Unchained - Do You Need to Think Twice Before Restaking Your Assets? - Ep. 488

Episode Date: May 2, 2023

In this episode, Sreeram Kannan, founder of EigenLayer, and Konstantin Lomashuk, cofounder of Lido Finance, discuss the complex world of staking in Ethereum's post-Shanghai ecosystem.   Listen to the episode on Apple Podcasts, Spotify, Overcast, Podcast Addict, Pocket Casts, Stitcher, Castbox, Google Podcasts, TuneIn, Amazon Music, or on your favorite podcast platform. Show highlights: how the staking ecosystem in Ethereum looks after the full completion of the Merge what EigenLayer is and why it's called that whether EigenLayer's model is more similar to Polkadot or, actually, to Ethereum Layer 2s what the risks of restaking are, for both projects and users how Lido is working to keep Ethereum decentralized and censorship-resistant what superfluid staking is and whether it could be done on Ethereum how EigenLayer incentivizes alignment between the delegator and the stakers whether there's a risk that Lido could become a governance layer for Ethereum what can be done to increase staking services adoption Thank you to our sponsors! Crypto.com Guests: Sreeram Kannan, founder of EigenLayer Konstantin Lomashuk, cofounder of Lido Finance Previous coverage of Unchained on Ethereum staking:  Shapella in the Rearview: After Major Upgrade, What’s Next for Ethereum? How Will ETH React to Ethereum’s Shanghai Upgrade? Post-Merge, If Lido Becomes Dominant, What Does That Mean for Ethereum? In the Recent Crypto Market Meltdown, What Role Did Lido’s stETH Play? Is ETH on Its Way to Becoming Ultra-Sound Money? Yes, Says Justin Drake Learn more: How Liquid Staking Works A Guide to Ethereum’s Shanghai Upgrade How to Stake Ethereum With Liquid Staking Learn more about your ad choices. Visit megaphone.fm/adchoices

Hi everyone. Welcome to Unchained. You're a no-hype resource for all things Crypto. I'm your host, Laura Shin, author of The Cryptopians. I started covering crypto seven years ago and as a senior editor of Forbes was the first Main Street media reporter to cover cryptocurrency full-time. This is the May 2nd, 2023 episode of Unchained. If you've been enjoying Unchained, please leave us a review wherever you listen to the podcast or share this episode with a friend. and spend crypto on the crypto.com app. New users can enjoy zero credit card fees on crypto purchases in the first seven days. Download the crypto.com app and get $25 with the code Laura. Link in the description. Today's topic is staking and restaking on Ethereum. Here to discuss are Sviram Khanin, founder of Eigenlayer and Constantine Lomashuk, co-founder of P.2P.org and Lido.

Welcome, SfriRom and Constantine. Hi, Laura. pleasure to be here. Hi, Laura. We are coming off a successful completion of Ethereum's transition to proof of stake. And that was through the successful Shanghai upgrade, which enabled withdrawals of staked ether. How would you each characterize the current staking landscape today? What are the main issues and what are the next developments you think are necessary?

Why don't we start with you, Sri Rom? It's been super exciting to see both the merge and the withdrawal. enables. So now Ethereum is a complete system, you know, for all practical purposes, anything you want you can do on the system. I'm quite excited about both the merge and the withdrawals because now I think from our point of view, what it does is you can now build general purpose systems on top of Ethereum, because staking, which allows for for both positive and negative penalties to be built on top and a general purpose programming language,

which is the EVM, come together. And so I think we have this fusion of like two superpowers. One is the incentive mechanism and the programming ability. So that's something I'm excited about. And I think there are a lot of very interesting developments happening in the staking space, liquid staking to begin with, and then DVT, which enables many, many validators to come together and participate in it.

And I'm sorry, DVT. Distributed validator technology, which is basically the idea that many nodes can work together and act as a single unit as far as Ethereum is concerned. But inside that unit, actually, there is like many nodes which are working together in like a union. So it's like taking a single validator in Ethereum, which requires 32Eath, which may be a lot of money for many people to participate.

And in further democratizing this, the idea that many people put in maybe one-eat and 32 of them form a group, and then they can all, like, from the Ethereum's point of view, they look like one node having 32-Eath, and they all come to consensus on everything. So these are, like, really fascinating developments, both liquid-staking and DVT, and I think I would like to think that eigen-layer or restaking is in this category of really interesting things, because now, with something like eigen-layer, what you can do is use the security of Ethereum and the power that, you know, these staking protocols bring to supply it to very flexibly to any kind of new protocol that can then be built on top of it. So that's that's what I'm

very excited about. Yeah. First of all, you know, like I want to congrats everyone, you know, like who participated in it. It's like developers, core developers. It is tool developers. It's a client's developers, you know, like and you know like this, I mean like it's too about two and a half year. So like beacon chain launch in December of 2020. And you don't like, and the withdrawal is like finished this loop of upgrade when all flow of staking, almost all flow of staking is working right now. And this is exciting for all the Ethereum community. And everyone, you know, like who believe in like trustless virtual machines that can work in the internet. But it's still your question was about challenges and also like not only like

good things, but also we have still a lot of issues. I mean, one is like a decentralization. This is I mean like an issue in staking and the question how this network can be even more decentralized. Second is like white labeling, you know, like this is a lot of nodes who use different brands, but running by the same people. It's still, you know, like M.V. Issues, you know, like that's still here. One of the things that can make liquid staking better and other staking products better,

when withdrawals can be initiated by users who staked. Because for now, it's impossible, and only validators can unstake. But this is still a lot, and it's a lot of work. what, you know, like what was achieved, it's really great, you know, like, and how it's happened without any, like, big issues. So congrats everyone. So I think now that we've made this full transition to this system, there's a lot of thought around different things that can happen using steak. And obviously, like with Lido, that was, you know, one of the early new ideas around

what could be done with that. But I actually want to start this part of the discussion talking about eigen layer because there has been a lot of chat about it. And so, um, Sri Ram, why don't you just give us an explanation of what it is? Yes. Just starting with the name eigen layer, uh, eigen in German for your own. It's your own layer. You can build anything on top of Ethereum security. So if you look at like the evolution of Ethereum, like when Ethereum started, the major like jump was the idea that anybody can build an application and tether it to a common security zone, right? Of course, the technical thing that was done in Ethereum was to create a general purpose programming language, but I think the fundamental innovation is not that. The fundamental

