Should Gnosis Become an L2?

All right, exciting debate, I audience, kind of say a short opening, a few opening remarks, mainly about what are we even debating today. So Gnosis Chain has quite a long history. If you've been in crypto for a long time, you might actually remember that as initially launched as part of the proof of authority kind of ecosystem of chains and that one was initially a bit experimental because it used DAI as a payment token and the name at that time was xDAI. Now completely taken over by Gnosis DAO and it looks nothing like it looked initially. But it's seven years in running, and today it looks like a, let's say, small Ethereum, in the sense that it's running the same technology, it's got its own beacon chain, it's got extreme solo staking culture, which I think is very interesting.

Doesn't seem to experience the same type of issues like we have with PBS and the big builders centralizing essentially some of the block building. 300,000 nodes, that's like a lot to me. Then quite impressive application ecosystem. I think the main ones are Circle and Gnosis Pay. So it's not like, you know, you just deposit your coins into...

There's a lot of things that are happening on Gnosis Chain. And maybe two more things that are important for this debate. There's a bridge between Ethereum and Gnosis chain. At the moment, this is like four out of seven multisig, but it's extremely public in the sense that we're talking about the seven public institutions. Each transaction going in and out has to be signed by four of them.

And it's all extremely transparent. VersaBridge Explorer, you can actually see how transactions are signed. And finally, maybe also an important fact, the market cap of a token, when we think about economic security, if that's even a thing, which we might discuss a little bit later, is roughly 500 times less than Ethereum, right? So it's like, you know, there's a big difference in kind of size. So that's the chain that we're discussing today.

And yeah, so maybe we can start with positioning statements, and then I will try to add some nuance to the debate and dive really deep. So how about we start with Martin? Yeah, for sure. Yeah, so I'll try to lay out the case why so far we have decided to explicitly not become an L2. And the first thing is what's actually the core idea of being an L2?

The promise, the theoretical promise is higher security and specifically a guarantee that if you bridge assets from Ethereum to Gnosis or to an L2 that you not necessarily solve that because if we look at historically the incidence of where money was lost in large bridge hacks, then actually almost 2 billion are lost, but most of the money was not lost because of keys being compromised, which is the risk that a multistake bridge brings, but instead of code issues, so some bugs in the code. And that is not solved at all if you are an L2. And I would instead even argue that risk is higher, is even higher, because the complexity of an L2 and I would instead even argue that risk is higher it's even higher because the complexity of an L2 is much higher so in practice most L2s are not by that strict definition those completely trustless L2s in practice they actually have multi-sigs that are often called security councils I mean that's what they're called and again I think it is well it's critical to have them but if you have them anyhow it's questionable how much additional benefit you would get from being that L2. So additional point people bring up is, yeah, you need to post call data to Ethereum for data availability. I would refer to our 300,000 validators and it's not 300,000 nodes, it's 300,000 validators, so I think we believe it's 8,000 or 7,000 nodes.

But still, a significant number that I would at least say, yeah, kind of data availability is not a concern. Another argument why we have decided against it so far is in a way unpredictable costs. When we start a couple of two years ago when we already had to make those decisions, we wanted to be costs on Gnosis to be less than a cent for 100,000 gas transactions. That was always our target. And at the time, of course, it was still pre-blobs, L2s could not offer that because the data availability costs of Ethereum were too high to do that.

That has currently dramatically changed. Currently, that cost is neglectable. But that's also extremely unclear how that will develop further. So it could happen or because the cost structure is so or the supply is fixed essentially it means right now we are still always slightly below the or the demand for data availability space is still slightly below the threshold, so costs are essentially zero. As soon as we hit that, it's extremely hard to predict where costs are and where costs go, and they could very well go to a point where some of the applications that we run, for example, the Synosis Pay, that definitely, again, demands this transaction costs of less than a cent for 100,000 gas.

That would simply be completely out of our hands whether that would be the case. So, in summary, questionable whether we would actually increase security in some form, a solid home staker culture that I would say data availability concerns are not a thing. And finally, being able to have a predictable, somewhat more predictable transaction costs on the chain have been reasons to not go the L2 round despite everyone else in the ecosystem saying that is the thing to do. All right. Thank you very much, Martin.

