of, you know, being a good cop if we can do it effectively, if we can do it in the right circumstances is very valuable. And L1 should not do that. I I 100% think L1 should be 100% censorship resistant. It should be sort of the wild west. Um that's what it's designed to do. But on L2, um we get all of these like every time there's a conference

like this, there's a hack before it, you know, some period of time. It's never been a safe 3 years, right? Um that's that's sickening and and I think we're going to scare people and not grow the pie unless we take pragmatic actions in the short term. And that doesn't mean, you know, Zora needs to do this in 30, 40, 50 years. I hope you know we've gotten better but in the in the interim let's help the people who we're trying to onboard. But I think that's a good point, right? So there's a distinctuation of L2s, right? Every L2 is different. We like for arbitrum ecosystem, there's arbitrum one and arbitrum one's goal is to scale Ethereum without compromising on its values cuz again like the these get hairy these situations become hairy of like who's sensorship,

right? So who's censoring? what government requires you to censor which then might affect your rights, your ci civil liberties as well as your financial freedom. It might affect like some countries or or group of people's freedom like the Canadian tra uh like tra uh trucker truckers that got all their money frozen. That's basically instilling that same value. That being said, like we have the arbitr stack and there's many different chains that configurations that you can deploy. So if you're an enterprise that needs to follow certain legal guidance, you can launch a chain that has different properties, right? So BK was referring to arbitrum one and and I think it's important to separate that there might

be like exactly like you said, there's different chains with different applications. There's different users or operators of chains that need different product market fit uh requirements. So uh let's uh move to uh to the second I think most important feature which is safety and uh my understanding is that one of the reasons why why you want to have L2 versus L1 is that it's extremely difficult to actually launch a secure uh decentralized validator set right when you are L2 you're essentially tapping to uh 300 billion worth of economic security uh whereas chains like plasma I mean I checked today uh you guys I mean

token is like bottoming uh I'm sorry to say that and uh the market cap is half a billion uh how can you secure more than five billion worth of assets with only half a billion of tokens worth security I mean is that a concern you know for you guys are you guys concerned because some of the D5 protocols have more TVL like like MA uh a um I don't know I mean I'm I'm I'm generally curious how you how you kind of think about it you Yeah, it's definitely a concern and like most, you know, new L1s or even L2s, we're undergoing a process of like progressive decentralization, right? We just launched mainet uh a couple months ago and right now we still run our validators and we're in the process of

you know going to a point where this is going to be a problem. Um but for the time being it's not the economic safety that's doing it. It's more of the operational and like software security that is doing it. very similar to you know a centralized or sorry like a single sequencer L2. >> So are you saying that there's like a closed small validator set right now running plasma? >> Yeah, right now the validator set is exactly that. >> What do you guys think about this problem? Because I don't know I'm I'm I'm genuinely curious.

>> Yeah. No, I think you have to attach to where the security is. I think you know years ago there was uh you know all these crosschain bridges and you could just exploit the weakest one and you had all L allc's in there for for a little bit. I don't know if it was a formal like era that we all agree on but you know there were smaller chains and and some of them got 51 attack and then you attack the rest of the bridge. That's that's not right. Like was always my dream by the way to to get enough Chiuawa tokens uh to get the majority of validators and go through all the Cosmos chains and like steal everything because I've got Chihuahua. I'm still secretly buying Chihuahua for all you know, right?

>> Yeah. I I think the economic security is hugely important, right? Like again, that's what contributes to decentralization. That's what contributes to removing counterparty risk. I think if you're not paying attention to that, you're putting users at risk. your the institutions that use this change should really be aware of this because for at least their risk or market makers at least for what percentage of risk they're willing to take or how their users might be affected it's it should be one of the top considerations and it goes to finality which I'm sure you'll bring up but it goes with finality too that economic security behind it >> I mean like some of you assuming are aware of governance attacks right we had

