ZK & The Future of Privacy

Transcript

Welcome to all of you to the uh third event of it and uh in this November we are focused on privacy like this November is the privacy season for us. So uh just quickly uh before the speaker what is at about is to create the most cyber punk and ethereum align like community that we have in the space. So make basically Ethereum communities like fun again building open source having conversation about uh important challenges and so on so forth and uh so yeah this is November season

so privacy um this is the third event and you can find uh the first one which I was talking about what means beum aligned and um the the event about uix comply about privacy on our YouTube channel youtube.comy and uh and also like this conversation will go to the YouTube channel and everything you can find like directly also there.

So this season uh we have uh today we have Lucid Samuel uh about ZK and the future of privacy strength and limit which basically it will be like a very interesting and also educational conversation about ZK which actually I'm excited to to see it so I I will try to be as fast as possible in this in this uh uh pre And uh next week still like on the November season we have um ion chains which is uh one of these guys that basically could find out uh who is the hacker of different uh uh

protocols and stuff which will talk about not just how privacy works but uh how it's very easy to find people if they don't have privacy. and you activity. We will do quickly like one onchain raffle as we did last time because like open source like from here it starts something that is like based randomness.

So uh the first onchain randomness which is like unpredictable and pretty much like secure and impossible to hack from validator and stuff like as much as possible especially layer one obviously in a stage zero layer two is pretty much impossible to prevent anything but in that case you can go in a trial with the team of the layer two. [snorts] So uh and as last time we did like the first this time we do the second one but with some more funny thing because today there will be 10 winner because we have the sweatshirt of privacy like last time and we have 10 of the t-shirt from the first event that we didn't give away and so let's find out how many numbers 23

right okay so let's do it so max number 23 10 winner and also obviously with based randomness if you ask 10 numbers it will give you 10 unique numbers is impossible that you get the same number twice. Yes. So the winner is the number 24 which is the guy who just arrived and sit without the sticker. Then we have the number four. So number Okay. Here.

>> Nice. So um and now we can start with it easy zero shoe. >> Um hello guys. Thank you for being here. Uh today we're going to be talking about one of the most fundamental um elements of crypto or primitive of crypto which is privacy and uh we'll specifically be talking about zero knowledge proofs which is zk and uh yeah um a little bit about myself I'm u I'm mostly a software engineer I just started doing cryptography and uh I'm with founder of cage or kadic protocol we basically help

um institutions in invest and trade our crypto assets privately and um I also work with project AON at Zcash. Are you guys familiar with Zcash? >> Yeah, great. So um I also help with um scaling Zcash. Uh we're working on something called Project Aken and it's meant to improve scalability of uh the Zcash architecture. And uh prior to that I've also worked at previous uh protocols that are focused on uh ZK and implementing uh programmable ZK applications. Uh so yeah so today we're going to be talking about ZK and the future of privacy and uh I'll try as much as possible to

um reduce the complexities so we can all like you know get along easily. Uh yeah. So before I go very deep into um the complex part of explaining ZK um Eric Ugs is [snorts] is one of the founders of Cyber Punk and he made a statement about 30 years ago. He said um privacy is necessary for an open city in the electronic age and this was way before blockchains existed. This was when things were mostly text, emails and uh you know web 1.0. Uh yeah so and I think one of the most important thing or important word in a statement is he used the word necessary not optional uh

because um privacy is basically our sovereignty. It's um very fundamental to who we are as human beings and um Ethereum basically builds one of the most you know fundamental applications that we use today. I mean in terms of blockchains and um you know that because it's a public ledger this basically led to what I call the transparency trap and um yeah so we live in a world where every payment every trade every mistake literally every relationship you have it's on the public ledger and um it's literally recorded forever and anyone anywhere can literally access them um via you know it scan or any scanning

tools And um you know bringing it back to what the cyber punks were basically fighting for like you know building towards for the future of the internet um they basically ne they never asked for like you know perfect transparency. It was more about selective disclosures. If I use anything you guys don't understand here just um holler at me. I would explain. Okay. And uh yeah so proves our way of you know um escaping this trap. Uh just to put it in context if your entire life is basically a is basically public it means people would be able to um access where your personal information is publicly people would be

able to know that you know you guys are literally here today publicly. You guys paid for a beer at the bar just because you use USDC and um it's available publicly. So uh this is what ZK basically um helps you with and uh yeah so moving into the topic of today are what the f is [laughter] what the f is zk so zk basically proves truth without actually revealing the evidence. It's basically lets you show correctedness uh without showing the data. It helps you prove a fact while um uh while keeping the fact hidden. Okay. So, it helps you reveal

