Privacy Redefined: The Rise of Zero-Knowledge Technology in Web3, Is It Real?

Transcript

good so a warm welcome to everybody for this amazing panel today I'm also taking a seat why not um so I'm Ben the Corbin the um head of ecosystem at CV laabs and the chairman at the crypto Valley Association as well as uh I run my I'm a founding member of uh of a ZK start of myself building ZK for Big Data so I'm super excited to have these many hats on uh and today we'll have a panel titled as you see privacy redefined the rise of zero knowledge technology in web three and if it's real that's what the the question which we're going to test let

me move a bit further forward so I see you great um so today on the panel we have uh Jordi balina co-founder of polygon on the on the far left uh I mean there's not too much to say about Jordi his name speaks for himself one fun fact I wanted to share with you about him that when he was 17 years old he created a virus that was tested in his own computer and the computer actually never worked again so it was a very good virus so uh this is Jord with us today it's amazing to have him H next up we have on jordi's right hand the next one for you on the left is even even Shapiro chairman of the board and former CEO of

uh Mina protocol and uh we also have the current CEO of mina protocol sitting on on the cross AC Foundation Mina Foundation sorry Mina Foundation that's the right way to say and um so before Mina um um even was doing research in the AI in robotics space and uh then he got excited into zero knowledge proofs uh to bring a better regulating and internet post AGI um he's recently pretty often here in uh in in the ARA in Switzerland because Mina Foundation has just moved their headquarters to Geneva to Switzerland so it's amazing to have you here today as well with us next up in

the middle uh is Anna Grace kvki from the co-founder and chief legal officer of K protocol Kum protocol is building a the future of decentralized trading thanks to an onchain order book with embedded prevention of front running made possible with ZK proofs and built with the technology uh of polygon miden and the infrastructure uh fun fact about her she is a lawyer first but she wanted first to be an engineer so that's why she's super excited on all of these topics and she really has a very deep knowledge also on the tech side I can approve that you'll he today as well so amazing to have you last but not least

uh on my very left we have gazi uh the VP of corporate development at at Zama building a fully homomorphic encryption and pioneering that entire space so it's amazing to have you today um oh a fun fact I found out about gazi which I wanted to share you is that he used to communicate with his cousins at the age of eight in coded letters uh to inv to avoid interception by their parents so it's like so this is the setup for today and um so the starting concept why we are discussing this here today is in the first five months of 2024 um in the ZK field so zero

knowledge Technologies we have witnessed an influx of over quarter of a billion in Investments signaling a significant shift towards privacy preserving Technologies u in the web free ecosystem so um as said in this panel we'll explore foundational aspects just to warm up I'll ask uh our panelist to do that in a moment uh but also we'll look into tandem Technologies when we speak of privacy preserving in webtree towards the end of the panel we'll question some some genuine paradigms and how the shift is going on and if ZK is speculative or not so my first question I'll throw the ball right away to you Jordi um could

you warm us up by defining zik technology on a very very high level maybe for the audience who has never heard of it I doubt but let's see and uh maybe go the less technical way yeah I'm going to try I mean good let's let's see um sometimes we talk about ZK but the the the Z is com from zero knowledge but it's a little bit confusing the terms uh the way that I like to explain zero knowledge is by what's verifiable computations at the end is this is probably the easiest way to understand what's ZK theide is you have a computation uh can be a long computation uh imagine for example I don't know training an an AI uh system or or just Computing a huge hash or

whatever whatever program that just takes imagine one one week in a super in a super computer uh you know some really huge computation okay and uh I want to so I I I do this computation and I give the result to you and I'm telling this is the result if you R the the the computation this is the right result how you verify that this computation is valid the the trivial way is okay you run the computation again so you just rent the superc computer you spend one week and uh you you do the result and you say okay it's valid okay that's that's the the way it works that's a trivial way but there is a much better I mean that's a different way to do that the idea is that I run the computation

of course so I do the computation I compute the result and besides Computing this computation I compute a proof that this computation is valid the idea and then the idea is that I give to you the the the result but I give you this small proof and you in just few milliseconds you can't verify you know for sure that this computation is valid okay so this technology this verial computation this is what we call about zero knowledge we call zero knowledge because I can I can give you the result and I cannot I can so I don't need I don't to to for you to prove that this result is valid you don't even need to know the the initial parameters okay okay this is a little

