How AVSs Enable Scalable, Secure and Customizable Execution | S8 E4 | Ludwig & Muto

Transcript

maybe a bit of a a controversial take here but I think that the idea of preon pre-confirmation right now um they're very important for interrupt but looking at the actual behavior of users on chain right now um the main really unlock that I think abss will actually have uh is less so the view of having pre-confirmation at the general protocol layer but rather at a application specifically hey everyone this episode is brought to you by SE the Blazing fast parallelized blockchain which is unlocking salonik performance for the vast ocean of eth devs out there now you're going to be hearing all about say

and their new V2 upgrade but if you take away one thing evm is here to stay there are some problems with it which we're going to get into later in the episode but say and especially their V2 upgrade is helping solve that so thank you very much say for making this episode possible hey everyone this episode is brought to you by Kinto the safety first L2 which is accelerating the transition to the onchain financial system they've got some very cool features like user owned kyc and Native account abstraction which solves a bunch of the onboarding and Bott back problems that many l2s suffer from so you're going to be hearing all about them later in the program but if you want to learn more and better if you want to join their launch program engine n GN you can click the link at the bottom of this episode

go check them out tell them I sent you all right everyone welcome back to another episode of bell curve before we jump in quick disclaimer the views expressed by my co-host today are their personal views and they do not represent the views of any organization with which the co-hosts are associated with nothing in the episode is construed or relied upon as Financial technical tax legal or other advice you know the deal now let's jump into the episode all right everyone welcome back to another episode of bell curve joined by my trusty co-host miles uh of Ry and today we've got Muto from ethos and Ludwick repeat guest from sella laps gentlemen welcome to the show thank you for having us it's great to be here yeah

thank you good to be here pumped guys this is going to be a really fun episode we've spent some time in the reaking season talking a little bit about what we could call the supply side market dynamics so that's everyone on the um they know on the ethereum side that's interested in seeking additional yield and what those Dynamics are going to look like what we wanted to talk to you guys about today is try to get a little bit more granular on the demand side of the market from the ABS perspective so each of you guys have built um on the AVS side of things and we wanted to get into why you decided to be an AVS and maybe Ludwick I could toss it over to you uh first maybe you could talk to us a little bit about what you're building at sella the initial problem that you're

trying to solve and then why you thought uh like reaking was the right solution for you um yeah so to give a bit of background on kind of how we came about what we're trying to build and why we decided to build build it uh we initially came into the market trying to optimize the entire industry of providing liquidity in amms so we were aiming to build sophisticated systems to facilitate very efficient onchain liquidity management um from regular users with also funds and larger institutional players but what we kind of came to accept is that regardless of how smart you can be um and regardless of how informed you might think you are

as a liquidity provider the the current implementation of amm just doesn't make it profitable it it you cannot really achieve that level of profit profitability that would be desirable because you are constantly getting arbitraged and offering stale quotes uh and how we see this exhibit itself in ethereum is that you have these sophisticated Builders and Searchers that are um consistently arbitraging the price of an amm against a centralized Exchange price because the centralized Exchange price operates in a faster capacity when the amm is only going to update every single block and so our um Baseline solution here is to create an auction system that pits these

arbitragers against each other so that they can compete to purchase the right to execute that Arbitrage against the pool similar to what you see in a block Builder auction where we're all fighting against each other as Builders to bid the most value back to the proposer but here the value is not directed to a proposer but rather directed back to the LPS that were suffering the economic loss in the first place on top of that we also aim to solve the sandwich problem uh by effectively batching all of the user orders that come after this arbitrager uh so they can form a single uniform clearing price where they will all get the same price and the LPS will all clear at the same price so that there is no reordering mechanic that can um exist where you can extract some

value from those uses all right so BL to say some of that back to you it's like the initial problem that we're trying to solve here uh solve for roughly is lever right there's a LPS especially on eth based dexes like Unis swap are just losing an enormous amount because there's a 12C block time and they're overpaying Arbitrage yours to uh reset the price right to whatever the price of like eth is on on binance say right so that's the that's the starting point of the problem part of the solution there is a is a batch auction type function but in order for that to work you need basically to select a quorum right you need to run an auction and in order to do that you need to select a quorum and I feel like

that's where the reaking part comes in right so the reaking is okay I've got this problem that I'm trying to solve batch auction is my solution in order to do that um you know in theory you could just decide to be a chain or a smart contract or something like that but really what we need here is this Quorum of auctioneers um just for for the for the audience's uh enlightment can you just describe a little bit about I mean don't any state Secrets here but like a little bit about how you actually select that Quorum and then what are these U you know these auction participants what are they doing or opting into yeah um so yeah focusing on the reaking side the core really is to um run this auction it's a batch auction but at the same time a dual auction where you have an

initial auction for the Arbitrage and then a batch auction for the rest of the order uh user orders uh but in terms of how we're selecting the Quorum there's effectively going to be a limited capacity in terms of how many validators we're willing to accept because we don't need thousands and thousands of validators because the state um that we're underlying operating on is on ethereum or any other chain that we might deploy to um The Operators really their their main responsibility is to ensure that the um auction happens as smoothly and as efficiently as possible so we can form the maximally profitable batch for the lp and we can maximize the amount of volume that we're clearing um against the am liquid on top of that against um the various like offchain

liquidity that might be provided in our offchain order book um so what they're doing basically tldr is you're receiving user orders you're you're propagating them around um various nodes will relay them around the proposer will be randomly selected each validator will send in their current view of the network which is the best Arbitrage that is bidding the most value back to the lp and all user orders that are valid that they've received and then the leader will merge all of these different views into the globally optimal View and then submit that on Shain um and we don't do a immediate verification of the Quorum certificate if you will rather it's an optimistic process where we verify that the leader is actually the leader of our

consensus at that round they get to execute and then we can verify things after the fact to ensure that there was no malicious manipulation from from that proposer within our round of consensus I'm curious was there any point in kind of um the Cilla Journey where you guys were planning on having maybe your own chain right that had no affiliation with igen maybe had no affiliation even with the eth token um and at what point did you kind of realize actually wait if we um need our own Quorum igen layer is now offering you know a way to access like a portion right of eth consensus which is really all we need this actually could you know gives us like the basics of what we need

but could maybe make the product better um because these these right uh operators are running full notes of ethereum already is that is that a way to yeah look at the journey in the in the very beginning um our initial intuition was actually to use LP tokens so it was to use the very collateral that is providing liquidity within the amm to secure the auction that is maximizing their profit in the first place which I still to some extent think is like the maximally efficient way to do things and lines in incentives in the cleanest way the problem with that is that you now enter um an accounting problem basically in terms of your consensus you have all of these different LPs they have varying

concentrated positions they represent different balances they're operating on different pools uh and you now have to account for the fluctuation in price and value of each of the these assets so for Simplicity sake really we we initially thought okay let's it's best to go with ethereum in the beginning um but there's also the like big boost in Economic Security that igen L provides in so far as there's such an abundance of reate capital today that we initially did not consider the optimistic model we thought that requiring such a significant stake on a per validator basis would um basically be a limiting factor in bootstrapping our network but now that we see that there's such a large amount

of capital available we realize that we can push these um minimum State requirements far higher and operate in this optimistic scenario which provides like overall better quality of execution for the users because you're minimizing gas costs at each batch so every single time a batch is cleared and executed on chain you're avoiding these um these costly signature verifications basically got it and um I want to transition Mew over to you here in a second but I just want to like underline that for for listeners and lwi maybe ask you to just say a little bit more more because what we're trying to draw out here is the additional features that you can get and uh from a from a reaking type setup and I would love to get a

sense from from your perspective of why was the why were you initially concerned about having some amount of security like some amount of procured security for each one of these validators that participates in your Quorum and what do you get um as each one of those validators have an additional amount of security like are there additional tasks that they can perform is there some amount of scale in the auction that gets unlocked like what do you get for having more restake Security in terms of more more restake security for us it it it really is bottom line um important on the per validator basis and because any of them can be a leader within around a consensus and the risk there for us is the theft of this lvr uh that is Bid

back to the user so if you do not have a minimum stake requirement that is larger than the big biggest Arbitrage that we've seen in the history of of ethereum they can and are almost incentivized to just steal that for themselves um so unlocking that larger pool and the higher requirements enables you to um open up the like trust that you're allocating to each leader so that you can wait after the fact to verify in case something goes wrong instead of having to verify everything all the time but apart from that um in terms of pulled like pulled security as a primitive here it it benefits from it in in the sense of lower cost of capital

