On The Brink with Castle Island - Andrew Cahill and Saurabh Deshpande (The Block) on Ethereum L2 Scaling (EP.323)

Episode Date: May 31, 2022

Andrew Cahill and Saurabh Deshpande of The Block join the show. In this episode we discuss: The Ethereum L2 ecosystem The varying scaling solutions and views on the communities with traction Views on native tokens being launched on these platforms in 2022 Optimistic rollups ZK rollups Learn more by reading the report, "Layer-2 Scaling Solutions: A Framework for Comparison"  

Today on the podcast, I sat down with Andrew Cahill and Sarab Deschpante from the block to discuss their new report on layer two scaling on Ethereum. So there's a ton going on from a development perspective on Ethereum L2s and the team at the block is all over it. I think you'll enjoy this one. We went deep on a number of the projects. So without further ado, here's my conversation with Andrew and Sarab from the block. Matt Walsh and Nick Carter are partners at Castle Island Ventures. All of these expressed by them or the guests on this podcast are solely their opinions and do not reflect the opinions of Castle Island Ventures. You should not treat any opinion expressed by anyone on this podcast as a specific inducement to make a particular investment or follow a particular strategy, but only as an expression of their personal opinion. This podcast is for informational purposes only.

Brought down by bad mortgage investments, Lehman, which has 25,000 employees, will be liquidated. The federal government loans American International Group, AIG, $85 billion. This is a different kind of market, and the Fed is asleep. The federal government is stepping it to stabilize Fannie Mae and Freddie Mac. the two mortgage giants that have been threatened by the housing crisis. The bank of England has pumped 75 billion pounds more to Britain's ailing economy with a new round of quantitative easing. You print a couple trillion dollars and all of a sudden people start to worry.

So out of this worry, we have something called the Bitcoin. Bitcoin. Welcome to On the Brink. I am very happy to have Andrew and Sarab from The Block. And it's an interesting day to be talking crypto. Luna and Terra, by the time this comes out, something will have happened. But it's a tumultuous day. we're here to talk about layer two Ethereum scaling now.

You guys just came out with a great report. So thanks for joining us today on the podcast. Thanks for having us, yeah, big fan of the show. Thanks for having us, yeah. I appreciate it. Well, maybe just set the stage, Andrew, in terms of what was the impetus to start to tackle just a view on layer two scaling with the block? Sure, yeah.

We kind of hit layer ones, I want to say, like six months ago at a time when that was like the talk of the town we had like, you know, Solana Avalanche kind of starting to gain some traction. And I think at this point, kind of seeing some of the challenges that Salana's been having and kind of like how the Ethereum community is kind of pivoting to like this multi-chain framework. Like at first it was this whole ETH too, like we're going to execute everything within Ethereum. And now it's kind of, they're like very set in this like layer two kind of path where everything is going to be executed on these separate chains, which is just and they're kind of at a development point where like this stuff is all ready in production. We're starting to see users kind of use it. And then we also have this kind of interesting thing we touched on the report, which is like kind of what's going to happen on the token side, right? Like a lot of them are going to end up, they have to find a way to pretty much incentivize more people to use their networks.

And that's going to involve launching tokens, which is always fun to kind of look at giving, you know, folks at the community opportunity to kind of participate directly in the protocol. When you guys write these reports, I mean, who are you exactly writing them for? Because as I was reading this report, I was thinking, you know, there's obviously a lot of investors. that would want to know this information. There's probably also a lot of people building things on blockchains that want to know what the right platform to build on is. So who do you think about when you write these reports? That's a good question.

Yeah, I say we try and write to like the broadest audience. So whether it be investors, enthusiasts, or people, like the goal is really just to kind of give people a lay of the land. Like we're never going to try and pick one, you know, a network that's going to work. It's kind of just giving people, you know, a little bit. In this case, there's a little bit of background of like the whole story. of Ethereum, like why it's seeing so much usage, kind of what issues that's causing for

users. And yeah, just kind of give the broad community an overview of the lay of the land and then just get as much data in there as we can as people always find that interesting. Yeah, the data is great in this report. And obviously, it's coming at a pretty interesting time. We just had the Ugo Labs, NFT situation a couple weeks ago with the massive gas spike and Ethereum. And so just the need for layer two solutions is top of mind for a lot of folks right now.

be curious how you guys each just contextualize where we stand right now in terms of the production readiness and the ability of some of these layer two is to actually meet the customer need. And just where is the industry from a health perspective on layer two right now? The biggest thing we saw in the report was like if you look at like the center of a lot of these networks is the virtual machine, which kind of dictates how the apps are coded and kind of how they run. And optimistic rollups have launched with like EVM compatibility. So we've seen a lot of like applications kind of quickly move over there that are like ready and up and running and people are using them on like the ZKK roll-up side.

