Are L2s Extensions of Ethereum? | Kyle Samani & Max Resnick vs. David Hoffman & Jill Gunter

I think the market demand for producing chains should be just taken as fact in this day and age we know that there's going to be many many many chains some chain is going to figure out how to aggregate all the chains some chain is going to figure out how to make these chains have stronger property rights and composability mechanisms between all these chains and that one chain is going to incentivize the production and aggregation of even further chains and that chain will become money bankless Nation for those of you

who did not make it to permissionless this year there are some pretty hot conversations that were held over those three days at the conference the spiciest conversation or debate I should say is what you're going to hear right now it's a debate as to whether or not layer twos are extensions or ethereum this debate is between Kyle Sani and Max Resnik teamed up on the side of no layer twos are not extensions of ethereum versus myself and Jill Gunter Jill Gunter from espresso on the side of yes they are extensions of ethereum some

people said that this was the panel of permission lists not my words this was a tweet out of the moderator Westy from blockworks it was also the last panel on the last day of permission list about four hours before my fight with Kane Warwick and I used this panel uh to get into war mode in order to prep for the fight and I definitely think that worked there were also lots of other super hot panels and conversations that were held throughout permission list Ryan's mainstage panel with Chris Dixon was super insightful I did defi main stage with Kane stonnie and Sam casman manian

from fra co-moderated by Mike Epolito and Ryan had an ethereum layer 2 conversation with Alex from zky Harry from arbitrum and rushi from movement Labs all of those conversations are getting released on the bank list premium podcast feed so let's go ahead and get right into the episode but first is a word from our sponsors yeah how's it going everyone I've been excited for this conversation for a while ever since we we set it in ston I'm wesy I'm on the blockworks research team uh most my reachers which been around ethereum infrastructure rollups and so this

question that we're debating today I've been talking about for a while and it feels like we there have been a ton of conversations on the Twitter verse on podcasts but I don't think we've ever gotten people with different views in one place at one time so I think this will definitely be super fun conversation hopefully we get some some hot takes uh as well and like I said end this uh conference uh with the bang here so I guess to start I'd love to have sort of opening statements if you will from everyone on the panel so we can get

a quick introduction of everyone and then if each person also wants to go into while I guess their high level thoughts or what they think of when they hear the question are l2's extension of ethereum so we'll start with Kyle here and make her way down sure hi everyone my name is Kyle samani I'm a co-founder at multi Capital we are seven years old investment firm focusing in crypto um and I do not believe l2s are in their current form as right the generally used

widely popular l2s such as arbitrum base optimism or not extensions of ethereum although there are other configurations that could be hi everyone I'm Max resnic I'm the head of research at special mechanisms group SMG is a Skunk Works organization within consensus which is the parent company at metamask and fura we do special projects we work with all sorts of things throughout and we particularly specialize in me transaction supply chain transaction Landing we also do a bunch of work on

the ethereum protocol and I agree with Kyle that l2s in their current form are not ethereum I love that you guys are already adding some Nuance to this debate uh my name is Jill I'm a co-founder at espresso systems we're working to solve cross-chain composability starting with l2s on ethereum so very relevant um and I believe that l2s are an extension of ethereum I think there's two ways to ask the question one is based on definitions the other way is to ask it based on the

goals of l2s versus the goals of ethereum and I think any way you slice it they're extensions hi I'm David uh one of the two bank list guys the other one's in this room somewhere um and I think layer 2os are ethereum are extensions of ethereum naturally uh and I think whether or not that that statement is true or false really depends on of course how we Define words uh all debates are essentially semantic debates I think ethereum all blockchains are property rights systems uh and

whether or not layer twos are extensions of ethereum uh is either true or not true based on how strong the layer one can promise assure your property rights on a layer two uh and so this is why we have things like layer 2 B uh the different stages uh fraud proofs uh ZK proofs uh the whole purpose of layer twos is to be a on toone extension of your property rights on ethereum uh and I think this debate is being had now because uh what what is not happening or

