EigenLayer and Restaking Explained by Founder Sreeram Kannan

[00:00:00][Music] foreign [Music] Gardens YouTube channel where we discuss all things blockchain decentralized web smart contract programming and cryptography Katie and I are your hosts as usual and today we are welcoming sriram Cannon who is the founder of layer Labs which is building egg and Lair and beyond that I Recon that you are leading the blockchain lab at the University of Washington and you've been teaching there for quite a while sriram we've got a lot to talk about welcome uh thank you so much Sam and Kailyn it's a pleasure to be here on the North Koreans podcast so we'd like to go with this uh fast get to know kind of a set of question when we're starting a podcast so could you succinctly introduce yourself and how you basically came to the idea of making like again there what's the background of egg in there please yeah I have been working on blockchain over the last five years my interest in peer-to-peer networks dates beyond that my PhD from like Masters and PhD from 2016 to 11 was in this peer-to-peer wireless networks when I got back into like uh this idea that potentially peer-to-peer networks uh and uh blockchains can be used to accelerate the rate at which we cooperate with each other so I have this phrasing that I say that just like the internet is the information Super Highway blockchains and crypto could be our cooperation Super Highway and so this became kind of like my driving thesis and I've been working on that from 2018 at the University of Washington as an academic and working on infrastructure layer like considers protocol scalability data availability the game theory around some of these things so that's what I was doing and one thing I found which kind of led me eventually to eigen layer is that whenever you want to build a new consensus protocol whenever you want to build a new infrastructure layer like an oracle or an authentication Service or a secure multi-party computation or anything sophisticated which needs root access to the distributed system that's not possible to build with existing cross networks every new innovation needs a whole new trust Network and somehow I could never wrap my head around it because the idea to me of uh or at least the biggest value of organization of blockchain that I saw was that it is separating who brings trust and who brings Innovation as a smart contract developer I don't need to be trusted I'm borrowing trust from the blockchain platform which is a huge value proposition just like separating Innovation and capital is the whole proposition of venture capital like somebody else brings capital I bring Innovation and we work together and create this very interesting thing blockchains for the first time enable the separation of trust and Innovation like blockchains bring to us as a developer I bring Innovation and then we throw this together and then like out Comes This Magic output but that was not true for infrastructure layer Innovation and that kind of bugged the heck out of us so that led me into building iron layer of you know viewing ourselves as kind of the customer for this thing anybody who wants to build new distributed systems the other thing I saw is like there was a lot of opportunities for smart contract developers to go Express their like creativity in one of thousands of existing projects or go start their own project but such an opportunity was not available to distributed system Builders you have to go work at one of the five big blockchains or oracles or whatever otherwise starting your own is very very difficult so anyway so that's the context of how we got it right now yeah so could you briefly explain what restaking is and why restaking is needed yes the idea that existing trust networks only serve one particular consensus protocol and one particular stack right on top of which you know you have this things like a virtual machine uh is how all existing blockchains are designed but the thing that we saw is no actually once you have the root of trust what is the root of trust of a major blockchain like ethereum is staking right like there is an amount of economic capital committed and there we are committing to participate in this network if I can take this taking and apply the trust to other things other people may build on top then I can actually make credible commitments for example let's say I have you know Sam has a new Oracle protocol and I'm saying hey you know I want to run Sam says hey here's a note software that you have to download and run to participate in the Sam Oracle and what I you know all the node operators and sneakers what they can do is they can download the software commit their Capital to you know slashing conditions on Sam's Oracle and then say that hey if I misbehave on Sam's Oracle I could potentially lose my deposit but if I behave right then I'll get like some reward which could be you know every time somebody queries Sam's Oracle he's basically earning a cent or uh you know away or whatever and then that comes to the node operator so basically that's the idea that of restaking is you're putting the same stake that you've already staked in the ethereum platform at additional risk in return for offering additional services and getting an additional yield based on that so eigenlayer is a two-sided Marketplace on the one side we have stakers opting in to this service and it's an opt-in platform you know of course we cannot force all ethereum nodes hey you have to do this who are who are we to say that but we can offer an option that anybody can opt into from the Staker side on the other side we have service Builders you know uh Kalin may have her own like data storage service and they say Hey you know everybody please download and run this data storage service for me and basically that's the other side of the market so we bring these two sites together so that another way of thinking about eigen layer is it's a Marketplace for decentralized trust we are buying and selling decent clusters if decentralized trust is the core essential