these types of attacks in the wild if you like and major D5 protocols are generally concerned uh like for example Madawo was always known from having this facility of the uh emergency shutdown if there was a attempt of a takeover I think this is hugely important frankly uh but also uh we have witnessed takeover of some chains right and that's also very interesting you know when you think about your assets uh how easy is to do that type of a governance takeover over um by malvalent parties um assuming that whoever is running the chain is always like benevolent you know may end up really bad I think but um yeah but that's just uh my take so um so you've

mentioned that for end users um if they want to tap to liquidity of the overall ecosystem uh they can use layer zero for example right you know they and this is the reality actually and I I actually agree like layer zero is hugely successful protocol. Um so uh for those that don't know how layer zero work uh this is the essentially interrup infrastructure that allows any app to define let's say they own security. The way you do it is for every single path between the different blockchain you define you know um essentially the security module if you like. uh and uh as you can imagine you know if you want

to be on 50 different blockchains you know there's like 50 square paths so as a result it's super fragile right it's extremely extremely hard to configure and because of that most of the protocols like fall back to default right so you're essentially secured by default layer zero security which is in most cases one EOA or two out of two EOAs right that's what you get uh this is what you are secured with in reality. Uh do you think this is a concern because like one of the main reasons why people choose L2 over L1 is that they can actually trustlessly tap into the whole liquidity of Ethereum, right? You know there is a way to build a trustless

bridge. Uh that's what makes you the true L2 and the rollout. There is no such way to tap into this liquidity of your L1. You have to use layer zero. Um so I'm kind of curious uh about your take here. I think it's definitely a concern for like integrators like us. Uh you have to take you know careful care to understand what these things are and you can't just sort of yolo your way into >> it's like super complicated and everybody like literally gives up. >> No. Yeah. I mean I I before I started my role at plasma I had no idea what was really involved in doing you know working with layer zero like at all the layers right but if it it it's not uniquely complicated though right we have lots of different things that are

in the stack of an L1 or an L2 that you have to make sure you configure correctly and you know have the right parameters understand what they are and I think as an implement that's just one of the many difficult things you have to do and then for the users They basically have to trust that you did that one correctly along with all the other things that they're assuming that you did correctly. And I mean they use L2B to figure out who's who's doing well. But they isn't a totally transparent or easy to understand way for normal end users to figure out whether you did this right or whether you did other things right. It's like they go on basically Twitter and like okay who seems like they're the coolest L2. All right. Yeah. Optimism arbitum.

Yeah, I heard those a lot. They're probably safe. I think that's like describes the thought process of a lot of users. >> Um, okay. So, since we're bashing layer zero, I'm sorry, by the way, layer zero, if you're here, you know, like we love you, but we need to address the elephant in the room here. Um, layer zero tokens, omni chain tokens. Um, arbitum is not just arbitrum chain, is also a stack and you can like create your own chain using the orbit stack. And at one point, Arbitum was promoting uh this strange in my opinion solution to create an orbit stack with omni chain token powered by layer zero as a gas token. What do you think about this idea Lumi?

>> This is like one one important thing to understand for bridges in general is one bridge does not equal just a bridge, right? Right. And I I was actually really happy about the announcement that you guys are like for L2B is refocusing on the bridges aspect as well cuz like users are not aware of the risk, right? And they should be and they there should be some education. So for example, like like if you have a wrapped asset bridge and you go from one to another, you carry indefinite risk. That's the user's risk. The risk is on the user. If it's intense with the LP um bridge like across then it's the risk is on the LPS and at least there are financial actors that can calculate risk understand this put that into the fee while the user

still say stays safe right going from one chain to another like besides L2s because you have a single source of truth just can't be by definition as safe as either of the protocols that it connects to right it carries them like that risk of the weakest link Right? There is like this is one of those things of like I might have my opinion on a on a design of how you should run your chain, right? But the reality is like we have the saying of your chain your rules, right? Different enterprises like USDC is a good example of that.

That's a mint and burn uh token the same way that uh the layer zero of standard is. So at that point you're basically trusting the security of circle, right? And at that point you need to be aware of that like it's not a natively minted asset. If you use USDC as your gas token like your counterparty circle, right? So there is some inherent like considerations there that like the chain operator at the least should be aware of and inform their users. It's going to show up in one form or another as fees or something else to the user. But it's it's definitely something to consider.

