zkSessions: DeFi + Privacy Panel with Tarun, Zac and Stani

Transcript

just just a sort of a preamble this this whole topic of define privacy i feel like came up through and i think there's most of you who brought this up on the podcast just like it was more challenging than a lot of people thought and with this entire event and this panel we hope to kind of scratch into what that actually means why is it challenging what has potentially already been presented today that helps to solve that and what do we think might still be off the table um maybe we do very quick intros to each of you uh in case this is the only talk that people are catching so stani do you want to start it's a short introduction i'm i'm stunning i'm the uh founder and ceo of the avi protocol

uh and uh ave protocol is practically a uh decentralized uh liquidity market burned by by the where you can actually supply liquidity uh in pretty much earn interest rates on whatever you're supplying at the same time you can actually use that supplied asset as a collateral uh borrow other assets or delegate that credit line for some other purpose to someone else and uh team building smart contract based for now for four and a half years uh before that it did actually a lot of uh

work on the legal industry and and by uh by education uh lawyer and before that i did some fintech work so to practically build fintech applications related to finance so very much about finance related stuff nice zack do you want to do a quick answer to yourself as well sure thing um hi so i'm zach i'm the cto of the of aztec um and we um uh we focused on delivering um uh transaction privacy to blockchain specific ethereum um uh so the ability to um anonymize cryptocurrency holdings um and longer term the ability to

effectively program private smart contracts and have um private um uh interactions both with like uh at a piece of paper level but also with um diva criticals and to do doodle with scale uh it's the end goal cool cool terrain [Music] uh i'm tyrone uh i'm a founder of a company called gauntlet we kind of assess the uh economic security and defend protocols um kind of testing sort of things that are outside of sort of just the uh you know types of things that you would get from an audit and formal verification um and you know previously i spent much time in

high frequency trading and then also in building asics um which is sort of how i ended up in this space um and i've spent a lot of time thinking about how the current designs are kind of very anti-privacy in some ways be more due to limitations and so happy to be on this panel mm-hmm you're also a sometimes co-host of the zk podcast also sometimes coaches know you're from the vk podcast also invest in a lot of uh kind of projects like stony so cool so the first question i actually

want to talk about is to get us all started on this idea of privacy and defy what like do any popular d5 applications ones that are very well like very you know known and used do they offer any sort of privacy today i'm actually i can throw this your way stani like does ave have privacy and if not why why not kind of yeah it's uh it's i mean the the nature of like defy is is about bringing transparency in in the sense you know transparency to the markets like if you look at traditional finance i mean it's very difficult to to quantify what is happening in this

this like not just payment networks by trading networks and you know if you look at just like centralized finance you might see that how much like cryptographic funds they have uh in balances like commonly but then you don't see that kind of like trading activity exposures and so forth d5 is trying to kind of like solve that as one of the uh byproducts you know because it's a uh you know server anonymous uh ecosystem and uh you have smart contracts and you you have to kind of like see what the code is what the transaction is transactions are uh you bring that transparency that is why like d5 products are uh like very good value proposition

in our protocol because you can quantify the risk that is in the protocol and and also like you you can see like uh the interest rates they're built on chain and and you know and how they are governed but that's the kind of thing i think like the first step in d5 has been like like uh this transparency by default but now that we are interacting more and more like we don't have it that's that's that's why uh that's what's the issue kind of like that we have been solving in with with define in dfi transparency now the the privacy is kind of like the next step there are some stuff built like on top uh i've seen this kind of like a

what it was like uh blender and another project called hush hush and and what they're actually doing is that you could you could you could withdraw and borrow from the protocol without uh you know revealing your identity in the similar way i store under cash and uh and then one idea of remaining retaining privacy is like use something like tornado cache and you know you do it before you go into the dfi and and so forth but i guess the problem is that when you start to accumulate wealth and value you know it's just like different tokens and so forth but you know nfps as well then your identity starts to expose a bit and when you actually change addresses and you start interacting with other addresses you

know the way you behave can be tracked and attributed and that's going to be very difficult so partly why i'm here is also kind of to learn learn more about uh like how these issues can be solved and there's been like very interesting solutions let's me to another question which is like what can analytics companies see and i and i guess from what you've said like everything if they want to throne yeah so like you know i think you and i hosted a episode of the zk podcast where we we interviewed um alex from nansen who kind of has