bit uh confusing this is a part of the zero knowledge but uh the easiest way to understand imagine that this computation is a hash it's a non-reversible function so I give you the result and then you can verify you can verify that this computation is valid that means that um I am proving that I know the inputs I know the the input of the hash function but I not revealing to you what's that input and this is the hidden part of the this is the part of the zero knowledge okay but you see this this this doity and how you are using the technology is exactly the same uh uh but this doity is what makes we will talk about later about the applications this is what these two ways of using the the technology gives you the two big families of of applications but zero

knowledge is about that it's about proving a computation or proving that you know something without uh about estatement without revealing the full statment okay that was a nice summary at the end thanks very much um I think it's a great kickoff uh we are today exploring also privacy preserving Technologies a bit uh Beyond or in in in uh in connection with ZK and Zama is known for being the Pioneer in the fully homomorphic encryption technology my question to you gazi is does it differ from ZK or is it complimentary or how does it relate actually to ZK could you just elaborate on that for us yeah so it's definitely complimentary when it

comes to privacy so in the example we had here there was a proof of computation so that you don't redo the compute uh I'll take an example imagine you are a VIP Club I need to prove that my salary is more than €10,000 for you to give me the password in the ZK space I have the attestation from my employer and I'll prove to you that the attestation is correct and that my Sal salary is more than 10,000 and you will give me the passcode the drawback is that in the process you know that my salary is more than 10,000 there is another technology called secure MPC you take your attestation you spread it across five 10 20 servers no one has the information

the full information their process and they send me back the passcode they don't know my salary but also you need to trust that these servers are not colluding they're not cooperating to get my attestation back finally the FH case I I take my attestation of salary I encrypt it I send it to you you run the check that the attestation is correct you run the check that my salary is 10,000 and you send me a passcode but you don't know any information you don't even know whether you sent me the passcode or not so all the information is with me but again the drawback is that I could have cheated I could have

taken an Grace's salary and sent it to you and you would know whether it's mine or hers and so we orchestrate these three Technologies in Zama to make sure that I can prove to you that I know the salary attestation but without showing you the data and then I encrypt the data and send it to you so that that's how we complement the three Technologies and at the end when I want to decrypt I use NPC for the decryption so we believe that they they go hand inand it allows to reveal the least information while preserving you know a functioning system a composable blockchain okay thank you super interesting thanks for the

explanation um and Grace I would love to have you on we'll we'll have a section in a moment on a use cases but you in Kon protocol you use ZK you don't build ZK you use ZK could you elaborate on just very shortly on where exactly do you use ZK and what is it for um ZK allows I mean we mentioned how it allows to prove that something was like a computation was run correctly and um I have to tackle some like a bit of about regulations here uh everyone loves regulations um regulations um apply for deterence basically you know that you're not supposed to do something and there's someone watching around the corner who

will make sure that if you do not comply with it uh there will be a sanction there's always like this uh sanction section at the end uh if you don't comply this is what will happen to you the way we are using zki proofs is to go beyond that and to ensure that we have effective prevention so in the case of Kum it's about preserving healthy markets and making sure there's no front running and potentially other types of Market manipulations are simply prevented because we will require um a ZK proof that ensures that the matching script was applied as intended instead of having someone just verifying Based on data that no one was front running and subsequently um just sanctioning the the bad actor so here if ever someone

tried to go against the matching script temper with it for example so let's imagine there's a centralized entity running the script and willing to put forward a specific actor specific Market maker instead of someone else or trade against you it wouldn't be possible because we wouldn't be able to generate the proof and simply um in the entire infrastructure the trade and the match will not go through so it's a way of um trying to prove that we can enhance regulations with ZK proofs and create better products by embedding those regulations and how the product operates okay thanks very much super super interesting we'll get to the use cases in a moment uh turning to even um Mina protocals Berkeley upgrade just happened

on like three days ago on the fourth of June and um in preparing I was just broing through the community and also the news and in a nutshell there was like Mina was cool but it's got even cooler now adding with adding this CK application layer could you just give us a bit of uh context and expand on that what is actually happening with you guys yes thank you yes um so historically so far zero knowledge proofs have been fairly hard to program with uh you had to you know you had to be a cryptographer most of the time uh and the existing blockchain infrastructure hasn't uh exactly been built for use with zero knowledge proofs so there is

like integration difficulties as well as just challenges developing with it uh what we've done in the most recent upgrade is added like zero knowledge smart contracts to mina that basically are are similar you can think about them as same as like evm smart contracts but they support zero knowledge proofs at like this deep level uh which makes it super easy as a developer to get access to like these privacy functionalities these infrastructure scalability functionalities uh as as just like a normal developer that's trying to build uh the other thing it does is Mino Zer knowledge proofs have this cool property called recursion which means you can combine these properties together you can say a user can produce a proof of say that they're their age or their income or whatever and then you can then put that into like a scalable ZK rollup

