What to Know About Zero Knowledge

Transcript

hey everybody I'm Albert and I'll be your moderator for this next panel um and so for starters I think we all can agree that it is an exciting time in ethereum and for me personally one of the biggest reasons I enjoy being a part of this ecosystem is the really special Extraordinary People I have a chance to work with and learn from and two of the best examples of this are up here on the stage with me today Barry Brian every time I talk with you two I learned a ton and it's usually fun too so I'm excited for us to have another conversation here today so the topic of this conversation is what to know about ZK and that's a pretty big

topic but I think we can make some Headway for me when I try to improve my understanding of something I try to look at it from a variety of Vantage points and when it comes to ZK I think there might not be anyone in the entire world with better Vantage points than the two of you especially when we talk about ZK in the context of ethereum's future one reason you two have such special Vantage points is not only do you do technical and product work but you also operate as coordinators you provide guidance technical assistance to a lot of other people and projects in the ecosystem so on that note can you share with the audience a bit more about you

know just what you do on a day-to-day basis I'll start with you Brian thanks so much for the introduction Albert and thanks again to edcon for having us here so I gave a really brief introduction uh during the talk on programmable cryptography but uh you know just to reiterate and give a quick intro here I'm Brian I'm the executive director of Xerox Park which is a relatively new research organization focused on application Level r d on ethereum and other decentralized platforms we're particularly interested in applied ZK and it's one of uh it's our currently our most major area of focus because we're really excited about the potential for ZK and related Primitives to really unlock the next

generation of uh ethereum applications for as far as myself a lot of my day-to-day is spent thinking about how to empower various grantee teams in our ecosystem with new project ideas with new educational materials coordinating between different teams to put them together and figure out like maybe team a has a gap where Team B has knowledge of and vice versa and they should be talking with each other and trying to holistically understand like what are the needs at the different levels of the stack from infrastructure all the way up to application um to understand like where might we need to put more either grant funding or educational resources or whatever else to ensure that the whole Space is kind

of moving in tandem cool Barry um hey I'm Barry um thank you Albert for the nice introduction um so today yeah so so I work with the privacy and scaling team at the ethereum foundation and we have like two major focuses our first focus is on privacy and our second is on scaling so today we'll talk mostly about the Privacy stuff or the like applied zkp stuff I suppose that privacy is probably is maybe not the best description of because we do some scaling stuff with gkps and we don't do other like proof of validity stuff with tkps so that's our main focus and like we're I'm excited to talk to you today more about like how we can apply zkp to different things in the

ecosystem awesome thanks so much for that and one thing I want to share with the audience is one of the things I really love about both of your groups you know privacy and scaling at the EF PSE as you guys call yourselves in Xerox Park Brian your group is what I've seen is that your groups are both very alike and very different and you kind of have this combination of being alike and different in the right ways so that the whole becomes greater than the sum of the parts which is something that I think has really happened with PSE and with Xerox park I think one of the core things that makes you guys similar is you both have a desire to solve the actual problems that exist as opposed to like settling for superficial or defensible Solutions and I've seen both

of you and your respective groups demonstrate a deep sense of responsibility for how you go about doing things and like a willingness to prioritize kind of the collective good here and so I just want to say that the broader ZK ecosystem is really lucky to have both of your your groups and uh you know on that note let's like dive in on the core topics here so one of the challenging things with ZK is it's like so damn interesting you almost don't know where to start and one way that I like to start in those situations is to ask people what is surprised them like you guys are two experts in many ways so what has been surprising to you over let's say the last three years you know it's been three years since the last edcon it's

been three years since a Dev con that's a long time especially in this space so uh yeah what has been surprising to you and Barry maybe I'll start with you yeah sure so like there's a bunch of things that have come up over the last couple of years that have been surprising the the first major thing that I was like wrong about before was I thought that like the zkbm was gonna be impossible to do and like any kind of practical level and I I think that that was because we there was some like one really nice thing about doing the zkvm project is it kind of forced us to explore the limitations and when we did that we found like a bunch of optimizations that we're now able to apply to over their Primitives so that was surprising and also really nice to