Obviously it's a challenge to kind of replicate that environment directly. So we're kind of at a point where you have someone like a StarCware that's kind of gone the route of kind of building out their own language and development framework for people to kind of launch this. And they've also like work directly with people to work with them to like kind of recode their stuff and port it over into this new environment. but then there's this whole separate cohort of protocols that are like well underway with developing these call them like ZK AVMs so these virtual machines that are compatible with Ethereum tooling so we have like I think we flagged that like Matter Labs has one that's in TestNet but Polygon who is actually the sponsor of this report they're incubating like

I want to say like four separate projects that are all working on attacking this this question from different angles of like how do we give developers the tools to be easily able to kind of port their stuff from Ethereum over to our chain, because that does sound like a huge competitive advantage in terms of building it out. You have to get users on these platforms as soon as possible to have a chance at, you know, competing not only against other L2s, but also like the broader kind of like layer one competitive landscape. You know, I don't know if sort of if you had anything else to add, but.

Yeah. I mean, in terms of the landscape, I think we could pretty much break them down into optimistic and ZK roll-ups, right? And as far as optimistic are concerned, the existing Ethereum tooling is, I mean, they can use that existing tooling. Wallets like MetaMask easily plug into, I mean, you can use them for,

like your arbitram, optimism, etc., etc. Right? I mean, you just shift to that network and the user doesn't feel any change except for the lowered gas fees. So as far as optimistic roll-ups are concerned, I think they are pretty much up and, running, barring the liquidity on various tokens, right?

I mean, if you look at the L1 pools on uniswap or any other decentralized exchanges, you can pretty much swap from any token to any token. But as far as the L2 is concerned, the liquidity is pretty much limited at this stage. So I think there needs to be improvement along those sides. Right. I mean, we've seen that optimism has already suggested that they are going to drop their token. I think the allocations are already stated.

So pretty much, I think, on. On optimistic role-up side, what we are waiting for is sort of incentivization for all of this liquidity to move on to the L2s, right? And we saw that with L-1s when the Dexas were competing against each other, that whichever text announced their liquidity incentives, right, and the liquidity sort of move from one next to another. So I think similar can be expected as far as the optimistic role of the concern. On the ZK side, right, as Andrew mentioned, we see that there are a lot of trade-offs that these guys have to make in terms of their design choices. So, if you look at two or three predominant players on the ZK side, you have stockware, you have ZKSink from Matterlabs, and you have, as Andrew mentioned, I think, four from Polygon. So within Polygon, their approach is kind of different. Hermes has chosen to build an AVM-compatible virtual machine, right, ZK-AVM,

and they are going to come out and launch with that, right, so that the existing Ethereum protocols can easily port over to do their roll-up. But if you look at stockware, they are preferring application readiness or ease of building applications over and functionality over the EVM compatibility. Right. So they have developed their own stuff with Cairo and they are building a transpiler, which will convert your solidity code into Cairo and compile it in Cairo's machine language. And again, as with Matterlap, it's a similar thing, right?

They are favoring EVM compatibility over functionality. So, Starknet recently announced their bridge from Ethereum to Starknet. I think yesterday or day before yesterday, it was in the last few days, they said that you could port over your assets from Ethereum to stockware. As far as the readiness of, like, Hermes has launched with their V1, which is just, I think, payments. We can just transfer tokens from one address to another, nothing more than that. And they are, I think, gearing up for their test nets in the next couple of months for the V2, which will be a more general purpose protocol.

