Privacy in Cosmos Live: Bringing Privacy into Cosmos Panel Discussion

Transcript

right so next up we have another panel i'm going to invite four panelists this time around i hope we have enough mics we have four oh so we will have to share one mic whoever's on whichever couch here okay there's one [Laughter] all right so the this panel is the title is bridging privacy into cosmos um when putting together this event we felt like a lot of what we were focusing on

was cosmos with a sprinkle of privacy there is a lot happening in privacy and so with this panel what we're actually going to be doing is talking to four privacy focused projects two of which are connected to the cosmos ecosystem already one of which which is deeply in it but two which are less connected and i thought this was actually a cool moment for the cosmos community to also get to know what's happening in privacy so let's start with the project that you did already hear about today uh penumbra do you want to just say a little bit about yourself and the project um all right uh my name is henry uh

i guess uh so i gave a little bit of a a brief overview of of what we're building in the the talk that i did but um uh the punch line i guess is that we're trying to build up a private decks and the thought process there is that [Music] trading is a really interesting use case for bootstrapping privacy because it's a case where people have a financial incentive to not disclose their information and so you could potentially change this kind of existing dynamic where a

lot of the time when people build private systems it ends up being kind of like okay here's our private x right and it does x and it's private and it's like kind of worse at doing x but like it's private and you should like privacy and like you know everybody thinks that privacy is good so you should use our product and like unfortunately i think that that doesn't really work as a as a as a product and the hope is that uh with trading it's a it's a case where privacy thought of as kind of control over the disclosure of information has this very concrete kind of

financial incentive and maybe we can harness people's like kind of naked self-interest to bootstrap this like private infrastructure awesome like can we actually be better at doing x because it is private rather than you know being having some kind of trade-offs that's awesome adrian maybe also mention your connection to the cosmos world yeah hi i'm adrian um yeah my connection to cosmos world is i joined as the third corps protocol engineer on cosmos in like 2017 and sort of built multiple companies out of the space including a validator

and yeah so what is our noma anoma is a suite of protocols and mechanisms for privacy preserving human coordination in a very self-sovereign fashion when you think of human coordination right that can mean very simple things such as being able to express access control on your wallet but it can also mean very complex things such as on an ephemeral level wanting to buy starbucks coffee and offsetting the carbon cost at the same time on a very permanent level being able to define that whenever you interact with the financial system you always want to make sure that you're buying fair trade things

and anoma allows all of these things to happen in a very programmable and intuitive fashion and i think maybe one of the main things here is that well one thing we have to realize is that we need way more information density in how we interact with uh the the world right when you think of something like an election or of the financial system it's very information reductive step where you sort of try to break down things into single points that you can coordinate around whether this is a buying choice in an election or a single price for a cup of coffee but real life is way more complex than this and we should allow people to be able to

express these like very complex preferences that they may have in relation to the environment where when i buy coffee i not only care about the price point of the coffee i also care about whether um it was sustainably sourced whether the co2 cost is offset um and yeah this is really the idea with anoma to allow very high information density um to be automatically computed over and so for people to permanently express their preferences in relation to the world in order to automate a bunch of these things so that i can say i always want to buy neutral energy and i have to do this once and then for the next 10 years it's just automatically done but i can also update these kind of preferences yeah and of course all everything in

privacy-preserving manner because we discover privately privacy is fundamentally important because a bunch of things start breaking if we don't have privacy got it emray sure hi everyone i'm emory um with o1 labs o1 labs is one of the main contributors of mina protocol mena protocol is building a private and secure way to access web3 it started as a you know using zk primitives as a as a project that has full node security with just a few kilobytes of data so that has been live on mainnet for over a year now as as a new blockchain and now in the coming months we're

looking to enable easy zk programmability where anyone can deploy using super easy sdk and typescript zk apps on the mina network and be able to share their data in a private manner where they make attestations instead of sharing all of their data onto smart contracts on the blockchain and when you combine that you know super small data full node security with this that enables mina to become kind of this user private user data layer to any other layer one ecosystem whether that's you know cosmos or anywhere else so that's what we're working on and maybe one thing about mina is it was one of the first if not the first to kind of

showcase this type of technique called recursion in zika i know it existed before and had been written up in academic papers but as far as i know it was at least the big first big scale use of that technique which is now being used by lots of ck projects all right alex uh yeah hi everyone first of all let me just say it's a pleasure to be here thank you for inviting me it's amazing to see uh you know i've been in the space for about five years now it's amazing to see the cosmos ecosystem just grow in the way that it has so um again it's a real pleasure to be here and be part of this event um i'm alex i'm the ceo of elio um alio is a platform for private applications um similar to you know probably some some of the things