Like my personal favorite design is ETH as the gas token and then you can use pay masters to pay transaction fees in USDC for example. But that's my personal opinion. Zikync seems to have always been like super like firm on on the security and using Ziki proofs and with elastic chain and whatnot shared security. uh you guys seem to be taking the hard uh route and I I I wonder you know how uh what is like your take on this problem because you know uh users seem to prefer cheap and fast bridges right they don't understand you know the risks and you guys are building this super sophisticated infrastructure with

real-time ZK proving to make sure that you know all the liquidity in the whole elastic network is safe um it feels like you're climbing a by mounting the hardest wall, you know, like like the most risky uh with the most effort. >> Yeah. Okay. So, first of all, it's indeed hard. So, to build all of this infrastructure, to build the key circuits, to build all of the smart contracts that we have, I it's indeed hard like it's a lot of work and but I think it works to do it when you have a lot of bridges and they have a different economic guarantees for user. You basically kind of you're paying for this. And when I say you pay for it's a user who using the liquidity from one chain to another. They need to use a

different entities across the chain that they're using and they're not following the core like part of the blockchain. They're not using the the execution environment that they are working on right now. but instead they need to switch from one environment to another, get liquidity on another chain which just requires uh to be paid for. And so I think the cool feature about layer 2 in general is that we inherit security from layer 1. And when you when one chain inherits security from layer 1 and another chain also inherits security from layer 1, they can trust each other. That's what we're building in Ziki sync. Basically we have ecosystem of different chains and each of them can trust

implementation of each other because they're using the same code base. So there's no difference of me trusting on ziki sync or trusting the abstract chain. So then when you have also fast prover you build this you know stack of technologies and then when you have it all together you can build a fast bridge and the cool feature about fast bridge is that money is not moving from one side to another through exchanges what not but they moving through the techn through the zero knowledge basically you proving the different chain is correct and you can use it for uh transferring funds. So it's almost the same as transferring funds inside the the same chain. It's become cheap and user will like that it's cheap and fast and not that it's secure but you

get it like just you know as a side effect. I do think there's like a future direction of like ZK's design is very cool, right? But it exists in a in a silo and I like ZK gateway. I like the interop. I was part of that process, right? But I I think like you're still building a silo in for security. Yes, you it's better, right? For finality, you're still trusting ZK gateway in between. So the risk is shifted somewhere else. there is like a ultimate design for Ethereum itself to consider over a long run which is enshrining onto L1 like a proof aggregation data aggregation layer you'll probably have to have shared bridges it's going to be years away right in the meanwhile like arbitr like we the user shouldn't have

to be aware of what cluster they're in or what chain they're on they're going to want to use applications so for now like intense based design where >> spicy sorry uh intense based design where like at least the risk isn't on the user assuming you're using a safe bridge that is or like crosschain bridge then you can at least shift the the risk onto sophisticated financial actors rather than the user itself and you can use assets across any chain. I'll just add that like I agree with with most of these statements, but I think the that you know, building out stuff because it's hard um is one reason we really want to do it because everything might not be solved by more ZK proofs everywhere, but um a lot of things might be and you know, I think we're in the

dark ages of sort of simplified bridges right now that don't use a lot of the ZK tech or you other frameworks or other interop and there's there's so many of them now. I think we'll look back in 5 years and think, oh wow, how how did anyone bridge like like they did back then? Um but having someone to champion these efforts is really really important because everyone else in the ecosystem learns from them and can improve on them. So I think it's worth doing. Uh, Usman uh we don't have much time so I'll ask a strategic question direct and I know you've been uh with plasma for a relatively short time but you know this is recorded so for everyone to know do you think there's a chance plasma will uh do the move which is to like do L1 for a couple of years and eventually you

know come back to daddy and you know become an L2. I mean, I think no. Um, >> oh boy. >> Like, I'm not gonna, you know, put it on record that this will never happen, but uh I would maybe put my money there. Uh what about you guys? Uh is there a chance uh you'll do the DYDX thingy and like uh detach from Ethereum at one point because you will feel that your ecosystem is so strong that you don't need Ethereum anymore.

