How Arbitrum is Pushing the Limits | Steven Goldfeder

Transcript

our goal here by the way there are a lot of interesting people working on a lot of really good people working on this EF is working this we just want we just want to bias towards action right I don't want to spend a year writing the absolute best spec and then another year implementing that because I think so much is here today I want to I want to say hey this is what works today we're building this right now which is what we're doing and we can continue to analyze it and continue to improve it and I want to be in all those conversations and calls and if in a year or two we have something better let's swap that in but this is a solution that that this is a problem that needs a solution today nothing said on zerox Research is a recommendation to buy or sell Securities

or tokens this podcast is for informational purposes only and any views expressed by anyone on the show are solely our opinions not Financial advice bachio Ryan and our guest May hold positions in the company's funds or projects discussed what's up guys welcome back to another episode of zerx research today I'm joined by Steven goldfader the co-founder and CEO of offchain labs thanks for coming on the show Stephen great to be here thanks for having me me um I think we should probably arbitrum is massive and everybody knows about you guys so I think I'll try and keep it as focused as possible and I want to talk about time boost interoperability and where arbitrum stands today um but before we get started it might be a bit helpful to get a short overview of what you guys

have been working at offchain Labs over the past few months absolutely so I would say there are broadly three pillars so um we obviously continue to contribute heavily to arbitrum and arbit development um some of the biggest new techical development uh developments there are arbitr from stylus which gives developers the ability to write smart contracts in Rust and c and C++ uh right alongside and fully interoperable with evm so you can have say a single application that's a mixture of you know solidity evm and and stylus code um number two is bold uh this recently shipped on mayet this uh was the is the new challenge protocol which among other

things uh opened up Val validation and arbitr chains for everyone so now anyone can be a validator fully permissionless of arbit chains which is a very very big upgrade when it come to comes to a security perspective because ultimately the validators are the one providing Securities and how anybody can participate in that uh another one is time boost um which is a new transaction order po ordering policy which tries to um sort of um give everyone what they want so appeal to both people who just want fast block times on the Chain but also internalize some of the MV value as well so that's uh basically on the arbiton side technical some of the big things we've been working on we also built prism which is the leading layer

one consensus client for ethereum um a lot of interesting things going on there a lot of P Das development um looking to increase the blob count so a lot of uh very interesting layer one development and a particular layer one development that's going to continue to help ethereum be the platform for successful layer twos right so so there's a an interesting merger of these two threads where U increasingly a lot of the layer one development is focused on um you know the rooll up Centric road map itself and becoming uh the dominant platform for ler tws and we've also been working a lot on interoperability ethereum is well particularly if you ask someone outside the ecosystem they tell you ethereum is fragmented today um I think a lot of it is is a narrative problem but there's also obviously some

technical truth to it ultimately I think it's a good thing um because it means there's a lot of different people developing different things and trying and experimenting right and the other side of that coin is things are a bit fragmented so the question is how can we build the best ethereum user experience possible not just the developer experience but also the user experience from onboarding to uh interacting across multiple chains in the ethereum ecosystem how do we make that seamless and intuitive and we've been doing a lot of work on that at off train Labs as well fantastic um I want to talk about all of those but I think the main thing um to get us started would be time boost um can you walk us through what the time boost um upgrade will be the proposal is and what exactly me it wants to achieve

um yes and by the way it's it's it's well timed this just went on live went live on tally which is the onchain voting platform so it's uh you know uh being actively voted on in the Dow in the final stages now right so started off a snapshot get approved and now it's actually in the onchain voting process so definitely a very timely time to discuss this is something the Dow is waiting right now and yeah so what time boost is you know it's funny because you actually take a step back there's always been this push it says decentralize the sequencer and I'm a big fan of decentralizing the sequencer but I think what people don't realize always is decentralizing the sequencer doesn't mean anything in a vacuum because you have to ask the next question is what do you want the sequencer to do or what

policy would you like the decentralized sequencer to enforce right so there's a sequencer which sequences transactions According to some policy and those policies can be en force in a centralized or decentralized matter and depending what the policy is you know the decentralization path looks very different so today let's focus on on the policy and there broadly there are you know before time boost there have been two broad classes of policies among sequencers the orb sequencer still today and always has done a first come first serve ordering policy which basically means you send transactions to it and it publishes the transactions in the order that it sees them there's another policy all the way on the other end end of the extreme which says so so sorry on the arbitrum side the first come first serve tries to

enforce some order of fairness right so those who come first and there will be those that argue this is not fair but ultimately you know it's a very very um basic concept of fairness that the the transactions should be processed in the orders that order that they arrived on the other hand you have people that argue for systems called what we call me meev auctions or mea which you may have seen over time which is the idea is no don't prioritize the order that they came in give uh the first spot to the highest bidder right so maximize the value for the tra for the chain ma maximize the value capture for the chain um for maximize the value capture for the chain in in ordering transactions and your average user you know a user a

regular user who just wants to pay say a little bit and doesn't actually want you know to pay a lot might see that his transactions get delayed or deprioritized and these are like sort of you know the two bookends here so on the one hand you have the thing that's maximally fair I would say to your average user they're going to get their transaction in quickly without delay and by the way that's one of the things people really like about arbit today that it's fast it's efficient and you know you have really fast block times Arbitron block times in the public network are 250 Mill 250 milliseconds which is basically the fastest blockchain out there you have other uh blockchains today that are looking and trying you're building all sorts of pre-confirmation to try to get the block times this low but arbitrum natively supports uh 250 millisecond

confirmations and actually even faster you have orbit chains which are arbitr chains that are uh others are running using the same technology stack that actually go down to 100 milliseconds a good example of that it's a chain called rehea that does that so the technology is really stable um at very very low block times on the other hand though you have the me auctions which is like great for internalizing me value because what happens is so so so what's the problem with the first come first served but the problem is there are still MV Searchers in the chain and they really want to get you know their their transactions in quickly so what do they do they try to shave off the time to the sequencer and they end up paying a whole lot of money exter internally to the system say to you know uh web or Cloud providers or

