The Race for Private Ethereum | Zac Williamson

David Hoffman:
[0:02] Bankless Nation, I'm joined with Zach Williamson. He is the co-founder of Aztec. Zach, welcome back to Bankless.

Zac Williamson:
[0:07] Thank you so much. It's a pleasure to be back.

David Hoffman:
[0:10] Zach, we're inside of a privacy meta. Aztec has been building in privacy for an extremely long time. I want to ask, is there like an arms race for private smart contracts? So like Zcash is firmly in the lead for private money, you know, like a private store of value. Is there a same, is there a similar fight happening in the world of private smart contracts.

Zac Williamson:
[0:33] Yes i would say there is kind of so There's lots of protocols that are advertising privacy these days, but that is a very nebulous term. That's poorly defined for a reason, because then all these outfits can kind of get away with using the word without really having quality privacy. But just as Zcash really is that private Bitcoin, private digital gold, there is a, I would say, a race to be the private Ethereum, where you have that smart contract capability of a blockchain network where you can define your own assets, but you have strong privacy guarantees as well.

David Hoffman:
[1:12] How do you position Aztec in this race then? Like what about Aztec makes it competitive inside of this arms race?

Zac Williamson:
[1:19] The thing about Aztec is that outside of Zcash, we're the only privacy-preserving blockchain that doesn't have backdoors in them, that doesn't have centralized actors, doesn't have third-party auditors that can control you or censor you. We're the only ones where you can have genuine private smart contracts, where you can have private data, private information, identities, credentials, whatever it is, and it just works out of the box. It's composable, and there's no third-party snooping on you and looking at your transactions.

David Hoffman:
[1:53] That feels like a big claim. So one of the reasons why I like Zcash is it's in the Zcash ethos of you know like no backdoor no compliant privacy no back door so that you know it's private for everyone else but it's public for whoever has this viewing key you're saying there's no other private smart contract or it's based chain no private blockchain that exists in a fully private paradigm that that does not have viewing keys does not have some sort of like trusted authorities ASIC is the only one yes.

Zac Williamson:
[2:24] I'm willing to make that claim If anybody disagrees, you're welcome to come at me. I'm very happy to have a public debate over this.

David Hoffman:
[2:30] Why is that the case? Why is there, if that's, because there's a bunch of private, projects out there that I'm aware of. Why is the seeming like back door, third party viewing key, seemingly the normal equilibrium for everyone else?

Zac Williamson:
[2:45] So there's two reasons. One of them is technologically, it's very hard to do. If your transactions are private, then that means that you're building these super complicated zero-longs proofs that needs to be extracted away from everybody. It adds different loads on your network. And the easiest way to solve this is by just having third-party actors involved to kind of help get those transactions on chain, so it minimizes the demands on your network. If you're relying on those actors, then they now have a responsibility to censor, to monitor, to surveil, because it's not decentralized. And the other one is, honestly, it's path of least resistance. If I want to do a private stablecoin and I want to integrate with, say, Circle, they're going to say, okay, great, put a viewing key into your code so that in order they can see everything and plug it into TRM and do all the compliance monitoring. Otherwise, we're not interested. And so generally, if you want to get integrations with other components in blockchain, you need to work with the infra that already exists. And that infra is used to surveillance and full transparency. We're taking a very different path, but it's the harder road and the one that's less traveled.

David Hoffman:
[4:07] That's what I've known Aztec as, the brand that you guys have, is like you guys have always been doing the hard thing. I think that's also exemplified it. when Aztec launched, it launched as a stage two roll up on Ethereum, or at least very far down the path of being pretty hardened and decentralized from day one. You guys have always kind of done the hard thing.

Zac Williamson:
[4:27] Yeah. Yeah. There was some minor, minor, minor debate about the stage two thing with L2Beats, but yeah, we intended to be stage two. The current version that's live now, it is stage two, you know, so yeah, straight off the top stage two, because for us, it's, it's important, you know, if you're a privacy network, then if you're not fully decentralized, then those, as I said before, those centralized actors have a responsibility to surveil and monitor your network. And we just didn't want to be in that position at all.