So, yeah, I mean, this is where we are at, where we are waiting for the kind of incentives to drop. and how the users and developers will move to where it is. I definitely want to get into the incentives, but maybe before we do, for developers, what is kind of the core tradeoff between building on an optimistic rollout versus ZK roll-up? How do you guys think about the most important trade-offs there on each side? Just as like a warning, like for researchers,

I wish I had more development experience. But I guess at the highest level, I'd say like if you look at someone like we flagged in the report, and considers their virtual machine like EVM equivalent. So they like to use the word of like copy and paste one click deployment. Like if you have an app that runs on Ethel1 and solidity, you can easily bring it over. And then you have like platforms that are EDM compatible. So if you look at Arbitrum's virtual machine, we wrote that at this, they're working on some

stuff, but right now it's EDM compatible. So you have to make some changes of modifications to that code and, you know, review it before you kind of deploy it in this new environment. And then on the ZK side, with the exception of Starkware, who has their Cairo kind of development environment live, all these other ones are pretty much still in like testing phase or development phase for actually deploying like general purpose applications in production. It's kind of how I'd see the whole like high level.

How about you, Surab, do you see this as a, these two coexisting at scale? Do you think that there's a world where one side of the, equation ends up being more substantial in terms of overall development in the ecosystem? Like Andrew said, I think we are not developers and I wish we had more development experience. But when I spoke to local developers in India, some of them mentioned that if you know solidity, it takes a while for you to learn Cairo, right, and deploy applications there. Whereas others mentioned that if you are a good solidity developer, you would probably be up and running with Cairo in a matter of a couple of weeks.

So I personally don't know which one of them is the correct view of what is happening with ZK Roll-ups. But yeah, I mean, that seems to be the trade-off where developers will have to relearn some of the things and then go on about building applications. You mentioned the incentives. I want to hop over to that. And so some of these L-2s have already launched their own token. Some of them notably don't have a token.

Some of them are rumored to have a token. How do you think about just the presence of a token in this ecosystem and maybe just set the stage for where we are in terms of L2 tokenization? Sure. Yeah, I can start. Yeah, so we looked at kind of six projects in the report. At the time we wrote it, I think two of them had already had tokens. And with all the different networks, it'll be kind of interesting to see how they actually involve them in the ecosystem. I think like the first thing we flagged was that in order to like incentivize people to run this node infrastructure, you're probably going to need. needs a token to do that. That's how all these other networks kind of work was like kind of

the first thing. Another like interesting and I think sort of you have some interesting perspective on this was kind of like on the fee side. Like if you look at all these L1s, pretty much all of them have their own native token that is used to pay the fees on the network, which from like a user perspective is definitely a serious consideration. Like if you want to use this network, you got to have a little stash of their token to kind of pay your fees as you as you do your your transaction. Yes, I don't know if you wanted to add on to that. I know you had some interesting thoughts on this. I mean, I think almost every one of them will end up having a token, right?

Because if like some of them have it, they can incentivize developers and users. And others that don't have it can't, right? Because they don't hold that power. Right. So just from gaining users perspective, I think there will be a point when almost every L2 has a token. If we break this down, right, I mean, all the ZK protocols supported by Polygon, they will already integrate Matic into their offerings, right? I mean, there will be some sort of utility for Matic within their ecosystems. As far as optimistic proverbs are concerned, like optimism has announced there are rumors that Arbitrum might.

ZKSink in their FAQs has stated that there will be a token. I mean, there are no too waste to it. And because of this, I think everyone will be. forced to have a token. And if you look at the design, right, I mean, Matick already has utility, optimism they've stated what kind of utility the token will have. As far as ZK Sync is concerned,

in their ZK Porter module, where the data availability is off-chain, which will be maintained by some of the guardians. So that seems to be a proof-of-stake kind of a design wherein their own native token will come into the picture. We are yet to see how stockware, I don't think they have yet announced what sort of functionality the token will have. We are yet to see that.

But yeah, I mean, I don't think these tokens will be used as fee on their attitudes because unless there is some mechanism wherein the user doesn't have to care about holding both ETH as well as that token into their wallets to make transactions, right? Because that adds some sort of friction to the overall user experience to have like two tokens. 8th for porting the assets over to L2 and then whatever L2 native token for the L2 ecosystem. And another problem that it could have is I think a lot of these L2s may end up being interoperable, right? So you can port your assets from arbitral. You can already do that with bridges, but there might be some bridges with the ZK roll-ups as well.

So when you have tokens that are used to port fees, I think it might add some sort of friction to overall user experience. when users are taking tokens from one L2 to another. So yeah, I mean, that's what I think I think about the overall tokens. The one thing I just add is that like when we talk about the tokens, it kind of raises this bigger question of like, what is the relationship between these L2s and Ethereum layer one, right? Like a lot of them today are branded in this way of like, we're working to scale Ethereum.