try to buy you know uh computing power in particular locations really to try to shave off uh milliseconds to get to the sequence more quickly so with today all that value is being externalized right all the who's capturing the value from MV it's not that it's not happening it's that um there are just paying others maybe they're paying their cloud provider maybe they're paying someone else so they're capturing whatever value they capture themselves and it's all going outside the system the chain is not capturing M auction the chain is actually capturing that value right your average user transaction might get delayed for a long time but the chain is going to actually capture the value of the M because if I want to buy priority I'm not going to go ahead and try to you know get better positioning and shave

off some milliseconds by getting a better connection Cloud connection or getting a better Cloud positioning I'm just going to you know bid dirting the auction and that's going to be my best my best path to to do that and the beneficiary of that is is the chain so the question is how can you like have the best of both worlds and that's what time boost tries to really achieve the last technical piece before we get into the time boost solution is uh another future of arbitrum today which uh will remain and that is that it has a private mempool right so the idea is that your transactions aren't publicly broadcast uh for others to see before they're actually included they're privately sent to the sequencer and the sequencer publishes it so if I'm looking at your transactions the very first time I know

of your transaction is when it's already actually published and what that means is there's no front running I can't try to oh I see that you're trying to do this trade let me pay a lot of money now to try to get in front of your trade in front of your trade um I can again I can blindly try to buy better positioning saying hey I think someone might be doing a trade like this I want to get in over here but I can't actually look at your transactions in front one them we think that's a really important property it's one that exists in arbitrum today and it's one that we want to be preserved in any solution so what time boost does is it kind of tries to straddle both of these worlds and get you know give you the best of both worlds and what it does is it says

arbitr is still going to have these really fast block times uh but there's going to be a notion of an express lane and so you can buy basically for a short period of time you can auction and buy an express lane which gives you a slight priority and the way to think about it is all the express lane transactions will get included in the next in the next block some of the non-express Lan transactions might get pushed one block later right so you're still going to and what that means is for your average user that's just playing their game or or doing something you know basic they just want to know they're going to have a really fast block time they're still going to have a really fast block time you know the worst cases they'll be included not in the very next block but

in the block after that but the block times are very fast and so it's still going to feel very fast to them for those that want to buy a little bit of an edge they'll be able to buy the express Lan ticket to guarantee that they will be included in the very next block uh of these transactions and what we see here is um you know a really uh ni nice properties for both people again those that don't care about ordering aren't going to get delayed right no matter how much these other people bid they're still know that they their transaction is going to get included very quickly they're not going to get delayed indefinitely and maybe the next block maybe the one after that but really that they won't get delayed past that but those that want to By Priority are now going to have an option to do

that for the chain so the chain is still going to is now going to actually collect a lot of the value that's being spent in the ecosystem anyway but today it's actually being spent externally and the chain is collecting zero of that value and with time boost um but as far as I understand so since there's a private men pool you can't do front running but you can still um do liquidations because you have the knowledge of after the block so for the next block I can include um healthy me like back running as well as like liquidating lending positions Etc right correct exactly so yes the way we like to think about it is there's different types of me um we think front running is you know the bad type of me one that we

think should be um eliminated to the extent possible and then there's like back running which basically says like you know hey I see a transaction here I want to get my transaction very quickly and maybe I want to buy an express lane ticket so I can know that I'll get my transaction immediately after your after after in the next block after this transaction or something like that um that we consider to be more healthy M things like Arbitrage uh again so it's the idea here is not to say all MV is bad but we think the idea of you know seeing you user transactions and front running them that's something we'd like to prevent and the current arbitrum design today doesn't allow for that uh and the arbitrum post time boost will not allow for that for that either that makes sense um what's feedback been like

and are there any like General concerns over the time boost proposal yeah I think that there's been a very very healthy um debate about time boost um generally um you know it's it's it's a new system that kind of deviates from you know the priority fee like systems that others are using today so anytime you introduce something new there are those that are going to have you know the uh fear questions or hey like why don't we just use this ABCDE this exists today and I think it's a very strong value proposition which it really provides you know some very very highly desired features but also um you know provides some of the B benefits of some of these other systems so I think it's a

very unique interesting place in the design space and ultimately though there are there are definitely people that you know have questions and see it differently which is why I'm very happy that it's you know going to the Dow and and those voices will be heard right I don't I don't know what will happen uh at the Dow honestly I think that there's a good chance that it gets pasted I think there's some chance that um some voices say Hey you know we don't want this and um it's not a win or loss to me either way I'm a proponent of it I think it's a good idea but ultimately the way we set up the system is that um the community decides and if the community decides hey you know we don't want this or not at this time that's okay the nice thing is that we've also built in I

think one of the one of the places of some uh questions or concern and honestly uncertainty because you know no matter how much modeling you do until the thing actually runs in public you don't exactly know is you know how much uh you know how how much delay can you buy right so I mentioned the default is 250 milliseconds which is basically one block time uh that can be shrunk down you could do 50 milliseconds of Advantage you know it can be it can be shrunk or expanded so uh the proposal and the way that it's written gives um you know the operator the sequencer the ability to uh you know figure that out in some range to find the optimal the optimal uh um amount of time for time boost and I think again there will be some I would imagine parameter finding and making you know once the system is

live but ultimately um I think that this gives a sequencer enough uh latitude to uh find a really really good uh set of parameters okay so uh since you guys have been in the L2 interrupt working group I'm curious what are the biggest challenges you've seen in achieving true interoperability for the ethereum ecosystem yeah it's a great question so I think first I'll say is I think many in the community view interop as this very very deep deep technical you know research problem I think that there's an aspect of that but I would actually argue that probably 90% of it is a ux problem and then there's at 10% which is a deep

research problem what I mean by that is there's definitely like a lot of deep research you can do around say bridging between chains deep interrup um you know even like arbitrary message passaging how to do this and Trust minimiz ways uh efficiently um Etc and we'll definitely get better than that over time but I think we already have a really really good tool set but it's not necessarily being presented to users in the most intuitive way in a way that you know they can just like know exactly what to do um when I speak to users you know there's some like just L2 Fair right honestly you can go to coinbase or binance or really any any exchange and you know you don't need to know anything