innovation is creating a Shad security marketplace. The idea that there is a common Shad security pool, like, you know, Ethereum nodes can actually validate. Now, any application that can be built on top. So I like to think of this as kind of like the first modular blockchain. So it basically modularized the idea of, you know, application as distinct from the trust zone, which is Ethereum. And when you see the evolution going forward, you see from there successive opening up of the innovation landscape. And I think layer twos are a major paradigm shift there. The idea is now that not only you can write an application and tether it on top of this EVM virtual machine, but you can have like a whole chain which you validate yourself,

and proved that the chain was validated correctly back to Ethereum. And so this was the layer two landscape, which broadened the scope of Ethereum. And I think, you know, the underlying ethos there is really actually, you know, inspiring that Ethereum took this step. It's a leap of faith to take because, you know, you come in and say, oh, now, you know, it's not Ethereum running everything. Nothing is running natively on layer one. All these layer twos are going to run it. and it requires like an immense faith in the power of free markets and open innovation to actually take that leap. And many of us thought that this is a pretty crazy leap of fate to take,

but actually just paid off massively because you have these absolutely powerhouse projects, which are now built on top of this like one framework, which would never have been possible if it was only Ethereum and one protocol and there's an enshrined like sharding or whatever other techniques. This is just much better. So I think that's the second step. And I think the way we envision eigenlayer is as an even broadening of the innovation landscape. So what cannot be built with just layer twos today can be built with eigenlayer. So what is the idea?

The idea is the root of trust of Ethereum is coming from staking. You're putting down stake and you're saying that you're validating Ethereum blocks correctly. So that is the root of trust. If we can take this route of trust and supply it to anybody who wants to build, let's say, a new consensus protocol. I want to build a new consensus protocol. I want to build a new Oracle. I want to build a new data availability. I want to build a new bridge.

I want to build a new MEV management system. Anything. And then say that, hey, already I have a route of trust here. Like people are staking and then making commitments that they are going to make Ethereum blocks correctly. And what eigenlayer does is ask the staker if they want to opt in to eigenlayer.

I'll explain later the technical details of how you opt in. But when you're opting into eigenlayer, essentially what you're doing is your saying, I'm not only validating Ethereum blocks, no, I'm validating potentially new services built on top of eigenlayer. The services could be any of the things that I mentioned, new chains, oracles, data availability, data storage, anything that you can think of, which requires N nodes to do the calculation. So now you opt in to such a system. And when you opt into eigenlayer, what you have to do is you have to download and run whatever other software.

So in addition to running your ETH node software, no, you can run all. all these other node software. If you're running eigenlare, you're not running all the services on top. You pick and choose which set of services you want to run on top. It's an opt-in marketplace. And so now what it does is whenever you have a new idea for how to build a new distributed system, a new chain, a new Oracle, any of these things, you don't have to go and find your own trust network.

You can just come and avail of this massive existing trust network, Ethereum, and then just plug your, like, system on top of it. So you can have, you know, all of these services, ecosystem services for Ethereum, which are powered natively by Ethereum's taking. So that is the vision that we have. It's part of why we call it eigenlayer. Anybody can build anything they want on top of this, you know, common framework. And we think Ethereum is the natural home for something like this,

because Ethereum has shown again and again that it's committed to the principles that you are very passionate about, which is open innovation, credible neutrality, decentralizations, ship resistance, all of these things. So that's the, yeah, that's what I can live. Would it be fair to say that it's kind of similar to the Pocodot model where there's all these different parachines, but then they share security at the base layer? Because that's what came to mind for me. That is absolutely a reasonable comparison. So there are some dimensions where it's exactly the same and some dimensions where it's not the same. One dimension where the PolkaDod model is different is PolkaDot is not optional.

in. So everybody who is in the polka dot relay chain will validate all the parochains. So there is a kind of enshrined mechanism. So there's no opt in there. That's that's one difference. The second difference is you can, because it's not opt in, you cannot opt into arbitrary like slashing things. You're opting into like a very narrow scope, which, which itself, you know, you can run basically any virtual machine, but you cannot run an Oracle with Polka dot security. You cannot run a new distributed storage protocol with Polka dot security. You cannot run a new consensus protocol with Polka dot security. So there is some constraint.

I would compare the Polka dot parachain model much more to the layer two paradigm. So in layer two's borrow exactly Ethereum security, but only for running new virtual machines. And so that's what, but it's a more enshrined mechanism for doing it, whereas eigenlayer is much, much more flexible. any distributed system you can think of, you can just run it there, and you could borrow aspects of Ethereum security. And sometimes you will borrow a subset of Ethereum security, but sometimes this is something insane. You can borrow a super set of Ethereum security. And I can come talk about

that later. Okay. I'm going to put a note to myself to ask you about that later, because one other comment that I want to make is I sort of feel like Lido is a way to reuse stake. for the user, but Eigenlayer is a way to reuse stake for developers, something like that. Does that make sense? Yeah. I don't know what. We are definitely doing complementary things. And I think I have like one point of view there, but I'm sure Constantine has some interesting thoughts. Yeah. Constantine, what do you think? Yeah, first of all, I should make a disclosure that I'm investor in Angular.

there because but I will still challenge, you know, like three RAM to make it more interesting. Yeah, first of all, you know, like we should not forget about risk. You know, like when you run Ethereum, it is a risk of flashing. When you start to run other protocols, it's also like adding your additional risk. And there are my thinking here that what is I'm going to are doing is like risk-taking feature. Yeah. So like some type of, for example,

for large protocols can do restaking by themselves. You know, for example, I don't know. We can imagine that TK Sync or like second layer they can just use withdrawals or ask users to give withdrawals from their validation nodes to them straight and do it. But for smaller protocols, you know, like and to and also like for big tail and also for some big one, Iguilare is the best solution can be the best solution. Why? Because they make a marketplace around that, so you don't need to bring additional stake and work with their stakers. You can just, you know, like use the marketplace.

And in case of our staked is theory of my thing that it's kind of reused inside of Aguilar, so you can not use ease. You can use steak teeth, for example. And also, you know, like you can. Also, like, it can be an option that LIDA can make any type of integration with Aguilair, you know, like more deep or like it can make straight with other protocols. So I, restaking is a mechanic, you know, like how you can take more risk. But don't forget when you get money for something, like when you get rewards, for example, for doing job, you get additional risk and nothing is free.

