ZK Study Club - ZEXE reading pt 3 (of 3)

Transcript

cool so this is the third installment of the sexy sessions for the Study Club we had the first two which were very in depth or Youssef went through all of his notes and shared a lot with us about the construction and today's session is supposed to be more of like not wrap-up but like we're gonna look at applications of it and we're gonna try to pool any up like open questions that we had around the paper yeah so welcome everybody I think it'll be a little bit different so we're not gonna do quite as like structured but I guess the first question out is just sort of like does

anyone actually maybe I can ask does anyone have is anyone open to sort of describing the decks example this is XE indexes and how that's special could have a go I just yeah cool but we're gonna start with that as opposed to the simple er c20 example I don't know something I mean if you want to the LC 20 example me just come each needs to come first okay so let's do that one so the basic idea is we want to restrict the the birth credit so records transactions

have to follow consume some records to produce others have to follow the death the predicate says I think they consume the birth credits or the things they create so if we want to create a token then we want the you know the birth predicate to say that this many tokens were destroyed and then this then it's moving tokens are conserved but of course that's a problem because you need to create product in the first place so the ERC 20 idea is we have this two things possible in the birth predicate there you can either mint or conserve and mint creates a new er c20 and it's sort

of become tsumtsum record and not quite you know how that works or something which gives which basically means that every time we run mint with this with the CRC 20 thing we generate a different token you know full supply of a different coat I've okay but they're gonna make it are they gonna make it invisible whether whether the transaction is a mint or not well I think we're gonna get to that in a second I think what you're describing here is look basic right it's gonna be it should be invisible yeah I can't tell what so if you if you conserves if you if you consume some other record which doesn't actually do anything except to

provide the randomness the seed for the the idea that you token and you can't tell yeah you could burn yeah you can you could do a burn of something Unchained and then refer to that yeah so you just doing some record and we don't care what's in it we only look at the the serial number which is going to be unique then you know the whatever the nonce is that's gonna be unique and then we use that to mint a press supply of an entirely new er C 20 token and once you've got so the one person who did the mint has aeneas has all the ASC 20 token and then they can pass it to other people but now in order to create this the same EOC 20 token the with this this

number you've got to consume exactly the same number so you can only create something if you're you can only hand it out if it exists this way you were to set you before you just start you sort of started describing two things as the birth yeah predicated as mint or conserves because that sort of is I mean this these seem very similar to like the mint and burn or the lock and unlock of like bridges and I'm wondering if this is related not really says they won't one thing here is that men only happens once

yeah and then we can come then we can conserve with the same with the same token the same token ID okay you say oh so you're saying it's mint conserve burn so and then there's like a light isn't many times but every time we mean we create another different type of token yeah and then thereafter once we created this token we could only conserve it oh I see so they're actually hiding when you create a new token yeah oh okay that's also hidden well everything's hidden this is this is yeah we saw I was I was imagining that you could do since we're saying we're gonna do the ERC 20 you know somewhere these first I thought that you were able to do birth

predicates in either mode mints or conserve but you were able to do both so you can have you know a growing you know token count over time if you want it I didn't really what you can't do that so the key thing is is that every time you do mint it bastes the ayah the idea the you know the idea which determines what the token is on the the record you consume all of which on the serial numbers of the records that are consumed in the same transaction and what that means is that every time you run mint you generate a different type ID XR yeah and that's part of the that's part of that record too poor okay yeah I see that but I presumed that the the I haven't looked how the amount works I

presume the amount is fixed by the the sort of contract mm-hmm by the the predicate itself so that people can verify that their work what the supply was right all right which which actually I just really enjoyed this idea of using a predicate on you know on the overall token because I don't see that at least in any of the smart contract ii stuff I've done in in etherium you know there's not least calling calling in a predicate for me is this familiar so I really like I really like that familiarity you know instead of saying oh yeah we have the records now colonel and said just talking about yeah