about nl2 you just say I want to use arbitrum and arbitrum is there like like ethereum like Solano like any other chain any other chain but there's definitely some uh fear among users like oh this seems more confusing because it's a layer to so I think a lot of it is a narrative part but also like ux you go to the salana ecosystem you know the Phantom on boarding you go to Jupiter you know exactly what to do you know what to you know what to use and it's very very clear I think because there are so many layer twos and because there's some of this like almost confusion without even using the product and just like is assumed complexity right so I just assume that using a layer two is hard even though it's not like there's there's a lot that we that we have to do there to actually package

up the user experience so a wallets is is is is a really really great place to start and say how do we um you know just get better onboarding experience for users or a lot of really good teams building in the wall in the wallet space today and I think just um integrating some of the Technical Solutions we know how to do today is going to be um a big step to that so we at offchain Labs recently announced something called the uh Universal intent engine and this is an intent based interrupt system and our goal is to basically get less than 3 second uh interrupt between any different any chains that are that that are participating in this with an initial um five chains uh as part five chains initially and hopefully expanding

much larger so permission list opens open set of contracts that can work today and on any evm chain and the idea is it's intent based right so the idea is that um it's using liquidity uh and intents but it also has something which is really really neat because one of the problems with these systems is like if you have a liquid chain you'll have an ecosystem of solvers that will participate and can actually make this system work but let's say there's like a new chain that launches today right there are over a 100 arbit from orbit chains in development solvers aren't going to have the incentives and the liquidity to be able to support all those so we added something called the um solver Loan program which is being uh funded by the Arbitron Foundation initially it's a really nice solution that basically says um you know you can

go on these chains and get basically interest free short-term loans to go ahead and actually know solve these intents and it makes it economically feasible for solvers to come in they don't have to lock up liquidity and fragmented ways in different chains you know the the liquidities there there's like a hub and arbitrum one so they can they can manage their their liquidity over there take these loans pay them back and the nice thing is it's extendable right so the Arbitron Foundation is not going to be able to do this for 100 200 a thousand chains but if you're launching a new chain yourself you can go ahead and participate and you know put up the liquidity yourself maybe in your own token and now have your chain basically easily become part of this and what the universal intent engine does it tells you on any chain

that pops up solvers know exactly where to go on that chain to find these intents right so we're basically trying to solve this problem um as stage one with you know providing the backend infrastructure to make bridging and interoperability really really easy uh between chains but also putting in the liquidity incentives to make to make the system work and make sure that those solvers are incentivized to participate in different chains I think step two here um is making sure now that you know this is nice these intent based Solutions and this Universal intent engine are nice developer Solutions what we really need is wallet infrastructure that just abstracts us away from users right what we need the experience is hey I'm a user on arbitrum I want to go and trade something on some other chain uh

maybe that's another chain in the arbit ecosystem even right maybe it's ape chain and I want my wallet to do the following I want my wallet to say trade this asset and AR for this asset on AP chain abstract away the fact that on arbitrum the gas token is eth on ape chain the gas token is ape and just do this for me and by the way show me my balances in one in one in one place right we have all the technology to do this but I think this is basically where 90% of interop lives and we're we've been pushing very very strongly ahead uh with this vision and really want to see ethereum uh interop look like that look like the user goes they can do anything cross chain in the ecosystem there's an intent based solution that works works in the background eventually again and

that's where the 9010 comes in maybe in in a year or two or three or four or five from now the back background infrastructure is going to swap out a little bit a little bit maybe it'll be more efficient maybe it'll be more trustless maybe we'll rely Less on intense and more on something else I don't know but we have enough today that we can already build an excellent user experience and packaging that user experience together for users is uh I would say um a large part of of the of the battle today and I'm confident that we as an ecosystem will do this uh we have the tools to do this we have the desire to do this you mentioned you know working groups there's a lot of different teams coming together that are trying to solve this and I think you know we are definitely doing a large

part in in bringing a lot of the tech to the table that makes this possible and I think uh together as an ecosystem we will solve this for users so just like I just want to follow up here with like a like a quick question to that so you say that um intense it sounds like intents more or less get us 90% of the way there in terms of solving interoperability and I've seen like um I've seen other discussions where they focus more on the research side of things so like like uh real-time proving some people say that like true interoperability won't be here until real- time proving is here and I guess in your view that this is essentially like a non-issue right like intents mostly get us there real-time uh

proving and stuff would only serve as like a complimentary or if you could just help me complete the view here I think realtime proving actually I'll say something even a little spicier I actually don't think real time proving helps very much uh for many of the way that people ways that people think it does I think intense is one you know good piece there's also you know tools that can be used for arbitrary message passing as well and there's a host of different solutions and different service providers you know we work across interop with teams like across and layer zero and chain link there's a lot of different and circle right there's a lot of different pieces for different particular assets that also can help so it's a lot of it packaging together and that's why you know really the wallet is helpful right if you're if

you're doing if you're using usdc you'll probably use you know um you know you know circles native you know ccip right if you're using uh some other asset you might use an nintend bra so so it's it's really about packages to users and not you know making a user say okay for this particular asset like what should what what should I use should I use a cross should I use layer zero is this an of it's really about packaging This Together uh and making it seems for users but your question though uh about real time proving around realtime proving there's a big problem uh that people don't really address and realtime proving doesn't actually solve this right so the the vision of real-time proving and I'll try to sell it for a second before I tell you why I disagree the vision of realtime proving is hey

you know we can just use zero knowledge based Solutions if chain a can BAS basically just give you a proof of what it did to chain B right so chain a gives a proof to chain B of the current state of the chain and it's real time that's all you need for interrupt right because chain B just ingests that proof and Don doesn't need to wait for anything else wonderful it can go ahead and process those balances and you can just basically pass these proofs along in arbitrary ways and chains just can trustless like do everything it's a great vision the problem is the problem is there's another piece so say that I'm chain a and you're chain B we're both realtime Prova bles zero knowledge chain I you I hand you a proof in real time it says hey um you know on my chain