find the other the other major thing that I think was surprising is just the depth of things that that dkps can be applied to that like when I started I thought it was mostly about private applications but it turns out that like as we Explore More we learn about like all the different places where you need zkps in order to enforce like computational correctness as opposed to privacy one example is easy remember there's a bunch of other examples of where you need that yeah I'll jump off of that actually and Echo some of the sentiment from Barry's comments around zkevm I think one of the most surprising things about the applied ZK space to me has been in general uh

first of all how accessible it is for uh especially math and programming generalists and uh you know we found time and time again that applied ZK seems like a really scary sounding topic and I think a lot of the reason for that is because it encapsulates a lot of things it encapsulates things from applications to infrastructure to Theory but when you focus in on one of those layers especially one of these layers that hasn't been as fleshed out for example the application stack you'll find that there's just so much low hanging fruit that actually someone you know with say like an undergraduate math background can quickly you know ramp up into and become really effective at and I think that you know as a consequence of that

the other thing that's been very surprising to me is just how fast things have been moving um I remember like you know three years ago if you were trying to build like a browser-based ZK app versus today if you're building a browser-based EK app first of all the developer tooling is just phenomenally better but second of all even on kind of like naive metrics around holding the same like level of let's say like difficulty in working the stack constant how much better are our approvers today than three years ago it's something like a three orders of magnitude difference and we see examples still across like the entire proving stack of where we can continue to push on uh these tools to be more efficient and even on things like you know sometimes we'll start building a

primitive that seems useful or important without quite understanding like where it might be useful so for example like implementing different cryptographic Primitives inside of ZK circuits we'll do that without even some sense of like oh you know like what is a specific application maybe we have got like one like toy application in mind that only kind of makes sense but once you build out the primitive and you take that to the people who are thinking at the application Level suddenly there's like you know six ten a dozen different things that people are like oh actually that unlocks is totally different thing and you you couldn't have predicted that in advance but it just happens so quickly even in a matter of weeks yeah I think I have to say that it has been remarkable to me to see how fast certain things have moved and this is

even from the perspective of being fortunate to have been around ZK for several years and I'm sure that's exciting for everyone to hear as well because you know we're all looking to build better stuff over time and I think a natural like kind of follow-up question here is just well then how do you see like ZK complementing the future of ethereum so I'll actually chime in here with some like start with the obvious which is you know ZK technology it it fundamentally enables better scalability and better correctness and you know boat scalability and correctness are very important for blockchains uh and of course ethereum specifically and one thing I've noticed is I don't think those are the first things that jump to mind for people when

they hear zero knowledge and that's a bit of a shame because if you look at where ZK technology is being applied a huge amount of it I would say the vast majority is towards things like improving blockchain scalability and improving the correctness of systems whether it's blockchains or or otherwise you know the zkevm stuff that you mentioned and we saw a presentation about earlier with high Chen is is a great example of that so you know we have that kind of as like a baseline for how ZK complements the future of ethereum but I feel like you guys have an even better kind of lens on that so um Brian maybe I'll start with you on that yeah I mean for me I think a lot of the most important ideas here um are the ideas are around

transparency and building something that doesn't you know that that people can't align on but doesn't have to be completely transparent um as I was alluding to in the earlier afternoon talk but this idea that you know in order to have something that we come to consensus on like the thing that we're coming to consensus on must be known by everyone is really really limiting uh and in fact I would say that like you know it's only a small subset of applications that we've seen so far built on blockchain that uh like we we've only seen a small fraction of the possible applications uh for blockchain as a result of just this like fundamental limitation that I think ZK and these these new kinds of cryptography are going to enable yeah very much so I really like the idea

of of like uh scaling consensus in a way that's not like just layer one contestants but it's like Community consensus around things like the other thing the major thing that comes up for me when I think about this is like scaling but also it's this idea of like we had so much like all of the data on ethereum is public and that makes this very rich data set that we can use to make like proofs about different things about us that like Brian gave a really good example earlier about the Ender Chest in Minecraft where you have all of these attestations and you're able to just like kind of compose them in a way that you want and in order to be able to component compose asset stations we need to have ckps and that's that's one of