but we still very much need attributable security and I know sham has has been talking about this a lot um but if we have like significant pulled liquidity but within say an Epoch um all of it gets slashed and you have a single validator that get slashed and then there's a chain reaction if we're not able to claw back that amount of liquidity and refund the users that that lost it doesn't put us in a better in a better position really makes a ton of sense mut let's uh let's let's transition over to you here can you just give um kind of same same way that we phrased the the question to lwig like can you just give us an understanding of what ethos is and what was the what was the core problem that

you were interested in in solving yeah yeah so I think the one liner is essentially ethos is a a Marketplace for commitments all backed by restake eth um and I think when we kind of started about this uh we really were focused on like what igen L unlocks which you've gone into quite a lot of detail on and I think for me the the core to it is essentially like programat programmatic access to security as well as decentralization and I think one of the core things and the reason why we started EAS was essentially recognizing that this was possible through igen lay but it wasn't actually as uh programmable as it needed to be and I think for for us in particular what we were focused on and mo mostly interested in is kind of seeing this delineation between different types typ of AVS so

you have stateless avss and then you have state full avss and I think in particular for us what we were focused on was this idea of stateful avss and particularly this idea of having ecosystems that are still able to kind of have their own flare on perhaps their own execution environment their own BM um or having their own kind of uh extensions to what is done on chain what their validators are doing for them but still having that be backed by uniform uh Economic Security level and so for us um generally being ethylin We Believe eth is the the best asset the best credibly neutral asset restake eth is essentially one way to get that cheaper and again more programmatic um and so that's essentially where ethos was kind of grown was essentially this mechanism to build These Chains backed by this

primitive um and then what we've uh shifted towards and are focusing more on is this idea of what are these actual extensions that you can do and how can we actually take advantage of this restake eth in order to effectively extend what's possible in cryptoeconomic systems today and so I think the the most uh preent example that right now is to build something akin to a commitment layer where you're able to take your validators that are running your chain or really just your abs but in our case a chain um and then have those validators compute or otherwise commit to doing some other work and then commit to those on chain and have them be arbitrated and so this is where like there's a whole system of things that we can get into in terms of how IG layer makes this secure using things like

Stakes Shure um but essentially what this allows you to do is build things like co-processors that can then readily commit on chain and then have challenge periods that are all effectively um coordinated through through IG lay and and Ethos is essentially the mechanism to to make that happen and in particular we're building the the first chain that will essentially enable that uh functionality in a more General sense and I think what so what you hit on is there's you know kind of the um vanilla value prop of of reaking which is the ability to stand up chains that are you know faster like in a cheaper faster way um and then there's this much more interesting sort of um you know a design space of of more important

unlocks I think that you're going after um really around interop and I I would love if you could just maybe speak a little bit more to what this actually means um in terms of like a generalized platform for commitments for any sort of like stateful service right so this doesn't necessarily mean it has to be like you know um an ethos powered like chain this could be just one part of let's say like a sequencer right um or like yeah yeah so I think in general if you want better interrupt there's a lot of different ways to do it um you've probably gone down the rabbit hole in terms of like shared sequencers and other kind of designs but I think for us what we've kind of boiled it down to is becoming

most important is this idea of being able to make a commitment and then having some credibility around it such that uh Searchers or just interop Partners can kind of rely on that commitment and so for us the the important components of what makes a commitment credible is uh one is like like having enough Bond or otherwise having enough Economic Security behind it and so that's where igen lay kind of comes in to to solve that and then the second part is needing essentially particularly for the interop use case you need a neutral uh Space by which these commitments can be arbitrated um and I think the the example that's interesting to talk through is like in the context of perhaps like Cosmos chains you can have two different Cosmos chains they can each have their validator sets maybe running if we forget I layer they just have their own

tokens they each make Bonds in their own tokens and they make a commitment um the ISS that you run into is that as soon as as lewood kind of mentioned earlier in like the lvr case as soon as in the interop case that it's uh more valuable to uh descent you will have essentially chains reorg or otherwise socially uh get around their their commitment and so that's where I think having a credibly neutral asset such as eth um and then having a credible neutral layer where these commitments can be made um and these commitments can then be proven against on igen layer is is quite important I think that last part is also more of a like implementation detail but also important which is essentially if you're making these commitments you need to have some way to verify the commitment and you need to basically say

like this person said or this proposer said that they were going to include my transaction in the next block I want to slash them because they didn't you then need to prove the state of that chain on particularly on eel one and I think there are just um good arguments around returns to scale for having a single place where that is happening and that's essentially what the ethos chain is is coming up to to be is that that mechanism where you can prove essentially the state of many chains onto one chain and then prove that state onto igen lay and do the slashing that way there's also various things around like returns to scale for uh bringing in a bunch of pull security I know sham has talked about a lot of that in the past attributable security is a whole other can of worms we can also get into yeah I

actually kind of want to get into that because I feel like both of you uh touched on that that I think you know the the full value prop doesn't kind of come through until you get something like steak shirt right which is again to the for the audience the difference between pulled security and Stak sh is really a um it's two things it's one you know this idea of like isolation right so having some sort of reserved Capital just for one project um and then the second piece is what happens when there's slashing right and with attributable security the slashed assets are actually redistributed to harmed users um and that you know gives you like it's maybe it's an underrated like

but very very powerful tool um and so yeah I think for for both of you uh you guys see this as like a core part of the longer term design um and maybe just to explain to the audience like why this piece is so important for your products um would be super helpful yeah with I mean without it we're just running like a bunch of validators with no underlying backing behind them uh and we're playing the decentralization theater really um like if there is no bond to be slashed and if there is no ability to claw back the funds that were lost um great we've burned their assets so we've dis

disincentivized sorry um this malicious activity but your users are still at a loss at the end of the day so it doesn't seem like really that great of a solution in my eyes at least um in terms of attributable security U the specific case that I've always been very concerned and I've talked at length with kaido and the igen L team in general um about is this idea of epoch based slashing and this was one of the initial designs that they came up with last year and that they decided to scrap and I think it's a great thing that they took the time to think about a better way uh of doing it basically the problem here is you have individual epochs where your accounting State based on these

epochs and so if a slash event happens at Epoch 1 um the update and the balance will only happen and take effect in eoch 2 but from my like initial intuition if I'm operating for an ABS in pop one and I get slashed and that it represents a significant portion of my balance all my my entire Financial incentive is to look at any other ABS that I might be also operating and try and extract as much value as possible because I know that the value of my bond is severely discounted now and so I would always try and extract as much as possible and try and be as malicious as possible every where but now you have a scenario where those other avss aren't aware of this

they aren't aware of this slash update that that that took effect in in this current Epoch and you find yourself in a situation where three abss or 10 abss are all slashing at the same time and they now have to spread like the the piie of that end stake which will likely not represent sufficient uh stake to actually um redistribute that value and make the your users whole and that like is I think a very large problem one of the toughest uh problems in designing the system correctly and I think attribut attributable security provides An Elegant solution towards that yeah um just to to jump on that as well I think one of the the core like Frameworks that I've been thinking about

with regards to triable security is essentially the difference between like what is an optimistic model versus a cryptoeconomic model I think lately we've been seeing this permeate a bit more in the the space where you have like ZK optimistic and crypto economic is like the three and maybe trusted Hardware is like a dark horse in there as well um and I think for particularly the way that we're thinking about like optimistic versus cryptoeconomic I think it primarily comes down to like a ux perspective where um if you take an example of a co-processor what you can do in an optimistic scenario this is kind of like how rollups work today is you can basically do some some compute commit to out- onchain and then you have a fraud proof window whereby anyone can challenge that and I think part of the the Reckoning there is essentially you

need to have some amount of bond to prevent there but it's like it's not enough of a bond as mentioned earlier it's not enough of a bond to make it like actually dis incentivizing anyone from committing uh fraud necessarily um and there's also not a lot of reason to do so because there's a long fraud window um but I think if you flip that and you go into a cryptoeconomic uh model particularly with something like attributable security you can get kind of The Best of Both Worlds where you can put up a much larger Bond uh which will allow you to essentially circumvent the fraud prooof window um and then particularly this is where like sture comes in in particular is if you can model the this the paper goes through all the specifics here on like the cost of um the the cost of an attack versus