David Hoffman:
[4:56] So when you come, when you talk to like institutions and you tell them, hey, like we are fully private and we don't have a viewing key for you. No one has a viewing key. Do you, does that scare them or does that actually interest them? Is that a carrot or a stick?

Zac Williamson:
[5:09] It would scare them if that's what our stick was, but it's different. The Aztec protocol at the network level doesn't have any viewing keys, but we're fully programmable and privacy preserving. So if you as an application, if you're some bank or some tradfire institution, you don't care about decentralization, you're like, yeah, I love censorship. And you just, you want blockchain because it's a more efficient way of settling transactions or because you want access to liquidity. Well, you know, you're in charge. When you're building your app, your code, you can put in a viewing key if you want. You know, Aztec is a neutral network. We don't get to have a say on what you do with it. It's kind of the point.

David Hoffman:
[5:43] So what you're saying is that if they want to have a viewing key, they can have one. They just need to build it into their application layer because it's not being given to them in the protocol layer. It's not on the layer one, but they can have it in their app.

Zac Williamson:
[5:58] Precisely. And I think where we differ from a lot of other chains is we are genuinely focusing on building solutions that don't have these backdoors. Generally, most other apps, they're like, okay, probably USDC, call it A. It's like, well, okay, fine. But A is backdoor and B, you've got a private token, but you're not a privacy preserving network because you don't have this concept of private composability. You don't have this rich way of writing smart contracts, financial instruments, whatever it is, where you can define your own private checks. Maybe it's, I want to do something with identity. Maybe I need to get a credit score from Web2. Maybe I need to check digital passport signatures. Maybe I am doing some weird trademark algorithm that's privacy preserving. And I want other people to use my smart contract as an ingredients in other products, like in a DeFi ecosystem, that facility doesn't exist anywhere else

Zac Williamson:
[6:59] in a way that's remotely decentralized, but it does on Aztec.

David Hoffman:
[7:02] We kind of exist in this equilibrium in crypto where like 99 plus percent of crypto transactions are done on public blockchain ledgers. And I think if you asked me back in like, you know, 2018 through 2022, it's like, hey, David, what will the future of privacy on crypto be? And I'm like, oh yeah, eventually it'll just flip. It'll go from 99% public to 99% private. The public side has been really sticky and we haven't really made too much progress in flipping the equilibrium from public chains to private chains. Where do you think that, do you think that we can make that flip to at least 50-50 anywhere close to that? And what will it take to get there?

Zac Williamson:
[7:43] I think we can make that flip, but I'm not surprised at where things currently are. I've been working on Privacy Texas 2018 team. And, you know, even back then, I knew that individuals don't pay for privacy. If you have two comparable products, but one is spying on them and sending all the data and doing godness what with it, and the other one is like super privacy preserving, the user will pick the one that's cheaper and has the better user experience. And this is very much a tragedy of the comments like situation where people don't really fully understand just how their data is being harvested and weaponized and used against their interests. And so it's hard to fight against that. My thesis has always been that privacy is going to win because it enables products and applications that cannot exist without it. It's not a feature, it's a requirement for a lot of different types of applications, financial instruments and assets that you don't see on public blockchains today. And that's going to be what tips the balance in my opinion.

David Hoffman:
[8:49] What are those applications? What are those applications that require like cypherpunk levels of privacy?

Zac Williamson:
[8:54] Well, I would say it's the majority of TradFi eventually, because if you want to do anything that touches the real world, so let's say you want to trade a real world asset like mortgages or bonds or something like that, then generally you need an understanding of who the holder is. You need some level of compliance because these assets exist in the real world, in a real legal system.

Zac Williamson:
[9:20] And so right now, if you want to tokenize those assets, you make some rather fatal trade-offs because you have to have a data custodian in the mix. You have to have somebody that understands who a user is and what they're doing and what their accounts are and says, okay, fine, you're allowed to trade this asset. But then you can't easily then use those assets in a DeFi ecosystem. I can't build applications and products that use those assets. And so what we've seen over the last year, one of the taglines has been the institutionalization of crypto. So what's really happening is crypto is moving off chain, it's moving into proprietary networks and other L2s where you can perform these kinds of checks because it's not going on a public network that everybody can see.

