Who’s who in eth2: Kain Warwick from Synthetix

January 26, 2022

In the sixth installment of our Who’s who in eth2 series, Protocol Specialist Elias Simos interviews Synthetix’s Kain Warwick about what motivates him to push crypto forward, his own eth2 validators, and what keeps him up at night

Welcome to Who’s who in eth2, presented by Elias Simos, Protocol Specialist at Coinbase Cloud. In this series, Elias interviews key contributors to the development and growth of Ethereum and eth2, exploring their involvement in eth2 and visions for the future. Here we share their perspectives on what eth2 means for the world from people deeply embedded in its ecosystem.

Who have we talked to?

The series includes interviews with notable builders, researchers, infrastructure experts, and leaders from the eth2 ecosystem, along with principal members of eth2’s four client teams: Prysmatic Labs, Sigma Prime, Nimbus, and Teku.

In this sixth post, Elias interviews Kain Warwick from Synthetix about what motivates him to push crypto forward, the power of marginal improvements, his own experience running eth2 validators, and what’s keeping him up at night ahead of the merge.

Q. Let’s start with a little bit of background about yourself, your involvement with Ethereum and, more broadly, how you got here. What was your lightbulb moment when you thought “Ethereum is going to be a big deal”?

Dating back to 2014‐2015. I was already down the crypto rabbit hole, and at the time I was running a payment company that was working with a lot of the brokerages in Australia — so allowing people predominantly to buy Bitcoin. I was aware of Ethereum, but I was probably still too much in my Bitcoin maxi phase to have really gotten it.

Then, around the time of the decentralized autonomous organization (DAO), I'd been reading a little bit more as that DAO process kicked off. I think there was just something around $150 million pouring into the DAO that made me think there's something interesting going on. I would say that was the lightbulb moment – after which I really started digging in and reading everything I could.

Then, in late 2016, I kind of went down the stablecoin rabbit hole. I’d obviously been aware of Tether for a long time. But it was reading about Maker, and reading about some of the approaches people had taken on the stablecoin side, that actually got me to build something in this space.

Q. More broadly, if we open the scope up beyond Ethereum, is there something in particular that motivates you to be in this space and to help push it forward?

So, I have this view that crypto has the ability to tilt some of the disequilibria in the world today. I think there are things that are fairly obvious that we want, that are really hard to solve when you have to give up so much of the coordination power to central authorities and governments.

The idea that crypto is a coordination mechanism — where you can create rules and everyone has to abide by those rules with a path to upgradeability baked in — really strikes me as something that I think is going to be similar in a way to how the internet, for good and ill, made information so accessible. I think crypto has the ability to facilitate these coordination problems.

And humans, while I think they may mean well, are highly corruptible. It's very easy for things to go wrong if you give people too much power.

Q. Do you think there’s a chance that we in this broader space and community will devolve into mirroring the structures of the old world, voluntarily or involuntarily? And if that’s the case, what can we do to prevent it?

I think there's a human tendency to defer to authority. A lot of people are very happy, in either certain aspects of their lives or many aspects of their lives, to just not have to worry about things and not have to think them through deeply.

That’s where the need for authorities emerges, and in my opinion it is kind of problematic. Because as soon as you start deferring to authority figures, those authority figures have the ability to abuse that power — whether intentionally or unintentionally.

Holding on to power is also something that is very human. Desire and drive as well. This is why I think the idea of smart contracts is so powerful — these coordination mechanisms that you can create where the rule sets are fairly rigid but not too rigid. I do think that in the absence of that there will be pockets of control that emerge.

This idea — of creating these coordination systems that are incentive‐compatible — has the ability to offset and maybe overpower that natural tendency of humans to move for authority or hand over power. 

Once those privileged positions disappear, there's no power to accrue at all. You just have these autonomous structures that are operating based on the rules that everyone agrees on. And maybe there needs to be some meta‐coordination of how we change the rules, etc. But I think that how we change the rules oftentimes can be market‐driven.

It's still so early, though. It's hard to say whether this just devolves back into the normal — regresses to the means of normal human behavior and human history.

Q. Still, there’s something really powerful in thinking about a lot of marginal improvements, sequenced one after the other, eventually leading to an absolute change…

In complexity theory there is this idea that complex systems have to emerge from simple processes. You can't design a complex system top down and just say "I've got this amazing thing" — and then just build it and let it go. 