the profit of corruption um if you can model out what the profit of Corruption of your system is you can essentially buy enough Insurance such that it is in everyone's best interest to take the the result of say the co-processor as instantly as it's posted um and then you can rely on the insurance mechanism to make any users whole on the back end and so I think that just unlocks like a whole different set of interesting ux um parameters and ux kind of like Styles going forward um and so that's in particular like what we're building towards um I think there's a lot of open questions on how exactly this works out but um I think that the ux unlock will be pretty massive there everyone want to give a big shout out to today's title sponsor SE now I want to talk to you

guys a little bit about why I think SE is cool from a design standpoint a big problem that it solves for eevs out there and then some cool stuff that say has coming up the reason I like say from an architecture perspective is again it's a very fast blockchain parallelization all of that stuff but say has essentially been custom building block space which is for Consumer apps and dexes now they have some very cool features which enable that so twin turbo consensus optimistic parallelization say DB all of this stuff allows you to reduce the time to finality make for very very fast transaction so if you're building a consumer app or a DEX this is basically the blockchain for you if you've been building in the evm you love the evm but there are some restrictions about it that don't support your app so maybe you can't do fast enough

transactions or it's not parallelized whatever it is you can now take all that stuff that you built you don't have to start from scratch and you can build it on say now recently they've launched V2 but also public devet so the way that you can follow that and keep up to date is go and follow say Network on Twitter all right thanks guys he everyone this episode is brought to you by Kinto the safety first L2 with use your own kyc and Native account abstraction that is accelerating the transition to the onchain financial system now we've been talking about Larry thinkink on this program and how he's been shouting the good word about tokenization to all of his buddies on Wall Street this is happening and the very cool part about Kinto is that they are built to accept a whole bunch of assets a whole bunch of

users from the traditional Financial system which current l2s have done an okay job but there's a lot of room for improvement for onboarding so it features very nice things like user owned kyc and Native account abstraction that solve some of the biggest blockers around mainstream adoption namely security and user experience you can become a founding member at engine. kin. XYZ if you believe in this onchain financial system join Kino's launch program and thank me later cheers guys and the link bottom of this episode click it now guys I want to change tax here a little bit and ask you guys about uh from a a branding perspective actually in sort of a squishier cultural alignment sort of perspective and something that we've talked about a little bit so far on the season and something that you hear quite a bit

about is this is especially from I think people that are more questioning the ultimate value proposition of uh reaking and whether or not Economic Security is a meme or not and they sort of say well one one of the big advantages of being an AVS is you're sort of buying into a narrative right and and part of this is you you hear this was there was some element of truth to this right when with eth rollups right where oh we're not a failed L one we're an e rollup and I actually think it sort of worked for them for the most part because an enormous amount of attention and eyeballs and VC funding and all that stuff was going into anything that calls themselves an L2 or rollup and I think the pretty phenomenal success of igen layer which say is sitting at 20 over

you know2 billion in tvl and we're doing a whole podcast series on it there's a ton of attention around this ecosystem and I think there probably are um you know some some teams out there that want to like Leverage that are plug into that um and so I'd love to get your guys perspective as Builders of ABS as of like how much validity there is to that narrative and then I've got a couple of follow-up questions after that yeah I'd I'd be happy to talk about this um briefly just because we we actually went through this exact kind of narrative um like roller coaster and so I think in particular the way I think about it right now is most particularly like there are a set of projects that do value this kind of like narrative alignment and I think um from igen Lay's

perspective selling IG lay alignment to them is valuable um however I do think um you do have to get you do have to basically like reckon on like who are you working with here I do think there is like somewhat of a negative selection bias in in these kinds of um mechanisms because you essentially get towards this like mercenary culture where it's like oh I'm just going to whomever gives me the best narrative at any particular time and so especially in in the current uh climate where we have things like um alternative restating mechanisms starting to to crop up um as well as just like generally different forms of shared security so if you look at Babylon and and Cosmos providing Bitcoin security there's uh initiatives around like restate salana all these different things um and then of course all of

Celestia and what's going on there I think you just um if you focus too much on that like narrative um drive from like an ecosystem perspective you end up in this mercenary culture where you'll just have oh I'm igen lay aligned one month and then next month I'm a different um ecosystem aligned and practically one of the concerns right now for for IG lay that happy to go into as well is like the switching costs actually aren't incredibly High um and so if you wanted to go from one shared security provider to another it's not actually the the end of the world and so you can do it and then I guess just to touch briefly and also to hand the mic over on on eos's perspective I think um primarily right now like igen lay has become like a very strong partner for I

think a lot of ABS teams on just becoming like a thought thoughtful design partner and so I think from our perspective it has been somewhat beneficial to be tightly aligned with them um however I think it's it's pretty important to build your own brand and be somewhat independent as well as also just pushing the state yourself such that um I think in eth's future if it ever made sense that we would be like agnostic to reaking provider or if we moved off of um Igan lay to another reaking provider because of any particular value Puffs like that's totally reasonable I just think right now like IG L actually has like a really strong Head Start and then also has the best feature yeah just to to hop on um on what muda touched on um I think yeah this idea of like narrative hopping and and suddenly

shifting towards building an ABS has like become quite an apparent problem um I think in the long run so long as iir is able to manage um who they're effectively attributing the the spotlight towards um that will um be sufficient to protect the brand and ensure that this is a permission system anyone can build on top of it um naturally people that are operating a grift will come we've seen this with ethereum we've seen this with any any permission system especially when the financial incentives are so high um but at the end of the day like as a a a protocol as a as a project in general you're not an ABS you're using igen's back and system right you

are benefiting from the tools that they have created and the brand that they have um effectively propelled towards attracting a very large amount of Economic Security a huge huge amount of State e and that is incredibly important and beneficial but your underlying success will depend on your ability to do something useful with that stake and so end of the day for a lot of these projects that that seem to be hopping over towards this narrative and that you can't really distinct a clear reason to why they're actually now an ABS and why there is like this Economic Security that has to has to happen to some extent um I'm I'm not necessarily that confident that that will like turn out

really really positively for them I think for fundraising it's probably a different story but in general like your end success you're beyond B Endor as a company is to produce useful and compelling products for your users and the entire Economic Security side of things should just be there to facilitate that and just to it just be there to ensure that you have this larger freedom of operation to be able to push some level of compute offchain so you can do Fanci things in this onchain onchain world apart from that I I don't think people should be leaning that heavily uh into it yeah it's kind of funny we were joking about this with shiram and in the first episode of this season it would be like if one of the games on your phone

was advertising as like oh we're an iOS based it's like who cares that's so irrelevant to to what you're what you're describing but let me ask you guys this the uh I could see on the one hand right now where we could argue about whether or not it's really sexy to call yourself an ABS clearly there are at least in the short term some sort of narrative benefits there I do Wonder though um on a longer term time Horizon if that could actually be something that maybe like a larger and more established project wouldn't actually want so for instance they might actually want from a branding perspective I don't want to be known as an ABS right I already have this brand name and audience and user base Etc I think that igen layer is the correct

backend for uh for my application but I don't want to go around calling myself an ABS and so I'm wondering if you guys think that that branding might cut both ways and whether or not there's room for maybe one of these new reaking platforms to position themselves differently yeah um just to to touch on that I guess briefly I think one of the the ideas I think through on this is that um it can be seen as somewhat of like a detriment to call yourself an ABS I think the the most um kind of preing example I think of for this is like this this concept of an L1 premium like we we've had this idea of like oh some Chains Don't Want to Be rollups because they want to be their own L1 because it attracts some premium by uh by being an L1 is that true or not it's actually

like an open question for me I think it's it's true for like the the barbell like the top um few um chains and then it's like extremely untrue for the rest of them um but who knows and so I think that that part of narrative thing it's just very hard to like disprove um that like being an ABS doesn't detract value and so I do think there's like that that is some component but I think to that end like there's nothing telling any of these projects or like even uh speaking for for ourselves here for for ethos to call ourselves like an ABS um in fact in most of like our our discussions like we really don't call ourselves an ABS I think there is an ABS that underlines what we do and then there's everything else that we do and I think that's the

the framework that we think through um I'm not like wholly convinced that any particular competitor is going to be able to like eat any of ig's lunch by like just not calling things abss and calling it like something else like you're going to need something more than that um but I do think it's it's more of a reflection of like the strategy side of this and to their credit like I don't think IG right now is going out of their way to say like oh please call yourself an as please call yourself an ads yeah I think there is one thing I want to touch on there you mentioned the the L1 premium right and and usually that is like a staking token for your for your L1 and there's you know been some discussion of like okay well what are you kind of giving up as an as an ABS

