Anna Rose (00:05):
Welcome to Zero Knowledge. I'm your host, Anna Rose. In this podcast, we will be exploring the latest in zero knowledge research and the decentralized web, as well as new paradigms that promise to change the way we interact and transact online.

Anna Rose (00:27):
In this episode, we once again explore the ripple effects caused by the OFAC sanctions against Tornado Cash, where my previous episode covered the events from a legal perspective. In this one, my guest, Martin Köppelmann and I chat about the impact this has had on builders in the community. Specifically, we look at how the Tornado action helped reveal centralizing forces in Ethereum. We talk about compliance over compliant censorship and the need for decentralization. Now, before we kick off, I just wanna share an announcement for one of our partners from the last zkSummit, Anoma. Anoma has recently released a new white paper that better describes their system and architecture. It also shows how all the awesome cryptographic libraries they've been developing fit together find this paper at anoma.net/papers. We are adding the link in the show notes, so be sure to check it out. Now, Tanya will share a little bit about this week's sponsor.

Tanya (01:16):
Today's episode is sponsored by Aleo. Aleo is a new layer one blockchain that achieves the programmability of Ethereum, the privacy of Zcash, and the scalability of a rollup. If you're interested in building private applications, then check out Aleo's programming language called Leo. Leo enables non cryptographers to harness the power of ZKPs to deploy decentralized exchanges, hidden information games, regulated stable coins, and more. Visit leo-lang.org to start building. That's leo-lang.org. You can also join Aleo's incentivized Testnet 3 by downloading and running a snarkOS node. No signup is necessary to participate for questions. Join their Discord at aleo.org/discord. So thanks again, Aleo. Now here is Anna's interview with Martin from Gnosis.

Anna Rose (02:06):
Today I'm here with Martin Köppelmann, the co-founder of Gnosis. I wanna say welcome to the show, Martin.

Martin Köppelmann (02:11):
Hey, nice to be here.

Anna Rose (02:12):
Last time you were on the show, we actually talked about CoW Swap and the work that you guys were doing around that topic. This was over a year ago. Today, the goal of the show is to talk more about like Tornado. We're gonna look at like the impact on the ecosystem. This is kind of why I wanted to bring you on, but I definitely also wanna hear about the update on Gnosis, like what's happened since you were, you know, last on the show. I know a lot has gone on, so it'd be great to hear from you about that.

Martin Köppelmann (02:40):
Yeah, maybe to follow up kind of from the last episode. So last time we talked about back in the day it was probably still called Gnosis Protocol. Then kind of was turned into CoW Swap. So CoW Swap now or CoW Protocol is now its own project and became independent from Gnosis and we did this with second very large project, SAFE, just now they started SAFE now and released the SAFE token and similar with CoW Swap. So we had this model that we incubated other projects as part of Gnosis, but eventually decided, okay, they should be their own things and they kind of should give back to Gnosis and to Gnosis token holders by allocating a good chunk, usually around 20% of their new tokens to the Gnosis down. So in a way, those, from my perspective, those projects are now kind of friends of Gnosis are kind of still associated with Gnosis or kind of tightly coupled with Gnosis, but still their own projects.

Martin Köppelmann (03:44):
Gnosis itself has focused since almost a year now on Gnosis chain. So we are running, it was previously the xDai chain and we merged tokens and kind of yeah, it's now the Gnosis chain and our main focus is here, on well, decentralization and yeah, it connects to the topic. So we want to make sure that the application likeTornado Cashcan securely run on chain. So one big goal is to have a very large and very distributed validator set to really kind of have this idea of people running a validator at home in all kind of countries, that is, yeah, the largest priority of Gnosis chain.

Anna Rose (04:38):
Interesting. Is Gnosis chain like xDai? I remember that was sort of the first time I had ever used Bridges. We had Igor on years ago talking about bridging, and it was in the context with that project at the time, it was actually, I don't know what it was, it was like a lock unlocked, two smart contracts. Was it a multisig?