ingredient of blockchains we need to have efficient markets around that and eigenlayer is the marketplace for decentralized trust I'm very curious to understand at a high level how do you implement that do you make my ethereum node double sign to punish me to counter incentivize me not to mess around if I subscribe through a middleware to other slashing conditions how do you build this in a secure Manner and how do you observe malicious behaviors on subscribe Services absolutely so there are two parts to this how do we observe and enforce uh how do we observe malicious behavior second is how do we enforce the penalties I'll start with the second one so that gives a context for how the enforcement works there are two ways in which you can participate in eigen layer number one is called native restaking or native restraking is you stake in ethereum you set the withdrawal powers to the eigen layer contracts so the first thing to be clear is eigenlayer is not a new L1 new network nothing it's just a series of smart contracts on ethereum eigenlayer is a series of smart contracts on ethereum what it does is it manages people's stake or restate right and in one way of restaking is you deposit your money into the ethereum smart contracts because you're an ethereum Staker but that means that since you're depositing your money on the ethereum contracts you can specify who has the ability to withdraw it and you set the withdrawal powers to the eigenlayer smart contracts and in the normal mode you would basically just trigger the eigen layer contracts and it would withdraw on your behalf and give money back to you right so that is what would happen if you had behaved normally but if you had behaved maliciously and I'll come to how that is detected if you had behaved maliciously on the eigenlayer contracts then eigenlayer could withdraw money on your behalf take a portion of your funds and only give you the remaining portion so that is the eigen layer slashing is actuated okay now how is malicious behavior proven and this is service dependent each service essentially rights as smart contracts that talks to the eigen layer of smart contracts and the service smart contract has three functionalities registration which of the nodes can participate in this particular service or you should have at least take 30 to eat right like that's an example of you know our condition but it could be like oh you need to stay 320 or you can say three right like whatever the set of condition is the second one is the positive incentive conditions which is whenever you you know you're served like let's say Kalin has a chain built on top of a wider layer she says you know whenever you serve the Kalin chain when you make a block I'm gonna give you like you know one heat right distributor to all the stakers something like that so that's the payment condition and then number three is the slashing condition which is the negative incentives it says how does uh slashing work uh on the Kalin chain and like Sam was alluding to one-way slashing could work is on a chain the biggest like source of error is if you double sign a plot like you sign one block on one you know block number 30 you send two different blocks and call them both block number 30 that is a re-arc attack and that should be punishable and on the ethereum uh smart contracts that uh the service contracts that let's say Kalin chain has it'll say yes if you had signed two signatures with the same header and same block number that's a slashing condition so that's an example for how slash in conditions can work is they are provable through ethereum smart contracts that something wrong has happened you know different Services need to figure out how they need to write their own slashing conditions if it's a data storage protocol you may say that hey I'm going to ask people to randomly show some chunks of the data and if they don't show their chunks of the data then you know they get slashed so that's a slashing condition for a different kind of like a service so each service needs to figure out how to do this and but the service talks to the eigen lab contracts and essentially as a Staker you have opt-in to you can opt into any subset of services you may like kln's chain but you may not like Sam's Oracle and so on like you can decide which set of things that you actually want to opt into you know just to conclude it I was mentioning that there are well there's one two ways to opt in one way is to set the withdrawal credentials and we call that native restaking there is another way to opt into eigen layer which is through liquid staking derivatives like so if you already had a liquid staking derivative you know the stakebeat from Lido or something else or the rocket police you could basically take that and deposit it into eigen layer contracts and that's just enough you know you don't need to set withdrawal credentials or anything so those are the two different types of opting into eigen layer so other than the oracles can you give other some examples of the main use cases for eigenlayer yeah absolutely so you know one way to think about it is as we see the ecosystem if you just like take the ecosystem backward to like before 2014 when ethereum came into the scene you see like Bitcoin was there and any new innovation for example you want to build a domain name system and then people said oh yeah here is a new domain name system built as a proof of work system called name coin now this is a whole new chain it needs whole new trust Network and then every new application at that time required a whole new trust Network and what ethereum did is to aggregate all of them into one umbrella and say that hey anybody can come and build a tap on this common trust Network and you know what we saw is over time many many dabs were built till essentially ethereum went to Breaking Point there is not enough block space to actually build more dapps because gas is going to be expensive and so on and then we saw this like disaggregation from ethereum to all these alternative systems somebody said hey I have