Justin Drake on Ethereum's ZK Future

Transcript

the last 10x, you know, might require this collaboration to go from a 100 mega gas per second to one one gigas. And we have seen from partners like Taigo that this is incredibly difficult to do in production. So really trying to make it, you know, out of the box to where these popular rollup stacks um can do this multi-proof system as soon as possible. How are we making Ethereum an institutional friendly protocol? Welcome back to the rollup. This is your home to learn from the biggest guests, the brightest minds at the tip of innovation. Our goal is to help spread the collective knowledge of everyone in our community.

We are breaking crypto out of its echo chamber and taking it to the real world. All righty, Justin, welcome to the show. Good to have you on. Um, and it's been a while. Yeah, long time. Yeah thank thanks for having me. Thanks for joining us, man. Absolutely. Um, so Justin, I want to kick you off with a question. I've been around in the ZK space. I've been like a novice. I did some research for about like four years now and I remember your name from these early like ZK days. I've I saw you in in group chats snooping around um ZK tech and whatnot.

I want to know like sort of what got you so bullish on ZK in the first place and then clearly I think over the last 12 months you've been going pretty hard and advocating for this technology. what sort of brought you back in into the fold of being bullish on ZK? Yeah, so I got into ZK in the days of of Zcash. I looked it up. Zcash was launched uh in October 2016 and just a few months later uh in mid 2017 I I wrote this article talking about Z Ethereum which was basically taking Ethereum adding all of the ZK tech and hence the name Z as a hat tip for for Zcash. Um and you know one of the ideas that I had was uh let's just sync the

whole history by verifying one snark. Uh and then another idea was let's just have one single up code this universal upcode that would verify a snark and that would be flexible enough for for all of the smart contracts. Um, when I wrote this this article, I sent it to Ellie Ben Sasson and uh he invited me to a conference in Tel Aviv that that he organized and it was kind of this surreal experience where one thing led to another and on day two or something I was having dinner with Zuko, Vitalik, and Ellie and a bunch of other people who uh later became founders of of uh of stockware.

Um and that was even before I joined Def Foundation that was just lurking around and part of my curiosity was peaked because I just found it so mind-blowing that you could take computation uh just prove it in a very succinct uh proof but also kind of hide the hide the inputs and have and have privacy. Um and then uh to a large extent what happened is that I I joined the foundation and worked on all sorts of other problems unrelated to snarks because the technology was not mature enough but I think we're now reaching a point where finally the technology kind of eight years later is mature enough and vitalic had this um item in his in

his road map with the merge the splurge the verge etc. One of which was um let's snifify everything. Snockify all of the things and basically snarks solve a lot of remaining open questions. So one of the big ones that I'm working on right now um snarkifying the L1 EVM that gives us two things. One, it allows us to dramatically increase the gas limit. I talk about the gigas frontier which is one gigas per second on the L1 EVM. It also really helps with decentralization.

The valid data sets can now be verifying proofs of validity uh on much smaller and weaker devices that they have today. So the the meme that I'm pushing forward is the Raspberry Pi Pico uh which is a a $7 piece of hardware um an order of magnitude cheaper uh than the Raspberry Pi. And that should be sufficient to participate as a validator and consensus participant uh on Ethereum. Um going back to the roots of Zcash, there's actually this very exciting um thing called a a privacy wormhole. Um so the idea here is you have privacy at L1 where the way that you you you opt in is by sending funds from an Ethereum

address to another Ethereum address which is provably unspendable. So from the point of view of the blockchain, it looks like a normal transaction, but actually you've just in some sense destroyed the balance associated. And the way that you would um have your your your funds um be recreated is by submitting a proof of burn. So proof that your funds were sent to an nonspendable address. Um and that that could allow us to have privacy uh at L1.

Um, snarks also help us with composability. I've I've been talking about this idea of the united chains of Ethereum where instead of having fragmented and and siloed L2s, we can have them all come together even with the strongest form of composibility which is known as synchronous composability. Um, that's something that we're going to see more and more as the tech uh matures. And another very exciting thing I'm excited about is this idea of removing the security councils for the for the L2s. One of the big reasons why we have security councils and governance is because um there first of all there's there's there's bugs in in in the

software. It's very complicated. And two, if you want to remain EVM equivalent, you need to have the governance to stay in sync with the L1 EVM, which is itself an evolving beast. Um, so there's this notion of a native rollup. Um, and we're going to have vast amounts of of L1 gas, which the L2s are going to be able to reuse. Uh, and by construction is going to be uh bug free uh and EVM equivalent. So you won't need the security councils and you won't need uh the governance. Um and then another thing that I'm very excited about just final point uh is postquantum security.

