Panel: Getting Started with Zero-Knowledge14th Oct 2023 | Sunway University | ETHKL 2023

hey good morning everyone um great to see you here bright in early on a Sunday morning you could have done anything else like sleep in was it raining this morning yeah but you're here so that's awesome um thank you so much for joining us so I'm deborra Nita I'm the moderator for this panel called uh getting started with zero knowledge the first session for I think there's going to be a couple of other talks focus on zero knowledge so I hope you stick around um a very quick intro I'm a senior account manager at Yap Global Yap Global is a PR um it's a public relations firm Focus solely on crypto native projects so we work with um some decentralized Finance projects and infrastructure if you've heard of Lio ens um nil Foundation Manta Network gcoin a whole bunch so we basically help our clients get into the media and get publicity um and I'm joined by our two speakers today they're both from the ethereum foundation I think ethereum Foundation needs no introduction um we have uh CCE who is a senior engineer at privacy and scaling exploration team and we've got one sub who is a researcher at privacy and scaling Explorations maybe you guys want to do a quick intro of yourselves yeah thanks uh I'm CC and uh I've been in crypto space for five years and and building ZK circuit for for almost three years and uh we're building the ZK ebn these days uh yeah glad to be here I'm and uh I'm working as a researcher just um recently looking at some programmable cryptog graphies and yeah happy to be here thank you okay quick question Who has zero knowledge of about zero knowledge hands up okay so everyone else has above zero 0.1 okay uh just want to get a quick sense of who's in the audience um if you are a developer in web 3 could I see Hands okay cool um does everyone else work in web 3 or are you are you okay students any students here okay nice uh anyone who works in web 3 but not a developer or maybe you're trading and okay nice so I think everyone's pretty familiar with okay everyone's familiar with web three topics in general okay well we're we're not going to go so super deep we're going to touch on some high level topics because that's what the rest of the day is for so we'll touch on um yeah some high level themes and maybe we can get straight into it but uh first to start off can um maybe one of you guys explain zero knowledge in a nutshell what's your goto to understand it yeah hope I get this one uh so Z you can think it as a digital signature for computations and for this computations you don't uh have to reveal like uh every data in this comput uh computation so you can hide important information you can build uh privacy applications and and uh a lot lot of uh scaling applications with uh zero knowledge yeah um and I can give a little bit more kind of theoretical some definition that uh zal prove is proving some relationship so the the most uh one simple example is your um transaction signature ecds say the litic curve signature actually you are proving that you have a private key for the public key and it's pretty short right so in zkp when the relation is too big the proof actually also grows but what we really would like to do is that okay we would like to keep the proof still small when you would like to prove a really big relationship um and also proving during proving the relationship uh we should input some values for example your private key and in the leptic signature you could hide your private key while reving your own only the public key and uh in zkp actually we should be able to choose what kind of information you would hide while proving the relationship is the what do Z keep you Works was it simple or yeah can you talk a look uh just um for the rest of the conversation zkp would I mean we just shown zero knowledge ZK and the p is proof so zkp we just refer to it like that when you talk about a relationship can you expand on what you mean by relationship or maybe giving a simple example yeah the simplest example is the ticar signature that you have a public key and you have a private key and public key is a derived from the private key this is a one simple relationship and actually we can expand this relation into a arithmetic circuit which something like a + Bal C and A * b equals c something like that actually and this summation and this multiplication exactly what we are doing in the computer so it means you can write some program and we can prove the input of the program is the output of the program you write and that's the relationship that we would like to prove here okay great um so I don't know about you guys but uh I feel like we've heard a lot about zero knowledge it's been hyped up a lot in the news or we hear about it on Twitter or just in discussions what I'd love to know is is it really a big deal and uh if it is why is it so important um yes it it's uh definitely a very big deal um I think uh it uh I think as you can get an impression that like it it is not a concept that it is easy to understand but it definitely builds a very uh different um different application different uh computation uh paradigms that we uh we haven't experienced yet before and uh the current like um if you re watch the uh the code the research paper and uh published for this topic in this years and the speed of topic like month by months like it is a very uh like people put a lot of money and uh efforts into uh researching this topic because it might solve some very uh deep problems that um that frustrate Us in in current digital world so that's that's really interesting