Datasets:

MasaFoundation

/

bankless_EARLY_ACCESS_In_Defense_of_Lido_With_Hasu

like 0

Follow

Masa Foundation 7

Speaker A: Welcome to bankless, where we explore the frontier of Internet money and Internet finance. This is how to get started, how to get better, how to front run the opportunity. This is Ryan. Sean Adams. I'm here with David Hoffman, and we're here to help you become more bankless. Lido has over 30% ETh staked right now. All of the ETh staking, 30% is in Lido. Should we be alarmed by this? Is this a threat to Ethereum's decentralization? That's where we start this episode today. But we didn't end the episode there. We also got into the more meaty topic. Is it even possible to keep Ethereum decentralized over the long run? The guest today is protocol researcher Hasu, and this is a fantastic episode. A few takeaways for you. Number one, we talk about the state of staking as it is today. What happens when an entity controls over one third of Ethereum stake? What about two thirds? Number two, we talk about the state of Lido. How can it become more decentralized? Is it on that path? Number three, we talk about this cultural phenomenon. In Ethereum. We've got idealists and we have pragmatists. There seems to be a cultural class between these two ideals of Ethereum and the reality of network effects and Moloch traps. What's the right balance? And number four, we end the episode actually where we start, which is this question, does Lido have a bad rep? Is that fair or is it unfair? David, this was a really interesting episode to me. I love having Hasu on. It's not his first time coming on bankless. Why is this episode significant to you?

Speaker B: The need for this episode has been crescendoing. I would say there has been a growing conflict with Lido, between Lido and other staking as a service organizations out there. And some of it is real and concerning. And these concerns have been elevated by some in the Ethereum leadership, some strong Ethereum community members, and other critiques of Lido I think might be narratives and might be opportunistic by other staking as a service orgs. And so we I want to really unpack this question in this episode, does Lydo deserve the bad rap that it gets? So this is the in defense of Lydo episode, where a Lido representative can come and speak on behalf of the vision for Lydo, which is not something that I don't think you and I have yet been able to articulate here on the podcast. And so what is real about the critiques of Lido and what is fake is something that we try to unpack here in this episode.

Speaker A: Yeah, and as always, of course, we have a debrief episode where David, I can't wait to talk to you about this. There's so many ideas in my brain about this. I want to get your raw thoughts that's available to bankless premium subscribers on a premium RSS feed. You can click a link in the show notes to access that. And as always, guys, this is the first step in a conversation about Lido. It's not the only step. So we'll certainly have future episodes taking the counter position to this. And as a reminder, guys, this is just one episode on Bankless about Lido, and I think we'll have future episodes to continue the conversation, maybe from some alternate perspectives. But first we disclose bankless holds investments in Lido and Rocketpool, and David and I have angel investments in some smaller staking startups. We're long term investors, not journalists. We don't do paid content, and there's a link to all bankless disclosures in the show notes.

Speaker B: We were really excited to host this conversation with Hasu. Like Ryan said, he's been on the show a number of times before. I have a ton of respect for the man and the way that he thinks. So I think this is going to be a very enjoyable conversation for the broader Ethereum ecosystem and hopefully can start many further conversations as well. So let's go ahead and get right into that conversation. But first, a moment to talk about some of these fantastic sponsors that make the show possible. Bankless nation I'm honored to welcome Hazu, an independent researcher in crypto, back to Bankless. He's been affiliated with a number of different organizations pushing the frontier of this industry. Previously research at Paradigm Governance inside of Makerdao, and more recently, strategy lead at flashbots. And also, the reason why he's here on the show today is he is also a strategic advisor to the Lido organization. Lido, of course, is a staking as a service dao that holds over 31% of all ETH staked, making it the largest staking entity in Ethereum and has been the focus of a lot of support from certain parts of crypto and also a ton of controversy from others. So today on the show, we're going to unpack some of the nuances about the Lido organization, the perceived risks, the reality of risks, all about Lido and what it means to be the largest staking as a service organization in Ethereum. Hazu, welcome back. To bankless.

Speaker C: I'm very excited to be back. David Ryan, it's a pleasure.

Speaker B: Yeah, I think this is a conversation that I think all of the crypto space, I think generally Lido gets a bad rap, and maybe some of it's justified, maybe some of it's not justified, but I'd like to actually explore the contours of that conversation and try to figure out what's reality and what's narrative and what's truth. So you ready for that?

Speaker C: Yeah, let's dive in.

Speaker B: So maybe we can start with this. Hazu, do you think Lido has an unfair rap in crypto? Like, do you think the perception of lido perhaps doesn't, it doesn't deserve the perception that it receives?

Speaker C: I think it's a nuanced question. I think that. So just to already kind of lay out what I guess the controversy is, the argument against Lido is that it's the biggest staking protocol, and because of its size, it can have a small but nonetheless possible impact on the validators of Ethereum to behave in certain ways. And there's, I would say, a secondary argument that says staked Eve is kind of the asset that is issued by the Lido protocol that represents stake on the beacon chain in a validator. And that asset is also becoming increasingly used and increasingly attractive to users, because if you compare it to Ethereum, you know, Ethereum is the best asset in the world. But if you have staked ETH, then it's Ethereum, but it's staked. So it's also earning four to 5% in ETH denominated network rewards from the beacon chain and the consensus layer. So it's like ETH, but it's also generating some return for you. And so because this asset is becoming increasingly used compared to Ethereum, I think there's also increasingly systemic concerns in the sense that if ETH is kind of the unencumbered asset, it's very unlikely that there's ever any problem with it at the smart contract level. And if there is, then it's very clear that what Ethereum should do about it. But it's not so clear about staked Eve, because staked Eve is a smart contract on Ethereum and it's a protocol that's not part of the Ethereum stack, it's something that is built on top that was developed by the free market. Right. And I think if there's a problem ever with staked Eve, then it's much harder, much harder decision that could potentially cause a rift in the Ethereum community for what to do about this, maybe.

Speaker B: To explain some of the reasons why this debate is so big, is I think if you rerolled the dice of Ethereum, you would have this debate no matter what. Just like how you said Ethereum, I totally agree with you. It's the world's greatest asset. But the only thing that's better than ether is of course, staked ether. And I think in this conversation it takes the assumption that eventually staked ether will be the dominant currency inside of Ethereum, the dominant form of collateral, simply just because it is ether plus more, plus more of itself. And so this is why this conversation naturally concludes is like which free market service provider is going to provide that version of staked ether into defi, into the ethereum economy. Would you say that that's an accurate interpretation of the future?

