Privacy in Cosmos Live: Panel - External privacy into Cosmos

Transcript

all right so yeah this is the last kind of piece of content panel before our lunch um to give a little bit of framing for this particular panel we did this also last year the idea with this was to introduce ZK projects to the cosmos ecosystem but also kind of get a sense for folks working outside of Cosmos what it would take maybe to get into it maybe as a starting point why don't we intro do a quick intro can you tell us a little bit about yourself and about the project so um I'm going first I guess I'm Florian I do engineering for over one Labs um to give a bit of background I guess of one Labs is the company that

incubated minoprotocol and we remain a still a very active core contributor to the protocol amongst other very capable companies so Mina is one thing and then we do have a second thing called snuggages which is a very powerful typescript um sort of like CK toolkit SDK which allows you to write Mino smart contracts your knowledge smart contracts and recursive circuits yeah cool and Daniel um so I'm here representing Elio but I don't think I was necessarily outside of the cosmos ecosystem I used to lead the protocols team at stake fish and yeah Elio is a you can call it a

zkl1 and yeah fun fact Matt Greene was my advisor in grad school and so I'd known about Elio and its papers or knowledge execution for a long time so yeah so so a lot of the work I do is very close to that I mean it's very much in the ZK ecosystem but you might be a good person to speak about this you moved a little bit from I mean you had been very active in Cosmos what's the change like when you when you step outside into ZK land I don't think I necessarily left Cosmos but I did I did rage quit my last company so but I I think um yeah I ZK so I had actually

been a part of I guess like the Z case I used to do bioinformatics research for a few years and we we use fhe a lot and then we had to use EK in that but it's really exciting to see ZK find its sort of Home in um immutability getting some Unchained verifiability and um the entire stack of composability and that blockchain offers though yeah and maybe to give some description like when we talk about the ZK ecosystem it's sort of a collection of it's sometimes it's like it goes from like applications who want to build on ZK systems using privacy all the way to like zkl ones and I think you know a lot of the

conversation last year and something we want to continue is this idea of like how do these projects either applications or the l1s actually Connect into the cosmos ecosystem is there any benefit to doing that I'm wondering if you if either of you have had thoughts on that especially in prep for this panel yeah so um I mean it's a cosmos panel or like Cosmos event after all so that's a very important thing um demeanor protocol the Mina ecosystem is actively considering integration with other blockchains for example there's the Neil Bridge which is a trustless CK Bridge which allows you to bridge the meaner State over to ethereum and you know that that's just one bridge and I

guess potentially in the future there could also be more purchase to other ecosystems and chains so that is one possibility so yeah yeah maybe yeah I think that's exactly right maybe another way to look at it is perhaps at the infrastructure service provider level let's say like the validators um I I think that validators have like this unique position to be able to see and um support a lot of different blockchains so Cosmos obviously and then ethereum polka dot tezos Solana and then Elio for example and I think they they validators can be a pretty effective bridge in both understanding the tech stacks of these different ecosystems and

then they're probably some of the best perhaps advisors in these blockchains and the ability to connect so I think there's a lot that the L ones or even l2s have to learn from validators and kind of seeing like right as an example right blockchains will rise and fall some some of them will disappear but then the validators are mostly always there if they don't die so yeah and I mean you just talked about bridging to ethereum and that does seem to be a narrative that's grown in the last year versus like literally since last year um but I mean obviously we're here in a place which which you know promotes more

like an IBC route do you think that sort of that look at ethereum is that almost like a bear Market phenomenon and a look back is it sort of like oh we'll go to where the it's still warm and not necessarily to where the future could be yeah so um I guess my personal opinion is that um ideally you would connect all the blockchains with each other it doesn't matter if it's ethereum or non-etherium um certainly at the moment ethereum is a very hot topic that's where all the I guess capital is where all the projects are but I know that for example in the community there are people working on Roll-Ups um on the Mina in the Mina ecosystem and

they are exploring a sort of like layer 2 interoperability Universe which I can imagine but also then interact with Cosmos look at potentially IBC integrate with that so I think there's a lots of potential but they are currently at the the Hot Topic is ethereum but I hope we can continue and then look at other blockchains connect the overall blockchain world so maybe yeah maybe just to add on to that it's it's interesting because um when there are these like sort of connections or Bridges to other ecosystems sometimes it's like if we connect ethereum then we hope that ethereum devs will come onto Cosmos and then develop