Anna described it you have a life cycle it yes I mean the interesting thing here so it's kind of like we definitely have small contracts at all we have small tokens because this is somehow this is how they do smoke spot execution with with the EXL model and but you can do other things with easier with these mental conservator see twenties like this next thing so the moment so the LC 20 looked to see which things consumed but it's sort of this looks at the the birth predicates of the things that

consumed and that means that the death predicate can be anything and I think this is what this this this decks idea does its we keep this may be this mental conserve creation birth predicate but we can use the DES predicate to of records to allow us to trade one token for the other for another yeah it's regularly right that you can have this token if you give it if you can create a token with a creative record in which you own

these tokens which which I have which you can conserve as long as you give me some other tokens right so I'd imagine that in if there is you know some kind of you I hooked up to this you know in the clear both parties would agree on the death predicates of the I killed the tokens that are being swapped well okay so that the party that wants to do to do this one thing they have to create this this death benefit and so in theory they could publish this and anyone could redeem it I don't know if that happens I have to actually look at what's all the details this next thing is yeah but

but the I think the proposal here was that it's possibly it has to be the record has to know who's swapping with whom in which case we somehow the two traders need to talk right I guess this was getting into the difference of this intent-based Dex versus and I can order book based Dex so whether the publishing was happening ahead or not like on an order book it would be listed exactly what you know what token and and what volume for what trading pair versus an intent based you know they need to figure that out out of band and probably it would be the taker ascending sending something to the maker but it's good see there it seems unclear

but in the order book based yeah the maker would have needed to you know kind of craft the outline of how it's working if I remember right and say that's it they say to specify that their properties they want and then near the end sort of on page 33 we have this this open book decks where the maker just publishes like what they want an information to contact them and then you take your plies more form of communication there and try and get transaction through X where it goes through a trust of their party everyone reveals everything to you and these have different privacy these two privacy

guarantee world one who gets what and one gets the other so with an open book depth we have okay what's trade what are the the things they claim they have here [Music] limiting so knows what the what the party is apartment apartment themselves we can get this if we can like anonymously publish that we want to do this but that doesn't mean that people can see what we've what we said we wanted or you can keep it secret apart from the maker what you said you wanted

but then the maker gets to know the operator gets to know who you are yeah yeah can't consequences and trade-offs as per usual what did you what were your thoughts on this guy on Dex's on then on this on this sort of construction that they had put together here I think I think it's not silly fascinating you know it's kind of it's so what we've been doing and ask is kind of coming at this from a slightly different angle but this because of was like we're constrained by massively limited computational theorem and less

restrictive to curse and one of the things that was being done frustrating us is while speaking on see like the confidentiality of values that are being transferred wrath and the other stuff is critical to present like to correct for amenity meals and genuinely no transaction block is quite appealing and the idea also hiding the offending the aliens that have been run through some of the legitimacy of a transaction is particularly like core is pretty cool yet I don't know if anyone like so Alistair has been on previous sessions just here wolf but actually I don't know if everyone knows what you what you work on so to say that quickly yeah sure so

I'm here with so Aztec is a privacy solution for different is your bottom so we use your own photography to enable the construction of privates tokens where the values being transferred are encrypted and like copy copy see um come even looked at coming know about looking at it but at the transaction logs he after interviewing fees of the notes and so we're in a anything is kind of the creation of private unless there's some more customizable through natural logic transaction semantics and language were deployed of the test nets we're planning a

production water p3 and yeah we're kind of actively researching and like manure on your on their knowledge podcast recently and I what I learned was that you're not using snarks they're using a difference your knowledge proof construction and so I wondered like given that like could you do a set very similar thing to this well I mean one of the things which is quite interesting is we can use excess justice we have a little bit more agency reveal it because we use because the beginner curve wouldn't need to be wouldn't need to have a look at I to joyously so so yeah that'd be some of the stuff is kind of

extremely relevant to what we're doing in aspect because what we're not a stop we absolutely can to bet as if it's not technologies and so for example something which were actively exploring is the idea of validating a batch of Aztec priests instead of stock circuit because divided a NASA crew requires a small number to cascade applications and one by the apparent check but that will anyway check see if you wrap their own that's not second you can perform another Nick Patrick external to the smart ii basically been basically music you can put you can bet that rico validates a batch of private transactions inside a sock circuit with