Alice uh says that she wants to go ahead on chain a Alice says that she wants to go ahead and transfer funds to chain B you see that proof you verify it checks out so now you're saying okay so chain B can go ahead and now um you know and and and now um accept that deposit if you will right so uh honor honor that honor that cross chain uh send now let's say the problem is is doesn't really work unless you trust me the operator of chain a or the sequencer or whoever that is whoever is producing that proof right what's stopping me from producing a second proof a second proof also in real time that says actually Alice is sending those those same funds to a different place right so you have

now I can produce if I'm the one producing that you know producing those proof say I'm the sequencer of that chain or whatever that rol is in that chain I go ahead I produce you two different proofs that um are two different histories right and if you will double spend Alice's coins and the answer what stopping you is these should be you don't want to actually just trust a proof from me a proof doesn't mean anything unless it's the canonical proof you need the L1 settlement time right so you need to actually go ahead and this proof is not just a proof of some chain some TR chain the transactions that I'm proving are included uh in the L1 block they have fin ity those transactions so I know that hey it's not just even giving me a proof of some set of transactions I actually know that those

are the corre correct transactions so I need to really two things I need to know that the set of transactions are final are correct and also I need to know that okay what happens what's the execution so the proof can provide me the execution of those transactions but I also need to know that this is the correct set of transactions because otherwise I can give you two very very different histories of my chain give you real-time proofs and they'll both uh you know seem correct to you you and you won't really know uh which one if you accept one I can go and you know kind of rug you and give you the second one after that so what you really need is finality and it turns out that L1 finality is not something we can do in real time right it takes 12 to 18

minutes to get to get real layer one finalities if you have these two ZK rollups even if they're giving you real time proofs until you have that finality there could be an L1 reorg right so even if you see it on layer one included in a block there can be a reorg and that set of transactions that you're looking at which you have a proof for is is worthless because the actual set of transactions that were included in layer one is a completely different set of transactions and you actually gave you know you credited Bob in your chain because you saw proof that said hey Alice paid Bob and ask asked those money to asked for those funds to be sent to your chain but actually in the real set that actually ends up getting finality Alice sent those funds to Carol and now Bob has these chains so so what you have

is one of two things either you need to wait for L1 finality or you basically need these chains to trust each other right you need to trust me not that the proof is correct CU that you can verify that I'm not going to create another proof but if you need to trust me that I'm not going to create another proof then the proof is not very helpful you're just kind of trusting me right so just trust me that I'm going to submit these transactions to to to the layer one so U that's why I'm bearish uh real I'm not bearish real time proving I should say because I think it's a great technology and zero proof is fantastic but I'm bearish um realtime proving as being a realtime interupt solution I think there's always going to be this optimistic period where you're trusting someone and therefore until we get to

like instant finality in layer one which we're nowhere near I don't even think we have paths to know how to get there until we get to that point um you know the Delta between instant zero knowledge proofs and zero knowledge proofs that take 10 minutes is actually not going to be um very very helpful for us that makes sense um I want to talk about this wallet thing that you brought up twice because I think it's I agree with you completely I think we should abstract as much as we can at the wallet level um especially with intents like completely I think some stuff is theoretical and we work towards like some theoretical interoperability but at the end of the day like intents are fantastic there's some trade-offs that come with relying on solvers but I mean

if you have to trust parties you have to trust parties um are you guys working on anything with regard to that like are you working with wallet providers are you working on um your own solution potentially or how do you see that like wallet space evolved to abstract away orbit chains but also abstract away like other Els or interactions between them yes we're we're working very very very uh heavily in the interoperability place and think that it has to be a it has to be number two things one it has to be a solution that involves many different parties right and that's actually one thing which I think I want to you know zero on a second interoperability is never going to be solved by one team and you know what we like to do is say how do we you

know build our technology but also work together and and help and facilitate you know the ecosystem in in solving this if a few years back and we still a partnership that's actually very strong today we announced a partnership with espresso systems on interoperability Tech um I could talk about the specific flavor what we're working on with them but I actually want to make a meta point which is and we announc then this is true then was was true then and it's true today the reason why we want to work with others is because you know if I went ahead and created the arbital interoperability stack I can promise you that like the op chains aren't going to use it and like polygon chains aren't going to use it and I very much think that in order you know to get out of our own way we need to actually enable

neutral parties work together put out you know really good infastructure together with with with others as well that we can all sign on and say hey this can be the interoperability stack not just for my particular stack not just for Arbitron chains but actually for ethereum and maybe even Beyond right and let's just unify as much as we can and I think this is a mistake that a lot of teams fall into where they very much try to build a walled interrup you know experience in their particular branded environment but we have a a larger environment that we're trying to interrupt that's called ethereum right so I don't need to create some other like multi-chain structure in ethereum my goal with arbit from morbit chains is to uh make them as easy to connect to any chain that that that that a user

might want to go to right so it's not just say hey if you're in the orbit ecosystem you connect really really fast to any other orbit chain there will be Times by the way where where from a technology reason we can have better Tech between two chains like that of course we should use that but more broadly you know unless there is a good reason for it I don't want to build artificial walls so if you know the the uh the universal intent engine that we mentioned is a great example it's just evm contracts and there's no reason that uh you can't launch it on any evm chain so I don't want to build a wall and say only Arbitron chains can use this I actually want to build the best in drop because if you're a user if you want to go from arbit to base you're going to

get there if I try to add like friction it's just going to make you more annoyed and maybe your your likelihood of coming back will be less right we're not stopping you from getting there so therefore I think the thing that we can do best is say how do I make that path for you to get in this case off of arbitrum the easiest and most seamless possible and not add any artificial friction at all so you know we're working very very hard we'll have you know more to share specifically soon you ready announce the first phase of of um of our of our work here which was the universal intent engine to make interop between any two chains you know fast and again our our goal here by the way there are a lot of interesting people working on a lot of really good people working

on this EF is working on this we want we just want to bias towards action right I don't want to spend a year writing the absolute best spec and then another year implementing that because I think so much is here today I want to I want to say hey this is what works today we're building this right now which is what we're doing and we can continue to analyze it and continue to improve it and want to be in all those conversations and calls and if in a year or two we have something better let's swap that in but this is a solution that that this is a problem that needs a solution today so and yes we will be doing a lot more so some Alpha will doing a lot lot more in this space directly as well um but um I'll leave it