can you touch on one or because you mentioned earlier that uh it can do something that's not been done before and it can solve some big problems that people are or developers are struggling with maybe one or two examples yes so uh I will start this example as a a digital identity one so uh as you know you probably uh have some like uh identity card that say uh that has your names that has your age that has your genders and uh you you are kind of defined by uh a lot of this attributes but a lot of uh um application of um this uh that requires only one or two of this attribute for example they want to uh the application might require you to prove you are uh uh citizen of uh uh somewhere but they don't require other attributes and usually we are oversharing the other attributes in our daily lives and that's why we don't have a good privacy but with the zkp we can um prove that uh yes I uh this this attribute comes from this identity card and identity card uh belongs to you so so you can uh prove only the attributes that is relevant to to the application setting but without like reving the other irrelevant detail that uh hurts our privacy that's really that that's a great point you want to add to it I like to add one more Point here for about the scaling Point too because actually C mentioned about the Privacy that we can achieve through the ZK but also if we have the zero knowledge actually it means like you already proved the relationship ship it means that you don't need to run that again for example what this is doing the for the zkm is that making a proof that the state transition of the ethereum block is correct which means you don't need to run the hold the reex execute all the transactions at all because we have a proof which can be verified in a pretty fast uh proving time a verification time so actually ZK is also very matters uh in terms of the scaling too so yeah so interesting so we were talking in like in terms of themes it helps with privacy it helps with scaling maybe we can touch on both of them a little bit um in terms of privacy so earlier we were discussing about how you know for the longest time humans have always used say weapons physical weapons to protect ourselves but now we have a very different kind of environment to operate in which is the digital world and we need some weapons for that as well so can we talk about how zero knowledge helps us to um how is it a digital weapon as you guys said earlier yeah so to add a little bit here for about the digital weapon actually not only does Z knowledge the digital weapon cryptography itself is a digital weapon and zero knowledge proofs the recent improvements in Zer knowledge proofs will let us uh use this weapon in a really really easy way regardless of the proving time and the verification time so yeah and why is it important to protect ourselves um because it's also related to the privacy because I think I can introduce some uh store now and decrypt later strategy of the quantum computation so some intelligent organizations are storing every data uh in the internet because they are based on the RSA signatures and uh if it's a store them into a stories for right now maybe 10 or 20 or 30 years later we're going to have a quantum computer which can help us just decrypt the RSA and then we can decrypt everything at the time so that is a kind of a strategy and also to protect ourselves from those kind of attack we need a much better cryptography to prot protect ourselves for that and um this is how we are protecting ourselves from those kind of attacks and this is why cryptography itself works as a our own digital weapon in this in this digital world sorry just to jump into that a little bit more what happens if we don't protect ourselves or what could happen um you're going to give your all information right yeah and also like uh that's a without cryptography like everything uh you see on the website you cannot trust it uh the the thing because we can trust the content you see on your mobile phone is because the website is uh signed by the uh the a signature by by the uh by the server by the uh people who provides that content and that kind of signature uh is uh managed by a a uh certificat that there's a uh identity um management like uh like like organization and and that uh that kind of uh defines our current landscape of like internet browsing and so as our live increasingly like uh like interact with the digital contents and like we are kind of shop onlines and like uh in our bills and like Banking online like um we we need more uh craphy tools to to make our uh like uh online lives safe yeah so it's also it helps to differentiate because it I mean it comes down to proving something is true so it helps differentiate what is true and what could be a scam maybe what what cannot be trusted yeah yeah yeah that's true yeah and and especially when when we uh think about uh scam usually scam uh is trying to exploit the current uh loop Poes in the in the in the uh current system that cryptography didn't uh defend you so uh usually scammer strategy can be summar like they try to learn more about you they G more information about you and the more information they get they are more convincing to you and then eventually you you you trust a scammer so so I think that's um very important like how how do we control like our our information to share and and and that that helps uh prevent scanners great thank you uh okay so we'll just move on uh a little bit more into what's happening right now in the landscape um what are some uh important developments you think that's happening maybe in the past 