Zac Williamson:
[10:07] But the downside is, A, it's not happening on real blockchains. But what you sacrifice is composability. You no longer have a world where I've got some novel, new, interesting asset or DeFi primitive, and I can plug that into all these other assets and protocols that other people have built, almost like Lego bricks. This is how DeFi summer originally happened. Instead, you're getting off-chain mega apps, really, where maybe it might be something running on Robinhood chain, or it might be hyperliquid or it might be Polymark. But generally, people will be using one crypto application. They won't be using anything else. And that application is going to be very vertically integrated. And I think this is a fundamental bottleneck on how quickly crypto can grow, how many assets it can access to total liquidity you can get on chain.

Zac Williamson:
[11:02] And I think privacy is the necessary ingredient to really solve this.

David Hoffman:
[11:07] Can you illuminate that just a little bit more? Because I think people, you called like an application on Robinhood as like off-chain activity, but people on Robinhood chain was like, what do you mean? That's on-chain stuff. What's the actual off-chain component that you're talking about? Could you make this a little bit more tangible?

Zac Williamson:
[11:23] Well, I mean, it depends what you mean by chain, right? But it's a pretty, like Robinhood is not a permissive, complete open ecosystem. Them. And generally, if there's a compliance credentials that you need, like KYC that you need to do, those records aren't on chain. So if I want to integrate with those applications as assets, I need a relationship with that data custodian, which is complicated. And if you have multiple different data custodians for different platform services now, I as an app developer if I want to use all these components in my app I now need relationships with all these entities for on behalf of my users it doesn't really scale And, you know, and this value is not being captured by, you know, distributed blockchains, you know, ones like Ethereum, where they're not like operated by relatively centralized actors.

David Hoffman:
[12:19] So it's the data that you're talking about. I think what you're referring to is assets and financial activity that is highly tradfi. Because like when I send you, Zach, Ether on the Ethereum layer one, there's no data custodian there. That's not what you're talking about. you're talking about tokenized stocks that need to do a KYC check or when you're talking about mortgages, there's some provider there. But how does Aztec fix this? Because you guys aren't a data custodian either. And so what about Aztec actually fixes this problem of allowing composability with compliance?

Zac Williamson:
[12:56] It's simple. You put the data on chain and you don't worry about it because it's encrypted and it's private. So the...

David Hoffman:
[13:02] Oh, and you can't put the data on chain in any other context because it's public.

Zac Williamson:
[13:05] Precisely.

David Hoffman:
[13:07] Okay. But how do you actually put data on chain? So like a KYC check, how does that standard actually emerge on a private chain?

Zac Williamson:
[13:15] Well, this is devils in the detail, isn't it? So we have to almost put it together from scratch. You know, we're doing this, World is doing this as well, but generally there's a lot of traditional flows. You can just replicate on chain, you know, like if you're doing a passport check, you can do that on chain. With a, you know, digital signatures in your passports. We already have them with something like ZK Passport, you know, or a lot of, a lot of identities around things like uploading documents, doing document scanning. Yeah, you can do that. You can do that in a smart contract if it's private. So it's certainly not going to be instant, but I think that this is the inevitable future if we want to preserve... That a lot of the value adds that made blockchain successful in the first place. If we want to preserve this concept of composability, which allows for a wide ecosystem of applications to be built upon shared primitives with shared liquidity, and you want access to that liquidity, this is going to be very important. It's going to be important to stop this fragmentation that we're currently seeing

Zac Williamson:
[14:17] in this siloing of liquidity.

David Hoffman:
[14:19] Maybe now's a good moment to talk about the ZK Passport product that comes out of Aztec, so that listeners kind of understand what that is and how that works?