a better consensus protocol somebody said no actually I'm going to do a better virtual machine somebody said I have this better chat security model and so on right so we saw the disaggregation again happen at the lower layer which is at the consensus layer now what eigenlayer enables is re-aggregation so any new idea that you have right you may have an idea for a better consensus protocol like sui and Aptos too and to today you were required to actually go and start a whole new chain and it's a very difficult thing to do you could have built soy or Aptos on top of eigenlayer so new l1's is an entire category that can just kind of be internalized at least to the extent people want to on top of eigen layer the second one is actually the thing that I'm most excited about is services for modular ethereum so one of the paradigms that we're seeing is you know instead of thinking this about this entire blockchain as like a monolithic bundle that does everything together and serves different functions together you can think of the modular Paradigm says that each module each specific functionality can be built on a separate distributed system and then we tie all of these together and then we get like the service that we want and all separate consensus the separate data availability a separate execution a separate settlement a separate decentralized sequencing and so on right and the problem this is a brilliant you know landscape because when things are modular then rate of innovation accelerates because you know people will compete on each module specifically but one problem with the modular ecosystem is when you tie up modules based on different truss networks tie them all together to form your emergent service the problem is then the service trust the security that you inherit for the service is the worst of the trust of all the modules right your system could be attacked if the consensus is out or if the data availability is out all of the settlement is out of these decentralized sequencing yourself or the Oracle is out any one module can basically like corrupt your entire system and the eigen layer view on like how to build a modular ecosystem once you have a common trust Network now anybody can compete to build modules on top and all of the modules are built on potentially the same Trust and then that means when you compose all of them you're not losing trust while at the same time preserving free market Innovation that essentially people are going to say and compete hey for example one of the first Services we are building is a data availability layer eigen da it is basically just does data availability and because it just does that you can actually do it much better and you can and innovate and experiment on it but we don't claim that we have the final solution for the data availability problem there'll be many people who will compete on either lab to be the data available similarly there'll be like a competition for who would be the decentralized sequencer because you know now decentralized sequencing is a major problem for Roll-Ups similarly you can have like a super fast settlement layer settlement on ethereum takes you know 12 minutes 12 minutes is a really long time and you cannot move data in and out of ethereum with very high guarantees uh because of the finality time of ethereum so that's another thing somebody could say hey I've got like some of the cool ideas from Suey or somewhere else and new academic papers and I'm going to build this super fast element layer that arrives at finality in one second but has the entire economic capital of ethereum behind it so that could be another kind of service you could build a bridge service on eigenlayer because you want to move data from all these different ecosystems in and out of ethereum and you know Bridge service is another example and then the last category which we're actually super excited about and uh maybe also of Interest particularly to note Guardians is Mev management when a snake as an operators opt-in to eigen layer they can actually commit to certain Mev rules why on ethereum not on a new L1 or new like service or anything you can actually modulate the rules that you've committed to on ethereum why because your restaked on eigen layer you can say hey I'm going to participate in this particular auction market and if I misbehave on some rules of this auction Market I can get slashed so now the other side of the auction Market can trust you or you can say that I am opting into uh this encryption service you know I'm going to only include at least use a portion of the block to include decrypted transactions of certain encrypted things and that's a service you could opt into you could opt into something we we designed called Mev boost plus plus which is basically you have Mev boost today where one of the problems that we're seeing is node operators don't have the freedom and agency to add transactions and they have to just they're basically selling the entire block space out and you could say I'm selling the first 80 of the block space out and the remaining 20 I can add my own transactions you can say you know even driven actions you can say that whenever there is a Liquidation on compound I have to First Take out the liquidation before filling up the rest of the blocks so this basically improves the time to Liquidation on these like D5 protocols if you improve the time to liquidation and you actually improve Capital efficiency so there's a range of things you can do with eigenlayer which completely modulates the ethereum ecosystem and what you can build on top of it one way to think about it to wrap it up is any innovation any idea that you have in building a better system can now be internalized back into the ethereum ecosystem so that's our kind of broad project that we're in it's an opportunity for node operators Builders different categories of participants different categories of actors and now I would like to explore you a bit further how all of this is implemented how all of this works Etc so you mentioned that these are middlewares