Vitalik, please give us a positioning statement. And like I say, I mean, let's just keep to the overall ideas, and then we will go deep into specific arguments that Martin has mentioned, and maybe a few others as well. Yeah, and so I think the way that I look at this, is so that you have two chains, and one of the chain is trying to ask the, or a system that is on one of the chains is trying to ask the question, what is the correct, I could say, root of one of the other chains at a particular moment? And you have a collection of different inputs that might correspond to that in some different way, right? And one of the approaches that you could take, which is like the base case, is always a multisig.

Though the other type of input that you have is like actually some kind of proof system, right? And to me these are like basically, probably the two main options that exist. And there is also a theoretical third option, which is like Ethereum itself, like basically deciding to like recognize the other chain and fork if the other chain forks, but that one's relatively less practical, right? And if you think about those two, right? Like generally, I think, especially if you're in a regime when two different systems, or when two different inputs are medium secure, the thing that is the least likely to fail is some combination of both.

And this is something that I think under a lot of different assumptions you can mathematically prove. So I think there is value, especially when we're in this current situation where sometimes code has bugs and sometimes people with keys have bugs. And there's large instances of things going wrong with both. I think the correct thing to do across really a wide variety of contexts is to do some combination of both of those two things. And I think the L2 beat stage one definition actually is like one of those really good shelling points for something that captures a good compromise in that direction.

And then I also look at the long term, right? And to me, in the long term, technology improves, but trust does not necessarily improve, right? And basically, yeah, we've already seen technology improve over time, right? We've seen us going from extremely basic things getting hacked all the way back in 2015 and 2016 to there being this core of even DeFi that's relatively stable and hacks happen, and hacks happen on the edges, and the edges push out and push out further over time. And so I think there is a long-term future where code is the thing that actually takes over in security.

And I would even say that the thesis that code is the thing that can take over in security, at least for what we might call relatively objective domains, is kind of the underpinning of the entire crypto project, right? In the sense that if that is something that you're not willing to believe, then there's even a case that, like, actually the whole concept of smart contracts is pointless and, like, even applications should become multi-sigs, right? So I think that kind of argument is the, I mean, the reason behind Ethereum doing smart contracts in the first place, including application layer. And it's also the fact that there is this progression of code improving over time. And us now starting to have multi-provers, different proof systems, trusted hardware, various combinations and so on is one of those examples of that trend, and we're roughly somewhere in the middle of that.

And I think also on this point, right, I think the core point for me is this idea that it's not a binary, it's a spectrum. And once you think of it as a spectrum, it just feels obvious that the correct point where to be right now is actually already not zero and pretty far from zero. And I think the not binary and it is a spectrum thing also applies to the scalability side of layer twos. Right. So like I wrote this or post on a plasma about a year ago.

And one of my key points that I made on plasma is basically like, look, theoretically perfect. Nirvana plasma is hard, but if you start with a Validium, so if you just start with a system that you have stark proofs published to chain, and then you say, well, what can you add on top to protect even a little bit of people's coins if the main actor running the chain disappears, it turns out you can do quite a lot. It's very easy to protect all coins that have not moved within a week, for example, and that's the majority of altcoins. And so I think with becoming a layer 2, there's a lot of these intermediate approaches that are worth considering as well, and given Godot's scalability goals, actually full-on roll-up is one of the last steps in the roadmap I'd recommend. I'm glad you kind of raised that point because I wanted to clarify for myself and maybe also for the audience that the core issue of this debate, we're debating whether the Gnosis Chain should become an L2 and I have a feeling that the community at large has a very vague notion what L2 even is.

Because I know that there's a range of opinion, if you like, from the most strict ones that would insist that L2s should really inherit full security of Ethereum, so only rollups and maybe Plasma, right? And rollup stage two, probably. I mean, for sure not validiums, for sure no some hybrid constructions. They don't inherit full security. They shouldn't be really called L2.

So let's call these guys the extreme one spectrum. On the other side of the spectrum are people who are saying there's not even such a thing because there are only chains and bridges and there's just anything in between, right? So the concept is kind of hazy, so I'm kind of curious when you say specifically Vitalik L2, what do you really mean? Can you define that for the audience somehow? And maybe Martin thinks that L2 is something slightly different, and who knows?