Zac Williamson:
[14:26] Totally. So it's a little app on your phone. So, you know, generally if you have to sign up for something like, you know, relatively important, you might be asked to download some kind of KYC app. What it does is, you know, you tap your passport to your phone and then it takes a picture of your face. You know, you go left, right, blah, blah, blah. And it matches your face to the passport picture and then it validates. Yeah, OK, fine. This is a legit passport. It's got all modern passports have little chips in them where you can get attestations. So you look you read the chip you check that the face match works etc um and great congratulations you've been identified as a human being only problem is generally that works today by you know all of that data including your passport information is sent to some random server somewhere where you don't you have no idea what's going on it could be left unencrypted it could be insecure usually it is usually this stuff gets attacked and hacked and obviously you can't put that on a transparent blockchain. What CK Passport does is it runs all these checks using zero-knowledge privacy tech. So what's leaving your phone isn't your passport data, it's a proof. It's a mathematical proof saying, yeah,

Zac Williamson:
[15:31] The phone's camera took a picture and I matched it to this passport's data. And this passport data has been signed by a specific government. And so I can now prove to you that there is a human being who is a citizen of a country. Or the claim could just be as simple as this person's not on the sanctions list and you don't know anything else. Nothing else leaves your device. And then that can then become a credential inside an Aztec smart contract that can be repeatedly reused and queried by a whole ecosystem of applications.

David Hoffman:
[16:02] Right, and so if you're an institution who is interested in privacy because that's how their current systems work, this is how you can offer them compliance. Or it's like, well, how do I know that North Korea isn't buying my tokenized security or whatever? It's a contract call to the ZK Passport smart contract that has a hook to the web app or the mobile app?

Zac Williamson:
[16:27] Yeah, precisely. or like the mobile app like uploads a credential onto the asset blockchain and then you can just query that inside your smart contract you know that's like that's kind of like that that's the um the the highest quality way of doing this you know there's much simpler solutions as well it could just be simply like you know you you run some off-chain process right and some guy that you trust gives like has a little stamp and go yes okay fine this person is legit and then but then with Aztec you can put that credential on chain and use it without it being linked to any of your your crypto accounts and for restrict like for restricted tradify type activities this is quite important because we're talking about like asset classes or or users where you know portfolios can be extremely large extremely sensitive you don't want people to know about what's being traded or how it's being traded you don't want yourself open to the kind of transparent analytics that exist on chain And you can get that privacy on Aztec.

David Hoffman:
[17:27] So the answer as to like, why can't we do that same system on Ethereum where you do a passport check or a general credential check and you can do it in the public context? The answer is, well, if you do that with enough activity, you can kind of Sudoku puzzle your answer into who that credential is. So the credential doesn't tell you, hi, I'm JP Morgan, but with enough activity on chain, somebody could like kind of just like trace, to figure out with enough data who this credential likely is associated with. And so that's why you can't do it in the public context. Is that right?

Zac Williamson:
[17:59] Precisely. And this also goes back to when I was talking about how privacy wins when it's a requirement and not a feature. For example, right now, David, would you be willing to take your passport and put it up to your camera and show the whole world your passport details?

David Hoffman:
[18:14] I'm not willing to do that.

Zac Williamson:
[18:16] Precisely. That's why it doesn't go on chain today.

David Hoffman:
[18:19] Even though I do have my passport right here. Yeah. Closed. Closed for the listener. Did not see it. Yeah.

Zac Williamson:
[18:27] So privacy is important in some respects. Things like that are important. You know, people don't want their identity to be stolen.

Zac Williamson:
[18:33] And so that flow I talked about, you can't do it on a transparent blockchain. You can do it on Aztec.

David Hoffman:
[18:39] I see, I see. Is there, I suppose there's also a conversation here about there's been just a growing crescendo of AI-driven exploits with a PHP exploit. This topic feels adjacent to what we're talking about now because like, oh, that's how you would probably back into who this credential is in the public context. Yeah. How does the rise of like data breaches or just the rise of AI capacities to exploit vulnerabilities? How does this conversation relate?

Zac Williamson:
[19:08] This is going to change the world and the world is not ready for what's going to be happening with all these AI attacks. What used to, maybe even six months ago, what used to require like very high skilled, highly paid labor to like to find all these bugs and exploits. Now you can spin up millions of little minions that are just as capable or almost as capable that can just scour the web for you for every single little vulnerability and exploit an attack and then find a way of using them and even autonomously executing on these attacks. Possibly even billions of these little minions soon.