relatively low computational livets compared to other alternative ways of doing that like bypassing criminals and yeah so that's a kind of little what I'm glad you shared that with us cuz I just I know you but I don't know if everybody else was aware of like what you're working on Alistair sorry I didn't mean to cut you off like that though did you want to continue okay well yeah so the detail here with this this exchanges so what's the death but it could look like we have this exchange or cancel predicate so that there's two things we can do someone can that you're the person you're trying to make it populist of this and Undertaker can use the exchange

mode of the desperate to give themselves the tokens are gonna be traded and give the they may cure be attendance they wanted and the exchange predicate checks that this happens all the make it can cancel by getting the tokens back in this case in case that's been broadcasted I mean canceling really means that whatever took an amount is really just sent to a new I guess how are they calling it here like it's going it's going to a new public key like it's

going to a new address at the end thing right yeah but the maker has the right that they can use their something signed with with their key to get it take David's awake which is not what have exchange works what is this burning is this the same as burning no neither exchange no counsel burned anything no it's simply instead of I guess instead of being able to so you need to broadcast our that you want to make it true grade and oh and you also need to consider the total and that maybe EDX rate has changed up or down doesn't matter and you as the person who created

the initial offer as the maker you say yeah I no longer want to do it what what I published so you need to conserve tokens but you need longer valid but there's been there's been publicly good so you have to move it to another address you you own well I doesn't need to be public they makin take it to agree on this but then if it take it doesn't so you do the exchange is to make us to be able to get the tokens back but I mean it when all this has been done when all of its been done out of bands you wouldn't you wouldn't need to cancel but in this idea that there's somehow the

queer book you know you still need to cancel radicular like they make it doesn't want to trust to take you exactly their identity is tokens you know you want to either we get the tote either we get the thing we wanted to trade for or we get our original tokens back that's what the work that the exchange will cancel thing guarantees right it's not whatever it's that they take it doesn't do it doesn't do that oh yeah so this is all in the order based decks I guess I'm also I'm kind of crossing into intent-based X which is maybe where I'm adding confusion because then intent-based X was simply as you know your you list you know the pairings

that you're interested in but there hasn't been you know any transactions partially created that gets shared great because it's supposed to be done out of them yeah yeah yeah so might like that you can stay in the order base necks and yeah I'm not sure if it's a limitation here so this exchange or council thing appears to have the the takers can be involved only the take you could do the exchange thing I'm not sure if you actually needed that or if you could do an exchange you could offer a contract which anyone could redeem if you just told them what you know the details I mean maybe it does need to be with the

takers so the makers crafting something for the tape or is that so that the taker can also you know it's like maybe maybe it's so there's a little bit of privacy and transferring at you know Harsha at Trent and then yeah so I think the series is that they need to talk anyway so even if we could do this but generically we have the public key the the taker would still need the makers public key and the make it wasn't want to publish that right along you know there attend to trade although yeah so they yeah maybe we have a thing where the makers were all public and the takers were right but then the previous trick that we picked up isn't it's know

when we were delegating transaction signing we're not able to use this this technique again where we were you know not showing you know the true public key we were like we were using this randomized double signature scheme we were not able to do that here that's that's a good question then because what you say is also when I was thinking that there was a you know maker maker makes the first moves but then yeah yeah I would imagine that the Senate has it work again so can we sort of like make the taker like this user it doesn't know this is the worker thing yes yeah well

so we picked up this trick from before we had this work where we were so we're delegating the transaction signing which is where the most CPU intensive work needs to be done and we were using yeah this randomized untrusted worker yeah yeah yeah and so in a way we're in a similar situation here just like Alistair said that you know the maker doesn't necessarily want to expose his public key quite yet you want to expose his intent to trade so I'm wondering if we could do a similar thing although I wonder don't they kind of say here that like they like through their logic they're preferring not this intent based

one like you just said I mean I have a few can feel like they're just like here's the two options cuz the way I for some reason I got the sense that this is options and then they give a different two options afterwards further detail mm-hmm no no a debate it's like the end that is this closed an open book yeah they just don't correspond to the well they're both or little based Dex's right I know the beginning it's maybe maybe lesson said about intent based X because as I as I understand it most of the communication needs to have in some