which then merges all those proofs together uh for scalability so it's like the super general purpose use of zero knowledge proofs that now is accessible broadly to people that want to program with them uh which is super exciting because there's so many cool stuff you can build and now it's it's possible to build on in production yeah super interesting and uh really looking forward how this unfolds in the community and keep us updated um let's move on because I think that's what interests probably the most of the people out here or or in the audience is ZK a kind of like a busz word like this is what we're also trying to uh speak about today so when unfolding this question actually um oh all eyes are looking towards use cases because if

there no use case out there if we being like dactic on a very high level let's be honest that's still not a real thing and it's rather a bull thing uh bull uh be yes sorry so um use cases Jordi could you uh maybe elaborate or expand what you started with on the families of applications uh of ZK could you just pick yeah here we we we can divide the ZK and this very broad uh definition but we can divide the the ZK applications in what we call scalability applications in general the use of this ZK are huge proofs when mean huge proof are proof that for example for all the rollups or

for when you want to do a huge computation I mean this verifiable computation sometimes you need huge computers huge infrastructure to build these proofs and uh um you don't care very much about the Privacy you are just uh Computing something and proving that this computation is is is valid okay this is one kind of applications and here is very much scalability is the the important part roll ups specifically and then we have uh another kind of applications that are more related to privacy in General these are applications that are run or used to be run in the in the mobile I mean in the in the users devices or in the mobile in general are small proofs but are proofs that are not designed for scalability they are more designed for for for

privacy for hiding or if you want for Selective disclosures s and this opens mean Here We Will We Will discuse about but here is what opens all about the Privacy vote or whenever you want to do something and you don't want to reveal all the information these are the two big uh uh topics and then there is other ones I mean he mentions multiparty computations and some specific applications in other places but these are would say these are the two big families uh of applications where you can uh put each of application more or less of the CK thank you super to setting the stage for the use cases I would hop over to to even to you private

vote selective disclosure for identity could you speak a bit of that what is actually happening in Mina and in the industry uh yeah so it's it's so once you have this privacy property you can use it for applications like that exactly so um private voting is is a very useful one I think actually it was used uh Punk pole uh was uh used actually here for the startup competition yesterday confirm that so being able to say like here's a proof that um that I have not voted yet in this poll but I am voting now and uh don't you don't know who I am but you can see um you can update the results of the vote uh additionally being able to do things

with identity like like like you're saying uh being able to say um I'm like for example been kyc but I'm not going to show you details about my identity you just know that I've been been kyc uh or Additionally you can do things like saying here's a proof that like I am a human with a unique passport uh I'm not like a bot or something that's been replicated a bunch of times um when you're like making an online account for example uh if that makes sense for all the different kind some of the different kinds of things you can do yeah yeah super interesting going hopping on to gazi um Zama what do you build what is there really actually built handson with using your technology uh your uh you've

been I know you've been building and and um very hard on the community and developer side you've been also hosting events and participating events here in in the zurk zugara um could you elaborate on that what is in brewing with your Tech out there so I mean we made a few announcements of l1s and l2s that are launching with FH and hopefully we'll see that in Q Q4 Q you know or q1 next year uh we are talking to a lot of funds uh either to tokenize their uh mutual funds or to tokenize stable coins I'll go back to the salary example and you know I may sound very materialistic but if I if you pay me my salary in stable coins it could be a problem if

the whole world knows about how much I make you know so you could give me that information and crypted and I'm only the only one who is looking at it now futuristically hopefully in two three years we we use today ZK to prove that I know some information so if I submit a trade I prove that I know the clear value while encrypted what I what we want in the future is to run the FH with ZK so I can prove that I have executed the F computation with the ZK proof so exactly what you know uh we're discussing about the proof of compute but at the level of an FH operation but it's hard work it's it's two three years of research so

we're not there yeah understood however I'm I'm I'm pretty sure that some handson uh teams are already building with FH yes something very close like a protocol like not just like a layer one could you mention like something which the audience could then Google up and see who's building with yeah so in decentralized AI for example we have a few clients who want to basically run AI protocols on confidential data and so the FH comes handy they haven't launched yet so you know if you Google you'll find a few names you know I'm not going to play favorites but there are at least four clients that made public announcements uh the reality though I'm

