Future of Staking | ETHDam 2023

foreign cool so um yeah I'd like to remind you guys if you have any questions for these guys do go and find them afterwards they're not going to bite and you may get a more intimate answer um for our next panel there's actually going to be four panelists plus the moderator so I am going to ask the panelists to three of you are going to have to get comfortable on this couch um so yeah without further Ado I would like to welcome them to the stage so please give a round of applause [Applause] oh should we put all four yeah let's do it nice and cozy the couch Factor hahaha how to get rid of it so very quick this is like a this is like the staking Community we're very close in it good morning and thank you for being so comfortable with each other my name is Pascal I'm one of the people organizing the local developer Meetup you can find all about it at ethereum.nl but I'm now here to introduce you to this amazing panel we have from left to right I'll just introduce you with a first name nikhil Paul Pablo and easy um you know the drill you can ask questions on slider323232 but first i'm going to ask each of the panelists who are you what's the project you're working for and what do you have to offer and try to do that in one longish tweet perfect happy to kick it off thanks everybody for coming and listening to us I'm Nick Hill from figment we provide infrastructure for staking effectively token holders who don't want to run their own infrastructure but earn staking rewards we run infrastructure on the behalf hey there my name is Paul Lansky and I'm representing dabnode which is free open source software that you can use and install in your computer to run whatever web3 infrastructure you want without ever having to touch command line so we have this beautiful UI that you say hey I want to run a validator click click click you deploy everything in your own machine self-sovereignty is super important the centralization and then you can start staking and one of the things that we're doing is we're integrating with all these guys here so not only you can run a solo Staker but also you can participate and offer decentralized infrastructure for all of the LSD protocols out there thanks um hi everyone I'm Pablo Alba from Diva so there was a liquid sticking protocol what we are doing a bit different is that we see DVT as a kind of technological Revolution for staking and we're building a protocol that's both completely permissionless anybody can join can be an operator can use it but also completely trustless we've removed all the sources of upgradability of uh kind of attacks that you could have and we Consolidated all that into a single solution where you can either stake or operate a node um and yeah I've been involved with a lot of the research and the foundational decisions for for the protocol hey everyone I'm Izzy I'm a contributor to Lido as the master of validators Lido is basically a liquid staking protocol headquartered on ethereum but also active on polygon Solana Pokemon Sama although the latter are being deprecated and phased out right now um and Lido is all about providing uh democratized access to staking basically so making staking as approachable as possible as simple as possible accessible and composable so so we noticed that your your offerings are uh complementary maybe to make that even more clear could you describe your target user you can also say the same as a guy on on the right or the left um but it's probably good that you you paint a picture of your of your typical user and maybe let's start from right to left now sure um so I guess one of the core principles of Lido is that anybody can be a Target user uh when you think about it from a Staker perspective so one of the first remits of the protocol was to make staking approachable for people with less than 32 each and from the other perspective obviously when you're not talking about about stickers you're talking about node operators up until now uh Lido has been a permission protocol in that node operators are curated by the Dao with the Advent of Lido VT which introduced the staking router modules will be able to be added to Lido which will basically plug in validator sets and allow a wide swath of different node operators independent people solo people douse communities to band together and also participate in the network so thank you for their distinction so user is both the validate on the one side and then the user in your case of liquid state on the other side and one is more business and the other more is retail user it could even be the same right you might have somebody that is like staking but also wants to operate validators at the same time so there's definitely also an overlap or it might be other Dows or communities that want to band together to pull their funds to run validators together or to stake together and then be able to do things with those build funds Pablo if they're targeting everyone who specifically are you targeting with diva so we we see users as a sliding scale of involvement so later was an inspiration for us in the sense that we want everybody to be able to participate on staking what we do is we make it really seamless that you can be a Staker but then you can use that as a collateral to run an operator and we remove the barrier to entry so we don't start a 32 East but we're actually starting at half an eighth Bond we made this decision by thinking it should be the price of an iPhone like if you have a good iPhone or if you have a laptop you you should be able to run a node and today you have the technology but what we're doing is we're creating the economic model for that and that was not easy because today most of the DVT has been thought as oh we're just going to split the key but it doesn't answer questions like how do we attribute how do we attribute rewards to people doing a good job or how do we penalize the people who are doing a bad job and I think we cater both stakers on operators um but I think the Innovation for diva is that the economics are very deeply integrated so they make that very attractive cool as well I mean