the really cool ways that the two complement each other yeah definitely like another framing that I have on this is this idea that like um ZK snarks enable you to make proofs that are stateless in some sense so um you know you can prove that you know some secret corresponding to like some password corresponding to a deposit note that is in a specific Merkel route um but you know what Merkel Roots do people care about making those proofs on right like it's a it's an interesting thing a lot of times like when we run educational programs and take people through the idea of like uh you know a cryptographic mixer one common confusion is kind of like well in the zkp can I prove that like the note that I'm

withdrawing is in the evm uh or like is in contained in like the evm state route and the answer is no you can't you can prove that you have some password that's contained in some state route but what is actually the Machinery to make it such that we've got like a source of State routes that everyone one agrees are the important ones and I mean that's almost the definition or purpose of a blockchain like the point of a blockchain is to come to consensus on data that we all think is like significant or important or unambiguous zkps are stateless you need to combine them with something that is stateful for those proofs to actually have meaning to people awesome yeah I think that's a really beautiful

way to kind of think about how these two technologies are just very naturally complementary in ways that we're really still just scratching the surface uh with and you know of course that process is going to continue um so kind of continuing with the sort of theme Here of what to know about ZK and my Approach of like taking different Vantage points one other way I like to look for Vantage points is to play the what I call the overrated underrated game right so what's underrated about something or what's overrated so uh let's start with what's underrated or just something that you think is underrated about ZK uh today barrier throw at you sure so so like one thing that I think is underrated is that like

when you're using like Brian already talked about how groups are really important and how we have this massive data set and like one thing that I think is overrated is that dkps will make people become more like collectivists in their thinking because when you're trying to when you're trying to participate as part of a group you need to think about like the signal to noise ratio of that group that you want to be you want to be communicating from groups with high signal to noise ratio and like in order to do that it people will will start to kind of carefully curate the groups that they participate in and they'll even start to think about like the ways that they communicate affecting the public perception of those groups

that's really interesting actually I'd love to dig in on that a little bit more if you don't mind because so when you talk about groups and sort of uh quote being more collectivists and thinking can you kind of unpack how like ZK works with these groups and and like why it would lead to that sort of approach sure sure so so like basically the way that it works is that like when you're part of uh you're part of this group and every member of the group is able to make signals about is able to make zkbs proving they're part of the group and each time they make a proof they can attach some kind of like message or communication to that so it becomes this fun kind of social media uh application we were able to make ckp's and you're able to prove you're part of

the group then the next logical step is to think that like okay well what's gonna which are gonna be the important groups in this new world and like with some more thinking it it seems clear that like the important groups will be the ones that have the the highest signal to noise ratio that have the least amount of like spam and just nonsense messages and the most the highest level of like just insightful and and thoughtful posts and discussions like and and like because every member of the group wants to be in the groups with this higher signal to noise ratio they want to be in the most prestigious group then every member of the group will start to try and make the group that prestigious and that like high

signal to noise but there's there's also this fun Duality where like while the group while the the nature of groups being shared make them become more collectivist there's also this like strong individualist Trend that's applied from the privacy because when you're probably when you're when you're making posts you're you're private so you're talking from the group but like the actions that you take can't be like attributed to you personally so it's this really fun Duality there wow that's super interesting and something I've really haven't thought about much until now so I'm glad we're having this conversation here and it's something I'm definitely going to be you know thinking a lot more about moving forward um Brian what jumps to mind for you from

the underrated standpoint yeah well maybe the first thing is just to add on to Barry's point because I think it's such a great Point like um we have a handful of of community members in particular who are thinking about like applications of ZK for digital identity um including like uh locksmen ayush sempriti um and Vivek who have been building apps like hey Anon which allows you to prove that you're a member of a group and then anonymously like tweet with the cryptographic proof that that's being attested to by someone in that group and one of the interesting dynamics that I think uh like these folks have been thinking a lot about is that there's almost a like common pool of group credibility that you have to take responsibility for and you know when you're speaking on behalf of the

group you're really uh you're really either tapping in into that pool of credibility and then altering it so I think that that point about the collectivism there is really interesting and it's not something that people are uh thinking that much about quite yet I also think it is something that we do have some history of understanding on like even with the internet itself enabling people to be anonymous and to build these like digital communities that aren't tied to your real world identities there's like a rich literature as well to draw on um maybe the other thing that I'll mention as uh something that I think is underrated is you know this isn't even something specifically about the technology itself but more around like the social environment surrounding it uh while ZK and applied ZK has so many