Zac Williamson:
[19:44] And it's going to change the web because right now, most internet applications are slop. Either it's AI slop or it's human slop, but very little attention is being paid to security

Zac Williamson:
[19:57] Because there's this principle of obscurity, because who is going to be looking at some random ass website where there's like a server in Utah that's running some old PHP code that's managing, I don't know, like a customer order book of like $10,000. It's not worth anyone's time to look into that. It's worth an AI agent's time to look into that now. And yeah, we're not ready for what's going to happen. The whole trusted operator model of the internet is going to get destroyed. And the idea that you can just trust websites with your sensitive information, with your card details, with your passport, with anything is going to rapidly erode. Blockchains are part of the solution here because blockchains, they've already existed and live in this highly adversarial environment where, you know, if somebody can attack you, it's going to happen. And, you know, as we've seen with the Binance hack, blockchain tech is certainly not immune to these problems, but they are going to become hardened and secure much, much faster than the rest of the internet. We're already seeing that Zcash, their circuits are formally verified. Aztec, we formally verified large parts of our protocol and we're not going to stop until we formally verify everything. So you've got basic mathematical theorems about what this protocol should be doing that you've proven map to the code and said that mathematically can't be any bugs there.

Zac Williamson:
[21:23] This is the environment that blockchains live in. And so we're going to see a move where people are going to stop trusting third party operators and they're going to want to use decentralized, hardened infrastructure, these ledgers where, you know, transactions are completely trustless. You massively minimize the number of third parties and middlemen you need to trust because that's going to be really, really dangerous.

David Hoffman:
[21:48] I've been looking for an articulation about why this rise of AI driven vulnerabilities, discovery of vulnerabilities, but also exploitation of vulnerabilities. It feels like it is a boon to crypto, but I've never really heard it articulated so clearly as what you've said just now is, well, the Internet, the Internet has always been in an adversarial environment. I mean, you know, Facebook, for example, Facebook servers gets attacked. You know, everyone's trying to attack the Facebook servers, and it's been that way for a decade. But that's really like the largest honeypot that's out there. One of them. Until crypto came along and then, you know, Bitcoin and the whole innovation of like decentralized systems exists with the assumption that people are trying to attack it all the time, every moment of the day, every second of the day. And that's been the way that Bitcoin and blockchain systems have been built under. And it's kind of been unique to crypto that we exist in this like inside out world where we have like the normal world, the web two world, where it's just like security. Like I remember Andreas Antonopoulos called this like the bubble boy world where the silos, all the valuable data was contained in this bubble, but they were really fragile. And so if the virus ever got in, they were, it was dead, you were exposed.

David Hoffman:
[23:09] Whereas like Bitcoin is like the sewer rat version where like, yeah, Bitcoin has 10,000 diseases on it in every one given moment, but it's immune because that's the environment that's born in. And what you're saying is like, because the cost to attack these systems has dropped so dramatically low because of the reduction of token costs. Well, actually the web two world now exists. Like the diseased rat is inside the bubble and all of these systems are now inside of our world. And this is a boon to decentralized open source code because this exists because this is the environment that we live in. But the question I have for you is, why is that any better for Aztec than it is, you know, Bitcoin, Ethereum, Solana, any other public ledger? Why does Aztec benefit uniquely more than the public chains?

Zac Williamson:
[23:59] Because it's private. And this is the really, this is the core part. If I want to migrate more of my like Web2 type systems, into something which is a little bit more hard to trust it, Maybe, you know, I want to use payment rails that aren't that aren't using Visa, I'm going to use blockchain and I want to do a lot more checks and queries directly on that blockchain layer because I don't want to trust that part of it. I don't want to trust some rickety server somewhere. I want to trust the blockchain as a, to quote a very overused phrase, the world computer. Well, How do you enforce transparency? Because the internet doesn't run on enforced transparency. We're now talking about a very, very different environment to the regular blockchain environment. We're talking about one where you have proprietary information, you have custom data, you have sensitive user information. All of this information is required to construct and execute these transactions, to create these records. You can't put that stuff on Ethereum. You can't put it on Solana because it's, you know, you'll get sued by 50 million people because you've just put all their sensitive data on a transparent network. It just doesn't work, doesn't compute. But it does compute on Aztec. And that's the difference.