One of the things that Synthetix challenged significantly was this idea of creating these perfect waterfall systems — where we're going to spend three years building this thing, and then we're just going to release it magically and it's going to do all the things that we hope that it's going to do. I think the tendency to do those things comes from too much of an adherence to immutability, and this idea that we're going to have to release something that, once released, cannot change.

If you go down that path, it creates an almost unattainable standard for you to achieve. Augur, I think, is the best example of this. We’re up to V3 now — and every time the same pattern emerges: they design something seemingly in a vacuum, then it's released and it just doesn't get traction because there are issues. But it's too immutable to be modified in a way that addresses the issues the users have. It just doesn't work.

At Synthetix, we took the opposite approach: building small components and then iterating. When we first launched back in 2018, proxy contracts existed. But I think we probably took it much further than a lot of people had, where we just put proxies in front of everything. Now it's much more common for people to have proxy contracts and multi-signature wallets (multisigs) that control things, but back then that was not the case at all.

The community was, at least initially, strongly opposed to this approach. People were like, "This is horrendous, you have total control over it." But we were still convinced that was the best way to scale the protocol to billions of dollars in value. 

Judging by the outcome, this was the right thing to do. But I should note that it gets to a point where you do accrue a lot of tech debt, and that process has its limits. 

I think we've hit those limits, and we've looked at how we then draw a line and start from scratch. Then we take the learnings and start iterating and implementing new functionality from a clean slate.

Q. The story of how you went from Havven to Synthetix, and to where you are today, is pretty much common lore. But what I’d love to touch on is when it became clear to you, from a builder’s perspective, that Ethereum will need to scale. Was there a particular moment?

Absolutely — in fact, it's one of those where I can cast my mind back to that exact moment. It was probably three hours into Black Thursday [when cryptocurrency prices crashed alongside traditional markets on March 12, 2020]. It was just one of those crazy days where I thought to myself: "OK, this is a three-sigma type event. We really have to observe and try to understand what’s happening here."

"To me, that was the moment when I thought to myself, ‘It doesn't matter what the demand is in any given second, it matters what the peak demand is’"

The problem was that so many interconnected positions across decentralized finance (DeFi) had been composed between Maker and Uniswap and all of these different projects. It was very obvious that we'd reached a tipping point in terms of the amount of economic activity that was there.

You don't notice it until there's volatility. This complexity via interconnectivity was building up over the course of several months and in some cases years. And all of a sudden there's a sharp move — and, as everything is connected, it all pulls together. 

Every single interaction requires block space. And at the time it became clear that there just wasn’t enough block space for the amount of transactions and the amount of economic activity that was on the network at that time. We saw the mempool just get flooded, and it took almost a week for it to get cleared out after Black Thursday. 

To me, that was the moment when I thought to myself: ‘It doesn't matter what the demand is in any given second; it matters what the peak demand is.’

It was such a volatile day that it showed what was going to happen in the future even on much smaller moves. Because the economic activity wasn't going to go away — it was just going to build up. 

It was basically from that day forward that we put 100% of our resources — that we'd been using to look at how to improve and optimize the protocol — into trying to get the Optimism integration going.

Q. Let’s talk a little bit about your decision to work with Optimism. What was it that prompted you to commit to building the future of Synthetix on top of Optimism, and how has your experience been so far?

I think there's two categories, at least right now, of DeFi protocols. 

On the one hand you have the ones that have scaled with some sort of off‐chain computation. DyDx and Deversify are pretty good candidates for that as they can really take some core actions off‐chain, like order-matching, for example. 

On the other hand you have protocols like Synthetix that operate purely on chain. Because all of that state is then on chain, we virtually have no option but to try not to break composability.

"That's just a good thing. It doesn't feel good to be the person the market is punishing. But it's still important that you're in a market and that there is competition, and there's multiple people and the stakes are really high."

Particularly when we were doing initial due diligence, it looked like zero-knowledge proofs, and the menagerie of different scaling solutions around them, were just not quite there. Roll‐ups weren't either, but it felt to us like roll‐ups were closer. 

Ultimately, it's people that build these things, and the Optimism team really impressed us. You need to have a level of confidence that the team you're speaking to can actually execute.

I'm still confident that picking Optimism was the right choice, even though every single scaling solution is taking longer. It's just a class of problems. 