on our side or or maybe it goes vice versa there's there's a pretty big like centralized exchange chain that right we had been in Converse like Cosmos some of the cosmos folks that have been in conversation within like the bridging was great right we could we could do that and it was just a matter of how we needed to upgrade certain things on their end but then the majority of the conversation had always become like oh like what what do you want out of this bridging it's like well we want your devs right we want like Cosmos devs to come over to build on your thing and then maybe like so I even though like I think ethereum is is great and we're getting a lot of this sort of connection point but then there's like the flows are going in in certain directions and

and perhaps like that bridging is that we want to like suckle on the teeth of ethereum and then have all this like you know extra liquidity come in or but which is not a bad idea right but then I think you know we we eventually are like Kumbaya in the beginning and then when someone gets big enough they want to be their own King and so yeah I just briefly want to add to that um yeah I totally agree I feel like most bridges are even though they may be too big they are like mostly leading towards one side so I think there's there's a lot of potential for a truly interoperable layer 2 world we're like chains interact with each other they they benefit from each other not only one Central I guess

chain and just a very big beautiful interoperability world I have a question that actually speaks to both projects in and outside of Cosmos which is how do we get people building private applications is it like what I mean I think both of your projects have a version of that but maybe even think past that what do you think we need to get more people in because like that like if you only have a small pool that's like moving in and off like around Bridges and still the same devs you're fighting for a very small pie so how do we grow the pie yeah so um we have heard that in the previous talks I think a very very big important thing is user and developer experience right on

the application Level for the users but also on the um sort of like tooling levels for the developers um if we only have a very I guess like centralized minority that knows how to use these tools then we will not have any more developers working with it there's like an entry curve a lot of learning you have to do to develop meaningful applications so it's important to make that more accessible also on other chains overall very accessible toolkits easy to understand so you don't have to be like a cryptographer or like only a couple of handful people that are already doing the same thing on five other chains and then just I guess copy pasting that over so like opening up the developer

experience developer pool and then I think users will follow they will use these applications if they are good which they hopeful are so yeah I also think that ZK itself is not just a web 3 thing as I'd mentioned right we were using it when we were doing just bioinformatics genetic sequencing research I'm just sharing data between Labs um I think that opens up this huge opportunity of like the web 2 world that has an opportunity to use like this Tech platform that is going to be on on blockchain and and I think a really good example is I think I don't think ZK is necessarily like a speculative value proposition that blockchains are providing that have

ZK functionality it's more of like a useful value like for example right Bitcoin people some people hate it some people like it and so like there's like that that argument that all Bitcoins Burning Down the planet a lot of people use the same sort of argument with AI you know using massive GPU farms and ZK is also using these massive GPU Farms that eventually turn into fpgas and A6 but then right like the the argument like so so for example from the climate perspective you know the Aerospace industry uses nine times more image nine times more carbon emissions than Bitcoin but then people don't like vilify it right as they do with it because there's value

that people have but with ZK I would say that's more analogous to the Aerospace industry in that the the value is not necessarily speculative in that right just like number go up but then people are using it right like people are using it in consortiums we were using it in like genetic like sharing data back and forth between labs and fhe like there's like a lot of gaming applications Etc ekml like hiding some of them and allowing people to be more comfortable sharing their data to train these models I think they're like I think the key point is that DK allows the web 2 world to interact with this technology platform that is native to to these to to the blockchain and that will I think naturally allow let's say clients to

come in or different Industries to also use this because it's a it's a necessary component and that that that special part is basically the the the rise of like the sort of pow style ZK miners that are going to have this Hardware that's now now much more efficient because so much money has gone into the r d to develop them um because right the value associated with these tokens a lot like what happened with how bitmain was able to get so much money into r d because the Bitcoin value went up and they're able to develop more efficient and effective Asics so I want to kind of take a little bit of a step back to some of you said earlier which was about the languages and so I