the result of something but not the impute that you actually made or yeah it's basically proof without disclosure and um yeah so oh yeah got it. So uh to prove the validity of a statement. So this is me trying to explain um the underlying factors that make ZK what it is. I would still go further along as we progress. Um yeah so to prove the validity of the statements without revealing the statement itself. Um there are mostly two parties involved that you know makes zik work. One of them is the prover. The next one is uh the verifier. And um in some instances we have um third parties

which are called um the issuer. So how does this work together? The prover is the person trying to prove something. Um the verifier is the person trying to validate that whatever the prover is proving is actually factual or real. And um the verifier could be somebody who actually issues the um um sorry the issuer could be somebody who actually issues the particular information that needs to be verified or proven. And um you know it could be the government, it could be anybody literally and um yeah so um I'll give you guys a real life explanation of Zika in the real world so you guys could you know really understand what we're talking about in

terms of provers in terms of verifiers and in terms of issuers. So uh what exactly does can does zk do and how does zk work. So um ZK could enable you prove that you weigh less than 100 kg without actually you know making people know that you're actually 99.9 kg basically you weigh less than 100 kg. So, uh, for example, if you, um, if you get a report from the ELS facility that, you know, this is your weight, you want to prove it to be able to fly or do something and you don't want to show your actual weight. You just want to show, yeah, I'm, you know, lighter than 100 kg, you could prove it. Uh, you could prove you're old enough to drink, uh, you're 18 and 1 month old. Uh you could prove

you paid your taxes without actually revealing your income based on the threshold. Uh you could prove you literally attended the stock uh just like we use pope. Um you could prove you're rich. You could prove you're poor. You could prove you're over 18. You could prove your credit score is above a threshold which I explained. And you could prove you're 6 ft and not 510. You know the Tinder conspiracies and uh so much more. So um a live um implementation of ZK is the Google wallet. Um if anyone here is based in the US, you can basically um upload your driver's license or your passport to the Google wallet. And um if you're trying to like get a pass of something, maybe

you go to a bar and they want to, you know, you want to confirm if you're 18 years, but you just want to show you're above 18. You don't want to show your date of birth. You don't want to show your uh country of origin. You don't want to show all of those information. you just want to prove you're 18 years. So if since you've uploaded your um ID to Google Wallet, Google Wallet communicates with the issuer which is the government that gave you the ID to be able to verify those informations. And once you just you know tap to the um to the bar or the club you go to. Uh some of them have devices, some of them just needs to see a proof that you yeah eligible to drink or something like that. you could literally just you know pop up your Google wallet and say yes

I'm eligible to drink and that show was like you know a real world use case of using zero knowledge proofs and um yeah so how is that possible how does this work without just you know how does provers work how does verifier works what are the underlying cryptography underlying mathematics that enables this um happen yeah so um basically proofs comes in two categories we have interactive proofs and non-interactive proofs. Um interactive uh proofs are basically the ones that uh we use mostly in web tree and um be and I'll explain why. So before I go there I'll probably explain

non-in groups first. So this is um [snorts] I want to use this as like a case study. So just imagine that you know there's a ring shaped curve and um there's a gate inside a ring shape curved that basically requires a password for you to go through and um your friend um you know you're the prover your friend is the verifier and um your friend wants to prove that he knows a particular information because he's the witness. He has to go through the gates. If he goes through A, he's going to uh have to go through the password verification and uh he could tell you I'm, you know, I went through A or I went through B while he's

inside the gate. And if he goes through B, he just he could go in and come out the same way. So, um what's the point of of all of this nonsense? Uh the point is basically um interactive is basically the prover and the ver the verifier having to communicate every time um on the validity of a statement or the validity of an action basically. And um yeah so okay good. So um like I said earlier if you enter through side A you need a password to get through through side B.

And if you enter through side B, you could just come out the same way without having to like use the password. So, and this is a good example because it shows like you know um whether or not you're required to use the password uh while doing this challenge. And um yeah, so this literally makes up um some of the core fundamentals of Zika proofs. Uh there's the weakness, there is the um the challenge which is the question, and then there's the response. And we also have um the witness which are witness provers verifier and like challenges. So this is basically not very um sustainable in terms of interactive proofs are not very sustainable in terms

of um blockchain applications. Uh why? Because they have a big limitations. They have a big limitation. uh the even after because most times even after the verifier is like you know satisfied that yes okay my friend knows the password and you know I believe is uh uh he went through the tunnel the communication is only between the prover and the verifier third party could not cannot verify that you know this guy's actually um know the information or have access to this information and um yeah so that sums up the slide and uh yeah So moving into non-interactive proofs um like I explained before there interactive

