[LIVE] Eigenlayer QnA with Sreeram Kannan

Transcript

bankless Nation we got a live stream for you here today we're going to get some questions answered about I think something that people have the most questions about which is reaking igen Lair ABS Networks timelines where's the yield coming from what are we going to do uh there's a lot of different questions I have a lot of different questions that other people have have and I have Source questions for the past two weeks from people around the Crypt sphere around the ethereum ecosystem both on Twitter in my DMs on foraster as well we are going to bang through all of these questions one by one by one I've got 23 questions total some of them are redundant so I'll ask them all at once uh some of them will be longer than others some of them will be very short um we're going to get through all of

them we're going to get you guys some answers to the questions that you guys have about Igan lay so I'm going to bring on two people out of the Igan ecosystem one you might not be as familiar with Neema bizar uh Bazi who I actually met at eth Denver 2019 uh and he has joined the Igan layer ecosystem the Igan Labs team and also fream Canan everyone knows freom uh so we're gonna bring them on in just a second but first a moment to talk about some of these fantastic sponsors that make the show possible Kraken knows crypto kraken's been in the crypto game for over a decade and as one of the largest and most trusted exchanges in the industry Kraken is on the journey with all of us to see what crypto can be human history is a story of progress it's part of us hardwired we're designed

to seek change everywhere to improve to strive and if anything can be improved why not Finance crypto is a financial system designed with the modern world in mind instant permissionless and 24/7 it's not perfect and nothing ever will be perfect but crypto is a world changing technology at a time when the world needs it the most that's the crack mission to accelerate the global adoption of cryptocurrency so that you and the rest of the world can Achieve Financial Freedom and inclusion head on over to kraken.com slbank list to see what crypto can be not investment advice Crypt trading involves risk of loss cryptocurrency services are provided to us and US Territory customers by pay Ventures each pvi doing business as kren cell is the mobile first evm compatible carbon negative blockchain built for the real world and now something big is

happening introducing the cell layer 2 it's a game-changing proposal that's going to bring cell's rapidly growing ecosystem home to a eum vitalic has shared its excitement for the cell layer 2 on the cell Forum so has Ben Jones from optimism but why the cello layer 2 will bring huge advantages like a decentralized sequencer offchain data availability and one block finality what does all that mean Rock Solid security a trustless bridge to ethereum and more real world use cases for ethereum without compromise and Real World Adoption is happening active addresses on cell have grown over 500% in the last 6 months with the cello layer 2 gas fees will stay low and you can even pay for gas using erc20 tokens but C is a community governed protocol this means that cell needs you to weigh in and make

your voice heard join the conversation in the cell Forum follow sorg on Twitter and visit c.org to shape the future of ethereum I'm joined with SRAM Conan the founder of modern reaking s Ram welcome back to the show hey hi David really excited to be here we're looking forward to do this together and uh also next to him we got Nema vizeri ecosystem research development over at igen layer Nema welcome to the show thanks so much for having me also n you're a you're a new face for us maybe give us a little bit of your introduction who you are and what you do at Igan lir yeah sure so I joined the igen labs team uh a few months ago um to lead the ecosystem

development um and AVS you know um uh teams that have been interfacing with the with the protocol um right before this I was actually a research partner at Poly chain which is how I join which is how I got to know about the protocol in his early stages and I just saw the ecosystem kind of flourishing so um I was super excited to hit up sh like hey what do you need to help with so I jumped on on the board beautiful and I'm enjoying having you here because I think I I want a lot of questions answered about the ABS ecosystem specifically so I'll be directing some of those questions to you some of the higher level more General stuff I'm going to throw to srom and I think that's actually where we're going to start this first question that I've got uh was

actually I think this is kind of a troll question but I think it's place it's a good place to start anyways this is from Perma weave why should I care about it s why should we care about Igan Lair different people care about I layer for different things if you are a builder You Know It uh opens up an opportunity for you to go build new things without having to secure a trust Network on your own you can build arbitary new protocols on I that's why you should care about it if you're a Staker you're staking in ethereum you're already putting your e at risk in validating the ethereum protocol might as well contribute security to other protocols and earn some rewards because of that if you're a node operator you're already running services for ethereum validation or some other

chain validation one of the big problems as an operator is every new chain that you validate you have to sell to a new community of token holders with igen layer what we are able to do is the same Community many many operations so that's the value proposition to The Operators if if you are building a rollup in the ethereum ecosystem you may be Bal neck by cost and bandwidth and the economics and IG da which is on IG layer is a solution that we're offering for you so that's the value proposition to the different sites beautiful a couple of these questions or maybe some of them will actually be um In The Same Spirit and so I'll ask them both at once and these

next ones Nema are going to you uh Udi Udi Suzuki reblock doe and also niski first question what are some unique use cases using Lan layer and then the second question it's unclear how much need there will be for avss beyond the early handful of examples where will the demand come from I think the spirit of this question is what are the flavors of avss what are the different types of services that can come from these things how do we even know that there is a there there neay maybe you can take this one yeah sure um so I mean right off the bat um a few things that are you know uniquely enabled uh by the protocol right uh let's just say bridging right the the economic security that that we're uh

dealing with here um can just enable much more um secure protocols um to enable uh cross-chain activities um the other thing that I'm personally really excited about um is the idea of co-processors right um where you have some cryptoeconomic Bond but in terms of what the actual functionality of the co- processor ends up doing whether it's generic or hypers specialized um that suddenly gets enabled by any L2 or even on the L1 itself right um uh th those are just a couple ones um but uh in terms of um another one that's really interesting I think um is what uh uh witness chain is doing right with with the proof of location and uh the

fraud proof Network right um you can have um you can tap into the power of uh the ethereum network to effectively uh deoy um it it trust minimize uh fra proof Watcher as opposed to having each L2 itself claiming that they're uh providing fraud fraud proofing service for you as well so I think that's going to make the community of of fraud proofing any kind of optimistic system very much more robust so one one one thing I would add to that is uh you know a little bit of a high level response uh you know when people think of the modular uh Paradigm the modular Paradigm is chains where Integrated Solutions for all the different things that need to be done on