at least what is perceived by the market is not happening is that layer toos are not becoming extensions of e value AC cruel uh I think perhaps that is something that is worth maybe talking about or focusing on but at the very least that could also only be temporary amen yeah there there's a lot of threads you mentioned that I definitely want to go off of first we should certainly Define exactly what we're talking about here because I think that's the first question most people have is what does it mean to be an extension of a theorum or to be ethereum um and it sounds like

you you defined it David so I'm curious because uh the people sort of on the anti side here said in their current state um so I'm curious from this side uh what would you say like what properties would you say make something an extension that like maybe in the current state they're not but they could be at some point in the future I have a fairly simple itness test for this which is is the extension thing it's an extension if it's doing something that cannot function on Main net and that

doesn't just mean like hey increase the throughput of an app by like 10x or 100x like that's still functionally the the same thing but but if you're actually unlocking like new behavior um or or new use cases that that to me is an extension uh and explicitly therefore not a replacement the mainstream l2s that exist on ethereum today are explicitly replacing a function that L1 did a year or two ago would you agree yeah I would agree with that I think my test is also about whether these things

are their own siloed state with network effects and when you're your own siloed state with network effects you naturally compete with ethereum which is a siloed state with network effects so that's another thing that I would add to my definition and then I would just add like go look at what the things arbitrum da is spending money on and it's basically spending money to take away the lifeblood of apps that they're currently operating on ethereum L1 and another thing go look at the fact that

uni chain just announced that they're launching their own L2 so that's a hugely popular app the Unis swap router is one of the most commonly uh used highest gas usage contracts on Main net and they're saying hey we're going to go away from mainnet and make our own L2 which again has siloed state which is not interoperable with anything else today today they explicitly have on their road map interoperability and composability is a goal so are we are we taking uni chain as it exists or I'm not

even sure the state and it exist that it exists in today or are we taking l2s at the word of their road map so like according to your logic gel if we put like one million TPS on the road map does that mean that ethereum is scaling like I I just think we live you have a viable path to get there we've been saying ethereum is scaling for years I think you would agree with that I would not agree that ethereum has been scaling for years I don't think we've done any I guess because you see L2 is not as an extension of ethereum I want to address what what Kyle said though I mean I

think firstly we can simplify the definitions what is ethereum it's a world computer it's a platform for Innovation and what are l2s they're applications they're specialized form of application but they're applications and I would contest the statement that l2s are only extensions of ethereum if you're doing something you fundamentally can't do on the main net I'm not sure that I really understand that um but I would say that regardless of that definition you are doing something on

l2s that you can't do on the main net you have a totally different Dynamic around transaction costs throughput Etc I'm surprised to hear that statement coming from someone who's such such a big backer of salana because what would be the use case of salana what would be uh the reason for a developer to go over to salana if that was not by your own definition a fundamentally different experience than you could get on ethereum that's what l2s are offering to developers today salana competes

explicitly with l2s so like I I agree with that but the the frame of the question is not base versus salana it's base versus ethereum yeah but you you're getting a fun why would any developer choose between these things why wouldn't they just go to eth mainnet you know as of a few years ago which was where all of the activity all of the users all of the liquidity was because they're getting a fundamentally different experience on salana and today because they get that fundamentally different experience via an L2 which then by your definition is an extension how how can

you have a fundamentally different experience and an extension those two are contradictory statements if it is fundamentally different it is not the same that's what I'm working off of Kyle's definition so you'll have to ask him well like Kyle is saying that he is not afraid to say that salana is competing with a the and I'm not afraid to say that and I'm I'm also saying that l2s can be competitive with the ethereum L1 as well and there are things that I would like to see on the L1 that are only possible on the L1 like censorship resistance and credible NE neutrality