applications to blockchains and cryptocurrencies unlike blockchains and cryptocurrencies one interesting thing about ZK is it's not really possible to make a bunch of money in a short amount of time and while that seems like kind of a silly thing to bring up it has some really like real impacts on who ends up coming into the space you know so I mean for anyone who's ever like you know sent a tweet that has like the word metamask in it you just see all the Bots and scammers and stuff falling out I think that the ZK space right now is really high signal to noise and that's something that I really value and want to preserve as a result of you know it's it's something that you can't quite get into just to build the equivalent of like a scam coin

I'm glad you brought that up because yeah one of the things about blockchains and I would say it's it's truly one of the best and worst things about them is that they're inherently financialized right like the innovation of a blockchain is really the introduction of incentives to like more classical traditional consensus Theory and incentive is sort of like a scientific sounding way of like saying money or things of that sort and uh it's I found that in the ZK space you know it is really one of the best if not the best places to focus on these like very interesting problems um but problems that are also still grounded in like pragmatism which is something that the blockchain world also forces you to do uh and I think that

combination is it's pretty much it's very hard to find anywhere else even in the rest of the blockchain ecosystem at times um so yeah I'm really glad you brought that up uh maybe now I'm going in the other direction anything you'd call it as overrated perhaps yeah so I think that when we talk about things that are overrated we need to think about it from different perspectives that I think that in the western world uh Financial privacy is something that's that's I think pretty overrated because there's a lot of people focused on that and it becomes this like big discussion in the ecosystem and I think that like there's just a lot of other like super interesting things that we can do that that should have like similar kind of

like discourse around them for example I think I think that like the ways that Western societies work is that like money is one tool but there are a whole bunch of other tools that that decide how our societies work what people think about and and those tools are like well how can people speak how do people get heard who gets hurt how do we decide what to do and like all of those things I think can benefit a lot from TK so so that's from the Western World point of view I think that that's something that's a bit overrated uh but I think that like in the global style that doesn't really that doesn't really carry over because I think that like Financial privacy is just such a huge important primitive to a bunch of people

who are living in like just just in a whole different world where like the most important thing in in the society is not really about discourse and like how we sort of organize together and it's more about just like trying to trying to just deal with a bunch more pragmatic practical matters that's really interesting I love the like kind of the nuanced differentiation of not just the topic but like where it applies on the world I think it's one of the things about blockchains in general as well that make them so interesting and challenging right is that it is a technology that can do very different things for different people in different places and uh one of the things that I know that we all spend a lot of time

thinking about is you know how how do the different things we do affect not just the people that we kind of are most familiar with on a personal level but people in other places in other situations and so really glad you brought that up something I'd love to unpack with you more in a future conversation as well um Ryan how about you anything jumped in mind is quote overrated yeah there's a few things so um I would say that like maybe the way I would frame this is like I think that this is very important like the thing that I'm about to say is very important but that it is oftentimes seen as important or drawing attention at the

expense of other things that are also really interesting and that's the symptom which I mean to say that this is overrated I think scalability currently like drives a lot of the like attention and funding into the space and again just to reiterate like I think scalability is incredibly important and like it has driven so much of like the research and Innovation however I think that the you know privacy or information asymmetry aspect has been undervalued in comparison uh because of something that I honestly think of more as maybe even potentially just like a artifact of of history or business considerations around um you know historically when ZK Tech was very immature when you did have to

in-house build your own proving system and infrastructure from the ground up the only you know business use cases that would really justify that were building out these like entire monolithic systems whether that's a roll-up or uh another like alt zkl1 or things like that those are the only things that can justify that level of investment to like building a cryptography team that implements everything end to end and as a result of that that's driven more that's that's pulled in more of the attention at the expense of for example resources allocated towards these more like modular systems or like interoperable systems that enable you know smaller teams of developers to build out privacy applications