applications inside that chain but the model Paradigm is oh maybe there are different modules maybe there is a data availability module maybe there's a consensus module and so on and normally when people think about the modular landscape they think of maybe there are three modules right maybe there is an ordering maybe there is a consensus maybe there is a um uh settlement layer uh data availability some very few handful of modules our vision is a little bit different you know if you look at the cloud era you see in in on cloud there are thousands of thriving software as a service solutions which are basically like very hyp specialized focused

services that if you're building a consumer application for example you know and examples of these SAS Services include like you know I want to offer a database as a service there are tens or even hundreds of database as a service the reason is there are different types of databases different types of you know there's a SQL database there's a no SQL database and all these different uh some optimized for AI some optimized for different purpose and so on so you have these hyp specialized software as a service solutions and what you do today when you're building an end user application is just concatenate a few of these software as a service solutions and then only build what components don't exist that becomes a consumer application that's our vision for what crypto applications look look like

these uh applications consume a bunch of services that are built on ION layer say and then tie all of them together and then offer it as a consumer application so what might this be and and Nema gave like some of the categories we're seeing a lot of interest interaction but the broader picture is you can think of let's say rollup Services right s you know your rollup you know ethereum is very deeply into the rollup uh road map what are the problems that rollups are facing can we kind of way find ways to solve it can we find ways to do sequencing instead of having a single Central sequencer can we have decentralized sequencing alt layer is doing its version of it espresso is doing a different version of it you can have um you know watchtowers which is

how do you know that the rollups doing what it is saying it's doing when you had a couple of big rollups then of course you can just um you know there's a lot of extrinsic incentives everybody learning Brock explorers watching the network you know optimism arbitrum a few networks people can watch but if there are thousands of these networks are who's going to be watching you need actually dedicated resources to be watching and incentives to make sure that they're doing the job that they're supposed to be doing so that's watching then you have bridging like there's fragmentation of liquidity across El tools we you know you guys have discussed this a lot in the show so how do we solve it you know can we have instantaneous bridging across these cell tools can we have new kinds of um um finality Services I I I come to final it

usually the sequencer tells me like hey you know it's final or like hey there is a I I promise you but instead can you have economic finality can you have me protection Solutions can you have an encrypted mle like what shutter is building these are all examples of the rollup adjacent services like some services that I can just integrate in as a rollup and then I get better as a rollup this is one category which you know meta category I would say we seeing a lot of traction but there's lots of other like vertical or meta categories we're also seeing Nema mentioned co-processors the idea of a co-processor is you're sitting in ethereum and I want to enhance the functionality beyond what like I can functions I can call inside

the evm can I call an AI and then I'm s inside ethereum L1 right call an AI and say Hey you know run like the previous for example I've run a DEX and I say that my price is now adjusted how much price I take as on Unis swaps adjusted based on on the competitive dynamic between how much Sushi is charging how much other dexes are charging and then I adjust my prices up and down the intelligent D5 for example can can happen if you have a processor like co-processor like that maybe you want to run a command which is actually like um you know a something that normally run only on a Linux machine so that that would be a general purpose Co processor maybe you want to run a command which is

a SQL Server Like I just call a thing and say that hey load this database a qu to the database and give me the response and I want that back on ethereum so these are the kinds of examples we're seeing inside co-processors but there are other categories like cryptography like can I run more advanced cryptographic like Solutions why do we need this because you know blockchains are usually like transparent ledgers uh that makes it verifiable but you know there's a trade-off against privacy so I want to have private but verifiable systems can I split the secret or information across the many nodes right so you need deep access into the distributed system into the consensus to to change it could be

secured multiparty systems trusted execution environments fully homomorphic encryption um these types of categories that we're seeing on on I so these are some of the broad categories the one thing which you know which is we haven't seen built already but we've seen proposals is how to modulate some uh commitments on ethereum L1 itself all these other things are outside of ethereum L1 but I'm a block proposer on ethereum L1 can I promise that I can do a pre-confirmation for a transaction I include in an L2 I'm staged so I I tell you as a kind of like a block proposer who's going to propose the next block I

promise to you that hey when my slot comes I am going to include your transaction so the L2 can have an instant confirmation with economic security so these are the kind of like range of things that we're seeing it's we did not invent all these categories most of these categories came Doo out of permission Le Innovation and that's what's really fascinating about this yeah one of the things that I think really gets people's imaginations going about IG layer is there it's pretty boundless the design space is pretty boundless but that also makes that leads into my next question here from Ethan uh Lipman uh if we have such a boundless design space and I'm assuming that also goes into the fee structure that these avss charge how will AVS rewards payouts

work if they can charge their fees in their own kind of network and their own token and ultimately goes down to the E Staker on the layer one what's like the supply chain of value flows from fees charged on the AVS Network while they're while the application is being its business going down to the eth Staker who has restak their ether uh is that a Nema question or is that a three ROM question I'll let uh take the first St at this but uh I'll try him with anything he hasn't covered yeah so the um the idea is that you know there is an AVS consumer like imagine an AVS like something concrete as an oracle right this is an oracle so an AVS consumer is an dap that is

calling the Oracle right so that's that's where it is starting but the DAP itself has a user right so I'm I'm calling a swap which or a lending protocol which needs to call an AR that Oracle now like triggering you know an ion layer AVS Oracle and that or so there's a fee that is being paid to the Oracle so the Oracle accumulates all these fees in a pot and the Oracle has a certain obligation to pay some fees to the igen layer stakers which is specified by the AVS model for example it says that hey for this much amount of stake I'm willing to pay this much amount of e or this much amount of my token emissions and that is paid period ially to the stakers and operators so

essentially it comes from either the user to the dab the dab to the to the AVS the AVS to the Staker uh and our operator and it looks like it's a lot of layers and like some people complain that this could be a big problem for um the uh this kind of an e modular economy but I think that's simply not true from our experience in the cloud like each consumer application in the cloud in integrates 15 SAS services in the back end these are all abstracted out that you know as an end user of a particular application you wouldn't even know that they're using all these things in the back end but they are and they have like you know different periods and settlement periods

