The (Anti) Superchain Thesis | Cooper Kunz - Aztec Labs

thank you thank you uh can everyone hear me okay who's in this General vicinity you can hear me back there cool um my name is Cooper I'm on the product team at Aztec Labs uh this is called the anti- super chain thesis uh go ahead and plug my Twitter you can follow me at Cooper Coons um also just caveat these are obviously my own opinions they shouldn't reflect my colleagues Aztec Labs Aztec the network uh if I say anything stupid it is definitely on me and not on them um so we'll kick it off what is a super chain I'm sure many people here are familiar with it I'm sure it's kind of a new terminology to some others um but we'll just Define the super chain we'll start with optimism they originally coin this as far as I know uh so we'll just use their definitions for now uh super chains can be considered a network of chains that share bridging decentralized governance upgrades a communication layer and more uh from their perspective you know they would say say a super chain is all built using the op stack uh later in their documentation they also Define it as you know a decentralized blockchain platform which consists of many chains that share security and a technology stack uh interoperability and standardization is you know key things that they articulate uh you'll notice one of these definitions has decentralized blockchain and one of them has decentralized governance uh those are kind of intentionally different definitions um and so for my take for this presentation I'm just going to steal optimism's definitions and remove the fact that you have to be using their software stack uh that's something that's convenient for their marketing but you can consider this to be the same uh you know in a variety of architectures uh so the for the sake of my talk a super chain is any network of chains that share bridging decentralized governance upgrades a communication layer and you know various pieces of their architecture um alternatively you can steal their other definition again and say it's a decentralized set of blockchains that share Security in a tech stack uh I would again say it does not have to be an OP stack uh and so we'll talk a little bit more about this today um and so this presentation uh contrary to its title is not you know exclusively about optimism or the op stack it's about any L2 that specifically is trying to build products that satisfy these kind of definitions um any you know variety of l2s have business models that are building in this direction and so we'll kind of talk about that um and so this is what the op stack looks like you know this is from their docs they have a governance layer a settlement layer an execution layer kind of a derivation layer I don't know exactly what that means but I'm sure someone from optimism can tell me um a sequencing layer and a data availability layer um some of these are currently implemented in the op stack some of them are proposed improvements to the op stack um notably you can see that within the settlement layer optimism is departing from optimistic rollups and proposing to integrate zero knowledge proofs uh which is a big win for anyone who's been working on zero knowledge proofs to the Past you know five or seven years like Aztec has um and so this is what you know a full stack deployment of the op stack looks like you can take these boxes and you can look at optimism and you can look at coinbase and uh coinbase's Bas L2 and they're using the same various pieces of this architecture um but the super train would look a little bit more like this so you could see you have you know n numbers of settlement layers execution layers and derivation layers uh they would share governance and they would share sequencing and standards around data availability um but you can have different types of execution environments finality terms Etc uh but you know the key thing is that these L2 super chains share governance they share some type of sequencing proof aggregation or data availability um and so this is you know my take on kind of how you can think of a mental model for the op stack there's one set of governance there's kind of one set of standards for data availability there's one set of sequencers usually uh but you can kind of mix and match other pieces in the architecture uh kind of aligned with like the modular thesis um and so this is you know a good mental model here um and so you can take this mental model and you can look at other l2s are there any other l2s that offer similar technical architectures technical Stacks or products um you could look at ZK syncs architecture they call these things hyper chains but at the end of the day it's a bunch of l2s and l3s that share standards around data availability sequencing and governance that can have different various pieces of modular components throughout them and so I would argue throughout this talk that ZK syns hyper chain is you know just one manifestation of a super chain um you could also through if you squint hard enough you could look at polygon's aggregation layer polygon's aggregation layer doesn't necessarily try to force shared sequencing on people but it does uh try to force shared finality and proof aggregation on people giving you similar properties of atomic inclusion and execution and so you could you know if you squint hard enough polygon's aggregation layer will probably share some type of sequencing some type of data availability and some type of shared governance and so therefore I would argue that this is again just a different manifestation of a super chain um you have to squint a little harder to look at arbitrum orbit stack they don't actually formally share governance and their business model is a little different uh if you deploy an L3 that uses arbitrum water Nova you actually don't have to pay rent on it but the other ones kind of do so it's a little bit of you have to squint a