The Bear Case for Shared Security | S8 E8 | Sunny & Toghrul

Transcript

I think an important question is here is not how much security is necessary but how much it costs for you to acquire the security because let's say if it cost me $10,000 per year to acquire hundred billion dollar worth of security then I might as well get it like it it doesn't really matter like if if it's super cheap for me to acquire that security I I'll pay for it like it it doesn't matter the bigger the number the better one only from the marketing perspective even if you ignore everything else the bigger the number you claim your security as the better it's going to it's going to sound and all the podcast Etc that you go to hey everyone this episode is brought to you by namada the shielded asset Hub rewarding you to protect the multi-chain so not only can

you protect your own data across any asset and any chain but you also get shielding rewards for strengthening data privacy protections for everyone so you can check them out at n.net and follow the domain net launch stick around you're going to be hearing all about them later in the program hey everyone this episode is brought to you by real a new permissionless L2 built on the arbitrum orbit stack by Define natives for Define natives for tokenized real world assets it's very cool chain highly recommend that you go check it out and they just started their Rewards program so you can type in re. rewards you can use the code in the show notes below all right thanks everyone home the show 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 BCR today I'm joined by my trusty co-host for the season miles um and today we're going to be talking to Sunny of osmosis and tog from scroll guys welcome thank you for having back this my third time on I think do I do I have one appearance per season almost I think you're I think you're Vine for the title Sunny of uh

most included yeah as a b curve guest it's good to have you back on um and uh yeah guys uh really excited to get into this one so we've been very kind uh to reaking as a category this season talked to a lot of different reaking marketplaces and lrts and avss about why we should all be excited about this and we wanted to use this episode actually as an opportunity to maybe Steelman the be case or maybe poke a little bit of holes in uh reaking or shared security and and why it might not be quite as exciting as as everyone might think it is today and uh too I'm actually gonna call on you first here because you um were the author of a a recent speech at modular day the title of which is shared security is a scop so

very strong title there uh maybe you can give us a a quick tldr of uh you know what you discussed in that talk yeah of course uh before before I get into that I would like to say that I actually don't think shared security is a s up I was just trying to be a bit spicy with the title John Sharp when inspired me to make more spicy titles when I make my talks uh so so I I think uh a lot of the hype around reaking and is centered around this idea of shared security and like oh for example with the IG layer oh you get the security ethereum or you inherit certain security properties of ethereum Etc but it's a bit misguided and it's also like a bit

they're overselling what they actually have BEC because uh what igir provides is economic security and you can say that it shared Economic Security with ethereum but but ethereum security is separate from the economic security of ethereum so the economic security is enforced by the capital state whereas the protocol security or like the security of ethereum itself is enforced by the full nodes so let's say even if you manage to take over the consensus of ethereum and let's say 67% vote on a uh on a head of a chain that has an invalid block any honest full node will just reject it so let's say if honest if coinbase runs a full node it doesn't

matter how many people vote for an invalid block coinbase will just reject it and so and so it's almost like a layered approach where Economic Security is your first layer of security where certain things rely on and then if Economic Security fails you fall back to protocol security which is enforced by full nodes and then if that fails you fall back to social consensus or Social Security where essentially you can go offchain into Discord or different forums or like conferences whatever and discuss how you can fix a certain problem so think of like the Dow hack for example that was a fallback because because the protocol security couldn't do anything in that case so ethereum

fell back on the Social Security and used the social consensus in order to enforce an upgrade that would introduce an a regular State up State update got it so um maybe tole the slightly more diplomatic but less catchy uh title of your talk could have been uh shared security is getting misappropriated and there are different levels to this actually and what we've been referring to mostly As reaking actually refers to the economic uh security layer of the of what you just gave us maybe uh son you can kick it over to you here because you know I know you have some some strong opinions on the idea of shared security as well and you know you're very active in the cosmos world where one of the earlier iterations of shared security or replicated security is actually live and

and in production so can you kind of give us your just highlevel um thoughts on shared security as a concept how well does it work in Cosmos and maybe comment a little bit on um toe gr framework yeah I think uh maybe I have a set of like hot takes on on on shared security and I don't know like maybe throughout the course of this we'll figure how to tie them into one cohesive like take but I mean a couple to just rattle them off like one uh things I think are being pushed to have more Economic Security than they usually need so I think like there's this idea like oh everything needs to be like as secure as like base layer ethereum and I

just think that's like not necessarily true right like you know you should figure out what is the uh cost of attack or no no what's the what's the benefit of an attack right like what Can you steal or what gain can you make from an attack and then Target to have some economic security a cost of attack that is higher than than that benefit right and so you know I think uh there's many ways you can think how how to measure that but like for let's say an app chain like osmosis like a very simple way to measure that is just at the value of the bridged assets on osmosis right and say okay hey we have you know $300 million of bridged assets on our chain let's target something that's like 25% more of

that we don't need billions of dollars of Economic Security as long as you have a way to scale up Economic Security as uh as needed uh so that's like maybe take number one take number two is I think a lot of the economic security why people want more Economic Security right now is um economic like all of this reaking and everything is less about Economic Security and more about political alignment signaling uh and so that's probably like why people like oh I want to use igen layer it's not they actually care about the E they care about ethereum alignment or oh I want to use Babylon not because I care about bit the amount of stake economic St security I just care about like signaling Bitcoin alignment right so I think that's take

number two and then take number three which is like you know the farthest on the nihilistic spectrum is just like oh all of this is kind of stupid and like Pro proof of stake itself is like fundamentally broken uh because you've basically gotten into a system where it's like major like almost all major chains are like being run by the same set of like you know probably 15 to 20 like operators and given a lot of these operators the biggest thing is almost all of these operators are not actually having their own Capital at stake right it's usually some sort of delegation whether through like an LSD in which the delegate people delegating that Capital are barely even having any

say into which validators or through D direct delegation where they have a little bit more say but like at the point that it's like not really their Capital maybe the economic security is not actually uh what matters and there's you know we're kind of running more of a proof of reputation system anyways and so maybe we should just accept that and like figure out how do we design a system that like a better proof of reputation system than like keep trying to focus on this like proof of stake that isn't actually working in pract or is not you know we're deceiving ourselves and thinking we're running proof of stake when we're not really yeah I I actually wanted to almost end this conversation on that but let let's dive into that and in addition to uh being very popular and having this be

the bare case for shared security we can have this be the bare case for vanilla proof of stake as well um and I I think one of the the almost assumptions of a lot of uh a lot of resta today is built on the idea that these avss need to have a permissionless operator set and I I I think it's probably worthwhile to just question whether or not that actually needs to be the case right or something like a proof of authority type system could actually do the job much better so me miles I'd also love to kick it over to you here because I know you and I have been chatting about this a good amount like what are some of the issues that you see like just at a high level with like vanilla proof of stake and then maybe we can get into how