Martin Köppelmann (04:55):
Yeah, exactly. Exactly. So that is also kind of nice connection to zero knowledge proofs because we are in the process of upgrading those multisig bridges to zero knowledge based bridges and we are working together with a team, Succinct Labs, and there are few more teams that are doing this as well, that we're able to create light clients as your knowledge proof. So you can run a light client of Ethereum in Gnosis chain and more importantly, you can run light client of Gnosis chain in Ethereum and verify it with your knowledge proof. And so that's how we want to upgrade our bridges.

Anna Rose (05:39):
I always understood the other way, like the sort of trying to do a ZK light client on the Ethereum side as being really challenging. Why doesn't everybody already do that?

Martin Köppelmann (05:48):
Yeah, it is, it is challenge because it is challenging executive.

Anna Rose (05:51):
Okay. So makes sense.

Martin Köppelmann (05:53):
So first of all, we are in Gnosis chain, we are using the same consensus as Ethereum. So we also have a Beacon Chain. So as soon as we were able to implement a light client of that Beacon Chain, then we kind of solve both ways, because yeah, in a way it's the same challenge. It would be different if you want to bridge whatever to, I don't know, Solana or something like that, then you would need to, within Ethereum, verify Solana consensus, which might be possible, might be not, I'm not sure, but kind of here we just need to verify in a way, within Ethereum, we need to verify Ethereum consensus or well Gnosis chain consensus, but that is the same, that is also Beacon Chain. So it's also Ethereum consensus.

Anna Rose (06:38):
Interesting. Because they have the same data structure on each each side. It may, it's easier to do.

Martin Köppelmann (06:44):
Yeah or at least you kind of only have to do it for, I mean, again, I think it's potentially possible for most consensus algorithms. But the good thing if you do it for Ethereum, you can also do it certainly for Gnosis chain.

Anna Rose (07:01):
Interesting. Were you kind of waiting for the merge to happen? Like does this only work once you have POS?

Martin Köppelmann (07:06):
I think the concept could be applied to proof of work, but of course it does make sense to spend, I mean this is still this will still take some time to kind of audit and kind of to get it secure and so on. Of course, that effort, we well, obviously we anticipated proof of stake coming, so of course we directly put the effort for proof of stake. We are doing our merge as well. So kind of still right now no change still on the old consensus algorithm, the kind of xDai consensus algorithm. But we are doing kind of in a similar way, we are doing our match, and then we will have two chains that both have a Beacon Chain, Ethereum and Gnosis chain and then we can build this nice light client bridge between those two.

Anna Rose (07:47):
That is so cool. This, this really does sound like it's gonna require a full episode. Let's do that when your merge is over and we can compare merges. I feel like that would be cool. All right, now let's kind of move the topic over to Tornado. I think at this point, you know, we're recording this about six weeks, seven weeks after it initially happened. This is very on brand for the show. I often like to record things once time has passed, because then we get to do kind of a historic look back at the impact instead of trying to guess. I know in those early days there was a lot of hot takes and you, like I mentioned, had created this thread that really kind of put them together and shared a lot of the sentiment from that, like, what was your feeling maybe as it first was happening? Where did you think this was going to go?

Martin Köppelmann (08:43):
Yeah, I mean, I think it was immediately clear that this is a challenging moment for Ethereum and that in my view, it would really be a test towards how decentralized Ethereum is and, or maybe this perspective is, or my perspective is Ethereum in my view should be this global neutral settlement layer, and how easy is it to tamper with that neutrality? Because I mean, it's, of course the sanctions are one thing, but if it is possible to easily kind of block some addresses, then at least technically you could do the very same thing. Let's say you have two competing DEX's and one wants to kind of, I don't know, kick the other out of the network. So if it's possible to, kind of just easily say, okay, those kind of contracts or those kind of addresses get censored, then it's certainly at least thinkable that is applied kind of outside the context of sanctions, just in the case of like businesses competing. And in my view, that is possible, then again, it really questions this hope to have this central platform.

Anna Rose (09:54):
Do you feel like it just, like it showed us examples of how censorship happened maybe for the first time, like really in our face, and like what I saw at least was like people started to have thought exercises of what could happen. And through that we learned like, oh, this is how one could censor all of Ethereum.