right and you can you have your own staking token what role does it play and then if it gets if you get big enough and if you're really really successful and your own native token is now worth say billion dollar right um do you graduate from the marketplace or you know are there things that are those value ad like sticky Services right that that keep you on the marketplace longer term um and Ludwick like I could you know maybe point it to you if let's say you know Silla is a massive success right and you have a huge market cap for your staking token and you have a lot of liquidity right which maybe opens the door back for like this LP token staking

thing are there you know are there reasons basically to keep you like on the platform or like what are the sort of reasons that you would want to stay on the platform yeah I think this idea of like stickiness towards the marketplace is is is very valuable in so far as IG layers is effectively creating these Baseline perimeters and there is a lot of research work that has to be done in ensuring that this pulled security is spread out efficiently enough where it isn't prohibit prohibitively costly but at the same time also guarantees the underlying security of a given chain um and then you have the concern of the implementation of that and ensuring that that entire uh like smart contract

apparatus is extremely secure now you have attributable security which is also if not more complicated all those things are not easy problems um and they take a very large amount of work and a significant um investment uh we like as a company want to focus on offering the absolute best experience for our users we want to focus on being the most performing decks on ethereum and on other chains we want to provide um cross-chain swaps that are hyper efficient we want to ensure that LPS overall um are getting a really good deal and are able to seamlessly provide liquidity without worrying about these various arbitragers that are very good at finding ways to extract cap like extract capital in any form um and so

that's a huge Focus for us we don't necessarily want to focus on all of these services that igen layer is already specializing at and is spending an incredible amount of resources into improving every and so for me like a huge part of the value prop as well is not having to concern myself with continuously trying to improve that part of the system and focusing on my core business logic and so for me that's an extremely valuable component of the system uh and in general touching upon the marketplace view um it's just a more efficient uh pricing mechanism for how much you should pay for security because you can Benchmark that on how much attributable security a competing ABS is is demanding for and how much they're paying for it

um whereas if you're operating purely your own chain um that Discovery mechanism in terms of emission might be a bit murkier right that makes a lot of sense and it does feel like you know one of the core value props of Ian lay is just being able to carve out more of your time and resources towards the core product right but also not having the limitations of needing to be built directly on ethereum so you can have your chain right um but you don't have to spend you know you can basically reallocate time towards the things that matter which is building a great product um I am curious though like I I know you know ethos is a tender chain um I'm not exactly sure with Cella but how much sort of like handholding work did you

guys do with the igen team to like say okay what is the right Quorum design for this for like my application what are maybe some consensus you know systems that I can grab off the shelf or is there something you know maybe that's a little bit more customized here that makes sense um so maybe like transitioning a little bit over to like okay great this makes a lot of sense let's go with I lay now what do I actually what is the quum design look like like what is actually you know what are we doing with as a chain um and yeah curious just how that Journey kind of looked for both of you guys and you know if that was something that the Igan team helped a lot out with yeah I think um there's a few different things to go

into here I would say like in terms of the actual design particularly as you get towards like a stateful ads as we kind of outlined earlier like there there are actually quite a lot of considerations on like what not not just the Quorum which is uh at least in in IG layer terminology is like somewhat limited but also in like the consensus model that you're running um you mentioned like we run run tendermint um I think that the primary value ad for us on running tender for our chain is essentially like uh customizability as well as like being ready more or less off the shelf to to do many high performance uh use cases um and I think the other part is that for for us for the particular design that we're building um you have to consider like what are the important Dimensions that

you want to optimize for so I think when you're thinking this through like um there are different designs and you see this in in l1s today where it's like some chains like ethereum um highly value having a very large and distributed validator set whereas other chains are happy to centralize this a bit more um or at least have like more high powerered validators and smaller sets um and so I think in general I'm somewhat more on the aligned with having smaller and more powerful of validators um powering particular applications and then building the right layers of control around them in order to make that um still decentralized and still censorship resistance as well as uh powerful and so from from that perspective um it made sense for us to to focus on on tender and as well to focus on our Quorum design which is

again like part of this IG layer construct of like where are you getting your stake from I think for us um we've essentially taken on a more neutral standpoint at least to to start which is essentially this idea that you would want to have kind of equal distribution of um Quorum weight by the amount of restake Assets in in each Quorum and just to or each strategy I guess specifically just to parse that a bit more it's basically this idea that you can restake I mean you can restake native eth you can restake staked eth um like through Lio or any of these other options um and so I think for for our perspective we're kind of happy to be somewhat distributed across there I think for different applications you might value different um uh qualities here so as an example if you wanted uh

to actually push for more decentralization like it was important for your application to have many different validators have them be geographically distributed and all these like more Ecentric um kind of qualities you might push for having a larger percentage of your Quorum come from native uh eth and I think that just generally pushes for like more support for like home stakers as well as this whole other ecosystem but I think in general for most of the abss that um we've been chatting with and also like uh generally working around is this idea that you would just want to kind of get your base level of security um and then figure out the application on top um and I think the one of the greater services that I layer can do is really just abstract like where the the stake is coming from and ideally one of the

things that I think would be useful for IG L to also introduce is some layer of observability as well as monitoring of like how centralized or otherwise is the distribution of these different strategies because I think it it kind of comes into like the network effect or otherwise value prop of igir is like not only is the tvl important but also like how distributed is that tvl because if like just spitballing here there were to be like a restating provider that came up and just had like a bunch of restak E from one particular provider like that would have very high tvl but it wouldn't actually provide a lot of the benefits that you would otherwise value um and so that's that's kind of how I've been thinking about it on our on our um and we've taken like very specific approach to designing our

consensus because it's we have extremely different considerations from like regular blockchain applications like we don't actually hold State uh um we are operating to maximize the optimality of a set of inputs that are only valid within a slot right um and so for that it was a complete redesign we first looked at tender um and we realized there are a lot of communication steps and the underlying schedule uh of consensus does not fit the needs of our application whatsoever and we really want to ensure that we can have maximum throughput and very low latency consensus because the amount of time

that it takes us to form consensus is a direct impact on how accurately um a market maker can quote because otherwise their orders are out in the open for for a longer period of time and um the risk that that quote becomes stale increases so to maximize the effectively the optimality for our users we had to go from scratch to build something that met our very specific uh requirements on the side of the help uh with respect to igen there it was mostly in in the re review stage U I've been working really a lot with kaido and just having someone who who's knowledgeable and and cares enough uh to review this stuff that is sometimes like

tediously cated um is like a super great value ad um but apart from that we really had to to to do things and Implement them in a completely different way so ju just out of curiosity a little bit for when you were did you did you wind up with something that looks like kind of a modified version of tendermint like hot stuff or something like that or did you just start from scratch and build something completely from the ground up so the the kind of process that you I like the T the how it came about was first we have tenament okay what can we get rid of to to shave off on latency okay now that we've gotten rid of this really doesn't look like tender that much anymore um okay now there there are these specific

considerations that relate to liveness we don't actually care about this because our proposal rotation happens through the schedule of an etherum block because we're operating a side card that is tied to neum execution client um so okay let's remove that as well okay now we end up with something that looks even less like 10 which is now basically completely different now how can we modify the underlying schedule of consensus to ensure that the proposer has as much information as possible from all of the other views of the network to be able to create an merg and output this optimal solution um and then from that um how do we ensure that we can do so in a very gas efficient way and that's when we entered in the idea of the optimistic um optimistic style of of

validation but initially the consensus itself was designed without the consideration that we were actually using igen lay and we had not um realized the value of opting in to this AVS side of things with restate eth um and so after that that that did have an impact on on our design because that's the time at which we were able to say okay we can confidently switch from this process where we're verifying and forming a two-thirds consensus on the validity of the output to a model where okay the leader can now just submit directly on chain because the state is so much higher that if there's a $10 million Bond I don't want to say like never say never right but uh realistically

speaking the lvr theft will not be greater than that and so we don't uh we don't have that as a concern and in regular operations things are just that much more efficient man super interesting um so lwi maybe we can get you to like just to maybe zoom out here a second and just ask for instance maybe some folks in the audience might still you sort of danced around this and give it that the latency was actually a really clear example uh for me to just uh get that talked through but you know for folks maybe who um aren't as much in the weeds or just want a takeway of for instance why like where would I use like an AVS versus something like a rollup and actually like today rollups you might look at that and say well that doesn't