that these guys have described you know this is uh things that potentially benefit from privacy and i really love the way that henry framed it is like you can't just you know you can ba you can pound the table all day and say you should care about privacy but fundamentally i think you know all of us you know should focus on how privacy makes something fundamentally better or enables something that that hasn't before so anyway alia what we try and do is enable take the z cache basically paradigm and extend it to more general programmability and i think maybe the other different thing about elo that i'll mention is we have a consensus algorithm that specifically tries to incentivize uh proof generation and with the hopes of making it you know more economical more efficient over time and uh and and to therefore bootstrap this as a

paradigm uh into the ecosystem more broadly and help integrate these private applications uh across web3 so when it comes to privacy projects i feel like there are different kind of levels of how private something is you have the model that zcash sort of pioneered which is to have a shielded or cloaked as i'm going to keep trying to use a portion and a transparent portion so part of the blockchain you can actually map like any other and part of it is hidden and i wondered for each of these projects like how pr like what parts are private and which parts are actually public this might be something interesting for people to

understand so maybe we'll go this way yeah reverse yeah by the way is cloaked the new term so i'm trying i like i kind of like it actually it's like star trek star trek yeah i'm into it cloaked and decloaked um yeah i'm into it uh anyway so from so i guess like maybe just as a general answer to the question um you know there's there's a couple different dimensions of privacy that you might care about as a user right so one is you have an identity that you're using on chain which could be your account or something and maybe you you would like the fact that you're right you know you would like that identity to you know be obfuscated in some interaction right so you don't want the you know one of the interacting parties or an observer to know who you are so that's kind of one

dimension uh and then the other is like what is the what is the interaction like what is the program basically being run and maybe you would want to obfuscate that program like for example if you're if you're participating in a dark pool or something like that um you know there's potentially reasons for example trading to you know want to hide the fact that you're participating in a dark pool because if you're trading in a large enough amount or you're a specific player say you're a large venture fund exiting a position obviously just that knowledge alone that's that you know a program is being accessed might trigger actions um that you don't want to necessarily you know trigger so i think those are those are um you know and then of course i guess the last thing and since we're talking about web3 is like the transaction amounts right so

sometimes like you know for example tornado is like you know it's it's the way it works is kind of more as like a mixer so you're hot you're kind of mixing the amounts um that are going in so maybe the amount or the specific data um or state at which you're which you're accessing you might want to hide too so so yeah i guess um i'll leave it there in in the desire to have it general or do you want to talk more about alias specifically no i will so i guess the question here is are there parts available that are transparent yeah so the model that i like and that we try and strive for at alio is is private by default so you can actually hide all of these various aspects and you know for those of you who are curious alio kind

of came out of this academic work which is sexy stands short for zero knowledge execution and the whole goal of that was to try and extend zcash effectively and enable both um you know privacy of the individuals interacting as well as the program being accessed you know i think one thing that often gets overlooked in conversations about privacy and private applications is um it's sometimes very convenient to have public state which is accessed and so you know i think this is this is sort of i think it has to be left up to the application developer i'm like okay what am i what am i trying to do here what parts do i want to let the user choose whether or not they want to let disclose what parts of the program state do i want to make public and the bottom line is i think you know

protocols in my view should strive to make that up to the developer of the application so you want to hide everything you want to hide some stuff maybe there's a selective disclosure and if you think about it just the last thing is if you think about that's exactly how our financial system works already it's like if i send em ray a bank wire it's not like i know his account balance right like i can tell him or i can tell the irs you know and selectively disclose what i choose to but by default it's effectively private i think this default by default private but by exception public is the paradigm which will i think it kind of have lets you have the best of both worlds also i don't mean to put i'm not trying to compare you on levels of privacy i think there are use cases where more privacy

is needed and there's use cases where actually it might be good to have either a selective disclosure or even a transparent anoma how what like what parts are private how does it work um you can look at this from a couple of perspectives right you can look at this what is private on the protocol side what is private on user sites i think starting from the user privacy by default really really matters because especially if you have shared privacy sets because privacy is a public good as in even like i benefit from a bunch of other people being private because it enhances my privacy in the system actually so we want to incentivize people to uh by default always be private because it

like it sort of it increases the privacy set that we can all rely upon this is also why it's like not a great model that we want to like have separate privacy sets like when you think on for example like tornado cache you have privacy sets between the different denominations and the different assets and like that heavily starts segmenting the privacy guarantees that you're getting so privacy by default is i think very important but then it also comes down to so users can effectively choose how much privacy they want they can be fully private they can decide to be fully transparent this is in the end up to the user because for some things you need to leak some data right if you uh so if you're pure execution environment that's not as important