for when they pay these fees to these SAS Services some things a monthly subscription some things a pay as a pay as you go all of these things are happening in the back end and that's roughly why we think that it is going to be okay that you have more complex value flows because our bet here when we're talking about permission L Innovation is really specialization right this is really what drove the SAS economy is and know if you just look at the broad Arc of civilization it is bends towards more specialization we all become more and more specialized in what we produce but more and more generalized in what we consume right and so these avss are going to be exactly the same they're going to be very specialized in what we produce what they produce but end users

are going to consume many of these things and pass a little bit of value to all of these and actually this makes for a very much more efficient economy because you know if you ask the question which is going to be the best uh chain it's a very hard question to answer if you ask the question who's going to build the fastest nosql database it's a much much more narrow question you can kind of zoom into like oh this guy has this experience that experience it's a B2B system and then actually they can go and build it so that's why I think it is okay not only okay it is actually great that you know you have hypers specialization and a sharing of value across all these layers through the modular

stack Neema anything you want to add there I I'm just fascinated by the fact that um this this permissionless Market uh is going to enable um all the various flavors of services uh I just view it as you know we've seen uh distribute systems uh these backend infra providers be able to do what they do over the past 20 30 years right um except now you have this cryptoeconomic angle attached to it um where you're building on a substrate that as long as there's a market for something no one's going to be able to tell you you can't kind of provide this this functionality so um it's it's very exciting yeah I think maybe part of the

spirit of the question is that like one of the reasons why everyone is getting excited about reaking is because we love ether and we love staking ether and we love the yields and we get more of it this way um but then there's also a tension I think maybe at least I suspect in AVS networks that also have a token where maybe a lot of the economy in that AVS network is the token and then that AVS network is paying in that token but then the E stakers want eth not the token and so in order to pay rakers eth that token has to be sold which goes against the incentives of the AVS network uh and so and and also then like are e resters just collecting a bunch of tokens as they yield like how how you have any Clarity that you can provide

here on this one R yeah I mean this is the power of uh perm list markets is you know we can't say whether this is going to be the market you know e stakers demand that they get only paid in E we will see that that's what happens if we see that yeah actually each stakers want all kinds of liquid tokens then that's what will happen if we see that actually each sakers want or willing to take even locked AVS tokens with a kind of like a Westing that will happen the the thing is we are very unopinionated about it and this is a this is an em we view this as an emergent property of what you know

it's like you know you can dial back you know say four years before the D5 summer and then if you go ask you know Justin or vitalic or somebody else at the ethereum Timo or somebody at the ethereum foundation what can e be used for right MH I don't know it can be used for you know in a D5 protocol like this it can be used in a in a d to provide liquidity provision but to earn like some Farm like you know all of these things are possible and we'll probably see a combination of all of these things and also stakers might segment by preferences there may be stakers are very pure and they just want to take e or they're not able to take for legal other reasons anything else there'll be

stakers who will take all these other things but going back to to the kind of high level thing about is there a conflict I think there is no like the idea is that the each staking if it is beneficial to an avas they're going to use it if each staking is not beneficial to an avas and their own token taking is more beneficial they absolutely should use their own token taking we are not taking a kind of a any opinionated position on it in fact igen L itself the contracts are WR to enable dual staking and multicum staking so you can stake e you're building an oracle you can have your own token and you can have e staking you can say that hey I'm borrowing some amount of security from a

neutral substrate and some amount of security from an opinionated like aligned substrate it's actually like game theoretically very interesting and useful to actually do that because you know the the the neutral subst is neutral and therefore more externally verifiable the neutral substrate can but can only be slashed on specific violations whereas my own token holders have exposure to just you know what is happening in in my system and that is a different kind of crypto economics and actually promoting this is what we think is the best thing right maybe the question uh that question specifically is like how to couple the income being generated from avss and then transfer it over to the stakers the E3 stakers maybe

that question is actually more appropriate to ask the LRT um level of the of the stack because they are the people actually more or less in charge of these flows so maybe that's actually yeah if I layer is any opinionated then I layer is not the actual place to Target that question maybe maybe this is a question for the lrts yeah we are also seeing interesting kind of payment management type Services emerging so somebody may say Hey you know subscribe to my you stake you know you don't even have to go through an LT un subscribe to to my payment Management Service I'll collect all these tokens I'll do an RFQ auction and then later I'll give you like you know your tokens in eat so these we view as again emergent structures rather than like enshrined

opinions so absolutely the lrts therefore not only uh LRT is a liquid reaking token it's a token that represents a position on igen layer but also represents an opinion on what set of operators should be used what set of avss you should stti to but also what rewards you want to collect and how you want to do trasy management of these rewards so the dimensions of opinion of a liquid reaking token are far higher than a liquid staking token right which is you stake stake in ethereum you're complying by protocol you have no price fixing abilities you just take what the protocol does so ethereum is this rigid solid underlying substrate whereas Ian L

is this very you know Anything Can Happen kind of a sub so it's funny that you say that because there's the variety of opinions that liquid reaking tokens could have which is good because there's so godamn many of them there's like 15 of them out there so maybe we'll we will see all those opinions expressed Nema these next set of questions I want to ask uh to you and these are all about um the logistics of avss um three questions here kind of all in the same Vibe can an AVS be aware about the state of other avss if an AVS creates a slashing event or slashing events how can other avss um be aware of that and become aware about the change in collateral for that's

being shared among abss and then are the scenario are there scenarios in which uh one AVS the slashing event for one AVS could trigger the slashing event of another AVS maybe you could takeen these and however you want yeah yeah I'll start with the first one um so the background image that Shon painted is these Services talking to each other right these abss um so in the same capacity um you can imagine how an ABS talking to another ABS has the capability to read some state right um now there's like onchain contracts there's o chain like software being run right but depending on the exact nature of what is happening for an ABS right

sure you can say like Absa uh their offchain contract whatever The Operators are running can watch for some State change on abp's contract right so there's no limitation in terms of like who's aware of what's and what's happening right um which I guess leads to the the second question a little bit in that the the design for slashing uh which is a feature we're going to come out with later this year um uh are still uh you know um we're thinking very thoroughly about how that's exactly going to work mechanically um to reduce the risk as much as possible um but then again um if an ABS uh operator can read a slashing event on another ABS right it can choose to um react to that under