Speaker C: 100%. And I really like how you said that. If you reroll the dice of Ethereum and you play through the history a million times, then I think in virtually every case you would have the same debate because you had ether. And then you go to proof of stake, you create the ability to have staked ether. And this always, always creates the situation where people will of course want to delegate their stake, two node operators, and while having it on the beacon chain, they still want to use it at the execution layer, they want to trade it, they want to collateralize it. And in general they want the optionality to do with this asset to do whatever they want and not have to wait a month in the withdrawal queue. And so you basically created this situation and there's no way to get out of it. So you have to deal with the consequences. Any chain that has proof of stake a and has kind of as bias for the native token to be money has this situation, and this will play out in every single blockchain ecosystem that has these properties. And so it is a very hard fundamental question and challenge that we need to address. And there are no easy answers.

Speaker A: I think presenting it as a market challenge is the right way to frame this. Too often I feel like in crypto, clearly, sometimes there are bad guys and there are good guys in crypto, but sometimes crypto becomes overly tribal and we attack people and individuals and entities. At times we should be attacking what we call Moloch problems, game theoretic type problems. What you're illustrating here Hasu, is sort of a game theory type problem of, well, who wants this? The market wants it.

Speaker C: Why?

Speaker A: Because the market wants to stake their ETH. If they are eth holders in order to receive the yield. And they want that eth to be as fungible and as usable as possible. So we have these liquid staking tokens that are staked in ETH that we want to use across other things. And so the problem with that is it creates centralization, it creates some sort of network effect for, let's say, the pool or the apparatus or the entity that can service that market need most effectively. And this almost resembles when you were starting to illustrate the problem or the criticism of lido. It kind of sounds like a too big to fail type of argument. It sounds a little bit like the problem that Ethereum faced in the early days of a lot of ether being concentrated in the Dao.

Speaker C: I'm so glad that you said this. I was ready to make the same in my head, I wanted to bring up the same thing because you already had this once, in the sense that there was the DAO, and I think it had between 15 and 20% of all eve deposited. Right. It was this insanely popular project, and then it had a security incident, and yes, it got resolved and there was a rift, of course. And I think that left a lot of scar tissue in the Ethereum, you know, community and the soul of Ethereum not to go through a situation like that again. But what then happened afterwards was actually Vitalik said, you know, please, guys, if you make another Dao or another ICO, please cap it at $10 million. You know, don't raise any more money because we don't want to create any systemic risk.

Speaker B: $10 million or 10 million east.

Speaker C: I think. I think it was. I don't know. I think it was $10 million. I can't remember.

Speaker A: It was close to one in the same back end.

Speaker C: Some relatively. Some relatively small amount, like maybe it was 1% of ETH, maybe it was $10 million, I'm not sure anymore. But the point was then when the ICO boom happened, the first projects were basically following this guidance, and they sold out some within one or two blocks and within minutes, basically, yes, they ended up respecting this wish and prevent the systemic risk to Ethereum initially in the sale, but it had all kinds of negative effects. So first of all, it was extremely unfair. Who could contribute because only the most sophisticated whales were able to participate because they had to stay contribute in the first block, and they had to be in a very particular time zone. They had to be at their computer at the right times and so on. And then once the shares started trading in secondary market, of course, they went to where the market would have cleared anyway because the supply was not matching the demand at all. And so I think it showed that it's very hard to kind of make these rules and that the market will always find a way.

Speaker A: Yeah, I think that's a great point, Hasu, and I want to tap into kind of two points you made. So one is, it seems like you can't enforce this on the social layer. Right, Vitalik or the community saying, hey, out of the goodness of your hearts, right. We'd request that you wouldn't pool more than 1% of ETH in one smart contract location. Right. Well, that's not enforceable at the protocol.

Speaker B: Moloch doesn't listen to that.

Speaker A: Moloch does not listen to that, wrecks all of the plans. And of course, the market forces take hold. So that's one of the points you're making. The other point I think that is important to make here is that this pattern of too much eth being pooled in one place plays out and has played out in other cases. So let's say a centralized exchange or a group of centralized exchanges start to pool a lot more eth. Let's say maybe just for trading purposes. So ignore staking for a second. Let's say a coinbase and Kraken and binance held 60% of the ETH. We found out that one of those actors was like an SBF type character. Well, then the Ethereum community has this dilemma of like, wow, a large portion of the ETH supply is now controlled by a single entity. And how do we get that back? I mean, there have been other cases of this, too. I remember in the early days of Makerdao when Makerdao was just like a wrecking ball and everyone was depositing ETH inside of makerdao smart contracts. There was this question, what happens if Makerdao gets hacked? There's a lot of commentary on Twitter saying, well, now Maker Dao is so big, it's the kingmaker. If maker Dao smart contracts get hacked, well, then that would cause, like, a fork of Ethereum kind of Dao style. And now we're talking about this with Eigen layer as well. So this pattern tends to play out. And I just want to illustrate what the problem actually is. So let's say Lido right now has 30% of all staked ETh, right? And so what portion of total supply is that? Is that like, you know, it's, uh.

Speaker C: It's 30. It's 30% times 20%, because 20% is the number of all EtH that's currently being staked on the beacon chain.

Speaker A: Okay, so it's 30% something percent. So about. About five or 6%.

Speaker B: 8 million ether.

Speaker A: Okay.

Speaker B: 7.9.

Speaker A: Okay, so 8 million ether. Right. And so if something bad were to happen to that set of smart contracts, like, what's the outcome of that? Let's say an exploiter, like a curse. Say there's an EVM style issue. We've seen that this week. There's some kind of exploit or drainage of the smart contract. Like worst case scenario here, then some exploit or some black hat has all of this ether potentially. I mean, that's like, worst case scenario, what could happen? And then the question is, okay, where does the community go from there? Is that one of the core problems of kind of accruing so much eth in one place?

Speaker C: Yes, absolutely. So I think this particular risk that you outlined could not really happen because the eth cannot be withdrawn from Lido because it is staked on the beacon chain. And the beacon chain has a very long withdrawal delay. So at least there's time to deliberate what to do. So you could not have a kind of curved situation. It would be more like the Dao situation, I believe.

Speaker B: I think the real crux of the issue here isn't that there's a honey pot of ether, but it is what happens when a staking organization has a certain amount of ether staked. Specifically, there are a couple of thresholds in the mechanics of the way that proof of stake works that gives an operator who has control of enough ether over those thresholds, gives them some influence over the protocol. And I believe those numbers are 33% and 66%. Hazu, can you walk us through, what can someone do, some entity do if they have 33 or 66% of all ether staked?