So, Constantine, what are some of the risks that you're thinking of for different types of projects, or chains that are thinking about using restaking for security? Yeah, first of all, you know, like, staking has, I mean, what is staking? Staking is a mechanic where person or like a protocol lock capital for some time and do some job, you know, like for example, running a node of Ethereum protocol. And then get some risk because, and get rewards. You know, like, and all these things are usually important, yes. And in case of Ethereum, the risk is slashing or losing of rewards.

And you can, if you, in some edge cases, you can lose all your ease when you stake your ease, yes. And when you restake, you add additional risk and it depends on the protocol. You know, for example, bridges has one type of risk. oracles has another type of risk. Layer 2s or like sequencers has another risk. So like... But can you give some examples of what those risks look like for those different types of projects?

Yeah, for example, if you secure some type of oracles or bridges, I mean, first of all, it depends how, like, the protocol developers will pack it. For example, and what type of risk, a risk taking will secure. If this, for example, a risk or a buck of protocol may be like restaking will not cover it. And the question is who makes this that decision? If it is a risk, for example, you want to delegate risk of some type of bridges and bridge

is exploited by some reason. For example, validators did this attack. Then you can lose part of your capital if you put it under the risk. And so on, you know, like. The question is, when you get reward, you get for some job or like for the risk. And if you get this rewards, the question is how to participate. I mean, in case of, for example, Aguilera, the user will choose with which protocols and which, like, operators who will run this nodes, you know, like we work with.

Why? Because, and for example, they can try to use the most safest or the most, most bigger, with one's biggest reward. So, Sri Ram, how would you respond to these concerns about risks? No, these are real. The idea that basically, you know, I think Constantine's here already, what are you doing with staking? And restaking is, we think of it just as making staking more programmable.

So you're basically like, now, you know, normally when you're staking, you're subjecting yourself to Ethereum slashing risk, right? Which is, if you don't validate your blocks correctly, then, you know, or there was a bug in the Ethereum protocol. There's an R of those two conditions, right? Like, then you would lose your stake, right? So let's take the simple scenario where the code is all correct. The, if under that circumstance, actually the risk profile is only you lose money,

only if you didn't fulfill the conditions, the covenants of the Ethereum protocol correctly. So in some sense, I think while there is risk, the risk profile of. validation is fundamentally different from defy risk. So in defy, when you're putting a putting your like steak teeth or any other token into like an LP into uniswap or a compound or any of these things, essentially you are underwriting price risk. The price can fluctuate and price risk is exogenous. Like you have nothing to say about it. You have no control on whether the price is going to go up 10x step, you know, or 10%. Like this is not something. endogenous to the person who's putting the capital in. Whereas validation risk, if you assume

the protocol code is correct, is endogenous. Like by doing more and more like by running a distributed validated group, by basically making sure you're running better and better protocols, you can drastically reduce the risk. So I think validation is a different class of risk and DFI is a different class of risk. So that's the first thing I want to say. Inside validation, you have like, like two aspects to risk. One is risk due to your own, you know, misbehavior. And the other is the risk due to protocol errors. And I think, but if you know how to do the operation for that protocol correctly,

then basically you can very much minimize the first risk. And what remains is the second risk, which is that the protocol has an error. And so, you know, we try to think a lot about, like, how to minimize these stakers exposure to protocol errors. In fact, the way I think about like how liquid staking and something like eigenlare or complementary is that, you know, liquid staking, the hardest thing that, you know, a protocol like Lido or Rocket Pool or any of these things have to do is to figure out how to imbue trust in the validators. Like that is the hard part. For something like eigenlayer, we see the complement is the hard part. The complement is how do you induce trust in a new distributed system, in a new distributed protocol?

That is the hard part. And that is our specialty in some sense. It's figuring out what is the protocol configuration, what is the slashing. And this is very, very particular to like, this is like a deep understanding of the distributed system and the slashing conditions that go into each of the distributed system. And you ask Constantine about like, how would you think about risks in these different examples? I'll give a simple example. Let's say an Oracle.

You can build an Oracle. So Oracle is basically like a price. Right. Let's say just you want to report BTC to USD price on Ethereum. And you say, okay, I want to get the Ethereum nodes to participate in this. There are many different ways of building the Oracle. Each of them lead to like completely different kinds of validation risk. So I'll give three examples of actually how you can build an Oracle.

One example is you will get slashed if your price report is very different from the majority price. I say one Bitcoin is like, you know, or one USD, whereas most people are saying one Bitcoin is 30,000 USD. then, of course, like, you know, I'll get slashed, right? So this is one example of slashing. For sure, babe. So deviation from majority is one kind of slashing. Another kind of slashing is, you know, when you all vote on something, then you bring in like a much more trusted group, you know,

which may be a Tao or like reputed people in the community, whatever, like some other thing. And if they say something different, then you could get slashed. So that's another kind of slashing. A third kind of system would have no slashing. It'll just take them aggregate as many distinct opinions as possible and then just present the majority opinion as like a price article. So these three systems have very, very different risk profiles. For example, in the first system, the risk profile is, if you disagree with the majority, you would get slashed. What if a majority quoted one Bitcoin is one USD and you quoted one Bitcoin is 30,000 US?

You are right, but you may still lose your fund. So the trust model is you are taking a risk on the majority opinion. The second model is you're taking a risk on the Dow opinion. Like the Dow says something and you disagree with the DAO. You may be right. The Tao may be lying out of their teeth. And you may lose your funds. So that's the second model.

In the third model, there is no no slashing. So actually, so this exposes like what I think Constantine was talking about earlier, which is risk is very nuanced when you're operating into new, of distributed systems. And calibrating and understanding this is one of the main roles we see for eigenlayer, is understanding how to build new distributed systems with proper slashing conditions where you are not subjecting yourself to undue risk. But of course, it's an open market. We believe in permissionless innovation where many people can come and build these things. And we try to balance, you know, how much openness we have with how much safety conscious we can be. And

There may be like two tiers where one tier is more trusted and, you know, has been more vetted and other tier is anybody can build anything. And there'll be like different risk profiles on these kinds of. So that's absolutely how we think about it. There is a large category of things where you do not need to take undue risk because, you know, validation risk like I pointed out, is actually much more tightly constrained than something like defy risk. And we think this would be the next frontier.