Martin Köppelmann (10:14):
Yeah and I mean, unfortunately, it suddenly stopped that what could happen. I mean, it pretty quickly happened. So we are nowI mean, as we speak, we are now in a situation where at least, or where we're roughly, I would say 50% of all the blocks that are builtare censored or kind of don't include transactions to those sanctioned addresses. And it's not at all clear that this number will remain at 50% there. Quite strong indications, quite strong forces that this number might very well increase.

Anna Rose (10:52):
I mean, originally, I know there was some ideas around potentially the validators being those that censor. Especially, I mean, this was sort of interesting timing too, because this happened before the merge, and there was like about six weeks left before the merge was gonna happen. And then, yeah, there was this thought like, as we go to proof of stake, does this make it more dangerous? Like if it was still proof of work, would we be able to like be more anti-censorship or more censorship resistant? Do you think that some of the narrative around that was tied to that merge? Like do you think it was almost folks who didn't want to do the merge, who might have been very loud about this?

Martin Köppelmann (11:31):
I think it is intertwined with the question of proof stake or proof of work. But it's certainly not that would certainly be a mistake to say it's only a problem of proof of stake. So you absolutely had the same problems or similar problems with proof of work. I think it's too early to to say whether proof of stake or proof of work is more censorship resistant or leads to more centralization. I have long advocated that proof of stake can actually lead to more decentralized chain at the current point. I would probably have to admit that it's more nuanced. So I still see clear advantages in proof of stake and to name the most obvious one I can run a validator and that's actually not too hard.

Martin Köppelmann (12:29):
I mean, it's effort, but it's the absolutely doable, kind of the bar for me to previously kind of run a mining rig would be much, much higher. So that is kind of one of the main arguments for proof of stakekind of to lower the entry bar for kind of self validators in a way compared to setting up a mining rig but yes, there are certainly also elements that are problematic in proof stake and just name one is kind of this coincidence that now kind of the main validators can very easily and be custodial exchanges. So previously kind of the miners of course, had quite some centralizing role or well, I mean, there were only a handful of large mining pools and kind of even maybe even fewer basic manufacturers, but then at least they were distinct entities from exchanges.

Martin Köppelmann (13:32):
So it was a factor of decentralization, but at least you had kind of many players. And now kind of in a way, the mining pools have become the exchanges, which is certainly a problem for decentralization. That kind of, all that complexity is yeah, kind of in parallel we had this shift from, or kind of the shift towards proposal builder separation and that is not people mistake this sometimes it's not necessarily a proof of work / proof of stake issue. It's just the thing that doesn't matter whether you are the miner, the mining pool or the validator, but whoever is supposed to build the block, we have now this concept, and that was previously possible, but it just became now more popular that you outsource kind of this actual activity of collecting transactions into a block to a builder and let the builder do it for you. And that is interesting because the builder might be able to produce a block that kind of, it extracts more fees or more value. And that is something, this could be done in both proof of work, proof of state, but it just recently became very, very prominent and now kind of, that's the second force of centralization that you have potentially just small number of those builders that actually decide how and when transactions get included into blocks.

Anna Rose (15:01):
Interesting. I actually, there's a whole episode we did with the team from Flashbots just a few months ago. And this, like, in that case we talked about this PBS or what is it? Proposal Builder Separation and actually how that works. I think the builder in this case is sort of those MEV bots right? Those are the folks that you, is that the finders, the searchers or the builders? I'm sort of blanking on that actually.

Martin Köppelmann (15:28):
Previously it was called searchers and previously it was possible to kind of just kind of find so-called bundle, so kind of small sets of transactions that would extract some value, but the rest of the block was still built by the validators now kind of the dominant model has switched to builders where kind of the searcher is now a builder and actually builds the whole block. And that, that is, in my view, a big centralization force because previously it was enough to kind of be a specialist in one small area and as long as you were

Anna Rose (16:08):
But now you can combine roles

Martin Köppelmann (16:09):
Right, but you have to combine roles. You have to combine roles because you only win. It's not enough to kind of say I know about these particular DEXs and I can extract some arbitrage here, you need to propose a full block. So, you know, need to know about everything to kind of in the end have block that is winning and yeah, the consequences that currently, very few builders win almost all blocks, and for some reason, those builders have decided to not include Tornado transactions, for example, yeah.