because you don't generally need to coordinate with other parties yeah but for example if you're in this model where i have like a cryptokitty and i want to sell it to someone or i have some bitcoin um and i want to buy concert tickets in barcelona for them um this kind of selective disclosure where you can discover countless parties with whom to work on achieving a settlement for mutual benefit you need some amount of disclosure in order to discover counterparties if you say everything is private all the time it's very very hard to have to discover counterparties around the world unless you say well we have like the protocols always the counterparty which is sort of the kind of amm model don't you also need it some somewhat for like oracles for like price right so a

bunch of other things in these systems also need to be public and i would argue that like proof of stake for example it's still a fairly modern thing there's an argument you made that this should be public because you want to understand what the security model of your chain is um same for bridges most likely bridges are still very hacky and it's like way easier to recover from faults if you have private state a public state instead of private state so i think we can think on privacy on the user side and there we want to give people the ability to be as private as they want to be but also to be as transparent as they want to because we can't really estimate the use cases that developers and end users have in mind so we should make it very cheap for them to

be private but not force it onto them and then on the protocol side as is tech matures sure we can be more and more private but sort of certain linchpins i think it makes sense to be public yeah cool and meena yeah i'd say mina has a similar approach where we're not super opinionated about privacy necessarily in fact mina did not start up as a privacy project it started more of as a project that kind of breaks the scalability trilemma by having this you know fixed size tiny node then you can you know scale it as much as you want um but what we're seeing nowadays is um and as well as having the best you know zero knowledge uh primitives and what i call the our aim is to be the most efficient zk state lawyer in web3 um

and what we're seeing now is that you know the best use cases there are about uh making for mena at least making user data private um so instead of you know if i want to use a defy app or an nfd app or a dao or whatever instead of having to make every action that i take public or every data that i need to provide public instead i provide attestations whether it's about my you know personal financial information or my kyc or my dow voting rights or what i voted right um where you know people then you know can make attestations instead of sharing the data and then also it's what we're discovering it's it's fairly easy to build um

layer zk roll up type layer tools on top of amina so in fact like some a handful of developers are using the alpha of our sdk and you know one of them has already started building like a basically a shielded transaction layer on top of mina for native mina tokens as well as third-party tokens um so what i would how i would position it is like if you have super easy to use powerful zk primitives then all this stuff becomes much more easy to build as an application ecosystem and that's what we want to enable at least for the mean ecosystem cool um so yeah i think there's like a little bit of a contrast in that um like penumbra is like quite opinionated

about privacy um and uh specifically like picks like a very very narrow and like um focused use case and if we rewind all the way back to the the way that you introduce the panel there's i think an interesting perspective there um so looking at zcash there's uh the the transparent part that was forked from bitcoin and then there's the shielded park and if you go back and think about the history right that project started um at a time when you know a zk chain had never been built and also at the same time the rest of the blockchain ecosystem was much much

less secure or sorry not mature sorry [Laughter] that you didn't have like you know ibc didn't exist like none people were just like talking about like oh what if we could do like cross chain atomic swaps wouldn't that be cool um and i think you had much more of a sense in which like each chain had to be its own universe and i think what's interesting about the context that we have today is that the development of cross-chain mechanisms allows

individual chains to have like some specialization for a particular use case and so for instance for penumbra right like having transparent transactions is not something that we're going to do but that's not because we believe that no one should ever do this ever it's because like you know if you want to go and do like a non-private smart contract you should like do an ibc transfer over to juno where there's like people that that's their whole thing right and our whole thing is like this like private decks thing um and and so i think that's like kind of uh i think a little bit of a different approach than trying to build a more general platform

um i don't think that that's sort of i don't mean that it's like a i don't think there's any like value judgment in there it's just like a different kind of strategy um and and to some extent it's also because like you know like elio exists right so like there's no point in us trying to build like this like generic uh like zk smart contract system because this is already being worked on right yeah and i i think i love how you frame that and i personally believe that the future is going to be application driven rather than protocol driven i mean all four of us up here arguably are kind of building protocols i mean henry you're obviously working on the most specific version of a protocol but you know i

think the future is naturally going to go to a place where people have a demand for a product the protocol underneath it will naturally be built to serve the product that people want right and i think in even an alia although we're building a more general platform and i you know i don't think that alio is the right choice for every web3 application and you know one of the things that different hs differentiates us from example for for example ethereum is we don't have a model of an on-chain vm which in some ways is limiting in some ways it's very powerful but in some ways it's limiting so i think it really depends on what you're doing similar to privacy like you said for some things like nfts and some for some applications of an ft's privacy may be important but for just like sim you know like you know

showing off whatever nft you have i mean kind of privacy is the opposite of the point right you want people to know that this nft is yours and you want to show it off you know and maybe that's a bad example but i think i yeah i think i think everywhere and you know i would encourage all of you who are at this event and you know hopefully at future events and um going forward i think about just these trade-offs being made at a lower level and how those percolate up to potential applications that you could or could not build because i think there's room there's a massive massive space to experiment and there's a huge number of different ways these protocols and products could look uh in this in this paradigm so yeah 100 agree actually on this product perspective i mean this is