applicable that should even be for something like realistically a lot of these things are kind of like middleware right yeah yeah totally totally and I think like um just to touch on that piece right now with permissioned versus non- permissioned operator sets um first of all it's worth noting that all avss today or at least all that I'm aware of are all permissioned right in terms of their operator sets it's yeah these are like baby steps as we kind of go into production but it's okay there's like 200 operators that have onboarded Dion lay here's the 20 to 50 that you guys have probably heard of they're running all these other systems go talk to them right and I in that sense is nice because it's just kind of

aggregating these operators in one place and making it easy for you to go get them um in Cosmos it's not that much more difficult but it is a it is a bit of a process right um then they come on board and you have a different set of actors stakers right delegating Capital to them uh you know to kind of repres present the actual economic stake right and then the question is okay if you know who all these actors are in the first place right you can if they do anything wrong they're they probably have legal liability right it's like the probably the strongest form of slashing it's like they'll go to jail if they do anything wrong um why are we bringing

economic stake into this right because it's not actually the economic stake that keeps them honest it's fact that all of their future cash flows are based off of their reputations and that's you know and or you know not going to jail is also uh you know because they're all docked right um so why are we really doing this whole economic you know Security on on top of it I think that's actually very legitimate to question I think it's only really makes sense when you say no no no this is just a like a baby step right we're going to make this operator set permissionless then you could have these you know Anonymous operators coming into your set you don't really know who they are right you can't track necess you probably could but you

you don't want to create a system that relies on being able to send these people to jail right so then you want something actually you know increasing the cost right or increasing the friction of an attack because you don't know who these people are we're not quite there yet but that's where I would say like Cosmos chains are today right anybody can actually enter those sets so it's good to have some sort of economic stake back them um but I think that that is a big part of like why this point in time you know you can really make a strong argument that that a lot of the economic security here is is overhyped um I have you know my own issues with with like delegated proof of stake mainly around the fact that when something does go wrong it's not the

people that did something wrong that are getting punished right um or something like with DX right where we can't actually you know have a programmatic way to s for something like me which is actually far more common and more of like a threat to ux and you know users than say a double signning or something like that right so like the lack of flexibility of what you can enforce and and then who's getting punished at the end of the day um so yeah that's a bit of a mini ramp but I don't know if you guys would push back on any of that or if you you know kind of are more or less aligned uh I think I think there are a few things here so so I think going back to to uh to proof of stake just vanilla proof of stake my my problem is that and

and I was part of that camp I think like four or five years ago I think if you go and look at my old tweets you can still find me saying completely opposite things to what I'm saying right now it was a bit naive to assume that people are just going to Hom stake the majority of the people so so I think to a degree the argument that all proof of stake systems at some point deg great to delegate of proof of stake system uh means that you should probably build a system that is that has delegation and shrin in it because it's unrealistic to expect everyone to run validators at home and be like and be and ensuring that everything runs smoothly they don't

have some bugs or anything and like because I like if you look at historically most of the slashing events in ethereum happened because people were running two validators for redundancy and then they vote on different heads and they got slashed and so yeah so I think I think with that regard I think it's better to just avoid building vanilla proof of stake if you're building just build delegation based proof of stake and in terms of proof of authority and like the permission then permissionless sets I think I think uh it depends on how you define permissionless and permission uh protocols so so

certain people Define a permission proof of State protocol as the one where others use their tokens to elect the validators and there's also permissioned where basically you get proposed a certain set of people who you who can run and nobody else can register as a candidate so I think I think if you have a permissionless way to register as a candidate so anyone can can register and say I want to be a validator delegate me some money then I think uh Economic Security makes sense because it your tokens almost act as like uh that they replace uh your identities like one token one vote essentially and it's a

popularity contest and the way you you select the most popular validators is is through your token and in that case um sorry and in that case Economic Security makes sense if the if the cand are permission then I I I don't see a reason why you need Economic Security you might as well just have someone elect them for you and just trust them and so yeah and in terms of slashing we had this conversation with sunny after our panel at modular Summit I I'm I'm not sure if with delegation you need slashing or not it's like a difficult question because from one hand uh obviously slashing helps because it adds

a big disincentive for people to actually behave honestly but on the other hand people who are getting punished are not the people who actually misbehaved uh or like the big proportion of the funds that are getting slashed but at the same time that disincentive creates an incentive for people to vote for people with good truck record and so it's a bit of a mixed do do we want to incentivize people to just YOLO into any validator that offers the highest API or do we want to incentivize people to vote for to delegate to people who have the best track record historically I guess that assumes some amount of due diligence and Intelligence on the perspective the part of the user which

not to be cynical is probably not a good thing to assume and I do think that part of the uh the the way that shared security or reaking is getting implemented today it what you're talking about there is principal agent problem and is you know if you had like a sliding scale of how serious that issue is maybe there's none of that in proof of work um or you know proof of stake as it was imagined initially in the early days of ethereum but as soon as you had the ability to delegate there's some amount of principal agent problem there and then as soon as you add reaking on top of that that it just compounds it right and however much financialization that we end up adding onto that process just makes that issue worse and worse right yeah um so yeah I I I think I I

kind of agree where it's like you know I think maybe I have more of an issue with the slashing part more so than the using vote like stake to like vote like I even in a proof of authority system what I really mean by proof of authority is like okay the current validator set votes in and out new members but the token holder set also has like some governance capability to do it or maybe it's like some sort of like two house system so it's like you know you still want some economic um you know you still have like the token holder set voting in and out uh new validator node so you know doing like whether it's uh you know oh you have 51% you vote in and out or you say oh it's like based off the

amount of tokens you give this much voting power these are just like different voting mechanisms I think maybe what I'm more talking about is like the slashing seems a little silly in the world where it's like no one's actually doing due diligence on these validator sets I think uh I I I I said it on Twitter like a few few months ago and I got a lot of slack for it my biggest problem with slashing is actually not that it doesn't add any any additional security I think I think it it'll be silly to argue that there's no added benefits to slashing is the complexities that slashing creates like think of how many centralized interrupts we got on ethereum because of slashing like like lsds could have been

enshrined into the protocol etc etc certain things could have been made much cleaner if we didn't have slash hey everyone this episode is brought to you by NADA the shielded asset Hub protecting the future of the multi chain NADA is a chain that's near and dear to my heart because in previous seasons of bell curve we've talked about how the Cardinal original sin of most blockchains is not thinking about privacy from the beginning but also because if we want to onboard a whole bunch of applications it doesn't work in a default transparency setting so the MAA not only brings about a better future that we all want but is going to help on board a ton of new applications so pumic introduces a novel shielding reward system the way that it works is if you hold your assets in the shielded set you help strengthen the mod's data

protection guarantees and collect nonrewards in return the model will initially support IBC and ethereum based assets but eventually the idea is for this to expand into a comprehensive hub for all assets really really excited about what the folks over at NADA are doing I highly recommend that you go and check them out at n.net you can follow their mainnet launch which is zoom TM very exciting tell them that I sent you click the link bottom of the show notes hey this episode is brought to you by real re. they're a new permissionless L2 built on arbitrum orbit for tokenized rwas so basically they've got this portfolio of real world assets that actually exist out there that generate yields they import them onchain and find ways to reward their users very