proofs which are just give you an example of communications between both of them and non-interactive proofs which are used in most ZK blockchain related applications. Um there are two types of non-interactive proofs. Um the first one which is very very heavy and one of the first to be implemented. Uh Zcash was actually the first to do a running application of zero knowledge proofs. Um so it's called ZK snacks and then there's ZK stacks. We'll talk about ZK stacks later. ZK snacks means zero knowledge sustained non-interactive arguments of knowledge. And um the ZK stands for zero knowledge which means not having an information about something. Uh sustained test because

they are small and they quickly verifiable by the verifier. So it doesn't you don't need to take so much process to actually verify particular set of information and non-interactive which I think I've explained it means there's only one round of communication between the prover and the verifier you don't need to do the back and forth every time just once and arguments means it's theoretically based on math very very extremely unlikely to be able to fool the system or be able to trick the system into like passing information that's not correct and I would explain why you know this is and um of knowledge can't be constructed without having access to the information and um I think uh just to explain further uh ZK snacks

they use a cryptographic primitive called elliptic curves um basically it's a method of creating and verifying this proofs and um one thing about zest is that during the initial setup phase The prover and the verifier must literally agree to use a shared key. Uh this is called CRS which is common reference string and anyone with this anyone with the access to this key would be able to have details or have information on what was actually verified and what was actually proven and um basically this is what makes Zik Snacks possible. Okay.

Yeah. So to break it a little bit further um zik snacks consist of um three major functions key generation proof generation and verification. So um zik snacks really comes down to um these three algorithms the setup which is the proven um and verifying. So everything else is scaffolding. So basically at the setup you create the keys at that point in time. Um I think I explained at first what the setup entails basically the prevent the verifier keys. So um this is where the system basically takes the circuit. I would explain this also. Um and um the system the the statement that

we actually want to prove you know we could say uh so when you see okay I think I've not okay yeah so for the setup we need cx w is the witness yx is the public impute of the information that you tell. So you create a witness automatically and then we have a proving key which the prover would use to construct proof. So this is W which means witness X is a public compute and um PK is the uh public key private key sorry. So if the um if the prover which contains both the public impute and the witness is

incorrect or the PRF which is what is generated basically at the moment you prove wouldn't be valid. So it could be true or false depending on so this is why when I said theoretical it's almost impossible to bypass or to full seek a proof is because uh there's basically math behind it. And um yeah so um a proven key which the prover will use to construct the proofs which is the PRF and um a verification key which is VK and um the PRF which is the proven key and then we have X which is the public input. So a verification um the verification key is what you could use to verify that the proof was

you know uh made and either the proof was valid or invalid. So um basically so at a very very high level um zik snacks are just compile you compile a proof um you prove it and then you verify it. Compiling is where you set up the information the public imputes and the keys and um proving is the computation where you where you generate the PRF and then um verifying is the verifying key which you as a user will be able to verify particular set of information. So just to take this back a little bit to our first example. So there's a correlation of um of thoughts or explanations. So when um

you want to verify that you are above 18 to be able to drink someone or the developers basically set up the proving um the done the setup of the provers and um so at the moment where you prove the witness is your issuer for example like the governments they know that they know that you proved and um the public imputes is the information that you uploaded and then The private key is the one that you are the only person that is able to like you know verify this particular effect of information and what the verifier just needs is the verifying key and um the PRF which is the proof that was so whenever a proof

is completed it creates an ash function um have you guys done a transaction before using Ethereum so you know it has an hash function on the on it stand so basically that hash function is on the verifier and they could you know verify the app function which is generate which is the PRF that was generated when the verification took place and uh X you're probably input and um yeah so uh I think one very very important thing to note is both the prover and the verifiers although explained in math they are all algorithms they're not um engineering specific algorithms or so um yeah talking about stocks I'm more of a snack builder. So I'm not a

stack builder. Uh I think stacks could be best explained as the cousin of um ZK snacks. They came very very much later. Vitalix calls them uh the cousin of ZK snacks. So snacks means or snacks rather means zero knowledge scalable transparent arguments of knowledge. And why is it scalable? It's um I explained that snacks were succinct which means they you know require small proofs. So rather than stacks being succinct they are scalable which means they could generate and verify proofs with like larger weaknesses. Um the previous the previous explanation I gave to you was um you only did like you know probably

just one witness to be able to um attest that this particular set of information is accurate. ser you could use a host of witnesses and um yeah and it does it way more efficiently than zacks and uh yeah they're transparent they're not basically private um by nature although you could build more cryptographic primitives that make them private um I think another very important thing is that there's no trusted setup um what I mean by trusted setup is when I was talking about setup at the previous slide I told you there was like a general generation place and the um setup where you generate the PK and C and X keys um for stats there's no setup

so it's basically publicly verifiable randomness to be able to generate the shared key and um yeah that's about story I won't be going deeper I'll probably publish a blog later on this in case you guys want to read and uh yeah so let's leave Matt for a little bit and then we'll go back to Right. So what are the use cases of ZK? Um Helon like I said Zcash use ZK for private transactions. Sorry I didn't attach as much logos as I should. We have a lot of applications basically using uh ZK in production today. I was just very very lazy. So start looking for their logos.