Maybe he sees already an L2 by some very loose definition, right? Like lots of training wheels type of L2. Yeah, I mean, I think we have to separate the mimetic and religion aspect of this from practical technical properties. And I'm not one of these arch-pragmatists who thinks that the religion is valueless, because I think we do live in an environment where there's lots of people who love just claiming that they are Ethereum-aligned and coming to conferences and sponsoring them and trying to attract lots of users when actually it's this trust me thing operated by people who are way less scrupulous than anyone that Martin would ever be willing to even hire for any part of his team. So I think having these, I mean, like social coordination points and like basically, yeah, like making hard lines and saying, you know, you're not, like you're not allowed to call yourself this unless you have certain properties.

Like that does have a valuable social purpose, right? But I think like here, I mean, if we want to get into this narrow question of what Gnosis should do, I would say step one is we want to figure out what technical properties it makes sense for Gnosis to have at different time periods. And then question two on top of that is like basically, what do you call it? And I mean, like to me personally right now, actually question one is the more interesting one, right? And to me, the core of question one is basically that if I have an ETH on top of Gnosis chain or alternatively if I have a circle on top of either ETH or some Ethereum layer 2, let's say, I mean, we'll use Fuel as an example because it's like the one that's stage 2 and so it's nice and clean in that way.

Then, like in both cases, what are the security guarantees that I have, right? And then case one is if the Gnosis multisig is evil in the sense of signing the wrong thing. And then step two is if the Gnosis, or part two is if the Gnosis multisig misbehaves by failing to sign the right thing and actually disappearing, right? There was, I think, even an incident a couple of days ago, right, of some leader two just going offline and like the L2B, it's kind of stage two withdrawal functionality actually being crucial in helping people get their money back, right? And like I think it is a spectrum between you get zero percent of your money back and you get a hundred percent of your money back.

And like I think there are lots of 80-20s along that spectrum, right? So like the concept of a Validium, right, that Starkware talks about is like one of, just like one of those examples on the spectrum, right? And the property of the Validium has is basically that, you know, a chain commits hashes and proofs to another chain. And this is like an extra thing that gets attached to the bridge mechanism. And then you have this property that if the bridge participants do not have the power to steal, but of course, if they disappear, then everyone's coins are frozen, right?

And then on top of that, you can add extremely basic plasma mechanisms that just say like, hey, if withdrawals have not been processed for a week, then you start allowing people to use old state to withdraw, and then potentially it's like the percent of people that get their money back goes up a little further and potentially goes up quite a lot. Or if you want to go in the other direction and say, well, what if the proof system has a bug, then what happens? Then you could make the argument, well, you could, what if the proof system has a bug? Then what happens? Then you could make the argument, so like, well, you know, you could make a validity, but you could make it be overrideable with, say, a three-quarter vote of the Security Council, right?

And like, if you actually write down the numbers, like if you write down a specific probability model of like what is the probability that each individual Security Council member either disappears, whether of their own volition or as part of a collusion attack, or because they get hacked, and what is the probability that they signed the wrong thing, and then on the flip side, what is the probability of a software bug, and then you solve for parameters, then you are going to get, I think, something stage one-like where it just makes a lot of pragmatic sense as soon as you can to add a proof system and give it some share of the vote. Yeah, so it's all about the properties, right? And I think it's undoubtedly that a lot of existing L2s probably have much weaker trust assumptions than Nozisave today, but what would be your mental definition of L2s probably have much weaker trust assumptions than NosySafe today, but what would be your mental definition of L2? Maybe one response to those improvements that we can make and that we are making. So we are all about kind of what is really practically helping, and here the improvement we are all about kind of what is really practically helping.

And here the improvement we are doing to the bridge is to have light clients as assigners. And that at least brings us to a property where kind of no human actively needs to be involved to kind of, yeah, no specific human needs to be actively involved to kind of just make sure things work and then we will still have this governance or kind of security council bridge in case, but kind of just from a day-to-day operation. What is an L2? So I would here go with the strictest definition and say, yeah, you need to have the full, inherit the full security of Ethereum. And if you would do that, then you would have wonderful properties.

Then you could say all the L2s could fully trust each other and you could have, if you would deploy a token on one chain, then you could send them to one L2 and then from there to another L2 and from there to another L2 because they could fully trust the other L2. In practice, that's impossible today because we have, how many L2s In practice, that's not, that's impossible today because we have, how many L2s are you tracking? It's like way over 100 in production and the problem is that we're launching them literally every day with one click of a button. Let's say just 100 L2s. If you would in practice today create such a token that would trust every L2 as if it would inherit the security of Ethereum then what you would end up is having the weakest security of all the hundred.