we also want to Target everyone what do you mean no I'm joking um ideally um we target people that want to run infrastructure that want to participate and take one more step into the centralizing ethereum knows his chain any proof of stake chain that exist now and that that might exist in the future we target people that consider decentralization something crucial something important and we target both people that are technical and especially people that are not technical so if you want to start participating getting rewards on uh ethereum staking or gnosis chain staking and you feel that your technical level might not be um maybe you've never installed Linux before maybe you've never used command line before maybe you're worried about upgrades and how am I gonna we take care of all of this so basically our Target user is a node operator that once um that wants soup to have something super easy and up until now we're targeting solo stickers but now we're going to start targeting any LSD and anyone who wants to be an LSD operator and we're lowering the barrier of Entry the economic barrier of Entry to staking by offering protocols like Diva like Lido V2 like Steakhouse like rocket pool protocols that are deeply integrated on how dabnode works you can just deploy them in one click and they require a lot less than 32 East to run a node let me drill down there so you try to make it as easy as possible for people to stake at home to run the raw machine what do you think works best do you want to try to convince them on ideological level or are you more able to convince them on the financial level I think we don't need to choose I think we can offer to both um it makes sense to run a note it makes sense to validate and stake on ethereum for yourself especially because with tools like that node you don't even need to dedicate time to maintaining this we we have auto updates we take care of all the devops um we we take care of all the testing new versions before they get pushed to the smart contract by the way the out updates are decentralized um everything is put in a repository and distributed by ipfs so we don't have a central server that's a single point of failure etc etc so basically um it makes sense for people that consider this Inquisition important it makes sense for ethereum as a protocol to have more solo stickers so definitely there's an ideological part of here but there's also a lot of money to be made people mind Bitcoin in order to make money and that's totally that's a totally legit way to go and stake ethereum because you want to make money and I think we can Target both and Nikki I would think that you're on the other side of the spectrum could you tell me a bit more about your type of customers that's that's part-on so it's it's a good contrast so we are on the other end of the spectrum where we're typically offering our services to institutional holders of tokens so folks that are typically holding large amounts of tokens um they may be running their own infrastructure and sort of use us in parallel so that you know again if you have a large holding you want it to be distributed um or it could be folks that have large Holdings um but don't want to run this infrastructure so that's a Target group um I would say it it varies a bit so it's it's directly folks or or token holders who are customers as well as folks who want to offer staking to their end customer so if you think of a custodian or an exchange we would be behind the scenes running their infrastructure to offers taking um so that that's our group I'd say it's um it's Global uh the things that are important to them are you know sort of having a large team we are shocked to compliant we're very focused on regulation so it's sort of the audience is people who are interested in those topics as well as having the ability to stake there's one more thing that I wanted to double click on with Nikki and an easy um it's surprising to me so there was recently this podcast with Laura shin and there's always this word custodial that we use so either you have control over your keys and then you have your ether or you don't have control of your keys and you don't have your ether and we talk about custodial when we send ether crypto to an exchange you guys are not custodial you guys cannot run away with your customers money can you explain how because some people actually think that you can run away with your customers money yeah it's a that's really a great point and the terminology we use is protocol staking right because taking can get banded together with other activities in D5 like lending yield farming so we're trying to push this terminology that this is protocols taking you control your keys you never give up custody of your assets it is you the token holder and the protocol that is interacting uh and we're just sort of uh facilitating that and the protocol is determining the staking rewards you are determining how much you stake and there's sort of no intermediary and and I think that's one of the reasons we are sort of very excited about it because it's if you think about what's happening from a regularity standpoint um and as well as some of the negative events that have happened in the last uh 12 months they've all been related to people taking custody of someone's assets and doing things that you know they were sort of not supposed to be doing so so in light of those we feel protocol staking is like a bright spot uh in the ecosystem yeah I think that's correct and maybe to add a bit of technical Nuance here which is perhaps something that contributes towards this confusion so when we talk about private Keys there's different forms of private Keys um I mean obviously like a lot of different forms especially on different chains but even in terms of like when you're thinking about staking on ethereum so when you stick on ethereum you have your ether which is controlled by the private Keys which control the address which is yours on the network but there are also the private keys of the validator these are two different things so when we're talking about non-custodial or custodial protocols we are