middlewares can be built by layer Labs by any developer and you can propose your middleware and they exist in the form of smart contracts and there are all sorts of predicates and conditions that can be defined through those I was just curious to understand how this would function for let's say you're trying to export your ethereum your security on an ethereum node to for instance secure a new narwhal and bull shark blockchain that's coming up how is this exactly going to work the communication yeah uh great question so when you're building a service you know service is just a generalized abstraction of you know it could be a middle where it could be a chain it could be anything when you're building a service on eigen layer the way you think about it is you are actually building two particular components one component is the Smart contracts on ethereum which which specify the three things that I said registration for that service pass through incentives for the service and the negative incentives for the service that's what lives on ethereum and everything else you know actual running of the novel and task or bull shark or whatever consensus protocol that you may have is all abstracted out into a node software that you also deliver so you are when you're a service you're delivering two things an ethereum smart contract which specifies registration payment and slashing and you're also specifying a node software which is completely arbitrary you could write it in C plus plus whatever you know you ship it as a Docker container you know many ways to do it and essentially what that does is when you are an operator so here is the workflow of how the the system will work if I'm a sticker then I say that hey I'm actually opting into this like new consensus chain okay that is basically when you're specifying that you're saying that hey I'm opting into this particular service contract right and the service contract specifies I'm going to get paid this much I'm going to get slashed this much and I'm going to kind of like this is the registration Edition and you have the dawn chain and then I'm going to download this node software separately off chain and then run it and you know the way the node software should be structured is that like if you run the default version of the node software you should not get flashed right like that is the padding condition so if you run the node software and one thing we take care in eigen layer is we don't want to slash nodes for lightweight things like off time we want to slash nodes for heavyweight things like malicious behavior lightweight things you just reduce the two Bots or like remove the node from that service or whatever but you don't flash very so that's the kind of core Distinction on eigen layer but the basic idea is that's it so when you're writing a service you're completely free to write very complex general purpose software as the node software but then you have to write like the payment and flashing conditions as smart contracts it's a bit as ICS where you have to run the cosmos Hub Gaia client and each child chain let's say you also have to run their clients in order to okay so I like to see it as a general purpose shared security Marketplace or system that's definitely very interesting thank you so the question I have is when ethereum is restaked on a sort of a reuse of Economic Security from ethereum itself then for the slashing that identifies the misbehaviors on ethereum can you explain how this is sort of different from fraud proofs on optimistic Roll-Ups that also inherit the base layer security oh great question so you know the broader question here is like hey there is already like a layer two Paradigm right which is basically trying to take the security of ethereum and Supply it to like some other things what is the kind of difference addition of eigen layer to this how does it like map between the two things so the great thing with layer tools is that essentially you are transferring Trust on execution at least the current layer to Paradigm is basically saying that I offload execution right virtual machine running and I make State claims on ethereum and when I'm making State claims that this is the state of the virtual machine on ethereum I either submit a cryptographic proof of Integrity or there is a fraud proof which then gets executed only when somebody challenges these are the two parallels but essentially what you're getting from this is execution layer claims what you do not get from this is maybe the interesting thing to start with and that's all the things that you can do on eigenlet what you cannot get from this number one you cannot get for example you want to run an oracle okay so I want to run an Oracle and I want to get ethereum trust right there is no way of doing a fraud proof that your Oracle input was correct or not so that's number one right but if the ethereum nodes you know which are restaked on eigen layer if they are running the Oracle you have two layers of trust so this is a general thing in eigen layer actually in any decentralized trust Network there are two layers of trust one is the trust that comes from Economic Security which is you know I have so much of stake versus there's also a trust that comes from decentralization oh I have 2000 nodes you know distributed around the world or ten thousand nodes around the world that are running this service and eigenlayer is a marketplace where you can kind of rely on both sides of trust and for example in the Oracle case the way you would rely on decentralization trust is you're saying hey thousands of nodes which are ethereum stakers are actually opting in and providing me this price feed and they're very difficult for them to collude together so likely a majority of them are doing the right thing and therefore it is helpful so that's one part of it the other part of it is you could create slashing conditions which are much more nuanced so for example when I'm providing this Oracle input I could say that if somebody challenges my like Oracle input then what I could do is then maybe