David Hoffman:
[25:11] One of the conversations I've been having with John from Venice is that like one of the best strategies to reduce your company's vulnerability of getting your data exploited is to just not put the data on the Internet in the first place. And it seems like that's kind of similar. It was just like, well, just don't reveal it. And the way that Venice's solution is like, we'll never ever put it on a trusted third-party operator service, Venice being the third-party operator. And what you're saying is like, well, any system with a third-party operator is going to be attacked, exploited. That's what we just talked about. But in addition to that, if you just don't reveal the data at all, you don't have an attackable surface area.

Zac Williamson:
[25:51] Yeah, I would even go further and say it's like, from the application's perspective, it's not their data. If I'm as a user, if I'm putting encrypted credentials, sensitive information is encrypted inside a smart contract and I hold the decryption keys, nobody else holds the decryption keys, right? They don't leave my device. They're not leaving my laptop. Then it's, you know, and I'm using some site like Venice. It's not Venice's data. It's my data. And so it's exactly the same. And this is all about effectively risk arbitrage, where you can use a network like Aztec to de-risk your application and remove like this thorny problem of potentially holding all of this data, which can get hacked

Zac Williamson:
[26:33] and attacked and leaked.

David Hoffman:
[26:35] So, Zach, we've talked, I think, a bit about kind of the macro tailwinds behind privacy and private chains, like the big underlying theses, the fundamentals behind why private chains are valuable. Let's talk about the short term. What is the incentive for the average bankless listener who's a typical retail ledger button presser? What's in it for them? Or for the people who listen to bankless inside of Wall Street and are a part of larger Wall Street institutions, why should they come on to Aztec in 2026?

Zac Williamson:
[27:09] Well, there's only a couple of months left to 2026, but why should they come on in 2026?

David Hoffman:
[27:12] In 2027, yeah.

Zac Williamson:
[27:14] Why should they come on in the next year? So part of it is, you know, there's going to be very little reason why not to. You know, already with apps like CK Money, you can just get privacy for free. You can earn your yield in L1 DeFi the same way you normally do. You just, it's private. And, So it makes you more secure. Right now, we're in a world where everybody's throwing all their data on chain. You know, everybody can see the transactions, trade histories, portfolios, blah, blah, blah. Another danger of AI is it's making it much cheaper to make sense of this data and to categorize and systematize it and actually use it in adversarial ways. And so the less data you're putting out there, the better, the safer you're going to be. But even with, you know, ASIC only launched in April, you know, we're doing a major upgrade which I think will have landed by the time this goes out. It's called Alistair Tierra. But it's still pretty new technology. But even with its novel state, there's still

Zac Williamson:
[28:10] New capabilities that you're going to be able to do on Aztec. You can't get anywhere else beyond things like simple things like private yield, things like private AI inference, where, you know, we're going to have ways of paying for inference without that leaking, without that going all over the internet, which again, I think in this very adversarial environment that we're working in, that's going to be pretty important. And, you know, we're also, we're ramping up what you can do on Aztec. You know, we're going to have private perps, private futures, and we're just growing that ecosystem of giving people, step one is just giving people what they already have, but just making sure that they're more protected than they currently are because people can't see what they're doing. You know, if you're a DGN trader and you're really good at it and people are starting to copy you, well, this protects you from that. In the near term, you know, I'm really bullish on being able to connect as what we're building with Aztec with private identities, things like reputation networks that will allow people to be building and deploying products and services that

Zac Williamson:
[29:21] Go far beyond what you normally get on transparent networks because they're going to be able to know things about you without that data really leaving your control. And it's going to open up a lot of new things, things like, you know, basic things like under collateralized lending and novel forms of DeFi primitives. You know, something that we're going to see in Aztec very soon, I'm pretty hopeful about, is a fully indulgenous private stablecoin with private issuance and settlement, things like that. So we're just going to be opening up Aztec to just a world of new capabilities that hasn't been seen before.

David Hoffman:
[29:58] When you think about onboarding new users, new transactions, new assets, Do you think that the idea here is to bring on a lot of existing crypto users? Or are you also trying to go to the world of Wall Street and be like, hey, I know you've never touched a blockchain before, but we actually have the solutions to the reasons why you haven't? Is it net new users or existing users?