probably the best tag data set in all of ethereum um of basically millions of identities so like to give you a comparison people who charge a lot more like chain analysis and sell to governments maybe have tens of thousands of identities um kind of mapped to individual users or funds or uh types of users and so i think the interesting thing that we're seeing kind of right now is that the beauty of being able to build things composably comes at a cost and that cost is sort of like a uh well composability increases the attack surface area but

if we get past that the second thing is it comes at this cost of uh you know you now have chain you now instead of having like a single dollar bill that gets observed in a transaction moving between uh personnel and person b having the whole history of like every contract interaction means that you have a much richer sort of like graph database of people's interactions and i think as as guillermo's you know talk showed and that that's like this paper working on kind of to kind of give a little generic framework for how you have to think about this stuff is that defy has this problem of for fair and transparent pricing or fair and

transparent liquidations you have to have a ton of public data you have to be able to make it so that anyone who wants to participate can participate but at the same time if you want private data you kind of need to make sure that you can't kind of reverse engineer the public data that's being given and the current mechanisms because they were built around the computational kind of like constraints of ethereum and also sort of the needs of of traders who may not care about didn't care about privacy at the beginning um you know didn't really think through this type of stuff and so at the end of the day there has to be a better way of doing what we do in the normal

world which is split up public and private data um and you know hopefully things like plonk and and other improvements will actually get us to the point that it becomes easy for developers to do that without you know having to understand what's happening i mean you know as much as i love you know learning about you know class groups and you know kind of more esoteric things that i didn't think would ever be useful in real life from my undergrad degree i also know that like 99.999 percent of app developers don't want to know what a circuit is

they don't want to know what a polynomial over kind of a semi-simple ring is like none of that all that stuff needs to be hidden from like app developers and we're just like not there yet when you talk about that disconnect between kind of the the app developers and the actual protocol folks who care very deeply about privacy like do you think that people care that's actually the question is like you you sort of said like now it could be useful it could be important to start incorporating or thinking about it but like do the actual users of the d5 protocols care have you heard that like that ask

it's actually to all of you yeah to go on um yes uh obviously i'm i'm probably quite extremely biased um uh but um i do think privacy is quite an essential component for um to for ethereum and public blockchains to really realize their full potential because the more the more transactions get put on chain the more the more of your the more people's the more of your life you're putting on chain um so certainly when ethereum you know in bitcoin were just starting out and people people were just um playing like it was it was just kind of um a fun project um then you know it was it was falling as purposes private you know

nobody knew what an address was like how to link addresses and identities but now especially when people are paying all the value on chain that's no longer the case um and i think if public blockchains and decentralized tribals are going to truly compete with um kind of the existing private kind of web 2 financial infrastructure um you need privacy because people people consumers expect you know people don't want that purchase distribution load the internet they don't want um like to people to be able to look at how much they're they're being paid the salary is getting stream through through through the theory of smart contracts um i think it's going to become more and more relevant the more uh the more value it accumulates on chain as you say that actually it makes me realize that like currently a lot of

ways that people still capture privacy or keep their privacy is to sort of interact with that web too to go back and forth but even like folks who send tokens to an exchange in order to kind of like centralizingly mix it even though it's not like actually private there is somebody that knows it most outside viewers won't be able to track it as well and that interaction continues on that level of privacy and i guess what you're saying here is like without those tools built into the web3 stack then you'll always kind of need that other part yes and it will always give um uh web 2 a unique value proposition compared to web 3 which is that

uh you can get um like you know as long as you trust your centralized service provider you can get a very good strong privacy um whereas with current criticals on its own that's not currently the case just add when people when folks started adding privacy into defy or even thinking about it what were those first problems that people encountered i don't know if tyrone if you could say something or maybe zach if you have an idea yeah um like i guess the first i mean i guess like the history this all goes back to z cash in 2015 when they uh i think they were the first team to really um convert um these zero knowledge proving systems

