Privacy in Cosmos Live: Zaki Manian (Sommelier)- The Unbearable Weight of Sovereign Multichain UX.

Transcript

oh yeah so first up we have sucky where are you there are you giving a talk on the unbearable weight of sovereign multi-chain ux exciting all right um I guess right sort of like the theme of talks from me lately has been ux um and ux has been on my mind because growth is on my mind both growing similier and the challenges of like onboarding new users primarily from eat the land um into an application but I've been basically doing a lot by because of similier and other activities I've been

very much more on like the application side of the of of Cosmos and the blockchain ecosystem lately and the uh and like what I have become much more aware of is like the incredible challenges of building like a user experience that can like onboard users in any meaningful scale um and how hard these problems are so just a little talk about like this is going to be a quick talk just about like challenges that I see and like not necessarily proposing Solutions uh but encouraging people to think about them more um I also would just say that like essentially Cosmos growth for Cosmos at this point is existential

um you know I would say mostly are sort of a monthly active user numbers have been flat for the last year and if we don't change that in the next year um we are not going to be around um so basically the sort of General theme about things that will improve uh in like the ux in Cosmos is when we've had account abstraction throughout Z in production for you know almost two years um but like relatively limited uptake um and so like why is that what can we do about that um IBC middleware is becoming more and more of a of a thing I think like one of

the more interesting ux things that recently shipped um is stride doing um uh stride doing one click uh uh liquid staking so like basically you start the liquid staking process but then all of the subsequent transactions are delegated to the IBC relayer to complete um and then there's this like the magic word of the moment intense which I'm mostly not going to make this an intense talk um so what are our challenges um gas has been like sort of a never-ending uh challenge for the cosmos user experience um most you know Kepler recently shipped like making it easier to use alternative gas uh gas ux could be improved a lot of

is just social coordination like agreeing on coins that can be used as gas and coordinating all the validators and the uh and where you where where needs to be governance to support a single gas token um it's still like a complete nightmare to you to like manage hot and cold wallets um multi-cigs are usually are Pace like the native Cosmos multi-cigs are still very like non-native to the ecosystem um signing multi-step flows where you have to do it and like deal with like various latencies are a complete nightmare so like I want to bridge from eth to uh I wonder bridge to from eth uh via Axor to osmosis and then do

something is like you wait 20 minutes and you do step one you wait 20 or 30 minutes and then you do step two and like it just shocks me that anyone would do this um and then whenever anything Fails Like failure and Recovery is like an even longer uh and more complicated process like I find it hard and you know I the like I have the whole architecture of the system in my head um and then privacy is going to make all of this like much much worse um so it introduced off-chain State um viewing Keys um basically every every Cosmos chain has been running the same uh staking and distribution implementation and privacy

basically inherently brings us multiple uh implementations of staking and generally like overcoming like the down only tokenomics of Cosmos involves building multiple implementations of this so from like a wallet and a dashboard point of view instead of having like 45 50 index implementations of exactly the same software now we're going to start seeing more and more Divergence in these systems um and then we don't really know what the Practical trade-offs are like metamask snaps seem to be like one of the most important things happening in Cosmos but like what are the practical trade-offs between like using a snap and a dedicated wallet like Kepler um

so things that confuse me are um like I don't know why Cosmos wallets don't come with like a built-in bot interface that like or in like first party Bots that come from the wallet providers that just will automate things like I don't know why like yield most is a separate thing from Kepler um it probably shouldn't be um I think wallets should charge like monthly fees um I think like the wallet monetization model is so far been basically like pay us as a first party project to like you to be like to get like user reach um I think we need to like definitely be moving more towards where wallets have

like direct relationships with their users um uh I think I left off the last but we'll forget that um the other thing is like session keys so like one of the other things that we could be doing a lot better in Cosmos is rather than like you use stargaze I'm going to like make a bunch of transactions why do I have to like sign with my Hardware wallet for every single thing that I did um I just want to bid on ffts um like we should be we should be doing more about delegating restricted capabilities to uh this is another basic use case of our uh of our Aussie stuff but like especially in the sense of like

