S3:E2 The Future of Ethereum Staking with Collin Myers
On this episode of Archebyte, we are joined by Collin Myers, Founder of Obol Labs. The Obol Network is an ecosystem for trust minimized staking that enables people to create, test, run and coordinate distributed validators.
Collin kicks off the episode by laying some groundwork and explaining the history of staking infrastructure and how it has led us to where we are today, post Ethereum Merge. We briefly discuss withdrawals and how they affect staking dynamics before diving deep into DVT - Distributed Validator Technology. Collin breaks down what DVT is, how it works, and what it can enable for validators as well as the crypto ecosystem at large. We discuss the challenges that at home validators face and Collin shares his advice for becoming an at home validator and overcoming these challenges.
We close out our conversation as Collin looks to the future and shares what he’s excited to see develop in the crypto infrastructure space in the coming years as we acknowledge that blockchain tech is at its core one incredible use case of applied cryptography.
📬 To keep up with the latest from Archebyte and receive articles and research from the rest of the Archetype team, subscribe to our newsletter: http://eepurl.com/iCApL2
- - - - - - - -
0:56 Welcome Collin
1:28 History of staking infrastructure
6:05 How withdrawals will affect staking dynamics
10:23 DVT: Distributed Validator Technology
11:49 How DVT works
13:45 What DVT unlocks for validators
23:07 How to become an at home validator
25:28 The future of crypto infrastructure innovation
👋 FOLLOW US
Obol Labs: https://twitter.com/obolnetwork
Today's episode will be focused on staking, specifically staking infrastructure on Ethereum in the past, present, and future. I'm joined today by Collin Meyers, the CEO and founder of Obol Labs where they are building out distributed validator technology to better the Ethereum Validator Network.
Collin is also a long time builder and a crypto OG, having started his crypto career at ConsenSys back in 2018. Welcome to the show, Collin.
Hey Katherine. Thanks for having me. Been looking forward to the discussion and can't wait to get into it.
There's not a lot of people that I think I can riff on for an hour about staking infrastructure, but I think you're one of the few, and I'm excited about this.
So before we dive into all the details about DVT, (explainer can come later), can you paint a picture for me and our listeners on the history of staking infra over the years? So how it got started to where it is today. So in other words, the past and present of staking infrastructure.
Yeah. So the way that we view these things is, well, I'll talk about them broadly and then I'll talk about them more in the context of Ethereum which is important. So like broadly proof of stake conceptualized around somewhere between 2011 and 2013. It's unclear to me exactly where it started.
Some said it was PURE coin, there’s a variety of different white papers. But ultimately, over time, like proof of stake continually as it advances, it advances in kind of the same area, which is it increases the amount of validators, right, that a proof of Stake Network had. So earlier versions of Proof of Stake were like 10 to 25 validators, which is like V0. V1 of proof of stake was like 25 to 100. V2 of validators was 100 to 1000.
And then most recently with Ethereum in proof of stake and ETH 2.0 it's more like 1000 to 1000000 is like what it can enable. So that's the spectrum and from where I sit as to how proof of stake has evolved over time. It's been a continual battle to figure out how to add more nodes to the network without it being too high of cost, without it increasing barriers to entry.
But really, it's like, how do we get more machines to work in a cryptographic way properly in the context of these two? Now, more specifically, I consider Ethereum to be the proof of stake leader. They're using, like, advanced cryptography inside of these two, which is called the signatures, really signatures in my opinion, are like the key to proof of stake scalability.
We wouldn't necessarily be able to do what we do without VLS signatures. And that big push from that V2 to V3 of like Proof of stake globally has really been driven by VLS signatures. So ETH 2.0 now in more specificity launched in early December 2020 at the original onslaught, I believe we have between 14 or 15 million ETH staked today in the network in total.