like zika snacks into something more than an academic tie-in that she turned into a real um product and on a blockchain um and certainly uh yes it's been very much a personal time to reinvent the wheel um uh but um uh to ensure everything is privacy preserving um certainly even even basic things like how do you how do you do value transfers you know like most like or like how do you like just do the most basic thing that um uh people take for granted these days um things like um i guess the main the main problems to be to to that are have been so statute to be solved um uh um uh how do you well maybe maybe i can talk a bit more about

i guess what we're trying to solve because this is the ones i've experienced with which is how do you how do you what the the main problem with zero streaming systems and privacy based around them is that these the pro the mathematical proofs that you need um to construct these groups are extraordinarily slim um to the point where like if you have a regular computer program and you want to turn it into a zeros proof running that prover will probably um uh be slower than running the actual the original program by about at least a factor of a million um and so that turns you know your your quad core two gigahertz um state-of-the-art uh laptop into you know like the equivalent of a you know home pc from the from 1980 with four kilobytes of ram uh and so uh to give a scale of the problem of

actually trying to like do meaningful number crunching in these trimming systems um and so that's where a lot of the a lot of the um the focus and attention it's been over the last few years on making these faster funding you have mathematical instructions that um uh mean the amount of work ethic to make one of these produces has been reduced um uh and then on top of that um you also have uh i think has been hinted here as this the incredible user unfriendliness of programming zero knowledge circuits um right now we're still in the stage where you really need to know what you're doing um uh that you need you need to have quite like a relatively strong background in cryptography so that you your bulletin which is secure um and uh whilst like the they're all the quality of tooling has made it's commonly used and balanced

lately particularly things like circum um uh and cksync um it's still it's still very um like you're still an extremely primitive landscape compared to um you know traditional programming languages so i guess that's the next frontier that kind of needs to be conquered before um we can start to get any kind of evil massive option amongst developers this actually makes me think of i mean one thing we did we maybe could do now we could have maybe even done earlier in the session which is somebody asked this throughout some of the talks which is like what exactly is defy and what is privacy in this case because defy as you've just described it if you talk about zcash defies transactions

of any kind right but when i think of d5 at least i tend to think of these like more sophisticated like paradigms or something like that and yeah that's i wonder do you make that distinction or do you see people i do i do it's just that um uh it's very hard to talk about private defy uh like fully privately like cryptographically private defend because it doesn't really exist right now like you can you can anonymize your cryptocurrency as a public defender but in terms of like an actual like uh using cryptography to enact a traditional d5 protocol traditionally you've only been around a few years but you know what i mean um it's uh i don't think um here's so many disagreements i don't think anyone's done that yet yeah in fact i would go maybe go a

little further to say that i've seen lots of um dubious claims of achieving it uh that have very little uh basis in fact or reality that they they kind of provide both the thing so i think we're kind of in the 2017 phase of private devi where unfortunately there's a lot of hype and not many real proofs or things that actually work [Music] but i i think the the simplest thing you know to your point your question of like hey you know i'm a centralized exchange user centralized exchange is a mixer for me

um one thing that i think you can't really guarantee is that the kind of like uh those kind of products that you get from centralized entities are actually going to evolve um as quickly as they do uh in in d5 so for instance like the kind of interest bearing tokens um you know obviously ava and compound are kind of the the first ones but you know now there's you know like urine tokens and stuff like that uh i don't think and maybe this is a bit conjectural but i do not think any centralized exchange provider would have ever offered you that

why because they take a much larger portion of the vig by offering you uh funding rate based products where they can kind of like capture both sides of that and like you as the end user don't get as much yield um and you wouldn't actually have this innovation into the kind of like weird structured product development that you had d5 that was like almost orthogonal to what happened in c5 until it grew big enough that they started copying it um and i think that that type of like in innovation cycle is extremely extremely unlikely um you know in the centralized world because centralized entities don't need to really offer you great structured products because if they did then they would be taking the

other side of that and taking a very large fee that's just like anyone can anyone who's ever kind of built these things or traded these things would tell you that like there's no kind of fairness in those products usually there's just a buyer who needs it really desperately for some reason like a miner who has a ton of weird exposure uh and the other side is like you know a capital provider who doesn't have to really tell you how they're computing like their interest rate or you know why they're charging a certain amount uh and i think the key thing about d5 is it has this kind of transparent flexibility in the pricing mechanisms and the revelation mechanisms to all users so all users sort of know uh how their pricing is computed