Zac Williamson:
[30:20] We're doing a phase draw. Right now,

Zac Williamson:
[30:24] We're focusing on Web3 users because they already understand the blockchain paradigm and it's much easier to communicate with them. We want to focus a lot more on hardening our network and making it a lot more turnkey before we start going to wall street you know like i said we formally verified major parts of our protocol i want to formally verify everything i don't want you know because when i'm talking to somebody who has very little literacy in blockchain they have zero risk appetite to like to like even the idea of having a private key is like scary i want a solution to them which is this turnkey and white label where they can get their team, they'll have some weird derivative or trading policy or whatever they want to do, right? Like they can just get their team. It's just not blockchain native. They're not, they're not like, they don't care. They don't know anything about privacy. It just works out of the box and it's secure. I've been in this industry for many, many years and doing privacy, doing the same thing because I'm not, I try not to be a bullshit merchant and I don't want to make claims that that can't be backed up. A technology like Aztec, I think, is unbelievably compelling and powerful for institutions and enterprise, but it needs to conform to their expectations, to traditional expectations, not to web-free expectations, which means we have a little bit more work to do there, a few more months of work to do before it's fully ready for that. But, you know, it's only a matter of time. Everything is falling into shape with our formal verification work.

David Hoffman:
[31:51] I do think that as exploits, it kind of feels like just the world is exploding as we are walking through it, or at least that's what we're going into, according to the AI researchers, which everyone seems to agree with. The formal verification is going to move from just like something aspirational to like something urgent.

Zac Williamson:
[32:07] Oh, it's already urgent. People don't realize it yet, but it is so incredibly urgent.

David Hoffman:
[32:12] I want to ask about the network, the Aztec network, like the economics of the network. One of the interesting things about Aztec is that you guys launched stage two, like fully decentralized. What that comes with is that the Aztec token is a proof of stake token in the same way that you would expect a layer one proof of stake token. It's just like a layer two on Ethereum. But because privacy is more, as I understand it, computationally intensive, how does just the economics of transactions on Aztec and the Aztec staking token for validating the network, is there anything different there because of the privacy? Or how do people think about this in the Aztec context versus what is kind of the status quo?

Zac Williamson:
[32:53] Yeah, so honestly, from the tokenomics side and from the staking side, it's pretty straightforward. Not much changes. You know, we have a token which powers the network. The Aztec token underpins everything on the network. All activity requires Aztec for gas. And we get the network security through staking Aztec if you want to be a validator. And all the governance decisions go through Aztec. When it comes to the economics of like operating the Aztec network, right now, the privacy requirements, they add a small amount of overhead because you need these prover entities to improved blocks, which you don't normally have on a network, but their computation requirements are pretty light. We've just recently made a governance change to fix what was effectively just more of an We hadn't quite got the economics right, so we were overpaying our provis. So that's now been fixed. Over the next year, we're going to be making some relatively minor changes to make proving much, much cheaper. So, yeah, I think the additional overheads of serving privacy from a network operator perspective are basically going to trend down to zero very quickly.

David Hoffman:
[34:08] What about just kind of the network when it comes to UX? for example. Is it fast? What was it like to trade tokens on Aztec? Is there MEV on Aztec? What about all of these kind of things?

Zac Williamson:
[34:21] Yes, there can be MEV on Aztec because typically, if you're doing trading on Aztec, you want to keep identities private, but the actual protocol needs to be somewhat public for price discovery and that does a lot for MEV. There are solutions around that using a multiple party computation and that we very much are exploring, but that's kind of a next level thing. When it comes to things like latency, transaction speed, that's where the privacy right now has a little bit of an overhead. On my laptop, for example, I need to generate this transaction proof because the sensitive data is not leaving my computer. Only the proof is. And that takes roughly five seconds in a web browser on my machine. It takes one second if I'm running a native app. So that is one of the costs. We're working to get that down. We've got an engineering roadmap to squeeze that down. There's a lot of low-hanging fruit, particularly opened up through AI.

Zac Williamson:
[35:21] So, yeah, there's going to be progress there.