yeah that's a great point I mean I'll never complain about improved scalability and efforts to improve scalability because like you said it's obviously very important but I think this also kind of ties back to something we were talking about earlier which is a lot has changed over the last few years and the ability to work in this more modular fashion is one of the like really consequential byproducts of everything that's improved and so it is I think worth drawing attention to the fact that that can be done a lot more today than a couple of years ago and as a result like you know in comparison that might make some of the things that really dominate a lot of the discourse frankly around like ZK it might make them a bit overrated as a result so really glad you you brought that up

um maybe now let's let's maybe one thing I'll also say is like to be clear I also don't think that the solution is to focus less attention on scalability and more towards privacy like for example you know the work that scroll is doing is absolutely incredible and um in working with PSC on the zke evm uh but rather you know building more of a narrative around the importance of these like uh privacy applications or information asymmetry applications uh and you know getting more people to think in that direction but you know not at the expense necessarily of one or the other absolutely I mean at some level attention is finite and I think that's part of the The Challenge and the fun of being in these spaces that are so interesting and touch on so many

different things is you know we have to constantly calibrate and recalibrate on really every part of the ecosystem that we're in so I think it's an excellent point um now I'd like to I guess zoom out a little bit I'm not sure where this is going to lead but I think it's always fun to ask people for predictions so I hope I'm not putting you on the spot too much here but I'd love to hear a prediction or two that you have for the ZK space we could be it could be something for the next few years could be something for the next few decades um Brian maybe I'll start with you on that yeah this one's interesting um maybe I'll give like a few ideas that I think might be like we might be more likely to see in the

near to medium term and then a couple of thoughts on the longer term um so well I'll start with the longer terms one thing that I think we are starting to uncover is this idea that uh ZK applications might not be a coherent product category um like in the same way that we don't call websites like https or like encryption applications because https and encryption are so ubiquitous or in the same way that we don't call I think you know another example of this misnomer like 20 years ago was this idea of like.com businesses nowadays we talk about things like you know e-commerce businesses or like a landing page for XYZ or uh you know a SAS business with a web front end we don't call these things

like dot coms or dot orgs because the idea of a.com or the idea of using encryption to secure traffic that's going back and forth across the Internet is just like that is water that's Table Steaks I think as we move towards uh this world where we're having a lot more of these like decentralized applications where we have these needs for Privacy Information asymmetry um or uh you know like stronger types of encryption those things are just going to become like table Stakes for the infrastructure that applications are built on top of um so that's maybe the first prediction is that like ZK apps like ZK is not going to feel like a specific technology for a specific category of apps it's going to underlie you know the next

generation of of decentralized applications we see built on the internet um as far as like shorter medium term predictions that might be more actionable I'm personally really excited about a lot of the work that's happening in uh and I'll say specifically instead of ZK identity I'll say ZK enabled digital identity systems kind of you know pushing along that Narrative of the long-term prediction as well as ZK enabled ml Primitives um I think that there has been a surprising amount of progress in the last year in doing things like you know putting machine learning models inside of snarks and figuring out like where are the appropriate applications for running those as well as like Barry mentioned uh both the like creation and discovery of groups and the mechanisms

to be able to prove claims about membership in groups both of those things I think will probably see something that resembles like you know at least some use case that Ordinary People find useful in the next couple of years cool I love that uh Barry what comes to mind for you yeah so when I think about this I think about like just on a on a longer scale time scale I'm thinking about like 20 40 kind of timelines I think that I think that at that time someone will have won a Nobel Peace Prize for something that involves ekps so let me let me give you an example of a potential future here so imagine that

like this kind of like uh back okay so during like International Peace negotiations building trust is a really important part of the process and part of that process involves like showing how many troops you have or how many nuclear weapons you have and like in the future I imagine that like a lot of these components will be to do with like software bugs right I have a zero day in this in this product and like so so like one potential future could be that like as part of these negotiations we both make zero knowledge proves to prove how many zero days we have and then we just we just reveal each of those issues and and manage the patch like some really

important piece of software that's incredible I'm so glad I asked this question now because I was definitely not expecting a prediction like that so um we will check back in 2040 you're not gonna hold you hold you to it but I I do love though the what's more important than the specific prediction right is the underlying thinking here and it's absolutely true like like you said you know trust and building trust and the process towards building trust is fundamental to relationships at every level so whenever you have technology that can play a role in that which blockchains themselves also do it's undoubtedly going to have an impact at some time scale on like how those things