make much sense for something like what you're describing because there's one centralized sequencer so you can't have this sort of rotating um you know set of auctioneers but there's really no reason why in fact like tons of companies like espresso are working towards decentralizing sequencer sets for big Roll-Ups and then it actually looks pretty similar to the construction that you have on right I big difference being these are ethereum validators that are running a sidecar as opposed to their own uh separate operators but you know just for for listeners like you just like underline kind of the two or three points like why be an ABS here as opposed to something like a rollup that could also make sense I think yeah you can view a rollup as an ABS in the sense that today we have a very broad

definition of what a rollup means but a very um very specific view of how that looks in practice and the current like centralized sequencer model that we see um longer term like as regulatory um concerns emerge um I think that there's going to be a lot of value in having these truly decentralized systems and from that you can call it an ABS you can call it just a rollup end of the day it depends on if you're opting into Economic Security that is provided through restate eth but zooming out like technically speaking you can really boil down an ABS to okay I'm interacting with this restake security apart from that like the sky is kind of the limit in

what you can Implement um so I'm not sure that like really respond like is a clear answer to your question to be honest but um for me it's hard to specify when you'd use a rollup and when you use an ABS because in my eyes a rollup can very much be an ABS and um they very likely will become abss uh but overall you can see abss today that you could consider rollups as well so for me the line is is far murkier than than it could initially appear and I think that it it just really depends on what underlying Economic Security they're um utilizing to secure the system because if they're using restate you could call it an ads but the underlying operational mechanism uh would be effectively the

same right and maybe like one easy call out is if you think the best version of your product only requires like doesn't need a state F sort of operator set right it's just a a stateless sort of network of operators doing a job and then you know um communicating back to to ethereum uh I think you know before Ian layer there was not really great options for that right it was like you could just have like a committee or like a multisig kind of you know running these running these operators but now you can actually get what looks a little bit like the benefits of a rollup right inheriting security from a much you know larger decentralized network but also just have a stateless sort of operator

set if that's going to be the best version of your product which it sounds like for Cella it is yeah I think yeah now listening to to to your comment here I I do realize I I actually do have a bit of a clear answer for for Mike's question in the sense that um when we're looking at rollups purely on the basis of okay this is a layer above a base a base chain and we're operating with this state that is effectively um interpreted through like compressed State on the basee Chain you have like exactly what m m was referring to these other requirements for various other applications that don't need that compression of state and they don't actually care about this idea of oh we're have our own custom execution

environment and then we're pushing State down to ethereum so then in case that something goes wrong we can interpret it or roll back State um for our example we're not compressing down state we're deciding what to do and how State on the base chain can change right um so I think that's like a slight Nuance where this idea of the ABS in that context specifically is far more beneficial because you don't really have the same requirements and you don't have the the burden of loss of funds or various other problematics that might um occur if you are a stateful um chain that might operate over a base a base chain if that makes sense yeah just to touch on that as well like I I very strongly agree with this idea of like if

you want stateful actions like a rollup is like a really good fit for this um and then I think the other thing to touch on is actually one of the things that that ethos is building as well is this chain using commitments and that chain will be a rollup um and I think it's important also to call out like what is the best way to secure that uh rollup I think everyone like agrees like decentralized sequencers will be coming eventually it's just a matter of kind of when and I think our perspective is actually that restake eth is actually an extremely good way to secure that centralized or decentralized sequencer um and I think it it boils down to like what is the restake eth actually protecting here and I think in the case of like reorg attx or otherwise um kind of inclusion uh Mal feance that you can

do at a rollup um these are much easier to arbitrate and otherwise like connect directly with an ABS structure rather than if you're running like a whole other chain there's like this this General mechanism where you'd want to like rely on social slashing and all these other things I think with a rollup because you have this um mechanism where you can have forced inclusion into the L1 um using something like restake e to secure the decentralized sequencer set makes a lot of sense um and so that's kind of why we're pretty focused on that um and I think in general I think it will be like the dominant model to at least bootstrap a lot of these decentralized sequencer networks I think to your previous question there will always be this question of like okay once you're bootstrapped you want to then launch your own staking token I

think practically speaking people like tokens so they're probably going to continue to do that um and then I think what will eventually emerge is some sort of like dual staking mechanism where you have your native token that is perhaps dominant and can be used for um perhaps even intercept slashing or social slashing but then you have for for most of the objective faults which are kind of most important here um you can slash with a large basis of uh restate these awesome guys wanted to get a sense um you know as we start to wind down here two more lines of questioning for you uh one is just on how are the basically how much to pay you know um so L we you actually started out with something pretty interesting as maybe like a floor for each validator which is opted into your auction as the amount that you

could conceive conceivably generate in terms of lever uh you know one time so maybe that's the floor I don't know like 10 million feels extremely safe based on what I know about lever but you know so maybe that that times however many auctioneers you want in the system but I think this is a big question about like how much stake am I procure how much restake do I need to procure overall and then what is the appropriate mix between I mean frankly today it's like ethereum or your native token but now igen you know is part of that and then eventually maybe outside of a platform like igen layer you could even get like usdc or something in the mix which I know that there are lots of technical challenges with that and what happens in the event of a slashing condition but frankly one

of the thing that's really attractive about it is that the value of collateral doesn't fluctuate right and so that was like a whole big thing but like how how have you thought about that like how much to uh procure in terms of like metering in a level and then which assets yeah yeah that I think it's it's a really hard problem because the tooling there is not you you don't have an outo box simulation that you can run and say okay how much how much e do I actually need uh for us the view was um and this is actually part of the reason why we we built brones which is basically an me tracing system that analyzes like occurrences of me on chain um was to actually figure out how much can be lost in the worst case scenario

how many valid like the main question that we thought about was okay how many validators do we actually care about how many do we want to have with within our network and on an individual uh node basis what is the maximum theft that can occur um and from that that should be the minimal stake requirement on a per validator basis in terms of how much you pay them in emissions I think that is really going to be a question of okay how high are the subsidies through ethereum staking Baseline and some form of igen layer incentive U and then after that period um kind of matures and the emissions go down how much do I competitively have to pay within this Marketplace to get the highest quality

of operators that I need yeah to to add to that I think um some of the things that we're thinking about L kind of mentioned like the lack of tooling I think actually this is like an interesting call out of just like there there really isn't a lot of great tooling here um I think it's somewhat application specific to actually model out like if we're thinking about Staker you you do need to model out like what is the potential profit from corruption so in Le's case like he's talking about meev and how we actually modeling that out in our case what we're thinking about is if if you're making a commitment to say like a co-processing result you need to basically model out what is the application value of this and and one of the core like um realizations from the Stak Shure paper is it's not the tvl of your application

it's actually this subset of basically like what is the within to get in technical terms like the reversion period of like what could be like reverted in order to harm users um and so doing that kind of calculation I think will always be somewhat of an application specific task like you will need to outline what this looks like and you'll need to opt into the insurance part but I think what's missing and could be generalized maybe as an interesting project to to kind of shepher forward is this idea even right now if you're going to buy even just base cryptoeconomic security from from igir you're not saying Hey I want $100 million worth of of eth you're saying like here's my inflation rate or here's my percentage that I'm willing to pay um and so there there's kind of like this weird like PID kind of dynamic where

it's like I'll I'll increase the rate that I'm willing to pay and then I'll get more security but then that's too much so then I'll decrease the rate similar to like the the um inflation scale for like an L1 asset um and so I think having more of like a layer of abstraction where any protocol and and this is what I think is like the really cool part of IG L and like the big unlock is that you have this programma um programmatic access to security um having more of like an open protocol where you can just directly say like my current TV or my current like total value at risk is X dollar and I then need to pay for that amount that also does um something you've touched on Mike as well as this idea of like not all of this is in E like some of this is in

different assets some of this is denominated in a able asset um and so doing that whole um conversion I think will be quite tricky and so practically the way I see this this working and this is what we're doing today is we're just gonna I don't know if overpaying for security is the right way to say it but we'll essentially have more restak eth than we strictly need um just for for extra security and then there will be efficiency gains as this gets um improved yeah yeah in terms of the generalizable tooling um and being able to communicate how much you need like how much you want and how much you're would to pay depending on the amount of liquidity that's allocated I think that that is definitely a super important thing that I I hope but from what I hear is focusing on um in terms of looking at