David Hoffman:
[35:23] What are you guys' goals? What are you guys working on in the short term? So you guys are launching Aztec v6, Aztec Tiara. Talk about the upgrades that Tiara is bringing. And then once those are up and running in the network, what are you guys working on next?

Zac Williamson:
[35:37] So as a Tiara, it gives a lot of security. It gives the upgrades, right? Lots more formal verification. We've run AI audits on our code 24-7 for months now, like throwing the best models at it. And we have halved our proving time. So basically an over 100% increase in the performance of our tech. We have much faster transaction confirmations. I guess it gives a soft finalacy where a sequencer is checkpointed onto Ethereum saying this transaction is going in the block. That's now between 6 and 12 seconds, I believe. So pretty Ethereum-like, which for privacy transactions, that's quite impressive. That was very hard to do, but we've done it. So those are the main things from the protocol side. Faster transactions, less load on your computer to make them.

David Hoffman:
[36:34] Yeah, and then what's the roadmap next? What are the things that you guys have left to optimize? What do you guys think, the work you guys are focused on for 2027?

Zac Williamson:
[36:42] Yes, the roadmap we're focusing on is, At the protocol level, a lot of it is security, security, security, security. We live in the most hostile adversarial environment you could possibly imagine, far worse than anything we've ever seen before because of AI. So get the thing formally verified. Make sure that it's completely ironclad and reliable. While we're doing that, you know, we want to do things like we want to enable faster withdrawals. We want to improve the, again, prover speed. There's still a lot of soft engineering we can do there. We want to make it cheaper to run a prover and basically get to real-time proving so that we can produce our blocks much faster. We want to improve hard finality from right now it's 40 minutes we want to get down to five minutes where basically you um you have a uh like the actual like a full asset lock proved and settled on chain on l1 that's basically the time limit for how long withdrawal you have to wait right now it's 40 minutes we're going to get it down to five minutes but then include a effectively a tipping system so as a user if i want a fast immediate withdrawal i can pay a little bit extra like 10 or 20 cents to get that proof made instantly and put on chain.

Zac Williamson:
[37:56] So, yeah, from the protocol perspective, that's what we're focusing on right now.

David Hoffman:
[37:59] Do you have any main CTAs or just a request for the listener? If they want to come on to Aztec, what should they do first? What activities do you have for them?

Zac Williamson:
[38:09] Go to zk.money and give that a whirl. By the time this podcast goes out, we should be also having quite a few other applications you can try. There's something called P-Inference for private inference. Give that a whirl. Pay for your AI privately and have it secure. Run a secure enclave so that you don't need to worry about all your secret information being leaked. We have a native web wallet called Nolo, give that a go. If you just want a MetaMask experience, just get to transfer and trade around tokens privately. We've got Nix, which is going to be private DeFi, private Yield. So Nix.xyz is another one. Man, I'm probably playing favorites here because there's probably a few I've not mentioned yet, which I should be. But yeah, just go to ASIC.network. That's where where we're going to be broadcasting what you can do with Aztec, why privacy is valuable. It's going to be describing the tokenomics in more detail as well, if you're curious about that.

David Hoffman:
[39:04] Zach, I do hope that this is the year, maybe this year or 2027 per your point, that we get off of 99% public transactions and hopefully we can get to 9010 at least by the end of 2027 and hopefully it goes far beyond that. So I wish you guys the best. You guys are really moving the ball forward. I do really like the point that if this is true, that Aztec is the only smart contract chain that has no viewing keys that kind of puts you in a league above everyone else.

Zac Williamson:
[39:32] Private smart contract with no viewing keys. Ask your AI. It'll agree with me.

David Hoffman:
[39:35] Yeah, I believe it. I believe it. Best of luck, Zach. Hope you guys get all the success you guys deserve. You guys have been working on Aztec as long as I've been in crypto. And so I think now's the time.

Zac Williamson:
[39:47] Now is the time. Thanks a lot, David. Thanks for having me on. It's been a pleasure.

David Hoffman:
[39:49] Bankless Nation, you guys know the deal. Crypto is risky, but if you make it private, it's a little bit less. You can lose what you put in. It's the frontier. It's not for everyone, but we are glad you're with us on The Bankless Journey.