very excited about tokenization and and you know these funds and Banks don't want to publicize too much but we have you know at least on the tokenization space we see a bright spot for ZK and F so it is a reality today okay yeah understood um thank you um turning to Evans Banks and corporates really had a narrow and limited uh Channel um of collaborating uh or any of you could actually take this one and so collaborating on data in general has its privacy concerns scalability concerns and and it's it's like a narrow Niche and how does zero knowledge or or privacy preserving Technologies at

current stage open up this opportunity I can go briefly but then anyone feel free to add on go ahead um yeah I think that the one of the Cor possibilities that's unlocked is yeah like you you know as as a business you probably don't want to put all of your your private data directly on chain uh for a lot of reasons um but with zero knowledge proofs you can start putting proofs about that computation on chain proofs about your sensitive information um and those can be like you know combined together to put a lot of information on chain at once as well so if you have a lot of users who may Andor have a lot of you sensitive data you can start thinking about making that interoperable with the onchain world with zero knowledge proofs which which

is which is cool um I would simply add you can share all of a sudden all the relevant data simply without all the rest because you end up sometimes submerged with a whole bunch of data where you only need one line of information out of a specific document the corporate world is the one where you have lots of certificates be it regarding the the entities the users and so on so you have whole files that you need to go through and it's a good way of simply saying this is complied with this is correct we can proceed with that transaction we cont this without having this to go for this entire file and also taking in some cases um unlocking potential for further automation that's reliable by taking out this idea of there is someone looking

and screening for this information and who's supposed to to check I mean one of the applications is like how can we improve kyc process but I guess we'll discuss it further when we get to the the challenges because there there's a lot of challenges there but there are other processes in the corporate world that can be improved and maybe one thing we haven't mentioned in the use cases all the health sector this is where we have a lot of sensitive data uh be it's for vaccination prescriptions can you get that drug from a pharmacy are you supposed to um this is where a lot of things can be improved and for use cases this is like a whole bunch of use cases that users care about because what happens with blockchain is we talk about technology that is exciting to us but

not necessarily to the end user with ZK um like not scalability part but at least um the Privacy part users will potentially get a bit more excited about blockchain all of a sudden because it's easy to understand okay understand Jory you want to to add to going a little bit deep on that but U we see the financial financial industry the quantity I mean the the huge effort that means keeping data of the users but actually for the real business I mean a bank they don't care who you are and you are the money comes from from the business perspective they need to do all the collecting all this information because regulatory things but uh many many business especially if you go more uh smaller

business that are based on financial they have a huge cost I mean huge things to do in something that's not directly related to their core business with zero knowledge and this is the the nice thing is that you can um guarantee that uh the fonts come from the right place that uh so you can warantee all the all all the all the premises that you supposed to come maybe you can guarantee that some information is deposit in some place so you can give all the warrantees that are required that because uh all these Financial systems are collecting all this information but without having

to collect this information you just need to proof that uh something that your money comes from right place or you just need to prove that you are see it for something else or you just need to prove that if there is an issue your all all your information can be disclosure somewhere okay you just need to and and this is the cool thing I mean this this selective disclosure so you just reveal you just give the information that you are required going to back to the F to the security systems I mean when I'm when I'm when I'm Contracting an insurance now I give to maybe I need to give them the my whole track record my full medical track record just to see if I'm apply or I don't apply to to an insurance uh things I can prove I just

can give this bit of information I can apply without revealing all my data all my data to the insurance company yeah so all these things I mean this this selective disclosure of the data this is what the the the the zero knowledge promise is you can do this selective disclosure of the data without H without um having to renounce to the security of the to the the main core why all this information is is selected and this is the I mean I think this is the biggest application of the of the of the of the Z yeah um sorry we have to if you want to speak about challenges we need to move on um okay I've got three topics you can grab whichever you like uh and go just take one one each we have uxui

usability who is using actually and you from of users perspective then we have the legal challenges I'm counting on you uh and a great but to everybody and also scalability problems and issues uh FL is yours let's use like two three minutes max on that so then we can just make a little glimpse into the future at the end can start maybe if you want in the go ahead I'm I'm going to I'm going to start very much in the challenges of the ZK Tech uh right now so I would say that at this point we are changing a little bit the era so until now ZK were really complex to generate applications on that were very very I mean you needed some good knowledge of the technical part of

that we are entering to the space that uh the ZK uh is going to be commoditized uh everybody should be able to build the program just in R and and and just uh checking and verifying this computation somehow so this uh commoditization of the ZK is coming here to mention that the hardware acceleration in CK is something that's really uh that's what we will will see in the next uh month and this uh general purpose virtual machines and this uh um this uh Hardware acceleration these are the M challenges and the the big things I mean expecting uh two orders of magnitude acceleration on the generation of these proofs is