talking about two different forms of custody the form of custody around the tokens themselves so The Ether that gets deposited into either a protocol or a centralized exchange or some other form of liquid sticking solution and the custody of the validator keys themselves the design of ethereum staking is such that custodying only the validator keys does not allow the node operator or whomever is running the validator or any somebody that controls a validated keys to run away with the money that has been deposited to that validator the only thing that allows them to do potentially and DVT is something that really kind of meters this risk is to get that those funds slashed so there's like the potential risk let's say of a hostage situation so liquid staking protocols and especially decentralized and uncustodial ones do a couple of things to meter this risk so the first is that the software itself um what runs on chain I.E the protocol middleware is non-custodial in the sense that when funds leave a user's address they get deposited into a smart contract that smart contract can either be immutable or if it's upgradable it's upgradable by something like a dow so we never say that if we are depositing our funds into like Ave and Ave is a Dao and there's a contract that's upgradable that is custody however if you deposit it first to a sex and in the case of decentralized exchange they actually take the money and then stake it that is a custody model there may also be custody models that are on chain in terms of there's a step an extra step on chain where the funds first need to get like hop before they end up on the beacon chain but liquid sticking protocols obviate that step and disintermediate it because it only happens based on code and the second thing with regards to validator Keys each different liquid sticking protocol has different model around how validator keys are custodied but the custody of those validator keys does not transfer into custody or the ability to rug users of their Eve um the guys in the middle so the guys next to you say we have everything is fine everything is under control how are you holier than them and I'm actually mostly mostly asking Paulo because in your case running your own machine that's pretty straightforward but how do you refine the trust assumptions that they have and I heard the word DVT it seems like they're stealing your thunder yeah well they've accepted so we needed to make something better than that so we we started looking at all the trust assumptions were there and we basically work really hard to remove as many of them as we could so for example the the risk of if you own a validator key you can misbehave you can disappear with today's specs it could be that you stop validating and your eth is locked for up to 10 years and only then you get it to the withdrawal address and only half of that so that is in my opinion a validator custodial key risk that we're trying to solve and what we do is we we try to go back to the principles of how a blockchain should be it should be done in a blockchain if somebody is sleeping nothing happens everything should continue right so what we did is we kind of thought what if we redesigned validation as a layer two and well principles what we have should we allow anybody to Caster your key well while we did is there's no key it doesn't exist it's always a multi-party computation thing if you use something like fireblogs um for example the key never accessed in one case is distributed um is generated in a distributed way with dkg so that's a really strong Advantage because it means now you can reach consensus on what should be signed and what we do is we use a two out of three consensus which means that only 11 out of 16 nodes need to be online and by model this is like an office where people can be constantly sick and as long as no more than 30 percent of people are sick any given day the office keeps working and I'm sure you've been to offices like this but what we did the numbers and it's pretty incredible because if you only have 95 uptime which is terrible the worst sets have a 98 in ethereum DVT with this params turns 95 uptime into 99.99 uptime and the other key of so clearly DVT can can allow homesteakers to do a much better job actually than professional validators just because of this reliability aspect and the second aspect is we try to avoid immutability so if you're a router even if for a split second your smart contract needs to Route something now if you have a governance takeover and the Dao which could be you by the way um says now we're doing this and we're going to rock all users you could technically do that so we've gone for more immutable Solutions which I think we we're all working towards right it's just safer and more I can do what blockchain stands for uh Paula had a follow-up question for you so you guys you built this software stack and you're kind of a king maker because you can present the user with the choices that you deem the most holy or almost how do you select the software that you incorporate in that node um we don't uh the dab node smart contracts are publicly accessible and anybody can use the SDK to publish their packages and as I said we have this sort of adapt store right we have this app store and you can make a package of whatever protocol whatever nodes whatever thing you want the users to run the rocketpool smart stack the execution layer the consensus layer um you can if you can dockerize it with our SDK you can turn it into adapt node package and then you can publish this into the smart contract and then every dab node user can go into the DAP store and we'll see that package now dangerous danger anybody can make a malicious package and can introduce and inject code that will take Keys away and we have seen because they have explained very properly um that the only thing that you can do with validator keys which is the only thing that is in the dab node machine is you can cause slashing you can cause havoc you can do a griefing attack um but you can't really steal so there's really not much point on it but still we need to sort of gatekeep