So today if you look at L1 we have various places where we use non-postquantum secure cryptography specifically BLS for the aggregation of signatures and KZG for the blobs. And both of these primitives can be replaced by SNOX which are postquantum uh secure. Introducing extended exclusively on Stark, a nextgen ps trading platform live today with 50 plus trading pairs from crypto to tradi. You get the best experience on chain today. Deposit in the community vault for attractive returns compatible and engage in unified margin spot trading lending and per trading exclusively live at

extended.app. app on Starknet today. Justin, when you're when you're thinking about the uh implementation of some of the uh actual details and these uh actual mechanics for Ethereum L1 and I kind of started this before you came up speaking about how it was great that the Ethereum Foundation has really kind of taken this approach of of using the the amazing work that some of the teams like Boundless and others have done. when you're thinking about actually implementing this this work, um how do you how do you envision the collaboration between the Ethereum Foundation and teams like Boundless to actually get these this type of uh you know progress forward to help the Ethereum road map, you know, accelerate.

Everyone always wants faster. You you know that better than anyone else, but just generally how do you view this kind of collaborative effort actually unfolding? Yeah. So all of these teams that are external to DF foundation are helping in two ways. First of all, in terms of performance, they're constantly coming up with new innovations and they're learning from each other as well, which means that they're lifting the all boats uh and moving at an extreme pace, you know, roughly, you know, 5 to 10x per year of of performance improvement. And that is absolutely fantastic. And it's it's something that, you know, deaf could not even dream of or or even think of of paying for just a few years ago

and now we're in some sense getting it for free. Uh you know paid by the VCs in some sense. Um the other amazing things about um the diversity of the ZKVMs is security. So these pieces of software are complex. We we are assuming that each individual ZKVM has critical bugs, has vulnerabilities. And the way we hedge against that is in the exact same way that we have client diversity for the consensus layer clients and the execution layer clients. And the the great news is that we actually have more diversity at the ZKVMs than we do for the EL and and CL clients, which is kind of remarkable because you'd think that a

ZKVM is a much more complicated beast than a CL or or an EL client. So in this very luxurious position, um yeah, does that answer the question? Yeah, I think it does. I wanted to Yeah, I wanted to plus one on ZKVM diversity. I think we've realized this as well like you know working zero network to boundless that we just need many ZKVMs to make these systems you know practable to secure tens of billions of dollars in production and something we're really focused on at boundless I hope to you know soon have an offering to deploy a rollup with three ZKVMs soon all at the same time we have seen from partners like Taigo that this is incredibly

difficult to do in production so really trying to make it you know out of the box to where these popular roll-ups stacks um can do this multi-proof system as soon as possible. Jacob, you mentioned, you know, $10 billion. I'm thinking 10 trillion dollars. So on Ethereum today, we already have a trillion dollars of assets if you count ETH and all of the other assets. Um yeah, we're planning for much bigger numbers. Yeah, I I agree. I do have this internal saying. It's brick by brick, right? We're going to hit one milestone. You know, we hit, you know, onetenth of a cent to prove an ETH transaction. I think we're about 1 billion TVL right now securing ZTEC ZK tech, you know, let's build it up and compound that success.

Yeah. And in the article that I wrote eight years ago, one of the lines were was in the last few years, we've seen incredible orders of magnitude improvements in performance. So even back then, we were on this exponential curve. Um, and when you think back then, a single Zcash transaction would take minutes to even produce. And now we're at a point where in a matter of seconds we can prove entire blocks and as you said it costs about 1 cent where an individual transaction is either a tenth or hundth of a scent. Yeah, it's it's really incredible to work alongside these other teams. Um, and sort of adop, you know, adopt this risk five standard that we originally came up and just see this compounded success. Um, seen incredible numbers

recently from, for example, uh, ZISK, which is the team, you know, Jordi Bina is leading right now. They've given a over a 20x improvement for everyone's executor and they've open sourced it from the beginning, which is just incredible for the community. And then super excited about air bender as well um where they they're posting you know industryleading numbers right now on top of you know I think Chris Zero is cooking some things as well. Another thing that the ZKVM teams have been very useful at is illuminating the right path forward from an a design and architectures perspective especially the the ISA and you're right risk zero was literally the pioneer uh that brought risk 5 and now it's an industry standard and from our perspective you know we