I think scaling is just really hard. As you get to the last 5% or 10% it can really blow out on you. But the end result is that I'm still really happy with that choice. I think the Optimism team is in a really good position to execute. 

But, I also believe in markets — and I think, if we're wrong, the market will punish us and it will reward someone who chooses someone else. That's just a good thing. It doesn't feel good to be the person the market is punishing. But it's still important that you're in a market and that there is competition, there's multiple people and the stakes are really high.

Q. What are your thoughts around roll-ups breaking composability of L1 Ethereum? Is there any first-hand experience you’ve had with Synthetix to that end?

Synthetix is interesting because we broke composability for exchanges, which was a pretty critical function within the system until a while ago. So maybe one of the ironies of this transition to a "less composable networking in Optimism" — I say in scare quotes — is that we actually will regain composability. 

Rather than using the L1, which has been kind of a speed bump to prevent front running of our oracle updates, the oracle updates will be so much faster that we will be able to actually forgo the fee reclamation mechanism completely, which is what breaks composability. So we're actually going to get some composability back!

Now, that’s all well and good — but if you're the only people sitting there, then there's not much to compose with. I think that's the broader concern. Which is why I think the social aspect of this transition to Layer 2 is so critical, and why I think people paint me as an Optimism maxi. Sometimes I probably am, and part of the reasoning behind that is that we need to coordinate on something. It's still better for all of us to coordinate on that and all go to that one thing, because it's far better than the status quo, which is Layer 1 right now.

That's why I've been so aggressive in advocating for people to engage with Optimism and be prepared to jump in and deploy to that network. Because I think that once we have social consensus among the different projects building on Ethereum, it will become much more obvious that we can remove that fragmentation.

But if we mess it up and it gets fragmented, then it's going to be problematic.

Q. I hear that. This is a good segue into talking a little bit about eth2. Broadly, how does the whole idea of transition from eth1x to eth2 affect Synthetix?

When I joined the Ethereum community, the light far off in the distance was Serenity. The interesting thing about that Serenity narrative was that it was somewhat all‐consuming. It was like, we're going to this place — this final state. Over time I feel this caused some confusion in the community of builders on what was to happen with the existing state that we’d all worked so hard to build up.

This is much clearer now, given the eth2 roadmap. I think people are much more comfortable with the idea of eth1x — that we’ll be able to live there, it'll work. We're going to have roll‐ups, and eventually the roll‐ups will get connected to eth2, and we'll get even more scalability. So there is much less concern and confusion I think these days than in the early days.

And for whatever it’s worth, I think one of the smartest things that the core ETH community did was to construct this eth1x< >eth2 narrative. I think that really helped to create something that was comprehensible to people. And now it has served its purpose.

One thing that we've kind of taken away from all that is that we've got a similar problem in this launch of Synthetix V3 — this idea of a clean-slate rewrite of the contracts. But Synthetix V2 is not done yet. So what we've come up with is a similar idea of Synthetix V2X, which is the existing implementation that's going to be running parallel on Layer 1 and Layer 2, that will continue on for the foreseeable future. 

In the meantime we are heads-down researching and working towards decoupling them from the effort of maintaining the existing protocol, and allowing them to work on R&D on V3.

So, one of the big takeaways for me is how important it is to communicate what these transitions look like.

Q. A little bit of on-chain sleuthing, courtesy of Nansen, points to the fact that you are actually participating in eth2 as a validator. Is that accurate?

That is accurate! I've got a Nansen subscription but I try not to look myself up, because I shudder to think what my addresses reveal about me these days. But I've got my public address, my kain.eth address.

Q. That’s the one. I only noticed because I was just building a bot on eth2 slashings and trying to add a semblance of an identity layer on top of that.

I'm like the most slashed validator on there, right?!

Q. You have gotten slashed? I feel I would have known if you were…

I kid. I wasn’t slashed. It was just offline penalties. But I think I was offline for like three weeks at one point. So, yeah, it's not anything to write home about [laughs]. But it's good that eth2 has this model of thousands of validators, so that bad uptime from a single operator is not taxing on the network.

Q. You’re touching on a really interesting point there, in that eth2 allows for smaller operators to get in, experiment, distribute the stake a little bit more, etc. How was your experience deploying and managing those validators?