know each project I don't think you mentioned actually Leo the language that Elio developed so there is this fragmentation happening a little bit in the languages the VK dsls I think I know at least 12 maybe more and almost every blockchain will have its own every ZK focused L1 will have its own language do you think this complicates it being adopted into other ecosystems do you think that if we had maybe even on our side a bit more of a standard that that would help to get these EK tools into the other ones yeah no I mean certainly um there's like you said plenty of dsls um some arrest like some are like typescript libraries

not necessarily as well like what's not good.js is um so and I guess it can be confusing at first for developers getting into that there's a lot to choose from not all do the same thing all require I guess slightly different like a knowledge of certain implementation details so I think it would certainly be beneficial to just overall improve the developing experience and that certainly includes the I guess DSL tooling layer how to write applications how to work with them right I mean if you develop an application on one chain you you want to make sure that however you can transfer to other chains you don't want to like rewrite

everything every single day for every other potential use case so I think it's it's certainly very beneficial to have a common ground if it's not on the I guess DSL layer then maybe on the like a circuit representation layer making sure that you can integrate with other Stacks other tool systems that includes testing and all that stuff so yeah maybe I this is maybe it's just maybe I'm being naive but I think that if there's a purpose and reason that people need to build these EK programs then you're going to have to just learn the language and suck it up I think I don't I don't think it's necessarily a problem that there's many different

languages um but right if if there's a language that you need to learn to build a ZK program that you need to serve a particular application or purpose or initiative then just learn it like I don't you know you just because it's like you have like this evm compatible ZK stuff and it's like maybe it's not fit for purpose but these dsls exist for a reason because they are fit for purpose to build these EK programs maybe I think what this makes me think of though is this idea and I've talked about it a few times on different episodes of like if you wanted to be private in a multi-chain context what you're describing is each side would have to like be right written in a different

language and I think that leads to a lot of vulnerability and actually I kind of want to even bring this to IBC and maybe also other bridging but like you know wouldn't it be kind of cool like I think IBC is cool because of the fact that it's this this thing that everyone can agree on like wouldn't it be cool if there's also a ZK language that was sort of agreed upon within maybe this ecosystem or like more generally I realize this is like this is a big ZK topic yeah maybe like you're saying maybe it's a standard maybe it's like the bridging State like the communication standard which IBC I guess can yeah Implement so no yeah definitely I think that's

probably two different types of um standards that you can think of one on the very high language layer like having the exact same language and then I guess something lower which may be the circuit representation or it may also be the communication between proof systems so that's also a big part we've only been talking about languages and they then compiled to their own or like use utilize their own proof system to do the proving of their proofs but they they can't really necessarily I guess interact with each other on the proof system so if you have a proof on bomb chain or like from one toolkit then you have a proof on the other toolkit what do you do you can't like you can't do anything with it so you essentially lost all the benefits of CK if you have two

this different proofs um so I guess it would also be beneficial to look at to a look at the integration level of proof systems and I'm not a cryptographer so I don't know how to do that but I think I would also be beneficial or maybe I don't want to keep shelling like validators but like I mean maybe it's like right because if there are these different ecosystems a lot of these ecos different ecosystems share the same values right there's obviously like The Usual Suspects but that maybe if they they are the same maybe they are the sort of like semi-trusted party that is able to communicate what needs to be communicated between those different ecosystems um but yeah yeah I think I think this is an idea that

until at least for me is still kind of developing of how and we we had talked about it like privacy on one chain privacy on another chain often coming in and out like going over a bridge just automatically removes the Privacy we had talked about like would there be private Bridges possible could IBC be made private could it if it Incorporated some and maybe it doesn't need the language so much on either side but yeah I do think there's there's a conversation here um I feel like we should have an IVC expert up on the stage to help us have that but um okay another question speaking of IBC is do you ever imagine implementing IBC could Mina or Elio

be and I like could you could you tap into this directly or do our privacy system is built so differently that it would be hard to do yeah um I guess we officially do not have any any plans at the moment but I can imagine that there are many community members building like compatibility layer tools and all that so I could certainly imagine that they take a look at that and and at least um take that as an ideal um certainly build up on that and maybe integrate with it but I guess we currently don't have plans in regards to that we are currently I guess more focusing on the approach that I mentioned and just to clarify what you were saying there is like you could have