So when you look at yield stacking, like where do you get additional yield? You get yield from Ethereum. What is the next level of like marginally additional risks that you can take that will actually give you like productive value? I think that's where we want to place ourselves is validation. And all the things. And I think this is another thing we want to clarify is we are not a D5 protocol. We are not a liquid token.

Like it's what Egonlead is building is very different from what a liquid staking is building or what a D5 protocols. So we do not envision on eigenlayer. The core feature is people set their withdrawal credential and then you borrow against. That's not what we're envisioning eigenlayer. What we're envisioning eigenlier to be is a universal validation marketplace. Come here, you can participate and get validated for many, many different things. And it's a double opt-in.

Double opt-in is staker has to opt-in to the service. The service has to opt-in to the staker. And both of them have their own subjective preferences. which meet and merge to actually make the market. So it brings much more of a free market dynamic than what we have to do. Oh, wait. So now I understand. So sorry, when different projects decide that they want to use eigenlayer for their stake,

then they also can select specific validators. And they don't have to just use all the same validators as Ethereum. That is correct. And this is what I meant by like we had an asterisk earlier, which is sometimes you can get more security than Ethereum, that's a rather insane thing to say. Why would you think that? Because when one of the things that you may want is I only want like 1,000 very decentralized

notes. Let's say this is some special use case. As an example, I do secret sharing. I take a secret, split it into small chunks and send it into thousand notes. Now, I really want to know that these thousand notes don't collude with each other. Otherwise, the secret will be exposed. So what I'm doing is I want to find like, you know, thousand dispersed notes. then maybe you know you select home stakers in Ethereum and I want to just send it to them because it's a very dispersed group.

They're not coordinated. I have a question here. Yeah. Like how you will choose solo validators or home validators because it can be a data. But if you will use it for ad drops or like for some type of utility, I mean it will be maybe it's already people who are running, you know, like thousand words and. using different IPs, different data centers, the same is air drops. People are, do like a couple people,

I mean, like, running or thousands of accounts, you know, like an arbitral, and then again, it's thisirdre. And nobody can understand because it's really difficult. And the same as these people even make KYC, you know, like for example, in different networks like Salana, for example, they have a validation program. And Salana is delegating to, you know, like,

unique validators, for example. But it's known information that it is some people who are running like 20 nodes for Salana and use different identities. And this is an issue. They just use maybe like all nodes or other white label service and cheat with KYC. And the question is, you know, like how in permissionless sense-resistant network, you know, like choose the data, you know, like that you can check. Yeah, this is a great question.

I think eigenlayer itself does not take any position on who's more decentralized or who's less decentralized. We want to be a neutral platform. That's what eigen layer is. On top of eigenlayer, we expect people will build more subjective services. One of the framework that we employ a lot when designing eigenlayer is a paradigm we call intersubjectivity. Intersubjectivity is the idea that there are like many. many subjective decisions is a consequence of many people interacting with each other rather than the subjectivity being enforced by the platform. So the platform doesn't say, oh, Lido is more

centralized or less centralized because, you know, in some sense it's more decentralized, in some sense it's less decentralized. And that is not a position for eigenlayer to take. But what we believe is that services building on top of eigenlayer can make that judgment. Each of them have their own judgment, their own ideas, and their own purposes. Somebody may say, I only want KYC, validators, and that's up to them. Somebody may say, I only work geographically distributed validators. I don't care if it's the same guy running it all over the world. One of the services coming on top of eigenlayer is called a proof of location service.

So proof of location service lets the nodes kind of like use latency and other thing to kind of figure out, use the Ethereum nodes to figure out the distribution, like the latencies and geographic distribution of other nodes. And so you can have very interesting protocols like this, which can offer services that other services may then rely on. There is no universal answer to this question. And just like each one has a subjective opinion on what is enough decentralization, that's exactly what we expect. And in fact, there will be places where people don't want decentralization. So this is another important thing we need to understand is for some of the things you want just economic trust.

I just want like $5 billion of like staking. That is all I care about. Because if that was wrong, I can slash the $5 billion. I don't care if it is one guy putting up $5 billion or like five million people putting up $1,000. It makes no difference. There are protocols like that. And there will be protocols which don't care about economics, which care about decentralization.

And there will be protocols which say that no U.S. validators. Okay, that's something that they want to use. They can use it. Some protocols will say, I want only U.S. validators because I want to be under the jurisdictional compliance. I think all we want to do is to enhance the express. in the free market. And we think of ourselves as a marketplace for decentralized trust.

And when you build a marketplace for a new thing, which is not at a commodity, one of the things you want to do is to grade the commodity, either yourself or how other people grade it and use it subjective. And as a protocol, we don't take the position of grading it. So we are uniform, neutral. We're neutral to the extent that we accept all eats taking. So that's the kind of scope. of how we think about it.

This is super interesting. It's like customizable security. I really like the concept. All right. So we're going to discuss a lot more about staking and security issues. But first a quick word from the sponsors who make this show possible. Join over 80 million people using crypto.com, one of the easiest places to buy, trade, and spend over 250 cryptocurrencies.

With the crypto.com visa card, you can spend your crypto anywhere and get rewarded at every step. Up to 5% cashback instantly, plus 100% rebates for your Netflix and Spotify subscriptions, and zero annual fees. New users enjoy zero credit card fees on crypto purchases in their first seven days. Download the crypto.com app and get $25 with the code Laura. Link in the description. Bet mode activated. The scorebed app here with trusted stats in real-time sports news. Yeah, hey, who should I take in the Boston game? Well, statistically speaking, Nah, no more statistically speaking, I want hot takes. I want knee-jerk reactions.

That's not really what I do. Is that because you don't have any knees? The score bet. Trusted sports content, seamless sports betting. Download today. 19 plus Ontario only. If you have questions or concerns about your gambling or the gambling of someone close to you, please go to conicsontera.ca.

With Amex Platinum, almost every purchase made with your card can be covered with points, including new tastes, new fits. and virtually everything in between. That's the powerful backing of Amex. Conditions apply. Local news is in decline across Canada, and this is bad news for all of us. With less local news, noise, rumors, and misinformation fill the void,

and it gets harder to separate truth from fiction. That's why CBC News is putting more journalists in more places across Canada, reporting on the ground from where you live, telling the stories that matter to all of us, because local news is big news. Choose news, not noise. CBC News. Back to my conversation with three ramen constantine.