whatever uh kind of uh you know path or like uh Behavior like like they just Implement so um it's it's as it's nearly like almost entirely up to the application devel ABS developers uh discretion in terms of how they want to um take slashing as a mechanism and as a reaction into account um and I guess the third thing which is uh can one slashing event uh have an effect on others um that would probably be a f a function or it depends on how the slashing mechanism ends up working in its final version right and the version that we're going

to release so um yeah yeah I I'll just add one thing to all of these things is every all of these events are reported back to ethereum ig layer is an ethereum Centric system right so any AVS which is actually there may be AVS which don't interact on a kind of like a routine basis with ethereum but all of the AVS state is reported back to ethereum on a periodic basis whenever it's reported it's that on ethereum it's on the ethereum contracts so now like I can write a contract which is dependent on all these other things like for example whenever a slashing is triggered on any AVS that's going to be reported to the I core contracts because it's reported to the i l core contracts I can read from

another AVS that yes that has actually happened and so I can react to it on a much more instantaneous basis so this is really why we think that you know having one coordination engine for all these Services is the best architecture rather than saying hey I have an LR let's say you know so this is a misconception that people have that oh you know now that the IG layer is there there and it's allowing lsts to stake somebody else can build another layer which takes lrts and then allows them to stake and so on but the the the right architecture is actually a much more integrated architecture which knows what's happening across all the different layers in fact this is one of the reasons we we keep saying that it'll be

even better in a future when ethereum can integrate igen layer more natively because the cross communication between ethereum and uh IG layer can be even tighter like how much of security is used by other things how much of security is reserved for ethereum all of this communication can be much much tighter if it's in the same system but inside IG layer all the state is on a common igon layer contract it knows when like your slashing has been triggered on any AVS it knows when State updates have been made on any AVS so all of this is routed through a common contract which is a common source of information and like Nema said you can pull from it as as often as you want but I think in the worst case PE you know all the AVS

contracts will pull at least at you know something like a few days period because you don't want to be desynchronized more than that and that's the kind of challenge or fraud proof period in IR moving on to a question from cloudy. eth shout out chrisy that's our community manager at bankless what happens if an igen layer AVS operator with delegated lsts get slashed what happens with the lsts where do they go sure want to take this yeah um yeah you know so there are two types of slashing that can happen on I so Nemo was mentioning that we have uh you know upcoming uh designs on the full slashing

mechanism but there are two kinds of slashing one is burning right you just burn burn the U stake the other one is redistribution right and the simple version to burn would be to just burn the uh you know LST this has some kind of you know different kinds of consequences for different people but one could also think of the U actually what is happening in the IG contracts is the LSD is not burnt the lstd is frozen into the contract state so you can't touch it you can't do anything with it it's just stuck it's like a stuck token and um but that's what actually happens in our V1 design but we are going

through some thinking on the different slashing architectures but the the thing is there's also redistribution and redistribution is I take the LSD and give it to the harmed party slash from the bad guy and give it to the harmed party so the other guy has the LST so that's that's what happens so that's why when an avas is whitelisting which lsts to use they should have some measure of Trust on the LSD because you know that's what's getting handed over across the different parties there is a more sophisticated architecture on burning which would be to like exchange the LST for e and then burn e m but you know uh we do expect that slashing is uh you know a rare event because operators have to be

malicious and they know they're going to lose their eat so they're not over optimizing for burning eat in this scenario Sheamus on farcaster orange Sheamus eth asks how will delegators operators and avss adequately be able to calculate risk the problems here will be incredibly Dynamic and involve unclear dependencies I'm not sure it's solvable let alone by all part ities routinely so I think it's a little bit of just the unknown unknowns about risks question uh Nema you want to take this one yeah this is a super interesting question um a lot of us have have spent some quite some time thinking about um and I guess to to approach this we can take a look at it from the perspective of each of these

different stakeholders right um for example um as an AVS developer what does r risk means to you right um is it um The Operators that are serving your abs um is it you know what lsds they have delegated to them um in terms of uh resters um which operators do I want to give my uh stake to right um what is the distribution of The Operators that look like uh for me to be able to assess who's riskier and who's not um and and in terms of operators um they also have this duty of being able to um or or wanting to rather look at the nature of an ABS and being and saying how complex is a logic for

serving this application right and how much does it how much loow does it bear on for example my physical infrastructure that I'm running right am I gonna run into um you know bottlenecks that affect my ability to run another ABS um so um all these all these risks um or things that um the various stakeholders have to take into account essentially compound right and depending on how you slice and dice an exact um combination um of these stakeholders you're looking at a different picture moving on to a question from a token terminal how do they this might there there might be no answer here this might be a market forces thing but we'll see um how do they see days you guys Ian

layer see the take rate in the ecosystem for example $1 paid by an AVS how many cents goes to rakers how many cents goes to operators how many cents goes to igen layer uh Nema do you have any answer here um this is still uh like being being thought out um and we we're taking a lot of community feedback from you know what ABS developers like to see um what operators are comfortable with like what ranges you know they're they their they think is reasonable um acrost what type of assets um so that's because it's such an nent space it's still something like we're actively figuring out um yeah that's that's that's where we're at yeah I I think you know that there are

dimension of pricing and that we thinking through the dimensions also carefully for example uh pricing need not be only for stake pricing should also be for the complexity of computation so we're thinking through like a all the other systems that exist and see like how we can absorb some of the best practices from there and also failure modes for example one of the things that happened in uh Cosmos interchange security is you know operators are running these other services by taking a cut of the fee other chains uh by taking a cut of the fee and if the fee volume is low you're expending a lot of energy but you're not getting like that much fee so you know

we should charge not only for stake but also for computation and so and when you're talking about stake there are also dimensions of stake where you know is my stake just going to be burnt for for this because that's a different risk versus is my stake going to be redistributed because there may be a different kind of incentive for that is my stake is my entire stake burn when I make a mistake in an avas versus is only a portion of my stake burnt when I mistake make a mistake in the zavas so there lots of different dimensions here and pricing will have to reflect all of this so it is still under like active design on the specifics of how the payments get distributed across all of

