Staking, DVT and node operators: who will run Ethereum tomorrow? | Panel | ETHDam 2024

[Music] for our next panel we have uh Emily from chain security who will be moderating and we' got got this mic for you hello can I call my panelists on stage please we got it right all right what time is it okay let's wait one hi everybody so we're still waiting for one panelist but I guess we can uh we can start I can start introducing myself maybe because we don't need him for that um I'm Emily from chain security we are smart contract Auditors we are the bug people if you've seen bugs everywhere feel free to get them um you know the word play that we find bugs and we give them to you um so yeah I am very happy to have here this staking panel I think it's uh pretty much what everybody's talking about at the moment with aen layer and reaking uh so we'll try to tackle this the three topics we're going to discuss today uh reaking in igen layer we're going to spend half of the time talking about this then 5 to 10 minutes talking about DVT distributed validator technology and then talking more in general about decentralization and solo stakers and if we we have a little time we would talk about rainbow staking but if you know if we're very fast all right um so I guess let's not wait for Paul um maybe can we start with some intros can you introduce yourself super quickly your name your role what is your project and who's your main client and what value do you offer for them you want to start you seem very in the sparting blocks that on can you hear me yeah okay my name is Ben I'm a senior protocol specialist at figment covering the ethereum ecosystem um so folks not as aware figment we're um an Institutional staking service provider we're on something like three dozen networks historically our Focus has been on institutional token holders um but we're broadening a little bit now for instance we have the Fig app so you know kind of DIY stakers can use that and effectively get access to our back end can you hear me uh my my name is Glenn I head a product at a company called Kiln um we're a a staking service provider uh we also do defi yield aggregation we we focus on many different segments so institutional um we have a a retail offering uh and and we also have a white label solution that um a lot of custodians and uh our wallets are also plugged into this is yeah it's working ER Raul I'm co-founder of Diva sein and also co-founder of theat network ER Divas is a li protocol based on DBT and the nectar network is a re layer uh actually a multi-layer re layer aiming to provide an alternative solution in the world of V thank you thank you okay so let's start with a quick overview what is reaking and why does it matter in just a few words just to kind of you know stretch the surface V yeah so um I think in the simplest terms reaking is about securing multiple protocols with the same stake this isn't a new idea so you look at some networks uh like polka dot like Cosmos you put things like shared security replicated security that's ultimately what we're talking about why this is so exciting now of course the launch of Igan lay at least the stage the second stage of igen layer the launch of Igan D and the ability to delegate to an operator but effectively it's securing multiple protocols with the same stake that's the idea of reaking thank you Glenn or rul you want to add something or maybe say it in in different words so that we really get it get the picture I mean um from my point of view I think it's the same like Ben said at the end of the day reaking is no more than security of the layers with the in this case ethereum or translating ethereum security to other layers and to other services so they can benef fit out of the of that network of operators of network of validators that we have in the network and also the security Economic Security that has been provided throughout the EO Network itself yeah and everything these guys said but I think also the inovation that it enables uh and the speed of innovation so the the new um projects that can build on top of this quite quickly without having to um create their own kind of source of uh you know their that that backbone so they can very easily build on top of that and and we're seeing that already all right so um the different actors in the EG and layer EOS system we have the resters we have the LRT protocols the avss actively validated services and The Operators can you tell me what are the goals and the risks of each of these actors and how they interact with each other maybe we can start with the rakers what do the raker want and what do they risk more money more money that is you know the usual no I mean put it on nicer ter terms I will say that the real at the end of the day is people who is willing to take a an additional risk ER delegating their e to other actors within the network so they are willing to delegate their e to operators so they can perform additional duties and uh yes obviously in a chance of for higher world and what do they risk more because uh actually what they are doing is uh well they are risking more because actually they are facing other dangers at the end of the day what the the difference with resen and what you are getting with resen is that you are enable The Operators to perform additional services to those abss so that increases the chances for that operator to be slashed and actually that increases the the chances for your e to being get slashed okay so they don't only risk getting slashed on ethereum but also through the different avss so increase slashing risk yeah but uh it gives you also exponential returns because each one of the abss