various assets as as Mike touched on that that does become quite complicated and now you have to take into consideration the available liquidity of those assets in the scenario in which there's a significant price fluctuation so in the beginning I think most uh of these abss are going to stick to one asset just because of the additional complexity that that these other assets bring in um but longer term I'd imagine a scenario where you're operating with a far more diverse uh set of assets that are securing your chain because the validators want different risk exposures at the end of the day on what they're holding I think this is like one of the most interesting sort of challenges and open design spaces um in reaking right now because we like at least now outlined like the types of Securities

that you maybe want right between like pooled let's say inter sub Ive something that can handle inter subjective work and then attributable which looks honestly more like insurance right um and then you have to you know there's another big challenge of actually being able to measure how much of each kind of security you need right um which like to your point is going to be super application specific um you need to find we need like a basically a lot of tools that make it easy for Capital to flow like more fluidly right to where it's needed at any given time hopefully like more programmatically than you know calling up an LRT provider and saying like hey we need more um right and then

you have to figure out how how to price it uh how much you should really be paying um and I'm super curious you think like in the end State the um abss will be paying more in Gross dollars for attributable Security even though the amount of security you know the like dollar amounts of attributable security are going to be a tiny fraction right of the pooled security I guess just from talking to you guys it does feel like you know ATT tributal is really the it's the highest value kind of security like nobody even totally probably can't argue that like insurance is a meme right um like it's actually very useful and so at the same time isolated Capital makes it really expensive right um so I'm curious

just to hear like how you're kind of Reason thinking avss will reason about the pricing side of this yeah there might be an interesting Dynamic where if you're a very high quality ABS that generates a lot of Revenue and and yield for your uh stakers that you can push for specific tranches of risk in terms of I willing to accept you as a Staker so long as you have restak with these other protocols that I deem safe enough um where the risk of having multiple uh slashing events within a s single Epoch that put uh my network at risk of not being able to um receive enough uh

or have sufficient collateral where it Des desensitizes an attack um Can Happen and in that context you might get away with paying far less for what looks kind of equivalent to attributable security but if that isn't the case I I definitely could see a scenario where tributal security is just that much more expensive and not even just on a on a dollar basis um I think we'll we'll really just see how the market emerges there for a long time I I I view this this Market as being mostly subsidized uh especially for State e that is paying so much yield right now um that I don't think that abss are actually going to be paying that much for a while and then

when we factor in i a incentives on top of that um I think we'll have a good time for a few years but then things are going to start maturing and the Market's going to get more efficient and things are going to get far more costly yeah yeah to that last point it does kind of feel like right now we're in sort of a like everyone's still dancing and like we'll see what happens when the Music Stops just because I think everyone right now is kind of um on this idea where it's just like oh there's going to be incentives for M and each abs and like you're just going to make so much money by being part of this and so I think it'll take a few as uh as you mentioned like probably years to see how this Nets out I think the one thing I

wanted to touch on in particular is this idea of um like you mentioned tranches of different abss I think that'll be really important to clarify and um otherwise calculate like what is the correlation of risk between different avss so the the most preent example I can think of is like an oracle AVS is probably like one of the most like in need of attributable security because you can imagine if you have an oracle um and then other abss are using that Oracle for other um categories you all of a sudden have like a massive red like Target on your back where if you just can attack that one you can get slashed once but you can um potentially reap like a bunch of different Rewards um and so I think it just becomes again somewhat application specific to figure out exactly how you want to um

categorize the security that you want to to pay for here and then yeah as you mentioned I think attributable Security will always come at this premium and frankly I think um it kind of depends on how these applications uh spec out but I think for our particular use case I think almost all of the the security that we use for the commitment layer will come from attributable security because it needs to have this insurance attached to it but you'll still have this like massive ideally massive um tool of Economic Security that basically makes the whole system safe and also is frankly like a um I'm going to hurt you really badly if you try and attack this kind of protection um even if I don't ensure everyone for that I at least want you to be like significantly more hurt that you're disincentivized to do this

yeah I think I think the Oracle one is going to be the most challenging right because there just so hard to track all the downstream sort of you know value at risk um but it makes a lot of sense especially when you know let's say it's like a co-processor or something like that something goes wrong a co-processor you know can't roll back the state of a rollup that's you know one of its customers right and so you just need something to be able to resolve but then I do wonder about like native stake in attributable security because if you want to redistribute the the collateral right they have to go through like the withdrawal cue that's probably not ideal um so that's maybe where like you match with lsts or usdc or something a little bit you know uh that that's better to

Ted for that for that type of security um so yeah there's a lot more to come here but guys I want to wind down by just trying to think through some of the second and third third order impacts of this new format of like an AVS and how that impacts applications you know historically you know if you wind back to even pre- rollups you were very limited in terms of B you basically could be a smart contract uh on ethereum right as you're sort of backend structure now with rollups that opened up an entirely new design space Al always you also always had like full stack app chains using like the cosmos stack things like that but now you've also got avss as well um and I would love to maybe we could zoom you know

three or five years into the future and what are folks maybe not understanding or thinking about how the introduction of this novel backend format is going to uh change the game um and you know we haven't really talked about pre we've talked a little bit about pre-confirmation this episode but it feels like this combination of uh like pre-confirmation play a story in here as well um so I'll just turn turn it over to you guys and you know what are some of the cool second and third order implications That You Don't See talked about a lot but you guys uh think is interesting I think the the world of defi is going to like go through an complete Revolution here because the like end of the day you could always

have a backend system that operated in parallel to your base chain to execute any form of arbitrary dogic but securing that was really hard bootstrapping the security of that was really hard now we have like a far easier and more direct mechanism to ensuring that Economic Security and as tooling gets better things are just going to get easier and easier on that front so may that be lending protocols dexes um per per Dees everything is going to go through this realization that wow I don't actually have to be as constrained in terms of computational resources as I previously was and so I can reimagine my uh application basically from the ground up

with this new found power in additional computational resources to do things far far far more efficiently um and so I think that's going to be like a huge huge Improvement for consumers in general and for the ux and and efficiency of defi as a whole yeah I think we were pretty strongly aligned there I think that's that's something that ethos is essentially hoping to kind of expose to more application developers is um rather than needing to build out like a particular abs for any particular kind of co-processing or otherwise commitment machine you're trying to build you can have more of a probably somewhat constrained but more readily available mechanism where I can essentially send a transaction onto the chain that says I

want to send an asset from this person to this person if this criteria is met and that criteria can be written in Rust it can be written in some other code and then that gets executed and then is enforced on chain and I think IG lay is is like or just race de eth is one of the core components to make that um viable and then I think the the other like part of the future that I would also want to highlight is this the story around interoperability and I think for us again like talking my bags a little bit is like I think it's important to have this credibly neutral asset in order to make something like this happen and so that's why ethos is kind of building the way that we are right now is so that we can have for this um

neutral layer where you can have commitments made between different chains all backed by the same asset and so I think that's where for us part of the future is ideally that we can bring pre-confirmation not only to to ethereum and I think we'd likely work with with some Partners there but also to other chains and have this idea where you can have um commitments made on ethereum that are then uh actually executed on another chain or vice versa you have assets held on ethereum that are then executed on another chain basically bring all these different worlds together with the idea that you have this neutral commitment space back by neutral asset I think that becomes is really important um there's a lot of like technical things to get built there but um that's kind of where I see this going maybe a bit of a a controversial

take here but I think that the idea of preon pre-confirmation right now um they're very important for interal but looking at the actual behavior of users on chain right now um the main really unlock that I think abss will actually have uh is less so the view of having pre-confirmation at the general protocol layer but rather at an application specific and again talking my bags a little bit here but um basically the idea that users really just in general want to do a single action or a set of uh very similar actions that interact with one protocol for most cases right

and so on that side it's actually far easier to have this idea of pre-confirmation at the application Level where you're a defi protocol and you have your abs now uh and you can easily precon um some swap or some loan um on your offchain state that is secured Again by this restate ethereum and then execute it when uh it's the right time to execute it when gas costs are low or whatever you want to optimize for right but given that this ABS now has control of that state um you can do that very easily and you don't really need these pre-confirmation on the base there obviously that is very limiting for interoperability um so that's a question that has to be solved and I think that