At the beginning, though, we were looking for 524,280 ETH and it was like a massive like what we thought barrier to entry. Everyone was really scared. Are we going to hit it? Because Genesis needed to be ignited by like that threshold essentially. So that threshold was quickly reached in early December 2020. I spent a lot of my time at that point in time getting deposits and kind of like helping get to the threshold.
And that kind of started the mission of like Ethereum proof of stake infrastructure over the past two years. The infrastructure has been bound to ETH 2.0’s functionality. So there's been a couple of different milestones. One of them was getting the genesis, let's get the deposits, let's have Genesis Ignite, let's start the chain. So that was step one.
And then step two was we need to build more and more security into the chain before we merge the whole thing. So it was like a year and a half of like deposits, deposits, education, user interface, discussions about withdrawals and kind of prepping for the merge. And then this past year, as everyone knows, we finally got through the merge, which now is taking us in the time period that we're in now, which is from merge to withdrawals.
And once withdrawals are enabled, it kind of really pushes the infrastructure space forward, also the product landscape built on top of it. So we're one week from Shapella, basically roughly one week, maybe even less at this point. And that kind of now takes the theory of staking infrastructure and also staking products to kind of its end game, I would say. There will be more innovative roadmap hurdles to get over.
But right now for the Ethereum infrastructure side, we've been on this nearly two and a half year mission to get to withdrawals and enable that, and that's brought about a variety of different products and solutions. And through that two and a half years, it's been the incubation ground for DVT, and we've always imagined DVT. It's very important when you're building these things for the record to stay in line with the core Ethereum roadmap.
We received good information early in the project from the founders of Optimism as like, this is what you should do. You need to stay on the roadmap and make sure that what you're building makes sense in the context of the greater machine, if you will. So for us, we've always wanted to roll out DVT, post withdrawals for like a variety of different reasons which we can get into later.
But yeah, we're at the withdrawal stage today. It's been about two years of waiting to get here and in aggregate almost a decade of Proof of Stake innovation and with withdrawals I believe that proof of stake is about to enter its next paradigm, one of which will be enabled by DVT.
Okay. I was going to ask you about withdrawals, because a lot of people have been talking about withdrawals, especially with the Shanghai upgrade. Like, how do you think withdrawals will affect staking dynamics?
Yeah, it's a good question. So we actually spent a lot of time on this internally. I used to read a lot of daily observations by Ray Dalio. If anyone who listened to this, as has read those back in the day, he kind of moved off of those. But I've been trying to create inside of Obol, kind of like an internet bond report or a market analysis, to the different staking products and their different tokens and the receipts and how they trade.
So recently we put together a really good analysis which is on our blog, and we have a company with a huge dune dashboard of all of the interesting metrics that we think are going to happen or how they would behave if the stakers behave in a certain way. And the big thing here with withdrawals is that it's the first time those who have staked into the network have the ability to like, comfortably switch products.
So what that results in is like a whole bunch of new stealth projects getting ready to launch before and or right after. And it also results in a bunch of protocols that exist today having to upgrade. And this is what you're seeing with Stake Wise V3. This is what you're seeing with Lido V2. And they have to kind of push forward a new, not just because of withdrawals.
Some of it is due to withdrawals, but it's also for those products time to innovate into the next chapter, and one of which can be enabled through withdrawals. The way I describe withdrawals, it's like the next great microeconomic event of Ethereum. And what I mean by that is over the past two and a half years, all this ETH has been staked on the network and then all of this reward ETH has been locked.
And since then we've invented a bunch of cool technologies. One of them is 0xsplits. The other one is different payment streaming services. And now what people can do is they'll be able to take their rewards from their validators and split it immutably in any direction that they want. They can donate to public goods, they can fund other projects, and that results in this microeconomic effect, where there will be a whole new economy built around rewards and where they go and what they fund and how they're used to build the ecosystem.
And like the way I categorize it is like, there is a stockpile of the world's most valuable asset, sitting there waiting to get reinvested back into the ecosystem. So you can think the way that we've approached it is what happens to LST prices, what happens to total stake between your at home validators and others, and how to stake kind of shift around.
