Private Transactions on Ethereum

Transcript

[music] [music] This is going to be a great discussion. This is the biggest topic um with respect to institutional adoption. So here we go. Um the closers uh of the day and probably one of the most important topics. So thanks guys. >> Hey. Hey. Thanks for having us. Uh glad to be here. Um, >> yeah, thanks for having us. I'm Jeremy, one of the co-founders of Aztec. I think we're just going to do a back and forth chat about the problems of privacy. Um, and there are a few. So, do you want to kick us off?

>> All right. Uh, so I'm working on a privacy wallet. Uh, we decided a few months ago with Vitalic to launch this uh wallet project. Um, we realized that wallets are the main tool with for people to interact with the blockchain and they lack privacy features. So you have a lot of um important information when you interact with the blockchain that you leak through regular wallet. You leak your IP, you leak your positions, you link your holdings and you leak your uh intents, you leak your strategy.

So it is very nice for transparency and like auditability to be able to like check everything, but it's also a big problem for um institutions for example. um no asset managers want to reveal their strategy. It would basically be like giving your edge on the market. And so we decided to tackle this problem and we faced a lot of challenges doing this and uh actually has faced uh the same uh challenges uh because they're building this natively.

Um so yeah, glad to have you here to to discuss that today. Yeah, maybe I can talk a little bit about our our history because we we didn't set out to build a a privacy network. We we set out seven and a bit years ago to put real world [snorts] assets on chain. Um we were trying to put corporate debt, big terminal A's, terminal B's on Ethereum. And we realized that, you know, this is just not going to happen without strong privacy guarantees. Um so yeah, as Nico was saying, like if you're a bank and you can see my portfolio, you see how much of an asset I'm going to sell when I start selling. it just doesn't work.

And so we started building kind of you know privacy tools back in you know 2018 um early kind of zk snark technologies um and we started experimenting with that and over the way we've we've built some um kind of innovations a lot of the kind of cryptography that is now used for zk technology Zach uh my co-founder built um and invented so there's a lot of challenges with privacy um I think the main one is that the state model is different Um, so maybe we could talk about UTXOs and uh I don't know how you guys are are solving that in the wallet but yeah what's the the state model for you guys? So we have to use the account model because it's going to be on

Ethereum. Uh but we will do it in a UTXO way. So basically one account per uh DAP, one account per interaction. Uh so this means like every time you will connect to to a DAP you will use a new Ethereum address. Uh so basically we're going backward back to uh to Bitcoin in the UTXO system. Um because it's simply like better for for privacy. Uh you don't want like it's it's really hard to imagine an account model uh that is fully private because you you link your your past actions. So you link metadata metadata and you can do some correlation analysis if you do this. Uh so yeah we went back to the UTXO model and one of

the challenge uh that we faced like was when you run a transaction on Ethereum you want to know what's going to happen uh you want to fetch some account balances you want to see uh what's happening on the chain and currently when you do that on Ethereum you do RPC calls and there is this very nice uh RPC provider somewhere somewhere on earth uh that gets this information about what you're going to do what you want to check what what you're reading on the chain. Um and so we faced um this challenge and we were like okay how are we going to do this in a private way and we ended up deciding that we need a small execution client within the wallet that's going to do it uh internally and have some private interactions with some

servers. Um and this is very similar to actually the private execution environment model that you have on Aztec. Um so yeah uh I realized [snorts] uh checking Aztec docs uh earlier today to to prepare this that we we we have a lot in in common on that front. >> Yeah. So we we kind of discovered this. We we used to have a product called Aztec connect. Um it was like a a payments roll up that also let you interact with uh DeFi on Ethereum. And I guess the problem that we're we're talking about here is if if you can ask an RPC node what's my balance then the entire world can see your balance. And so like there's a fundamental like problem with public blockchains which is like if everyone can see everything then

they can see it. Um so um when you try and hide that um you end up with like almost a harder problem which is you now need to query all the state on the blockchain to figure out which state's yours. Um so in order to not leak any information, you know, you have to test every piece of state and say is this mine? Is this mine? And that doesn't really scale. And so you end up having to build what we think is like we call it the private execution client. It's like a light client that sits on your device. And it knows how to ask for the right pieces of state uh to make this actually run on your phone and be be feasible. So it's taken a few years to kind of get the mental models right, but I it's validating to hear you come to

the the same conclusions, I think. >> Yeah. I wish you had the the EVM so I could just copy paste your work and and save a lot of time. >> You can use Noah. It's open source. >> It is. >> So um one big topic these days is like the institutions coming and um talking with Ash yesterday I realized like this is privacy is the link between the two edges of our community. You have like the cipher punk in one extreme side of the community that have like very strong privacy needs and you have this institutions that are on the extreme other side of the community that have the same needs. Um so really interesting

to um to be in that moment where uh we have privacy uh taking the the main stage and yeah I really hope that Aztec can uh be that layer two that that will be like fully private and enable uh fast cheap interactions fully privately. Um, so question for you. What are the next uh ZK money uh type of experiment? Are you going to do this again? Uh, when am I going to be able to do deny position without my friends knowing that I got liquidated?