12 months what are PE what are developers working on what are projects working on that you think is really interesting okay uh so uh so one thing I uh see is the uh zkl the uh zero knowledge uh machine learning it's uh still very um early early project but uh I think uh it uh one of the interesting uh application you can build is to uh prove that like Twitter used uh a certain algorithm to to uh serve the content to you um because it's useless that people uh the Twitter publish their algorithm because you don't know if they actually use the algorithm to serve the content to you but dkml can prove uh to to you that a certain uh tweet is ordered by that algorithm and I think that helps increase the fairness and make the uh social media companies uh more accountable and I uh one of the uh researchers who's uh building dkml is uh Kathy who is also in uh this confence and is she here in the room oh yeah oh hey Kathy interesting yeah so yeah please talk to Kathy and like learn more about dkl great I think uh Kathy you're speaking later yeah okay what time is your 11:30 if you want to learn more about ZK ml stick around folks okay uh any thoughts yeah also uh I have a talk right after this discussion and my talk is also related to the topics what's happening in the recent ZK world uh and just to share that here briefly that uh we are going through more for about the programmable cryptography not only for the journalist proofs but also handling the multiparty computations and Etc and also uh we are going through actually there was a pretty big uh paradigm shift from the Univar polinomial to the multivari polinomial that I will share in the talk later which lets us uh prove something in a much much much faster uh compared to the univariate setup which means um yeah I mean you can think like uh before that to prove the ZK evm it takes like how many hours does it take maybe like 10 or more like like like for yeah yeah for current performance yeah yeah it's kind of very maybe less than a day right yeah yeah almost yeah yeah but actually we should be able to do that in a kind of a millisecond or second yeah this is uh what we have achieved in the recent uh progresses and yeah actually this is a really big shift so we are trying to make everything provable uh in an easy way so which means that our goal is something like you write some program and code and making it as provable any kind of program because now we see that okay it might be pretty feasible and so the shortening of the time frame so before used to be you said just under a day now we're making significant progress and proving is becoming very very quick so why why is that so important I mean will it help for say user adoption of certain kinds of apps cuz you know users don't want to wait so long maybe it makes it more expensive with the previous Hardware let me give you some example you are um developing a racing game simulation ground Trace mode simulation game and then you write every source code and people run the source code in your Play Station and run that actually that's not provable but we can make it provable and it won't take that much time so which means that if you run the program and provable then you can prove that your gr small uh Sim Rising recorded kind of in one 1 minute and 24 seconds with which is faster than Marx first up or something I mean yeah this is possible so people can write something make it provable and we can use it in many way interesting thank you um earlier you mentioned zkm um so I think we've I don't know if everyone's familiar with ZK avms but it's been covered in the news a lot we've seen big players like polygon and consensus introduced this maybe can we touch on that a little bit what is ZK evm and why is it so important I think earlier we touched on you know why Zer knowledge is important we talked about privacy maybe this is more towards scaling so yeah can we break down what ckvm is and why it's important yeah sure um so uh so don't be foed by the ZK part in the zkm so it's it's actually a scaling project it's not a privacy project so uh as you might know like um you uh when you interact uh with ether you actually uh or if you send a transaction on etherum you are actually running uh a computation on etherum and that computation is uh performed by the uh every like four nodes like if you run a for no on your home you uh process those computations and verify if the computation s uh carried in a block uh is correct and and but uh the the current way your uh not do is to uh reexecute the computation and that's very inefficient and that uh that limits the uh how many transactions how many computations you can do on evm and so what ZK uh the ZK SAR can uh ZK proof can provide is um we compress this computations into a cryptographic proof so the idea is verifying the proof um directly tell you the result of the computation so you don't have to rerun the computation again and uh that helps uh uh scaling and that helps us to build some uh better like clients uh for example if you're using metamask or everything uh like that uh you have trust uh the balance that uh infura shows you and if you can somehow validate that uh balance information then then you you don't need that layer of trust and there are other cool stuff you can build with ckm and uh for example like uh you can build a proof or exploit that uh some of granty was uh exploring um which is when a hacker uh try to hack a contract and they they could store a million they don't have to do that they can prove that they can do it and Clan the Bounty and fix the system and like we can turn this blackhead hacker into a