um and they also have a very clear revelation mechanism of like how liquidations or or kind of like slashing type events occur um and that kind of public observability of those types of events just doesn't happen in centralized exchanges servers got ddosed you you we we we we liquidate your positions instead of going into the insurance fund right like how many times have you heard that in this industry probably a million right over the last 10 years um and so i think d5 kind of goes a very different route by having like calculations and pricing and like public observability but this is also why private privacy is very hard like the reason it's kind of novel

is because it provides this public transparent observability of exactly how things are computed um but on the other hand like making that private would also somehow like reintroduce some of the problems that you have with the centralized vendors so kind of figuring out how to partition the space of what the public portion and the private portion is i think that's the problem to solve before you can actually get the like you know the app developers and the zk people kind of in the same plane if that makes sense at least isn't that actually like one of the promises of zk though like exactly what you described the provability i mean i've heard that example emre who spoke earlier from mina that's one of their examples that they

highlight often is like provability of underlying code using zero knowledge proofs knowing that like it was implemented correctly or there is something in it that is correct i know it's maybe a bit far off still but but that was i know one proposal yeah so i mean i guess the there's still this fact of the matter that the mechanism by which people are trading or ensuring that like a system is economically stable needs to kind of commute in in the mathematical sense with the the zk operations and right now that doesn't really happen because like the a lot of the economic safety guarantees of d5 protocols rely on public revelation mechanisms um so i could make a zk proof that

proves to you that i computed the correct interest rate without leaking the interest rate but then i noticed that there's a transaction that's large uh and that changed the interest rate and so i know i can kind of try to impute statistically like side channel type of things so like until you actually have like a mempool also and like a transaction processing model that hides timing data i i actually think it's a little bit we you're we're still not quite there yet which is why this gets back to the question of how do you partition what you want public and not and that's actually still i think the thing that's necessary to make the economic mechanisms kind of have zero commutator with the zk type of things

yeah yeah just to follow on from that i think it's it's having having um provable computations is slightly different from having pro like data privacy um certainly using using snarks um to kind of reduce the computational burden on blockchains is extremely valuable and something which i think we're going to see a lot more of in the future um because you can you know you can have a very complicated defect protocol with very difficult and expensive expensive calculations to perform where which isn't feasible on ethereum today because it you know costs like say a million gas in the smart contract instead of that you could just send a zero knowledge proof to the blockchain going yeah look here's like is the interest rate here's the here all the metrics and i've proved i calculated them correctly um but as taran said um it's you you um for most pretty much every default

protocol to work you need quite a lot of public information to make those computations and if you add privacy then that information goes away so it doesn't work um although i do think i do think the solution that well um just uh this is just my my my points on how the future is going to look is that the way that we're going to achieve privacy developers is by the separation between public private state in my opinion will be um uh identities of users will be hidden and um uh in order if you want private private transactions you need uniformity of transactions so if every transaction looks the same um and you don't know anything about the identity of the user then you have very strong privacy but you can still perform the calculations you need

um to run d5 protocol because you have all the basic you know you you know the transaction volumes you'll know you know how um what's being put into and how to take be put into and take it out of the protocol um uh yeah it might be that's that's probably how things are going to look if um uh if we're going to achieve private defender in the midterm yeah i mean one of the one of the other kind of building blocks and i think you kind of touched on that there is this idea of like i mean could could you make private oracles like aren't oracles kind of like trusted because they're not private like you have to be able to somehow trace that or is that or would that also give away things if you

if you like if you have a public oracle but a private system like does it yeah i don't know anything so one thing i will say is again a lot of this boils down to like separating the time dimension from when privacy is realized so like the time dimension of the transaction being processed and propagated through the system versus like hey there's a proof that it was computed correct and what time that arrives because the timing differences actually leak a lot of information and so you kind with the with an oracle you kind of want something where you can do a vk proof that the correct answer went into the calculation like oracle update comes in imagine it's almost a homomorphic encryption type of thing so

stronger than ck proof so i i take in the oracle update encrypted i'm able to operate on the oracle update compute the compute whether uh ave positions are liquidatable and then after a certain number of blocks public publish the public oracle update and the proof and kind of show that hey the calculation of the liquidation liquidability was done a block ago or two blocks ago and it's correct and here's the data we use you can verify so you have to make sure these timing things are actually hidden because that's that's not a trivial thing either that's like where most of the side channel attacks for these such things come from hmm there was a few questions oh go ahead