And then the more kind of macro one is, what happens to the microeconomics of Ethereum staking and where do all the rewards begin to get funneled in the ecosystem? Because I really don't believe that most of them will get put in people's pockets. I actually think that a majority of it will get reinvested back into the community, which will have a profound effect in our next cycle.
Yeah. And I guess time will tell to see if that happens. But I think your line of thinking makes sense, especially as more people get onboarded to staking, I imagine the rewards may not necessarily be enough to cash out, let's say, on a very frequent basis but rather reinvesting or recycling, or it's good for the network to be actually economically might make more sense.
Yeah. And there's also like this whole group of at home validators that can't pay for their machines right now, because if you're at home validator, you don't have the benefits of the liquid staking tokens to get liquidity if you need it. So like I also am quite bullish on this next chapter of at home validators now that not only ETH is cheaper, but the fact that you can pay for your operations.
I think like in Western worlds, we take for granted how expensive or hard it is to run an at home validator in other countries. We have a super diversified team, we have 25 people spread across 14 countries and we are running at home validators in homes out of Jordan and India, Honduras, Cape Town, all these kind of different abstract countries.
And honestly, it's not profitable for people to run at home, especially if they don't have access to their rewards. So when this happens, I think there's going to be a whole new trend and movement towards running nodes at home because now you can actually see benefit from it.
Yeah, well, this is a great segue to actually talking about DVT, because I do think it does change/ improve different types of staking. So bringing it back to DVT, specifically, what you guys are working on at Obol Labs, so in your own words, what exactly is DVT?
Yeah, good question. So I’ll keep it super simple. DVT enables you to run a validator across multiple machines. It enables you to take a validator which today can only be run on one machine. It can enable you to run it on a number of machines, actually, however many machines that you want. Now, it's kind of being bound by what's profitable or what the latency requirements are.
People think of validators as either a public or private key pair or a person or an entity. But with DVT, you can take a validator and you can run it with a community. You can put different people, different entities, different locations. So at its core, you can run the validator across multiple machines, across the Internet, permissionlessly. And with that, we're now just learning honestly the different things that you can do with that or the different benefits that you can have with it.
One of the biggest use cases today for DVT has been collaborative clusters. So groups of people coming together to share validation rights just in case someone's at home validator goes down or just in case someone has a problem in the cloud. It's kind of collective action, which is very Ethereuml in nature.
From your description, I will say it almost sounds too good to be true. And I imagine a follow up you get often is how? You know, like how do you allow for this way of one validator to be run simultaneously on more than one node or machine? Like, what's the big unlock that allows for this?
Yeah, that's a good question. So technically, what enables it are the anchor backed VLS signatures, signatures that kind of allow this aggregation mechanism cheaply and quickly. So like we're able to maintain ourselves as the middleware and we're able to kind of quickly bring together groups of people in the amount of time that you have to do your duties.
Because VLS signatures are homomorphicly additive. And what that means in normal speak is that they can cheaply and easily be divided up into a bunch of pieces. And then cheaply and easily be recombined to perform a duty. So the signatures are kind of like the crux of how this is even possible. And then the second piece of it is DKG. Today, like a validator key is an individual key.
And with inventions like distributed key generation, this is what actually powers the whole thing, right? So, Katherine, if you and I were to share a node, we would do a DKG together. You'd get half the key. I get the other half of the key. We then use these key shares to, like, stand up our own machines. And then at our own machines, we can pick whatever clients that we want.
And you and I can say, look, you run these two minority clients, I'll run these two majority clients, and then we'll have like a good balance and stack across it. So without VLS signatures and without DKG, none of this would be possible. And those are two like heavy cryptographic areas of research and study. And that's really kind of, you know, bounded is to like what exactly is DVT?
And it's really like the cryptographic research phase of how to make a validator as trust minimized and eventually trustless as possible. So like a DVT validator is bounded in cryptography and like really what Oboll is, it's like a research project in applied cryptography, really.