Speaker C: I don't have all the numbers in my head anymore. It has to do with. It has to do with basically preventing the chain from finalizing. Something about the attestation committee. I'd have to look it up. But I mean, the overall rule is very clear. All consensus protocols rely on honest behavior to a certain threshold. And how do they encourage that? By creating the right incentives for this honest behavior to occur. In the case of ethereum, you do it basically by forcing the stakers to hold what is almost a form of equity or stake in the system itself. So if the system itself gets attacked, there's this implicit assumption that your eve would also become worth a lot less. The slashing is almost like an additional thing on top. So proof of stack would work also without slashing. And there are some, some staking protocols that don't have slashing. But of course if you can also add slashing, then you don't just have to carry it. You also have the stick. So it is kind of this extra layer of security. So that said, you're totally right. So someone who has stake above a certain threshold, they can perform various kinds of attack in theory. That doesn't mean they have the incentive to do that, of course. And the second thing that I would point out is actually, so in the case of Lido, I think this doesn't like you have to really stretch kind of the argument for this to apply in the first place, because Lido is a protocol for distributing stake from stakers to node operators. But under the hood, there's 29 different node operators right now and many more that are being onboarded right now and staking modules that are being developed. So this number is going to go up. But already today, these 29 note operators are not near the same party, nor do they hear like, nor do they listen to in many cases, you know, what the Lido protocol says, nor could Lido even unstake them today. So this is something that we can talk about. So this is not a feature that's yet enabled, this ability for the DAO to say, this node operator didn't do what we want, so we are going to exit them programmatically. Today, Lido can only say, you bad node operator, I will not give you any more stake that is coming in. We have seen in the past, in many instances, Mev boost is a great. MeV policy in general was a great example. Then there are other examples around what to do about OFAC when that happened, where we saw a lot of controversy between node operators that had very different views. They actively contributed to what the policy in given situations should be. And so it's basically not possible for Leida to say we are making a policy for what node operators should do and for node operators to just accept that policy. So they are very opinionated, they are extremely aligned with the health of the network, as is the DAO itself. Because when you look at the economic model of the Dow, it makes revenue or it makes money in perpetuity. If ethereum is healthy and ethereum thrives, and if that's not the case, then that's basically suicide for the DAO. So I think some of these, what is kind of really the risk that I am most focused on, and we can talk about that. I think we have some ideas here about how the DAO can mitigate that. But the risk that I see primarily is actually someone coming into governance from the outside, some attacker who maybe buys up a large amount of ado, or who find some other way to sneak a proposal through. So, like code injection, like, we have seen all kinds of attacks on Dao governance, right? And they actually, you know, do something to the smart contract of staked Eve. And maybe, for example, they find they print staked Eve from thin air. So there's more staked Eve now than eth on the beacon chain, and there are many safeguards. So, for example, there's a hard limit on how much Eve can be generated per day to really limit the risk that this can happen. But nonetheless, I would say this is really the biggest risk that I see on a short term basis.

Speaker B: I think that is perhaps the crux of this conversation and why we're having this conversation. It's about the relationship between Lido governance and the staked ether and its node operators. I want to really just quickly tie off the conversation about the thresholds and proof of stake, because these two things are coupled. Of course, if one staking entity passes certain thresholds of protocol control, then all of a sudden the governance of that entity becomes very important. And so that 33 and 66% that I was talking about, these are the two thresholds for proof of stake. 33%, like you said, is how you get finality. So you need 66% of all ether staked to, to be in consensus with each other in order to have finality. And so if one entity controls 33%, they influence and control finality. Not too long ago, we had this Ethereum non finality event, which was a couple of client bugs, which was an accidental non finality event. And if one staking operating entity had 33%, they would be able to induce this. If they so chose, they would have.

Speaker A: And that just means stop the blocks from finalizing, David.

Speaker B: Right. The blocks still propagate, but they are nothing. They are not beholden by the weight of ethereum proof of stake. They can be reverted without ether being slashed.

Speaker A: That's what you can do if you have over 33%.

Speaker B: If one entity has over 33%, that is what they can do. And then if you have 66%, you basically are Ethereum. That is like a 51% attack for proof of work. But in proof of stake is 66%. So if one entity had 66% of all ether stake, they get to basically choose the outcome.

Speaker A: You mean like reverse blocks or that sort of thing? Bit more. Are we getting over skis here?

Speaker B: Yeah, if I start talking any more in this direction, I'm going to trigger some of the Ethereum devs. But this is the direction directionally correct.

Speaker C: They can probably do some kind of reorganization. I would agree. What you cannot do still is make any kind of invalid state transition, because.

Speaker B: That'S the correct way.

Speaker A: Correct. Yeah.

Speaker B: You cannot sign transactions on other people's behalf. You can just do a lot else though. And so these are the risks, right? And so this is the protocol risk. And so importantly, it's worth noting the philosophy that Ethereum has about governance, which is to not have it on chain at least. And so there is no governance attack for Ethereum on chain specifically. You would have to process through all the all core dev calls. But if one staking as a service entity, perhaps Lido has more than 33%. All of a sudden that off chain governance philosophy actually becomes routed around by an on chain entity called Lido. And like you said, Hazu, there's this LDo token, which is the governance token of Lido. And so perhaps if Lido had over 33% and it currently has 31%, the off chain governance philosophy of Ethereum becomes ignored and routed around by an entity that has on chain governance with this LDO token. And so this is where this becomes systemic towards Ethereum at large, where if one entity had on chain governance like Lido does, over 33%, then all of a sudden lido gets to govern some of the rules of the protocol. I think this is part the contours of the conversation as it relates to Lido critiquers.

Speaker C: What would you say to this? I think if we want to be precise, then the concrete risk is that the Lido Dao makes some kind of governance decision, introduces some kind of policy that mandates that all of these independent, not operators who have their own completely independent motivations and stories and utility functions, that they all collude in some way, they behave in unison to destroy their own livelihood, basically, and attack Ethereum. I think I couldn't imagine how that would happen just from knowing them and none of them individually would do something like that, let alone all of them together. I would also find it incredibly difficult operationally how you would coordinate something like that.

Speaker B: Right. So inside of Lido, there are 29 node operators. And what you're saying that even if Lido governance voted to nuke Ethereum, you would still need to coordinate the 29 or maybe 27, almost the complete majority of all lido operators, you would need to coordinate them to even follow through on that?