creatively uh so one of the cool things about real is that it's the first L2 where 100% of profits from transaction fees and fees from protocols deployed onto the chain acre back to the L2 governance token and currently I mentioned this at the top of the episode but they have a reward pray program going on for early users of the chain where 10% of the rdwa tokens are being distributed which is very cool so you as a user can go over there and provide concentrated liquidity trade borrow and leverage tokenize rwas including T bills real estate and tokenized basis traits so again head over to re. rewards and you can use the code found in the show notes here tell them that I sent you thanks real for making this possible can you say more about that Togo why why does slashing what

explain that relationship between the existence of slashing and not in shining an LST uh because if you have an LSD uh that is fungible in between every single validator then if one validator gets slashed everybody gets screwed so and and if you avoid that then it's still it's still a bit of an issue let's say if the validators offer different apis Etc then you still have to design it in a way that is clean but but but assume that you don't have that problem and every validat on chain offers the same API then uh then you don't really have a problem you can just enshrine LSD into the protocol and that's it right and you

can actually end up with potentially a a stronger system in that sense you by picking a methodology right to distribute stake and thus distribute voting power and and you know block proposing power right um the only thing blocking that right now is I think togl you hit it is that if you socialize slashing across the entire set it's going to be a large portion of that of stakers that aren't happy right that some of their delegations are getting sent to validators they've never heard of right they would just much rather they'll never get fired for choosing like coinbase Cloud right on behalf of maybe their client something like that right um but if you kind of go that

route I guess what is the point you know Economic Security then right in a like an inin LST like sort of system is it just to raise the cost of attack if there were to be something so it's not nothing it's not that raise the cost of attack so I think uh I think Justin Dr had a good framing of it where it's like there's the cost of attack which is like how much you will lose and then like iot what the word he used like the you know the barrier to attack where it's like okay you have to get this much once you get this much it's not going to cost you anything because you can do the attack for free but it's like oh so I think like I think Pro stake system should have a way of staking that does not involve slashing right so that's way you

can maximize the like barrier to attack right and then the actual slashing I think you know it's only worth doing it if like it's up to the if it's the validator actual uh stake that's being done like put down so like you know in my opinion like a better way to do stuff would probably be like hey every validator has to buy a like validator slot right by like putting down like stake right they say oh okay here's 100 you know each validator has to put in 100K to buy to buy one validator slot but then like it's up to governance has to like token holder governance can vote okay do we accept you becoming a

validator because it's if you think about there's not that much benefit to having like you know coinbase run hundreds of validators like no okay you run one validator okay you and then they have their own economic state that that's actually slash and I think there's actually ways of taking that economic stake and making it um you know even more so this is like a little bit of a silly blog post I wrote like four or five years ago you know I will preface it I do not condone condone this but I was thinking about like okay how do you magnify the like value you know the slashing of validators BAS off a limited economic stake like oh you put down 100K and it's like sure you can lose that 100K it gets slashed but what

can you do something more and I had this this BL post like oh you could take that 100K and open like an assassination Marketplace against that validator okay obviously that's a little ridiculous but like I do think there's ways you can instead of just immediately burning that economic stake of the validator right you can maybe somehow put that to use in some other way to like harm the validator to increase the actual like cost cost that they have of misbehavior and maybe that's just social maybe that is like in in practice more just like broadcasting to as many chains as you can that this is a bad operator you know exactly I think like what these reaking things should be doing is that like hey if one operator gets slashed on

one not should they just use that lose the cash flows on that operate set right they should like it should Auto broadcast to all the other sets that it will like you know probably remove them from or at least highly suggest to them to remove that validator I I think the problem with reaking protocols like I I think if you if you're talking about base layers so ethereum or like salano or anything else slashing makes less sense additionally because you can always fall back to social consensus so let's say if something goes wrong you actually don't need to enforce slashing on chain you can socially slash or socially reorder to be honest this is what salana does like salana does not have automatic slashing Avalanche does not have automatic slashing

yeah yeah yeah like I I a lot of bigger ones actually don't have slashing it's ethere that mostly has slashing but but my point is a bit different it's you can always fall back to that yes it's messy yes it's going to take time but you can you can always resort to that the problem with reaking you cannot do that because because it's it's enshrined into ethereum so in order for you to actually to be able to fall back to the protocol or or Social Security you actually have to convince the ethereum community to basically back you up and do some social slashing or I don't know like a reorg on the ethereum side and so then it makes

more sense to have slashing for reaking protocols because you only have Economic Security of course you can just Fork away and say screw ethereum we're going to be by ourselves and screw this security but then what's the point of rest staking protocols so so so and that that's why I think in in in case of free staking or staking protocols in in in this context it actually makes sense to have slashing even though there are disadvantages obviously to it let me I'll add one thing where it's like okay the economic like to be more specific it's like the claim is is not that Economic Security doesn't matter I think it's more that Economic Security matters a lot less than the actual decentralization of the operator set

right so you know to to take a counter like you know let's take another example of where like you kind of have a proof of stake with slashing system which is like on these like cross-chain intent systems right you you know you have an act you you want to do an action you know um you send uh a market maker eth on one chain and they send you eth on another chain right but like you want that market maker to have like economic stake bonded that is slash because of two things one that market makers can often be anonymous and two it's a single person that's like has to do an action and that's when it's a single person that's doing an action their Economic Security actually matters a lot right

but then when we're talking about consensus protocols those are you know the number of people that need to collude in order to like do something gets like should be higher right and like you know this idea of like oh a conspiracy is harder to maintain like as as it grows right like you know a conspiracy among like two or three people is easy but like a conspiracy amongst 20 people is harder a conspiracy among 100 people is even harder and like basically to make a consensus level attack work in any like modern consensus protocol you basically need like conspiracy of at least like a third of the validator set um and so like and and realistically it's usually more than a third of the validator set and so you know the point is like hey instead of

optimizing for Economic Security uh we should instead be optimizing for validator like validator set size and this like pure direct proof of stake with like slashing of the delegators is like is actually antithetical to optimizing for validator set size yeah I I was just wondering when you when you say optimizing do you mean increasing it or decreasing it uh increasing it you want to increase the amount of people that you have to want to increase the amount of people that have to collude in order to attack this system yeah because I've been hearing a lot of arguments lately that you don't actually need a lot of people to participate in the consensus you can have like 50 and it's

fine so that's why I just wanted to clarify maybe 50 is okay right like I don't know like okay so onethird of 50 is like what like 17 or something uh so it's like okay maybe that's enough to uh that that's maybe that that is what you need that that's high enough to barrier to collusion I I think it depends on what you're trying to do right uh as a protocols as well like there's probably I mean probably much more impetus for ethereum to have a lot more decentralization at its valid than than than an application layer right and I think what we're what we're driving towards here is I would love to get both of your guys thoughts on the market size for like reaking as it's conceived of today because here here's the here's the