Is there a reason why someone may not want to or need to stake via DVT? And then conversely, you know, for folks who are looking or considering, like where does DVT come in to really improve different types of staking?
Yeah, I'll selfishly start with the latter because it's a better answer. The first one's a bit biased, but we're still looking for reasons why you wouldn’t want to, to be completely honest with you. It's not clear to us why you wouldn't. We're a bit biased looking for feedback on that from our listeners as well. So like to the benefits. DVT is for all validators, right?
Of all sizes. You can be a block Damon or figment. You can be an at home validator, you can be a small group of people. So what we've spent the past year and a half on is testing the software with different user groups to figure out what the benefit is. And honestly, like over time, the benefits just kind of keep increasing.
So I'm happy to spend some time say, on the different user groups. I'll start with liquid staking pools as they've been the earliest adopters of the technology. For example, Obol received a grant from Lido before we ever did our first fundraising round and Lido was one of the first large scale testers of the software with us. Well, I guess like with staking pools, first and foremost today, they have different architectures.
So I'll walk through like a couple of different ones because they're not all the same anymore as they once were. Lido Specifically what they're trying to do is like move forward with this V2 architecture, which is super cool. It's a module based architecture. We kind of liken it to like the Maker debt ceiling model.
Someone proposes a module. That module has a certain set of configuration. It has kind of a cap and then stake rolls into it. So like in the Lido sense, what we think makes sense for them is to like Host the DVT module and like what that enables them to do is inside today of liquid staking pools. Let's just use Lido as an example.
There's either 21 or 25 operators today. All of those operators are just running their own individual machines. If they were using DVT, they'd all be running the same machine or the same validator across their own machines. And the benefits of that are that any one of those validators in that pool is now not able to self slash. They're not able to act malicious and they're not able to sign any incorrect messages.
So what DVT does for more mature pools now is it actually, DVT is finally being used for its Byzantine like properties. And like it prevents Byzantine behavior. And now pools have gotten to ten, twelve, $15 Billion in size. So as you decentralize a pool, one of your main objectives is like, how can I prevent my validators from being malicious or acting Byzantine without having to include too much economic overhead.
And for liquid staking pools, due to the cryptographic properties of DVT, it enables one single validator to not have power to take the pool down, which is very cool. The other use case is also, a lot of liquid staking pools started with a fixed set of validators that were permissioned. And most of them were professionals.
And if you want to build the most resilient liquid staking pool for Ethereum, everyone agrees that you should include at-home validators, SME validators, smaller people into the mix. However, if you're the liquid staking pool, you're like, Hey, there's a lot of value at risk right now. Running a validator today is still pretty hard. So how do we, like, build a bridge inside of that ecosystem? And that's where we're doing a lot of cool testing with Lido today. It's like between the at home validator and the professional and actually watching them come together and trying to get them together to run these clusters.
So that ultimately decentralizes the validator set, gives rewards to the average person rather than just an entity. And it's really powerful and it's decentralization effects. Another example of this is Stakewise V3. Stakeweise V3 has this like vault deployment model, which is super cool. So like anyone can come in and design and deploy like a DVT vault. You know, you can just design and tailor your own vault.
And with the base layer being DVT, and then you can abstract however you'd like on top of it. This is where DVT meets Defi, and to be honest with you, this is kind of a huge area of research that has a lot of greenness to it, if you will. It's a greenfield. That's why we're interested in working with stakeholders and you spend a lot of time with them because I think their role model really pushes it into Defi.
And then as a means of that, it pushes DVT into Defi, which is quite interesting. Those are the 0 to 1 innovations of DVT. Those are like the most innovative things that people are using it for, their use cases. If we back up a little bit and just talk about how the technology can be used from 9 to 10, that's also beneficial and that's for the validator entity themselves.