there for right now yeah and so actually I kind of want to touch on that so um off chains played a big role in ethereum's development uh both with this Universal intense engine and also with the prism uh client that you mentioned earlier and so just like I'm thinking on like a greater scale uh in the ethereum ecosystem Beyond Uh ethereum's current problem area of scaling what are there any other Primitives that you think should be enshrined into ethereum for greater versatility of the network I would say um you know other than like things like continuing to iterate on like pre-compiled and the

like I would say that that you know the current development path you know is is the is the right way to go where we have a very rollup you know a rollup Centric road map and allow a lot of the say execution Innovation to happen you know in in the in the inside the rollups now if there are like pre- compiles that will make all rollups life life easier and I think that's something we should absolutely do and look at if there are new cryptographic pre-compiled and the like things we can enshrine inside ethereum but I I get very nervous when people talk about ining ethereum that we're going to actually stifle Innovation if I look at ethereum comp to the other ecosystems the thing that it has really nice is the flip side of this quote unquote fragmentation coin it's the fact that you have you know dozens

of independent excellent core development teams building on ethereum with slight variations because they're not the same people not the same teams and don't have the same ideas but often have you know very similar at least overlapping Vision if we stifle that Innovation and we got this some what our competitor's ecosystems look like you'll have one or two teams building out the entire ecosystem and you know what you'll have you'll have less fragment a you also have less diversity you'll have less experimentation and I think ultimately that's a bad thing right so um I'd like to see ethereum you know continue on its path and not enshrine like deep execution uh unless it's like really like a common denominator to benefit all rollups like hey if you could say hey all rollups need the evm pre-compiled so like let's figure out a

way to do that in a way that you know as metallic said he put out a blog post about Native rollups in a way that's extendable so now you know we have this base thing which everyone can use on efficiently and now you can add these other features that's fantastic but if you look at ethereum even in execution environments you have us that are supplementing the evm with our Rich from stylus um and by the way that's not a New Concept if you go back to ethereum history there was something called ewm which was the exact same vision wanted to do that on ethereum we've uh you that project uh was discontinued but not for you know really mostly for technical reasons U we've you know accomplished that today and you know I'd love to see that eventually go back and you know become part of core ethereum if the

community is interested in it you have teams like uh um you know uh fuel building the fuel VM you have the starware team building Cairo you have teams like Phoenix building a fully homeomorphic encryp encryption encrypted version of the of the of the evm the idea is that this Innovation is very very powerful and if you fast forward five years from now I can't tell you for sure I have my ideas but I can't tell you for sure which one of these paths will have one out and which will be more dominant than but what I can tell you pretty much for sure is that the evm as it stands today um will be improved by then and we'll have taken Innovation from one or more of these paths and it will become part of the core evm and that's why I think it's so important not

to uh cut off this experimentation or discourage even this diversity on ethereum yes it comes with a price it looks like fragmentation you have different teams different Visions doing different things and you know it's a lot less uh you know it looks a lot less cohesive and say another ecosystem where one team's calling the shots and there's no one else you know dissenting and everyone's building towards the exact API but you know this is a place where I disagree with SE toi who says hey there's only six interesting smart contracts we don't need to execute you don't need to innovate on the execution layer today I think that's a very uh I think that's an incorrect Vision because I think there will be so much Innovation and by the way don't just look forward and take my guest Look Backwards five

years ago today right like completely completely a different world I remember when we were initially raising funds for offchain labs is a little more than five years ago about six years ago um and no one wanted to talk to me by the way the word rollup didn't exist um and initially people didn't want to talk about something new because they their question that they had was is it a state Channel or is it plasma right so the notion was back then everyone knows that scaling is going to happen via State channels or plasma and if you don't agree with that then you're wrong and I don't even want to talk to you obviously that's not the way that things evolved and now we have something called rollups and state channels and plasma look like things of the past although there's been

a little bit bit of resurgence in interest in plasma but but but still you know rollups are dominating the day so I think if we Look Backwards it would be foolish to say that we know today exactly what the future will look like and the only way that we know that ethereum will be safe in the future and by safe I mean continue to be relevant is if we allow that Innovation to happen and the good news is that Innovation is happening primarily on ethereum so I don't know exactly which thread will wi out but I like all these smart people with different ideas doing their things in the ethereum ecosystem yeah uh yeah so like the ewm like ethereum 1.x that's a deep call yeah um so just to touch on the um on the safe portion of what of what you've said um

ethereum is known to be likely one of the safest chains right and just like in terms of like the cons the sheer number of consensus and execution clients right but at the same time something can be said that these clients might like slow the time for like ethereum upgrade and many have said this online so I'm curious to know like how would you from the um from the perspective of someone that like helps develop the consensus clients uh balance the product update like timeline for ethereum so one of the nice things about the layered architecture is we can have innovation in different layers and not actually have to have a dependency on them so I think it's a good thing that

the consensus client layer moves slowly what I think that would become a bad thing if we said hey we can't do any Innovation anywhere else or we can't have execution layer Innovation because until we figure out this consensus please so you know even in the consensus client right we've done a lot of innovation there right the whole move from proof of work to prove a stake the merge was a you know years long process maybe that was you'll say that was slow but I think that there's an aspect of taking getting it right understanding ethereum is is a very mature project today and we can't just you know push changes you know willy-nilly we actually have to test them make sure that they're um really respon the responsible thing to do um and battle tested before we actually put them live on mainnet so

there's definitely some friction there but again the nice thing about the layered architecture is it's not really slowing down Innovation on these other layers right when when etheria moved from proof of uh work to proof of stake arbitrum also moved from proof of work to proof of stake but it didn't have to do anything right it just benefited from that layer underneath it and arbitrum has been able to evolve itself and add stylist without any input without any approval from you know the execution layer right even internally the prism folks you know didn't have to say okay we're going to allow stylist to happen because all that it is is smart contracts that happen to ethereum and that's actually the benefit of of the layered architecture so um I think it's good to have certain layers that move a

little slower and certain layers that move a little more quickly and that to me is literally the entirety of the benefit of the rollup centry growth map if you take a different ecosystem which doesn't have this and you wanted to make some execution uh change you're going to have to actually do that coupled together with the consensus client and have to make this whole systemwide change and test all these pieces together whereas now we have very very well-defined apis if you don't change you know the evm layer do whatever you want on the consensus as long as we think it's a good idea on that layer and everyone else will just inherit that I think that's how we allow for both you know the software to aify but also to have uh innovation uh where we think we still need rapid um progress what's up