Speaker C: Yes, because it is really, really important to understand that all of these note operators, they run their own local infrastructure, they run all of the hardware. They run the validators. They decide what block to build, they decide what block to attest. They make all of these decisions. Lido has some policies. So I'd like to encourage people to think about the Lido protocol as this thin middleware layer that gets eth from stakers. And then it has some key for how to distribute that across different node operators. And then the node operators, they basically pay the reward back into the protocol, and then 5% goes to the user, and 5% goes, sorry, 5% goes to the node operator and 5% goes to the DAO, and 90% of it goes to the user. And so what's kind of really going on here is that the Lido DaO is the organization that develops and improves the protocol and makes sure that it's secure and that it's hardened and so on. And all they can do is they can make policies for how they think the node operators should behave. But the node operators are not compelled to listen. For one, they can exit, they can just not do it. And Lido doesn't have right now any strong way to punish that. And neither today nor at any future point will they have the ability for a node operator basically to take over their hardware and say, this is the block that you should build. This is whether you should attest to another block or not. And this is, I think it's an interesting case study to understand. The most hands on that Lido has been about. This is the MEV policy. And so what the MEV policy says is there was an open, there was a period, basically, where Lido said, our proof of stake is coming. So we are now upgraded. So the beacon chain had already been live, right? And then there was the merge. Right? And so maybe six months before the merge, Lido said, okay, so we know the merge is going to happen. What does that mean? So the beacon chain validators are now going to be responsible not just for the beacon chain, but also for the execution there. What does that mean? That means there will be, you know, they will like order actual Ethereum transactions and there will be meV. And so we need to have some policy for what to do with this MeV. And, you know, should we extract meV? Should we not extract meV? If yes, what is the mechanism? Because there are a lot of questions around this. For example, if you say, well, every, every validator can do what they want, then you get into the problem of Mev hiding where the validator can just extract. The notorator just extracts the MEV and says to Lido, you know, they extract, you know, the block is three etH. And they say, oh yeah, look at this one eth block that I mined. Here you go, Lido, here's the one eth block.

Speaker B: And then they pocket two ether.

Speaker C: That would be stealing from users, right? So Lido has the responsibility to make sure that there's no way for the node operators to kind of break this like 95 five split of rewards. And they also wanted to basically look at what Ethereum wants. And so, you know, the Ethereum foundation and flashbots and the community, there was a general kind of acceptance. So we want, you know, MeV is something that exists on Ethereum. You know, there's no way to just make it go away. But what is really important is that we keep the barriers to entry into the validator set as low as possible to encourage solo staking. And so we want to make it very easy to join a permissionless and competitive market where you can outsource the building of your blocks too. Right? And so that was the idea behind proposal builder separation and MeV Boost. And so Lido was actually the first of any exchanges of any second protocols of everyone who publicly said, we will support the idea of proposal builder separation, and we will onboard Mevboost, we are mandating the use of Mevboost for our node operators. And here's like five mevboost relays that you have to include, and here's five more that you can include if you want. So the first list, the must include list that it guaranteed that it set a baseline for basically the minimum value of the block that the not operator was seeing. So Mev hiding is not a problem. But then there was also the ability for them to include additional ones. So for example, if there was, once there was the Office for Foreign Asset Control in the US had put tornado cash on their list of sanctioned transactions, then some relays were saying, I'm fine mining these transactions, and others were not. And many node operators were saying, we do absolutely not agree with the idea of using only relays, or primarily relays that do not mine these transactions. We want to mine all of the transactions. They were also given the opportunity to do that. What the Lido Dao was trying to do in that case was really make a policy that would balance the interests of all the parties, the stakers, the Ethereum community and the node operators and the research community as well, which is kind of trying to plan what the roadmap of Ethereum with regards to MeV was looking like, and this is a great example in two ways. One, I think it shows the thoughtfulness that went into this trying to really having a long discussion period. Anybody can make a proposal for a policy and then it got voted on and it was like a couple of months process. And the second thing, it shows how the lido dao makes decisions. It doesn't make any ad hoc decisions. It tries to create policies that have a lot of deliberation. And then these policies, they kind of exist in perpetuity. And maybe once per year they can be revisited and amended. And if something doesn't work, then they get updated. Right? But this is kind of the, I think, a good example for how Lido Dao makes decisions and how it thinks about these decisions.

Speaker A: So I think that's an important point for listeners to understand. If really is lido is not one kind of entity. You can't look at Lido the same way you might look at all of Coinbase's stake or binance's stake. If lido gets this right. And what you're illustrating in this example is it is a loose confederation of a whole bunch of node operators. 30. You said 29. If it gets this right, then what you essentially provide is this very lightweight, credibly neutral middleware where, for example, all node operators inside of the confederation, again, 29 of these, they have the ability to do the OFAC censoring relayer or the non OFAC censoring relayer, right? They can choose and they have full autonomy. They can also withdraw from the confederacy at any point in time if they feel like Lido governance is going in a particular direction. So if they get it right, that is what they become. And the reason I want to contrast this is, again, I want to go back to kind of the macro point here of like we are trying to slay Moloch, right, as a hard thing across crypto. I think all of the good actors this space want more decentralization. I think sometimes we get in this more lazy habit of painting things as bad or good. It's either bad or it's good, right. When reality is much more nuanced, reality is there's bad, there's good, there's better and there's best. And I want to draw the contrast between. So Lido having 31% of stake, that's not ideal. I would say the ideal is that 100% of staked ETH is all solo staked. That is the perfect ideal to maximize decentralization. We don't live in a perfect world and perfect ideal. I will say for myself, it is better to have a lido with 31% of stake than to have, let's say, not throwing one exchange under the bus, but a binance with 31% of the stake. I like that better. The reason I like that better, it's not best, but the reason I like that better is because it's a confederacy. It's kind of looser middleware versus a binance, which feels a lot more kind of like a single entity, single person controlled. It's CZ. Right. So just making that point, I think, is important at this juncture.

Speaker C: And I mean, all of the decisions and the policies that are being made and the way that the protocol functions in Lido is completely transparent. And if as a staker, you don't like it, then you can leave. If, as a note operator, you don't like it, you can leave. But in binance, as a user, you have no idea what they're actually doing, how much money they are making. You only know what they give you. And they don't even have any node operators. It's just all of their own infrastructure.

Speaker B: I think the relationship between a lido node operator and the lido Dao is really the crux of this issue, because we talk about, sure, it's an independent node operator is a part of the DAO, but it's also independent. And I think the question of how much voice does a node operator have independently versus the DAO is the big issue here. So if the DAO makes some vote that is harmful to Ethereum, can a node operator actually achieve independence from that vote? And so you said, yeah, they could just leave, and they can not be beholden by that vote, but then they stop getting paid, they stop getting the reward. And so there is some sort of enforcement, coercion, coordination that the DAO holds. So, like lido, it's not decentralized. It is an organization. It's called Lido. It has a name. It does coordinated things. And so, yes, there are a. It's a collection of parts, but we can call it a name. It's got a name that means there is an entity here. How do we think about the independence between a node operator and the influence of the DAO?