So the way that Eth 2.0 has been, and an Ethereum validator today has been designed is a single public private key pair and a single machine. That's it, right? And if your node dies, you're incentivized to just bring it back on because the inactivity penalties are too severe that some of the pros who've been slashed historically have done, is what most people like in web2 do, which is called active passive redundancy.
So you have your validator machine and it's running, and then you have a backup machine that has the same set of private keys on it. So if your first one dies, the back one comes online and it saves you. There's a lot of false positives that get created in this. And like historically, the majority of slashing events have been because people have been pushing for higher performance by breaking the rules basically.
And then if they get in trouble, they get slashed. And today getting slashed, like still could mean you lose all of your ETH. So for these actors, these large centralized validator entities, they have built now, some of them billion dollar staking products that have like 10X too much hardware inside of them. And what DVT enables is what's called active, active redundancy.
So now these large operators are like, oh, wow, like, if I'm going to stake 15,000 ETH if I don't need 60 nodes to do it anymore, I actually just need four nodes using DVT, using active, active redundancy between themselves. So that decreases their cost of capital, that improves their margins, which is kind of the classic operating efficiency improvements of DVT.
But more importantly, one of the new things, which is super interesting, is DVT has begun to be used as a mechanism to get better insurance for validators. It's also been pulled into the SOC 1, SOC 2 compliance discussion of how these entities increase their security. So at the end of the day, it's also very important to note that DVT is a security protocol.
It's not this like middleware lottery thing. It's not here to get you more yield, or any of that stuff. Right. It's a security protocol at the end of the day. And it takes a blockchain network, a certain maturity to accept that as an acceptable piece of software to put in the stack and say Ethereum’s the only one that's there.
So, yeah, for the institutional validator, there's been like a lot of benefits that we are just learning every single day. Honestly, as they test it, they run it around the world in different locations. They figure out their costs, they speak to their insurers, which has been quite beneficial and encouraging honestly for us to see that happen. It's really institutionalizing the effort, which is great because we want everyone in the network to be able to use DVT.
But as a project we have to kind of cater to that market and figure out how we work with them. And then third and final is the at home validator. Honestly, running an at home validator today is like not really so easy. We've done the best we could, but like it's not about really setting up the machine. It's about internet connectivity all over the world is the at home validators problem.
And so at the end of December, we launched the world's first DV. It's run entirely at home, and it's run between Ireland, Estonia and Canada. There's two nodes in Ireland. There's one node in Estonia. At one node in Canada. And about a week and a half ago we had our first block proposal and one of the nodes in Ireland has a couple of other validators on it, not DVs just regular validators.
And when the proposal came up to propose, that machine did actually literally at the given moment it died. And then the other backup nodes are the only reason why it was able to propose. So for us, it was kind of like we were blown away and actually showed us exactly what is the benefit to the validator.
And the fact is, if you don't live in Western countries, the internet's bad everywhere. It's not profitable to run an at-home validator because you could lose internet, right? And you can miss a proposal. And proposals are now becoming a bit more rare. So yeah, the at home validator has its own kind of benefits that are still being worked out.
We actually have a second cluster not announced yet, but happy to talk about it today. It's actually a Southeast Asia cluster. Right now it's being run between Australia, New Zealand, India and South Korea. And now we're testing out the local Internet in these areas to see how it would work at home. And this is all being run by actual validators as well. So yeah, the at-home validators are like our tail end, that's our end game.
They need DVT, the network needs more of them, and we're learning from those user types every day.
Yeah. A lot of our listeners I believe are operators and crypto companies. So, you know, folks who work in the crypto space who probably would be pretty interested in setting up like an at home validator. So what's that home validator stack tack today? If someone were to come to you. Hey, Collin, can you help me set up an at home validator? Like, what’s step 1, 2, and 3?
Yeah, so before we started Obol actually, myself and our head of products, we both worked together at ConsenSys prior to this and we helped the Ethereum Foundation build what's called the ETH staking launchpad, which is the primary user interface for at home validators to select their clients, to get educated, to do their initial deposit to pull down the repo and the deposit contract.