So to really get to this great promise that L2s would bring I think it's a completely different approach is necessary. And if I would wear my Ethereum hat and kind of, well, Gnosis has been a project, kind of a contributor or kind of to Ethereum for, well, almost, well, 10 years now. So if I would say kind of from an Ethereum perspective, I would argue what Ethereum should do is have native L2s. And what I mean by that is L2s that are maintained by Ethereum open source developers that are all standardized. So we have, let's say, 100 instances of them.

So there are 100 L2s. They are all standardized. They all are what are called based rollups. So they just use Ethereum as a... But they don't have any multisig.

They use the Ethereum governance process of hard fox to upgrade if necessary and that would bring, well, that would bring finally the promise that in a way Ethereum had a couple of years ago with sharding when the plan was still to have I think even 1024 parallel shards that would all execute as similar to Ethereum. So in that world so that is what I would say Ethereum would do but on the flip side for Gnosis that would even mean if Ethereum would go in that route it would become even less relevant to be a non-standard L2. Because if you would have then those, if Ethereum would ever do that and have those standardized, internalized native L2s, of course, it would make it much weaker than to be a kind of non-standard L2. Is this something that you kind of like expected, that we're going to end up with hundreds of roll-ups and L2s being launched with very abysmal trust properties at the moment? Yeah, I mean, at the beginning, we were definitely worried about the opposite problem, which is making sure that we'd have at least five decent ones.

Right, I remember those days. It was a super small L2B website, and now we're struggling, keeping up with everything. Yeah. Yeah. And I think the base native roll-up topic is definitely interesting.

I mean, I think at that point, there's this question of how much actually needs to be standardized and even how much needs to be enshrined in layer one, especially in the long term for something like that to happen. I'd say, yeah, I do think that there is value in there being layer twos that are not EVM layer twos. I think, like for example, the work that Intmax has been doing with Plasma is like one, a really great example to me, right? Like they have an architecture, it's like a hybrid Plasma rollup like instruction that requires about five bytes of data to be put on chain per transaction. And somehow, because of that, they're able to be fully privacy-preserving and based at the same time, right?

And so that's an example of something really impressive that L2 is doing in a way that fully inherits the security of Ethereum without even being EVM. So I think that niche definitely exists. I mean, I'd say, yeah, I mean, from Gnosis' perspective, right, there's obviously the other question, which is in a world where Ethereum does have based roll-ups, how does Gnosis see its own long-term role in the ecosystem? And I think the answer to a lot of these questions will end up depending on that, too. So I'm curious.

I mean, first, I don't see Ethereum currently doing that unfortunately. But if Ethereum would do it, we would need to think what we do. Gnosis does a range of things. So we have the chain, but we also do various applications. Gnosis Pay or Circles, proper L2s again.

But of course, on the flip side, in that world, or one thing that's unique about Gnosis right now is that we actually have a beacon chain and that we actually have, therefore, also blobs. So it is also, if the world would demand even more block space than the block space that Ethereum could do by having 100 native L2s, then it would, of course, allow to do the same strategy for Gnosis as well with native L2s on top of Gnosis. I actually remember your argument from one of the previous conferences that you were arguing that we cannot scale Ethereum DA to handle the world's potential future demand. And also, you know, in one of your latest posts, you said that maybe we should, like, at one point, do the radical pivot with the DA if we want to handle, you know, all the world's future load. So I think, you know, this is an interesting argument.

I mean, would you... How do you see, like, long-term future, right? Can we actually scale Ethereum that way if eventually we run out of ideas how to get the DA that will have all the potential future use cases? Right. I mean, I think there the big question is basically if Ethereum is not able to have a crazy big DA, but Gnosis is, then what is the property of Gnosis that right, is that, like, I'm already predicting that, like, at least for latency and sequencing-related reasons, we're going to see L2s, like, bifurcate into, like, on the one side, L2s that are basically relatively okay with being centralized sequencers, but then having governance for rotating those sequencers around over time and basically making sure they're not abusive, sort of the standard kind of democratic structure where you have a president and the job of the vote is to kick out the president instead of voting on every single bill.