something i'd be doing at the protocol level um where we are building a sort of the first chain in the normal ecosystem is going to be called amada to just focus on a very narrow use case which is multi-asset privacy to for any connected asset where it's just like you want to have a single very large shielded pool that has eth that has dye that has atoms that can have effectively anything that's bridgeable um at the moment viabc and they all just live in the single shield pool you can do proofs on in the web browser in like three seconds um and you can even move an nft into this this is a cool thing like for nfts you kind of don't have any good privacy solutions but if you have a multi-asset chiller pool you can have an nft in the

same shield pool and it benefits and it sort of inherits the privacy guarantees you get from all the other fungible assets because an external observer it's indistinguishable whether you move your nft or some die if i can add one last thing wrap everything up maybe not to be the odd one out um i would say like and this is a bit preaching to the choir giving this is a cosmos ecosystem but like there is no doubt it's a multi-chain world going forward um and like every chain inevitably whether it designs itself to be or whether it says oh i'm just the most general smart contract platform ever even in that ladder case they're all going to start like naturally finding their optimal you know frontier if you will like i don't know maybe ethereum just ends up being you know the

asset settlement platform going forward right but it's not going to be the the privacy layer right um so i i think we just have to accept that you know it's a it's a multi-chain world at this point it actually comes with a lot of benefits to end users as well like if you think on building privacy tech on ethereum you inherit a bunch of transparent stuff that ultimately becomes a massive privacy leak for you but whereas if you sort of look at the entire stack in an integrated fashion you can very quickly build very good products that like deliver specific use cases for people that are hard to build in sort of a fully generalized manner sort of speaking of that connection point we did a panel earlier today where we talked about bridging we mentioned

this idea of like what happens with a privacy zone or privacy chain once it has to bridge away and in i guess like currently in all cases the minute you take that leap outside of your l1 you will inevitably be revealing something but is there is like have you been thinking about that as like are are you thinking of creating extensions of what you're doing like extending the privacy space into those bridges i mean at this point the ideal thing would be to have cross-chain shielder pools actually let's call them shielded soups because they're not cloaked okay nobody's picking up chain

cloak soups all right um no but so i mean this is trying out here this is still like early research right to get to cross chain pools or cross-chain chiller pools um i think for now honestly if you start braking um because you brake linkability quite a bit so if you go into private ecosystem and then you go out of it you can go out to complete new addresses there they are not linkable to your old addresses that used to go into the private ecosystem so to me personally i think there's not a lot of need to build super private bridges at the moment because as long as like you're just sort of taking public routes between private things it doesn't really matter the

routes that you're taking up private i mean eventually leaks are metadata but this is sort of a problem we can worry about maybe in like four years from now okay not right now yeah we have to i mean we just have to consider this right on the privacy side there is the amount of privacy preserving tech out there right now is so limited like it's tiny and i think we should really focus on making private applications of reality before you sort of start thinking like how does this need to look like in the perfect world right because the kind of the default case right is that like a like a layer one like a leo for example could fun if it was connected to some other ecosystems could kind of be effectively a mixer right so you could go from let's say solana to alio and then to ethereum

or something you could b it basically functions the same as a mixer um you know and and these bridges oftentimes the focus is just about passing those messages and in some ways like you need that information to be public for security reasons or just for functional reasons um but i think it's an interesting unexplored design space um you know because again they're we're talking all of us are kind of more privacy or zk focused projects but you know if we're talking about bridges between say solana and ethereum or something um i think it is kind of an unexplored design space to kind of figure out you know could you without kind of having an intermediate layer one protocol be a mixer like is there a way to kind of make that make that private yeah i'm

not aware of anybody who's kind of looking at that but it's interesting it's hard and just to sort of mention something it was in the oh do you want to say something before um just as a like a riff on that i think that um i i don't have like a lot of optimism that that type of design would work out and the reason is precisely because if you have a private l1 you can use it as a private bridge i think having i i think that the development of a private l1 that can talk to transparent chains is going to basically obsolete the

development of any kind of private bridging between transparent chains because why would you do that when you can just like you know go through any of the private zones something that came up in the previous panel as well was this idea that when we think about privacy and bridges we have i mean i have had a tendency to think of it as shoot like cloaked wait i gotta keep my own term going uh hidden tokens basically all it's always token transfer token transfer but as uh sergey mentioned what if it's messages that actually have actions on those destination chains that are private then it's not so much like the

i mean i don't know if that also reveals as much as if you were doing token transfers in and out i don't know if you've thought about that i i guess i think that part of the reason that people are are focused on token transfers is that it's like a simple thing that people haven't done yet right like in order to do um like private messaging between applications like you you first have to like have the private applications right so i think that part of the the kind of i i don't know that that's really about some kind of like narrow mindedness or that people aren't imagining this thing