>> Okay. I think that we get much of Ethereum right now. That's the properties that we get from security. We get liquidity. There's so much benefits to be on on Ethereum right now. There's no way we will detach from it. This like a lot of contradiction. If you want to detach from Ethereum right now, it would be a lot of work and also it doesn't like make sense at least in the current state. I don't oversee that it can happen to any of current layer 2 that they will have so much of liquidity that you know like all of the layer one liquidity will be there. So they can start thinking about building the same as on Ethereum and at the end of the day you will need well you will need to do the same hard work and then Ethereum did in the past. So it just like doesn't

wors and it's really hard to do it socially. I I think the answer is absolutely yes, right? Like the only way if we knew a better way there would be great. Like the only way to sec like actually scale blockchain to tens of thousands, hundreds of thousands, millions of transactions which real products need is to do the L2 strategy, right? like you if you just increase the amount of throughput on a L1 blockchain like DYDX then basically you're just having to dump a crapload of it's it's not a fixed cost. It's a variable cost and Salana for example basically inflates somewhere around5 to 8 billion a year that is dumped on the community that they're not really aware of this.

at some point that starts getting unsafe or it starts getting easier to capture the economic value goes down and and you can attack the protocol that way which translates to user risk right so I I I think we'll all converge on uh it might look different it might look like sharding it might look like L2s basically the same concept the idea is somehow you get a shared layer of of security and verifiability from underlying construction Yeah, I mean I think everyone's going to stay ETH aligned as much as possible. Uh for Zirket probably that's always going to be the case. Uh but going off as an L2 to do your own thing just will add more confusion. You know I I've also heard stories of other L1's you know building

out new consensus mechanisms and then you know the exchanges come and they say well what's the tip of your chain and they can't answer it because they did something really cool and something different and it just adds technical complexity and that doesn't grow the ecosystem. So doing that is a really really really big risk in my opinion and uh it will not help us get more users because if people can't talk to each other about one unified language which I hope is Ethereum but maybe it's not who knows um we won't grow as a as a space and I think we're all here to grow the space so I I don't want people to be super different. Sorry my last question because uh we are at the time and I'll uh use it to do the shameless promotion.

Uh do you guys love or hightail to beat? So, how do you see our role in the space here? And should we like look at L1's as well? >> Yeah, I'd love if you looked at L1's. >> L2 beat is amazing. The the value they contribute to the L2 ecosystem is tangible. Like they work as coordinators between they work as the police on the beat. Not really police, but you get the point of like I really appreciate working with L2B on a regular basis. And I think it's really important to have this sort of like independent actors looking in and holding people accountable.

>> Yeah, can agree more. It's a it's amazing project and I really like that how much you contribute to the ecosystem and to the end user. >> I think you need someone taking a look at other things. So, you know, I'm glad it's you guys and glad I'm doing it. I think mostly love is the right answer. It's kind of amazing how how this community actually want uh wants to be polist right themselves. That's that's probably what makes us different from Trefi. Uh that want to do they whatever they do, you know, in very obscure way, right? And opaque. Uh at the end of the day, we all want to be transparent in what we do and uh and we want to be honest. Thank you so much. You know, that's been a pleasure. That's been lovely.

>> One question from my side. BK, when L1 beats, you don't have enough work yet. So, >> yeah. I mean, external police helps me do my job, right? I mean, look, Bartk's roasting us every day. How do So, I need L1 beat. Um yeah, we'll think about it. First, we have to fix interope. This is much uh like uh yeah, that's on our agenda at the moment. Once we fix interrupt and then we fix assets and then we fix RWAS for you guys and we fix wallets and we fix some other things, we'll go to you. >> Okay. So, next week, thank you. Cheers. Thanks so much.

[applause]