And then on the other side, L2s, they go in the other direction and actually become based. And with based roll-ups, I think, especially with native roll-ups, there's also an interesting aspect of this where there is a spectrum of how much standardization and nativeness actually happens, because there's very minimum versions of it how much standardization and nativeness actually happens, right? Because there's very minimum versions of it where basically all that you're doing is you're standardizing a precompile that does zk-snark verification. And then there's the question of... And there's even ways to design it in such a way that people can come in and basically plug in their own coprocessors for their own non-EVM features, and they just have to provide the provers only for that part instead.

But outside of that, basically, I think there is room for a lot of diversity within based roll-ups, is one of the things I'm saying. And so in that kind of world, the other natural trade-off that you get is basically that if there is a limit to the level of DA that Ethereum can safely have, then you would have L2s that just use that DA directly. And then you would have other kinds of L2s that like take this higher scale approach. And then I could easily see very realistically that they will end up using some... Basically they're using centralized sequencers to give pre-confirmations for lower latency, and then they'll also use some kind of plasmic instruction, and then possibly using some kind of alternative DA to give as much security as they can for scalable use in the meantime.

So I do expect that full spectrum to exist. I think both issues are valid. So given that we're slowly running out of time, my last question to Martin, considering that the goal is to get rid of the multisig bridge so eventually we're going to end up with the trust minimized bridge it comes down to this vague issue of social alignment if you like with Ethereum whether you're going to follow the Ethereum forks, you know, these types of questions, right? So how would you answer the question today, is Gnosis safe today? Not Gnosis safe, Gnosis chain, obviously, today aligned with Ethereum, or is this like a fully-served ecosystem?

No, no, it's 100% very aligned. I mean, so we use the same... So you don't have to be L2 to be aligned. I mean, is it for me to say whether we are with Ethereum aligned? Maybe for Ethereum to say, but we at least...

I mean, so obviously we use the same clients. We are one of the few... I'm starting to say L2. No, we are not L2 2 but we are one of the few EVM chains that actually runs, has 4 clients running it, most have only 1 or 2 we we Gnosis has half of its treasury in East so I think in many ways I think we have a strong open source ethos. Everything we do is open source.

So I think in many ways, we are value aligned with what I would say are Ethereum values. But yeah, we have still decided yet to not post our call data to Ethereum and make the claim that we are in L2 for that reason. Right, so maybe now we can take some questions from the audience, if you guys have. I see some people on the left. I think this is not...

I guess I can stand up. Hi, Martin. Hi, Vitalik. Yeah, so basically, it's a little bit more of a practical question, but, like, usually I mostly use Ethereum L1. But usually I will use, like, a different chain, like Gnosis Chain, for, like, two reasons.

Like, one, there's, like, something useful to do there, like Gnosis Pay, I think is a perfect example. And then the other one is kind of, like, a risk diversification thing, like, for my assets, no? So basically, like, like assuming that you know we get over like the like the current bridge risk and one thing that I've sometimes like asked or haven't been like a fan of in like Gnosis chain is the fact that it uses like DAI as like the fee token because basically yeah it makes it like very subject to like whatever happens happens to MakerDAO, or well, Sky now. I mean, for example, now I think it still stays DAI and whatnot, but it can already be there. So basically, my question is kind of like, for Martin, and then maybe Vitalik, you can have an opinion as well.

What do you think of using DAI as a token, and do you think it's still a good idea? And yeah, I know technically it's very difficult to swap, but yeah, just your opinion. Oh, it is possible. No, I mean, I think I was historically never a fan of the idea to force people to use your token as a form of money. And I think it's much more convenient to use stable coin or to kind of to onboard.

If you onboard with $100, you get $100 and you're immediately able to use them. So I always liked the idea to be able to pay transaction fees in stable coin. So at the point, DAI or DAI well now SUSD or still DAI was and is still the biggest somewhat decentralized stablecoin. So I think it's the best option we have. If there is a better option we would take that better option.

But that being said it's super important to note that the security of the chain does not depend on DAI. So if DAI fails, the chain doesn't fail. So, of course, if you hold DAI, you are, of course, some, I mean, if you hold DAI, you are, of course, affected, but the amount of DAI you need to hold to pay the transactions are tiny, so you can happily hold other assets on the chain and you are not affected by what happens to DAI. Should L2s use ETH as a base token, in your opinion? Most of them do.