whatever comes into the official map so we have a public dabster where anybody can publish things and then we have like um a nicer way of presenting this and the only thing that you need to do is to contact us make your code open source and then you're in the urine domain in the main App Store now um you've you've called you've mentioned King maker and I want to say that for us because of who we are and from the beginning one of our design principles is optionality we are actively trying to not be king makers to always offer the users the possibility of using whatever software whatever protocol they want and of course there's an order if we need to integrate it we it might take longer time to to put one on the adapt store than another but uh it's open source you have an SDK how about you create the package if you want your software to be there faster that's the advantage of the an open ecosystem that's open source thank you I saw two questions coming from the audience uh we still have some time I first want to throw something uh to you there's there's been a lot of discussions taking the past few months issues around Mev issues around security of protocols protocols being attacked issues around centralization what keeps you awake at night so what is your staking dystopia so what can it terribly go wrong and let me go from left sorry yep you're left to right yeah from our perspective what concerns us is regulatory overreach if you think about what happened with uh Kraken in sec that if you if you look under peel back the onion it was actually logical the the action that they took but they're sort of at least in the US this element of you know going after everything crypto related so our fear is you know some sort of regulatory action that that just pushes taking offshore from from the US which I would say net net is not good for the the ecosystem right that's a pretty big Market a lot of interest so to be honest especially in light of the the recent protocol upgrade chappella and uh prior from a protocol and technological perspective we actually feel very good like I think the community is really like delivered and established itself is silly the The Regulators now we feel like need to understand what's happening and come on board and not not sort of throw roadblocks could just why are you different from Kraken so if you think about Kraken um and anybody's if you've staked on Kraken you are depositing your your tokens onto Kraken right so similar to Izzy's example you're giving up custody of your asset that's one second you're not interacting directly with the protocol it sort of crack in is the intermediary so for example if the reward rate on a protocol is 10 um you know Kraken may be giving you eight percent and keeping two percent um so your your interaction is is thrown intermediary versus sort of I think all of us on this panel what we're offering is is direct interaction with the protocol um and that's what the Regulators took issue with that one it's custodial you're giving up custody to sort of you know it's effectively a security if Kraken is determining what reward the users uh receiving they're effectively offering their security there's actually there was actually a question from the audience so what are the risks for stakers uh regulatory risks it's unclear to me if you mean staking provider or individual user but let me throw that to you guys so Nikki you're pretty optimistic with the way you organize now you're clear let's suppose we're not in a clear yet how can some of you guys make sure that we continue the ethereum blockchain we can keep the ethereum blockchain running yeah I think the best way of doing this is making sure that we don't put a Target on our foreheads and meaning if if they do come after big companies if they do come after figment after staking providers after those those have those are companies that have a name and they can therefore you can enforce action over them if you have solo stickers or people running on their own homes in different countries different jurisdictions in different places it's it becomes really really hard to follow these people and force them to shut down their machine because you don't know who they are you don't know whether I have a validator or not you don't know if my validator is in the Philippines or in Spain or in turkey or in Costa Rica and it becomes really really hard and so in terms for so for ethereum the best thing is to have solo stickers to have people running the centralized Hardware in their own homes but it's not only good for ethereum it's also good for protocols as well it's also good for for Lido it's also good for Diva it's also good for all of these protocols who can have these node operators being equally distributed and maybe easy I'm sure you'll agree Pablo or maybe something more spicy to add um no okay uh easy um is this an issue that comes up within the line of Dao I mean do you guys um aim to be geographically distributed sure yeah that's that's a really good question so when we think about like what makes a good protocol and what makes a good validator set um we don't and when I say we I don't like necessarily mean contributors to Lido I think we mean ethereum as a whole we're not only looking at things like performance and rewards we mean robustness of the network right that it can withstand liveness issues uh non-finality uh you know people like regulatory capture in certain cases so the only way to really have a robust network is to have it be decentralized decentralization means a lot of different things it means having a lot of participants independent participants it means them being well distributed geographically like you're saying jurisdictionally in terms of infrastructure so not everybody running the same setup or the same client um and also in terms of not using the same type of infrastructure providers so not everybody running on the public Cloud not everybody running in the same data center in Germany when they're if there's like a thunderstorm like it goes out um and probably not everyone running even like the same chipset architecture on their um homesteaker