we have this like much bigger Dao which you know of course cannot meet for every single Oracle input but you know when there's a contested Oracle input this bigger Dao meets and decides hey this is the right thing and we're going to go by this answer and then some of the other nodes could get slashed so there are the configurations of how you can inherit the economic security and the decentralization based security are quite enormous I'll give you another example of a service which is not a normally a layer two but something we're building is data availability all these Layer Two Solutions actually offload execution like we talked about but they publish data to ethereum itself right basically you need to publish data why do you need to publish data is because for somebody else to understand and interpret that you're doing the right thing they need to see what data you're using for your computation even though the computation is not done on ethereum the plain Text data is posted on a theme and now what you could have is but the problem is now the data bandwidth of ethereum is limited and on eigen layer because we are inheriting the same decent Life Trust potentially if all stickers opt-in and the same Economic Security you could actually build a data availability layer on eigen layer and that's one of the first services that we're building where you can actually inherit both the decentralization and economic security from ethereum for data availability uh another example is this is one problem that we already see with all the layer tools is how do you get fast finality let's say I'm a cryptographic like a validity proof based roll-up and I want to get fast finality I don't want to wait for 12 minutes for ethereum to get finalized before I can declare to my users that are final what you could do is you could build a super fast consensus protocol on eigen layer right and if all eats take opts in then essentially what you get is you know you write your data to this layer to the Superfast finality layer and this layer then writes the data back to ethereum eventually but this super fast finality layer has for the entire Economic Security potentially of ethereum opting in so you get instant you know one second finality times but at ethereum security so restaking completely changes the landscape of what a layer 2 solution is you can basically build anything that had to be a new chain that had to be in your system back into ethereum so that's the the flexibility of of iron that's crazy so in the case of a layer d a in this data availability solution I monitored how much bandwidth is actually used by the client itself the software and you take advantage of the Delta of bandwidth that's not being used I suppose that cloud providers are going to hate you as finally what you subscribe for at gcp or Amazon is gonna be used at its full potential so that can create some Scandal hezner is gonna fire much more than I mean just to clarify the eigen da position actually eigenda is designed for very low node bandwidth requirements for node okay okay what what is going on the idea of eigen Da is that and and this is not our original idea this is actually uh already on ethereum's roadmap and you know the great ethereum researchers have already figured this out is on the roadmap of ethereum called donksharding and the idea is that every node doesn't need to download the entire data which is what is going on today even with the upcoming upgrade called eip4844 or photo done starting every node will download every unit of data and to give you like a sense of how bandwidth is utilized on on ethereum and pretty much every blockchain today is the node bandwidth requirement to participate in ethereum validation you may need something like two megabytes per second that's 16 Mbps right megabits per second or two megabytes per second and you need two megabytes per second but the system bandwidth ethereum's data bandwidth if you just use ethereum for writing data the system bandwidth's only 80 kilobytes per second much lower than the notepad single node bandwidth right so why is this going on why are we not utilizing the node bandwidth properly it's not because we are leaving a lot of slack it is because the overhead of the peer-to-peer Network and consensus and all that is quite high and that takes up pretty much all of this like differential between the Max throughput and actually the node bandwidth so what we're doing with eigen da and is already on the roadmap of dog sharding but we just have because we are an agile layer that's just doing data availability and we don't have to move this big ship called ethereum we can actually build some of these ideas where every node doesn't touch all the data every node touches only a small portion of the data but the data is encoded and cryptographically committed in a way that even if a lot of the nodes go offline you can retrieve all the data from the chunks that are stored in the remaining nodes so the design spec that we are building too is each node only needs a small fraction more than the ethereum node bandwidth line instead of two megabytes per second you you may need another like half a megabytes per second which is nothing but the system bandwidth that we can actually offer is because we are spreading data through all these nodes the system bandwidth is something like 10 megabytes per second so it's a so that's why we call it hyperscaling ethereum is because you horizontally scale the data so that like nobody really touches all the data everybody touches only a portion of the data but they know they're getting the right portion of the data because it's both Erasure coded as well as submitted using this kcg polynomial commitments so that's really what uh eigen da is so hopefully we won't trigger the problems that you're talking about here I think we're bottleneck by ethereum capabilities definitely but it's a legacy blockchain but let's say we're talking about Aptos and Aptos has its own Roll-Ups or Solana which have much higher bandwidth requirements then my joke would be valid but