because you're simply not like creating a partial transaction let's say ahead of time so as the maker you're not saying I have you know 20 of token a and I will trade it for five of token B and then you know partially creating that transaction for anybody to fulfill or even already knowing a taker that says yes I will make that intent-based is simply saying I have token a and I want to trade into token B and then I would have been communication yeah you know needs to occur yeah yeah Ted likes it it's very simple yeah but going back to your thing about the sort of untrusted

worker and this sort of Rand or it was at randomizer Bowl ticket just signatures what role like who what role is the untrusted worker here is this sort of like that centralized third party when you just described as like a third party who's also like paying attention to that is that the worker in this case okay we're kind of okay this could work if the delegated worker works at all done whether whether we can hide from the trader the public key yeah yeah there so the reason that the untrusted

worker came up again is because during during that description of setting up you know this delegated scheme we were able to say all right there's two there's two addresses that are being used now because it's an untrusted worker you don't want to have to share I can address at all with them that they could use later to link back to other activity on this like you know in this particular blockchain let's say so what you can do is you're creating you know like a one-time address that's stemming from you know a real key pair that you know owns that record that is gonna sign on that record so the question was could we do the same thing so it's what we use

for the delegated workers so presumably we might be able to use it here as well so that the maker and taker don't have this funny game where you know whoever moves first is exposing ya an address but I'm entirely unsure which is why I'm asking yeah well they didn't say so this might be a good question actually for the group that da okay was there is there anything else you wanted to say about about the Nexus Hollister no I think that's what I understand about it I have

some questions regarding the exes and also the years C 20 construction because in a in such a private setup I'm like with the sexy also like with aspect for instance from Zack I wonder how the price signaling mechanism should work actually if I don't know the supply strategy because the ERC contract might be private and I don't know much of the volatility of what's being traded on Texas with this ear see 20 how should I just have a clue about the pricing talkin the the first thing

is you should know the supply strategy so in particularly the the that you need in order to own this token you have this bird predicate and that will tell you what the mint mode of that will tell you what what tokens were created you know it gives the rules for when we create token so the idea of the mental conserve is at one point we created loaded of tokens and the actual contract itself will tell you how many tokens there were the first predicate that you use for every minute or concert but it's critical that you use in conserved mode for every future transaction has in its code for the mint part how many tokens

were created so what you're saying is the the minting strategy is actually public is this true yeah the minting strategy is entirely we create a million tokens so every time we run this mint or conserve contacts in mint mode we create a million tokens of a new token type and then that supply is fixed at a million well we can always all-knower they're ever after so you get my childhood and what about the demand and supply at the Texas because this is obviously completely private to the outside world so say with these two

types of decks and one of them we publish our intent to trade and that's you know that us with orders in in public exchanges that gives you an idea of what the demand is what the exchange what people think the exchange rate is yeah I think what you're saying is it is public what amount I'd like to buy and to run a price I'd like to buy for the for the right guess it'll give several examples of no find the section kept so they get several designs so the attempt

both decks you know the makers published their intention to trade so that it just like with the other decks that they they may publish it but of course they can do it seriously and in the sort of open book example it's more or less the same we published the basically the record okay oh yeah the only base Texas yeah we don't need to know they take his private key they take his public key in order to do it so both the open with Dex and the intent based X we it's actually public

what people want to trade now in the closed book Dex where we trust some operator then only the operator even knows what the orders are I see okay and then their conclusion would be that as X is four for Texas is actually not a way to to prevent front-running is this is this correct do you think right at least this and try to find but what they say I mean the point is that you don't know which currencies are being transacted

oh you he means because you've got to add it an advertiser sure yeah yeah I mean this is I think this is why the intent based Dex is interesting or or why it's still talked about at all is that it does protect against front-running since there's no public there's no public broadcasting done of what will be traded other than no pairings of tokens that that makers are interested in so there's no chance to front run because the transaction the transaction that hits the blockchain is performing some swath of tokens whatever yeah the templates is people that they

may could take a negotiator back what the prices say and then we never know what is sure yeah yeah please well that actually raises another question of like why we use this so like your your idea there that it's this could potentially prevent front-running is there other things that this set up I mean I guess it's private but I'm wondering like what are the other reasons why we've won a construction like this so there's a fine thing you could do if you I mean I'm not sure people are going to be very happy with Texas where they don't see a market price coming out of