but just that it's like there's a big gap between imagining the like glorious zk future in which everything is private and like you know maybe we could build like one like private thing that actually clearly has product market fit okay first yeah i wanted to add there um i mean to henry's point like this is my pet peeve it's also like tokens is the only thing that exists in web3 at scale today like there's like it's all speculation right um and to build these you know these applications that are about so much more than that um you do need some sort of privacy at the state update mechanism right because um like for example in how how mena approaches this is um and you

know mina is by design you know is basically any chain can run a full node omena within itself but what the other chain is hearing from mina is just various state updates into a specific you know application that's running on mena so how that state update is achieved that's where the magic of zero knowledge happens where you know like the the proofs get posted on chain and they trigger the state update as opposed to logic being running the logic running on chain and the chain determining um in a transparent way how the state update happens so i think there's this also kind of to the off chain versus on-chain you know distinction where the more you can do some things beyond the token world off-chain the more privacy you get

to preserve there's also other benefits you know such as operational efficiency but i think that's a paradigm shift that we all need to start getting into as an ecosystem i think henry has a response um i'm not sure that i agree that it is always better to move things off chain and but but i think that there's actually um there's a kind of an interesting way to look at sort of like differences between projects and like i don't want to sort of well i'll just sort of say things and then if i say things that are wrong then i can just be immediately corrected but um if you think about for instance the

the way that uh sexy which elio is built on the model is is almost like every transaction is like a micro roll up right like there's some smart contract state and the way that the execution happens is that someone is going to execute the contract and then produce a proof of the execution and then submit that to the chain and it's like you know bundled up but but effectively it's like a a roll up but at kind of this like arbitrarily small scale and the the thing that's really cool about that is that right all the actual application data doesn't have to be on chain but the problem that it doesn't really solve is

like who is actually executing the smart contract and so there's kind of two possible alternate approaches to that one which we're working on is like okay well we'll zoom in to a very specific type of application and see how we can do some kind of like additive aggregation of input values and then another is um with enoma part of the the this was like a really cool thing that like clicked for me recently so hopefully it's correct um like the reason that there's like it makes sense to have a bartering system that's built into this

uh system is that it's like you're kind of bartering like different people's like potential state updates that is actually a very good way to describe this yes yeah i think that's a i really love what you said and by the way for everybody who ever wants to get a someone building a privacy protocol in knots just ask them to whiteboard you in a simple in five minutes how you would build private uniswap and you will see people do many backflips because either typically i mean it's complicated right because as henry's pointing out and let's just take this xe model right you're taking this off chain you're doing an on-chain you know uh computation who who is running that computation and zexi it's the user right but uniswap is kind of

built on this idea of a smart contract which requires an on-chain vm right and so you know you have is but on the at the same token the on-chain vm model typically is the account model and the account model is very hard to make private right so you kind of end up in this in this weird space so anyway it's a fascinating question to ask people for any of you in the audience the last thing i'll say though quickly is i think there's a world and we haven't touched on yet is i wonder if there's a world in which you can hybridize these two models right one thing we're looking at at a leo potentially is to have some concept of like a hybrid off-chain and on-chain computation where you can imagine a user basically kind of almost like meta transactions where you construct a transaction and then you

forward it to you know some approver or some other entity that basically sequences and or finalizes the transaction and maybe this maybe this paradigm enables you to hide the things that you care about hiding but it still gives you this advantage of having concurrency and composability which i think are the main advantages of an on-chain vm model um to maybe clarify this a little bit it's technically not the vm it's like so the reason why you can't build uniform privately is because you've unchecked like someone needs to hold the state uh right like and in like unisop on ethereum the contract holds that state um because the and the reason why you can't do it privately is because the person that constructs the proof is the

person interacts with the contract and that person necessarily needs to hold the shared state in order to be able to construct the proof i think long term there's actually some interesting research we can be doing and i think some of us are already doing around fully homomorphic encryption and how we can have shared private state sort of for longer periods of times and you can even think more like i'm thinking now between the message passing right you want to invoke some uh cross chain like you want to invoke some other l1 in a private manner what you could actually do is like how long do you need these messages to be private if it's just for front-running protection you can just need you just really need like a couple of blocks privacy where it's encrypted before they're executed

so there's sort of a breadth of design space out there how we can think on how do we make more and more of this ecosystem private and zk like zero knowledge proofs aren't the and all um there's a ton more open research that's probably going to take another 10 years to materialize into production-ready apps yeah i do so i agree the state matters but also i do think the vm does matter for a model like sexy right like if you're creating a proof about a state that even if it's public you just put everything on chain just publicly just make it public data but even then you know you have these records that can only be updated by one person at a time right so like unlike ethereum this is solved because every node is a sequencer right