Anna Rose (16:46):
And maybe to explain how that happened, like this is because of them using specialized MEV software, right? Where like the MEV, like the censorship is actually happening at the level of the team that's building some of the software that they're running, right? It's not them individually choosing to be censoring.

Martin Köppelmann (17:07):
Yes and no. So, so there are, again, two layers. So there are builders that actually build those full blocks and they, they send their blocks to so-called relays and currently they are a handful of relays, five to seven. And some of those relays, including the most prominent one by Flashbots themselvesthose relays sometimes require to censor. So kind of a few Flashbots would not accept a block that that includes a sanction transaction. So kind of on that level you can have censorship, but even if there is another layer that is not censoring as long as the builder still decides or the most successful builder decides to censor, then even if you ask all the relays then you will still pick the one that pays the most, so as Flashbots coincidentally also builds a successful builder, then even if you ask nonsense ring kind of relays, you will still, if you just go by who pays most, you will still in most cases pick the Flashbots built.

Anna Rose (18:17):
This is really interesting where I know that initially, so going kind of back to those initial conversations, like it really was this idea, I felt like the focus tended to be on the validator-custodial relationship. I actually want to define that a little bit broader for anyone who's not aware. Like when you talk about it and being more separated in the past, I mean, I sort of understand this because we have the ZK validators an independent, not attached to a custodial exchange validator, but maybe it's good for us to understand like what that separation used to be and how it's more tied.

Martin Köppelmann (18:50):
Yeah. So in proof of work, kind of the most successful mining pools just happened to be, usually they were miners initially themselves. So the dominant mining pools were separate from kind of centralized exchanges. Nowof course what happened is that centralized exchanges well happen to have a lot of well Ethereum kind of deposited from users. So they started offering this very convenient service to users to say to your user allow us to use your Ethereum and we will use it to validate for you and kind of give you some of or most of the rewards and of course that leads to a situation where quite a significant percentage of the current validator set of Ethereum comes from centralized exchanges.

Anna Rose (19:42):
Exactly.

Martin Köppelmann (19:42):
So those are of course now in in a quite powerful position because, I mean, exchanges always have been in a powerful position but now they got an additional power that they also build a good percentage, should be currently around 30 - 40% of the validated on Ethereum are coming from centralized exchanges.

Anna Rose (20:02):
And this may be for listeners who have tokens on exchanges, it's sort of for them to understand, like when you just say like, staking rewards, if that's like an option, usually that's just going to the validators that are run by these exchanges. So this is where we talk about that sort of centralizing factor. It's the ease of use, it's easier UX, you don't have to run your own infra You're not, I mean, in other networks where you would delegate through some other interface, you don't have to do that. It's all done for you. I mean, one interesting thing about Ethereum though is because it's like there's no delegation. Each, like there's only 32 ETH per validator. Do you feel like the average user still has some alternative if they don't wanna participate in this kind of centralization, but they don't wanna run their own infrastructure? Beause it's kind of hard to do like a full, I know you're running one, right? And it's like easy for you, but I don't think it's like easy for everyone to do 32 ETH on their own

Martin Köppelmann (20:58):
No, of course, of course and in all fairness, even if the 32 ETH would not be the problem, and just a small advertisement on Gnosis chain, you can do it with one GNO that's currently $150, you can become a validator. But even then I have to admit kind of running a validator is still, it's not ton of work. But yeah, in fairness, it is work. So you need to kind of check at least from time to time and from time to time you will need to do something. What are the alternatives? The alternatives, there are still kind of, there's a spectrum. So kind of, if you don't want to do a centralized exchange, the slightly better thing you can do is use something like liquid staking provider. Yeah. Lido is the most prominent one.