little bit harder but at the end of the day these are a bunch of 2s and l3s sharing a technology stack sharing sequencing sharing data availability and sharing some form of governance um and so you can see that for three of the most popular l2s right now these are all multi-billion dollar projects they are all betting on this kind of super chain thesis right they're all building in a direction that uses a bunch of l2s a bunch of l3s shared sequencing or finality and shared governance uh and this is a spicy take so this is my personal guess for why this is the winning business model uh building protocols is hard you have to have a really unique product offering to get differentiation building protocols as a service is way easier you can just punt that down the road and hope that they get adoption themselves um I'm going to go out and say that I think a lot of l2s don't see what they're building as public goods they see them as profal business Endeavors and they want to make money from them um a lot of l2's Founders are coming from web 2 and they want to license their software and make money from it that's the way that it works and and uh maybe the spiciest take is that a lot of these people are trying to play Infinite hopium games with their Community uh all it takes is just one more L3 and all of our problems will be solved that one's going to be amazing that one's going to get millions or billions of users and they're going to solve the problems for us and we will extract value for the right of them using our open source software um and so this is my personal take on why this is a popular business model for l2s today um it is fueled on hopium and so how do super chains differentiate right if they're all kind of building the same model right they want everyone to use their text stack share sequencing share governance um the current landscape they're all differentiating on speeds how fast you have finality and inclusion the cost of transactions is like a huge argument in L2 is everyone's competing to having the cheapest proving system these days uh ecosystem and developer experience is not still you know on parity throughout all of them some support different pre- compiles or different features um eventually I do think that they will start to battle on customizability so your ability to bring your own sequencing set or proving system on governance surface so can you opt in or opt out of different portions of governance or reduce your governance surface where you choose to do so um and how good are they at uh doing let's say business development or paying people to be part of their ecosystem and build their ecosystem around let's say one of these super chains versus the other and so my guess is that like how transparent evm blockchains differentiate right this is a majority of them but you can see one access is costs one access is safety or security as soon as you're above some threshold of speed and safety and costs um let's say you're average transaction is cheaper than a penny uh no one really cares after it's that cheaper than a scent right no one uses pennies anyways and after you're some above some threshold where you're no longer worried about your user funds getting stolen uh you'd be in this upper right quadrant and I would argue that if you're above some threshold of safety and costs you'll be completely differentiated on your brand your ability to do business development and have sales cycles and I think that you know polygon and base have been wonderful examples of this they are incredibly good at Business Development um and so I think right now people are debating about where they land on these axises on this grid for L2 Services eventually everyone will be cheap enough and saf enough that you will have to differentiate on your brand your ecosystem and otherwise uh you can also consider a trade-off space of sovereign to non-sovereign you know for Aztec having a governance minimization is a priority which we'll talk about in a second but you can see that like if you want to be as Sovereign as possible you're not lant on any external governance systems you should probably just build an ALT L1 uh if you want to validate your proofs on ethereum but have your own da or something like that build a sovereign rollup we think that the ideal space is kind of in this ethereum aligned L2 sequencer space uh Aztec plans on having its own sequencer Network and potentially a token uh but still using ethereum for da proof validity and other things um and then I would argue that no matter which of these chains you build with your sovereignty will always be minimized opt into some type of super chain uh you'll have to by definition be opting into the governance of that super chain if you did not like the governance of that super chain you probably would not pick it because that's one of the only things that they're differentia on um and so you know if you don't care about your sovereignty and you're happy picking backing on you know other people's success like super chains are probably a great option for you if you're not a very well-known project if you're having problems bootstrapping your own Economic Security or otherwise they're probably really good options for people but if you can differentiate if you can build good Services if you can get people to uniquely care about you you probably care more about your sovereignty and so yeah what are most super chains eventually going to be differentiated on right now it's cost and speed eventually it'll be brand and reputation and their ability to effectively bribe people or do traditional Business Development uh how many grants you have and how good you are giving out grants honestly will be uh the thing that wins the Super chain business model um and so this is actually the one of the last slides I have I know this is a short talk but this is actually the formal uh restatement of my anti-u chain thesis uh I think that there