that you cannot expect to see from a US entity centralized sequencer and you know you want to see the Canadian truckers be able to use the chain when they want to get their money out of their account then you probably shouldn't rely on a regulated centralized sequencer on us east1 or on some box somewhere in a westernized country you need more than one box contributing you need decentralized sequencing and you cannot basically

provide the same guarantees eth L1 provides without doing that I think a large part of this conversation uh comes down to the competitive competitiveness of Layer Two Chains to the host chain like the the layer one chain uh when base grows does it subtract from ethereum um which is yes probably the answer is yes it also subtracts from optimism it also subtracts from ZK sync it also subtracts from all chains actually actually there's no really difference between when arbitrum winds

it's pulling away from salana too uh and so there has been this ongoing uh development architecture in crypto even with and without ethereum you see this in Cosmos where you have a variety of chains with like no real ropes between them no real connections uh you have polka dot with many many chains with like very rigid connections between them and that really didn't take off and now you have ethereum with like what I think you guys are saying very loose connections but connections nonetheless strong property rights are uh continued

across chains and at least when an ethereum layer 2 wins it takes away from salana more than it takes away from the aggregate of ethereum because at least it keeps it inside of the ethereum ecosystem so when you can keep value inside of etherium uh yes the layer 2 the rollup Centric road map is designed to have all chains be competitive with other chains inside the ethereum Centric road map but at least there's some aggregate value created that's retained by the net ethereum ecosystem that is pulled away from the very disconnected

layer one landscape okay but let's let's kind of take this view so so is Avalanche which is an ebm chain L1 part of the ethereum ecosystem by this logic is monad part of the ethereum ecosystem do you have to just like do you just automatically become part of the ethereum ecosystem by posting State rates or do you have to post use use ethereum for da there's no clear line that there's no clear line it's a set of concentric circles uh but I think there is a clear line as to where property rights are given assurances to uh and I think where uh property rights

insurances stop is somewhere like Avalanche an evm chain manad and then way further away is salana a lot of great takes going on right now um curious your thoughts on sort of interop like let's say ethereum and all the RS agreed on one inop standard that there was good ux to basically easily switch assets from one rollup to another to etherum L1 and you have this Co cohesive ecosystem like you did salana across these rups would that change your guys' mind like would that be like would l2s

in that case like sort of be ethereum and be part of one ecosystem or would that not matter at all I mean it's like what what does it mean to to be ethereum like David you can say are property rights you know functionally equivalent and like okay that that's one element um another element is do normal people perceive it as as the same another would be like is there some general sense of um credible neutrality uh another would be um if you want to get late like transaction latency for inclusions

whether you live in India or the Philippines or in Mexico like what what what are those guarantees um so there's like different properties of of the systems um oh and then and maybe the other one of like again answering the question is X ethereum uh you know the most obvious one in my opinion is like is it pring meev value to to the base asset um and so there's different constructions of various RAB looking things whether they're espresso or arbitrum or Bass or

uni chain or base rollups or whatever and and all of these systems have different versions of some some of these properties um I think in MA Max's in my definition none of the proposed systems by any of these groups are functionally equivalent to ethereum they may have partial equivalence on some properties um but none of them are 100% functionally equivalent by definition if that was the case then I wouldn't be on stage right now like there would be no debate but like very obviously they're

not functionally equivalent yeah and I guess like just to come back to our opening statement it's like in its current form today is exactly what we are talking about we are not talking about what is on the road map because we work in an industry where road maps are memes okay road maps do not exist ship code the code that's shipped is what the thing is not the code that you promise to ship in the future uh words on a page are not code on the compiler but ethereum has had a pretty coherent Road mapap they had the pivot like 2019 to the rooll of centric road map but

nonetheless there was some true nor star that it kept uh when it pivoted to the rooll of centric road map there was one Central North Star we just decided to go get there a different way and so anyone who's like investing in crypto based on current conditions is not going to make it uh like things are going to change in the in the future uh I think ethereum's on it was on his roloc Centric road map in 2021 when Sal has its big rise as we realize that the layer one is congest congested and we really only had polygon as a side chain to ethereum with no actual strong commitments of property