Martin Köppelmann (21:46):
I'm also there somewhat critical because actually Lido also has now roughly a third of all Ethereum. So I also consider that a force of centralization, but they certainly, I would say they're certainly in a way better than centralized exchanges because they try to do it, or they largely have set it up non-custodial, so they don't really control or they don't control the ETH and they actually try to then work with many validators, they are not even a validator themselves. They kind of just delegate it to I think other teams and entities

Anna Rose (22:25):
Who are in other jurisdictions too, probably like a little bit more around the world

Martin Köppelmann (22:28):
Right, then kind of the third thing that then kind of if you want to go more decentralized, then there are solutions like Rocketpool where the requirement is stricter or it's kind of the number of validators is much higher. So Rocketpool has a permissionless model. So in Lido you need to be selected as a validator. So at the end it's, yeah, professional entities that do it in Rocketpool, anyone can become a validator, but to do that you have to bring half of theat least bring 16 ETH yourself so it, yeah, it's certainly more decentralized, but it also scales slower because kind of people cannot just dump a lot of ETH in there. You always need kind of the same amount of Eth coming from people who actually stake themselves. So kind of those are maybe the three, the three steps and the forcethe most decentralized thing would of course then be to run your own.

Anna Rose (23:35):
So I wanna, like, the reason I wanted to clarify that too though is like you just sort of mentioned there's sort of these two places where censorship can happen. If the validators censor, like could something like Lido do the censorship or not really, would it still need to be the validators that they've delegated to actively doing it?

Martin Köppelmann (23:52):
Right, so in that sense, there are three layers. So kind of, there's one, maybe the Lido or kind of Lido is maybe the exception and I would say Lido has of course, soft powers about over their validators because they can simply say as soon as the validator doesn't do specific things. And the most obvious thing is I mean, if the validator would've, for example, funnel the rewards to themselves instead of to kind of Lido, theoretically validators could do that, but of course there's the soft power relationship where Lido can, I mean, stop using this validator essentially and and there are some details around kind of when this is exactly possible. But anyhow, so then the second layer are the validators themselves, and they generally, they are in control. However, unfortunately, many have kind of chosen to give up this control about whether or not to censor by using a builder.

Martin Köppelmann (24:55):
So if you are a validator and you basically just say, I blindly kind of, I mean, so how does it look like if you use a builder, then at some point it's your turn to build a block, but now you're just querying this API and this API gives you a hash in return, so you don't even know kind of what block you are building. You don't even know really whether you're censoring or not. You kind of just get this hash and you sign it blindly. And yeah, as it happens right now, kind of this hash will most likely come from builders that is censoring. So effectively you are in a way participating in censorship.

Anna Rose (25:33):
This is interesting and so, like, actually, just to define this a little deeper, so the custodial exchanges, are they running builders or are they actually doing sort of vanilla validation usually?

Martin Köppelmann (25:43):
From what I have seen, it's a mix. So they don't run, I mean, they don't run builders, but they, and this is all kind of still unfolding so as far as I seen it cracking is using the, as far as I can tell, basically building their own blocks so running a vanilla validatorwhile Coinbase has at least started to use builders, so Flashbots.

Anna Rose (26:07):
I see. So custodial exchanges might also be using this software, but not always, and I guess it will depend. The liquid staking validators, are those often running their own builders as well?

Martin Köppelmann (26:19):
They're never running their own builders, but they are

Anna Rose (26:22):
Using builders?

Martin Köppelmann (26:23):
They're validators to use builders and yeah, so at least for Lidothey kind of madekind of rules for validators what they are supposed to do. And yes, they are supposed to use builders, which essentially means profit maximization, because that's really what you care about.

Anna Rose (26:44):
Yeah. I mean, this is what's so interesting in teasing this out, is you are seeing like the ripple effects of one action on, like maybe unlike other, other kind of concepts or like when people have talked about this censorship resistance, like it was to me at least like somewhat theoretical, but here you can actually follow, like, here's the sanction, here's sanctioned addresses. These groups followed suit and actually like censored them. And how does that actually impact overall the ecosystem's ability to censor or like censorship resistance, basically, like, and so where, where would you say we're at now? Like what percentage, like you sort of mentioned this 50% is, would you say majority of this is because of the use of builders and it is, are you saying that like 50% of the validators that are out there are now using builders and that's why that's happened? Or are there like various forms of censorship happening?