these things yeah and and to add on top of that the the reason why I said we're still so early in how this might look is because it's a problem that I think the broader industry is thinking about right when you look at salana and like you know uh fee markets or like you know the concept of the discussion on multimarket multi-dimensional fees right like the notion of what a fee even is around what type of resources who is providing those resources all that is yeah it's just being explored yeah yeah yeah right um I remember that that's still I think an ongoing conversation even when we talk about resource pricing inside of the evm as well right um F.E asks how long does it take to unstake uh and then I'll add

myself why is there an unstaking waiting period at all uh stre you want to take this one yeah yeah uh it takes seven days to on stake today and why is there a 7day period is because you know when you stick and there are and you provide security to Services what don't want is you stake you provide security to your service attack the service unstake and then go swap and get out right so unstaking should have a lag greater than the lag within which any misbehavior will be detected MH so that's why we have the unstaking period but it provides a un um uh unintended and you know maybe not that unintended from our

end an unintended benefit which is actually fund safety so you stake a lot of funds in ion ler and you know one of the things you have to be worried about what if it gets hacked what if people just like take all the eat and go that would be like an absolute disaster one of the things that protects it is to withdraw any unit of money out of the ion layer contracts is going to take a week I know already you've queued the withdrawal you have to wait a week this slows everything down to human time scale no monitoring there is paing there is a security Council comprised of you know uh 13 members who are outside the ion layer code Team all of that can actually kind of come in and

step in to do something in case there is a problem but that's that's a kind of like an unintended side effect of uh the core reason which is for fra proofs L coco. eth uh he writes great songs sometimes we put them into the rollup he asks how big do we think the premium will be that AVS is put on solo stakers is there a good enough way to determine who is a solo Staker he's asking because there's this is the one way that I can see that igen lay will actually increase ethereum's decentralization um s do you have an answer here yeah What premium abss will put like I said we are like a little bit of

a open market type uh project so we don't know what will happen but there are clearly absolutely clearly there are places where you know avss just want more decentralization than just one more stake right you know I want to sequence transactions for a private rollup I want to do like uh you know where the so the high level humanistic is where I cannot attribute false and slash for them more stake doesn't mean much but many independent parties consistently verifying and coming to agreement is actually the benefit so Services which have that structure imagine I take a

secret and then spread it on like many many nodes if it turns out coinbase is all these nodes then like you know I didn't like spread my secret to anybody like the guy has the all the copies of the secret so it's useless to like when I'm doing Secret sharing when I'm doing sequencing essentially for any nons slash false I want to have a decentralized committee and Okay so that's that's the types of things where like uh types of abss where they would absolutely want more decentralized operators the question is how much uh you know so one question is how much premium which I said you know it's emergent from the free market but the other one is how do you detect whether

somebody is a solo sticker I think this is a harder question but it's one of those questions where you know Nema mentioned earlier like one of the services building which is a proof of location and you know you have to be in you know many different locations so there's a geographic you know um diversity benefit for example that that itself so witness itself is an AVS but there could be other avss which not only track like operator location but do something more opinionated it could be an oracle that is basically a decentralization Oracle which looks at all the statistics of you know stake distribution you know historic uh correlation with exchange wallets all of these things and then puts in like a

mercel route of the are the most likely to be decentralized and this is not um you know opinionated from the IG layer point of view but this is a service that somebody else can say yeah I know I trust this opinion and therefore like I'm going to recruit stakers from this set or operators from this set again emergent Downstream rather than ins something that I think is particularly interesting on this front is uh when you have solo sakers irrespective of the kind of Hardware that they have right um one thing that we noticed thing is a lot of networks uh rely on this uh component of a keeper or or a relayer right um where the computational bandwidth of what you're actually relaying is not

super high but depending on where you're located um you might want to have like many many of them spread across all over the world right so the more of those you have within your system maybe it's latency you care about maybe it's like sensorship resistance whatever it is um if they can if you can go through multiple routes from point A to point B through these um you know lightweight maybe not even lightweight solar stakers um that's a very emerging property as well I think maybe part of your guys' answer is that it's actually not necessarily being the binary of oh you are a solo Staker you are one person but it's more about um how much net decentralization do you add to the system when we measure decentralization

in a variety of different ways um and because there is no good way of just like being a solo Saker actually only has is only one bit of information like there's other bits of information there um okay so there's four questions here they're all about the same thing we kind of have touched on this but I want to dive head first into this subject Jenia asks how do they respond about the risks of reaking being a daisy chain of potential risks Clairvoyant asks what doomsday scenarios have the team simulated or envisioned chain link God says how does or should the Igan reaking ecosystem approach risk management particularly managing the second or third order effects of lrts lrfi cascading slashing conditions AVS selection priorities leverage caps Etc

and lastly what's matter CFA says it looks like a lot of rehypothecation in trafi how do you quantify the risk I think this is what people's like first gut reaction goes to when somebody explains what reaking is like oh you take your e and then you stake it again and then you can also keep on doing that over and over and over again and people are like oh okay great so like you know cascading liquidation um what I'm sure you guys have answered this many many many times what's like the simple way to explain um the the the scenario here not that this isn't a risk but just like how do we provide more knowledge about what's going on yeah I I think there is uh you know we put out a

detailed risk document so I think that's maybe a good good place to point it we can put it up on the I'll put YouTube link afterwards I think that'll be awesome but um the high level thing is number one um the thinking of I think the term reaking is what throws people off if you know the the exact same concept for example already existed for many years in Cosmos called interchain security right it's called interchain security so nobody thought of the exactly the same questions but here's the the thing that when you stake um so the the the right

way to think of this is Shar security so imagine there are 1,000 protocols Each of which can have $1 million State okay so this is One World another world and you take the same fees and then now aggregate them into a$1 billion pool which is restak across thousand protocols which world is better let's say in the First World um to attack a protocol you need $1 million capital and then if you do attack it you're probably going to lose half of that 1 million or something like that in the second world to attack any one protocol you need one billion dollar Capital when people are comparing the two scenarios usually what they're doing is they're they're thinking there are 1,000 protocols with