So um we have Zcash for private transactions. Um they also use snacks. Uh we have to cache. I'm sure a lot of us are familiar with tornado cashache. Uh there's ZK money, there's um HC, it's an Ethereum L2, you also use ZK, there's Maiden, there's a host, there's int max like a host of um ZK production and then we have ZK for scalability. So we have ZK for privacy and then we have ZK for scalability. arms and basically stacks are used a lot for scalable uh solutions. You know, upgrading the blockchains, being able to verify information faster with multiple witnesses and all of that. Uh there's um ZK sync doesn't use snacks by the way.

They use um snacks. Um Polygon EVM, I think they use Snacks. And um there's Cairo. Cyro where Cairo is stacknet. You guys are familiar with stacknet the Ethereum L2. Okay, great. So um yeah, you'll be disappointed but yeah uh so stacket is a very very famous ethereum L2 they use um zero knowledge stacks to scale the blockchain and um yeah for identity solutions we have ZK passport we have um very mo ID we have um Google wallets like I mentioned and a lot of cases so um fortunately unfortunately we are only talking about privacy so I'm not going to be talking about like other

use cases of privacy of um ZK outside of outside of privacy related solutions but basically outside of this ZK could be used in healthcare could be used in um analytics data could be used in um um it's currently used by Adril the defense tech company um so yeah it could be using as many fuses as possible for LT tech you could imagine that you want to um probably go to do a blood test but you don't want them to know your origin or whatever you just want to give them a particular set of information and you want them to only see that information you're giving them so there's way a lot of use cases for zikk

but primarily for uh privacy private transactions I could send you a dollar and nobody else on it would know that we actually made a transaction I could do it anywhere to anybody in the world uh you could scale the blocks transmitted and I could prove uh I could protect my identity with zero proofs. So uh yeah so yeah um let's talk about the beauty of the mathematics. I will try as much as possible to scale down some of the components that u makes zk possible. So I started with sudoku because I think we are familiar with sudoku the game.

Yeah. So um basically I'm using sudoku to represent circuit circuit are a fundamental in in zero knowledge in zero knowledge proofs they are um constraint puzzles. So basically if you're playing sudoku every box has to and every column they have to be um they must satisfy the rules of the game. You can't put a random number in a random box. So um u execute works the same way. You have a secret solution and the system checks whether all the constraint are satisfied without ever revealing the solution itself. So imagine if you have a um a sudoku puzzle and there there are

numbers placed everywhere and you could actually verify that those numbers are actually accurate or inaccurate just you know based on VK. So circuit works in a similar way. Um yes and the next one is um elliptic curves. So elliptic curves are um they're basically like the uh playground. So that's where you use structured algebraic addition and multiplications um to behave in predictable ways um and predictable and verifiable ways. And um then we have pairings. Uh pairings are like magical to nails basically sorry basically between two mathematical

spaces. So you could use pairings to compare um you could use them to create information across across groups and like you know compare information across them. And um you know they're what makes they're basically what makes um ZNA succinct and you know they they help with like smaller proofs because you could do easy comparisons as they scale and um they also help with fast verifications and um so uh don't worry if you guys are not like really deep into cryptography the important idea is that math is beautiful and um this structure is basically you know obviously what enables ZK proof pro that um prove the truth of something without actually

revealing information about itself. And uh yeah, when most people hear ZK feels like magic, this is basically because of algebraic circuits, pairings, elliptic curves, uh these are the fundamental um primitives that basically makes Zik possible and um yeah, I have this actually. Okay, great. So um yeah, we have all seen Zik is very very powerful. uh it's a uh technology that doesn't rely on human errors. It relies on mathematics which a lot of things that we basically use today your devices has a lot of um mathematics based implementations and um yeah but it also comes with its cons basically because

implementing ZK in production is very very hard and it's very very messy. It's uh and I'll just go over some things that people don't like to talk about. This was supposed to be more of a depth conversation, but I'll try to um break it down in a way that we could actually understand it limitations and not just its limitations, but its limitations with the implications of tools or products that we use that use ZK cryptography. So, um yeah. Um sorry, I use programs ZK. I don't know what that means. It just sounded nice. Uh so basically so all the problems that happening um that happens