Speaker C: Yes, you are completely right. If a node operator decides to leave, then they don't get to take the stake with them. The stake is with the Lido protocol. And the Lido protocol distributes stake to different node operators, and it also aggregates the kind of the bargaining power and so on of all of these different stakers and really gives them the best terms, right? So there is a reason why Lido, you know, Lido note operators basically get 5% of the staking reward and on Coinbase they get a 25%, right? So this is, this is kind of the difference, right? Because Lido bundles the demand and thereby it has a strong bargaining power against the note operators. But you're completely right, bargaining power can also be abused. And this is why it is extremely important that the stakers also get a voice in the decisions that would affect them. So for example, if the Lido Dao wanted to say, node operators only mine empty blocks from now on. Okay, this is something very stupid, but it's just an example that would clearly be bad for Ethereum, because everybody now 31% of blocks would be censored and the capacity of Ethereum would go down. So this is, I mean, it's clearly not in the best interest of node operators who now make way less money. So they would resist, but maybe they don't have a way to resist. But for stakers, it's also clearly bad because they make less money and also they are afraid that their eth will become worthless. And so the Dido Dao has developed this new concept that's called dual governance. And it basically means that the governance responsibilities would be split between the LDO token holders and the stakers, that is the stake e folders. And it's not too dissimilar from other kind of bicameral governance systems that have come, I mean, around the same time or after that. So for example, there's the dual house system in optimism, where you have token holders, but then you also have kind of like just elected, not elected, but nominated kind of groups of community participants who, who are kind of trusted, right? So, and these are kind of meant to balance each other out. So maybe the token holders decide something bad, but then the elected officials, they can still do something, right? And you have something similar. You have it in arbitrum, for example, there's a so called security council, the ARP holders, they still have to upgrade, for example, the bridge contract. But what the Security Council can do is it's a group of, I think, eleven people. And if I think six of them vote together, then they can also kind of turbo upgrade also everything. And they can also pause the system. Why? Because sometimes it's necessary to overwrite kind of what the majority is saying, right. Or the majority can't act fast enough. And then it's also important to overwrite them for security reasons. So you do have kind of systems like that, and I think what makes them so powerful is that you really introduce new checks and balances that make the system much more robust. So the LDO token holders could no longer make any policy change that would go against the wish of, let's say, more than 2% or more than 5% of the staked e folders. And for something that is transparently bad or stupid or anything like that, it basically gives the stakers a chance to just veto the proposal and even get whoever made the proposal and voted for it slashed. So there's now a big disincentive to even making kind of a malicious proposal in lido in general Haaso.

Speaker A: So just to clarify, does this mean effectively that I think there's like over $14 billion worth of St ETH as the lido staked eth token that's out there? Does this effectively mean that, that that pool of capital all has a vote here? Is that what that means?

Speaker C: It cannot change the protocol. It can only prevent changes to the protocol.

Speaker A: You get a veto vote.

Speaker C: Exactly.

Speaker A: And what's really cool here, I will say this is one cool feature of crypto. This is why crypto organizations can operate different than corporations that we're used to seeing in the real world. This is akin to Apple iPhone users, all of them having a vote like veto power, at least over some corporate kind of decision making. So it's not just the Apple shareholders that get to dictate all of the rules. If you have an iPhone, you have some contribution over governance, and I do think you can't do that with iPhone users. Our system is not set up to do that, but we can design these types of governance systems in crypto economic protocols. And that, to me, is actually very exciting. This feature hasu, by the way, it hasn't been introduced yet. So we're talking about something in the future, is that correct?

Speaker C: Yeah. So it's something that has been in the works for, I think, nine months now. It's now in its fifth iteration. So it's progressing through the research stages. I would expect that it might go live before the year is over. So it's pretty mature already in its kind of research stage, but it's not live yet.

Speaker B: So earlier, when we were defining the problem statement, one of the problem statements that people give towards Lido is that Ethereum is this no on chain governance protocol, and then if Lido has over 33% or any staking as a service system that has governance, has over 33% of each stake. Then all of a sudden, the non on chain governance of philosophy of Ethereum becomes invalidated. And what this mechanism is doing is it is giving the staked ETH holders a veto mechanism to suppress the on chain governance nature of Lido in the instance that that needs to be expressed by staked ether holders. And so it is going closer back towards the no on chain governance philosophy of Ethereum. But I will say it doesn't get all the way there it is still, I'll call it a band aid, not a fix. And so maybe, Hazel, you can kind of talk about, you said 2% or 5% of the staked eth holders need to vote in order to veto a dao vote. My critique here is that, well, the product of staked ether is meant to be a hold and ignore type of product. And so while it is a feature that staked ether holders have the veto, it is also a bug, because now they are responsible for checking the Lido governance when they might just want to put staked ether inside of Makerdao and then go have a vacation for five years. And so I'll say that this is a step in the right direction, but I won't say the problem is of the fact that Lido now has on chain governance is fixed. How would you respond to this?

Speaker C: Yeah, I mean, if you compare it to the status quo, which is. I mean, they, they also have to monitor ado governance, but they have no way to stop it. It's a clear improvement because they still have to monitor. But they can stop it. Right. And only a small, very small quorum of them would have to watch.

Speaker B: And do we know what those numbers are? 2%, 5%.

Speaker C: So off the top of my head. So the reason I give two numbers is, I believe. So you want kind of. You want a number that is so low that it's relatively easy to trigger, but not so easy that it becomes possible to grieve the protocol and the other stakers. Yeah. So I believe the way that the fifth and latest iteration is designed is you need 2% to trigger kind of negotiation phase. And then this gives time for other staked e voters to both vote up or vote down the veto. So you can also say the veto was unjustified, and then your vote would decrease the voting power. Right. Let's walk through an example. So let's say 2% voted. This is bad. And so that's why we entered kind of the negotiation phase. Then after that, someone else comes in, two more percent comes in and says, I agree, this is bad. So it goes to 4%, and at 5%, the final veto would be triggered. But if someone now, in the meantime comes in and says, no, I think you guys are just trolling, this is actually good, and they voted against, then it might go from 4 million back to 1 billion or whatever. So it's possible to vote both up and down, and it's possible. So I might get the number slightly wrong. It's also possible that the mechanism might still change because it is under active development and it's being kind of researched and audited by multiple different parties. But I believe this is roughly the.

Speaker B: Idea, and this conversation is all in vain of discussing about what happens if Lido governance is oppressive and an individual node operator disagrees and would like to exit. This conversation happens first. It is a line of defense that allows for the node operators to not have to take that responsibility on themselves. We can stop any sort of oppressive governance vote from going through through the veto power of the staked ether token. But let's take this further and take the example of that shield. That protective barrier of the staked ether veto gets broken, and all of a sudden we do need to discuss about the coercion defection incentives between the Lido governance and a node operator. You said that if a node operator wants to leave, they can't take their staked eth with them, and so they lose their golden goose, they lose their income. Can we just talk about the parameters of that? Say I'm a node operator and there's an Lido governance vote that gets passed, and I disagree with it, and I think it's bad for Ethereum. I want to leave. Do I stop only getting new ether distributed to me by the DAO? Do I have to give up my entire supply of ether that staked in my node? What are the parameters here?