So we talked a lot about restaking, and I just want to fill out a little bit on the liquid staking side because there's different types of liquid staking as well. And I wondered also now with withdrawals enabled, like what else is on the horizon in the liquid staking? Marina. Constantine, can you take that? Yeah. First of all, I think when withdrawal is available, it is like, first of all, the market can be more competitive, you know, like, why? Because it's more for new

liquid-staking solution, for example, it's more easy to have some liquidity to withdraw from them, for example. And also like another thing that still, you know, like to build like decentralized network, to build healthy network, you need to features. And some features are still not here. For example, like I mentioned about withdrawals that can be run by smart contract or by the user initiated withdrawals and some others. Because, you know, like I think that it is a lot of great developers who want to build liquid staking, who already build in liquid staking. And they still has a lot of assumptions or like issues, not because they don't want to build a great product or like to

to build more decentralized product or to be a better product because it's just impossible because they depend of the protocol. I mean, the withdrawal is available. It's already like one, we have a couple more features and it will be more simple. Also, I think that people will try to take more leverage and their cheap ease landing will be, I mean, finished, you know, like because if your is not staked or like if it's put on the lending protocol, people will take loan under the state this and loan some is and stake it again. And so the lending price will go up. And it will be like almost the same, you know, like depends on the risk again.

Because as I told you again, when you put in any Defy protocol, in any like platform, it's always additional risk and we still speak about risk more. And also I think that we are restaking will be available. This is also like a great feature. And yeah, I think that I mean that this is the right future to, first of all, like keep Ethereum permissionless and censorship resistant because, you know, like Lydon, for example, has a purpose and mission to keep like Ethereum permissionless. and sensor resistant.

If not Lider, someone else, that's great. The question is that what we need is to keep Ethereum, because when you stake, you get, I don't know, depends on how much is stake, and depends on how much transaction was done. You have about 4% reward, for example, per year APR. But that's how to say, that's good for long-term people who are really ready to commit to Ethereum security.

But if, for example, you make your network centralized, if you make it like permissioned and censored, for example, by MV or any other partner, you just disrupt the network, the value in what you believe, for what you build, and make from that just state machine or like database. And what I try to say here is that all this development and mature of, Ethereum Protocol, make more competition, more features, and make it like so decentralized, unbelievable, you know, like, and so great that and unstoppable. And yeah, so like, this is what's happening and really make everyone who is contributing,

who is working on it happy. So, and then one thing that I was wondering when I was doing some research, because also there's talk of like super fluid staking. So can you just define that? So superfluid staking is this idea that, so normally the way liquid staking works is you go to, you first go a stake in Ethereum and then have a liquid staking derivative issued on top. Superfluid staking was an idea introduced in, I think,

to go in the reverse direction. First you stake in an application and then use that to stake in a protocol. And this was kind of popularized by osmoses where you can take a liquidity provisioning token and then go and stake that in. the code protocol. Of course, we cannot do that on Ethereum because, you know, to do this, the code protocol needs to respect that particular, like, you know, application token. And of course, Ethereum's not going to allow, like, you know, some other LP token to be staked into Ethereum. So that's not a thing that that can, is possible. But something like that is possible

on eigenlayer. Like, you can take a steeth or a steeth-eat like a curve or some other LP token and then take that and then deposit it potentially into eigenlayer. Because now what happens is because you have steep, you're getting the Ethereum yield. Because you have the liquidity provisioning, you're getting a yield from that. And then you're potentially getting a yield from like eigenlayer. So this is like a maximal yield cascade. Like you're going through like all these three things. And again, things like this, eigenlayer does not take a position because we don't want to say,

what is the risk profile. Steeth to Eith is maybe lower risk. Maybe steeth to U.S.D. C is a different kind of risk. Maybe some other thing is much more risk, right? And this is up to each of the services building on top of eigen layer to decide what kind of like, you know,

staking collateral to accept. And so this is what would be called superfluid staking is basically the idea that you are going through like a DFI or some other protocol and take a kind of like a derivative of token and then state that. And the way eigenlayer is structured is all kinds of tokens in the Ethereum ecosystem can be state.

And that's, you know, the vision is kind of be a security hub, like validation hub for all of the Ethereum like ecosystem. So one thing that, and this is, we kind of talked about this when we were talking about like the tradeoff with the risks, but I wondered, you know, just hearing all this, it makes me think about how kind of like every little change you make could be an attack vector. And especially like in DFI, we've seen the more composable things are. Like there's a lot of kind of unexpected ways for exploits to happen. So, you know, obviously EigenLayer is on Tessna at the moment, but I'm wondering, like, what are some attack vectors you see there? And so we'll start with that for EigenLager.

I think the kind of things I am personally most worried about are basically because to the extent that we are completely permissionless, Igan layer is completely permissionless. What could happen is somebody comes and builds a protocol on top and it either has like a malicious slashing condition. That would be an attack vector. Or that protocol itself is just badly designed. It's like a Luna or something. Like, you know, it just uses the eigenlayer collateral, uses that issue, some other, like, crazy stable coin, and then, like, it has some leverage loop. These are absolutely unpredictable. And I think that I would completely agree with you that just like restaking was unpredictable.

And then once it's that, it's kind of like MEP or liquid staking that, you know, you don't see it. And then once you see it, it's like, oh, of course, it is needed and it is important. And, you know, you don't know what are all the downstream consequences of that. I think that's the same thing we feel with eigenlier. The way we try to mitigate this is by, like I mentioned earlier, having this two-tier system where one tier is more closely guarded and much lower risk. And now if you want to build something on another tier, which is completely permissionless,

but the first tier will be permission, right, in the sense of like if there is a committee and the committee has to approve these things. And to convince the staker to go to the second tier, you know, most stakers may not take that step. But, you know, if there's a very strong protocol, very interesting protocol, somebody might. And so those are the kinds of risks, at least we want to kind of segregate them, rather than, you know, everything is aggregated and it's very confusing for people. And, Constantine, what about you?

What are some of the different attack vectors that you're looking out for on the liquid staking side? Yeah, when we look on any, like, smart cortex, contract development, I think the biggest risk is smart contracts in any like application, like Aguilare or like any other, you know, like, and of course you, I mean like any or Dow or entity is trying to make the biggest amount of audits, you know, like, but still, I mean, this is, I think, the most dangerous, the more complex code base, the more risk it is. Right, but is there anything like similar to a 51% attack that is kind of a, known thing that you guys are trying to avoid?