sort of challenge when I look out into the world and I think the way that a lot of people still think about reaking especially the way that I was thinking about it going into this season was hey you basically rehypothecation some of the ethereum um uh ethereum's staked eth right uh so you lower the cost of capital for that and then all these other like l1s or app chains or whatever get to uh bootstrap security but then when you look at the major categories of who that might be it's like well is an Al does an ALT L one really want to use restake eth like maybe um maybe but there is this like weird squishy sort of alignment sovereignty even even just like brand perception issue like does an Al one want to be seen requiring eth to bootstrap like probably not if they're

in it for the really long term and then okay well applications right like these middleware like app chains things like that maybe maybe they're going to ultimately end up being users of this well and that's that's where the proof of authority argument is actually very relevant because suddenly if you have a permissioned operator set oh do I really need this restake security it's kind of nice for the headline but like how much good is this really getting me so you kind of go and then even for this is the the sort of Symbio view of the world which is there are these big blue chips maybe it's Athena or maker or something like that and there's a reason for they're trying to decentralize um some internal components of the protocol but you know maybe I can use my own token for that right so you know initially

this enormous Market size kind of looks a little smaller when you dig into the categories and I would just love to get a sense of you know order of magnitude here I mean how how big do you see the the opportunity for sort of reaking as we've been defining it today yeah I mean so this kind of goes into my second Point what that I started with like I think it's mostly about political alignment signaling I think the market opportunity right now is like you are you know I think ethereum you know a lot the larger crypto like communities they're more than just like they are communities right and so it's like okay you do reaking with aan L to just like signal alignment with the ethereum

community in order to get like you know podcasters to talk about you and like get like integrated in you know people in that ethereum uh social like Circle to like promote you right I think that's really the market opportunity I think that's probably the case with Babylon as well you get like a bunch of bitcoiners to promote you um long-term though it's like I I agree like it's I think what's going to happen is people will do like you know Target minimum VI viable alignment right it's like I will pay the absolute minimum possible in order to like nom claim alignment but like in reality any economic secur like paying anyone costs money right like you have to pay these eth stakers or these BTC

stakers and it's like you know you probably want to you know if you have a protocol that's actually earning Revenue right you want to like internalize as much of that as possible and so you'd probably just try to use your own uh token as your staking token and pay that as much as possible right so it's like you'll be like oh okay yeah I'll give like .5% of my staking rewards to this thing just to claim alignment but like really 99.5% of your staking Rewards or or Revenue right ideally it should be revenue is going to uh your own uh stakeholder said I I think an important question is here is not how much security is necessary but how much it costs for you to acquire the security

because let's say if it cost me $110,000 per year to acquire hundred billion do worth of security then I might as well get it like it it doesn't do really matter like if if it's super cheap for me to acquire that security I I'll pay for it like it it doesn't matter the bigger the number the better one only from the marketing perspective even if you ignore everything else the bigger the number you claim your security as the better it's going to it's going to sound and all the podcast Etc that you go to uh so I think from that perspective actually I I think babone is in a better place than ethereum I lay and that's primarily because with with with ethereum you already have a way

to to earn uh yield on your eth with minimal risk through staking in Bitcoin there's no way you can earn yield with minimal risk and so Babylon introduces that opportunity where you can earn yield with minimal risk and I think with babyon you don't even need to pay that much for people to actually opt into that assuming that everything works as they describe and like it's it's extremely secure Etc uh whereas with ethereum if you're already earning like I think at the moment it's like three and a half per yield per year on on staking connecting to ig lir after all

the like all the points Etc like stop stop being emitted how much are you realistically gonna earn probably not that much extra at at additional risk so I I think I think that's why like co-opting into something like iir would cost you more than co-opting into something like babon but yeah but the question is just like where is the equilibrium how much uh how much minimum yield people are willing to accept in order to take this like even if it's minimal risk the minimal risk and that will determine how much security is actually necessary that I guess if you were to point to like um you know one of the

harder I guess uh value props of reaking to debunk right now is cost savings I think like if osmosis had a huge inflow of bridged assets and wanted to increase the amount of Economic Security you know from today tomorrow from like 300 million to a billion um you would need to pay more of your osmo supply to get that in osmo than you would need to get that fill that Gap right in let's say restake Bitcoin um and you brought maybe you need to pay like less to the Bitcoin holders because they're earning no yield than you do to say like an eth holder right that has some maybe more opportunity cost or something like that or is already earning yield um and I I

do want to get your thoughts on on that piece of it right now because I'm talking to like definitely some Cosmos Builders where let's say osmosis what was in the beginning of the supply schedule is supposed to be I think 20% of osmo supply for staking right um and let's say 10% of that is can be considered like you know ecosystem dead weight loss just be via like commission right operator commissions getting sold so really 2% of that osmo is for sure getting sold some other portion is probably getting sold but could also be looked at as as you know uh rebasing right um versus with Igan maybe and Babylon it's it's looking like you know

rough benchmarks like that 2% which was sold via you know operator commissions maybe is your entire cost for that same amount right and is that the way that you are looking at this as well where it's you know this is at the very least a way to get a higher number of Slash you know economic value um or less like I think if that's probably the hardest thing to argue against here um even though there's lots of other reasons to be skeptical yeah so okay operating in the current model of proof of stake I think yes that makes that that that sounds right um and this kind of goes back to like a

combination of my first and second points which is like one you should know what your target security is and then two you should use you know your stakeholder set is going to be more aligned than external parties and so the example I like to give is like okay you're trying to build an army you start with like your own citizen right that makes up the bulk of your army that should make up the bulk of your Economic Security um then in cases where you need assistance right oh the amount of bridged assets on osmosis like spiked right we're like oh okay we need more Economic Security right now right so then it's like okay what do you go to next is like your allies right you say like oh okay uh let's go you know Axel

is one of our closest Bridge Partners right let's get Axel Axel tokens to be re St on osmosis to provide more Economic Security then finally in your army if you like really need the like final boost of security you you bring in Mercenaries right and I think the Bitcoin and E stakers are mercenaries they don't really care about your chain that much or the success of your product they're just here to like you know they'll provide the security they'll probably dump the rewards immediately uh to compound their Bitcoin or eth but like it it it'll uh give you some you know it it'll it'll close the gap until you can bring figure out how to bring in more of your native uh Economic Security so I think that all makes sense in once again this is all in the model where

we're still like sticking to this idea that slashing actually matters and all this stuff right like if we if we go back to like the you know maybe none of that matters and I don't I don't know if that like the getting that more Economic Security actually impacts anything one argument I would say that I have against the model of like giving your native token to let's say Bitcoin state Akers or E3 stakers uh is the fact that that only creates sell pressure so it might be cheap initially so let's say it's cheaper for you to to do those emissions initially but at some point because it only creates sell pression and there's no buy pressure there created uh the cost of your native token

is going to drop so you'll need to increase the emissions and at some point you'll get to a point where the cost of paying to the external uh staking network is going to be more in percentage wise of the circulating Supply versus what you would have had if you had native staking so I think it's useful for bootstrapping the network but I don't think for a lot of networks it's useful as a long-term tool do you guys think that there is a chance that for something like and this is the other by the way challenge with just as we you know airing all of the challenges with shared Securities there's an adverse selection problem right where you and you see this a little bit in in Cosmos