out yeah i definitely like i was just stuck on thinking of the the point that the mentioned like where where kind of like you draw the different lines of which what what part of like defy or let's say uh data should be public and what should be uh private and and even more like who decides that because the way we have now d5 what what makes it very interesting is that i could just go into these protocols and actually i i could see like what's the liquidity there what's the level of uh risk and and and quantify whether whether actually uh i should that should be depositing there

is it is is the free support the reward and that's like kind of like a uh interesting part of the whole uh d5 and and then if you think about it like on to what extent you should uh uh you should have the privacy and who decides that that's like a uh that's a very difficult uh part of it i guess like you could always go the the road that you let the governance decide and and so forth you know where uh where you should where you should add those kind of like a privacy features but it's it's not that easy yet but i definitely would like to see more

uh initiatives in in december's finance that you know we we try to add some some privacy in certain functions and i i'm kind of like i'm even like somehow uh surprised that there isn't that many you know tornado cash like things going on that that's applied but also like i think what's the issue is that you don't need to solve things on application level necessarily you know you you if you can have some sort of privacy with let's say a a mixer and you can operate like globally in d5 uh with with the just practically going through

a mix it's kind of clunky solution at the moment but still then you don't need to like build privacy specifically on various of uh uh protocols and that that could be one thing but it's still like very very interesting to see like how things will evolve there was a question actually in the chat here um which is directed to aztec and zach and that is would aztec layer sort of the zkzk roll up would that support interest bearing tokens or does it currently only support trans like direct transactions so currently it only supports transactions to be honest it only supports right now ethereum and

die uh but that's mostly for um legal reasons um uh which we're getting legal advice on whether on on the tokens that we can support um as for supporting interest benefits in the future absolutely um it's quite hard to do uh because generally you want to if you were in the private setting you want to do things like obscure the age of um a transaction you kind of need that to calculate interest rate um the the way that we're going to do it in the short term is effectively you have a public interest collection of interest bearing tokens that everyone can see but um there that position is owned by a large number of people and you don't know the relative holdings um and you don't know the identity and you don't know the identities of the

people holding it either interesting so you see the pool see the like the participants yeah uh exactly so like the idea is imagine you want to swap some ethereum for like like in just maintaining like rv um so yeah it is you deposit a theorem into uh uh roll up um and then um you would aggregate say a hundred transactions like um of these deposit transactions like hundred like of these d5 transactions together inside the roll-up so um observers can kind of see this pool of pool of ethereum being filled but they don't know who's filling it um and then once that call is filled with let's say a hundred individuals you then send that transaction through to ave the um the other network and then get a big block of rv tokens back in return and then a smart contract holds

that as a custodian um for the roll-up um and so um uh yet no one knows who who um holds positions in this [Music] uh your music timing delay of source would it would you wait for that to fill up for that process to start or would it be like whoever's there at a certain time you'd have to wait for it to fill up so so this is this is the process which will um become much the the music principle will improve significantly with greater liquidity and use but yeah the way we constructed it you'd have to wait for it to fill up okay yeah one one one thing i think that happens in the normal world that i think people and

if i don't don't take advantage of it and partially because of ethereum kind of be issues but also partially just like this might be expensive to do is that like when you do a secondary offering for like a large set of fixed-rate instruments you usually kind of provide some type of like priority uh interest waiting or like a swap where basically people only settle the swap at the end of like an auction closing and the earliest uh participants in the swap kind of get like a slight booster um and so like there's a lot of economic designs you can do

to try to like incentivize filling these types of pools that you know i think if you draw from traditional finance you could actually um make it seem close enough to the end user who kind of doesn't really care that much about one second interest rates but they really care about like two hour interest rates or something cool yeah there was actually a question directed towards you as well about um kind of going back to those earlier points we were talking about from francois should developers embrace a mission of educating the public on the consequences of the lack of privacy in the financial world