will be less than 25 meaningful rollups or layer tws you'll be able to debate any of these and this is controversial um I think that less than 25 projects will consume specifically more than 90% of ethereum Da or relevant resources um I think that these Pro projects these top 25 will want to differentiate and not necessarily compromise or collaborate with others I think meaningful projects may be okay paying rent initially but they will not be okay paying rent forever in the term of soft sofware license fee splits or Mev shares um yeah through centralized sequence or fees Mev share or otherwise and they will have a strong desire to internalize value capture however possible uh again if you're one of the top 25 projects meaningful projects do not want to lose sovereignty these top 25 projects should be probably as far to the left of that Spectrum as possible um they do not want to inherit additional governance or Surface risks um and this is maybe one of the most key points by the time super chains really get prioritized in engineering road maps I actually think that there will be better architectures I think there will be better solutions for those projects that are not top 25 and not able to be Sovereign themselves um specifically I'm kind of bullish on the ethereum shared sequencer and the ethereum pre-confirmation calls that Justin Drake and others are talking about right now um the ethereum shared sequencer actually gives you a lot of the strong properties in term including Atomic inclusion uh shared liquidity fra M mation reduction Etc that the super chains are promising the difference between using the ethereum shared sequencer uh is that you're ethereum aligned rather than optimism aligned or aligned with a given L2 so I think by the time L2 teams actually prioritize this in their engineering road map uh it'll be a slightly outdated architecture um and then I think the last point is that those that do opt into being a part of a super chain if they are successful will eventually opt out and leave whenever they reach critical mass or the ability to do do so uh at a certain point it no longer makes sense to give someone else 15% of your uh sequencer fees uh for code that's already been open sourced for a decade or something like that um so this is the the formal definition and I think anyone could debate me on you know any one of these five points and I'm happy to do so um uh and it's 100% possible this is me right now right like I'm like people barely only want two to three evm chains why why do we need hundreds or thousands or millions of them right like uh it's possible I'm mid curving the hell out of this I don't know um but let's look at Aztec Aztec is a differentiated rollup at least in our opinion and let's talk through some of our design decisions uh Aztec has a fully programmable Aztec virtual machine the AVM uh it's brand new no one else has anything like it uh it's a decentralized layer too anyone will be able to run a sequencer anyone will be able to run approver it's privacy first what we mean is we have native account abstraction we have a lot of fancy zero knowledge and client side proofs all accounts on Aztec are private smart contracts by default and your account contracts can actually be encrypted uh so no one can see your signing schemes or otherwise um we have 100 million in funding from Top teams including andreon Paradigm varant and a bunch of others so we're well capitalized our priorities right now include privacy that's always going to be our top priority safety security credible neutrality and governance minimization if these are things you care about I hope you like as Tech and what we're building uh if you don't let us know why and we'll fix them our priorities right now don't yet include being the fastest we're not trying to be the cheapest we are not trying to have a rent seeking business model we don't really know how we're going to make a bunch of money uh that's not what we're prioritizing right now what we're prioritizing is bringing privacy to ethereum in incredibly neutral way uh and we currently have no intentions to launch a super chain or a hyper chain or whatever I would say if you look at l2s within our relative market cap that's a pretty contrary intake um so you know we have to ask the question should Aztec launch a super chain are we like mid curving the hell out of this like should we launch something with shared sequencing in a ZK evm or whatever else we can dream up should Aztec deploy Aztec as an L3 to other l2s um and provide you know a shared sequencer among them um building a centralized cheap and fast network is honestly way way easier than what we're trying to do today um I think we would be really really good at if we wanted to um and if we don't do this are we mid curving L2 business models we talk about this all the time um obviously I don't think so I think our strong commitment to privacy credible neutrality and shipping something no one's seen before should win out in the long term but it's quite possible that we're wrong um and so yeah you should start building something new on Aztec today something no one's seen um we have a really easy tutorial within 30 minutes you can build private onchain voting uh deploy it to a local node uh you can join our developer community and we aim to have a test net soon uh so with that hopefully you understand uh a little bit about super chains and why Aztec doesn't have one uh but maybe someone will convince us that we're wrong and we should pivot I don't know um I have four minutes so I'm happy to answer questions if that's allowed I'm getting a nod does anyone have any questions cool uh well thank you everyone for your time I'm actually going to be on a panel beond the base layer after this uh with Mike neuter at the ethereum foundation and Porter Adams at zky uh so if you want to chat head over there and I will be hanging out there all night thank you [Applause] all