especially where it's like well you know I don't really feel like I need to be a consumer chain so I'm not going to be um there there are very high quality like stride is a good example actually of a very high quality um consumer chain where there's more of a deep strategic rationale for them doing that but generally that's kind of the challenge I could see that in in ethereum land as well where let's just Envision some hypothetical future hopefully it's not hypothetic for too long but there's an app with 10 million users you know it's like I don't need any of the stuff that you're describing really and it's uh maybe more of like a a middleware solution that no one's really heard of that uh like needs help in the beginning and so you know one one of the challenges I think with the shared

security model is that it it it lends itself to being a little bit of like a bootstrapping mechanism that people use in the beginning and then it's almost like SNL right where like people launch their careers there but then they never really stick around most part no yeah what what like actually keeps them on the marketplace longterm um I think that's one of Ian Lay's like big questions and challenges that they're thinking a lot about right now like where can you find some synergies right or like you know um Beyond just go to market and marketing um but it does like bring me to this question of are there ways that reaking can actually improve the core product and the core value prop um and that to me is where like I think this insurance idea comes in right or

Stak Shir where if you are paying for something that you know if anything goes wrong is going to get actually redistributed to your users to make them full maybe that actually is the more compelling long-term play here and reaking today is just making us think for the first time actually really hard about how much security we need right and what we're getting from it and then that leads us to the conclusion that we don't need we need less but we need to do more with it right with what we have which is actually you know reimbursing users if something goes wrong I'm curious do you guys see that playing out you know yeah I mean I think this this kind of what I mentioned before of like you know the idea of burning the steak is kind of silly right

this probably like more useful things you can be doing with that stake that's being slashed right and like an example is actually like uh on osmosis we're working on this like novel fast bridging method uh but like and we're g to let valid or validators basically or use their like osmo stake to like enable this fast bridging but if it if they like mess up or something their state doesn't just get slashed right it's actually used to like compensate the user that was like that was harmed so I I think like yeah you should be actually figuring out smarter things to do with that steak than just burning

it and I've got a qu maybe Togo I want to te this up for you as well because we've been we've been um pushing on shared security as an idea a lot this episode but I think it's actually some of the challenges like what miles just described like the adverse selection problem people wanting to leave the ecosystem it's probably a relevant question from your seat as an L2 as well right I think there's a lot of questions around like hey we want a bunch of app chains to come and build but then once these apps get a lot of users what's to keep them in the environment and you can see like a bunch of different L2 Stacks asking that question maybe we can make these apps interoperable with one another or like whatever it is but I feel like it's a similar sort of

question so I'd be curious how you've thought about that from the L2 side of things and then how you think about keeping apps like in an ecos system in general I I think I think uh I genuinely think the the longterm uh vision of how you can keep the apps in the ecosystem because as soon as the app grows big enough and also for a lot of apps synchronous composability is not really a selling point like think of like also it's not really it's not really a problem right now because all of uh most of the popular apps right now are mostly defi Legos so they're dependent on one another but and I'm going to use this buzz word let's say we get get Mass adoption one day and like we reach this

like 1 billion users most of the apps are unlikely to be defi apps they're going to be more like Normy Centric apps I don't know nftd based ticketing I don't know certain apps that don't really need interoperability and that just creates more reasons for those apps to basically separate and launch their own chain especially if they have their own token that creates an incentive to create more utility for the token because now you have staking you can secure the chain uh your app chain via your own token Etc and I think the way to partially solve that problem is via I I I called it modular block space which basically allows you to separate block

space into uh into chunks that that each app controls and then they can Define their or ordering rules they can Define their saking rules and Order however they want create a batch and then add them into this picker block that way you can create an incentive for people to actually stay within your chain and not go elsewhere while also uh while also basically creating utility for for for for their ecosystem and their token to actually capture some value rather than giving all the value to the chain that they're deployed on and by the way if you if you want to ask how it's going to work I have no idea it's just like something theoretical I was like I was

about to I was I think we go down we could go down a rabbit hole there um but yeah I I I pretty much agree with all of that I think it's a slightly different question than like you know what keeps apps in for reaking specifically I think that's kind of a broader you know how do these things actually grow and evolve and is there a typical life cycle to what this looks like I've really enjoyed how few people actually know what poly Market's built on and M you know the fact that they don't have their own chain yet it's like okay well um when do these things really matter right um and maybe we'll see that play out over the next year or two or maybe we don't but it is slightly tangental the there there's a lot to be set actually if you um if you look like

one sort of interesting exercise if you this like modular versus integrated debate that's playing out in crypto right now if you look at the backgrounds of a lot of people who are advocating on one side you can can see like this is kind of like the cloud versus on-prem like fully integrated stack type approach and a lot of the people that are in like advocating for more integrated approach like worked at Apple you know the people on the more modular side is like you know uh Google or something like that and so I I think that there are multiple different ways that this could work um and that there are there's like a path dependency um element to this as well so I don't know people use a lot of like this is the best way and the only way and it almost never plays out like that in in practice

um I don't know if to you had something to say but no I was just going to say that I think eventually the paths are going to somewhat converge I think salana we can already see like to a certain degree I I I did this talk a few moments ago AG go called modularity is a meme and so is everything else uh and and you can see how celesy is trying to become more integrated by basically talking about integrating zero knowledge proofs into the protocol Etc uh and allowing uh bridging to trust minimize bridging to be enshrined into the protocol and you can also see how salana a lot of there are a few apps on salana that are currently building l2s like GTO is

modular modularizing the block space Etc so I think I think eventually we get to a point that the like the modular approaches are not going to be much different from integrated approaches they're going to somewhat converge and salana is not going to look much different from ethereum even though there will be differences but I don't think they're going to be as massive as what they are right now yeah I I tender this is the all we're all building the same thing meme um so maybe we stop Gast Ling each other shout out uh Josh Bowen for that um I I wanted to ask um maybe uh I mean or miles you want to talk a little bit about like more flexible the need for more I guess the last thing on my end get your guys thoughts on it we've

talked about the insurance piece we think we you know that actually is there's something there right you can do more with less um and we've touched on you know perhaps having like uh different types of slashing or different types of tokens like right not everything is Slash and stable um I'm curious what you guys think about the inter subjective work token um or be a token that maybe can you know support more types of faults um you know we touched on let's say like dydx in the beginning of the call that I've done a lot of work with them on how to handle me and it it's all social slashing right um but maybe there is a way to create uh have have different sorts of staking

like maybe the native token right um is subject to more types of slashing or can use its own governance right to do something similar to an in subjective work token whereas like the external restake token does not really touch any of that stuff I'm curious what you guys think about you know what's been put forward with Ian and in this idea of okay again vanilla reaking maybe not that interesting if you really think about it being able to support more types of faults including faults that are maybe impact users more often on a day-to-day uh whether it be liveness or me or something like that um I'm curious what you guys think about it if there's you know a place for it um longer term

yeah I think so uh I guess a couple so please correct me if I'm like mistaken on about this because I feel like I didn't I didn't actually read the whole like 50 page white paper don't worry about nobody has read the whole thing okay um I mean it sounds like the tldr is basically like oh we there's some things we can't use eth for because we don't have the ability to Fork eth like uh and we want to sometimes use social consensus to like just Fork the system and like that so and so we're gonna create a new token Ian to to do that I mean I think the idea make sense of like hey you want to Fork uh sometimes you need to like have social consensus I guess what I two things I don't understand one is like um like okay then