it but if you add that then that leaves you very open to market manipulation I mean and this is this idea that it could prevent market manipulation now I think it would increase it if you add a price so the point is is that you know i yeah i'm saying this construction though because here you're saying it doesn't have that yeah here there's no market manipulation because it doesn't leak any of the information about the prices but if you added if you did some kind of I don't know strategy for trying to I mean you could conceivably have it leaked the exchange rates in some aggregated way over a bunch of transactions and I can almost see how to do that then then it would be fairly open to to market

manipulation and much worse than one that didn't have the privacy which is just from you but but ok but but is that true because I mean if we so let's forget about intent-based because mostly we're talking about an open book and closed book we're not losing we're not losing the ability for aggregators and the open book decks right because the the information that is revealed is assets and the amounts that were involved I could I'm just reading directly from the paper on my own page 33 at this point so if you wanted to have an overview about you know what is being traded for what and at what volume so you're able to see you know exchange rates moving over time you don't lose

that however what you do get is trade anonymity great so and then then a closed book decks it's the same world as any other centralized decks right now where you're trusting that the rates advertised are correct and they're gonna be exposed by whoever's running that closed book so we're going back to front running you know on an open on an open book decks I don't know I mean yeah I guess because that the assets are being revealed that somebody who is you know a minor in this blockchain they could still a friend run and the closed book decks the only person who could front run is you know an inside operator so it sounds we from really you know

from a distance it sounds almost like the exact same thing that you get now except in the open book decks case you have the anonymity between the maker and the taker nobody learns about their public address in other blockchains right now that are that would have something like this like you know kind of you get to create tokens on top of this blockchain usually addresses that are signing and sending transactions are owning or have approval of tokens and so there's a little bit of metadata leaking about the fact that this address this address held tokens and just did something with them just transferred them so I you know as an overview it seems like really it's the thing that

you're getting is exactly the same by during open book decks case you have an on the MIDI between the maker and the taker and that's that in my opinion is something good because yeah for example if you wanted to do this with the RC 20s and the theory on you you know you can't do that because you would need to have an address owning tokens send that in a public way unless you're adding you know unless you're doing some extra layering on top yeah yeah makes sense makes total sense I'm kind of I guess I'm kind of waiting to see if people who are doing these

things in aetherium will eventually get to that point because yeah I just you know I like it I definitely liked aetherium and I enjoy the amount of experimentation that's happening there but this is something that is quite tough because we're doing this account based addressable so you're using an address for a long period of time so not only are you showing all the token transferring that's how you know at the Dex level but also address for address you know maybe how long people were using that or you could do you know some where things like like with Bitcoin where you're doing transaction analysis and seeing which addresses are participating together over time yeah so I think being able to hide your address with would be useful

when it's just general members of the public using this you know you don't have to think about whether you're leaking all that time I hope my current roommate is sneezing behind me you come you myself total meet ourselves was next yeah what's next stable coins yeah I trying to work out what's going on here so what they were doing is so instead of this mint or conserve we have a mint to

enforce in which we said they have some centralized authority that can do enforce policies that do something with these coins but without so though the central centralized authority can do create these policies it and AB can enforce them it doesn't know who actually has the coins or other things about them well but it was saying something about having a white list or a black list somehow wasn't it yeah or simpler simply a black list but yeah I'd imagine you could be the opposite to guess this is this is for all the the

sto type type things where you need to have some some kind of approval I think that I'm curious but here with the stable coin is whether or not these are like what kind of stable coin do they have in mind here because there's these I mean there's the stable coins that are based on Krypto there's stable coins that are based on like what is a collateral so yes there's there are the algorithmic stable coins which I think are still pretty unproven and then there's those that are based on collateral some are based on like fee add collateral and some are based on crypto collateral I think what is all the way impossible

from what I can tell to make an algorithmic stable point in a zero a lot of private worlds because the other is you can't transfer Isaac get against an asset value and that if you're like if because the price operations your your onion are pasteurized certificate as well because estimate erased then you you need to thought be able to forcibly sell it off and if you don't know how much collateral we have or what it's worth or have a sell it because you don't have the secret keys to it then then that just seems a little bit unlikable I'm not sure that's the case in particular looking at this I can't see why we can't do a a day type cd' are you know silly playing a lot of scene