with ZK they're not just random are they basically they basically comes from all the primitives um underneath it. uh the prover cost basically when you're proving a particular set of information it's very expensive because it relies on um very very heavy maths amount amount of maths um and they're really not a flaw there just a consequences of mathematics so zik snacks also rely on like elliptic cs like I explained um MSMS FSTS um very very large circuit like the sudok I explained and um you know so proofs are basically expensive uh to generate and um then

there's a lot of bugs and constraints fragility I would explain this so uh what I mean by constraint fragility is that a lot of um the circuits that I explained earlier again they are very very fragile so if one piece of if there's a one if there's one piece of information or one piece of constraint that goes missing and um or or one mistake in a polomial commitment is broken the whole system is basically like dead. So that's a very very important thing and um you don't even see it because it just happens silently. So um there also metadata leaks uh because ZK doesn't cover your

um it doesn't cover the network layer. It just encrypts particular sets of information. I would go on to explain this further and um ZK basically only hides values of information. So whenever you make a transaction it obviously hides the value but then there are other things like the timing the routing uh what channel the transaction used to take place and um you know who related the transaction. Um you guys know of routers like um um what good routers like Jupiter on Solana, Metura, Unis swap. Yeah. So uh these are routers. So if you do a transaction between yourself or you try to do a um a

trade. So ZK could hide your addresses, could hide the amount of your addresses which are the values, but it might not if you don't you create new encryption to be able to encrypt those information. uh those things will still be available and you could get dxed. So and uh yeah um on I think I've explained su fragility um basically you don't uh I think from a developer point of view you don't just debug code as a developer you basically uh debug algebra so if you're not really good at math it's very right very w and um yeah there's also the trusted setup risks uh this is a very

very huge uh problem that I love people taking ZK to uh mainet as faced. So at the moment of creating proving and verifying keys if your setup is basically compromised you know somebody can like you know forge valid looking proofs um and then you know get access to every information that gets u transmitted over the blockchain and the user experience is horrible. Uh but then some of these problems are actually getting fixed by the way. Um yeah uh ZK um solutions and public currently being used at large. I personally worked at

min protocol. We um basically um reduce the amount of maths developers have to do. So you just use your normal TypeScript. We created SDKs that you could use to build applications that rely on ZK without actually understanding the math behind it. Uh so we just abstract the matter where you just have to do normal engineering. So um what else? So yeah all the problems with ZKS are just side [snorts] effects of the same mathematics that makes the powerful. So a lot of us would not really care about this part of ZK. We just want to use Z related app apps for the benefits that it gives to us. So this is why I had to like you know

implement this slide. It's um the illusion of privacy. It's all an illusion. So um and basically what I want to talk about here is there are a lot of wallets, there are a lot of applications that promise privacy and they are actually truly not private because and um because most ZK based applications they require something called a viewing key and this viewing key is a special key that can decrypt all your transactions. So ZK alone and because you need to be able to view your balance. So that's what the viewing key basically does for you to be able to, you know, you're on Metam Mask, you want to see the amount of money you have. Um,

if it's a private Metam Mask, you're going to need a viewing key to be able to decrypt uh the ash and be able to see your balance. So that's why viewing keys are like essential in most ZK applications. So um so the problem is most of this wallet they use just zk and they have access to everybody's viewing key or to all the transactions. So if they get sanctioned by the US government or by any government they could literally just you know give access to all your informations in public because they can decrypt your transactions. So your favorite ZK wallets or your favorite privacy wallets might not be truly private and um some you know some are like exchanges, custodians,

compliance sometimes even the RPC nodes or even the operators that have access to your viewing case depending on the architecture of um the applications and there are a lot but I wouldn't be mentioning them. D Y R and uh yeah so in this applications we can see your transactions um because of ZK setup some people still deploy their ZK applications on AWS and when you deploy there they still get to be able to you know decrypt your transactions. Some Zika applications have compliance partners to prevent the sanctions that happen to Tornado cash.

So this uh compliance partners can literally encrypt your transactions. Um some of them actually even linking K during you know backup or when you're doing exporting transactions and um some of them relay your transactions through centralized relayers. So if you see a privacy wallet that uses uh unis swap or lei or um stargate or any of these routers and you didn't see anywhere on their desk where they have something like you know encrypted mess or encrypted routers set up it means there's a lot of possibilities for them

to be able to decrypt your transactions. It means you're not fully um you know like you thought you were. So um some hardware wallets also store your uh viewing keys in plain text because they are not customized for private keys or customized for private wallets. Um custodians of your wallet can also see it. Um yeah basically everybody can see it. So you just have to make sure you're using um the right wallet. And uh why does this happen? I think I explained it before. It's the way ZK works. uh you need a viewing key to be able to do balance display. Um some institutions must maintain auditability.