Speaker C: Yeah. So in the future, if you wanted to leave, then you would basically have to exit, or the DAO. The protocol would programmatically exit these validators from the lido.

Speaker B: DAO would programmatically exit a single lido node operator. So the DAO can exit a node operator.

Speaker C: Not today. Not today. But it's on the future roadmap that the DAO can programmatically exit the node operators. And why is this important? Because otherwise, the flip side of this happens where a node operator can misbehave and just say, what are you going to do? In order to hold the node operators accountable? It is important that the Lido protocol can also withdraw the stake. And this creates an incentive for them to be performed, which I believe that will be especially important once the number of different node operators in the protocol goes way up. Because while today we are at 29 division, in three years at least, my vision is for there to be over 5000 node operators. So the lido of the future will look a lot different than the lido today.

Speaker B: So just to really define the incentives here, the power that the Lido Dao has over a node operator is the income stream. The 5% of staking revenue that the node operator gets from all staked ether inside of its node. And so if someone is worried about the centralization effect of the Lido governance, really the power that governance has over node operators is the incentive of the income. Is there any other variable or is that it?

Speaker C: No, that's pretty much it. So it's getting access to more stake that is coming in and losing access to the stake that is already allocated to you.

Speaker A: Okay, a quick question before we talk about entrance, because I'm curious how we get to the 5000 number that you're just talking about and when that's practical. But before we get there. Okay, quick, I guess a quick question on this. Is Lido forkable today? Like is it open source? Could people fork it? So I'm wondering about a scenario, let's say a section of Lido governance and then a section of the node operators sort of turn evil and they start wanting to censor transactions, do all of these things. And there's a group of defectors who are like, we are, you know, we're the good node operators. And so we will exit, we will defect. And these are, you know, ten defectors. And so they defect. Can they defect? Fork Lido. Go play the social game on crypto Twitter and start saying, hey, these guys are censoring. This is evil Lido. We are creating a Lido classic and Lido good lido. And Lido classic, which is of course good lido with, you know, all of the more saintly and hallowed node operators. They do good Lido and it's a completely forked version and they encourage the community to exit from Lido and to stick with them. Is this a scenario that could play out? And that's why I was asking, why.

Speaker C: Is Lido forkable Lido classic? Barry Silbert's vision? So let me turn the question around. So if you, if you forked the USDC smart contract today, would it be forkable?

Speaker A: No.

Speaker C: Why not?

Speaker A: Well, because you have dollars in a bank account.

Speaker C: Okay. Yes. So Lido is the same. Yeah. So it has the eve, the on chain staked. If it only represents the staked that is also off chain from the perspective of the ethereum execution layer because it's on the beacon chain and it's in these, you know, hardware machines. Right. That are validating.

Speaker A: Yeah, but you could you fork the code and then just two step fork the code, have a different Lido classic over here, and then encourage all of the stakers inside of Lido to exit withdrawal. And I guess there's a penalty, there's a cost because everyone has to withdraw. So there's some capital costs in doing that and then redeposit into Lido classic.

Speaker B: Yeah, but then aren't you just competing with every other staking as a service Dao startup that is starting from square one?

Speaker C: Yes, but, yeah, but Ryan is right. So yes, technically that would be possible. Yeah. So it would be possible to kind of write over validators, but it would have to happen voluntarily, right. From kind of the site that remains quote unquote in power. They'd have same as they could, you know, you could fork USDC and you could say, oh please, here's my bank account. And then, you know, circa sends you half of the money. Right. So if they do that, then that works. And maybe you could even program something like that into the smart contract. I think it could be possible. I'm not that technical, but it sounds like it might be possible.

Speaker A: Well, I am not ready to launch Lido classic today, but maybe at some point, who knows? Let's ask about that question of we have 29 right now. And the concern about 29 node operators, even if it's not one entity, there's 29 is this confederation, is that it becomes kind of an OPEC like kind of a cabal. There's some kind of monopoly behavior.

Speaker B: So we can count to 29.

Speaker A: Yeah, we can count to 29. It's, you know, 2029 is better than one. But you know what's better than 29? 5000. And you mentioned the number 5000.

Speaker B: How can I can't count, if I.

Speaker A: Can lido, make it permissionless to be a node operator inside of the Lido network? And. Or how do we get to 5000?

Speaker C: Yeah, so I mean, first of all, I think instead of kind of creating an OPEC like situation or a cartel, so far, I would say Lido Dao has the policies that it makes is about breaking this kind of behavior and preventing it because that would extract money that otherwise would go to stakers. Right? So for example, these ideas around preventing MeV hiding, making a really robust policy, this all serves kind of in fact the opposite kind of anti competitive, like stop anti competitive behavior to make the market better for the stakers. But yes, thank you for the question. So today Lido has 29 node operators. In the future I said in maybe three to four years it will have 5000. So where is it coming from? Lido introduced a new version. It's called Lido V two. A couple of months ago it had two main features. So the first was it enabled withdrawals from the beacon chain. So someone who holds staked eve today, they can actually send the staked ETH back to the protocol and they can trigger an exit. And then actually the e from the beacon chain is being unstaked and it gets to them. Right. And the second part was called the staking router. So in the past there was basically this like one smart contract and anything that went into these 29 was distributed automatically to these 29 node operators. And what happened with the staking router is a fundamentally new architecture where you have kind of a new centerpiece and you can imagine it a little bit similar to like a, you know, a makerdao where there's money coming in. But then you have two different modules and they can all represent different staking strategies or different node operators. The 29 node operators, we call that. Now that's the permission set. And this permission set is curated by a committee that works for the lido DaO. It's called the node operator management committee. So Nom and maybe you have seen Izzy Doros. Izzy. So he is the leader of that committee norm. And this permission set would become one module in the staking router. But you can have many modules all coexisting with the permission module. And anyone can add new modules in a permissionless way. And then you have money coming in from stakers and there has to be now some new distribution. So basically modules would become added to the distribution function of the protocol. So if you zoom out three years, maybe there's 100 different modules and they have to be approved by the DAO. But once they are approved they have a basically a risk limit for how much if can go in. And then any if that goes in is distributed across the different modules. So what modules are possible? So we already said, okay, there's permissioned module. This is very good for basically professional node operators who have off chain reputation, who have really good performance numbers. And so it's basically the risk reward is kind of worth it. You say like we can, if you do something bad, then we know where you live, we can sue you we going to get that? It's non custodial, but you get the idea. It's like recourse. Exactly. But you can have many other modules. So there's two modules that I would highlight that will most likely be added in the next year. One is a permissionless staking module. So it's basically rocket pool, but in lido, so you can stake. Anybody can basically be an autopreader, but they have to put up a bond and then they basically get the other half of the stake and they can stake that.