rights connecting there uh but now we have uh and so this was the era of many independent siloed wal Garden chains blossoming out of ethereum proving the use case of or the idea of a successful rollup Centric road map uh the PO the pokka dot model the cosmos model but this time actually working and thriving uh and so we had the growth of many independent Silo chains now we have what is I think the era of the cross-hatching between these chains uh we have Jill from espresso doing some of that work here uh that is one Frontier of the

composability of networks but this one canonical bridge is the first step to producing composability but then the rest of the shape gets filled in by many different Technologies and I think that is going to define the next two to three years of ethereum I I just want to build on that to come back to this question of what are the goals of ethereum and what are the goals of l2s I think we've named three different goals throughout this conversation so far one would be to create censorship resistance applications now that's not the goal of

every L2 certainly the goal of some l2s there are l2s that include pornado cach transactions there are l2s building explicitly for even more kind of cyer punk values than exist on ethereum today uh on the ethereum main net um and so I would say that on that front the goals are aligned and compatible between the main net and the L1 uh and the l2s at some of the l2s building upon it a second goal that we've named now is composability I think censorship

resistance and composability are basically the two reasons that you have as a developer to build in web 3 or as a user to use web 3 apps composability to me is honestly still a big question mark that is where I think there is a risk that l2s go off to build their own wall Gardens that is why I'm working on espresso to work to solve composability and then there's a third question which you mentioned earlier that I think we've all been skating around which is the question of value acrel and that is the

real reason why everyone is sweating over this is because I think Bitcoin salana are each up like 40% in the last few months and ethereum that Meme where you're poking it with a stick saying do something I think the in the current state of ethereum we do have kind of this like blank step of and then and then there's a perfectly composable ethereum and there's like a missing step there but nonetheless we do see signs of internal composability across uh what Al glowski is calling the Clusters uh the super chain the elastic chain uh there

are it's a pretty coherent yeah or orbits yeah there's a pretty coherent path for for having internal composability across these systems uh and once the super chain once the elastic chain fixes internal composability of which they are highly motivated to do and they have the war chest to do the R&D to do that then we learn a bunch of lessons to apply it's like what worked and what didn't didn't we have like five different of these Frameworks all tackling this problem maybe there's five different solutions maybe there's one different solution we're going to find it and we're going to export those Lessons Learned to fill

in the gaps between the Frameworks between the arbitrum orbit chains and the super chain and that will take time that will take a lot of time but that is the path that we're on and I don't think that that's look like a crazy incoherent uh vision of the future it's not defined like yes everything is possible sure you didn't describe violating laws of physics so like okay it's possible but like that that's not helps like like you're just saying I'm not proposing breaking law physics I mean there's another thing you said David which is that it didn't work for Cosmos and it didn't work for hulot this time is

different man we have but we actually have the chains now like it is currently working I you know that like Arrested Development meme I mean first of all we'll see what happens when the liquidity rewards dry up on arbitrum and the the these War chests run out we'll see if it actually works when the tide comes out we'll see Max what would it have to look like for l2s to be an extension of e theory what about okay also the other thing David said was that we've been consistent on the road map but then the first thing he said was actually we shifted the road map from sharding and if it was basically

sharding I think that would be like pretty pretty consistent with being ethereum but it's not charting it's a separate team building a separate system which is not composable with separate economics has a separate economics and its own token that it's Distributing to its team that does not depend on on eth at all and it's it's making money for itself and for its team members and for its token holders and not for eat yeah I definitely want to go there to the economics and value cruel which we have been skating around a little bit so I

guess that brings about the question that I hear a lot which is rl2 is parasitic uh to the L1 so I'll poee that question to Kyle in a counterfactual world in which the L1 was otherwise scaling and sequencing yes um if that is not your counterfactual world then no but like that that's kind of like implicit like the question implicitly is being asked against that counterfactual otherwise you just say we're scaling the magical thing but like you need to position it against something would you guys say that they