there are going to be these two emerging forces in the sense of this push for um Global pre-confirmation that operate across uh this state that is interoperable and these more specialized and um specific ones to Applications themselves and they kind of clash um but the Simplicity and the application side of pre-confirmation does give them sort of Head Start in terms of implementation because it it's just so much easier to implement and if the behavior of users demonstrates that they care 99% of the time about that and about being able to pre-confirmation Across the 10 different layers which I think is most often the case um I think we'll see more of a an incentive to

actually solve the application side of pre- confirmations rather than the base layer challenges yeah 100% agree and I think actually that's like compatible with with ethos Vision where we see a lot of applications building out into their own chains you see this like app chain D building out and I think that's where we are pushing towards there's actually a lot of customizability and Technical work we're doing right now in order to make commitments for applications that there are that they that are their own chain much more expressive as well as much more like rigidly enforceable such that you can Define your protocol by the fact that I've made a commitment and I must abide by that commitment it's like part of my protocol now I think that's like a very interesting and exciting um vision for the future the one part I would say I maybe disagree a little bit

is around like the user for this I think I agree that it will be a user but I guess in general I'm like somewhat becoming an intent Maxi I think intents will essentially abstract all this um and such sets like yeah user might only be interfacing with one particular application but the intent solver might be a much more sophisticated entity that wants to get all these different um commitments or otherwise guarantees across different apps and different chains and then use that to get the best Capital efficiency or otherwise like the best guarantees yeah I I think the one small difference here that in terms of what I was referring to in terms of like these apps specific um protocols is I'm referring to app specific protocols that live on a single chain that is an interoperable chain that is a like general purpose

chain and so we're really talking about getting the access to the underlying smart contract uh and less so having this app chain specific state that kind of isolates you uh into your own execution environment because as a swapper I don't really want to move to your own chain just because I have to execute on your decks except if you're giving me amazing Point rewards but apart from that like I'd much rather have my assets on ethereum or my L2 of preference so I think we'll see that more in terms of we don't want to burden ourselves the complexity of having our own chain we'll build on a specific chain but we will gate the access to prevent any form of leakage and and ensure that things are sequenced how we want them to be sequenced but I I do agree with you in terms of the

abstraction through intent that's definitely like a trend that we're seeing now yeah agree and I guess playful um like constructive uh feedback I would also think through is like I guess the the one thing the case I would think of is like if an application is giving me a strong pre-confirmation on like ethereum State like how do I know that they can actually like uh abide by that commitment obviously they can give me a commitment that it will be in the next like the next block that touches this application but but stopping say some sophisticated actor from just censoring all ethereum or censoring that particular application and so I I do like very strongly agree with you that like application specific proms are like very valuable I just think there's also this set of other things where you don't necessarily just want a guarantee that

it will be included in the application eventually you want it like now in that application you want it in the next block and because you care about this broader context and that's where I see like these these things interplaying but I do generally agree Jens this has been a ton of fun I'm sure we could have gone on for another hour and a half on pre-confirmation Alone um but new lwig this has been this has been great guys if folks want to follow you or find find out more about your project or whatever that is which you highly recommend that everyone does what's the best way to do that for each of you yeah please um I'm on Twitter at zerx Muto um feel free to reach out uh happy to jam on any of those I'm on Twitter as Ox van Boven um and otherwise sella is on Twitter as

well guys this was a ton of fun um miles and I enjoyed it immensely I'm sure the audience did as well we'll have to uh do it again sometime soon absolutely thank you so much for having us thanks guys that was aome awesome all right partner excellent episode there that was great to get load wig and Mew on yeah that was awesome and glad we're doing these like wrapup you know syns again um I think it's always kind of helpful just to reflect a little bit on on what the big takeaways are but yeah I think those are the two of the most forward thinking you know AVS builders that have been working really really closely um you know with their with the igen team and and building out their designs kind of know

exactly where um the space is going um and know exactly how they want to use right uh reaking to really unlock like the best version of of their product in terms of sella and um kind of like the highest leverage I think you know approach to building a generalized platform on top of um ethereum in ethos um so yeah yeah lots to get into but let you start off so I think one of the I would actually direct listeners as well about what I really wanted to get out of this episode was okay we know there's a lot of hype around this abs and how much is it narrative versus how much is it an actual design decision in real we knew it was real already but it was really helpful to just hear from the Builder

perspective because so often in reaking you're hearing from the supply side you're hearing from the staking side like the VC the marketplace side you're almost never hearing from the application Builder many such cases in crypto to be honest so the the first most important person that I heard talk about reaking actually was Zaki in the episode that you and I did with him and shiram about a year ago describing why at sellier he valued reaking so highly which was basically just the pain of not recruiting the validator set but keeping them so lowering the cost of capital for being a uh a a Staker to smelly was like super helpful for me to like a big light bulb moment and then one of the the other light bulb moments I had with was

listening to lwig describe obviously with something like an auction that he needs to run with Cella um the the sort of latency difference that was like a huge just light bulb for me so it was really great frankly to just hear um from Builders like the the feature sets that why it has to be like in ABS um and the unlocks that they get like that was a huge that was huge for me totally because I think yeah going back to that that episode a while ago we we started to you know really understand so the the cost benefits right it's just cheaper faster you know you can get get to Market faster and and frankly like it's also your re like internal time and resources like you are can basically

forget about that right you can just Outsource security really quickly and focus on your core product um and then I think today we got a little bit more into okay how can you use reaking you know from a feature standpoint right like security is security for for any chain um and it doesn't you know the operator said and the um you know the the capital securing it doesn't always necessarily play into the product itself um right but with Cella it was okay they wanted to build this product um you know they wanted it to live on eth L1 right but to date what you're able to build directly on eth L1 is super limited um just the

amount of like logic that can go into it what can be done you know onchain versus offchain is pretty limited and and Ludwig's point is there may not be as big of a need you know to move to a rollup or something like that if now you can have really like a lot of the the core logic um running offchain completely right in parallel with this like you know an IG layer secured or just a restake secured chain right um and everything can settle back down to eth similar to a rollup but you're just you know no longer forced to be a rollup so that was one of the most interesting parts of the discussion for me as well just asking Ludwig what is

the difference really from the app why why would you build as a as an AVS and as a rollup and I was a little bit surprised by how squishy the difference was in Ludwig's eyes like he doesn't see a really sharp distinction I guess the the big difference really has to do with how stateful the applications are because like really zooming out what a rollup is is like hey I want to build this application this decentralized environment but the computing environment of ethereum is way too expensive for me to do that so I need to Port over to this other execution and Computing environment and the way that you do that is you can compress a bunch of that state batch it and then submit it down to ethereum to make it much cheaper right now there's like 4844 and blobs and different pricing for those

resources but whatever that's that's the way that you do that U and so there's been this huge push from members of the ethereum community to hey I know you guys all liked being on ethereum but uh like let's move up right like we we all want the cheap gas we want the nice properties of ethereum being able to to Port back here but we want it in this cheap Computing environment and he didn't really brand it like this but Ludwig Ludwig's solution is hey actually you know what I heard him say was people actually want to be on ethereum lots of people out there probably don't want to get pushed up to a lay or two they want to just be able to swap and buy whatever meme coins they were doing before or or whatever um on the ethereum environment and he sort of found this this nice way

of you know the the auction doesn't need to be stateful you can do all that Computing offchain and what you kind of have as a user now is like I can have a lot of these nice properties that you get on a rollup but I can remain on this environment of ethereum that I prefer which I thought was super interesting because what are we seeing right now with rollups it's we're getting to the point where you know the biggest issue is UNF fragmenting liquidity right and all of them are really wishing that they at least had access to view ethereum L1 State then and en rollups also you know are all competing for like L1 liquidity but they can't actually access L1 liquidity right right so what if you can stay keep all the benefits of the you

know being like directly on ethereum in terms of shared liquidity and and having you know a view into that state all the time but instead of you know doing all the heavy computers a rollup you do it as something that's running in parallel as a restak sort of consensus uh either a chain or just an oper Ator set right so then you get kind of the benefits of in terms of like increased you know compute power and the you know kind of giving superpowers to these to these smart contract apps right um but you don't have to lose sort of all the benefits of shared liquidity and and uh and L1 liveness too right um yeah I think that's a big unlock you know where

this you know what it made me think of was I wonder if this OB huge benefits for ethereum right but I could also see something like this working on a chain like salana where the frankly like at some point right we're going to max out the compute on main chain and there's going to be a push to bring some of that activity to like an offchain Computing environment like there have been like drift I think or uh yeah drift has an L2 what is it yeah zet Zeta Zeta Zeta sorry has an L2 but it has a really taken off and I wonder if part of that is because there would be a big narrative violation for the salana community they they've really hung their hat on hey we don't we want to keep everything on one layer but an interesting interim solution would be