We spent about a year with the EF designing and implementing a user interface for the at-home validator and today I still think the ETH launchpad is the best place to run at home because it's educational, it has all the tools that you need, you don't need to grab information elsewhere. That's probably one of the biggest problems with Eth 2.0 and just Ethereum in general is like the decentralization element of it results in misnformation.
You know, what's the source of truth is like a really hard thing. And the eth launchpad, it's open source you can tear it down, there’s no back into it. You can run it locally, set up real validators and as a result of seeing that progress and honestly the backbone of the stake in the network has come from at home validators to the launchpad.
In my opinion. We built the thing called the DV launchpad, which is a kind of multi-party multi user interface to create distributed validators within a group of people. So the launchpad has four user flows and it's migrating the existing validator. It's like I got a normal one, but I want a DV, it's create a DV from scratch alone solo.
So maybe I have four different machines or I want to run a combination in the cloud and then there is be invited to a group and start a group. So there's four user flows and there's a user interface that enables people to easily set these things up. So yeah, I would recommend for normal validators today, continue to use the Ethereum launchpad and then if you're interested in DVs as we get closer to our beta launch, you can use our large test net, but not today for business purposes.
So check out the launchpad and follow Obol for any announcements soon. Awesome. Okay. Last question. And this can be about staking or it's going to be about something else. What are some areas within the crypto infrastructure landscape that you are excited to see innovation in the next 3 to 5 years?
I mean, honestly, Obol’s really bound to improvements in applied cryptography, right? Like we are living in an applied Cryptography renaissance and I think a lot of people forget that, you know, people are always like, you know, what is the killer use case of blockchain? And everyone's been talking about email and trying to liken a blockchain to it. But I think what people are forgetting is that blockchain is the killer use case of applied cryptography.
It's kind of like way deeper than that, right? So like everything that we do as Obol is bound by cryptography because we are an applied cryptography project. So right now, like we're in a ZK renaissance, which is awesome, there's tons of money flowing in. Some of the valuations are ridiculous. There's cool projects, there's innovation, there's researchers leaving universities to fund and start projects.
And ultimately, like, we look at our future and it's very bound to ZK. Honestly, at the end of the day, like if you want to make a validator more trust minimized, we need like collective improvements at the applied cryptography level which we will borrow and apply it. So, we've been actively trying to partner as closely with projects like Axium or any of these other lower layer infrastructure cryptography projects that can help us scale.
Because at the end of the day, we believe that DVT is like the validator end game, but DVT is a 20 year research project, but DVT can't progress without massive improvements in applied cryptography. So I would say, like what we spend all of our time on now is like, you know, what's up in the ZK world, is it getting funding well?
What are different teams working on? How can we partner with them? How can we eventually help fund them right? To help build on top of Obol and help us through these things. So yeah, I would say that applied cryptography is the renaissance that we're living in at the moment and everything will be ZK in this next chapter for better or worse.
Great answer and I love that quote. What was it? Blockchain is the use case of applied cryptography? Nice. I might steal that line someday.
You have permission. I stole it as well. I certainly didn't make that up. I don't know where I heard that, but I came across my radar at some point over the past year. Also kind of it was one of those like it's like, oh, that's exactly what we're doing. Like blockchain? Its cryptography actually, blockchain is just a use case of it.
Yeah, that's true. Okay, well, I think that was a very wisdom, information packed 30 minutes. Thank you for taking the time spent with us and again, listeners, make sure you follow all and all they're doing now. Thank you Collin.
Yeah. Awesome. Katherine, thank you so much for having me today. Shout out to all the listeners. We got some big news coming out this year and if you want to run a validator, come find us and we can help you out.
Thank you for tuning into another episode of Archebyte. We release new episodes every Wednesday. Don't forget to check out previous episodes if you've missed any. They should be available wherever you get your podcasts. If you have any questions or ideas for future episodes. DM uS on Twitter @Archebyte. See you next week.