Al use a parasitic symbiotic yes they are parasitic to ethereum they are also net net generative of value inside of the ethereum ecosystem and so when arbitrum attracts the marginal user they Attract it to the ethereum ecosystem and by the way the dominant asset on arbitrum in all of the major l1s is eth so there's demand in these ecosystems for eth and there's Pathways to exponentially grow the user base of ethereum from what we've seen anytime

historically on any chain and that transaction volume and uh the the growth of the user base the growth of holders of be including long term holders of e who will probably still do that on the L1 all of that is value acreative there there's a the rooll of centric road map will succeed or fail based on the tension between defection of chains away from ethereum or coordination incentives towards ethereum there's needs to be a Harmony there a yin and a Yang of like we need to allow enough uh flexibility

enough tolerance for these layer tws to issue their own tokens to grow their own wealth to provide their own utility but enough incentives for that value generation to stay inside of the ethereum ecosystem and this is actually in start contrast to interestingly both Bitcoin and salana which are like kind of like monolithic chains where Bitcoin is like don't build any value outside of BTC the asset prune everything away from Bitcoin to focus on BTC the asset and salana is the same way which is in my mind net stunting is a ceiling on the total value uh because you're actually

being very rigid on the net ecosystems that can be born and so whether or not the roip and trick road map succeeds is going to be because it allowed systems to be parasitic of the layer one yet still be contained inside of the broader ecosystem I think you're touching on something really important which is incentives there's a reality that developers are incentivized to launch their own chains and so to the point of counterfactuals was coinbase going to launch base on salana as an application

if they didn't build it on ethereum or were they going to launch it as their own chain I I'm not part of the base team we can go ask Jesse afterwards but I think in that counterfactual world they would have launched it as their own chain instead of keeping it within the ethereum family which l2s have enabled them to do so I guess I want to clarify my position which is I do believe that l2s are bad for cryptocurrency I just believe that they are parasitic to ethereum in their current form and that do you mean do you mean that they're net

parasitic to ethereum as in they get they take away more than they give back I think they're currently taking users and giving back nothing so I think like from a like how much are they taking away well that is a positive thing and then how much are they giving back well it's like one cent a month on blobs I think that's uh what about their reservation demand of like how much eth is on arbitrum like I I actually don't know looked it 100 100 like yeah uhuh quite a large supply of eth on arbitrum what what value does that provide to the eem it would have been on eth if it

wasn't bridge to arbitrum sorry it would have still been on eth if it Bridge tar would it yeah but it2 have grown the usage of these networks I'm saying that we should have the counterfactual is we could have just scaled the L1 that's a road map and layer layer twos are the largest buyers that will ever exist of eth because of how big the layer twos are going to get l2s are not buyers of eth what what no users the deposit contracts as net

buyers of e that that's not a buy that's a move it would have been on the it was already on L1 by definition previously that is the only place it could have ever come from it was already there the the net value of these ecosystems the value that they provide to the world is in addition to the ethereum ecosystem and that value is creating reservation demand for eth asset so the value is in addition to the ethere I thought they were part of the ethereum ecosystem for for all that you guys have said in this panel so far this debate so far that we

can't talk about road maps road maps don't exist your counterfactual depends on a scaling road map for ethereum that has never existed your counterfactual is that all of that demand for for eth would exist on the eth L1 if it were not for arbitr I I actually I agree with that statement I don't I don't think there was ever a plan to scale L1 I mean they talked about charting for a while uh I don't think there was ever a serious design skill issue um yeah I I agree with that but like that is the