Martin Köppelmann (27:41):
It's not yet a hundred percent clear. I mean, it's just empirically clear that that a lot of validators have started using builders and it's empirically clear that most of the time that's block built by Flashbots. Okay, now it's unclear whether Flashbots is just the most successful builder and kind of just the most profitable and people just care about the profits and therefore they always choose Flashbots because they're the most profitable, or they specifically choose Flashbots because they censor. That's not necessarily clear by just observing the current state.

Anna Rose (28:21):
Didn't Flashbots also release sort of an open source version where you could change that? So like, wouldn't people be able to be running the same thing without that censorship?

Martin Köppelmann (28:32):
Again, there are other relaysthat do not censor and they are somewhat being used. So currently 20% of the blocks that use builders come from other relays that don't, I mean that some of them also censor, but not all but I mean, that's not enough because again, you need also the successful builders. And if the most successful builders or the kind of the most, one of the most successful is by Flashbots themselves. So as long as that runs only on Flashbots and you again ask all relays, even those that don't censor, you will still pick the one that's that censors because it's the most profitable.

Anna Rose (29:16):
I see. So this impact, I mean, so far what we've talked about primarily has been this act of censorship in what you're even saying. A lot of it sounds like since a lot of these actors are it's game theory, they're looking for the biggest profit. Like they're not, you know, necessarily acting out of trying to be compliant or not. How do you think things could actually be changed to be more decentralized, like, to be more censorship resistance in, in the way it's built right now? Like, do you, would you almost call for like a competitor of Flashbots to not be Flashbots and not be doing that, but be as successful? Like is that kind of what you'd need?

Martin Köppelmann (29:54):
I would instead try to focus on solutions that makes this building process less relevant. So in my viewdecentralization would come from a lot of validators that actually produce their blocks themselves. I feel like having a large number of validators that is supposed to help against censorship resistance because you say if you can't censor them all, but if this large set of validators all just uses three builders, then yeah, all you need to do is censor those three builders. So in my view, what we need to come to is to make it possible to build your own block without, and of course, the downside currently of course it is possible to build your own block, but yes, currently it is very likely more profitable to outsource this. So in my view, kind of one core goal should be in a way to reduce MEVso to reduce the amount of money you can get extra by building the block in a specific way.

Martin Köppelmann (30:59):
And yeah, there are a handful of, or there are plenty of proposals out there how to do it, and they would all kind of tighten the freedom of how you can block. So currently you have a ton of freedom. So in your block you usually have 2 to 300 transactions and there are absolutely no rules. So you, for example, you don't have to sort them by gas fees. You don't, I mean, you can really order them in any way and every rule, every additional rule of how to how to build the block would at least reduce the freedom and therefore reduce this kind of additional leverage you have in extracting, maybe

Anna Rose (31:41):
Although I'm gonna throw back to that episode we did, because what we talked about there was like using threshold decryption or like some sort of cryptographic way to reduce MEV. And what they were saying was like, you can do that, but it's like whackamole, like you'll bump that down and MEV will pop up somewhere else.

Martin Köppelmann (31:57):
Yeah, I don't buy that. So I mean, you won't get it to zero, but you can absolutely reduce it.

Anna Rose (32:04):
And do you feel like there's not, is there not enough action on that? Do you feel like people have sort of like accepted MEV as like a fun new ecosystem and they're not really trying to fix it?

Martin Köppelmann (32:14):
Yeah. I mean, it's also quite profitable.

Anna Rose (32:16):
Yeah. Interesting. So I, I would like to kind of hear from you sort of other ripple effects that you've seen since that OFAC action and since the community has started to react.