effectively because they just decide okay hey the memphis and this by the way is why front running is a problem right because if you're the sequencer you're like hmm that looks like a good trade i might just take it um so this is why again like this is like i think there's a really interesting kind of tension between these different properties that we're trying to achieve i want to bring this back to cosmos the other side of the private privacy and cosmos uh how do you i mean i think in the case of penumbra there's very clear connections already to cosmos and ibc but yeah how are you how are you all thinking about actually this ecosystem is this an ecosystem that you see ourselves bridging to or working with bridging projects to get kind of connected to and how could maybe people

in the cosmos ecosystem start to think about your projects um yeah i mean from the enormous side we are normal wants to bridge to everything at the moment by far the best bridging tech out there is ibc so i would really say everyone should be building ibc adapters for every other chain um like people should be out there building ibc adapters for nia for flow for pocket art you name it pretty much any fastphone ltb of t chain someone currently should be building an ibc adapter because it makes not necessary to connect it to cosmos but to make all our like all chains connected to each other because this makes the ecosystem much more composable if you can all of a sudden do a penumbra alio

a norma amino hub because all these chains just have a common interoperability standard this is the way i think about bringing privacy to the entire blockchain ecosystem just by getting people to implement ibc um because it's like it's a great protocol but what about like an evm like what about f like evm to evm do you think it should also be like like evmost to moonbeam or should both be ibc is that the most effective way to do it yes this is the thing like um ibc has very little actually i think has no dependencies on the specific state machine like it it's it has a dependency on the consensus as in it needs to be a fast final tp of t consensus it doesn't matter whether you're running

an evm on that consensus or some zkvm that's secondary um this is like any bft chain should be implementing ibc to just talk to other bfg chains yeah in the case of um mina agree with that by the way although i know the nomad folks are doing some cool stuff there too for evm but side note um so in the case of mina as i mentioned briefly um mina is designed to be a full node whether you're on a client you know a traditional client or you're on you're under blockchain right so we're incentivized to bring value to every other layer 1 or layer 2 ecosystem out there by providing this kind of user data privacy layer

to the other chains out there whether that's cosmos whether that's ethereum solano what have you in fact um there has been uh trust what we call a trusted zk bridge from ethereum that's been in progress for a while um hopefully there's going to be some sort of you know demo or overview in east amsterdam in the next week next weekend and you know that by nature is default for any other evm chain as well so for example atmos in the case of cosmos but then also is not vm dependent either it's just engineering that needs to be done for an efficient smart contract verifier of the canonical state canonical dk state of of mina so our goal is to you know basically just bring the some of the unique features that we

can to every other specifically about zk to every other um layer one ecosystem out there what what would someone in the cosmos ecosystem be able to do with that yeah so let's say you're a you know an amm or let's say you're a lending pool on you know atmos or cosmos or whatever um and you know you don't only want to serve over collateralization you know for your um for your lending logic you want to be able to underwrite users for that you need data on your users you know whether it's their identity or whether it's their financial background how much salary they're making and let's be honest as much as we'd like you know 99.999 of the world is not using blockchain for their daily stuff other

than maybe buying and selling and speculating so um you need to be able to ingest that data in a private way because no one is going to put their you know this is how much i'm getting from my employer every month in my bank account you're not going to put that on youtube unless you work in blockchain wow and apparently it's all public yeah but um i i personally don't think that's a viable feature so um yeah so there you know mina can provide this rails where users are easily able to bring um using you know zero knowledge um easily able to bring in attestations about their data whether it's their credit score or their credit history or anything else about you know whatever the underwriter might require and then have that automatically settle

down to let's say fmos just because it's the straightforward answer here um automatically settle down to atmos because atmos would be running in full node of mina and be aware of all these state updates so that's kind of the end-to-end flow that we're imagining here it's such that any other layer one application can benefit from this zk attestation superpower cool i think it's similar for elio so i won't rehash all that but i guess the one thing i just quickly want to highlight and full intellectual credit to the of one folks and isaac and um you know and the folks over there you know this idea of this recursive proof right you can one of the things about that snarks let you do aside from cloak or conceal is compress right and this is about compressing a lot of state into a small

proof this is how roll-ups effectively work and but i think there's it's overlooked as a tool for bridging right it enables you to take a lot of state or a current state and compress that down as something that's relatively lightweight this is exactly so you know effectively most bridges are implemented as light clients on the destination chain if you have a very light light client then um that can be a very powerful tool that you know again like mina again or the of one folks kind of came up with this concept others including cello were you know we also are have a kind of a recursive like client that that we use at alia but i think in general this is a concept that can be adopted i think is the future for bridging across the space is using this compressing property of zero knowledge