>> Yeah, it's a good question. So, so we've been kind of building this for many years now. We're now we have a fully decentralized test net and um mainet's coming very soon. um that network will launch um with kind of inbuilt bridging functionality. So you can as a user take a transaction on Aztec and bridge it to any kind of EVM chain all L2s or or mainet. Um so we kind of think of it as a bit like a privacy layer for for Ethereum. Um you obviously have to get an Aztec account to to use it. Um but we're working on ways to kind of make that easier. Maybe connect your Ethereum like L1 wallet. Um, but I think we need some standards here to help. Um, at the moment we have like the ERC20 standard,

but there's no like private token standard uh on Ethereum. In fact, I think one of the early EIPs uh was written by Zach and it says confidential token standard and it was written in 2018. Um, so maybe we can revive that and then we can actually get some some real privacy on Ethereum. But yeah, early next year you'll be able to kind of recreate Aztec Connect again and make kind of like a crosschain bridge uh to to all chains. Maybe the next thing we could chat about uh we got like 5 minutes left is like privacy sets um and how you think about um what is the privacy set for uh an asset like if you have a if you're an institution right and you want to do a trade that's $50 million $100 million how do you hide that in a in a public ledger have you

guys started to think about the privacy set for the wallet >> yeah so we started to look into like existing uh privacy protocols on Ethereum And we are still facing this this problem that like you can't hide on Ethereum like the the state is public and the balances are public. So in a smart contract that is like a privacy protocol you have a certain amount in this contract and currently it's not big enough to like take what you what you just said like if I come into privacy pools with like $100 million I'm probably going to be bigger than the entire pool. So like it's really hard to to hide my intentions like if I if I

withdraw from the pool then you will know yeah it was me doing doing all those interactions. Um so I would say Ethereum mainet the current state of of uh privacy protocols on mainet or like is good but it's not big enough to uh to to to hold uh those big orders. Uh and the only way it it it can work is to make it grow. So like people should you all should use those privacy protocols uh because by doing this you enable uh bigger fish to enter uh the contract and to enter the pool.

Um and we are facing this because we're a public blockchain but so we have to use those external protocols uh to to to make this work whereas like in Aztec you you will probably not have this. Yeah, I guess like Zuko from Zcash has like a a famous saying and he's like you can't give privacy to like an asset that's in flight. So you can't take something that's public and then decide to make it private. You have to either add a time delay, which is how a lot of the the pools work on Ethereum, or you have to kind of, you know, pool it with other assets, but you can't just take a transfer and suddenly decide to make it private if it started as a as a public um transaction. So we we realized this um like after speaking with Zuko I think

in 2023 2022 and we realized that like the only way to get like an infinite privacy set is to start with everything private by default and then let developers kind of conditionally reveal what they want to. It does mean that like you know you have to kind of add things onto the EVM and so we have a different execution environment a different model but in that world the way you can get like a very big privacy set is you can natively mint something privately. So let's say you have a stable coin instead of minting it on L1 uh where you can see how much is minted and and and withdrawn. If you mint it privately where you can't see the total total supply but you have a proof that the supply is is backed you can now

start to do these trades that institutions want to do where they're not constrained to one privacy set. So that's one of the reasons we we kind of think it needs to be natively private but we have to balance that with kind of also you know going where the capital is. And so that's why we're on Ethereum because it's it's an L2. So, >> and funnily like this links to uh [snorts] monetary uh policy in the in the traditional world. And uh in a sense they managed to achieve uh this because they're not able to measure uh the total amount of supply. So it's it's a very different uh um thing. But uh because they're incompetent, let's say, they managed to achieve this this privacy by not knowing uh the current supply of

money in circulation. Uh so yeah, it's it's really interesting because like you you did real mass and and real hard work to to manage to get this result that they just managed to to to got by being incompetent and having a weak system. So yeah, thank you for doing it better and doing it well. Um really good really really good to have this and uh this is beautiful because like this uh layer 2 is going to enable to have this private uh natively uh private uh set set of action and this is the beauty of like the layer 2 road map like we don't have to make ethereum fully private and we don't have to go this uh to go do all of this and we can stay a public blockchain and have this built on top. Uh so really really good

uh to have Aztec uh very like excited to to to see what's going on how it's going to unfold. Uh a lot of layer 2 have launched uh over the last few years with very similar uh go to market to just like be cheap scale. Um maybe I'm going to say something a little negative about uh this. When you have 500 uh projects doing the exact same thing, it is hard for people to navigate. uh it it is hard for the projects to to be successful and uh because you have this uh clear go to market and clear strategy clear differentiation and an amazing technology behind I I really think you you will be one of the most successful L2 and really looking forward to that

launch. >> Thank you. Um yeah I don't know what to say after that. [laughter] Um, no. I think that like the way we think about it internally is that most L2s have they're competing on TVL, right? So, you have unis swap on every chain, you have a on every chain, and there's only so much liquidity. Um, for Aztec, we we don't really care about TVL because we we have block space that's unique. So, we can like bring the privacy to other chains and hopefully augment them. So, I think most L2s are kind of, you know, they're competing with each other and they're all competing for the same slice of the same pie. I think hopefully we're one of the first L2s that can be like net positive for another L2 and actually or Ethereum.

So, that's that's the hope and I think with like cool privacy wallets on L1, we can also start to make Ethereum um a bit more private than than it is today. >> Absolutely. um as soon as you will launch and like the timing is good for us because for now we we will be just mainet uh we we plan to have a demo by dev connect and uh and after that we will start to add layer 2 um so this will be the the right timing and obviously uh we'll try to to to add Aztec as soon as possible because this this just makes sense for a privacy wallet uh on mainet so yeah thank you so much for doing all of this and thank you for having us and organizing this beautiful event um was a pleasure pleasure to talk with you and thank you

for being here today, guys. >> Yeah, thanks for having us. [music]