looking as long as we have this central authority which in this case will be just giving the exchange rate so if we somehow come trusted source for the exchange rate that sort of maybe gets published on Unchained somewhere in the plane and that these things could somehow read then we could do a CDP get some there are it appears even if you have the exchange rows so if you crazy if you could punish like the the collateralization ratio will CPM having public without leaking much information you wouldn't need to publish that at all you you you your your CDP would so we have these these these lock tokens yeah so we have we have this record you

know which which stores which has our locked collateral and stores how many tokens how many other stable coin there has a lock earth type thing in record the amount to die that we we took out now if the exchange rate is it is too low compared to these then we just then we can't get RF back or that's not absolutely well yeah the way CDP's actually work as they like automatically but I don't like collateralized once you hit a certain threshold no but you can force you to do that you can force the user to do that because you can you can put the minimum you can put the minimum

threshold achieve during a certain period of time into the miracle tree and then they won't be able to refer to anything in the past except to that that you know is too low or whatever I'm accepted to do this light right now I would just make it work when we we have this this thing the dopes so what we can't do here is is let someone else do something with your earth no you can't do that but you can angry you can force it to be locked yeah yeah the locking is not quite enough though because the way it stabilizes the price of die is that if you come in to collateralize your collateral is then you I mean this what going out this is a smart contract will sell off your commercial to buy died out of the out of the air and then do system

and you probably do that in parallel because your record you're all you're a lot deeper obviously that's like an algorithm there's no access to these in secret required to spend outs so there's enough if they have enough stake in the thing probably it just becomes expensive they probably just need to have more stake in the things so that they're not they're going to choose to sell off instead of yet instead of lose much more money by not selling off because they're then because yeah but you won't say you're relying on the under for people a lot of the clash of basketball things have you become outcasts relies then you actually don't have an incentive to sell off because you're done idea is worth more than your collateral that depends on I

as value will disappear no end of the right way you look it yeah okay either die play no no I mean in this type of zero knowledge thing yeah I'm saying if if to you you you stake much more than an then you then you then the actual value of the die or the whatever the 0 DK tie thing yeah then and then if and then if if they can't actually you if you don't do the don't don't do the unbond don't do the like sell-off thing when they when you would have to then you can't get any of it back oh you pump your collateral back but you pop it for my how much would die because

then I mean they like die send it transfer at that records and Australian and you want to practice you can't have a history of what point is here we don't we don't confiscate we don't make you a collateral you'd make your collateral unrecoverable you know we don't we don't look it we don't take away your money when Wed we're we're sort of 50% out of the exchange rate you require a much bigger collateralization it's a sort of force people so to the right thing to do is to is to redeem yourself or or increase your own collateral where rather than end up in a situation where you're saying is unspent able well right now they actually do 1.5 that's the job they can collateralize for the amount so

you could probably also do something so deal here which may be the peel don't want to do I'm sure you can do something like that yeah that gets who did this analogy I kind of wonder here if this is more like I mean it seems a bit like a thought exercise I don't know if it's a thought out I know also know I've noticed in their presentations on this paper they tend to focus on the Dex's I actually have never seen them go in-depth on the stable coin so I wonder if it's more just like the IDI initial thoughts about it so you know the main thing I'm actually interested in this for is the just the fact that they have the the pairing curve the curve on top of a curve thing that they this

construction that they already have in the paper because this lets you do this lets you do we want it for its it lets us you would let us do certain kinds of about availability stuff that we want and have the chain zero-knowledge so like you can you can aggregate and verify a bunch of snarks in another start and that other snark is actually you some proof of validity for a huge chunk of of things going on so this is so the the primary interest to me at least is that actually an implementation even if it's just academic quality of a you know a curve on top of a curve it's somewhat efficient yeah I haven't seen too many of the presentations on on sexy but I