that's absolutely absolutely actually you can build like absolutely crazy requirements on top of these systems but one principle you know this is at least to the extent that we are building and we are we want to on both cells is one principle that we are using is how do we essentially get scaling without sacrificing decentralization and this is a this is just an aesthetic that we aspire to but eigenlayer ends up being a permissionless open platform on which anybody can build anything even on ethereum you could say that hey if you're opting into my super chain you need to have 10x ethereum span with and that's perfectly allowable on eigen layer it's just that that's not the design space we are optimizing for ourselves but I'm sure other people will push the boundaries on so many ways on top of it amazing and that also introduces I believe a new cost Dynamic for security the price of security because when you bootstrap an arbitrary blockchain well you have to agree with some corporations some entities some people to bootstrap some node which is going to costs cloud and or defense if it's self-hosted whatsoever you need to incentivize them you have an emission distribution curve inflation etc etc but here people have can just subscribe to as many slashing conditions as they want and it's a bit less painful than I mean they still need of course to offset the cost of the new node that they're going to run on on their provider or at home but I really think that this also introduces a new cost for security and you can really bootstrap new systems with distance security at an affordable price right yeah absolutely so here is an example of the cost of security today right when you think about the cost of security we are securing a network there are really two distinct things that are going on one is the operational cost of actually running nodes right and then there is a separate cost I think these are both things Sam just alluded to I'm just separating them out so it's kind of obvious and Strikes strikes it on our face is there's the operational cost of actually downloading running maintaining the node operations and then the second part is there is the capital cost because you know I have 100 million dollars securing my network these guys need to earn like a 10 APR so that means I have to be paying them out like 10 million dollars a year either in inflation or actual fees eventually right like we can't all live on inflation forever and uh that's clearly doesn't work so we need to have enough fees to offset it and if you look at the two different uh terms today the operational cost versus the capital cost the capital cost is by far dominant and I'll take an example of an extremely chain like Solana to like illustrate this a cost of running one Solana node is maybe something like ten thousand dollars annually just ballpark order of magnitude and there are maybe a thousand notes right and essentially what we're seeing is ten thousand dollars per year times like thousand nodes it's like 10 million a year that's the cost of running the operational cost of running the Solana Network and you compare that to the capital cost of staking you know let's say Selana is like 10 billion dollar and 10 billion dollars State at a five percent seven percent ten percent APR we're talking about one billion dollar in payout to the stickers so essentially there is a huge differential between the operational cost and capital cost and what eigenlayer does is you amortize the capital cost across many many operations so these two can become more at par right so I have the system and I run like hundreds of chains hundreds of different things so then like the two things can become kind of comparable over time so by leveraging the capital across a bunch of different operations you know that's the kind of core principle of eigen layer of restaking in across multiple services so absolutely agree that the cost of security which is dominated by Capital cost because we staking is using it across multiple Services reduces the capital cost per service in fact I would say this is the precise Dynamic that has led us to a Smart contract World Imagine pre-etherium every new service is its own blockchain so it has its own cost of security every new application is its own work and what we saw is once you had a common system on which many applications can be built the cost of security of a Dap is actually very little you just pay when you need it right you are renting security you are renting a shared security system but the flexibility of programming adapt was only at the level of hey I have to write an evm application which fits in within this block space once you've removed that and have a chance security substrate that can be used in very very flexible ways you know that's really what eigen layer this is basically the Natural Evolution of Shad security from ethereum being like a common smart contract platform to ethereum being like a shot security substrate for any distributed system that can be built and just like one of the positive feedback loops driving applications was if you have other applications then the applications can compose with each other on top of this common trust network of ethereum similarly when you have distributed systems built on top of like the common cross Network on eigenlayer they can all start composing with each other with very limited new trust assumptions and so that's that's what we're hoping will will happen I was wondering about the impact that this could have on what is already established in terms of inflationary rewards to existing node operators what if like you introduced a new party for for security the the new network that you can subscribe to how would basically this impact let's say Legacy node operators how would this have an impact on the protocols to economics I think that's uh definitely something that could be interesting to observe yeah absolutely we're very excited to see how the the free market has ways of finding out things that we cannot anticipate and you know we're absolutely excited about that