$1 billion each and then on the right side there is 1,000 protocols with $1 billion restake but that's just not the right comparison at the same fee parity if there are 1,000 protocols each paying a certain fee to maintain $1 million stake you pay the the same fee you can actually sustain a$1 billion in a Shar security model so security is growing as you get more protocols it's a very like obvious thing once you stated like this but that's what's happening okay so you have much bigger security that is supporting all these things okay now you do lose something in the second world so the second world is better in the obvious sense that to attack anyone protocol you need like 1 billion but there's a little bit of a deficiency in the second world which is there is no unique attribution of this 1 billion to

any one protocol or you know how much is going to each protocol which is what is leading to all these questions about oh you know I don't know how much I'm getting there's a lot of risk and all of that and the way we solve it is by a principle we call attributable security attributable security is the following concept it says that an ABS gets this benefit of pooling but in addition AVS can buy a particular ability to slash and redistribute a portion of the funds let's say AVS one you know out of this $1 billion slash AVS one buys $25 million worth of attributable security which means if something goes wrong on that

AVS the igen layer protocol maintains this complex calculation which is what people are thinking how how is this going to happen this is going to be in protocol in the I protocol maintains what we call a solvency calculation calculation ensures that whatever happens how many other protocols trigger slashing simultaneously you will be able to redistribute your 25 million your attributable portion now suddenly all this question about like over leverage and everything else is just simply gone because each AVS has a unique specified attributable portion of slashing independent of what happens in all the other abss so this creates compartmentalization I know exactly what

how much I can redistribute if I slash uh and so this is actually the you know you know one of the tightest system of cryptoeconomic security it is not a leverage system with you don't know what Leverage is happening you know what it's not like that it is pooled security now you so what is the benefit you're getting the benefit you're getting is in the second world now even though I have only an attributable security of 25 million to attack my protocol I still need a$1 billion Capital I'm just much much more secure in the second world than having my own $25 million pool in the first world so that's that's the first picture about shat security and attributable security this changes the Dynamics of how people think about avss

compartmentalization so another like point is between IG layer so we can do this inside IG layer but to do this between IG layer and ethereum because you know we don't build ethereum so we have limited interfaces to ethereum but the nice thing with ethereum is eum is actually extremely uh high security so there's a lot of stake in ethereum $70 billion or whatever today's pric is there but you know there's a lot of stake in uh ethereum and the ethereum's security requirement is not all of this 70 billion and this is a whole other like discussion to understand exactly how much security is needed because staking is not fundamentally protecting

all the value all the tvl that is sitting on ethereum the staking is only protecting the value in Flight in in ethereum so a whole detailed discussion but what happens is IG layer as the amount of assets restake between ethereum and IG layer increases let's say it's 100% of ethereum stake is restak on Igan layer Ian layer will not sell attributable security for this entire 100% it'll take a back off say something like 50% we don't touch we're only selling insurance for the remaining 50% so what this does is basically the there is a buffer that's maintained to make sure that ethereum it's it's an accounting for making sure that ethereum security budget is at least that much

and the remaining is attributed to all these avss so that's the way we deal with that and then finally lrts right initially when you know one year back we were thinking about lrts we were actually worried that you know lrts can have all this financialization and risk and there is some truth to it but there's also the other side of it which is imagine there are no lrts so that's that's how layer is natively there is no liquid reaking position or like a fungible position but what could happen is I'm a Staker I stake in ethereum and ion layer and then now I have a position and you know people want to borrow against this position so somebody could

potentially build a borrowing protocol against an nigon layer position this position is non-fungible and unique and and so the only way if my position is now liquidated the only thing that you can do is now you have to go and trigger a withdrawal from igen layer and IG layer triggers a withdrawal from ethereum so what is happening is if you did not build a layer of LRT then anything that's happening in the financialized world has ripples that propagate down into I layer into ethereum you have to withdraw that money from I layer I ler has to withdraw that money from ethereum to honor that position and suddenly the AVS security fluctuates ethereum security fluctuates all of these things happen

with the liquid reaking position what happens is somebody has an LRT that's being lent and borrowed against they get liquidated the other guy just snatches the LRT there is no impact on ION layer or ethereum so financialization in a smart contract system is going to happen and if it is going to happen how do we mitigate and buffer it against you know ripples through the system I think this is the first thing that is already true is lrts actually create buffering for igen layer and ethereum okay now where is the actual risk I think the actual risk is in what people call looping right you go and know deposit an LRT get

eat and then go and you know or leverage reaking or leverage staking that's where actual risks are so this is now going into I stake I I borrow against an LRT and so the first thing is you cannot restake an LRT into I layer like that's not a thing this is again like some confusion in people's model or understanding of IG L I can't just go to go stake in puffer or something get an LRT and then stake the LRT back into I I doesn't allow for that so that's not a thing okay so and people are like oh I'm going to restake the restake and all this stuff like okay that's the first you know Baseline thing but this could

obviously happen in a lending Mar Market that I I take an LRT and then I borrow e against it and then I go and stake it and so on and this is basically a Mis uh assignment of risk in these Landing markets and that's what we should focus on why is this Landing Market allowing you to borrow an eat for an LRT with like you know only x% collateralization rather than like y% collateralization why am I a able to loop loop 50 times rather than only Loop two times so these are the kinds of questions that you know we should ask lending markets this is mispricing of risk in The Lending Market has nothing really to do with I layer and so but we are going to be like putting out some of these modeling

things so that people understand what is actually going on MH IM do you want to add anything to the uh the daisy chain of RIS conversation those the two points were pretty much what yeah what I was goingon to say so beautiful okay ju just one more thing on that is smart contract risks which we didn't talk about right like I I talked about earlier when talking about the withdrawal lag that the igon layer core you know e state is one way it is protected is because of the withdrawal lack because you know there's a time to pause and you know do security Council interventions and all of that but on the AVS slashing you know AVS could have a bug in the SL slashing