a roll up on your system that would then have IBC integrated potentially yeah yeah I mean potentially right and then that could also integrate with other Roll-Ups other IPC compatible change and then you could do that again and just keep going I think if there's a very specific purpose then the bridge should be should be built like if there's just like a bridge to just have a bridge then like as he was saying like is what direction is is the things going yeah but I mean the ZK prover ecosystem like these pow style ZK miners they're free agents right they can they can provide their Z their proving services to any chain

right if there are the cosmos ecosystem has a has the ability to kind of like use that then right that's just you're just sharing um infrastructure providers and then they're in is like the closeness that you can have between the two ecosystems that's actually that that whole world like what that will actually look like I've obviously heard a few times it described but maybe we can just talk a little bit about it which is you have the sort of transactions and the blockchain happening over here but the proofs are potentially generated in a different setting Often by a validator-like entity I know Alias built quite a lot about that around that yeah just just for some context ileo is like

this hybrid POS pow style system or the POS is just the block building and consensus and the POS stuff is uh the pow stuff is this like ZK mining right you just you do your chain off chain proofs um and and it's fascinating because right you're getting to see these two different ecosystem these infrastructure service provider ecosystems interact so just coincidentally right like when I was at stake fish our sister company was F2 pole which is like the largest Bitcoin mining pool at the time and so I was quite familiar with both of the infrastructure providers on both sides um and so now we're beginning to see because I'm I'm set helping set up the validator and prover ecosystem in Elio

we're beginning to see like the old like all the all the infrastructure providers kind of take a take a look in and and what I think is the most fascinating um is that you're going to begin to see like these sort of hybrid organizations where you'd have both the validators take on like the ZK mining capabilities and the ZK miners take on the the the the validator capabilities but I mean I've seen both sides and I honestly think that it's much more difficult to take on the validated responsibilities so I see more validators taking on um you know the the plug-and-play ZK mining um stuff but yeah I I think it's uh

it'll be interesting to see that the ecosystems interact yeah I also Mina has something like that as well no um sorry I blanked for a second oh no worries um yeah we have something called snog workers which are essentially like a prover infrastructure Network um they are providing the proofs for transactions right so if you send a transaction um to give a bit more of background um you send a transaction it's being proven using a snark broker and then recursively I guess merge together with other proofs and then you end up with one main proof sort of like um testifying the correctness of the entire chain so we we do have something called snark workers which is that

network of approvers doing that proving and then there's something called the snarket place very snarky names um where you can like they have like some some incentives right if you run a snark worker you you generate a proof and you get some yeah sounds good cool well I've reached the end of my questions but I do want to open the floor up for more questions if anyone has any um so we can oh sorry my bad [Music]

uh hello um so uh at penumbra we uh do IBC permissionlessly to everywhere but of course IBC is transparent and uh one of the things that we've been trying to think about is um how uh how do you manage to do private IBC um without uh having a problem of like if one of the counterparty chains misbehaves then you leak uh you you could potentially leak the you you could get contamination from a counterparty chain because ordinarily in IBC you have like circuit breakers but doing a circuit breaker with an encrypted uh or or zero knowledge IBC connection seems

tricky I'm wondering if you have any thoughts about like what directions those might go in yeah I feel like maybe like maybe I I don't know if this is answering the question but like Elio has like this program registry that allows for differential privacy and if there is a need for a certain counterparty on another ecosystem that does require some selective information then like that can be that can be opened up but did that yeah so okay yeah maybe we should chat offline about that I'm curious yeah I mean you're certainly the expert on that um

like an important question to actually talk about that's what I think I was trying to Hint it yeah that Vision but I yeah I think maybe I don't know if we are the right people I feel like we I feel like we need to solve this problem um in some kind of way whether it's through through like homomorphic encryption or um or NPC or something so that we can like contain a malicious counterparty chains in order to have like real private IBC that works this is a question actually I'm going to ask you this but like have you and this is maybe the the counter IBC argument would it be easier better do you have more flexibility if you actually do a bespoke bridge that uses one of these techniques is it is it harder to try to shoehorn it