proofs to potentially make it easier to run full light clients on you know as destinations of what whatever bridge um my confession is that i was listening and then forgot the question the question was how are you connected to cosmos but for you oh pretty yeah i think for us it's like fairly straightforward um uh i think that part of the kind of like meta uh there is that not meta the company no okay we can't use that word anymore by the way i i've actually succeeded in not having that be part of my okay okay so i wish i was so lucky yeah um so part of the thought there is like

about you know how do we sort of have like a small scope that we think we could like definitely build um and i think part of the reason like i mentioned a little while ago is like the idea that if you have really pervasive bridging you can intentionally like spec specialize and like narrow the scope that like one project needs to do and so for us i think sort of like how do we like what is our relationship with sort of like everything else in cosmos it's like okay well like everything except for like the four like bullet points that were on one

of my slides of like stake transact swap and market make like literally everything except for that you should go and use another chain and that strategy only works because of sort of intentionally conceptualizing ourselves as part of this like interchange ecosystem rather than being like trying to like we're not trying to build an ecosystem we're just living in one cool all right i feel like i've taken up a lot of time with these panelists i do want to open it up to the floor i don't know if anyone has any other questions i do have more stuff if you don't but i wanted to share the option so yeah this is

questions for our zk and privacy experts hey guys good panel sorry couple questions i guess first i'm curious of what what percent of applications do you think will end up being privacy preserving in the blockchain or web3 ecosystem right because traditionally you know like in web 2 it's been an uphill battle and the percent is actually pretty low kind of curious from the demand that you've seen um what just yeah just give me a number and kind of think through this and i think a second question that i have is around security and how you think about it kind of you know zk privacy preserving tech in general is very complex right and like a lot of credit to a lot of the builders curious how you approach security how

you think about you know audits and things like that and how do you think about catastrophic kind of failures when privacy or when there is a bug when the privacy is leaked and compromised kind of how you think through those things thank you yeah um percentage is the first one yes so you mentioned the example of web 2 and not a lot of things are privacy preserving web 2. i would actually push back against this and say the model is different right because your bank is private to 99.999 of the world population it is public to the organization the same is true for twitter facebook pretty much all web two things are privacy preserving against the general

public unless you're like a nation state like if you're the nsa then you're not private against the nsa but pretty much against everyone else you're fully private and i think as a result if we want to build web 3 we really have to think how do we get to a similar model where very very few people have access to all the data because currently the model we have is ethereum and everything is public and just think for a moment that like world finance moves to ethereum tomorrow um and everyone is fully public like this is not like that it's web 2 public it's like no everything is completely public on every human on earth and the next day someone sort of in the middle of nowhere with

internet connection can run a global financial mass surveillance program by having an internet connection it is like we have to be way more privacy conscious in web 3 than in web 2 because we don't have a bunch of institutions that effectively guarantee our privacy for most parts against the general against the rest of the world so i would say like we probably need i don't know practically speaking 99 of all applications event on web 3 eventually need to be private because otherwise this is like a significantly worse security guarantee and privacy guarantee than web 2 for most normal users of these systems i was going to say the exact same thing well not the exact sorry no no but what i would add to that

is um like at the end of the day blockchains are a global state layer right that's what they are and not everything needs you know a global state layer either right so it's it's in my opinion it's going to be a part of you know the internet and if you cannot somehow preserve your privacy with within this spec spectrum then you know you may not need to be using a sort of a website landscape but then also on top of it what percentage of web3 if we ask it that way i think it also depends an average user i would say yeah like definitely 95 of it will end up being private because i don't see anyone in their right mind sharing this much information you know on an ongoing basis um but maybe as a you know in a tvl

sense maybe it's less than that because um ultimately if you kind of as a user you interact with some private layer and then that settles down to again maybe ethereum ends up being this global settlement layer for all the value and it's not private and that's fine because what what you're seeing on ethereum doesn't give you information on a per user basis it's aggregated so that might end up being not private by definition but also you know private in some in some other sense it's not linkable yeah i think the first two so i agree with what these two gentlemen said on the pri on the you know kind of importance of privacy i think yes a second question though which is around complexity um i think it's a great point and you know remember one of the hard things about private

protocols is that if something goes wrong sometimes you can never know i mean zcash very famously had a bug and was it ever exploited do we know i mean probably not but we know it hasn't been exploded yet okay well we know that the return style hasn't been broken so but the point is like you know that it's these things could go undiscovered and that's a little bit dangerous but on the other hand i think in some ways it's more secure because you're you're putting more assumptions on cryptography whereas if you look at like a lot of web 3 you know there's so much i mean one of the greatest innovations or arguably the innovation of web that enabled web 3 is this this game theoretic security right but when you start layering game theoretic game theory on game theory you know at some point the assumptions you have to