I think generally I would say yes. I don't think ultimately the difference matters too much in an economic sense, especially now that fees are getting much lower and we're basically talking about fees being like one cent for even regular transactions. Like I think, there is an argument that it's like a mimetic thing more than anything else, but I do think transaction fees is just one of those places where ETH as an asset makes sense. Maybe it's just a UX issue. Okay, so we still have time for maybe one quick question.

Hi, my question is regarding another metric, more like a success metric for Gnosis Chain. Gnosis Chain, I would say, if you compare it to other ecosystems, has been medium successful. It doesn't have a lot of users, average users. It's just another L1. So you have the choice of becoming either just another L1, as you are now, or just another L2.

And I think, my opinion, and I would like you guys to challenge it, is it's easier to stand out as just another L2 by furthering, for example, shared liquidity or some technologies that are on the table, based roll-ups. These kinds of things make it easier to stand out and get more attention for Gnosis Chain and be more successful in terms of daily users than just being another L2. So the question is, what do you think about this different approach on why one would become an L2? Yeah, no, I do agree. If we would just become or if we would somehow do things that we can then make a plausible claim that we are now in L2.

I think that would probably not affect the attractiveness of the chain probably at all. I would say at this point. Because now we would then be L2 number 127. And maybe we rank... It's always top 10.

So yes, if we would go the L2 route, we would certainly try to do something more interesting and yes, certainly then things like synchronizing the chains to a point where you kind of were brainstorming about this idea of semi-based chain. So that would mean, so based L2 is essentially an L2 where the blocks are built by Ethereum itself. And that can bring enormous advantages because you can have synchronous transactions that affect state on Ethereum L1 and on the L2 in one transaction and you have automaticity because the L2 closely follows or one-to-one follows Ethereum state and in theory it would be possible to construct whatever you want to call it, semi-based L2 where Gnosis would do where Gnosis would synchronize its slot times to Ethereum but do six seconds instead of 12 so we would have one slot in parallel with Ethereum where we share the same state then we have one slot in between where Gnosis validators would do that independently, and then again we would have a shared slot. And that, yeah, that would potentially be something, but bottom line, just being NEL 2, I don't think that would... I think if you wanted to grab a niche of interesting technology space, then the whole EVM Plasma direction is sitting right there waiting.

And I think it's a big opportunity, because right now there's basically only one other team that's doing anything with Plasma at the moment right and I think the reason why is because like that stuff is technically hard and you have to think hard but on the other hand to have a you know a lot of respect for your guys's technical ability and so I think you actually can pull it off right and if you're going in the direction of saying like eventually you know you wants you want to have a light client on Ethereum. And then eventually you want the Security Council to go away. And then, like, it just feels like there's this natural progression where at some point you say the next step is also that the light client itself needs to sign off on, you know, transitions to the state function. And at that point, you're, like, literally a Validium already. And then there's just this natural path.

And then you're something that has actual security guarantees. And at the same time, you're not bound by the L1 blob data market. So I think it's an option that's really worth considering. Okay. Vitalik, I'll give you like four minutes maybe to try to sway voters.

We had a vote at the beginning. So we don't know how many of them voted for Gnosis to sort of like stay and wait in the sidelines, if you like. So what would be your argument or arguments, you know, for these people who voted for Gnosis not to be an L2? Yeah, I think a lot of people think of like, you know, to be or not to be an L2, you know, like that is the question as being this like very binary thing, right? And it's this, but I think as, like I think we've seen and as I've tried to argue, like not only is it a spectrum, I think it's a spectrum on which as technology progresses over time, it actually naturally makes more and more sense to start slowly shifting rightward toward being more L2-like.

And I would argue that that progression is one that's actually not too far from Martin's existing thoughts and roadmap. So basically, multsig bridge today, then actually L1, like, a light client inside of a smart contract tomorrow, then, you know, throw away the security council, like, keys, stage three, and at some point, like, actually add in validity proofs of the entire state transition, and at that point, you know, you're a validium already. And I think that is already like non-zero on the scale of L2-ness. And then there's just the natural things that it makes sense to add over time from there. And like, that's, it's like a natural progression where I think every single step makes sense.