counteract to compare against like it's just like it's the obvious counterfactual but that never existed again it's called salana it does exist we are using would Bas have built on been built on salana bass is just just it's literally a fork of Optimus it's the same [ __ ] it's not an application at all it's another L l2s are forms of applications no they're not okay it's literally and you copy and you take the same Unis swap contract and you copy paste it there if you just copy paste the same [ __ ] 10 more times like it an

application that's like half the applications in crypto but if I'm coinbase and I want a way to bring users on chain which as I understand is the fundamental goal of Bas what am I going to do am I going to build my own chain build an L2 on ethereum or go deploy an app on salana I'm I'm gonna either build a chain or build an L2 none of the above you can just be a broker and you can coin look coinbase today you have coinbase the UI whether it's the retail app or the Pro app or prime app or whatever and you've got a database on Amazon and like the the front end

happens to point at the back end obviously cuz coinbase owns both of them um and like the back end is there but the front end could point at any other place and that other place or other places it could point to could include ethereum mainnet it could include Aram it include salana mainnet it include [ __ ] Avalanche who cares and like so saying that like they needed base to exist is incorrect they already have the users they have the kyc they have the apps on people's phones they can just tap into liquidity anywhere the basis is by definition unnecessary I mean they literally forked optimism like they

don't it doesn't have to exist at all can someone Let Brian know let Brian Armstrong know yeah actually I would love to chat seriously so one thing I I do want to get back to is you guys mentioned sort of like eat the asset and one thing yeah like with this relationship between l2s and L1 whether they're parasitic symbiotic sort of all comes down to to eat the asset because that sort of is the exchange right like l2s are paying the L1 to post their data and in return they're sort of allowing e to proliferate on these l2s so I guess I

I want to get your your guys' take on sort of eth the asset is it enough uh to have it on these l2s is eth money on these l2s like where's the like actual value crew for the elong from E I think you're gonna have a very hard time in this conversation because Kyle doesn't believe in money so there's a very easy limit test for this so so the only assets at scale that are quoted in eth or in Soul are nfts nfts in fact on no chain are quoted in in dollars they're only quoted in Native assets but like nfts today are down 907 8 9% something

like that and if you look at liquidity pools the vast majority of quote pairs across both CI exchanges and defi are against dollars there are a couple of pools that are e nominated like it's not zero but I bet you across all e L1 and E l2s I bet you e quotes as a percentage of LP volume percentage of um t and as presenta of actual volumes I'm getting guessing is on the order of 2 or 3% the number one pair on ethereum L1 is eth usdc the number two pair is e USD so in

this industry we're going to just settle for the United States dollar as the unit of account because that's not the Cyber fun future I'm looking forward to what about the fiat currency crisis that has underpinned this entire industry for the last decade sorry what crisis the fiat currency crisis what what crisis what's the crisis the one that that we keep on printing all of them how we print did $16 trillion in 2020 what what what is the I understand you described the set of facts but what is the crisis the devaluation of Fiat what's the crisis people being able to store their wealth soore and maintain their wealth dud

there's an infinite number of Assets in the world you can own saying like last four years Kyle huh if we Haden you can buy real estate in New York City it will go up you can buy real estate in Dubai it will go up you can buy gold gold you can buy silver look I think the real thing that people have consternation about when it comes to value acral and the economics Etc let's just name the elephant in the room it's that l2s have tokens too that's all

there is to it and fundamentally I think that this is a question of whether you think that on balance l2s being on ethereum versus whatever counterfactual you want to pick even with them having their own tokens is that going to be a net positive or net negative to ethereum ethereums whatever you want want to call it business model protocol model Etc I don't see a case why the pi can't grow and those tokens and those tokens having

value can't coexist with ethereum's value growing granted that is not what the data shows over the last several months and so this is a forward-looking statement and maybe you know my counterparts here will say no road maps no future but I think that there is a path for l2s being value creative to the eth asset even while they have their own tokens I think that arguing that eth is money in 2024 or 2022 2024 is harder