but one idea that we've been tying around with is of course you know when you're paying each takers for node security you don't have to necessarily pay need you may be paying in your own token or in some other way and it's a free market so stickers will opt in if it makes sense to them and and this is going to lead to like a whole interesting dynamic of how this how the tokenomics and the structures of these systems Yeah you mentioned briefly about the Mev management and the current sort of like system and architecture of Med boost plus plus so from what I understand is that blog proposers at the current moment have to make the entire full block and I was just wondering could you explain why there is this limitation what kind of problems that can occur from this and what like how does eager layer come into this and solve this limitation yeah thanks Dylan this is a really interesting Direction so essentially you can think of on ethereum there is only one major slash encantation which is double signing ahead if I sign a header and I sign a competing header I'll be slash right like that is the kind of dominant primary slashing condition of ethereum and if you think about this the way Mev boost was architected which is so what is the purpose of Mev boost I am a block Builder who took a bunch of transactions and arranged them into like a configuration and I want to make sure that the block proposal will include this configuration this entire bundle together because you know I have arranged this bundle in a certain way and in order to maximize Mev or profits for some users or whatever the thing that I'm doing and the important guarantee that the Mev Market has to give is that the builders transactions when somebody says that they're going to include it they are included as such not broken and reorganized to the favor of the proposals block proposal so that is a very important thing and the way flashbot system Mev boost deals with it today is by saying that hey I'm gonna send you like there's this node called really the relay will take all these bids from different Builders and it'll say hey here is the here's the bid and this is the header of of this particular block that this Builder is making you sign on it okay why are you making the node sign on it before seeing the block is you know if I sign the header and if I try to revert my position and say that I'm going to sign something else then I'll get slashed by ethereum so essentially the entire flashbots architecture is using the one enforcement mechanism that is there on ethereum which is you will get slashed for double signing a header and then designing an Mev system which provides some kind of guarantee on both sides which is hey I'm going to take this block and send it to you and then you basically then sign off on it right like sign off on the header and once you sign up on the header you don't have any ability to change or manipulate the block because then you'll get slashed on eigen layer what you could do is you could say hey when it relay sends you something you know you could say that I'm going to include this in the first 80 of the block or 90 of the block and then in the remaining one I'm going to do whatever and because the slashing condition is not the one slashing condition written by ethereum for its own protocol development now you can write a slashing condition specific to the Mev Market as long as takers opt into it you can say hey if you opted into this slash information then you could include the Builder's block as the prefix on the first portion of your block and then at the end of the block you can fill in whatever transactions you want and this could be to maximize your Mev to express your particular preferences whatever it is that you want to do as a Blog proposal and this is really powerful incentive mechanism because you know now you can add additional negative incentives particular to the Mev Market rather than piggybacking on the one slashing condition that ethereum design for a different purpose and so that's really why I think uh eigenlayer opens up the possibilities for Mev management so actually quite a large design Spectrum in fact we have seen pretty much every Mev management solution out there can be fitted retroactively into ethereum using eigen layer because for example you can say I'm participating in a threshold encryption market and and say that hey there is this group which is going to send me this encrypted transaction I sign off on this encrypted bundle then basically I'm forced to include the decrypted bundle otherwise I'll get slashed so you can opt into like very nuanced specific conditions so so in this sense agonair is allowing developers or validators to sell portions of blocks in the Meb market right and so this is also increasing the decentralization of in the mov Marketplace is that that I understand the current yeah absolutely and one of the ways to increase the decentralization of the air maybe Marketplace is to have firstly multiple different kinds of Meb mechanisms all operating together because I could say I'm selling some portion of the Block in an Mev auction and some other portion of the Block in a in a threshold encryption and some other portion of a block in event-driven actions like selling using it for liquidation and so on so you can kind of pre-sell portions of block space so that's one like superpower the other superpower is you don't have to sell all block space to one Builder one of the things like we are worried about in the Mev supply chain is if there is one Builder or one relay or like some dominant thing then they can dominate the censorship Market because you know there's just one and they could say I'm not including your transaction I don't like you or you're not paying me the bribe or you're not paying me my extortion fee whatever it is and so we don't want that and for example you can create a decentralized building by saying that I'm actually mixing bundles from multiple different nodes and then I'm include because first 20 of blog space I sold to Sam and the next 20 percent I