want to be picking the simplest the most credibly neutral tech um and we don't want to be making the these hard choices but now you know what you guys pioneered is an industry standard and it makes the decision so much easier. Justin, I'm curious from from the Ethereum roadmap perspective, what is the what is the next kind of big milestone that you have in mind that we need ZK for like in order to get to why you know we need ZK and that's that that is a big thing that uh is very impactful for the protocol. Uh you know you know it could be at any part of the of the protocol. curious like what is the big thing that we're waiting for that will be unlocked by you know the work uh from

teams like boundless and others. Yeah. So I think at this point people are familiar with the gig gas frontier and you know bringing more decentralization. I think two things which are new is one this idea of the privacy wormholes like for a very very long time you know Ethereum was saying just do privacy at higher levels of the stack and I think the technology is becoming mature enough that we can have full endto-end verification of these systems so that we have very high confidence that there's literally zero bugs in those systems. If you rewind the clock and uh look at at Zcash a few years ago, they had this critical bug

which allowed anyone to mint the Zcash token and and be undetected. Um and that is extremely scary for Epha, the asset which is trying to be a monetary asset at global scale. And so we need to have this high conviction on on security. And on that point um one of the next big frontiers of kind of ZKVM technology is going to be simplicity like Mogarov complexity like how many lines of code is your whole thing um and you know how aminable is it to formal uh verification because right now we're leaning on client diversity but in the future

there's definitely a possible uh outcome which is that we pick a single ZKVM we end to end and for formally verify it and and that's the one um and instead of diversity we lean on formalification will that be boundless will that be others we will see but I I appreciate that answer curious um Justin as well uh when you're when you're thinking about um the the recent let's just say like shift in mentality with within the Ethereum Foundation as well as just broadly with the sentiment of the community. Hey, nice to meet you.

We've seen obviously institutional adoption of Ethereum. We've seen a lot of these players want to come in and tokenize assets on Ethereum. Uh Fidelity recently uh just, you know, came in is now looking to challenge Black Rockck potentially with their uh US treasuries on chain still muted. And you've mentioned privacy multiple times in this call and I think this has been something that you've been prioritizing, you know, mo more recently as you've said. uh you know deeper in the actual Ethereum protocol versus kind of either outsourcing it or putting it at a higher layer of the stack. when you're thinking about the, you know, the institutional adoption of the Ethereum protocol, what are some of the more important technical lifts that are on

the agenda on the road map that you think are particularly appealing for institutions who would like to allocate to buy ETH knowing that these are coming or would like to tokenize assets on ETH or on Ethereum or, you know, are doing things with stable coins, etc. when you're when you're looking holistically from the institutional perspective of using Ethereum, what are some of the most important updates? And you know, we may have covered them u but you know, I'm just curious what are some of the specifics that you think are appealing to large capital allocators who operate, you know, within their own set of regulatory environment, within their own set of compliance, etc. Uh how how how are we making Ethereum an institutional friendly protocol?

Yes. So you're right. Um there's been a flurry of institutions that have started to look in into Ethereum and one of the the common tropes I guess is that uh privacy is a a real pain point and at this point no one has has solved it. But if I were to say something about Ethereum is that at least it's trying to solve the problem and it's trying much harder now than it was a year ago. And we also have pioneers like like Aztec that have had this large amount of conviction and have been you know building for for over half a decade now.

In terms of deaf putting more of an emphasis we uh just announced a rebrand of PSSE. So it used to be privacy and scaling explorations. Now it's privacy stewards uh of Ethereum. And so we're just focusing only on privacy. So that's a team uh of um of roughly 50 people and um there's going to be some some some big announcements related to PSC. Uh possibly uh some changes to leadership. Uh we're also going to see much more transparency. So right now the whole Ethereum Foundation is on the ORC chart with the notable exception of of PSSE.