We're an Ethereum scaling solution. But at the end of the day, like these users are going to be able to withdraw directly from centralized changes to these L2. And when an L2 has a token, everyone that's in that community is going to have incentives to kind of look out for those token holders, right? And then the question of like what happens on ETH in the back end, they could become less important over the longer term. You know, like, as of today, one of the things I read about in their report was that if you look at Starkware, like, they're settling all their transactions, like all the L2s post their data to L1, what they're doing in the future is they're going to take like their permission applications which are in like the Stark X environment. and they're going to use their Starknet as like the data availability layer. So instead of like going directly to Ethel1, they're going to go to their L2, which then goes to L1,

which definitely will like help them in terms of performance. But it kind of raises an interesting question of like, where will their allegiances lie in the long term? Right. The other thing we also kind of flagged was this. There was this little, it's like a small little grant program, but there was like a little research being conducted in terms of like, could you do something like this on Bitcoin? And it sounds like the consensus that like in its current state, like no, like Bitcoin is not amenable to verifying these kind of proofs. But it is an interesting consideration of like people are going to be able to decide where they want to settle their transactions that are executed on these L2s, you know.

And with the tokens like that, that could become more interesting and apparent to folks. That's fascinating. Do you think there's a world where we see the emergence of new L1s as a result of some of this density that's built on L2? I don't know a ton about it, but we did mention there that there are like kind of two networks that are already kind of doing this. I think you guys have put out some podcast. It's actually about like Celestia and I think Polygon has their own solution like a veil. But we didn't do a ton of work in the report on that.

So you guys hit on kind of a core question of mine around the tokens. And I see the idea of launching a token for customer acquisition to be able to essentially subsidized developers. So, Rob, as you point out, introducing a payment oriented token. just introduces friction that it would be the equivalent of every time you went to a Starbucks, you could only pay with Starbucks dollars and you had to redeem dollars at the door. It's just a terrible customer experience. But it sort of begs the question of, well, what other utility could a token have that actually

makes sense to have a token? Like, why have a token at all other than a way for a development team to monetize their work? So when you think about the principal utility functions of tokens on L2s, what have the most validity? Yeah, I can start with like the optimistic roll-up side. So like the way these systems kind of work is that they're like predicated on people submitting like claims that if there were something wrong that happened on the on the L2 that someone would pretty much submit something to L1 to kind of dispute the validity of that and to actually like drill down to what the correct state is at the end of the day. And right now all that is being done somewhat like implicitly. There's no like explicit financial incentive where like if this then that. If someone posits this and then I prove that it's wrong, I'm entitled to some sort of financial award. And it sounds like that's going to be a critical part of it on the optimistic roll-up side is like, how do you make sure that people are incentivized to actually monitor these things?

And then how do you actually, on the flip side of that, like penalize people who are, who could be trying to like kind of subvert the system? Various data availability solutions that ZK. Roll-ups might explore. They probably would want to embed, like, like they might embed that. token incentivization to maintain those solutions. So over and above what Andrew said, I think that could be one of the use cases for the token. And I mean, as far as the governance goes, right, I mean, we don't really know, at least

as of now, we haven't seen a successful case where defy token has been used as successful governance token to the intentions that were initially desired. And also, I mean, in terms of sharing revenues with the users, right, I mean, that could construct it as security and whatnot. So barring somewhat embedding the token into the data of liberty solutions, I don't know what other use cases a token may have over and above what Andrew said. Andrew, you hit on some of the tooling around having centralized exchanges, give the ability to offload into L2 directly.

How would you just grade the industry right now in terms of just the tooling around getting onto these L2s, building things on them, and where do you see that size? of the market going over the next couple of years? Yeah, I'd say it's still, and it's very early stages. Like anyone who's kind of used this will know that, like, one of the things we wrote about in the report was about how bridging is definitely like a big friction. I think that's still the case for many of these things. I haven't been keeping up with all the different exchange integrations, but when you have

to like go on to an exchange by Eith, bridge it over, and then with both of them, like with both optimistic NCK roll-ups, the withdrawal process is also very, very, just like annoying for users. So on the optimistic role of side, it's like a seven day window and they're obviously working on a ton of ways to like kind of bridge that and you have a bunch of like liquidity providers that are going to get you out at 99.