I get you want actually like social consensus like different tokens for social consensus of different things like at modular Summit the uh fellow from igen Lair was like talking like oh you can use inter objective token to like you know oh there's like a debate about like who which rap album was better and like okay like this requires inter subject but it's like why would I want my the same inter subjective token doing that as like determining da like like correctness right it's like okay there's different levels of subjectivity like yes I understand da is like you know it's hard to cryptographically prove it but like that's way less subjective of hey was this data made available than something like oh which rap album is better so it's like okay you probably need multiple subjective

tokens for different uh use cases the other problem is that like hey this intercept objectivity works when there's no touch point to an external world right like in a problem with like a prediction market system right if you're betting with the inter subjective token as well that's fine because you can Fork the world and you know and one of the one of them the the token people are betting with like doesn't go like goes to zero as well but in the case that you're like betting with usdc or eth even right like even if the like Oracle token knows how to split itself into two you still have some fixed resource in the in in the center and this is like

always the case with like social consensus as soon as you like you know even with like we're talking about consensus protocols right where it's like yes you have social consensus until you've introduced bridging now you still you know the bridge still has to make a decision or you know Circle still has to make a decision which usdc to support or you know inter subjectivity kind of works in very isolated environments I think it works in certain cases but I think the use cases are quite Limited at least I hope the use cases are quite limited because the same reason why I'm bearish on on chain governance I'm also bearish on a lot of use cases for for uh for the inter subjectivity and I think like an example

of a use case that actually makes sense to use uh social slashing or like uh some form of foring social working is the Oracle problem like there's no way you can deterministically prove on chain uh what the prices that are external to the chain are so you to to a certain degree you want to uh have some form of social governance that that oversees that so for example what chain link does they introduce proof of stake but they're like oh actually like proof of stake doesn't really solve the problem so they have a committee overseeing that proof of stake committee that basically slashes them if if if the proof of stake side of things fails and uh same same is true here like like you can you can use

igen layer for in in similar scenarios but my problem with a lot of the in coin voting on chain that decides things is that uh especially with slashing Etc is that because it's subjective it it affects it can affect a wide variety of users so so let's say if you have a general purpose chain for example uh and let's say only 20 20% of the total value on chain is your native token the rest is like usdc some meme coins etc etc those 20% of the Tok of of the value decides what the chain does is it a good model I don't think so I think I think

it creates issues because then essentially those those tokens are the second class citizens that that have no say obvious you can say that there's always social consensus on top of it you can just Fork off and say screw you we're going to go our own way but that's not that's more easily said than done and so you you get this like two class system you do but do you think also that something like honestly one of the reasons why I have challenges with just onchain governance is that it's actually just onetoone token voting and it's a pure democracy and that like just doesn't work at at any scale really um and so it seems like we either like even just the way that we deal with faults in in a

reaking context is on the Spectrum here of uh you know non-sub like a double sign or something like that all right that we can slash and then yeah and then it goes all the way over to just like okay everyone just kind of take a vote on it and it just it feels like there's actually centuries of you know organizational psychology that would tell you that um regardless of human like I feel like that's the area to be innovating on is just the structures and degrees of representation ET ET because yeah I think I think the reason why I've been at least challenged with seeing onchain governance in practice is it just degenerates into like onetoone Democratic voting which doesn't work at all wa you mean you mean that U proposing building an ape theme a board ape themed hotel in Bangkok is not a

great idea that is probably the best example of what I mean yes uh yeah I agre yeah I think my general take here is I think there is like um there is a place for these inter subjective you know something that can support inter subjective faults I'm just not I'm not totally convinced on the universality of the token applying to it like for sunny for for osmosis it would be the difference of like if the the protocol owned Oracle system right had was submitting bad prices probably would want that to affect a class of stakers and have some process to handle that probably doesn't need to affect let's

say if osmosis had tons of Babylon right like restak Bitcoin it probably could be a little bit more narrow I don't think you want to mix those two things up but you know there is messy stuff here and I would say like with again back to the dydx me example that's probably even too subjective to be handled by something like uh you know a forking inter subjective work token or something like that but um yeah there's a place for it it just might be you know it's unclear if it's local or like this Universal approach is going to work better I think one of the things that I I think that we kind of didn't really touch on but I think that what's kind of more excited about reaking is actually being able to use it to like change the L1 itself

right I think like so far we've most have been talking about like oh reaking and then like just giving Economic Security to other systems right but I think what's actually kind of more interesting is like saying like hey you can add slashing conditions to the operators of the base system and I think that's actually where like reaking is you know that that was actually the original use case of reaking that I that I had heard and like I think that's a little bit more interesting because uh you know you can you can put like like in the DDX case right like if you can figure out how to add like certain like slash ability to like me uh things right or even if it's just like hey you can slash them via governance right that's

still I guess you could say a type of you know restating one of the ex like things that I so maybe I should have mentioned this earlier I'm I'm an adviser to Babylon uh and so this is like this whole Bitcoin reaking was like an idea that uh Dave like I came up with with David and one of the there was two reasons I did it one was to like introduce staking to bitcoin but the other was I event you know I had this dream of creating a specific app for Bitcoin which is a fast uh pre-con Network for Bitcoin so like you know one of the things that like uh you know I'm a little bit of a Bitcoin Maxi but like it annoys me that Bitcoin block times are 10 minutes and I'd like to create a system that you can you know

Bitcoin miners uh eventually hopefully stakers but miners at least can stake and like add give themselves more slashing conditions of saying like hey if they've accepted a transaction they'll get slashed if they don't do it uh you can do stimul stuff on ethereum I think that's actually where reaking is cool where you can add slashing conditions to the L1 get more guarantees from the operator set then the L1 is itself is willing to ask right and and vote extensions in Cosmos are kind of like a mandatory version of adding you know new jobs um yes okay so you think the like opt-in version of vote extensions that we're seeing Oh that's interesting I've never actually thought about that we could actually just create like you know right now the vote extensions are very like

structured data where it's mandatory for everything but like you know we could just put an arbitrary data field and then like like allow that to be opt in like actually this is kind of what um in Bitcoin bip signaling like you know it's an optional thing miners can choose to use like the bip signaling to like when to activate soft for or this is basically like um Drive chains is this idea for Bitcoin of like hey you build this like optin like miners can choose to be an operator of this bridge so yeah I think we that'd be cool to like create a version of B extensions that's like more optional and based off of reaking uh I I I I think I think such an idea reaking to enhance the protocol