So they need viewing keys to be able to do compliance uh to prove that yes uh we made $1 billion this year and uh this is you know our transactions. People on the public can't see it but you know the tax agencies can see it. They have to. Um also some snack schemes are not designed for selective disclosures and um proof to basically only hide what the circuit hides. So this is a very very fundamental part and um uh I think one way this has been uh getting fixed is uh something called client side proving. So basically you don't prove um uh when you

use a privacy hack for example ZK passports well that's why I listed them there if you use them the proving doesn't happen on their servers so there's no way they get to see the private keys actually happens on your device so uh they found there are ways they were able to like reduce the amount of proving size so your phone can actually do proving or your laptop or your devices can actually do proving but this means it's it's a little bit large. For example, for an identity app, it's like 500 mgaby for just, you know, scanning passport. Doesn't make sense.

So, yeah, that's just the uh cost of it. Also, um be careful about compliance rappers. a lot of mixers today or a lot of um pools today or tornado cash duplicates they actually have compliance and um this compliance actually have the viewing keys to be able to see your transactions and uh yeah I think that's about that so one important metric to leave zk basically hides the values the nodes your balances uh state transitions and your addresses uh it doesn't protect relate to transactions. It doesn't hide um timing correlation. It doesn't you

know decide who gets to see the viewing keys. Uh it doesn't help you in any way with RPC leakage and you know relay user link. So uh this doesn't mean ZK is not um uh useful. It just means there it requires like other primitives to be able to make it actually sustainable. And some applications are actually using this other primitives but some are not or many are not basically and yeah so possible solutions how do we solve this um on the computation part uh you could combine ZK with MPCs um NPCs are multi-party computations so it's basically um instead of one weakness to

be able to uh do like comput computations NPCs could connect multiple independent witnesses or like let's just call them validators to be able to you know it's a decentralized way of proving rather than just the operator being the witness and um yeah so it means no single laptop can you know compromise your privacy uh there's also FHE I personally love FH a lot it means um fully homophobic encryption and um okay so just imagine this um you have I have private data of everybody body in this room um automatically submitted to a form but I don't I use the private form so I can't see your personal

informations but I want to run a computation or an algorithm I want to use Google analytics on this particular form data I could actually use put this data on FH and be able to actually do computation on this data blindly without actually seeing all your informations so I could do an head count of how many people are here I could ask the data for um informations about you know where are people from here, what countries do they come from and I could do all of this blindly without actually being able to see all of this information. So um Z um FH is still very very experimental. It has mostly been research papers until just a few years ago. So um I don't

particularly trust it yet for production cases but excuse me a combination of ZK it's a very very good uh use case that's and that basically you know removes that part where somebody has to hold your veing key or an operator or the custodian has to hold your vehic basically and then there's z T means trusted um execute execution environment or something similar to that. Um so TE's are not mathematics.

T's are um machines. They're devices. They're like servers. Um Intel SGX etc. So they do private computations of information and they are fast. But the my personal problem with CE is again this is a personal problem is because it require it's it's dependent on hardware and it means if you store your applications data on an hardware if a fire gets caught up or somebody targets that particular data center it's really so um not as secure as Zika itself but yeah it's fast it's uh does basically everything is meant to do but it requires hardware or it's dependent on

hardware. Right. And um yeah, okay. Yeah. So I talked about the fact that ZK basically encrypts values and not the um network or execution layer and all of those particular information. So there are other ways you could actually um there other primitives you could implement alongside zk. So for kag myself my startup we basically use something called private intent. It means whenever you do whenever you want to let's say you want to do a trade for example uh you select um Ethereum to USDC uh you want to put a leverage of 20x you want to use a cross margin of this amount you want to do all of this

so these informations are passed to the to something called we call solvers encrypted so there is literally no public channel of your messaging or the intent of what you want to do and that's what we call um private intent. So the way it currently works is uh they are cross messaging protocols like across like wormhole uh when you want to do a transaction they use intent. Intents are basically just um I mean intent like the word in the English word intent it means a request to do something. So the user experience might not feel like you know you are passing a note you might just be clicking buttons but the back end

understands it as an intense. So uh we use this intense privately and then we have private meles which means um even if we want to we don't use any third party um routers but even if we want to uh we use only we use them privately. So we could we could um use a router through a shielded pool through our servers to be able to do um private transactions of any form that you want to do. And then um there's there's also mixets um basically it's it's mixnet rely on they protect your IP uh through RPCs they uh protect the timing the routing