We're going to have an orc chart for for PSSE and we're going to see all of the different sub projects and I think um the this cluster of people of 50 people is is going to be shifting gears um and um and I'm I'm I'm hoping for very good things to to come up first of all in the short term and be announced at Defconnect but also looking forward uh in in in 2026. I want to I want to double tap on that is like I'm super excited about anything there because the reality is that ZK is not going to stop at like scaling blockchains. It is this fundamental gamechanging privacy technology for the rest of the world where I truly believe

that you know we saw this shift from like if we remember HTTP to HTTPS that allowed you know online payments to happen allowed more secure um business to happen on the internet. I think we're headed for HTTP ZK. Um, and it's going to be just a safer world for everyone to live on the internet. I really view ZK as this counterbalance to AI the same way you saw like modern key pair cryptography be a counterbalance to this sort of big data movement in the early 2000s. Absolutely. Yeah. HTTPZ is coming. Um and from a tying perspective I think it's perfect because you know now we have you know the nation states the the

regulators the institutions comfortable with blockchains and I think to a large extent it's because of the lack of privacy. It's because we had full transparency uh that we were able to get our foot into the door and it's a little bit like a Trojan horse and now that in some sense we're comfortable in our position uh we can uh start pushing the boundaries with HTTPC. Yeah. And like a little insight here is you know many of our team members they go to open source cryptography conferences and they go to internet standard conferences and when they go to these conferences these are the same conferences that you know the early ideas of HTTPS and and modern key cryptography have happened people are describing the properties of ZK in the

next generation system they just don't realize that this technology is coming along um so fast and you know you see big players like Google starting to use it you see big players like Facebook and Microsoft have these giant ZK research teams. Um I think you know Microsoft has a research team of over 100 plus people at this moment. Um the technology is is coming for the internet. Another topic that institutions are interested in is uh quantum like they are extremely risk averse and the I basically have a meeting with an institution every week to talk about postquantum and the great thing here is that uh Ethereum is arguably like the most advanced blockchain pushing forward

postquantum cryptography. One of the things that we're doing is looking to replace BLS, which allows for a million validator signatures to be aggregated in one single epoch with a postquantum equivalent. And here we're actually using the snark machinery to do the aggregation of hashbased signatures. Um, and we're in a very privileged position where effectively the vast majority, let's say 90% of the ZKVMs have hashbased snarks at their core and we're able to reuse that technology and enshrine it uh at the filament one and this is part of the lean consensus uh effort that that we're working on right

now. Justin, one last thing for you. Uh, and I appreciate the the the institutional focus and just like this this kind of framing. It's it's an interesting paradox where, you know, we've built a lot of this open source technology, you know, in an industrywide perspective. And people are arguing, you know, the how institutions are going to come use it, whether they're going to extract all of the value from it, some of the value from it, none of the value from it, how they're going to be kind of planning about it. Regardless, uh, everyone is very clear that everyone needs an institutional strategy. We've we've al we've obviously got Etherealize which has been amazing. Um and we've al you know and we've seen just you know guys like Tom Lee and uh you know Joe Luben get out there and put their necks out publicly for everybody. Um so that's

been great. And just to kind of wrap this up it's obviously boundless mainet day. Uh you know I'm not asking you to tell us that boundless is the best thing since sliced bread but I'm going to ask you what excites you about the boundless protocol? Why why are you as a as a deep Ethereum researcher excited about the protocol? Absolutely. So, um, one thing that I'm excited about these these these proven marketplaces is the opportunity for them to become collaborative in addition to competitive. So right now the marketplaces is basically an auction house uh where we can have the most efficient the most you know economically efficient uh provers uh win out and

compete against the less economically efficient ones. What I'm hoping for, and I believe this is something that Boundless uh has been working on for a while now, is this idea of collaborative marketplaces where you have multiple provers that are working together towards the exact same uh proof. And the reason why I'm excited about this collaborative aspect is it will allow us to dramatically grow the the gas limit. you know, the last the last 10x, you know, might require this collaboration to go from a 100 mega gas per second to one one gigas. And the reason we might need this collaboration is because for decentralization, we're putting this hard constraint on the uh prover power

draw. So, we're limiting it to 10 kilowatts. That's the equivalent of a home Tesla uh charger. And um you know it allows us to have let's say 16 uh GPUs. But if if to get to one gigas per second requires let's say um 64 GPUs then we might need four such uh clusters to work together to to generate the proof. Um, and so just selfishly from the perspective of FML1, I'm hoping that we're going to see more of these these collaborative aspects.

Justin Drake, our our local uh Bitcoin security researcher in the house. Uh, appreciate your your time today, man, and uh looking forward to uh more of your talks and just more of the development from the Ethereum side. Um, and thanks for being a part of this this awesome day to push Ethereum's ZK efforts forward with Boundless and many of the other ZKVMs. Uh, it's always a pleasure to have you on and look forward to having you back. Absolutely. And congratulations, Jacob. Thank you. All righty, Justin. Speak zoom in.