itself is more useful than something like Bitcoin because the community there is a vers to changes and basically you basically have to think of a way of creative ways of how to extend the protocol in order to add more features without actually Chang changing the core protocol with something like ethereum I'm I'm I'm I'm not really sure that it Mak sense I remember there were discussions at some point that we can use IG layer to remove the centralized relays and me boost but why not just enshrine it into the protocol at that point if it's if it's a core part of the protocol and almost I think 92% of the validators use me boosts at this point it's it's almost a core part of the

protocol you might as well just enshrine a version that doesn't need centralized relays I mean mostly because I think ethereum core has been like slower and resistant to adding new like protocol upgrade and and basically it's like the rest model for this is saying hey let the market like choose what they think is like important protocol upgrades rather than like etherium cordev yeah it's like a canary Network for like potential you know like EIP upgrades um you I think like L2 pre-ms or another one that we're seeing as like an early potential use case but and you know there's another one which is on Solano which is like by the Picasso team where they're basically having validators like

you know p uh salana doesn't have have like like cin proofs in the same like you know Merkel trees they're basically having a side chain to salana run by the salana validators that is creating like meriz proofs and I I think that's that's an interesting one I mean selling block future block space right um could be one where you know it sounds sounds like a feature you want to add and then we see what happens if you can buy like five blocks in a row and then maybe we decide that's a really bad thing to inshine right um it's but but yeah I think that will be slower to like you know you need a lot of proposers to actually make these things useful but super interested to see what happens there that's

actually kind of an interesting challenge as well if you move to less of a top down uh sort of system led by devs and more of a market based system then let's say you create block space Futures you open up the world of multi-block me and you say hey okay guys uh we're not going to enshrine this but at that point who cares uh because it's already being done out of protocol and it exists although now just saying that out loud I'm not even sure what you could do to stop that anyway so um the joys of being permissionless um guys uh I you've been super generous with your time here um I appreciate you coming on and uh just as the caveat I think uh you know this we appreciate the folks in the shared security Community being a good sports about this I think

everyone on this on this call is actually uh very excited about what this is going to open up as a design space we just want to have a more pragmatic um and maybe like poke at some of these ideas that we've been talking about all season but um sunny and and togl uh folks obviously Sunny you're basically an allar of belr at this point and togl is great to have you on but if folks want to find out more about you guys or the protocols that you're working on or follow you or whatever what is the best way to to do that uh you can follow me on X Sunny a97 yeah you can also follow me on Twitter I'm not sure if I want to spell the handle it's quite long but yeah you if you type in to girl scroll you'll find me on there

we'll link it in the show notes we'll link it in the show notes for folks want to follow too um awesome guys uh appreciate you a lot uh thanks for coming on this was a ton of fun hope you do it again soon thanks for having me thanks guys miles that was a fun one man that was um yeah I love that conversation I think like you know this was supposed to be I don't know if you want to call it the bear case for reaking I just think it was like the intellectually honest conversation about where we are in reaking today and you know where we think it's going um as well as it kind of turned into a conversation on like broader delegated proof of stake systems which I think is like a big part of um I think the impact

that reaking has had already is is making people think a lot harder about this stuff for maybe like the first time um you know something as just as fundamental as how much security do I need and how much should I pay for it like what am I actually getting right um it's making people I think think twice uh a little bit about some of the properties of delegated proof of stake how we can improve it the fact that we can probably get more with less if we really like you know can't put some effort into coming up with better ways to do this um but yeah so there's so much there so wherever you want to start let's start with the discussion around

proof of authority I found this very interesting and maybe to te up on this this is I feel like there's this repeated pattern in crypto where you know there's a big breakthrough be it the initial proof of work Bitcoin setup or proof of stake or whatever and that is designed for a very specific set of criteria right you're trying to This Global Bas layer money compute system and there's all these reasons why proof like vanilla proof of steak makes sense for something like ethereum but then we move one layer up the stack um and we start to build things like either L2 or apps and the immediate immediately what we do is we carbon copy what happened on you know the main chain and then we start to ask these questions well actually you know the the vectors that

are driving you know what what the underlying like the kind of backend here should look like are very different from L1 and but you know we don't really want to you know we don't want really want to make this a private blockchain because we're all in crypto and that doesn't sound very nice or very good so let's kind of come up with all of these different contraptions so that we can get around being honest with oursel and just saying hey this would actually be just better with a permission validator set and then after two years we all just kind of say yeah this would be better with a permission validator said and some of these other bells and whistles maybe weren't as necessary um and so that's one way that you could look at this and I think that what I would love to see as well and I think what we

started to talk about in this conversation is the very different level like backend setup level of governance whatever it is for applications and middleware iders versus something like uh ethereum and so maybe you can talk a little bit about like this whole proof of authority I think that's a really good way of putting it I think uh to date we've had this bad habit of starting like grabbing a system that we know seeing the properties that that system you know provides to a different chain like say it's Bitcoin or maybe it's ethereum grabbing that same system and hoping you get the same properties um when I don't I think you should be doing it the other way around

right start with the properties that you want out of a system then think about what your product actually is and what the operator set does right um what it should look like to be the best version of your product while retaining and then and how do you get those like properties of let's say trust minimization censorship Assistance or how about like no you know taken to the extreme of Stak sure like no honest user should be harmed right how do you then back into a system where you know users can be reimbursed if something goes Super wrong um and I think what we've seen is you know we started with ethereum said okay this thing proof stake like Works should work well on paper since people are

running machines and putting up their own Capital uh then we see just especially with you know like it we'd be lucky to have one system that is important and big enough where people are actually willing to put up their own capital and run their own Hardware right um but most of the time it just devolves into finding workarounds to to you know do something that is more realistic right with delegation so okay now we have like delegation built into these systems um you know maybe I'm an achain maybe I'm an oracle or a bridge or something like that um do I need as much you know Economic Security or do I need to make like should I have the same

security system as say Solana right a delegated proof sake system it's a huge base layer um or you know does my bridge operate way better when there's actually fewer operators in there right and I think people are realizing oh actually no this product like works really really well if there's actually less operators in it right um as long as we can get those same properties where we feel like it is secure right or that there is a way to reimburse users for any sort of issues um I think you know especially in the early days of any project that's not trying to become like the next Bas layer money of the internet or something like that it's just building some sort of

system that works better with its own consensus or its own operator set you probably start to land on POA like or proof of governance you know as the starting point as you know for most of these systems right um and in that sense the economic stake that they're pairing with this proof of authority system is is you know probably more for GTM purposes than it is actually for functionality with the product I mean it does help to say yeah we have a billion dollars of Economic Security but you know is it is that is that what's making your operators actually you know stay honest or is it their reputations right and the fact that if they you know screw

up your protocol or do something malicious they're going to lose all their other business um so I think I think that's probably if we were to poke one thing you know at at the current state of reaking or at least if I poke at the current state of reaking um find one thing to poke at sorry would be hey these are like all gated you know operator sets that you know are basically have their reputations at stake because they're not putting up their own Capital um those reputations are super valuable why do we need the you know some third party delegator Capital to be like I mean it does increase the downside risk to those reputations like if they screw up