No, I mean, like, it can be like just a buck, like an NED5 protocol, you know, like that is I think there's the risk. The second, of course, when you like, I think like, also like the question, I mean like for three realm here is like how they work with their nothing on stake attack, for example. if for someone delegate there is stake this to some operator, they're running some software

and they don't have stake and they, for example, start to really big and they have, for example, a large amount of MEP or they can collude. So how you work with it? Our approach, so basically the question

is, how does eigenlayer kind of incentivize, the alignment between delegator and like the person who stake it is. And basically the operator and the staker. And Igenlare takes a very purest view. It's like Ethereum. We do not give any assurance of any operator. Just like Ethereum, if you're staking, if you're delegating, it's on your head.

Like you don't delegate to people that you don't trust, right? It's the same model on eigenlare. We do not give any assurances. That's why it's very different from a liquid staking protocol. A liquid-staking protocol has to think through the delegation risk and understand whether it's real-world trust, whether it's insurance, whether it is, you know, crypto-economic security, some set of, like, features that actually lock the node operator into correct compliance. This is not, this is out of scope of the eigen-layer design. eigen-layer allows you to specify who you're delegating to, but the trust, you need to trust the person you're delegating to otherwise to not delegate. and we see because we want to keep this as close to the Ethereum protocol as possible.

So the way we think about eigenlayer is just a staking extension, programmable staking extension of the Ethereum protocol. So it's very, very, very similar to the Ethereum protocol in that we have no delegation incentives. So any delegation incentive is either built by the free market, like Lido built kind of liquid-staking protocol on top of Ethereum, protocols like that may basically just opt into eigenlayer and then like essentially solve that problem for the market or you know there will be a delegation ecosystem that emerges which which is you know which solves that but it's out of scope for the native eigenlayer itself provides no assurance on

delegation and then one question that I had about Lido is and this has been something that people have talked about for a while but there's been concern that Lido could like effectively almost become like a governance layer for Ethereum. And, you know, currently out of all the liquid staking tokens, it does have 75% share. So are there things that Lido is doing to try to mitigate that? Yeah. First of all, I didn't finish with risk. And that was the second risk that I want to mention.

Like one of the risks of liquid staking is, for example, because the logic is our of Ethereum is still how to say in progress, you know, like, And for example, you don't have some features. You can't, oh, I mean, it's all, I mean, it's really difficult to, to make now fully like, pacification of the protocol. Yes, when, and so you have some proxy to upgrade this protocol. And now, for example, all the protocols are not all, but almost all, has a upgrade feature of this smart contracts and they can, you know, like try to attack, you know,

in this way. So this governance token, they has this power. For example, online the forum, on LIDA research, you can read the research about a feature called dual governance. And this dual governance feature can give this proxy, this or

I mean, this is in my opinion, big idea, you know, like, why it's a big idea because before it nobody had a solution for agent issue. And here, LIDA can give their permission to upgrade

or to vet the upgrade of smart contracts that connected with stake in Coethyum to stake its users. And it means that if Lido will try to attack Ethereum or like Ethereum users, the state Ethereum users who really like has something on stake, you know, like they can stop this upgrade. It's not in production, it's still in research, But as I know, guys are working on it, you know, like actively.

And this really, like, powerful. You can, you know, like, for example, think about 1970s when Nixon canceled gold standard. You can imagine if other USB holders can have a VETA power to not make it. What they would do? Would they, like, vote or not? Or they were scammed. And before, it wasn't possible to program. you know, like regulation.

And I think what blockchain give, but not a lot of defy protocols or other protocols use it right now, they give an option to make technology regulation. And this is a big revolution. When you can regulate something by writing smart contracts, and it's even like better than like current regulation, because in a like real regulation, sometimes some people make hard decision, like to cancel as standard to because they have like, I mean, some big reason for it. Yes.

And a lot of other people, they lose value because of that. And in case of, and with this dual governance feature, it will, if for example, somebody will implement it for use the dollar back then, it would be impossible. And that's what really blockchain and make it. But the question, yeah, is it risk? Yes, it is. Yeah, and one other thing that I think is a risk with Lido is, since it follows a custodial model, you know, for the user that could obviously result in them losing access to their ETH.

So are there plans to address that? As I know, Liddo is not non-custodial. And I mean, it doesn't a custodial solution. But users, they don't retain control over their private keys. Yeah, no, I mean, users deposit, like, light is a middleware. You know, like, what is the line? Light is a middleware. You use a smart contract system.

You put IS inside. Then, like, this is staked with some of the validator in the list. And then you can, I mean, like after the LIDA to upgrade, when these rolls will be available, you can be on stake this and get your ease. In general, it's self-castity solution. So the whole time the user has access to their private keys? Yes, yes, of course. Oh, I don't know why I thought it was that they are giving it to the validator.

Lider run by DAO, and it is a smart contract system. When you stake, it's going to validators, but validators need to initiate withdrawals, you know, like to withdraw to the smart contracts where the user can withdraw. But it's fully like self-cast no validator. Validators can still is that state with them.

They can slash it on. Okay, interesting. For some reason, I thought that there was a point in time where they did not have access to their private keys. So I guess I had a misunderstanding about that.

No, in the beginning, you know, like when lied along, it wasn't an option when smart contract run, you know, like can hold withdrawal keys. And when LIDA launch, it was a key that was created by 11 different founders of Big Defi protocols like Curve Maker, like Curve Maker and others, you know, like they make, and they have a BLIS signature. It's like a multi-seekon beacon chain. that hold like about 10% of all E's that was there.

But when withdrawals were available, these people rotated. I mean, they can't withdraw, but I mean, they signed a transaction and rotated these keys to lighter smart contracts. Okay, okay. So people retain their private keys and they're just interacting with a smart contract. Yes. Okay, okay. Thank you for clarifying that.

So let's circle back to Sri Ram earlier. You said that you could go into something called super staking, if I remember correctly. I don't remember the term. What was the term? Superfluid staking or no? No, no. It was earlier than that. You said, oh.