guys um welcome back to ZX research today I'm joined by Antonio from dydx thanks for coming on Antonio hey thanks so much for having me what would you say is a North star for dydx yeah so right now there's two things that are important trading volume is the Hallmark of any exchange right it's like that's um users putting through real dollars and volume real dollars and fees Etc but one of the things we are very focused on right now is the number of active Traders on dydx perhaps even more than the volume right now um and it's really just trying to make the product more accessible um make sure that it's not only really sophisticated traders that are trading

on this thing and that have access to this thing um and and really just trying to improve that by orders of magnitude to make dydx more widely used um and that'll obviously play into trading volume eventually and into the bottom line but it's probably users I would say is the Northstar metric right now perfect and what would you say is your general outlook for 2025 both for dydx as well as for crypto yeah so I'll start with dydx and then we can zoom out but with dydx as I mentioned really in q1 we're focused a lot on quality focused on a few new things like this simpler trading experience um then longer into this year I think you'll see us starting to really

develop some novel things on the product side again what is that exactly going to look like again I don't totally know yet I have a few ideas um but I think you'll start to see the form factor of dydx diverge from what you've seen in other exchanges and that's going to be intentional um I really want to take a shot at developing a product that is new and that's again really uniquely enabled by web 3 in crypto generally um obviously the price action has been pretty positive for the past month or two I'm not going to make any wild predictions on where the price is going to end up by the end of the year or so hopefully we're in the earlier stages of a bull market um but we'll see on that and the these things have a way of

teasing you and taking a little bit longer than you think but the the biggest thing that I'm always excited about much more than the price action is the development action and what types of new things are being built within crypto I have been a big fan of pump fun just from the sense that it's a really novel type of product that we haven't had access to before um there's a lot of things going on with iteration of current products dydx like a uniso perhaps a hyper liquid some other things um but yeah just really excited to see what can be built with crypto social f is and another trends that I'm excited about and just really trying to make crypto something that the real world pays attention to because I think crypto

right now is something that's become fairly Niche and that's fine like there's nothing wrong with that but I got into crypto 10 years ago because I really felt like it has the potential to change the world and have we done that as a space yet no not really like sometime we we sort of did or we on the precipice of that with like FTX and SPF and all these sort of things so that's what they promised but that blew up um and now we still have great products there's still a lot of opportunity for trading that's really important obviously I'm you know running a decentralized Rd exchange um but at the same time I am really excited about the potential of crypto to Branch out Beyond and just become something that everyone cares about and everyone can pay

attention to and even if it's under the hood plays a big part in their daily lives I'm going to shift away a bit from Technical and I want to talk a bit about the L2 landscape um sure obviously over the past year and a half things have changed quite a bit I think Bas is has eaten a bit into the market share of arbit trim um and we have other fast EVMS launching l2s like Mega E and monat just launched death net I don't know when they're going to launch their main net yet but um how do you guys plan on competing with um these other chains both on a speed um level I know you guys have the fastest block times I guess activity level makes more sense um and on the other hand one thing I feel like arbitrum had going on for it was defy

tokens and legit product tokens for a really long time um the market shifted quite a bit since then I would say like since the days of arbitrum Defi and L2 Defi and now we're getting more more attention into random tokens um is that something that you're concerned about do you think attention needs to shift back into arbit in some way any thoughts on how to bring back um more activity my thoughts are we have to have like very long-term uh convictions and what's important and what's not important right um that being said you can't ignore the shortterm right so if you're just going to look at the 10-year Vision uh you actually might not survive to to to see that 10-year Vision out so you can't

just say all right I don't care about anything today I have this other future in mind you just need to build towards that you have to engage with the shortterm but Al ultimately you with a long-term Focus I think we as an ecosystem actually did a very very good job at that right so arbitrum you know Meme season didn't take off in arbitrum right so let's just call us Spa spay here right it didn't happen there um ultimately I don't think that's a bad thing we didn't go very heavily after it like you know other other other ecosystems did um there are some structural reasons such as the front running protection and fairness which I think made it less um um attractive for some of these developers to launch some of these things in arbitrum and those were values that we didn't question and

say all right let's let's get rid of that to to sort of you know make this uh more attractive for a particular type of token in a particular type of time we had long-term convictions on on the importance of certain aspects of the software and the bets we've taken we've built I think very very strongly towards those and actually like you know if you look at if you look at today it's funny we're having this ation today as opposed to say three months ago if you look at today we're actually seeing some of that you know come back if you look at literally go to defi Lama today arbitrium is again the number one token the number one sorry uh L2 when it comes to Dex volume right so it has for a while because of the meme activity it wasn't but when it comes to the real

defi economic activity that was always shining through an arbitrum as the meme stuff quiets down again look at today um and it's been that way for most of the past week arbitrum has you know reclaimed it its spot and it's you know I think it's like number four of any blockchain it's been number three it's like you know basically ethereum arbitrum salana BC are are basically um you know taking those top four spots so that's like an example of not getting too clouded right three months ago you might have said oh my God like memec coins it's it's losing all but like it's okay like have a long have long-term conviction about the bets You're Building I personally am long-term convicted about a few bets one number one is I want to say fi right the merger of Defi and trafi we're seeing a lot of

this tokenization um stable coins are doing really well in arbitrum just a week and a half or two weeks ago usdt announced the usdt zero expansion where arbitrum is the Hub now for basically all new usdt chains that are you know other than the four originally natively minted chains we're seeing a lot of this Innovation um as well there's going to be some other interesting um other interesting stable coin activity that we're that we're seeing in arbitrum um um I think you know stay tuned for for for for some uh more announcements month on on that end uh when it comes to additional new stable coin adoption and we're seeing that happen uh very strongly in arbitrum so I think arbitrum is doing well there another place where arbitrum is doing really well this is still a nent industry but what is gaming