privacy enabled applications as well right if we're going to be using like different cryptography different like that may or may not have Ledger support like how are we going to do that without being able to like delegate restricted functionality to like sort of more short-lived keys um and then finally we have like the sort of meta thesis which is that like topology's in a single chain world are really hard like how do I get from like I want to use an app that takes you know that takes weth and I have like a bunch of different liquid staking tokens a bunch of like other tokens how do I get there is already really hard because like swapping pools liquidities slippage

depth now when you add on top of that like multi-chain ux um this becomes a nightmare um so you know basically the reason I have been running around intent pilling people is probably multi-chain ux is going to run into fundamental barriers until we get to the point where we have an intense settlement interface like what anoma is working on but that generally seems to be to be like where we eventually go but I think what we see in like the sort of IBC middleware world is like an early form of intense sort of expressed by stuff like squid and like the stride example that I gave before we're like you can delegate work to IBC rewares

um you know we do have Searcher markets like skip does exist it does have a search or Market we do have this IBC relayer market like these things do exist and like need to be extended in leveraged to improve ux so thank you very much [Applause] thank you Zaki are there any questions yes okay so it sounds like a lot of what you're talking about would be addressed with if it was more competition on the wallet side uh and I I mean maybe that's not all of the problems but don't you think I mean it basically do

you think there's a specific enough effort to give more grants to create more wallets that maybe then compete on these different aspects on the ux side I guess the like to me the question so like I guess I didn't like I okay so my question with this is is it's like so wallets have been like like the wallets have pursued models where they're like hey like we're going to be like a least common denominator over a lot of different things like that's basically been the the wallet model but like sovereignty means like a continuous drift away from least common denominator like the you know like the other example that like is

there is like so like the primary difference between like evmos and Polaris which is the bear chain evm Fork is like evmos is like let's have a bunch of Cosmos SDK message type and then have an evm message type Polaris is like there's no Cosmos SDK message types you have ever really interact with you just interact with the evm the cosmos SDK is a layer that sits underneath an evm for like doing cosmosy things um but that means that like the least common denominator among like Kepler wallets like just like isn't going to work like you're not going to be able to see your staking and delegation and even then you wouldn't be like it's not it's not the standard Cosmo staking module

anymore right so like I think we have like a real challenge that like um is going to have to be explored which is like what are what is the right way of even constructing wallets into this world because like the answer might just be that there's like a bear chain staking snap and that's the answer right um and like the least combinator denominator like wallet support all the same thing thing just like doesn't work and like the same thing is like how like how do you get privacy support into Kepler is like sort of a completely unanswered question hey

um we've been exploring some cross chain ux stuff that uh that is you know sailing onto this topic um and in particular we're we're trying to do like multi-chain flows where um we want to make sure that the the funds are recoverable um in the middle yep and one of the things that we're you know running into is basically um alternative key derivation paths on different chains yep and this the complexity of this just explodes when you get into like the session key World yep and I I don't I don't know how to to kind of

square these two um it it's really important for multi-chain automation to to be able to have uh you know user custody across all these different chains but um you know we're kind of like moving further away from from that with um there are like bip 30 so like HD path thing is this like hierarchy of hashable values um and there's probably like so like the origin story of this whole sub k thing like I will briefly digress and hopefully be like is like once upon a time uh the treasure people and The Ledger people came up with a standard

um uh which uh they were like okay our Access Control mechanism is like different blockchains will have access to different sub trees um in this like Merkel giant Merkle tree of of like private key values so you'll have your mnemonic at the top and then you'll have different sub trees one thing that like is sort of an interesting question that was like what how I think we should probably be doing session Keys is basically have like a sub tree for session keys right and so you could have some sort of local thing that like gets the root of the subtree value on your local device and then passes it into a less secure

environment so you never use those key that like sub tree for like high value key storage but like you could use that subtree but like basically it would be like turning the like Ledger um uh uh The Ledger uh like Treasurer Access Control mechanism a little bit on its head um and using it for like non-hardware key storage and then having like applications that can like you know it's always it you know like the general policy should be like uh we could even have an like an a ledger app that like derives that session sub key and like makes it available to the app um like so there's there's stuff that we could do there it's like a bunch of

coordination overhead um the alternative is just like separate mnemonics for like separate session keys and like long-term storage keys good any other questions okay thank you very much Zaki foreign