um you know i i think like in some utopian ideal world sure um uh and from a more practical standpoint i don't think developers end up being the people who communicate messages to the masses uh writ large uh i do think maybe they they kind of should communicate figure out how to communicate but i think it's gonna take more than just kind of developers communicating the role of privacy it's going to take like many different people uh in these systems kind of involved uh uh yeah so i think that's kind of uh that's it i i i think like at the end of the

day you know the robin hood stuff kind of shows people what happens when you don't have privacy in these financial systems um pretty starkly right like people people didn't even realize it um and i think uh to francoise second part of this question i would just point them to that and just be like look you have a live example of what happens but do you also think do you think if more privacy was already built into this would that also encourage people from traditional finest actually a barrier to them entering the space now or they don't care or they don't know i i think it's a barrier but stani

probably knows the best because you know he's he's talking to to to people trying to to use ave who for big capital prize but i'll say like at least from the more like hft hedge fund side i think one of the reasons kind of the traditional players have lagged behind kind of uh native players like alameda type of people uh it's because they're too afraid of this privacy thing uh it's a hundred percent like you know you go talk to some people that they'll definitely be like ah like everyone can just see my holdings yeah uh and like it's like an old money versus new money thing and by new money i mean like last three years

i agree i agree like that's the like especially on the institutional side like that's a uh that's a that's a concrete issue you know once you kind of go all the uh the risk involved let's say like whatever there is the the risk of deposits the underlying let's say stable points and and then like the the kind of like the last passion is the privacy there and and that's like that's a bit uh challenge like i i think like the natives uh native uh more institutionalized uh crypto market participants uh for them of course it's not not that of a pickup deal but when you are explaining to to a new audience that's

that becomes a bit of a challenging but uh one thing is interesting is that uh you know it we have a lot of like bespoke things here like the whole device is kind of like the spoke you know the transparency and everything like that can be explained but you know the [Music] there's definitely like a big issue if if if uh you know if it's not just actually like the for this marketplace it's not actually just the fact that uh you know you will see like how they transact just for the sake of transactions you know and the announcement and so forth but actually like there will be like the

native participants who will see okay they're transacting this way so let me just grab whatever they're trying to do and front-run that market participant and being in this kind of like an ecosystem for them is kind of like uh a bit of challenging and it's uh i i think that they did the front running is a it's a bigger issue like kind of in a commercial environment i actually this this reminds me a little bit about how like early on in social media where it's like younger people sharing all their information and older people being like what are you doing that's terrible don't do that they didn't care and now the adults have joined facebook and everything sucks but anyway just a

thought uh i i want to follow up with a question that andre asked here in a private d5 setting shouldn't it be possible to reverse the paradigm i think he's referring to what you spoke about zach or you'd keep everything private by default and then periodically publish important information like liquidity or interest rates without tying it to individual transactions yeah that's the goal um it's quite hard to do though um because you for example if you want to keep everything completely private um but say every say a week publish an interest rate how you how do you compute that interest rate because if everybody's positions holdings are private encrypted they need to share that information with somebody that can then um aggregate that data to compute an interest rate

um which obviously leaks a lot of information um you kind of it it's not very decentralized because you have to kind of trust someone with your data um to aggregate it together um so that's why it's kind of it's quite hard to keep everything private by default i think like conditionally release aggregated information um yeah so um uh there are ways around that using multi-party computations um another approach is just being very um uh kind of just pretty defining very clearly what information um every like um every transaction kind of keeps something some information private

and some public um is it is at least an easy way of doing things like if you if you keep every identity private um and you make sure that um every transaction is relatively uniform um like within a within a like uh you're in an empty set it's like every transaction is like say 100 die um if you're doing some some diet transactions um where if you want if you want to send more than 100 you just send you you send multiple transactions but they all look like they're covering from another side entities so um so nobody can make them um it's it's a little bit clunky but it's um so far i think it's like the only the only way i know if that's practical without getting into some really advanced um uh cryptographic multi-party npc

protocols that aren't uh will be very hard to implement on ethereum yeah and i would i would add one thing which is there's kind of some like uh you know i in my head i conceptualize it as kind of like a kind of lower bound type of thing like a heisenberg type of thing uh where kind of there's also a risk versus privacy trade-off so the more private your transactions the harder it is to actually do kind of large-scale liquidations so under like something that happened this morning right where kraken had this flash crash where ethereum went down 60 percent uh there you know there's a ton of liquidations caused by that