Martin Köppelmann (32:29):
Yeah, I mean it was certainly interesting to see how different players reacted. One thing I found quite problematic was that also kind of some RPC providers have started to censor even kind of requesting or kind of requesting state of a contract, which in my view is crazy. That's almost like saying kind of you go to ISA scan and it would say you can't, we don't give you the answer of what's the state of this contractors because it's a censored one. I saw one RPC provider. If you would include kind of just the address of a kind of one of those censored addresses in any transaction. So you wouldn't even necessarily do a transaction to this address. You would kind of just do a transaction to another contract that would contain as a payload this address, even then they would, would censor that that message because it kind of contains the forbidden word kind of. So it's completely insane because like then it would mean even if you, like, let's say you wanted to create a blacklist, you wanted to comply and you wanted to create a blacklist. So you couldn't do that because, because kind of you would need to tell your contract, this contract is on the blacklist and this transaction would not be accepted,

Anna Rose (33:58):
Make it blacklisted

Martin Köppelmann (33:59):
Because it contains this forbidden thing. So the sad reality is that, yeah, that those news simply caused a panic, I would say and I think in many cases projects completely overreacted. Yeah, I mean it's understandable. It's kind of the guidance was minimal. There was kind of just this announcementand yeah, many of those practical questions were, were not answered at all. I feel like with the little bit guidance coming out from from the relevant authorities, I would say in many cases, project have completely overreacted and have done things that weren't at all required. But yeah, again, it's kind of understandable that if there's uncertainty, many decide to better be safe. In most cases, if you have Ethereum wallet, you don't run your own Ethereum full node and of course you could and ideally you would.

Martin Köppelmann (35:02):
In many cases you don't do that, but instead you point to a external node. And that is called an RPC provider and the most prominent oneis Infura, but there are plenty others anchor and more and some of those I'm not exactly sure what the state of Infura is, butI think Infura also censored transactions, I don't know, but I've seen others censoring even requests, that would just ask state of this contract certainly would censor transactions to those addresses. But yeah.

Anna Rose (35:41):
Interesting. So this, what you're saying here is like, this was not an ask of OFAC at all, and yet they sort of started to say like, just to, you know, make sure that we're not in any way infringing in any way, we're gonna kind of overstep what we actually need to do any sort of mention, even if it's like a query, not even like necessarily a like of transaction with one of those accounts. We're also gonna censor that.

Martin Köppelmann (36:05):
So as far as I understood, kind of now the situation is if you are an American citizen and you want to withdraw or you had kind of deposited previously to Tornadoas far as I understood, you can request that and, kind of, well, quite likely you are actually allowed to do it. So if that is allowed, then, I mean, then you might have practical problems actually doing it because again some of the underlyinginfrastructure might make that impossible for you.

Anna Rose (36:39):
Will they roll it back now that they know? Because like what you're talking about here is OFAC actually issued some clarity, surprisingly on like the day after I released the last episode. So we didn't have any of that info when we did it, but yeah, OFAC released an update of an FAQ where they actually did outline if you had money in Tornado and you're innocent and you can prove that, how do you actually get it out? So there are ways, but what you're saying is like even to try to access that because other members of the community, other sort of tools have blocked any address touching it, you might not be able to succeed at making that request.

Martin Köppelmann (37:16):
It's definitely not the case that Ethereum is totally censoring and it's impossible, but it just became quite a bit harder because many tools you would usually rely on might not be working. And yeah, and to your previous question, I think I hope well, I saw it in some places, but I hope it will also happen further, that some have kind of reacted immediately, maybe over reacted immediately and now are now peddling back and kind of are now doingare becoming more permissive again. But of course, it showed where the choke points are, where the forces of centralization are.

Anna Rose (37:53):
Do you think it's kind of a good thing? I know this a bit funny to say, but it's like it really, for me at least, did also shine a light on exactly where those weaknesses were and yeah, it brought up a lot of things. I feel it also, I mean personally, it also made me kind of go, okay, where do I stand on a lot of this stuff? Like that was something maybe in the past I've been, I don't know, just going along, learning about tech, not really thinking about, So it's kind of useful

Martin Köppelmann (38:19):
No, for sure, for sure. I mean, I think it's probably too early to answer that, but yeah, it certainly has kind of increased the awareness of those whole issues. It certainly has to some extent brought more attention to, I don't know, projects like All That Node, that try to make very, very easy for individual or well as easy as possible for individuals to become validators. But I think it's by no means clear at this point, whether the willingness to make Ethereumneutral platform will be kind of strong enough to overcome, yeah, certainly, simply centralization forces that clearly do exist.