question whether or not it will actually work right i mean there's there's this concept that's kind of been around for a long time thinking about you know yeah like ethereum validators are all supposed to run re-run every state transition exactly step by step by step but do they like or do they just take whatever node they got it from's word for it right if they're like okay this block came through i do i really need to re-execute these transactions i mean they're kind of not really incentivized to so i guess if you're thinking about does ethereum provide me as much or does any model that that relies on game theoretical consensus actually provide me the security i'm thinking i'm getting i don't know i think it's a much less definitive answer than a cryptographic proof right where i mean the idea is if you believe the

math you know of course there's there's flaws found in these things all the time but fundamentally it's a much much smaller window if you can design it right now these systems are brand new and of course because like ask any cryptographer they'll be like if it hasn't existed for 50 years it's broken you know what i mean like that's the attitude and it's it that's an attitude for a reason but you know i think we're going to get there and and i think the the fundamental assumptions that these are resting on i think in some ways are stronger i think on point two uh we should be doing way more formal verification of cryptographic protocols because they give us fairly high assurances that these systems are correct i think we shouldn't this is sort of why i'm terrified of this um one chain to rule them all narrative

especially around ethereum right that like we should design these systems that are interconnected where components can fail independently we shouldn't build like one global financial system on ethereum or on any other blockchain for that matter we should have many different blockchains and they should all be able to catastrophically fail but when they fail they only kill themselves they don't kill everything else whereas if you have like imagine a model where like everything is settled on one security model you all of a sudden have this problem that like well the security model may fail due to cryptographic assumptions um and all of a sudden like global financial sys the financial system just breaks like this is not how we should be building way more for resilience and

that necessary means that like we need interconnected systems that can fail independently and sadly most of the ways that most money is lost in web 3 today are just like way more benign than that i mean it's almost a good problem if you get to the point where someone's exploiting this like deep down dark fissure of a protocol because you know there's so much low hanging fruit unfortunately um but yeah no i think i think it's it's a it's a reasonable concern yeah um that i i think i kind of laughed a little bit when adrian mentioned the sort of like if we imagine like the you know tomorrow the entire world financial system like is on ethereum like if you think about like how that actual that ecosystem actually works like that

just that concept is like really hilarious um but uh to the first part about privacy like because of that baseline of public blockchains just being like this complete like information free for all i think that the the kind of the bar has gone a lot lower than it even was in in in web 2. and so with respect to the the privacy aspect i think that as um the deployment of zk systems uh matures we're going to have to start getting into more nuance about like okay private to whom right if you have um

a model like uh mina or ilio you could actually imagine like doing a lot of like the kind of you know web 2 level of privacy but just with this extra assurance that actually all of the data was like executed correctly um because you can commit that onto this chain and use that as the settlement layer and i think we're going to sort of see this like spectrum of um different like levels of of privacy um where maybe in some cases you know there's someone who's like relatively trusted maybe in other cases you know there's attempts to try to minimize that through engineering um on the second point about the kind of

assurance of the cryptography um this is something that i'm actually very optimistic about but not for any kind of technical reason it's actually a social reason and that reason is that the if you look right like the the bug in z cash right like you know that that bug was not exploited because at the time that it existed there were like 12 people in the entire world who like even understood that could happen and they all worked on z right like it's both reassuring and not assuring but yeah right but but if you look at what's happened in the last um you know say like four or five years and in large part because of

like the leadership of of of z cash right like now you have this whole ecosystem of all these people um who are developing like really really good production quality implementations of quite complex cryptography right like the fact that mina exists is just like incredible right and it's not because there's some like technical advance that's driving that it's because there is now this like production of people who do really serious real world zk stuff and that is only gonna keep growing and so i think like that like force is gonna like wash out

like like all of the technical details the formal verification like people are gonna build the tools like it's gonna happen because there's just it's momentum it's there and that's the power of open source last point because we're one very quick thing one when i explain like why i'm excited about web32 other tech people non-web3 attack people i agree with like the point about like more crypto cryptography professionals and more better cryptography but also like this whole space is an amazing trojan horse to get cryptography into the hands of everyday people right like no one in web 2 has a pub like very few people have a public private key pair in their hands but anyone with a metamask you know extension because they want to buy some nft they get this public private key pair and and they can then use it for

other things which i think is amazing amazing all right sorry yeah this was amazing thank you so much for doing the panel uh i hope you guys enjoyed it i know we're a little bit late but yeah thanks so much thank you for having us thanks and i think if anyone is interested in learning more about these teams you will be sticking around i hope i do want to let everyone know especially those who came later we're actually doing a dinner later tonight here so this event ends around 5 20 now because i think we're 20 minutes late we're 30 minutes late 5 30 and then at seven we're also doing a dinner here and you're all invited so i just want to make sure you know that it would be very cool to hang out and we

might try to do karaoke do you know this okay that i mean some people didn't you weren't here for the intro so i thought i should let you know okay cool thanks a lot