gives you rewards okay thank you um something to add on resters or can we move to the LRT protocols no I mean everything everything R said but you know I think the the rakers are just widening their it's not really an attx surface but they're they're they're opening that attack surface to to you know to their State uh cryptocurrency so it's just just you know adding that additional risk and I guess just just kind of one Nuance to add um obviously from a Staker perspective perspective um you know as mentioned already rewards are very interesting in terms of you know why why stakers rakers would be interested in reaking but I think there's a slight Nuance here and so folks that have been following ethereum for a little while you've heard of things like me burn minimum viable issuance this idea that you know the ethereum core Community ethereum core Community being ethereum foundation and core devs can reduce the reward rate for validators I think that in addition to providing more rewards um reaking has the potential to give Stak more agency back by opting into reaking you can choose the operator and you can have more of an more of a more control over your risk reward profile so I think that's really important okay so um we're going to talk more about the operators as well and I think Kiln is an example of ion layer operator um the LRT protocols what do they want what do they risk and can you give some examples of LRT protocol yeah I can I can jump in so um there are a handful of lrts you may have heard of so ethery Renzo Rio and I think from a high level um the concept of lrts are relatively straightforward think about an LSD liquid staking protocol or liquid staking token the concept with an LRT is very similar you deposit your eth or your LST on the LRT protocol and you get back a liquid raking token but unlike an LST um the LRT protocol is not just staking your eth they reaking okay so they're going to be staking on igen layer and then they'll be delegating to an operator so I think on the surface what's happening looks very similar to an LST but one of the key pieces underneath um is managing which operators they're delegating to and then understanding which avss they're involved with because as already mentioned every additional AVS you're securing or every time you hear AVS or actively validated service just think protocol that's built on Igan layer um but every time as a rester every time that you're securing an additional ABS you're adding incremental risk so so these lrts will have to manage that risk and understand it to just know all the avss that they're they've opted into through delegating to operators okay yeah I just wanted to add that I mean to what the Ben has explained perfectly and just to speak a little bit about my book that one of the main difference as you were comparing them with the lscs and you were talking about the increase risk that the EVS Expos to the I mean to the actually the whole ecosystem one of the main difference between an LRT and an LSC is that an LSC you are fully dependent on the ethereum network actually you you're building on top of the etherum network with an allergy at least if with the ones that are doing native ones they are building on top of hien layer of on top of carak or on top of us or any other of the solutions that are out there so actually the the dependency and the risk is changes dramatically in a sense because you are not building on top of ethereum itself you are building on top of something else and yeah I lay is at at its infancy right so it's not yet battle tested so this is also an additional risk um so we talked a lot about avss already um Can Can we Define these a bit better and again same thing what do they want what do they risk so ABS actively validated services yeah I think I'll just jump in real quick and then if anyone wants to add so every as I mentioned already when you hear actively validated service just think protocol built on Igan layer um unfortunately I love to Define that even more but the the thing is with the IG layer I think there's or AVS is built on IG layer I should say there's something like 60 today give or take I don't know maybe a dozen or something like that at various stages of development and they run the gamut they can be simple app chains on one side all the way to full-blown protocols on the other like you could build ethereum on IG if you really wanted to um and additionally these avss are not just limited to the ethereum ecosystem most of them have focused on ethereum related services to l2s and things like that but you know you can have things outside of ethereum I mean at the end of the day the avss are just looking to use some of the security the so-called trust layer of ethereum to secure their protocols that's really the only limitation um there are some other ones as well in terms of what can be used as that security so whether it's eth or lsts or ERC 20s but aside from that it's pretty open design space for avss thank you um you wanted to add something no no no no I mean is perfectly clear I mean to me I mean the best definition that I found about an ABS is anything that want to boost TR a network so literally sounds good all right so moving to operators and maybe this time Glenn you can take it uh what what is an operator what do they want what do they risk yeah so I guess the operator um in in the igen layer sort of world with abss is very similar to to an operator on ethereum who's who's running a valid dat so so we're running uh the software to validate the transactions or the services for for each of those abs um protocols or or services that we've opted into um so they they provide the they