Speaker A: What is that bond denominated in?

Speaker C: I don't know, it would depend. So I don't know that the module has been developed yet, so I have no insight into that, but I would assume that you do it in Eve. So when I look at Rocketpool, I always thought it's huge. Kudos to them, basically, I'm a huge fan. I think it's an incredible project and really good that it exists. I don't agree with all of the design decisions. I think the use of RPL as a collateral to me feels like it's just like shoehorning in the token in order to pump the value a little bit and make the community happy. I think it's not actually the best choice if you just want the node operating to be as efficient as possible, because you really do increase kind of the collateral that is needed. And so yeah, I think that this is something that Lido could do better, to be honest. So you only require e vessel collateral? That's what I would do. Other than that, I think it could work quite similar.

Speaker A: So that would be the permissionless pool. So we get the permission pool, the permissionless pool. Was there another type of module?

Speaker C: Yeah, so the third thing would be a blended version between the two, and there would be a module that has both professional node operators and for example, solo stakers in a permissionless way, but secured by distributed validator technology, or DVT. So what DVT is one validator is usually operated by one party, but DVT allows one validator to be operated by a committee of many parties. So, for example, you could have a validator that's secured by a committee of five different parties, and maybe two of them are permissionless and three of them are professional. And so you could kind of have a module that really blends the best of both worlds. Anybody can join it, but it has kind of the uptime and the performance of a professional note operator.

Speaker B: Yeah. The thing I like about DVT, we've called it squad staking on this show before because it's a fun name and it's an alliteration. The simple way to put it, I'd say, is say you have one single central operator who's just extremely good at their job, and then you have four hobbyists. Well, you give all the ether to the professional, but then the professional node operator, in order to actually propose a block, needs to have transactions signed by the hobbyists. And so it's a nice way for the little guy to check the power of the big guy. And you're saying that these can be blended into a single module, a single committee. And I think this V two version of Lido, the pattern of how it's designed, fits other patterns that I've seen across the crypto landscape, most recently uniswap V four. But then also Ethereum's roll up centric roadmap, where you have this main hub, and then you have things that are free to express themselves hooked into the hub. And the hub just coordinates the layer twos, for example, or the Uniswap hooks or the Lido modules. And so rather than having a monolithic Lido with 29 node operators and node operators have voted in or out, instead you just have these modules that can express themselves as they see fit. Either as a DVT system, it is a single operator. It's any sort of expression of a staking system. And it's just up to Lido governance to vote in and out these modules. So this is the pattern that I'm seeing, again, like expressed across different designs, that I'm also now seeing in Lido V two.

Speaker C: Yeah, I mean, it is a design principle in general, but in particular in lido, that we are trying to push complexity to the edges. So you want basically the core, which is the DAO and the protocol. You want this to be extremely simple and easy to understand, easy to audit, easy to secure. And if you have any operation that is more complex, maybe because it needs to be, because it's maybe more economically efficient, you basically want these to happen on top. And that's why I think you see this design pattern now emerge as a very dominant form of design pattern all across defi.

Speaker B: So with ethereums like roll up centric roadmap, a lot of the EtH maxis are like, hey, all of those layer ones are going to just become a layer two on Ethereum. Would you say that all of these, there's a bunch of startup staking as a service systems that are out there, there's the long tail would you say that if you're a Lido maximalist, and since Lido does believe in being a monopoly over staked ETH inside of Defi as a token, would you say that if the Lido maximalist take would be, hey, all of these staking as a service orgs are just going to become modules inside of Lido?

Speaker C: Yes, I think that Lido is moving now to a platform model. The staking router is basically perfect for that because Lido has the most liquid staking token and it has all of these integrations. It is the most battle hardened, trusted staking protocol. What it has with this is basically the power of distribution because Lido has this eth coming in. And this basically solves exactly the problem that if you want to design a staking protocol or like a staking provider, what you need, you need basically to get Eve. What you have is maybe the skill to use it, but what you don't have is the power to get it. And so I think this creates a huge opportunity for Lido to become a platform where other people who may today choose to launch a new staking provider to launch on their own, instead, they will build it on top of Lido and they can set, and this is something we haven't touched on, but something that the modules also allow is you can set your own economics. For example, something we see in the bonded model of rocket pool is you need to pay a higher fee to the node operator in order for them to participate. Why? For one reason, because they need to put up a bond. Yes, so they have a cost of capital. But second, because they might overall be less professional and smaller in size. And so any fixed costs they have, they're harder to amortize. And so overall, they have basically for every block they make, they have a higher cost of production. So you need to pay them more. Yes. And this is something that, for example, can be represented in a module. So the module can say, this is a permissionless module, but instead of 5%, it pays 15%, and somebody else can say, oh, this is a permission model and it pays 12%. And then maybe the DAO has to choose between the two modules. So I think the gist is there will be free competition for how to set the fee. And this allows someone to come in and build on Lido as a platform and say, this is what I do. This is the fee I charge. Please consider me for the stake distribution. And it allows new models of staking that basically haven't been possible before, also.

Speaker B: To happen on top of Lido, one of their main incentives as to why somebody would start a staking as a service DAO or organization a new LSD is because you get to issue the token. You get the LDL equivalent for that particular protocol. Right? The RPL equivalent for that particular protocol. Are you saying that these orgs would be able to retain their token as well as also being a part of the DAO?

Speaker C: Oh, so it depends what. So you said RPA, but do you mean I.

Speaker B: No, I mean the actual. So going back to the comparison between like the ethereum layer twos, like why don't all just layer ones become a layer two is because it kind of neuters the power of the layer one token if they just become a layer two, or at least the perception of it. And so one of the incentives for why one might start a new staking as a service system is because they get to issue their own governance token or their own version of RPL or their version of LDO. But if they just become a module inside of lido, is there room for an alternative token? Because that's one of the coordination defection incentives between many different LSE systems and just one LSE system.

Speaker C: Yeah, you can absolutely have your own token. So imagine you have a module where you say I'm a node operator but I use eigen layer and for example I am a sequencer for this new layer two that's about to launch. So I have another income stream. So give me some eve, you know, I will do a good job and I will also generate not just the revenue from ethereum but also the revenue from this new layer two that's about to launch. So that's a possibility, yes. And I charge 10% for that. And so there's 10% going into the, you know, going to the node operator and 5% going to Lido and in this case then it's 85% going to the user, right? And of course if you are the creator of that module, you can for example decide to distribute your 10% to your token. So in terms of how the modules are designed, you have really complete freedom. You can even have governance on top of your own module if you want that. I think if a module is, yeah, it would kind of, you know, increase the trust surface of the module, which is a consideration in whether to include the module or not. But in theory all of these things are entirely possible. And something like having a token that for example has a cash flow from your module that introduces no additional security concerns for Lido. So I think this is something that we will absolutely see in the future.