um arbitrum has definitely the top out to uh mind Shar and gaming and I would argue one of the tops in all in all in all blockchains and that's the place where again it's not the most popular bet today it's not the BET of like mem coins and AI coins you know or maybe it's not the bet that was popular or two ago but it comes from a long-term conviction and when the market like didn't gaming didn't take off in in a large way as soon as I thought it would be we didn't we weren't like all right let's give up on this bet no we were like doubling and tripling down um you know there's the gaming developer conference in San Francisco in a few weeks we're going to have a very large presence there uh as an ecosystem because we have a long-term conviction

in that bet even if that's not the case today and that's because if we look where the the long-term value is so again if you look at these core metrics you know today arbitrum has um been number one in in tvl on the Chain you know in in overall Bridge tvl from day one um yes it's definitely gone a little more competitive but also um you know we don't uh manufacture any any of our uh metrics at all so what you see is is is really the truth it's organic it's what you get yeah it'll definitely be competitive but I think having long-term conviction and staying true to those is going to hold true and again we can already see that today where these metrics where arbitrum had been lagging for the past few months is already becoming you know rising to the top

again as some of the what I'll call like shortterm Cycles you know die down yeah that makes sense um I think we actually see this quite often because real world assets and real world use cases and gaming takes a lot of time and long time and lot like a lot of effort and capital to build out um often it's like a criticism for l1s and l2s that focus on them um one question I wanted to ask and I I've always wanted to ask this to somebody from offchain Labs you guys have potentially one of the strongest like I guess strongest um most dynamic and like kind of competitive governance forum and governance system to the point that do you guys at offc Labs ever worry that it might be negative for new

applications if a lot of applications control a lot of governance and um votes Etc so I've developed a very strong uh opinion around this really and it's it's little over the past few months and some of this conversation has happened in in in public some of it has happened there was uh just at Denver there was an Arbitron delegate day some of this is happening in private as well among delegates and I think that there's uh and again it wasn't me pushing this necessarily I think the ecosystem has really come to this um sort of new understanding and I think we need what we need is um we need to really differentiate between different types of actors in the ecosystem right so there should be arbitrum aligned entities and then there

there are like Enders and service providers I think one of the problems that we see in the ecosystem today is that there's no clear you know line there always so you'll have um and this is something which I had this you know I won't say the details right now but if you look on my Twitter I kind of had like this whole public thing a couple weeks ago where you you had a a a service provider who was you know doing something you know who was had a job to do something for arbit from Defi and then you know in the other breath they were just talking about all the other ecosystems are doing this and like if you look at me at offchain Labs if like my head of product or my head of strategy was also the head of product the head of strategy and my competitor

you know that would be a problem at the same time by the way like I look at the lights here if the utility company is also servicing my competitor that's not a problem I don't expect exclusivity for my utility company or my ISP or any of those Services even things like my my Auditors right I don't mind if you know those that are auditing arbitrum code as long as they're professionals and good at what they do are also auditing you know competitors they're not paid to be uh you know strategic and operate operators for arbitrum they're paid to do a particular service and I think so I think there's been honestly some uh confusion and conflation of these roles and what we need to see is a delineation of you know a separation of these roles where there are service providers and vendors that do specific roles and then

there are those that are operationally and strategically pushing the Dow and also I think a strong honestly conflict of interest policy which is and we've seen actually this this happen very well a really good example is when the GMX folks actually voted to fund synthetics uh right so a competitor but they were able to understand that in this role it's not about them hurting their competitor they actually want to do what they think is best for the arbit nucle system and you know GMX folks uh coin flip has been actually fantastic AR arbitr delegates so I don't think that there is like a again hard hard line that need to draw on all these things but but ultimately I think that um what we need is in the Dow we need that those that are making strategic and operating decisions of the Dow and those are

incentivized to do so are like all in and want arbitrum to win the same way you know again if you guys were at blockworks but also working at misari they'll be like what's going on like that doesn't work similar I think for a offchain labs and similar for the Arbitron Dow you need to make sure that the right people are making the right decisions and that we're not like incentivizing the Dow has these programs to incentivize delegates and pay delegates it's not paying them or incentivizing delegates in ways that sort of you know doesn't ask for that Allegiance and doesn't doesn't require that commitment you know like any job uh even if it's not an actual job um you know it needs people that are aligned and want the Dow to win I think that that that's the fundamental route I

think that will answer all those questions so when it comes to how can new projects come in and align themselves well the answer is like if we have the right set of people at the Dow um and the right set of people doesn't mean they necessarily agree with me and everything it just means people that are there with the right incentives for the right reasons that they want arbitrum to win whatever that means to them and it might be a different thing for me that's great because I'm not them and I'm not the Dow and we want people with diversity of opinions and diversity of you know thought and how arbitrum wins but the place where I don't want diversity is I want everyone that's making decisions to want arbitrum to win and to be aligned for that outcome and if that's the case and it's the best

thing to incentivize new projects and new protocols for Dow and not just entrench you know uh existing ones then I think that U we'll get to the right decisions but again I think the answer is really one step earlier making sure that the incentives and the alignment of the people making these decisions are what you'd expect in any organization so I actually I'd like to follow up on this point as someone that's pretty involved in the da um I've seen some like talk lately about um about like preferences to certain protocols so like just like uh the Dow should have like a preference to arbit from aligned protocols like native protocols versus like something that's multi-chain uh something like uh or I won't give an example but like I kind of

see this in a way being conflicting with like uh embracing more of a multi-chain interoperability like uh ecosystem right and so like how would you in your view how do you reconcile like these two things like trying to like build out a native ecosystem alongside like the whole like allowing an arbitrum user to access any protocol on um from anywhere if that makes sense it makes a lot of sense I think ultimately you know it's going to be both of those if you look at we mentioned arbitrum Dex volume before if you look at you know where's that volume coming from the number one this again just today so number one is unis Swap and number two is Camelot so you have right there right the and by the

way an arbitrum is not just like another Unis swap chain you know if you look at Unis swaps volume where does their volume come from it's probably like 2 point something billion yesterday comes from ethereum and 600 or so million come from arbitrum so arbitrum is like definitively like a very very important chain to Unis Swap and actually a massive driver of their volume right even the fact that it's about a quarter or so of of ethereum volume and the their number two chain today is important so um that that's to say like I think arbitrum is in a very strong position that it's not just like another generic deployment for these protocols but actually even for the multi-chain protocols you know pretty important and pretty exciting and I view you know I assume that the incentives will be aligned there but of course there's also