provide the the software and obviously the the operator is running that um you know is is running the validation service on on on those different things so okay thank thank you so I think that was kind of all the important actors right in the in the reaking landscape did we cover that yes I think so yeah perfect perfect so now um what do you think about the current hype around igen layer did they finally launch and how do you think that's going to impact the market um if I have to jump in I will say that uh actually Eng ler what I mean they've done great and what they have done is toose a new market vertical that they we weren't aware of at the end of the day I mean I don't want to dig too deep into that and well first I think we have to know that Egan ler has not deployed yet at least not fully I mean they have deployed something but h no no I mean yes I mean it's it's fly functional is valid I mean it's awesome what what they are doing but the one of the core functionalities of Rin is slashing and they haven't implemented yet for example um but I mean they've done great and it's something that at of people this Wen really they didn't believe in like H even six months ago and they have realized that once you understand the economics of Reen actually it makes sense and it can be extremely beneficial for ethereum Network I mean if you if we can leverage the risk with the benefits that we can get out of it I think it's one of the best opportunities that we can get in order to escale thether network yeah and just just to add to that I mean obviously they've I can so so reaking as a concept I think um is is here to stay and it's going to become a probably a foundational part of um the the ecosystem uh IG L won't be the only one the only game in town and we we're already seeing that and I think there's a real risk to ig layer that the other uh projects that are coming through can see what maybe where they haven't done so well and and do a better job in that area so around user experience the way things are implemented the way they execute um it's not to say that IG lay aren't going to aren't aren't uh aren't good because I think they are um but it's clearly still in development and making great progress I think just just the one thing I'd add um specifically about aan layer um and you know it's it's it's a good question have they launched because you know as we all pointed out yes lots of the pieces have launched lots of pieces of I layer have launched um but there's a lot that still has to be sort as as as ra mentioned you know slashing is a big one you know I think about this I kind of it makes me think about the launch of Beacon chain um so you know it seems like it's kind of following that minimum viable product and there are these stages you know when Beacon chain launched um you had the two two networks operating in parallel and effectively what's happening on Beacon chain is it was coming to consensus about its own State there were no transaction payloads in other words it wasn't doing the heavy lifting of main net which is still proof of work then you had the merge so that was proof of stake taking over and then about this time last year you had chaella so that validators could finally get their 32 eth back so it's it kind of ethereum kind of Beacon chain I should say kind of evolved in this manner it seems like ig's kind of taking a page out of that book where it's like you know it'll there are these stages to get to the final product and that kind of seems like where we're at with igen lir and with with that lack of information on rewards and and penalties and things like that there's there's additional risk when things are opaque there's of course more risk one last thing that sorry about that one last thing that I I would like to add is as you were mentioning like the competitors that are coming behind a leer are going to learn about from them and trying to implement things better I just wanted to point out because at the end of the day we are building a sort of competitor to I leer but at least in our case we are not framing it in in in the sense of competition itself so okay we are building an alternative to leer because we think actually that the ecosystem needs Alternatives and not everything needs to be concentrated within one single agent or one one single uh protocol in this case uh but at least in our case personally I never frame it in in the sense of competition that I'm going to stract value out of layer to take advantage out of them everything that we are building and I think most of the teams out there is open source is for the community and U yeah it's a although it's easier to speak in terms of competition I guess yeah I think I chose my words wrong I yeah alternative is is what I should have said not not competitive do you think we would at some point get into discussion the same way that we have now for liquid staking providers where some are bothered by the dominance of some of some actors do you think we're going to get to the same thing with igen lay and the the Alternatives is it a risk to have one main reaking provider um obviously I mean I'm taking this one I mean that is one of the reasons why we have built that I mean it's not that we are going to get to that phase under which we are going to have a one the dominant within the market we are already there I mean the until two weeks ago there was just aen leer and also when you were talking about H Ren you have spoken about the aen Leer ecosystem because it's the only thing that is out there so that implies a monopoly somehow and I think is what we should try to avoid because actually they have become really really big I mean