Speaker A: So I want to go back to something you were saying, you know, just, just now in this vision of Lido becoming a platform, like this is definitely the perception in the Ethereum community. Like Lido is going for it, right? They're not throttling themselves, they're not stopping at, you know, 10% staked or 20%.

Speaker B: They're not self cap or a 30%.

Speaker A: Staked or 40% staked. They are not stopping. They are going for a platform play. They are also saying, at least the case you're making is, hey, this is not as centralized as you think, even if we become a dominant, the dominant staking platform within Ethereum. Let me ask you, though, we kind of set this up as a, there are some criticisms in the community about Lido right now, and I think there's the perception among critics that this would not be a good thing. These changes that you mentioned with the different modules and, you know, permissionless pools, the ability of Lido stakers to veto and have some vote in themselves. Do you think this addresses the critics critiques, or do you think there's still something to them?

Speaker C: Ryan, in order to answer that, I want to take kind of a step back. We've already kind of said in a few different ways. So, for example, really, the Ethereum community would like for there to be maybe ten different staking protocols, and they all share the stake equally. And we would also like for there to be no way, for example, through any on chain action to influence the governance of the protocol or anything that happens at the MeV layer or the staking layer. So in my other role in flashbots, I hear the same things, just like in a different color as well, all the time. And this shows it's not a lido problem. This is kind of a general problem that Ethereum is dealing with and has to deal with, because increasingly there are things outside the protocol that are affecting how Ethereum works and, or that are growing really big to the point where they can become a systemic risk for Ethereum. So there are other protocols that also have large amounts of stake, not stake, but tokens in their smart contracts on top of Ethereum or that are like majority gas users and so on. And so overall, I think the first part of my answer is the Ethereum community has to find a way of kind of dealing with applications that are built on top of ethereum, that are growing really big, because I completely get it. It's really scary and there are no simple answers. So you cannot just kick them out or somehow throttle them in some way, because that would erode the credible neutrality of your platform, which, I mean, if ethereum loses its credible neutrality, then that would be really bad. But at the same time, you also feel like you cannot just do nothing because you see kind of the threat creeping up. And really the situation is not good. And there are real things that can happen, some of which we touched on. And so I think it's basically really complicated. So when I started thinking about liquid staking, then I kind of had a three part thesis, and I want to briefly go through it. Why? Because I think it's still really relevant today. Like, nothing pretty much has changed about this for the last two years. The first part is that there's a kind of a natural desire to divide capital and labor. So if I have stake, then I want to delegate it. I don't want to do it myself. Some people do, but they kind of, you know, they are in the minority, you know, and what that means is that at the limit, all stake will end up being delegated. So maybe today. So we saw a report, excellent report from rated.net work. And they were saying, you know, based on their estimates, there's 6% solo stakers today in the network, you know, and 94%.

Speaker A: That's it? Just 6%?

Speaker C: Yeah, yeah. And I mean, it was a high guess, so the number might be way lower than this. And so 94% professional note operators. So what happens as more users hold ether through their brokerage accounts, through exchanges, ether goes up in value. So people already get more freaked out about self custody as well and so on. I think this number actually goes down more. It doesn't go up. I think it goes down at the limit. Maybe there's 1% solo stakers. So this is definitely, I mean, a trend we have to fight, but it's not a trend that you can reverse. Like you have to deal with the fact that the number of solar operators would be very small and there will be delegation, which means you have this principal agent conflict, which is just basically a name of describing this whole conversation that we had. The second is that of all of the forms of staking that exist, liquid staking is by far the most dominant, the superior form of staking, and will, over time, crowd out all other forms. Why? Because it is effectively, it is like staking, but you get all of the liquidity and it's non custodial and you can basically turn it into any other form of staking. So if you have a liquid staking token and you can still give it someone else to custody, and then you have it like non custodial, right, but you get all of the capital efficiency and the optionality of using it however you want. And the third one is that these liquid staking tokens actually compete as money. Why? Because ether itself competes as money. Ether is the best asset in the world. People want to transfer it, they want to use it as collateral, they want to use it as defi, they love using it, they love holding it. And so the same applies to lsts. So any LST that competes in the market competes on the same terms as Eve. And that creates extreme network effects. And when you have an industry with network effects, what that means is the more people use it, the more valuable it becomes for other users also to join. So when I want to stake my ETH today, then why should I stake with staking protocol number 100, or even number ten, when I can join the biggest one that has the most liquid asset, where I can use in the most places, it is the most liquid. If I ever want to exit, it's instant. I can sell it, I can collateralize it everywhere, getting the best rates. And industries that have these network effects, they usually have very strong winner tick, most, if not winner take all dynamics. And there's virtually no counterexamples to this, you know, and you can say, well, but not in Ethereum. And I say, you know, I just don't believe you, because the other two trends are just like, they are too strong, you know, and Ethereum itself, if it was a country, then I would say, okay, well, you have like the rule of law and, you know, you have decision making. And if some company gets so big that it's like 50% of your economy, maybe you can exert some pressure over it and you can keep it in check. But the value proposition of a country is not credible neutrality, but in Ethereum it is. And so Ethereum is in this really hard place about dealing with projects that get too big. And this doesn't just apply to Lido. I mean, you mentioned it applies, I mean, it applies to eigen layer. It used to. Opensea was really big before blur came along, like uniswap, metamask. I think these are like all projects in some way, they dominate basically kind of their respective niche, and they are really powerful. And if you are like part of Ethereum, the Ethereum community, then you basically wonder, well, what should I do about them? So I remember very distinctly in my role at flashbots, metamask had 70% of all of the transactions of Ethereum. And these were not mostly sophisticated users. These were users who they are kind of. They don't really know what they're doing. Many of them, they don't really care. They wouldn't even know if, for example, Metamask did a block builder and wins every block in Ethereum and extract all of the value and then sets the price wherever they want. So they would never do that. So we know the team make well, and they incredibly ethereum aligned. I'm just saying this exists everywhere. Like this kind of rationale that you have power somewhere and there are network effects in every layer of Ethereum. And so I think this is basically, if you look at these assumptions, then for Lido to grow, to keep growing to a place where they are kind of dominant, is not just an option that is viable, that has to be thought about. If you believe in network effects, then it's the only option, because between being big and being nothing, there's nothing in between. You know, there's no place where you can be like 10% or 20% and you just do you chill, you know, it doesn't exist, you know, because like if you're 20% or 30%, if you don't keep growing, at some point someone else will grow bigger and then everybody will go to that player and then you will die. And this is, has been my view for a long time, and this is still my view today.