guess still what I like is the this idea that you have these predicates and that if you you know Minter conserve will conserve it also be extended then into this idea of enforcement so if we're somehow you know wanting to do funding curves or or other other such constructions and somehow somehow if I could be entwined with an important policy for another record and which may be maybe it could because as I was understanding from the previous parts of this paper you can have transactions that are of a certain size but they're kind of leapfrogging each other so you're able to use the result of one transaction to continue going into

another and then and then so if you were to have somehow you know this you know locked amount of another record and somehow you could make that available to this stable coin or the other name that they gave it is what the centrally managed asset I think maybe maybe at least for me when I'm reading that it is the second name for me it's just you know the stable coin means that someone's actually responsible as opposed to there are tokens and they're just given out to the users of this blockchain instead there is you know an entity that is somehow responsible so they need to have an important policy in

place but maybe it could also lead to things like yeah maker like let's make your diet set up but again at least for me and what was really cool is that these predicates give the opportunity along with this record structure to maybe touch multiple things at once whether whether or not that's possible I'm not sure that would be that would be really fun to play with though to see if you could have enforcement touching multiple records not not just the ones they will come yeah I wonder if like a lot of these applications a lot of the language and concepts they're using existing by things at the moment but I wonder if like when you talk about token exchange

and you talk like what I what I'm constantly wondering about this is like could you then use it more for like transaction token exchange like where it's not it's not actually tokens it's not actually about exchanging value even we were basically using some of these structures for like more of the infrastructural stuff where you actually want just some sort of like movement to be private or that this is sort of what I started to envision with it I don't know maybe I'm completely off-base but that was just that was this idea that like I've noticed this a lot in general when when these concepts come up people will start to apply a lot of different techniques to them but in fact like what they're actually doing is like

a future version whatever the Dex is evolved into and maybe they're no longer just like exchanges per se but they're more like almost like interoperability place which I mean they kind of could be or are but then there's other things you could do with interoperability and I'm wondering if this would somehow play into it but I don't know I haven't thought that far yeah I mean I get that I get that vibe maybe not directly from sexily but this idea that yeah you have picture chain Atomics locks I mean that's that's definitely an active area of research maybe because a lot of people have been mining Bitcoin for a while and they're going all the way the a pseudo anonymity rate or for various

other reasons but I don't know I think I mean you will probably hear me say this a lot but privacy by default is a really good idea which is one of the reasons I really enjoyed this paper it's kind of making that the main the main point like all of these ideas here are how can you make it so that transactions can happen you know on a on the public piece of infrastructure with you know privacy preserving as one of the first ways and I I think that's a worthy names that he was you going cool does anyone else have any concluding thoughts because I think we're a time actually you guys like the paper yeah well I find it surprising that the title

is actually enabling decentralized private computation and the only applications they came up with this simple talk may exchange I find this a little small I have to say compared to what's possible with this with the scaling technology so for the application part I'm pretty disappointed what do the others think it's an interesting model right the question is that I'm sure that it allows some very interesting applications that people haven't thought of yet because still to get your head around and that's again a sort of speaks to what I said there were it's like I think they're using examples that people will understand because the

other examples don't really exist yet yeah anything is also a function of the fact that um what you can do is actually used to like the amount of computation you can perform as relatively limited because of pre-construction plans so like keeping the examples so quite simple understandable things they'll maybe kind of help set expectations the you can't you know you can't validate the other the machine learning algorithm in sexy not just yet she's you know I'm thinking so / - I think I'm saying that hopefully I'm saying that correctly he's in our group and he was a co-author on this and I was thinking it would be fun like in general I was like I was hoping to interview him for the podcast and see

if I could like dig into this kind of stuff and see with him like I'd like to ask him that question that's like is this you say computation exactly what you just said well like you said computation and your example really wasn't and is that because you're just trying to describe I mean maybe it's like outside of the scope even of this project to try to describe that if that application and this is just like yeah here's the one attachment here's the system here's this limitations here's an example that we can say now maybe there's some others in the future yeah I think pratyusha is on these econ 1 and we will grab and torture him yeah definitely you know beer until he talks cool alright

I'm gonna shut down our recording here you guys are welcome to stay on longer afterwards if you want but I think just for our audience thank you guys for paying attention this was the third part of our sexy study club sessions and I yeah I'm definitely curious to learn more about this and to do more of them so yeah I'm gonna stop now cool