importance of having uh a native uh ecosystem and making sure again it's about alignment right making sure that projects that are all in an arbit and want arbit to win because um you know certainly with with protocols that are multi-chain they'll just follow the users you know more more immediately more immediately and feel like okay where are my users where if they're an arbitrum that's great but like they're less like necessarily bought in you know def fact you know in in in an automatic way to uh the arbitrum vision so the answer is like everything else I think you need a really strong balance between these two I think we want arbitrum to be you know the best experience in the in the way that you're cross chain and multi-chain behem us arbit should be a very important chain

to them and one where they know that their customers can come and they'll do well but also I do think having a homegrown ecosystem uh of innovation is important because I just don't want arbit to become this chain where it's just like these established protocols but the Innovation is happening elsewhere and often when you when you you know when I'm building a new protocol you don't no one starts off as a multi-chain you know Behemoth they have to start off somewhere so it's really the question of how do we incentivize those that Areo their very first place to build how do we make sure that their arbitrum home that arbitrum is a comfortable place for them and I think we have to make arbitrum uh open and fair for all of them and actually if you look to arbitrum history um this is really enshrined in in the values that that

brought brought arbitrum to where it was at the very very first early earliest days of arbitrum arbitrum went live uh to developers in May of 2021 arbitrum only went live to users in August of 2021 and there was like a lot of uh there were a lot of people pushing in say July of that year in July going hey we're ready open this up and we said no we really want to give everyone a fair and open shot to be ready at the same time and let's go back to the de example on the same day that Unis swab went Live Sushi swap went live and another Dex you may have not heard of uh called swapper by DX da went live as well and the reason I mentioned that is it wasn't like it never was us saying like

hey we're going to roll out the red carpet for you know this particular decks and and make them win it's like hey no we're going to try to be fair and create an open and competitive ecosystem for everyone and if unisoft is Win winning on layer one there's a very good chance that they're going to win on Layer Two because they're probably winning for a reason they know what they're doing they know how to execute but it's not going to be because we unfairly advantaged everyone so the the best thing that we can do is create an environment where it's open and fair and everyone thinks that they have a fair shot and by the way I think if you look more broadly at the Arbor from you system this this um you know neutrality is actually something that's helped it a lot right so if you look at stable coin

volume today or you know stable coin market cap arbitrum is one of the top chains or stable coin market cap it's one of the only chain that has you know significant you know measured in billions of dollars of both usdc and usdt and the answer is because you know whereas there are some chains that like are usdc chains and like usdt would not be welcome there there are some chains that are usdt chain right we don't expect to see a large usdc balance on Tron I wouldn't expect to see a large usdt balance on base because these ecosystems are sort of natively aligned with the with these arbit has always been a position of we want to do really well on our layer we want to really be the best infrastructure layer but create opportunities for everyone else to create open competitive Innovation and

we see that happen right with with assets as well as well as protocols where they can come and have their best TR doesn't mean that the little guy it doesn't mean that swapper is going to out compete Unis swap it doesn't mean that the little ones will always you know do that but it means that they're not going to be unfairly disadvantaged they'll have the opportunity to execute and what happens sometimes you know uh the larger ones wins sometimes you have something like GMX which came to arbitrum when was you know actually came from BSC originally it was basically had very very had no traction at the time and has turned into one of the most important protocols in all of Defi and that's because you know we didn't say oh well you have to be successful to get an advantage in arbit it's like no if

you're a team that's willing to build and try you're going to have a fair shot some will rise to the top others won't but it won't be us you know putting our Thumb in the scale and deciding who the winners are perfect um that makes a lot of sense Stephen before we end it I um I have one last question what I guess upgrades or in initiatives are you guys working on at opan Labs over the next few months that you think people need to look into that they should read more about um Research into and what should they be looking forward to over the next six months or maybe a year yeah so I think we we've covered most of them today I would say Stylus is definitely still a sleeping giant today we're seeing some really interesting usage of of stylus uh one example is a

team called Renegade Finance which is doing onchain dark pools um and the cool thing here is these have about a ton of zero knowledge proofs in them maybe about a dozen zkps verification for every transaction because it's doing this like private dark poool the irony is you know because stylus allows you to bring in off-the-shelf uh you know zero knowledge cryptographic libraries written in other languages that are about 10x more efficient on average than say solidity it's actually becoming even though we're not a ZK rollup on the application layer um stylus you know makes arbiton the most attractive place for say you know zero knowledge uh protocols and zero knowledge applications again because you can bring in these native cryptographic libraries that are just much much more efficient

and have been around for many years that being said today the Innovation that we're seeing is mostly in these areas like Renegade where they're doing things that are fundamentally imp possible without stylus right they're bringing in libraries that you know doing the amounts of zkp in this zkps in this example that they couldn't do otherwise I think what what what's underexplored today is stylus as a scaling solution right because maybe maybe you want to do something that's running just fine on chain today but it's computationally expensive you can say hey let me rewrite some of these components and stylist and I'll get a 10x benefit so that I think we've seen less of today but I think that's going to happen more and more over time the foundation is investing heavily and has Grant programs you know targeted at stylist developers so I think we'll see a lot of that of new

innovation come out there today there over time and I watch that I continue to watch uh things like time boost as I mentioned uh watching that vote seeing what happens is interesting um if it does go live on chain which I you know I personally uh think there's a very good chance that it will and it's something that we are proponents of um then I think it's going to be interesting to see how that develops over time and how that those markets um you know stand themselves up and and come and and uh and um you know just keeping an eye on that I think is going to be very very interesting um and then really just in the on the side of the ecosystem development you know you'll see at least from my team but I think just from the ecosystem more convictions in our core

bets right so defi trafi uh stablecoins rwas and gaming are going to be pillars that doesn't mean that we won't be uh interested in emerging Trends as well but it means you know you always have to balance these things and we're you know very much going to be in the Camp of we are a very very long-term vision and even if there's another cycle where you know we're still not valuing these things as much we think there's a lot of value there long term and we've been uh pretty consistent on that for a very long time perfect um Stephen thanks so much for coming on this was fantastic thank you so much for having me [Music]