information and all of this information and um yeah they are great. So the end game basically would be making sure you have encrypted states, encrypted execution, uh zk verified consent uh correctness and smart contracts where no one can see what's happening but everyone can verify the result. So um I think a live application of this I mean they test net still but um is Aztec the Ethereum L2 and I'm not um posting any project. I've just personally built with our tech and they actually deliver on they've been building for 8 years. So they actually deliver on all their privacy permits and that's probably the only thing I'll talk about and uh yeah

um good good shall I proceed? Okay, great. So uh yeah personally uh ignore my very very nice fancy name uh I personally think crypto is not crypto without privacy. I think privacy is a very very fundamental uh value that makes crypto what it is today. uh it doesn't make sense that you know we exchange freedom and sovereignity for transparency and um yeah um privat means privacy is not so thank you everyone some of them >> that was a lot

>> sorry >> no I mean like it was amazing like a lot of information actually >> okay thank you >> it was great really it was a lot [laughter] to handle because ZK is kind of like very complicated for for most of like for example for me they came from DeFi it's a it's a very complicated new topic and uh and things >> but were you able to understand anything >> actually yes actually yes I finally like connect some dots that it's one month that I'm like struggle to connect so thank you so much for >> you're welcome >> for this position which was incredibly educational. So now uh we are going into the podcast

section. >> Okay. First of all uh thank you for presenting the way you did today because honestly like this is a subject that I think could put a lot of people to sleep >> and I was particularly exhausted today. So the fact that I was able to retain all this information when I'm not a very technical person says a lot about the way you present. So >> thank you. Um uh I want to start with a more grounded question and then I'll go into kind of a wild out there question. But um you were talking about basically the weakness with ZK which relates to the viewing key and you talked about the two layers of the computational layer and then the execution and network layer and how these two interacting basically affecting the complexity I guess of

being able to read this viewing and then you also have the compliance factor and all that. Who would you say right now is solving this problem the best in terms of preserving privacy all across the board in terms of also like hey maybe my government just doesn't like me maybe whatever like I'm still protected using this application. >> Okay. Um again there are a lot but I'm only going to say or mention names based on my perspective and based on what I can publicly tweet about or talk about. Okay. So there's bias. Okay. >> Yeah. So um um Aztec is a very very great infrastructure. They're an Ethereum L2 but they still don't solve their uh view key problems. So which means if you view an Aztec um

application developers still have to decide whether they want to um you know secure the viewing keys part of things or not decide whether they or not they want to uh be able to view users transactions. So it's still they still it's just an infrastructure. So it's dependent on the applications layer. Uh but then they have another product being built on Aztec called Tacio. So what Tasio did was to combine ZK and MPC together to be able to do private shared state. Uh so there are multiple witnesses your validators network. So um know so if you build your applications on hashtec using tasio you get to be able to um prove to your users or prove

technically that you're not you know um handling the computations or provenence of particular set of information >> like centralizing the solution. >> Yes. Yes. Exactly. And then for sending and receiving are again buyers the Zcash is [laughter] >> really really [laughter] >> so yeah but um uh I think one of the most fundamental things about people in Ethereum and um even in Zcash is we all believe in open source. So which means anything I say can be verifiable by code can be verifiable by the white paper by the architecture. So you could see it in yourself. You could see by yourself

rather than you know what someone says. You can do your own research and like verify this particular set of information. But uh Zcash has very very strong crypto primitives that cannot be decrypted by literally anybody and they're basically um almost fully postquantum. So in the sense I mean that's a topic for another day but yeah. >> Well great thank you so much for that answer. U my second question like I said is a bit out there. Okay, >> mind this a little bit, but um I heard an interesting topic on a podcast recently where they were talking about, you know, of course where we go next with AI and you've got the people doom thing about it, talking about things like, you know, and I'm basically summarizing this information, but you

have an AI one day that's 160 IQ and it's like 2 years later it's 10x that, 10x that, right? You have AIs that are solving problems we didn't even know existed, you know, and coming with up with really unique solutions. And one of the the fears of crypto funks um crypto fanatics cryptographic proofs is if they're made by man can't they be broken by a higher intelligence and people have even argued do you think it's possible that one day AI will be able to crack Bitcoin right will AI be able to crack zero knowledge groups what would be your your answer to something like that >> okay great so um I mean this is why my name is Lucid ZK okay so zk is boxed by mathematics

Mathematics is logic. It's algebraic. It's additions and multiplication and subtractions. So you can't crack ZK basically or with AI with IR intelligence. It's literally math. But because they use computations and it is um probably inscribed on the blockchain or all of those factors um quantum could basically get to the stage where it could deppt but then literally almost everybody is already preparing for postquantum cryptography. So most applications Ethereum is basically working on a 4-year road map with Lin Ethereum to make Ethereum um postquantum proof. So uh yeah quantum is still light

years away and then going back to your topic on AI. AI basically I mean it's very very intelligent but then we are not talking about developers uh code we're talking about mathematics. >> So yeah it's a little bit of issue there. So what I think another way I can uh translate your question just for audience knowledge would be um because your public because the blockchain is currently transparent and public with AI surveillance has even tripled x or 5x.