and they are doing great but uh when you say the product that you launched you're talking about nectar yeah yeah okay um all right so we are uh already past our time to talk about reaking but there's one thing I'd like to discuss still uh it's about the risks of reaking and specifically people are talking about this Cascade risk um so I'd love to hear kind of a concrete example of what this Cascade risk could look like um a catastrophic failure type of scenario can you describe step by step how that could happen me in bathroom [Laughter] maybe no I mean maybe an operator can describe it better than myself ER it will be again as of now with the grand implementation of fan layer H what they have launched to main it it wouldn't impact that much because they don't have F slash in implemented one it is no no I mean like imagine in two years you know everything is implemented everything is deployed um what could happen maybe Glenn we haven't heard you too much maybe is this something you'd like to to walk us through or or Ben yeah I mean I I so to be frank I'm not thinking about the Cascade risk to be frank I I don't see it as overly likely but so the scenario would be something like if there were a mass slashing event so you're already seing a lot of interest in igir so I think there's something like 4.5 million eth that's been engaged on igir whe whether through the LST Road or through through the direct path um in other words creating an igen pod and then activating a validator um so that the idea is you know if if Ian gets so big um and there's this Mass slashing event what does that do to the security of ethereum itself and so this is this is one of the criticisms you see when you see the debates between sham con and founder of Igan Lair and you know Justin Drake or others from ethereum Foundation the idea that you know ethereum has this so-called social aay ethereum core Community I should say has this social aay where if there's a mass slashing event on ethereum the core Community can step in and try and unwind that if it's not an attack on the network and there there would be a reason to do that um and the concern from ethereum Foundation is that there is this Cascade risk potentially on igen layer and they they the ethereum core Community could be called on to step in to bail out risks that are taken on Igan so I think that's high level to Cascade risk um I think couple things to say there first is um there is this so-called slashing veto committee on Igan now again we don't know what slashing is going to look like on Igan yet which is a bit of a paino for folks but apparent there's there is the slashing slashing veto committee so IG layer itself has its own kind of social layer if you like um so they could step in and and and negate this this Mass slashing event if we're a bug or something like that not an attack on Igan layer um the other thing too there are researchers at at ethereum that are looking at ways to kind of solve this issue and one issue um which is kind of related is that there's an asymmetry about um eth stake so Igan lay uses an oracle process to see how much eth a validator has on Beacon chain but Beacon chain doesn't have the same V visibility to post exit balances on igen lay meaning if there was a mass slashing event Beacon chain wouldn't necessarily know about it and that's a problem so if this interest you just asn't aside check out uh barnab bay mono he's done a piece on this he's ethereum researcher um this asymmetry I think that is something that's important that's going to have to be worked out um in the medium to long term thank you some other risks of reaking we should touch upon before moving to DVT CU we don't have that much time left I I I think just one thing sorry I don't know Glen if you want to okay well just I think the obvious thing and I won't go into too much because I've already talked a little bit but um the obvious Point here is that um you know you're securing additional Protocols of the same stake you're necessarily stretching the stake so I think that's kind of the one of the top kind of headline risks there is Nuance there though um so that's another risk I guess just to be mindful of yeah and maybe maybe not a a huge risk but uh you know people chasing yields um and with the with the uh Alternatives or the competition between the different lrts may be offering uh unsustainably higher yield which you know could then obviously you know is not sustainable so um you know if they if they're cutting Corners if they're taking other risks maybe higher risks that are hidden um you know that could obviously affect directly affect users who are who are reaking um maybe I would add that I'm really curious to really understand how those obni asset Protocols are going to work uh because at the end of the day that is a pure risk management I mean the the moment that you are putting different assets you know that actually to validate in ethereum at the end of the day you have to convert it to ethereum one way or the other because way otherwise you are kind of backing that within a so I'm I'm really curious to see how that is going to be implemented and how I what are going to be the risk associated to that all right so um I heard that DVT distributed validator technology can help mitigate some risks of risk-taking um so in a few words what is DVT and why does it matter um DVT is distribution validation technology obviously and makes basically I mean what we want to call it or we want to explain it as a layer to for valid valtion at the end of the day through DBT what you are doing is in a peer-to-peer manner coordinating different people in order to confirm one