hey small subsets of um salana you know validator sets like maybe do this decks you run this extra side card to right you know do this this deck type stuff like that would be I could see that actually taking off in salana so you get and then you get and then it's kind of a Best of Both Worlds because then you do get all the benefits of being on one sort of layer and you don't have to doesn't feel as messy yeah totally totally and I think that's a a little bit of what ethos is trying to do um in terms of extending all of the benefits of like um of uh of pre-ms or any sort of like General commitments right Beyond just ethereum um because I think we talked a lot about like this idea of native reaking as a like a right you

know these use cases as feature additions to ethereum right so it's okay there's a bunch of activity happening directly on L1 um maybe we can be more expressive about like the commitments that proposers are making directly with the apps right on on ethereum L1 but I think there's you know inevitably inevitably going to be like a lot of multi-chain activity and not all of it is going to be you know settling directly back down to eth or having a direct connection to eth and I think um you know what ethos is trying to do is basically have this like sort of very neutral platform where you can make any generalized type of commitment between two different chains right it could be um like an ethos powered you know Cosmos

chain communicating with like an arbitrum you know the arbitrum sequencer right just the arbitrum sequencer throws a little bit of stake into ethos and you know now you can basically uphold have something you know backing those commitments um so that was interesting to me I I think yeah go for it what did you think miles of that discussion at the end between lwig and Muto about app specific versus more global precom yeah I mean I think it's super interesting I see like the puck definitely going towards app specific you know pre- comps I think this is similar to what we were just talking about on salana right or it's like you basically want to be able to um

you know get a little bit more value back right from the L1 be from the like the value that your app is actually generating um so I do see that absolutely playing out but I think that you know we're we're just really talking about like L2 pre-ms right now for you know kind of like Fast um sequencing and um you know basically getting past like the centralized sequencer option but I I do think yeah or it's it's at the end of the day I think we're gonna get more and more into this like app specific you know uh commitments yeah I think the point um of just the it's been very tough to nail down a timeline on when we could expect something like pre-confirmation in the

way that I think a bunch of the info providers on eth want that to be the case my sense is that it's closer to years it's definitely not months away it's probably years and I think the the implementation timeline on app specific versus Global Pay yeah play a huge part of the path dependence here in terms of what ends up I agree with you um yeah I think the other big um I guess interesting big takeaway from me is how much uh both of them care about attributable uh security and stake sure basically Insurance like it I think it honestly seems like you know uh both projects are going to find more value

right um from stake sure than pulled security and so I think it's you know people should recognize that we are in like V1 right now of what reaking can give you and uh maybe V2 is like the igen token or maybe that's like v1.5 but but once we get to Stak sure and like attributable security it's it's a very different ball game um and I do think we'll see reaking providers kind of diverge a little bit there in terms of hey some people think this is really really important and you should offer this within your product and others are going to say I think V1 is like really useful like pulled security maybe having it just be a little bit more isolated

instead of pulled with everything gives you those benefits um and the Crux is what happens when it gets slashed like is there a way to actually redistribute the assets to harmed users um because yeah I think if that's part of this you know question of Economic Security and whether Economic Security is a meme like I said this on when we were recording like I don't think anybody can argue that you know insurance is a meme like it's Capital that is going is there to actually going to protect users right in case anything bad happen instead of just slashing some random delegates Capital um but yeah I think that's something to keep an eye on because I think it's a big part of igen layers

road map obviously and I don't think anybody else is really working on it yeah so also maybe to maybe to start to wind down here the other part that I thought was very interesting to hear both their perspectives on was some of the moat that having your own like the some of the the the walls that igir can build around its ABS ecosystem and my my sense today is that it looks more saslik the mo and from from that I mean it's that's actually like powerful right like we told you the story before a million times we have the wrong CRM we made the decision five years ago it didn't really make sense then it certainly doesn't

make sense now but the pain of ripping it out is too much to bear we've got too much other stuff going on so it's just worth it to pay a little bit of extra money and have it not be quite as good because the disruption to the org would suck so that's that's one that that makes the product sticky where I think potentially I don't even know how they would do the equivalent of this but they could move from like a SAS SAS stickiness to a potential ecosystem Network effect would be the equivalent of what Apple does with their ecosystem where you've got your I your your airpods you know your Mac computer your iPhone everything just all syncs up it works really nicely and you know the trying to rip that out any one component

of the ecosystem out it's like just not worth it I might as well just pay you know and I think the question for Igan lay is going to be what are those components of the ecosystem that will build that will make it make my product even more stickier and graduate from this kind of sass ecosystem um yeah STI yeah I I agree with that um I think what can you do to basically like keep these even the most successful customers like on the platform um I think something that it's becoming very clear like at this point they can actually help Drive distribution for a lot of avss right it's like a lot of these abss are selling services to

rollups uh Igan da is gonna you know have a lot of rollup customers and it's about you know actually maybe probably explicitly like incentivizing Igan da customer to use like five or six other avss right and and maybe giving them some sort of like bundled package around that and the second piece is just um launching first party avss right that are complimentary to your thirdparty avss so here is a first-party AVS that you know gives like interrupt benefits between other avss right like maybe that makes these bundled packages even

stronger or maybe this is a first- party AVS that um decreases cost right for other avss um those are the sort of things that you know if you move off of V lay you no longer get those benefits right so if you look at you know the migrating off or graduating off of igen layer as something that is going to hurt your distribution right and it's going to maybe even um result in like new costs right or like lack of cost savings from some sort of complimentary service that Ian layer provides um that's the sticky stuff I think um so yeah we'll we'll see how that plays out but I guess we're gonna have to figure out what that is like one

of the one of the concrete I and I guess this is something that Igan L is goingon to have to consider from frankly their own you know Runway uh Capital allocation Focus sorts of decisions but what is the amount of tooling that they want to build on their platform versus what do they want at to just crop up as infrastructure around them exactly yeah um like frankly something to as a way for avss to meter out security and make those sorts of decisions if I was igen layer I I mean they've got a million things but I would try to build that because guess what I've got a token that uh I've got a token that I want people to use and so what better way to write the you know what better way to

incentivize to do that than create the framework and write the rubric for how ABS is should think about it in the same way that Google and Facebook have wrote playbook for how you should do attribution I mean it's genius what they've that's that's the real lock in that Google and Facebook have um yeah yeah I totally agree it's always challenging where it's like first party versus third party you know lrts are one example right where Ian L like could have built in LRT you talked to S about that like people thought they were and they didn't and could argue that Igan has you know a lot more tvl now because they didn't um and yeah I think there's you know all just answering this question of how much you should pay how

much you need and then how much you should pay is probably yeah like the could be one of the stickiest sort of tools that they can you know put out on the platform I can't tell how I feel about this this design decision which really pervades crypto of being really flat and neutral and thin um there are some instances where I love it like Lio tons of like minimize the surface area for governance be really thin really neut neutral aligned with ethereum but on the other hand you know leverage comes from solving complex problems and if you just decide I'm not going to solve any complex problems I'm gon to push all of that you're inviting someone to take leverage over you or to have leverage

over you and so yep I don't know General observation agree like um you know ethereum has the amount of activity it has today on L1 because of how you know neutral it was right but I would AR there's a lot of people I'm probably in this camp that in a perfect world ethereum would have um put forward you know an an inter op standard uh right before pushing out all the activity to rollups now I know that like in hindsight you know wasn't wasn't really realistic to do anyways but like now we have seven different AG layers you know popping up that are all trying to solve

kind of the same problem um and would have been great if there was like a enshrined you know ZK agay from that's built by ethereum now of course like that would be a little bit opinionated because now you have to build rups in a certain way that can leverage that right but that's the sort of trade-offs that I think igen will face too um and I think they lean more towards like being neutral and thin um and yeah I think I think that leaves room for like others to counter position potentially that are going to be a little bit more opinionated yeah all right partner anything else or we can we can probably that's it that is it yeah that was a lot of fun all right all right see you next week see you next

week hey everyone want to give a final shout out to this episode's title sponsor say now there are a whole bunch of really exciting reasons to be building on say V2 outside of just parallelization I want you to head over to say. to looking into building on theu devet again click the link at the bottom of this episode and head over to say. start building something today