So right now even if you try to use mixers or you try to generate multiple addresses you still make certain mistakes and um because of AI's intelligence you could literally map out all your transactions together faster than you know a normal researcher normal auditor will be able to do with yeah manually. So it means surveillance level as like you know 10x which is why you need privacy like what's up. So >> okay so basically too to come back to like the mathematics answers just like no matter how smart AI gets it can't make 2 plus 2 equal 5, right? [laughter] >> That was great. Thank you so much.

>> You're welcome. You're welcome. >> Great. Hello to the audience. Good to be back. Thank you for the presentation. I do have a lot of questions. However, we are running out of time. So I'll hit you with the main one. When we're talking about privacy, we have to talk about Monero, right? So, as of this moment, why should I use ZK or Zcash instead of Monero? Why is ZK better than Monero? Can you explain that? >> Okay, great. So, um I don't think there is a better or better than I personally like Monero, but even prefer their current upgrade. the current I've

forgotten the name of the new upgrade you're working on, but it would make uh Monero even way better than it currently is. But Monero uses MixNet, which is uh basically mixing your transactions to be able to um um remove direct pointers on, you know, what transactions you made or how you sense money. But then the last person that came, we were talking about, you know, AI becoming way more smarter to be able to, you know, de encrypt your transactions or, you know, understand the mixes and the patterns enough to be able to like, you know, show traces. Um, while Zcash has,

um, again, that's why I said I was very biased. So I am biased. [laughter] So yeah, but Zcash has um um Z addresses which are default privacy. They also have um Z addresses which are transparent and um yeah they have fully default private transactions. So rather than using a mixer and I understand Zika snacks it's cryptography not mixing and opuscation. I would personally prefer cryptography and mathematics protecting my assets over mixing sequences. Do you understand?

>> I just have one more question. It's kind of important one. You mentioned that you know you rather trust this encryption to mathematics instead of anything else. Going through your Twitter yesterday, I noticed that you said something very interesting and I want to highlight that. Okay. You said that human layer of privacy will collapse sooner than this technical layer of privacy, right? So what can we as humans do so we can improve our privacy? So we can then the technical part can help us even better question. [laughter] >> Okay, great. Thanks for reading my um so when I when I talk about human layer, I was basically talking to developers. Um I was trying to like make developers

understand that you could build the best infrastructures and the best architectures, the best encryption encrypt encryption schemes but if your user experience is fable um people would make mistakes. Humans are prone to mistakes. So you have to optimize your applications or your building process in a way that um reduces the amount of human effort required to make privacy you know work and um yeah so going back to your question I think that was one of the major reasons why I tried to like moved into limitations so we could actually do research ourself to understand uh I mean that's why I was

putting about this to understand ZK to understand it limitation to understand when people there are Ethereum tools that actually showcase ZK solutions and then they have the approvers and they are uh they have operators being able to decrypt your transactions but the only good thing is at least they're transparent about that fact so if you have an understanding of what operators need because most of them don't speak English uh if you have a understanding of what operators need uh you can then you um be able to know that yeah these are the good guys, these are the bad guys pretending to be good guys and then you could follow our profiles like web three privacy now they they post about and blog about and

talk about ways you could actually like you know protect your privacy way better um you could switch over from using centralized applications um I personally don't use any Google product I use um five or docks. Um for um my VPNs, I use Nim. I could pay with store, I could access, I could pay with some Zcash, I could pay with Monero. Um uh so yeah, uh it's it's a and then for charge, I still use Chajibility. So, but then there I still use some other privacy uh related product. there's there's nearly on GBT there's um there's a host of different

GT products so basically it's just having to I mean the privacy space is just booming recently so it's still a long run there are people who have been defending their privacy so much but when most things are already defaultly transparent but it's a little bit hard to get shielded which is why you know there are a lot of buildathons and people and tons of people building new applications so I think I've been talking too much but yeah the best way to do is just Follow more educational resources where privacy has a very good uh resource place. You could see what Vitalik to see Vitalik's privacy st you could see my personal privacy stack in full and like other people's privacy staff. You could ask questions and you know get answers and uh yeah

>> thank you for answers. Yeah.