PCS so Lido has like the Dow has enshrined a couple of principles with regards to this obviously one is the constant increase of the validator set and V2 and the staking router is like the evolution of that to include as many independent node operators as possible whether they're like solo stickers or larger groups and the second aspect of that is for the permission set like what is actually done and what kind of priorities are placed so um there is reporting that's done every quarter there reports on like how are node operators distributed geographically how much stake is allocated to every node operator because for example in a permissionless node operator site you can't guarantee that there isn't a very top heavy allocation of stake in a curated set what you can do is create sticking algorithms that distribute stake in a more Equitable manner which means that you have more robustness in the case of something like regulatory capture you see that node operators are constantly improving in terms of client diversity running all the different clients lighter was one of the first liquid sticking protocols to no longer have a prism super majority and is actually one of the protocols through its node operators right that is contributing the most to certain minority both from the execution layer and the client later uh like statistics so there's chainsafe which is the creator of lodestar which is running currently around 6 000 validators using lodestar an ethereum client written in typescript which is like an incredible feat just by itself and that's something that has been enabled through their participation in the Lido ecosystem yes we continue the nightmare scene so nikia's nightmare was regulation what's your nightmare what's your dystopia um ah good question yeah so I think what is he saying is diversification is super important right because the moment you diversify you're making sure that there's never a majority risk of like us or present and that's super important right but I think when we only think about that we're still taking a damage and we're just saying we're going to reduce that damage the way our portfolio manager reduces bad investment risk right like oh I can lose this company because it's only three percent of my portfolio but I think the unique advantage of what we're doing with DVT is that because we have a consensus now this consensus can be made of diverse people and because just in the same way that I showed how DVT could improve uptime from 95 with the same operator set to 99.99 what we allow is because these consensus needs to agree that means that we can probably improve like reduce the risk on attacks like this by 500 to a thousand times and again we we're always saying validator key doesn't mean you can steal the keys but you are the property management manager and you can run a lot of things and maybe but maybe you get mandated like imagine or staking becomes illegal the way heroin is illegal well what do you do if you seize a validator you have to burn it um I think that's my nightmare scenario hopefully we never get there but we're designing for that dark future in case we ever have to maybe just add a little bit to what Pablo was saying so like my nightmare scenario here is that that the technical environment doesn't um do the things that it needs to do to allow liquid staking like protocols to ossify as fast as possible in order to prevent this this nightmare scenario so DVT um and creating like a validation layer in terms of a layer two on top of the protocol is one way to solve this and we should have more than one way so another way is triggerable access right which is something that we saw a recent EIP about um my fear is that like this stuff doesn't come fast enough so that we can say that liquid sticking Protocols are feature complete once they've added this functionality and then they can start ossifying and removing Dao levers or Dow upgrade ability or or what or anything like that and the other like really really big like nightmare scenario is that there's a bug with a to Beacon chain deposit contract because that's that's like I don't know how you come back from that reflected um we're almost done let me throw in one more last audience question will eigenlayer at to the risk I mean there's already risk is eigen layer useful or is unnecessary risk just anyone who has an opinion on this um I think we're in a bit of a same situation as with MVP boost it is something that we at a consensus level there's no way to to do PBS yet therefore we had mbb boost that came out from an external group to cover a need that makes arguably it it makes um something that I I think MAV is a bug not a feature and we should be working towards eliminating but it is impossible as of now therefore a emu boost was made to to supply or to provide for this um for this sort of problem and and I think it's overall very positive because it allows solo stickers to access Mev Revenue that otherwise nobody would trust trust them with um because they would they could just run away with the MVD opportunities themselves right so overall effect very positive but it is not a consensus it is not something that the ethereum community has decided that this is how ethereum wants to go and it did have a bug a few months ago or a few weeks ago and it screwed up the network for a while eigenlayer we are in the same risk is something that's outside of the protocol it's something that we haven't decided it's something that we're just introducing into our kitchens and it might break and there is a big risk but is it necessary is it positive yes just like Emily boost it's doing something positive that people want but it is a huge risk I'm afraid the panel is done and it only now starts getting spicy so would I suggest that you guys you talk to the speakers aft with so many things that we couldn't discuss like economic pressure to centralize the the influence of Mev the influence of the network effect of staking tokens but there's a whole weekend to discuss that and want to thank you speakers and want to thank you audience and have a good conference and hack them amazing