And it's something that makes sense to just do over time as technology becomes more trustworthy, which I expect it will because there's just this massive ecosystem that benefits from that happening and is capable of executing on that happening. And if you do that, then the end state of that I think is Gnosis like having all of the properties that qualify it as an L2 in every real sense of the word and at the same time being like very close to things that I've heard you expressing interest in already. So Martin it seems like you're like trying to ride against a huge wave, if you like, right? And that's never easy. It takes courage, I think, to be contrarian.

So, for people that voted yes, Gnosis should become an L2, what would be your answer? I mean, how would you want them to reconsider? Yeah, I would say what we care about is building a credible, neutral platform, and we would say Ethereum is one, but there are not many others and too little others. And we would argue current, today's L2s are not credible neutral platforms. If you only have, or the way platforms can, in a way, extract platform power then they don't kind of create suddenly a fake post of someone.

They don't create a post under Vitalik's name that Vitalik never did. The way they extort their power is by changing API rights, by changing pricing, by changing the algorithm, all those things. And if we bring this analogy to the L2 world, then I would say L2s are not credible neutral platforms because so far they all, or essentially the big ones all, have single sequencers. And that sequencer has all kind of rights to can easily do things like, say, make specific transactions more expensive or even block them. And yes, then theoretically, you have the, as an application, you have the chance to go through one week for whatever transaction inclusion process.

But practically, your application is dead then on that chain. So long story short, what we really put a lot of effort into is build up our home validator community. Just I'm curious as a show of hand, how many people are running an Ethereum validator community. Just I'm curious as a show of hand, how many people are running an Ethereum validator here? Cool.

How many people here are running a Gnosis validator? Okay, okay. It's pretty much the same crowd. So yeah, we tried to achieve this or come as close as possible to the credible neutral platform that Ethereum is by putting lots of effort on having multiple clients, having lots of people be validators and also making a decision so far to try to as much as possible not go the proposer builder separation. So I'm not a fan of the state that Ethereum, in Ethereum 90% of all blocks currently are built by exactly two entities.

On Gnosis we still try to push validators and home validators to build blocks. So at the moment, you can even make the claim that the block building is much more diverse on Gnosis than in Ethereum. Again, we also don't have magical solutions. We might end up with the same economic pressures and so on, not making too big of a claim here. But overall, those are the things that are important to us.

And of course, practical security is important to us. And so far, we felt that slapping over this L2 claim will not will not necessarily increase practical security with all that being said of course we care about Ethereum we care about being Ethereum aligned and I will also not exclude the possibility that we will somehow go that route and at some point as we improve things, go the route Vitalik described. But bottom line, to start, it's most important to us to have our active own validator set and this strong community of people actually caring about about yeah, I mean, also kind of the cultural element of just being involved yourself and running a node and being a validator. So will your opinion change if Ethereum actually delivers, you know, ambitious roadmap with things like, you know, single slot finality and the base spring confirmations and, you know. Incursion wise.

and, you know, interoperability standards that, you know, will allow all these chains to, like, exchange messages together trustlessly. I mean, maybe it's just too early for you. I mean, I would certainly say if things change, we will, of course, try to find then a niche that we think is useful and clearly we want to if there is meaningful interoperability clearly we want to be part of that but again I feel like that's still far out unfortunately When do you expect far out, unfortunately. When do you expect or do you have a hope that interoperability standards will be a thing and once you become part of a greater ecosystem, you will benefit from it? I expect to see the basic interoperability standards start happening in the next few months, right?

And then I think from there forward, the question is how deep do you want the interoperability to be, right? Yeah. And I expect that to improve over the next years as both L1 and L2 is mature. And, of course, I expect inclusion lists and single slot finality in those topics to also get really start to become serious development like concerns once PECSRA is done as well. Beautiful.

Let us conclude this debate by voting once again and really quick, I misspoke so we're not going to take lunch after this. Martin's going to give a 15-minute talk about circles, so please stick around. All right. So you can scan the QR code and vote again, or there should be a link. If you can't scan the QR code in the back.

back and to reiterate at the beginning we had 28 agree that no sir should become an L2 we had 36 disagree and 9 abstain votes are on this I'm not sure if you can see it. Both are live on the screen. I can't see them from the stage. No. Okay.

Sorry. That's. It's looking like we have a consensus around disagree. We have 45 disagree. We have 22 agree and 8 abstaining.

9 abstaining. All right. Nine abstaining. All right. Yeah.

Okay. Everyone, please give a round of applause for the fantastic debate.