somebody some random folks just get slashed and that makes everybody hate those operators even more right but it's marginal it's marginal I think um until these systems start evolving yeah go for it I've got thoughts here on so I I think one again just in the the the only issue with saying hey this is all relying on people's reputation anyway so we might as well just sort of formally you know explicitly state that that's what it's relying on is there's a pretty long demonstrated history of business operators and financial institutions that are actually willing to you know willing or extremely negligent about their personal rep you reputations at the expense of pretty

shortterm uh decision- making or profits so I do think probably at the maybe one way of framing this would be in the beginning when these startups are all very small and The Economic Security isn't really getting you that much anyway maybe it makes sense to have more of these like POA type systems but maybe once something either gets big enough or I think the other thing as well which will eventually complicate this is once it becomes multinational or Global um then it makes more sense to have uh something like Economic Security and reaking because you know one of the things that at least I've been thinking about quite a bit is the context of how much should we be building these systems to be 100% insular cryptographically secure versus is relying on legal

jurisdictions because let's just be real you know these boxes and these operators have to exist somewhere and uh one of the one of the the nice things about the US is you know for the most part it's a very high functioning uh you know protects property rights good judicial like legal system Etc um but the thing the reason I would say that you can't rely on like the US legal system for an L1 is there's it's supposed to be Global and so you do need more sort of cryptographic um assurances there so element of how mature you know these These are but there's like some non-obvious tradeoffs there it's like okay um you let you go with delegated proof of stake because you do think like

you know having some economic stake uh especially as you grow is going to become more and more important um well what could happen is let's say you use you know that economic stake you connect that to like voting power or how much work you know these these operators do so like the more stake you have the more weight you have in the system or the more work you do in the system everybody chooses the top 10 validators that they know and love right now you have 10 operators that are like you're relying on way more than you would let's say if you had no stake and you had some sort of like basically the equivalent of what like an LST does right where they they distribute work very evenly across the

whole set right um and you just have governance kind of update that methodology or system every so often to make sure you're getting the properties that you want right soon as you bring in stake then it's a question of okay how does stake actually influence like the the Quorum design or like you know other parts of the protocol and you're kind of giving that you're offloading that choice right to stakers who will choose names and that they know and you know trust right so there's some like nonobvious stuff here where it's like oh well in this sense proof of Authority actually would get you those properties maybe even better than you know delegated proof of stake if everybody just chooses five operators to stake to and that's you know represents the

majority of the work and Trust in the system um you know again like look at dydx it's very very topheavy um if they had some sort of system that was not based off of you know voting power with tokens um you know driving like actually who's proposing blocks maybe you end up with a more diversifi set of operators that are you know are more distributed work across that set of operators um at brings its own trade-offs though and to your point when there's literally nothing at stake you know we haven't really seen that um an approach like that taken to the extreme yeah it's it's such a good point and uh there it maybe there I mean even

we we talked about this a little bit but there's there's really no such thing actually as truly permissionless um it's everything is permissioned in a sense and you could think about that as being the mechanism of enforcing who gets to participate is either a monetary one um which is how it works in ethereum today you have to have 32 eth to be a validator or one which is explicitly up to governance yeah and I see pros and cons with each one of these appr because you're absolutely right dpos you end up with a very topheavy validator set which might not actually even be who the best validators are but on the other hand onchain governance is messy I mean we got into in the episode but totally totally I think where this comes to the Head comes to a head um is L2 sequencers let's take know we didn't even bring

that up in the episode and that was the best yeah probably would have been a good time to do it but let's say when we're gonna have different systems right so there could be an L2 that chooses hey let's lean on governance governance elects a set of operators governance elects a methodology to rotate those operators or some sort of you know way that's like geographically distributed you know get the properties we want right and there will be other l2s that you know go after a more uh traditional proof of stake um sort of system where they say we don't want governance deciding the you know operator rotation methodology we want stakers to you know delegate to one of these

operators based off of like you know how they delegate that will drive the distribution of work you could see you know these trade-offs kind of play out in in um in real time over you know the next like six to 18 months as this stuff happens maybe maybe the systems that go with this proof of governance actually have like a better distribution of work better Better Properties maybe they don't because they have a single point of you know failure with this onchain governance thing we know or token governance which doesn't always work great either but um yeah I think that will be a super interesting like something I'm paying attention to um where I don't think there's a right or wrong answer it's just like we'll see

you know different approaches and it'll be interesting to see how they work out last last thing that I thought was uh an interesting conversation here we can I know you got to run in a couple minutes but the idea of building an ecosystem where you keep your abss or networks and I what I would just call this there's a lot of bundles as business models in crypto um that are like the L2 stack bundle if you're the op stack or arbitrum orbit or whatever it is you're trying to keep chains in your ecosystem and you're adding you're doing value added Services across the chains and interop has kind of been the idea um of how we're going to monetize like make these apps composable with one another um and it seems like that might be the direction that the reaking marketplaces are going as well I am just very curious

to see what feature of any will be the really sticky one uh that keeps networks within a specific reaking ecosystem but I think that was the last like interesting call out I thought from the episode yeah you you know it's not super clear that there are like extremely obvious Network effects to say like the next marginal ABS that joins you know the aan layer platform makes the platform more useful for all the existing avss right I think you could make the argument that there's some sort of um you know benefit for the consumers of ABS is saying like oh yeah there's more services all in one place right um but I agree with you and I think tog you know calling out the same sort of

dynamic with l2s right um or even salana like you get all these benefits by sharing State together then eventually you know the successful ones want more how do you create a system that you know lets them have that level of like control value capture feel like they're kind of controlling their own destiny but still providing you know a reason to stay long term not go like full L1 on your own right um super chain is one approach um you know I think what he said is basically have like uh all these sequencers basically creating you know their own mini blocks and then having all that batched into one big block um I think you know we'll tease tease a

little bit today it's a little bit similar to something I started working on a few weeks ago um in a in a new project that will yeah exactly um tease later this summer but um I think that's where things are going right you want to have a Bas layer that continues to be complimentary but allows these you know um allows the ecosystem projects you know to kind of graduate without leaving the actual platform itself right yeah all right miles I know we both got to run here but um this was really fun dude we've got one more episode this season uh it's it's a project that we have alluded to and there have been many rumors about um we won't blow up their spot but many of you can probably guess

who we're going to be talking to with this last episode so yeah yeah yeah it's going to be a blast it's be a blast all right partner I will uh see you bye hey everyone hope you all enjoyed today's episode wanted to give a final shout out to namada again a team that is near and dear to my heart they are building the shielded asset Hub which is protecting the future multi-chain talked a little bit about how that works in the mid roll but for now what I would highly recommend that you do is go check them out they're in their main net launch phases you can follow along with everything at n.net you can start rewarding Yourself by shielding your action and strengthening data privacy for all across the multichain so thanks again to MA for all the great work that you're doing