validator so it's a decentralized network over the ethereum network in order to perform validation duties so I guess reducing the risk of each uh each of the participants exactly I mean in terms of risk yes obviously the the risk and the H one of the benefits or the major benefit that it brings is more resiliency is through using DVD actually even if onethird of the of the house upset of a operators or validators is down the clust still works so it's way more resilient and I guess just just kind of doubling down on what on What R said I think you know one of the early blog posts uh from one of the DVT protocols they use the analogy of of airplanes so you can have a single engine aircraft that's single validator when you think about DVT all you're doing is making that a multi-engine aircraft so if you're flying along in let's say a four engine commercial aircraft you lose one of the engines the plane doesn't go down that's the highle concept with DVT for folks that aren't as aware um the other thing to highlight there with that analogy is that it's active active redundancy meaning that you know some operators today will have um a failover meaning you run your primary validator stack and then if that goes down you spin up a backup um that's you know passive uh redundancy but active active means that that all the validators in that node let's say you've got five validators or five operators in a node in a DV cluster um they're all doing they're all operating at the same time just the way that the same way that the engines on a multi-engine aircraft are are behaving um so from a high level that's that's DVT and I think DVT is also interesting in the sense that uh you could have this uh these kind of big liquid staking providers for example Lio that implemented this DVT within Lio so even by uh keeping this this ecosystem you can decentralize it while staying in it right um so actually chain security audited this staking router for Lio so this is why I have that in mind um so how uh can DVT mitigate the risks of raking um I think is uh double down again on what Ben just pointed out you are eliminating single B failures uh and I mean within ecosystem is that redundancy that been was mentioning if you are implementing DVT within V actually what you are creating is a safe Network or a safe layer uh that can take very the heat of a maor cling event it won't go down a straight way to ium validator but if you have multiple sets of valid validator you are spreading the risk between the whole network itself so by reducing slashing risks it reduces the the risk of this massive slashing event right okay interesting all right something else on DVT all right so we have four minutes to talk about decentralization and rainbow staking seems perfectly fine um so Glenn you said that staking is a constantly evolving industry and that decentralization is a constant journey is not an end State um it would be interesting to see what are the key decentralization metrics that we that we should use and basically how do we know if we have sufficient Staker decentralization yes as I said the it's it's a journey it's not it's not an end State because every time we you know like it's it's everything's evolving so we see Lio had a you know was quite dominant and now that's come down a little bit but it's still very dominant um with reaking we see we see more centralization sort of appearing but I think the ethereum community and uh e like the the ecosystem in general is quite good at regulating itself um so when these things happen like the recent um uh the recent things about the uh consensus or execution clients that that they came up forc a lot of operators to to change their uh or to expedite their um their shift to um you know um a more balanced uh more balanced usage of of of those execution layer clients um so I think it's it's just it's just a constant thing that we're always looking out for these new Innovations like like reaking as I said is is you know brings a lot of people together but we still we we still see centralization happening within that but I think I think you know us as a community you know we can see where those where those kind of risks and that that concentration is and and it it tends to even out over time um so I think I think just a few things to add um you know when you think about decentralization I think I think it's a nuanced conversation um so you know you you think about you you know Glenn mentioned the the client diversity for instance um I think that's important but I think there's as always as I say Nuance so um you know one of the one of the issues with client diversity is that we don't actually have certain numbers about which clients folks are running so on the execution layer side you know we have um self-reporting and we get a decent idea from those self-reports who's running what um but there are of course errors information can be dated on the consensus layer side it's more or less the same there been some methods some fingerprinting methods that have been used to try and determine which clients are being used but it's the same problem so I think this is a problem that you know um hopefully developers will will sort out but basically find a non-s spoof way to um to um illustrate which client you're using or to admit which client you're using I believe nethermind is working on something here um but but yeah some certainty around the numbers would be really helpful so we know where we're at and where we should be moving to um the other thing is you know when you look at certain kind of centralization vectors um I think you have to ask yourself how are they centralizing things so you know um you know