sell to Kalin and now I have a decentralization of block Building inside a given block this is another very interesting thing actually vitalik gave a talk on this at the science of blockchain conference at Stanford and he mentioned like how you could use things like eigen layer for uh for this purpose now I would also like to get a better understanding of how to get involved with egg and layer as a 3D developer or as a developer in general I mean I say solidity because right now we're focused on solidity at node Guardians and we wanted to know if there are resources available or if we can actually start building right now I mean there are tons of alt l1's and they need their egg in their contract on ethereum who's going to build it right now like who has started the work how to get involved yeah so we will have a lot more public facing information on basically the contract structure of eigen layer on the apis to build you know new services on I can layer all of this will be forthcoming in the coming month or so and we're starting a discourse place where people can debate and discuss some of these ideas and structures both for how to improve eigen layer as well as like what can be built on top of Wagoner including you know how do how do you design slashing conditions for a new kind of chain or a new kind of service or a new kind of Mev management or data or all of these things and this will be out in the coming couple of weeks so essentially we are starting to gear up our more public presence essentially to engage with the community much closer and we view this as a Community Driven project where our driving thesis for eigenlayer is open Innovation we want everybody to come and build on top of this platform form it is not a platform where we're saying hey we are building and we are the final Authority or anything like that so that's that's where we want to be but it's going to take some time for us to get all the documents and stuff in shape for for all of us to work together on so whenever this middlewares and contracts are being deployed there isn't let's say an egg in their protocol that is collecting fee for leveraging the stack that you are offering it is open or you're going for a design where there is a fee that is taken to ensure the viability of how speakers understand more how this is going to be articulated and if there is a need maybe for a token unless you're going to talk about it I was just curious to know how this would be built yeah yeah so we are still exploring the different kind of economic models but the base layer of the economic model which is it's kind of like uniswap right why is it kind of like unit swap on uniswap's two-sided Market one side is liquidity providers another side is Traders you is the same thing and Traders are kind of consuming the service of the liquid liquidity providers right so that's the unit swap market and the same thing is happening in eigen layer on the one side is these takers right like they are supplying Security on the other side is Services which are consuming security and the natural model is like potentially you know they are paying a fee and the system the eigen layer system could take a kind of like a fraction on the fee so that's the general kind of Direction but precisely how we will design it you know even uniswap hasn't figured it out today so it's any any other point or message you'd like to to share with the community or key element on the road map or thing to look forward to yeah as I was mentioning we're going to start opening up our public presence so those of you who are interested please follow the eigenlayer channel on Twitter at e-i-g-e-n-l-a-y-e-r we have a variety of different openings to potentially participate and join an eigen layer but particularly on on the engineering side developers who are building interested and building distributed systems have worked with node softwares things like that on the developer relationships we have we're looking for people so broadly anybody who's interested in kind of like turbo charging open Innovation and letting other people build new crazy things uh athenia please ping us on the angular Channel I was interested in more yields and then I realized that that's going to be a nightmare for a devops to subscribe to 40 different let's say altel one security schemes but it's definitely something that I really look forward to see actually major that's definitely very exciting to see yeah it is going to be quite interesting to figure out how the experience of The Operators at stakers are going to be as you run 40 different Services what are the first profiles what are the yield profiles what are the operational costs there's a lot of things to figure out so people who are like intrigued by these things please reach out to me if you want to come join us because it's amazing I already imagine some ways with kubernetes and Cloud to basically in one click bootstrap 100 instances of different let's say services at once and creating new standards for devops infrastructure for distributed systems and all sorts of clients exactly and one of the things I mean the main points we find in node operations is the software life cycle right like how do you to upgrades and if there was a place where this was all bundled together like Sam saying you know and you just download the latest one every week and that's just lets you be compliant on all the 100 softwares that could be something quite interesting yeah that's uh some good food for thought I mean it's been a great pleasure to have you on the show as usual I leave all the relevant links in the description and if you have any questions for sriram uh test the comment section right below it's been very cool I'm sure we'll probably talk in a future podcast to see how all of this evolves and yeah I'll see you very soon thank you everyone and if you like that podcast feel free to drop a comment or subscribe and have a wonderful week thank you Sam thank you Kelly and it's my great pleasure to come talk to you all looking forward to be uh be back in the future [Music]