Anna Rose (39:06):
It does almost seem like it creates a bit of a schism where some groups used it to identify places where they could work to decentralize further, and others were kind of moved more in the direction of like, how does one comply more?

Martin Köppelmann (39:20):
Right.

Anna Rose (39:21):
Do you actually feel that, do you feel within the community that sort of split?

Martin Köppelmann (39:25):
I feel more the split to question kind of what is an acceptable level of centralization or censorship. So a lot of the, not the guys, the Flashbot guys and kind of others who are currently out in positions ofthey say, okay, if things would get worse then, then we would do this and that, or then we wouldkind of, stop offering the service and so on, and again, I think to me the main question is, or they seem to say, yeah, as long as, I don't know, 20% of the blocks are still not censored, and that's fine. To me that's not fine, to me, kind of the current situation was like roughly 50% of the blocks being censored is absolutely not fine. But yeah, I mean, that is, that's not a black and white answer, I guess.

Anna Rose (40:17):
Yeah, I think the other thing maybe to note is like, you're not based in the US and we learned in the last episode what it means to be a US citizen in the case of these sanctions. Although I did hear like anyone around the world could still get in trouble for interacting with them, but like, especially for US citizens, it's more intense and a lot of these teams, like, do you actually see it a little bit as like a US and non-US? Are most of those teams that are complying happen to be like US entities or they have a lot of employees there?

Martin Köppelmann (40:44):
Maybe, although kind of the guy who's actually in prison in Europe, in the Netherlands and not in the US and the Tornado developers that are in the US are seem to be fine. So it's not that clear, although kind of, of course, on the sanction, on the sanction issueof course that's certainlymore problematic for US entities.

Anna Rose (41:07):
So I think it's time for us to sign off, but I I wanna ask you sort of a last thing on like what kind of advice would you give to people as, I mean, there is sort of a worry that we're gonna see more actions like this. What do you think when these things happen, how do you think people should be reacting maybe differently than they did this time around?

Martin Köppelmann (41:25):
In the actual situation? There's only so much you can do, but certainly like you can ask you the question right now in whatever position you are of how would would you act? And okay, kind of what is your power in a way and do you want to hold this power or other ways that you can give up this power?

Anna Rose (41:50):
So you're basically saying to anyone who has, who could potentially be impacting something like this in the future, those who can centralize those who can make these decisions to ask themselves if they wanna be in that position, and if not, how can they like no longer need to be in control of these things?

Martin Köppelmann (42:05):
Yep.

Anna Rose (42:06):
Basically it sounds like you're like decentralize everything.

Martin Köppelmann (42:09):
Yeah. I mean that is the one thing, at least on such importantkind of infrastructure components, I would say yes on many other things that's not required, but kind of this very core concept of does the transaction get included into a chain. I would say absolutely we need to decentralize everything.

Anna Rose (42:33):
And actually something we didn't get to talk about, and that was something that came up through this was did decentralization actually save it in this case? Not really. In turns out like decentralization can help certain parts maybe of the stack keep their neutrality. But in the case of Tornado, I mean they were also a somewhat decentralized entity. There were smart contracts, there was a DAO

Martin Köppelmann (42:56):
To be fair, it is still possible to use Tornado and in that sense, decentralization helps, but it certainly became much, much harder, if not impractical for absolutely most users.

Anna Rose (43:11):
Cool. Martin, thank you so much for coming on the show and sharing with us your perspective on this. And like I mentioned, it would be great to have you on again at some point when your merge is over so we can talk about Gnosis chain a bit more.

Martin Köppelmann (43:25):
Good. Thank you so much.

Anna Rose (43:26):
And I wanna say big thank you to the zkPodcast production team, Tanya, Henrik, and Rachel. And to our listeners, thanks for listening.