contracts that is really only buffered by The veto committee and initially we are going to have a single veto committee but over time you may have a Marketplace of veto committees like each each Staker and operator decides what kind of like veto committee they want and avss can work with them only if they agree on it's like a kind of like a mutually trusted party or like a arbitration agreement right so that that's actually one one thing I would say in this discussion is uh there is probably a misconception around how operators uh view their position in the ecosystem right um there will be a sweet spot in terms of how many abss they want to opt into um in contrast to how much

risk they will they're willing to take on right because obviously if you go beyond this threshold Beyond this like sweet spot you're increasing your risk at the you're in at the risk of losing yields right using like if you go down you can't really have that anymore so um the whole concept of oh yeah each operator is going to opt into every single ABS out there probably not the most practical scenario so um yeah on that note that actually perfectly leads into a question that I have so this one comes from me um I want to talk about uh possible design architectures for avss that have slashing as like uh are more permissible slashing whereas like I think we talk about a many abs

networks and we the general context that we talk about them is is like we want zero slashing events for which there will be many many services where like that is the optimum zero slashing events but may maybe there's a landscape to talk about ABS networks in which um people being slashed is actually a part of the way that the system works and it functions healthily like I could for example I could imagine if we all remember the old augur design the uh the old prediction Market where people would stake their augur tokens on the particular outcome of an event and if you staked in the minority you were slashed for being wrong and it was an oracle service and so somebody needed to be slashed in order for system to work

at all and as a result of this system we would had this Oracle prediction market and uh this thing would function uh so I could imagine an Ian layer AVS Network like that where the ABS role is um an oracle uh and I could imagine there are other AVS constructions in which a normal functioning healthily healthy operating AVS actually has frequent AVS slashing events and why are we okay with that well because the yield compensates for that and more have we we thought about sort of these constructions uh s you want to go first yeah I this is a really really interesting question and uh I should say this is out of scope for our current igen layer design and the

reason is you know systems like augur have a tyranny of majority you know you slash the minority based on a majority opinion what of the majority was actually adversarial mhm and then like tried to impose their like wrong opinion on the minority and get the minority slashed what really protects augur and augur as a system itself doesn't have it and the reason it doesn't have this problem is they had a way to Fork the auger token if something bad happened yeah and um you know this is where this is the boundary that vitalik has drawn for us that hey don't

do stuff assuming you know ethereum will fork for you and I think it's a really good design principle of like you know drawing a boundary so it's one thing to consider avss where you could get slashed you know a small amount in the normal mode it's another thing to consider an AVS which has a tyranny of majority and I think we just simply don't go there the avss should not have a way to slash a minority of people based on the majority opinion because if the majority colludes and tries to take on the minority then it's kind of like a big systemic instability okay and the other really important property I think the so whether avss or operators get slashed in

a in a normal mode is still okay but I think the main thing that we want to kind of like have as a property of I layer is that if you're an operator and if you're honest and let's say you can even add honest and effective then you will not get slashed right if you're honest you should should not get slashed as the bar but you could even like lower that a bit and say oh if you're honest and effective which means diligent and doing your stuff then you should never get slashed I think that is the the because that's what retains all the properties that we want about let's say an LRT right when somebody's holding an LRT what are they actually holding they are saying this is this would be equal to one e if I trust that the operators

are honest that's a really really nice kind of uh mental model that if I trust these operators are right then this will be equal to one one e because you know all these other things won't happen but if we start loading un uncalculated risks into this layer so this is again like a very sharp boundary we only want to do validation services and we only want to do validation Services where risks are what we call endogeneous if you are an operator it should be under your control not to get slashed that's a very very very important problem property because once it's like that then all these like Cascades and problems that people have about like I they don't show up but if

you don't do it like that then you have all kinds of problems and the system kind of like blows up in ways that you can't predict okay so that that sounds like a very firm line that you are drawing that you are saying Ian layer will not cross that line it will not cross the line of uh allowing for Effective honest uh operators to get SL to be slashed absolutely I think that is that is what is what makes it uh all the rest of the properties of the system to actually work okay okay and I think you also kind of implied that that same line is the line that vitalic was labeling in his blog post don't overload ethereum consensus that's correct in fact the example he gave was a USD to Brazil Oracle and and something crazy happens in Brazil and then you have Brazil North

and Brazil South and now you know a majority is voting with Brazil North and a minority is voting with Brazil South but actually they were all honest and then and suddenly you know you're slashing all the guys who voted on Brazil South based on the Brazil North guys and then like you know it's it's a know tens of billions of dollars of each slashed for just being honest and doing your job but but just not knowing that this thing can happen so that's the overload that overloads ethereum social consensus because then what happens is right now you know do do we save them do we not save them the shelling point is not clear and it just like confuses everybody so that's why we draw the line in there is because objective attributable

slashing for malicious operators that is what I can layer is intended for intended for yes and also IG layer is also intended for permissionless Innovation and at some point won't won't these two things Collide like if permissionless can I go build my subjective augur layer yeah great uh great question so the way uh we mediate this Is by The veto committee so the you know if I'm a Staker right look at what I'm trusting I'm trusting that either the AVS is correct or the AVS contracts are correct or the veto committee is good this is what I'm trusting when I'm opting into

the system that's the flow of trust and so it is up to the the the slashing veto committee to make sure that it is able to understand and like veto the slashing decisions correctly so which means it also has an onboarding function because it needs to know what is it going to slash or not slash when such a decision comes up so there is an onboarding decision which sits with the veto committee so there is a bit of permissioning there but over time like I said we're going to have not one enshrine veto committee but inter subjective veto committees each of them you know there can be different groups that different stakers trust and you know they mediate different oper you know different avss and what this does

is in in that world each of those vetto committees still needs to onboard those avss because that's how because they are in a in a position of mediating this dispute they need to know what they're mediating that needs to be like objective and concrete and to the users of the AVS you need a kind of like a clear understanding when you will you know get slashed and when you won't get slashed and the problem with the majority um kind of a so so in in a permissionless world it's possible that some veto committee you know onboards Services which have this kind of properties but those committees you know will be less trusted and less stake will be staked through them that's what will