uh the problem is if you're forcing people to actually have to spread out their transactions over time and in smaller chunks you actually increase the chance that the system doesn't actually stay solvent and so there's there's like the very fine line you have to walk here about how much you keep private and how much you don't keep private because overall risk that the smart contract holds at any given time um at least for leverage systems yeah yeah completely um they're turned off because yeah like with at the very least if you want privacy you're going to have to pay with latency um to like aggregate transactions together or wait a certain time the amount of time so people can't link a deposit in withdrawal and yeah that doesn't uh yeah that that

for specific yeah for leverage systems or um trading environments where like you know if things are high frequency you can't wait 10 seconds um for a transaction to settle that that is a problem follow up to that uh oh andre was asking do you want to mention some of the names of those vaporware privacy platforms i i'll just say i'll just say do your own research but anyone who's saying they're offering private unisof balance or curve very likely isn't as guillermo's talk has illustrated so and there's

there'll be a paper kind of doing that type of stuff um so anyway yeah but maybe we'll still be nice not call out anybody here um i i want to say first of all so there is one last question from raul which is the best resources to learn about various zkp algorithms but that i would say probably wouldn't necessarily need to be here i mean this is get into the podcast follow the newsletter we do we definitely have lots of resources to help you out there zach i feel like aztec probably has a platform full of zero knowledge proof documentation or maybe you know some places to send folks you know i must confess that that one of

the problems with this space is basically everybody who's involved in zero illness proofs um at this level is very busy kind of building a technology hour is not spending a lot of time writing documentation unfortunately um i mean it depends on how how how in depth he wants to go like one yeah obviously um yeah like the zero podcast like they're not like absolutely it's about in my opinion it's the best introduction um if you want to get into a deep dive i mean dan bernard just recently um he's a cryptocurrency staff that he's recently published a very comprehensive uh kind of um guide introduction to serialized proofs in general um uh i believe you can find it online i think it's uh i think it's um uh just a pdf that's released for free

um it's an extremely good text it's quite hard to get into and it doesn't cover zika snacks yet because they're a bit too modern um but it certainly is it's a great theoretical background that's so large level yeah and as alan mentioned there there is a zkp online course in the works but that's not out yet and it's still a little ptbd because of it got very delayed by last year's kerfuffle but there is something that is in the works but in general i would say like get involved in the community you can also ask in the telegram chat there are i know there are platforms and group like uh repos where they've just added some documents so that's that's some ideas um and actually i think that's a really nice point though

to to wrap up our day uh i want to say a big unless you had something you wanted to i want i want to give one person a shout out if i saw an audience which is arielle who wrote some blogs on the zcash website in 2016 or 15 which like eventually led to me quitting my job so thanks for writing this very cool sorry i should have given you a shout out i'm sorry yes those are those those are those are really good blog posts for like the basic vk like you want like you wanna if you like know enough math but you're like i don't want to like read this code it's like it's too complicated but you kind of want like a high level

really good you know uh pedantic but not in the weeds introduction i think that it's like one of the best things i've read as like a first i'd know nothing about zero knowledge proofs other than like bitcoin maximalist calling everything moon math without knowing why yeah james asked if that was ariel gabby's on yes they were talking about arielle who is who is zach's colleague at aztec cool i don't know if you wanted to say something zach but i'm gonna keep going um okay so yeah the next thing we're gonna do if you're into it we actually are going

to head over to gather.town and hang out a little bit if you're still interested in maybe catching up with any of the folks here i know i'm encouraging zach and stani and tyrone to head over there as well so maybe you can keep that conversation going for a little while i think we have the room for about like an hour or so and uh it should be fine for everyone to come if there's any issues let me know it's actually the same as the zk sum it's the same room i just kind of repurposed it there's different banners with zk sessions now uh the password and link john just put into the chat here and you also should have received an email like about 20 minutes ago uh with those details as well so i hope to see you there

i'm actually gonna sign off i will be getting these videos out to you uh over the next little while i have a plan for guillermo's um and yeah thanks everyone for this you know coming to the first dk sessions and bearing with us through all of this but i do think like these these challenges tend to make the crew a little tighter so i'm really glad you all came along cool see you later then