Yeah, what was it? Yeah, I was saying that usually it's a subset of Ethereum security that you would get from, from Egonlayer. But, you know, there are places where you could get a superset or more than Ethereum security. Yeah. How is that? work. Okay. So if you just take decentralization, then it's even easier. For example, you know, one of the things people say is, okay, you know, there are 8,000, 10,000 distinct validators in Ethereum, but, you know, a small number of them, like, control a majority. But you could just,

you know, and you could run a protocol on top of eigenlayer where you only have, like, the more decentralized committee. So if you just measure on decentralization, your protocol can actually have more decentralization. If you ask how many nodes are needed to get to 50% of your protocol, that could be a much larger number. So that's one simple way in which you could get more security than if you have more decentralization or more effective decentralization. And this is, I think we think of this as an incentive for more decentralization to occur because, you know, the more decentralized nodes exist, then they have this potential ability to earn these additional rewards. But it is not easy, like I think Constantine pointed out, to measure which are the more decentralized nodes.

And it's a subjective thing. And it's not something eigenlayer is itself planning to do. Whereas other protocols may be built whose specialities in doing it. And depending on whether there is demand for decentralization versus demand for staking, pure economics, different people may use different things. And in fact, I think even Ethereum itself, right, so when we talk about, oh, is liquid staking a risk for Ethereum? It's because the protocol has defined some rules, which is basically, you know, how much ethists take determines how much reward you get. Like that is the protocol rules. Everybody is just playing by the protocol rules.

It's not like anybody's like modifying the protocol rules. Everybody's playing by the protocol rules. But the protocol has no internal way to detect, say, decentralization and reward it. But certain aspects of the Ethereum protocol require decentralization and censorship resistance, for example. And because the protocol. is underspecifying the constraints because the protocol has not specified that it needs decentralization and because decentralization is not objectively measurable. And that is what is leading to these unseen second order effects, like which we do not know how to control. And one way that something like

eigenler can play a role there is because it is adding on additional incentives that can drive like different demands for different kinds of nodes, you know, both existing protocols can, existing stakeholders can opt to become more decentralized because there is an additional reward for being more decentralized, whether that is geographic, whether that is stake itself comes from distinct sources, there is no common point smart contract risk. Whatever it is that the applications define as decentralized, that is what will end up as being an objective. But also, I think there there's a counterforce, which is some applications may want enterprise-grade validators, like, because they just, that's what they care about.

They care about real world trust. And the emergent set of features from a system like eigenlayer is quite interesting and unpredictable. And I don't know what it would be. But the thing that I want to say is we are letting the market value the different aspects of trust differently, like the aspect of economic. the aspect of decentralization, the aspect of KOC, these are all different aspects of trust and they'll be valued very different by the market. Super interesting.

All right. So I know Eigenlayer, I think, expects to be live on Maynett and Q3. Is that still the case? That is correct. Okay. All right. So why don't we just talk a little bit about what it is that we're looking forward to when it comes to staking or like what developments you're watching in the staking slash restaking world? for the next, I don't know, through the rest of the year.

You know, like, first of all, I think that MEP is really important, you know, like, and PBSS, I mean, it's because for now, it is their market, you know, like, where people are tried to extract MEP, you know, like, and there are, it is really layers who give, you know, like, where blocks to validators and validators or science's blocks, but it is a lot of discussion around that. For example, you can look on sushi drama that somebody like stolen is, then they use our into an MEP, you know, like, and I mean, and I think that this is one of their like piece that should be fixed.

And it is a lot of work that should be done there on protocol level. And maybe even like this MEP should be burnt, you know, like, and then it will make the situation much better. And yeah, I think Justin Drake is proposing that. Yeah, yeah, yeah. I think it should be discussed with, you know. Yeah, yeah. I am, for example, think that it's a great story.

Why? Because in other way, if it will be like parties who make a decision around that, that can be a mess, you know, like, so it should be done on a particular level. I mean like regulation quality, you know, like because it's a lot of questions about like legal status and discussion. Yes, that I want to touch, but I think it's important. And also, like, tax threat, threat of staking that will make really adoption better. And I mean, like... What are the issues there with the tax treatment?

I mean, like, it's not clear, you know, like how to tax, for example, staking rewards. Should it be daily or, like, what type of, for example, you know, like you have different type of liquid stake in KER tokens, you know, like that. For example, Acure value or you base. And it's also like not clear clarity right now. Better clarity and better regulation will make it more available, you know, like for people to participate in security, you know, like of Ethereum. Also, and also like more developers will build more solutions and that will be better for everyone.

Yeah. So I mean like it's also like this withdrawal initiation exits, you know, like by smart contracts and by users. That's important. I mean, small features, not like innovations, but that should be done. And but in general, I think that staking will be really how to say for me. Of course, I mean, it will be like different other markets that, for example, as we are mentioned, like data availability, maybe like it will be a decentralized sequencer market. It will be like MIV markets, oracles, where it will be used.

But again, and but I don't see, you know, like in this mechanic, a lot of innovations, you know, like it's, I mean, it's not a lot of what we can do there. It's just a mechanic. But what are interesting is, and what I really like always try to mention is risk that when you like take more risk, I mean, and when it's composable, you should think about different attack vectors. And what Aguilera and other people who will use risk taking, they need to make a lot of research and make a lot of innovations, you know, like how to work with this composable risk.

what will happen if, for example, your operator slashed 100% and you're running another networks? I mean, maybe this is edge cases, but it can happen. And all, like, you have like a big network and a lot of is restaked. And it's start to find a lot of value because they can, for example, slash this network and then you slash a lot of validators on Ethereum that also can happen. So this is a, I mean, different composition of staking with other restaking efficiency. Yeah, that's definitely interesting to look. Yeah, this is super interesting points.

I think how to understand composition of risk between Ethereum and like other things that could be built on top. I disagree with Constantine on the premise that like things like MEP and Oracle or not, that in a way, I think there's like a lot of crazy things that can be built on top of these things. But the base layer needs to be safe and secure. I also agree with Constantine on that. That actually the risk management on the base layer needs to be very sound. Only then you can start building other things much more safely on top. And a basic question there is how much of leverage that you can take.

And what is leverage, I think, is itself not quantified? For example, right now on Ethereum, we have $35 billion of staking, and you have, you know, $500 billion of assets. Why is this okay? It's okay, I think, but it needs a theory for why it's okay. And we need to understand why it is okay. Is it over-liverage? Is it under-liverage?