perfectly valid and this will enable a lot of applications in the scalability and in the in the Privacy part thank you next up even go yeah we go left right on the uxui uh challenges like particularly around selective disclosure it kind of requires like uh new UI patterns that users will have to get used to and they'll be build into applications uh because often you want to store the data locally on the user's device and then have them make the proof that gets sent out um so for example we're looking at augmenting wallets to include um uh sensitive information stored in your wallet that then you uh your wallet similar to how you sign a transaction can make a proof um but there there's like new new interactions that'll have

to exist like that for this technology to to do its to do its thing okay the legal challenge go ahead I mean J touched on the upon quickly the lots of the information we store currently is because of regulations requiring them so there's a lot of CK applications are being developed but that cannot currently be uh implemented or at least used by some entities simply because it wouldn't allow them to comply with those regulations uh so you need to take into account what are the current requirements would we still be able to fulfill these requirements by using specific applications of ZK proofs and like one of the challenges is how can we push and Lobby for the regulator to take ZK proofs into account I know that Jordi

would love that to even make it compulsory to use and request CK proofs in some cases um so the major challenge is is what I'm building also accepted by the regulator so for example if we mention uh because it's always a Hot Topic how can we improve kyc by instead of storing all this data simply requiring a CK proof and making it faster more convenient more private great but a lot of these applications are not compliant so if you are a regulated Financial intermediary you cannot replace your current kyc process simply by ZK proof because you need to store the data you need to be able to access it if ever there are questions that come in um so just figuring out all of these things will be um is a

challenge currently because you may have a great technical idea but the use case is not fully fleshed out yet from a regulator perspective and for to convince people who don't want don't care about privacy this is what I wanted to mention earlier we also care about a lot about sustainability and having all these uh data stored online or in paper is not exactly sustainable so we could maybe push for ZK as a way of achieving some of the sdgs okay that's nice one yes Andi has similar uh challenges in terms of Regulation Hardware acceleration I would say Hardware acceleration will probably EUR to be behind ZK as in you know there are projects funded by daa but they haven't seen the light yet uh maybe to add one is cyber security because as you work

with banks or with funds at one point you need the sign off of the information security team today we started the process of getting standard so if FH will become an ISO standard but we also need the approval from the nist so that we're part of the list of approved cryptographic tools that Banks can use and that takes also a lot of time okay but we're getting there yeah it's great toar so it's actually if I understood correctly um ZK is is a mathematical proof so when we speak about Regulators or accepting it it's actually accepting clear logic uh the question is when and how and who can make this happen and it's great that I think all of these all of you in this panel are really forcing

into that bringing use cases building on the infrastructure and building on the tech and going onto these panels it's amazing last question very fast a short sentence from each of you please I want to hear do you think it's ZK summer now and the second thing is um how long will it last I would say I would say that uh we haven't seen yet the the the the summer I mean the the explosure the explosion of the ZK Tech we are very close to that the yeah there the I think this Hardware acceleration and the this

commoditization of the technology is when we will see that all the all the all this deck in its highest um view yeah I think ZK has really big potential like to be um as big um if not bigger than but probably integrated with cryptocurrency um so I think that like it kind of depends how far it goes and I think we are kind of in in Spring given like the the potential of where it can can lead to yeah definitely spring we have like the flowers starting to get out of uh their uh the major yeah they start to bloom all of a sudden it's thanks to projects making it easier simply to build with ZK proofs because we have a lot of ideas everyone I think everyone by now knows like this is what

you could do with ZK but then how do you go from the idea to the implementation and we are in the phase where we start to figure out the tools and how to build with them to improve those tools so that those ideas can become reality and all these use cases can be implemented and we can hope for a summer where these projects are out there raising reaching out to users and start being used I hate the seasonal uh cyclicality because Summer means that winter will come later which I hate you know I'll would say privacy in general as a theme has been growing the way I see it is it took 20 years for https to pretty much replace HTTP 20 years hopefully it will take less yes but you know we can see

growth we hope that the growth can accelerate but it will definitely get there it's just whether it's 5 years 10 years or just six months who knows okay so I would conclude that we are in the spring rather not in the summer yet we're in the spring of ZK and privacy preserving Technologies and going towards a Everlasting summer right amazing thank you very very much for your listening thanks for our panelist thank you very much