you can say well there's a lot of stake moving through this one entity um but if that entity is also um onboarding DVT for instance and making that a standard for the whole Market is it a good thing or a bad thing and so I think I think there's just my my point is just that there's a nuance and like on the DVT side for instance you could theoretically have a very concentrated Network where you know the majority is running on G and on the consensus I don't know whatever client let's say prism the majority of of eth is run there but with DVT if everyone were running through a validator cluster you could change effectively what the diversification looks like so my my point here is not that divers client divers diversification is not important but uh it's it's a very nuanced conversation and that's yeah I think yeah and obviously there's a lot more to it than just the the client diversification there's a whole bunch of other stuff which we all know about like where these nodes running are they all running in Amazon are they all running in Google Microsoft um you know there's a whole bunch of other stuff that comes into that so it's uh it's important all right we have six seconds super quickly what is Rainbow staking super quickly just because we mentioned it and we can't leave the audience without knowing first and foremost in in in like a few seconds here barnab mono check out his post on Rainbow staking that's where you're going to get the most information from a high level it's just recognizing that you can't force the market into the way you view the world today it's a it's the way that validators look on ethereum it's a relic of design decisions that everyone's going to look the same they're all going to run validators but rainbow staking is about segmenting those groups there are large token holders there are large operators there are Solo stakers and rainbow staking at the end of the day is about recognizing that and segregating each of those into different duties different penalties different rewards that's high level what rainbow staking is all about and you can see that with me boost and proposer Builder separation it's the same basic concept you're separating the roles and you're further segregating folks and it's just recognizing what the market is already telling you great for 50 seconds very very nice all right so that's all the time we have thank you so much awesome thank you no no no we got we got a couple questions if if you're if you're ready all right cool so our first question here is reaking hasn't really taken off in Cosmos and polka dot why will it take off on ethereum uh I I think the proof is in the pudding I mean I don't know if you say is it going to take sorry oh the point yeah everyone loves points right admittedly that's I'm I'm yeah not a huge fan of points but I think I think if you look at at the data I mean 4.5 million eth has already been engaged uh on Igan layer it it some of that's probably related to points I don't think all of it um certainly not on the direct side LST is potentially just in the sense that you can move more quickly than you can on the direct side um I don't know if you guys had yeah I agree with Benner and I think the real the real um uptake or the real adoption will will show when when we get closer to slashing being implemented to see who who runs away take their points cool uh compared to the ethereum network rewards uh what do you think would be the most lucrative reward Source in reaking I think don't all jump at once I think yeah I think it's too early to tell um but I think in the early days the the additional rewards that come through reaking are not going to be that high just because there's a they'll have a a a smaller amount of rewards to distribute to a larger number of people so it's going to be diluted at least in the early days so I think it's difficult to to predict cool and what would happen to your protocols if Igan layer has a critical bug um to my protocol in particular nothing yeah yes exactly all right uh should Igan layer kill the validators withdraw the eth if there is a slashing event or just keep the money in a big immutable treasury um I I don't I don't love the language kill the validators um Fair obviously um but I I will say that I think that you know aenir is in a in a unique position to encourage the behavior it wants to see on the network so we were just talking about client diversity um Glenn mentioned Geographic diversity things like that um so I do think with the tributal security model if Igan itself retained some of the eth that's taken as penalties they could also so potentially encourage certain behaviors like they could encourage home validators they could encourage client diversity things like that so I do think there's something there to keeping it and then you know not I wouldn't say donation but but payments made to encourage healthier things that you'd want to see on the network I think is possible and interesting all right great we'll take one more question real quick what hap what will happen to delegations if igen layer has problems do ations actually I mean that the seers are I mean if has problems is what I was speaking about actually as a delegator what you are doing is taking the risk of those delegates so obviously the market the whole Market of whoever has deposited or interact with will be facing that risk and will be eating the consequences I guess that is one of the good reasons why those people should diversify and use another things that are not taking later awesome well thank you so much everyone for your time and for your for your knowledge let's give them another round of applause [Music]