If it's already over-liverage, maybe we should not build something more like I get there. And that's correct. And that's the view we take. Actually, we think it's under-liverage. You can actually use security for more things. As well, as more staking is coming up, we don't think, like, the 35-bileged. layer of ETH is the stake. It's going to expand massively over the coming years. And the

questions about how to divide, like, the risk across these different layers, how to make sure that there is a common view, like if something happens bad on some layer, how does the other layers respond to it, how to make sure that nobody assumes that they have more security than they actually have. So this, I think, I want to separate two minor parts. Like one part of eigenlayer is what we call pooled security. Pool security is, let's say there's $10 billion of stake, and there is 10 protocols built on top. All of them, basically, these $10 billion is validating all of these 10 protocols.

So the shared security of each of these 10 protocols is $10 billion. Why? Because to attack any one protocol now, I need to attack the $10 billion. This is already, like, really useful. And in fact, this is Ethereum's model today. Why do I say that on tower of Ethereum, let's say there are 10 applications. Like each of the applications don't have attributable security. Uniswap doesn't get like out of the $35 billion.

They don't get like $10 billion just for uniswap. That's not how it works. It's just shared across all these different services, all these daps, and there's no attributable security to each application built on top of Ethereum. That's the same way that eigenlare V1 works. There's no attributable security to any one application built on top. The V2 that we're already like doing research on, which is, you know, trying to address issues like composability of risk.

has attributable security. Like, if you have 10 protocols built on top, each protocol can have a portion of the 10 billion, depending on how much they're willing to pay for it, attributable to them. So when it gets slashed, they get access to a portion of the slashed funds, depending on how much premium they're willing to pay.

So understanding these markets, understanding their, like, fundamental effects on Ethereum itself, these are all things we are actually quite excited about. And also, how Ethereum's staking and rewards can itself be, be tuned based on an understanding of the second order consequences is something we are quite excited. Yeah, one point I forgot to mention, you know, like, is a DVT, you know, like distributed validation technology. Why? Because first of all, I mean, first of all, you know, like some professional validators are already used. So they have teams, for example, because, you know, like,

always the question, who is holding the keys, you know, like, and to not hold the keys by one person, because if, for example, it's a big staking company, like, for example, P2P, you know, like, and the question is, and I am, like, a founder of it, but I'm not holding all the keys by myself. And the question is how to hold the keys. It is either, like, one way is to have a hardware secure model, you know, like that holding these keys, but it's still a lot of attacks that can be done internally, you know, like, oh, like if somebody will get access to your machine. And for that, you know, like a large validator that already use vouch,

and they also like distribute the key to many, like DevOps, for example. And they are, and to sign the block, I mean this DevOps sign it together. And so nobody has one key, you know, like to, for example, destroy this validator or like to make the double sign. And this DVT technology will go to the masses, you know, like when more, you know, like the problem of staking that, like, more complex the staking, the more it's centralized, you know, like, and this is a big issue of staking. Then, for example, polka dot staking is really complex. Why? Because you need, I mean, how it's work, how rewards work, that it's distributed to, for example, top 300, like, node operators in the same amount. And it means that the best strategy is to distribute to the last one. And because of that, the last one has the biggest APR.

And for example, P2P, because it has a good technical team, build a special software that every time, every epoch, redistribute this stake to, you know, like last one. And that's how P2P talk, for example, 10% or even bigger market share in PolkaDOT. Is it bad for centralization? Yes, bad. And it's really bad, yes. But for like some companies, they try to execute or take more revenue.

And the question is, first of all, it should be open and people should disclosure and different tasable data that nodes they run. And when, you know, like we will have, for example, Shardenk in Ethereum, it will be even more complex to run a node. And this, you know, like how to say for staking providers, it will be more harder to run the notes. And maybe it will be like examples where part of the keys will hold by professional validators like P2P blog diamond like another's. And part of the keys will be run by users. For example, you can run your note in your home and if it's just switch off by some reason,

you know, like the note is still signing because other parties are signing. And for example, somebody from P2P like or call you like Laura, how like what's happened. You need to switch off you walk back and they're paying for electricity or what else. And then it starts to work again. And I think that innovation in DVT is still, and also you should understand that you can't use one DVT solution because if it is a bug there, the part of the network can go down. And so it should be like different players on DVD market. It's the same as with clients, you know, like you have in Ethereum, you have a lot of clients. And if one drop down, you can move to another client. And also, I want to mention that it is like great companies, for example, in DVT, it's like OBL, SSV and others who is working on it.

And also it is like company who are building boxes, you know, like for users. It's like Depnode and some other, yes. And they are great because they give a person who has not has skills to run an old by themselves. And I think that to make it really like network decentralized, We need more innovation here. Maybe we need more funding here to, I don't know, you put a small server in your home, put in the plug and it starts just work.

And yeah, I think that is important. Yeah. All right. Well, this has been a really interesting discussion. There's honestly so much that we could unpack. I'm sure we'll be revisiting issues around staking and restaking in the future. But in the meantime, where can people learn more about each of you and your work?

Please follow us on Twitter at eigenlayer. There's also a discourse forum, forum.orgon.org. And at Discord, please, you know, which is linked from our Twitter. So please go ahead and follow on these things. My personal Twitter ID is at Sri Ram Cannon. So please follow that.

Yeah, I think Twitter is the best place, you know, like where you can read and also follow and communicate. And because, you know, like I'm building a lot of different projects. And we're building, for example, decentralized RPC and so on, many others, yes. And I think that, yeah, so like soon we will announce, you know, like a general like entity, like investment entity, cyber fund. And that, you know, like we'll try to make all the communication, you know, my communication and my focus will be there.

Okay, great. Well, it's been a pleasure having you both on Unchained. Thank you. Thank you so much, Laura. Constantine. Pleasure. Thanks so much for joining us today to learn more about Srebrom,

Constantine, Egan, Lear, and Lido, and the state of Ethereum staking. Check out the show notes for this episode. Unchained is produced by me, Laura Shin, without from Anthony Yun, Mark Murnock, Kevin Fuchs, Matt Pilchard, Zach Seward, Juan Aranovich,

Sam Shreveh, Ginny Hogan, Jeff Benson, Leandro Camino, Pamma Jim Dar, Shoshok, and CLK transcription. Thanks for listening.