work in the future and I love how you made it kind of concrete with uh with a specific prediction like that because I think that really helps Hammer home this point otherwise it's easy to be like yeah trust and negotiation and so on but uh not that's that's great um well so in the in the remaining time I guess here I just want to hit on a couple of things real quick maybe love to hear a little bit more about how you got involved in whether it's like ethereum or ZK because I think for a lot of people it's like always hard to sort of imagine how they can get into the space and what I always tell people is honestly people kind of just find their way here a bit and uh maybe maybe your stories can help like sort of shed more light on how it might that might happen

um Brian I'll start with you yeah so I guess the uh I mean one of the funny things about coming back to edcon here is that edcon is actually the first like crypto event that I ever attended um this was back like four years ago in 2018 uh and Albert was the one who pulled me into that edcon so it's just fun to you know be back here and uh and chatting on stage about ZK like as the space has grown like immensely um but I think that the uh you know story of how I got involved with ZK stuff was a series of seemingly improbable coincidences I would say that like really highlighted a couple of interesting things that well I guess

I'll explain so basically um in 2019 I happened to go to a conference on zero knowledge like on a whim I found out about it on like a Friday I think Albert told me about it and then on the Saturday was when the conference was starting so I like flew over and then at this conference I like learned that like okay ZK snarks now had like multiple different tool stacks for actually like building end-tent applications with anyway so this was really cool to me just like on a purely intellectual level I was like oh nice like that is I don't know like the progress in the cryptography here is just like interesting so I went home and I happened to then uh walk into a bookstore two days later uh where the three body Trilogy was on display and I

picked up the three body Trilogy and I think I was supposed to be like doing something else that day but I like started reading it and I just like ended up reading the whole book then I came back the next day and I read the Dark Forest cover to cover and I came back the next day and I read death's end cover together so this is this is a Sci-Fi Trilogy by Leo susheen um that's centered around this thought experiment called The Dark Forest uh which is this idea of like you know the game theoretic Dynamics between different space-faring alien civilizations um so I got to thinking about this and I was like oh like you know I like just went to the ZK conference I also just read about this like Dark Forest like you know premise for these like sci-fi books um and I guess actually like one thing

you could do with ZK cryptography is you could make like uh zero knowledge crypto Dark Forest um and then from then you know we sort of realized that like oh you know ZK is like possible to do on this thing called ethereum that I was also like had been thinking about for the past year and also the tool stack just happened to be barely good enough to make the proofs work in browser and so it's just like all these interesting coincidences which really highlighted to me it was almost like you know everything in this space is so interdependent on everything else and it was almost like uh you know we were like getting like a bump in a set and we just had to hit the spike in terms of building this Dark Forest game that ended up spinning up a lot of other interesting experiments

cool thanks for that I think we got you know maybe a minute left here so Barry would just love to hear you uh take us home here sure okay so I got involved oh right so my journey into crypto was like a little bit bumpy so in and I'm like I didn't really know I didn't really know what was happening but here's some things that happened along the way in 2015 I deleted my Facebook account I I was really I I started to become concerned about like the power that people had to like manipulate me and like our discourse into doing different things so I deleted my Facebook in 2015 and like I didn't realize it at the time but after that I felt this like loss you know I I lost

the community that I had been building for years and like I still think about that today I mean I still I still don't know which of the people I went to school with have kids now or not and like anyway that was my that was the first thing that happened I'm like about six months after that I I got really interested in ethereum and I wanted to build like uh a purchasing cooperative for t-shirts and like the idea was people would pool together their money and they would get a bulk discount on on like some T-shirts I'm like I didn't realize it at the time but like both of these things kind of connected into my like search for community that I lost my community on Facebook because I felt like it was

being weaponized against me and I I kind of went out to search for this other other community and like the ZK stuff and the scaling stuff was all just like oh my search for Community we need to do x y and z it's really interesting I actually didn't know that story so I'm again really glad I asked so Brian Berry to the edcon organizers thanks so much for joining me up here today hope everyone enjoyed it thanks Albert thank you Albert