You know, I used to run Bitcoin miners back in the day. I think it was much easier to set up an eth2 validator than it was to get a CG miner working. 

I just think the reason for those uptime penalties was that I was a little bit blasé about it and didn’t monitor uptime as much as I probably should have. And ultimately, I think the main culprit here was the Roomba running into my machine and turning it off. And I didn't even notice! It's not the most redundant set-up in the world, but it was more for fun than anything.

Q. Let’s dive a little deeper into what prompted your decision to run validators on the Beacon Chain. Was it experimenting with eth2 and being an early adopter? What was it that drove you to actually build your own set-up and be there early?

I think it was a combination of hubris and just wanting to participate. I wanted to make sure that I had validators running for Genesis. I stayed up until 3am I think for the actual launch. 

It's a moment in time — a ceremonious one, at that. eth2 is going to be the thing that in 20 years’ time we're going to look back at and you'll forget that there was a whole eth1 era for five or six years. That'll be kind of ancient history. This will be the "I was there when this network launched" sort of moment.

Q. Most DAOs in Ethereum hold a significant amount of reserves in ETH. Besides the path dependency (raised in ETH, collect fees in ETH etc), what purpose do you think this serves?

Firstly, it’s worth noting that the very fact that DAOs hold ETH reflects a shared sense of responsibility. One of the biggest problems in coordination is when you have a diffusion of responsibility. How do you maintain a sense of responsibility within public goods? This is antithetical to the way Web2 has developed so far — i.e. we all use some parts of, say, AWS, but we don’t all hold Amazon stock, nor do we need to.

With ETH it’s a sense that no single entity is running this network, no single entity owns it. Someone owns AWS — there's a guy over in Seattle who's running the place — whereas with ETH, I don't think people think that Vitalik [Buterin] is sitting there looking at the master node, making sure that we're all okay.

Q. Dropping ETH in the pool for all to enjoy…

Exactly. I think we all need to make this work. I think there is a sense of responsibility there — alongside the financial alignment and the direct-incentive alignment — of just a cultural alignment and a sense of ownership.

Q. Given what eth2 brings to the fore, then, do you foresee DAO treasuries participating as validators in the network?

It would be weird if they didn't. A lot of treasuries are already yield farming. When you think about that as a process, there’s risk there, as in there’s obviously volatility risk when you hold ETH. 

But I feel there's a sense that holding ETH is almost an ideological thing to do. You don't want to sell your ETH, you want to hold your ETH; this is the place that we live in, after all.

So I think there is a component of: if you're going to hold the ETH and you're going to sit there for a while, getting yield on it just makes a lot of sense. It's almost the sensible thing to do. And even though there is some risk to it, it's much less risky than yield farming, and it creates an alignment with the network.

Q. What do you think the path of least resistance is for these DAO treasuries to actually participate? Is it through staking pools, like Lido? Is it running their own validators? Is it going through someone like Coinbase?

Good question! I don't know that a DAO has the ability to take a non‐fully‐on‐chain approach to staking. And unfortunately, a DAO just can't run its own hardware. So almost by necessity it needs to outsource that to something that is trustless and non-custodial. Most custodians probably wouldn't let a DAO sign up these days. That might change in the future, but I think that's a bit of a hard ask at the moment.

But taking your treasury and dumping it into Lido or some other on-chain staking solution, and getting a liquid asset that you can get out at some point in the future, may be the only socially acceptable approach for a DAO to take.

Q. Switching gears a bit, with the merge approaching, what is it that maybe keeps you up at night sometimes?

Well, the big unknown at the moment is how the other scaling solutions (like roll-ups) will work out: how smooth it's going to be, and how much consensus there is around the different approaches. But I do think that the fact that we have the beacon chain operating is something. It's a bit of a zero‐to‐one moment. I think the rest of it feels like it's almost inevitable, whereas there was definitely a sense before that this wasn't even going to be a thing.

Q. Is there something that particularly excites you about the future, and in light of the merge and thereafter?

What excites me is a future where it's all working seamlessly together like a well-oiled machine; all of the fragmentation and challenges have been shaken out, and all of the things we're hoping for are going to play out will play out. 

That’s my view, but it's definitely going to be a painful process. It's not going to be smooth over the next 12 months as we try to navigate all these different scaling solutions and get through the merge.

But I remain hopeful that we will come out of this with a good outcome.

Interview by Elias Simos