white head hacker and uh and that that helps the the system like become more uh safe and yeah finally I want to cover that like ZK EVN is um just one uh uh it's just one part of a bigger uh uh set of uh ZK VM like the goal is trying to use zkp for to prove arbitrary computation and which one up can cover more uh yeah I mean right so I mentioned that our goal is prove everything so we can our maybe the final goal is proving the the computer itself maybe in the future maybe we would like to prove the quantum computation too because it's all about mathematics pols so because what we would like to do is just proving the relationships and for right now our goal the next goal is proving the making the computer itself as provable and it's a easier subset of proving computer computer is proving a VM which is subset of the computations the um and the subset of the ZK VM is is ZK evm and also which is a pretty good uh use case just proving ethereum change is correct so yeah this is uh about the zkm and zkm so from computers to VMS to evm so VM so well evm stands for ethereum virtual machine right okay gotta yeah because you know eim is also computer W computer that we have only one in the world so ZK avms are like the first use case um for for proving a well a programmable blockchain as we say yeah it's a very good starting point uh to try to prove make making a computation provable because evm is pretty simple set of some VM okay great um what could happen after we say we go through this use case of proving ethereum what could come next what's like what's bigger what's better than that after the proving the ZM MH um yeah that is the case what I just mentioned the ground TMA racing simulation game The Proving everything every relationship mhm okay great um we also so we've talked about you know what's happening right now in terms of zkm is being discussed a lot what else are developers working [Music] on uh like earlier we talked about uh there's a lot of updates happening with developer tools because there's still some problems around that so maybe you want to expand uh yeah okay um so yes um yes so the ZK space is uh still early and we are um uh we are Laing a lot of developer tool like um it um and so far like uh current Uh current Z applications you see in production are are written in a language called uh circum which is a short for circuit compiler and and and in this years we we witness a lot of new proving system coming out we have some like folding proof we have like recursion proof and we have lookups uh which which I cover yesterday and uh like uh C uh not yet have can do like every of these so we either need new language for that we either need to extend the uh circum for that we also need like different kinds of uh tooling to improve the developer experience uh for example uh z x build a very cool application that you can program SE on a website like tools simple like this can greatly uh improve the the the life of the uh that pris what was the name of that um uh ZK repo how do you spell that z k p L.D R .ev okay interesting any other yeah you know for the benefit of the audience any projects that you think are cool or doing interesting things that maybe they'd like to follow as well want to share uh what like interesting projects uh that you they're doing cool things they're developing cool things besides for example ZK repple um yeah so so um uh for example I think uh H to is also one so how do you spell that uh h a l o hello to Tool yeah so this is a tool built by uh zcash and uh it was the uh it is a tool we used to build the zkn and it was the tool not only available to uh two years ago that can do lookups and and uh weet the Plage proving system and it needs uh more developments and more more improvements yeah because uh if if you uh write T to right now you'll feel like so frustrating and and so like like it it's not the best experience to to write a circuit yeah gotcha one up yeah so I mean for right now uh ZK itself is a very in the early days because of that most of the death toolings are um does not have some good experience for you mostly so because of the and also the main reason is because uh hold the proving systems and all the Primitives are evolving and improving so fast so every day every month we have a new POC new minimum version of working so because of that the death tooling experience is really bad for right now so actually we can think about that into uh types for example if you are just only focus on the applied applications that maybe actually um you can just try to use the circom and like Halo to Etc but also you might be also interested in contributing to deliver a much better depth tool Lings right and in that case maybe actually recently because we had a big paradigm shift from the univariate to the multivariate so we are trying to make everything using some uh um multivarious setup and there are two very big libraries which is some something is say artworks and the other one is a uh ZK crypto and those two Frameworks are kind of a base of the how to implement uh this multivarious stuff and maybe if you can try to look at into that maybe you might be able to contribute to the dep experiences too cool thank you so much well we're out of time um this has been very interesting I hope everyone's knowledge is above zero um if it wasn't earlier but we're going to learn so much more for the rest of the day if you want to follow what they're up to you can check out psse dodev website or Discord um and yeah thank you so much for joining us I hope you enjoy the rest of the day and yeah thank you thank you thank you thank you so much