Whatever on the dollar today. But all that stuff is still like manually user initiated. Right. So like it's kind of hard to say now like what this could look like in like two, three years. But there's definitely like a ton of greenfield for these developers to go in and kind of clean up this experience. Because right now it is like quite frankly, much harder to interact with these applications.

That's like abstracting from the whole fact that like, we don't really know what the grand like L2 landscape is going to look like, right? Like all these different applications are going to be deployed on different L2s. And like, is there going to be a couple winners? Are they going to be kind of spread out all over the place? Because then you have to consider like, how are these apps on different chains going to interact with one another

over the longer term is kind of what I'd say. It kind of reminds me of the early days before, you know, Coinbase was ubiquitous across cryptocurrencies where it just was very difficult to get fiat currencies into the most basic, you know, top five coins. You sort of have the same issue right now in L2s. It's just very difficult for, you know, OLA person to use these things, is my sense. I think when we started writing the report, none of the major exchanges was allowing any withdrawals to endos, right? but by the time the report was released, I think Binance and whoopi, at least as far as I know,

allow you to withdraw your aid directly to arbitral. I don't think they have a bridge to optimism yet. But once you are on arbitrium, the decentralized bridges like hop, they allow you to move from arbitram to optimism for fees, which is like less than a dollar or so. So, yeah, I mean, that sort of thing. I think exchange withdrawals directly to L2s,

I think will be pretty much the norm in the next couple of years or years or so. So the seven-day withdrawal window, right, I mean, that too, if you're using their own native bridge, right, if you're using the optimism or arbitrum bridge, you take about seven days to exit L-2s. But, I mean, as Andrew said, right, there are these liquidity mechanisms which allow you to withdraw for a few dollars. So, yeah, I mean, in the last three, four months itself, I think the experience has improved a lot, and I don't see why it won't improve in the next couple of years. So as we wrap up here, I'd be curious, as you guys were writing this report, what

you were the most excited about just around the overall L2 ecosystem, or we had an into an L2 summer here? What's on the horizon and what gets you guys excited? Yeah, I just think how fast everyone's moving is probably the most exciting thing. Like, I think especially on the ZK front, like people, for a long time thought that this would take a really long time to get these virtual machines up and running. But if you look at something like what Starkware has already kind of accomplished, it's pretty amazing.

Like they poured over this D-Y-D-X application over and it's been very successful. And there's just a ton of momentum on the optimistic roll-up side too. Like the numbers are getting pretty substantial. Like there's still like one of the things we looked at was like the total value locked, like how much should I moved over. I think at the time we published the report, it was around like $7.5 billion, which is still pretty far below even like the $30 billion. that's been brought over to like other L1s from Ethereum.

But I mean, the numbers are trending up pretty quickly. So it should be really fun just to kind of see how this all plays out over the next couple months here, especially with kind of some of these token launches thrown in there. Hopefully they're not delayed too much given kind of like the market turbulence we're seeing right now. But it's hard to say. Yeah, well, will there be any buyers for the tokens, I guess? I'll see.

So Rob, how about you? What's the most exciting to you right now in the L3? two space. I think talking to all these teams and like hearing their perspective on what their mindset is, how they are making the design tradeoffs, because I think almost every protocol has made some sort of design tradeoff when they were thinking about scaling Ethereum. And another thing was, I mean, all these elements launched and everyone started saying that probably L2Z Ethereum may not make sense because these elements are already ready and

they have captured most of the users and so on. And I don't think many people were expecting ZK Rollabs to ship this quickly. And maybe when the market is hot again, these protocols will be ready to serve users, wherein users do not have to think about cash fees to make transactions. So yeah, I mean, that part was pretty interesting. Well, I think you guys did a great job of condensing a ton of information into a really easy to digest report. Where can we send people to learn more about the report and to follow your guys work. Sure. If you go on the blockcrypto.com, there's a report tab that has all of our kind of

sponsor content that's free for the public audience. So that's a great resource for for folks to go to. It's probably the best way. And then we're also on Twitter. We can we can share our handles, I guess, if that makes sense. That's great. Yeah, we put it in neat show notes. Well, thanks for joining us today on the podcast, guys. Really appreciate it. Thank you for having us. It's great. Thanks for listening to another episode of On the Brink with Castle Island. To find out more about Castle Island, visit castle island.V.C. To listen to all of our podcast episodes, please go to On the Brink-Podcast.com or just click on the tab in our website. Thanks for listening.