happen so the the thing is by allow by allowing for most of the genuine use cases of you know I layer which don't create these risks through the normal mode you prevent the risk because most of the yield is from genuine useful Services which think through these risks very carefully and somebody is going to build these things which don't think through this carefully and we want that to be exposed to the Staker and the way it's exposed to the Staker is none of the big slashing V committees are going to onboard that avas now I have to go and Trust some other random veto commed and like hey why are you doing that so that is how we stratify risk and make sure that people understand what they're

opting into beautiful that that actually has given me a ton of clarity about about this whole system uh getting back on to the uh the questions here as we wrap up getting closer towards the end I'm going to put these two questions together um aroon asks will they eventually switch to their native token staking to secure the network or is it always going to be staked eth always and also green curry asks what's the point of the token so now we are getting into the theoretical igen layer token uhal IG layer token doesn't exist so we'll we'll review this question with if and when it exists all right so we're punting on the on this question uh that brings me to my next uh my but just one

one important thing the reason we have each staking as the kind of central thing in Iger is you know this is a this is important to understand for you know I see a lot of people for example asking hey when you have a lot of different ders and fees go to all these ders what is the role of e and I just want to like take a minute to answer this here and the role of e in our view is e is the money that is programmable right you know you have the famous phace programmable money right so e is programmable money and what that means is as e becomes programmable money it goes into all these different rollups and layer tools and other places where it is being used and because these are

different necessarily distinct zones when a lot of the risk is now priced in E right for example I move value between one roll up another rollup that is priced in E and then you ask what is the right collateral to secure a bridge which moves eth from one chain to another chain that collateral should be like e or e e ped as closely as possible and so the fundamental role of e is e becomes this money that goes into all these different chains and actually doing something interesting there but then what ion augments that is with if I want to move this risk around you want to buy insurance or Security in the

same unit right like if all my liabilities are denominated in USD I might as well get credit in also USD like that is how I match is the lowest volatility Peg if my um if I'm moving e around on all these rollups but I have insurance and some other token then there there is a relative volatility of the token to e that I need to price and that just makes it more expensive to have other tokens of security for moov around e but that could be other things that are happening for example USD usdc becomes the dominant unit in all these different chains is a different world and in that world somebody may say I want to stake usdc to secure you know a

usdc to usdc transfer because that is the closest volatility Peg between that so again view this as emergent you know market dynamics MH and that's why even in I you know we are initially focusing a lot on each staking but you know who knows op token can be staked and you know for an OP token like sequencer OB token can be stake for an AR token finality and so on so these will all be happening also but I think the fundamental role of eth as money is what makes e as collateral pingpong each other to create this very power ful Network effect so that's that's our vision that would you say uh eth is like

the Bootloader for the igen layer system as an asset and then other uh tokens like ARB for arbitrum op for optimism can then kind of uh enter the game when that is when the system is ready for it I I think that's that's possible because you know many of these different uh uh systems may want for example the way they they might start off is they say dual staking with ar and and e but unless the the belief is that you know any of these other tokens acquire the Mone that e has been able to acquire risks will be primarily denominated in E and therefore e as collateral still dominates in terms of utility but

there'll be other tokens take for purely like alignment and other purpose and they they are the team that build it they have to choose what token needs to be St so that that is absolutely going to be a free market right so long as eth is the money of the metaverse this is probably where it's going to go moving forward if the metaverse demands something else Ian lay will serve that demand cool uh last couple last questions two more questions um Gabriel Haynes we all love Gabe he asks of course when theoretical token that's the answer okay all right David me asks when main net oh yeah V early Q2 early Q2 early Q2 all

right is there any other questions that that's the end of my questions are there any questions that you guys wish that I had asked what questions should I have I have asked the questions around risks were uh really good so you know another time maybe there's a there's an opportunity to dig into just like you know in in a detailed discussion on some of that because I see a lot of people don't understand the models clearly enough to to know what is actually happening um but I think they were really good some of the discussion around for example can you have an augur article I think your your questions around that were really awesome um all right guys well if that's all the questions that which it is I really

appreciate Nema it's good to reconnect and I'll probably if am I going to see you at East East Denver absolutely and a lot of the team from M Labs actually hell yeah yeah yeah we'll definitely see you at e Denver as well absolutely y right Bank list Nation you guys thank you so much for for coming on and answering all these questions from myself and the community I feel much smarter about Ian Lair bankless know the deal crypto is risky staking is risky reaking is R risky you can lose what you put in that was a terrible pun but we are headed west this is rentier it's not for everyone but we are glad you are with us on the bankless journey thanks a lot mantle formerly known as bit Dow is the first Dow Le web 3 ecosystem all built on on top of Mantle's first core product

the mantle Network a brand new high performance ethereum layer 2 built using the op stack but uses IG layer's data availability solution instead of the expensive ethereum layer one not only does this reduce mantle Network's gas fees by 80% but it also reduces gas fee volatility providing a more stable foundation for mantle applications the mantle treasury is one of the biggest Dow owned treasuries which is seating an ecosystem of projects from all around the web 3 space for mantle mantle already has subcommunities from around web 3 onboarded like game s for web free gaming and buybit for tvl and liquidity and onramps so if you want to build on the mantle Network mantle is offering a grants program that provides Milestone based funding to promising projects that help expand secure and decentralize mantle if you want to get started

working with the first dowed layer 2 ecosystem check out mantle at mantle. XYZ and follow them on Twitter at zerx manle are you launching a token is it already live how are you managing the legal and tax for providing token awards for your team toku simplifies everything thing about managing token Grant compensation and you can get started with them for free you'll have access to topnotch Legal end tax support to handle the distribution and management of tokens for your team toku caters to every step in the process from userfriendly legal templates for granting tokens to tracking vesting periods and calculating withholding taxes toku understands every Grant structure token purchase agreements restricted token Awards restricted token units token options and all the other

ones toku is already simplifying this today for leading companies like protocol Labs dydx Foundation Mina foundation and many more you can learn more about how toku can help you streamline your token management and get started for free visit toku tok.com bankless or click the link in the description below