into an IBC setting I think it I think probably the thing that we were talking about would not be IBC it would be another inter blockchain communication protocol like but but I don't think it would look like the the protocol that already exists um so it is different yeah yeah and and then you have these problems of like if every chain has its own shielded pool with its own proving architecture then how do you uh how do you do proofs that like translate between like you know you're using Planck I'm using growth 16 or whatever like what what do you do I don't know flap of standard on it try to get a standard and then you have then

you have 16 standards instead I mean I think IBC as a standard for this ecosystem is incredible given how early it was thought up and and I think like in the ZK space it's not just zkdsls it's proving systems it's like even like circuit-sized techniques within a proving system because you have things like Planck but then you have like variations on Planck um since then Halo 2 as well yeah yeah cool well any other questions oh maybe one more right there hey um this is a significantly more basic question

um so like you mentioned about ads talking about kind of like Bitcoin and you talked about um Aerospace and kind of just like how there's some like the ways that ZK is actually is quite useful for getting end customers outside of web3 I wonder like what do you think are the most like suitable end customers for ZK like and how do we approach those better so I think when you say customers I think the way we tend to talk about is like use cases so like one big one would be like id zki id zkml so this is where using a blockchain is sometimes really useful for immutability and consistency or sorry but I think that at least in the zkml case there's

already examples where you're using ZK to prove that something is accurately done within a modeling training and stuff like that like making sure that the data that was Fed was correct that doesn't really require a blockchain anywhere that's just using zkps so maybe yeah to speak on that when I mean zero knowledge proof technology started in the late 80s and it was sort of theoretical kind of under the radar people did try to use it for passwords I think in the early 2000s but what happened was like you melded it with blockchain all of a sudden all this like excitement maybe Financial incentive a lot of money poured in libraries got built mind share was grown and so it's grown like all the

zkp systems that we know today a lot of them were developed in the last five years the one the sort of the common proving systems the snar the the plant the snarks actually the snarks the Planck the novas um but um but yeah like what's funny is like it accelerated through blockchain and now because these systems are so good you can start to see them spilling out into these other realms sometimes Tides still back to a blockchain sometimes not but yeah I I I'm very excited to see where this is going like the topic of ZK outside of blockchain is like fascinating and not super well specced out yet the only one I mean the one I

can point to right off the bat is this like machine learning like showing that the data was correct I don't know if you guys have other examples learning like that stuff and and I I do think that blockchains uni or like the cryptocurrencies in their Associated blockchains particular use case here is that you have like this speculation and then the value increases and then you're able to get a lot more money into the r d that goes into developing more efficient and effective ZK Hardware because if the end goal is this like ZK Asic or some people come like zpus or something like that um then you have like this what what you know like what was not possible in the

80s or the early 2000s we now have um more accessible and potentially even commoditized to to serving whatever different use cases there are Bitcoin is I think the perfect example right we were able to just do it on like crappy Hardware obviously the difficulty went up but then we have very fine-tuned efficient and effective Hardware that exists now and not later because bitcoin's value went up and they're able to use that money to to support it so um yeah just I very quickly um just to generalize that I think there's probably a lot of potential in the web 2 world when it comes to proof of model so for

example CK learning that's a proof of model how do I know that the thing you're giving me is like truly that awesome AI thing and you didn't do it yourself that would be awful or maybe things like how do I know that you're really offering me like the the best price on like a flight website comparison um how do you like how do I make sure that you're not artificially pushing the price for some reason potentially or maybe peer-to-peer interactions with users how do I know that you are who you are that could be chats potentially even games if the technology is Advanced enough so yeah I think the like this the use cases

have been expanding and the ways of using these things and I think there's a lot of Uncharted Territory so there's actually a lot of space for creativity and trying to use them in certain cases seeing if it works I do want to just mention like some other cryptography that also is starting to be like or has been for a while but tied in with ZK so sometimes people use the term ZK as this blanket term for advanced cryptography so you have things like fully hormone or fully homomorphic encryption or multi-party computation um so you what we are seeing more and more is like these systems trying to incorporate pieces of those as well maybe combining them and there's going to be some very powerful interesting

things that come out of it I think I think it's fair to predict that all right cool thank you so much for staying for this morning session we're going to break for lunch and thank you to our panelists for the what is it external privacy into Cosmos panel yeah thank you thank you [Applause]