Privacy: The HTTPS Moment for Blockchain

Transcript

Hello. All right, I guess I guess we're on to the next talk. Hello everybody. Good afternoon. My name is Zach. I am the co-founder of Aztec. A privacy preserving layer two for Ethereum and I'm here to talk about how privacy for Web 3 is going to be the HTTPS moment for blockchain. And how I think and and how Aztec's work connects to all of this.

So, privacy technology transformed the internet. I think for those of us who were around in for Web 1.0, this is is is fairly obvious, but maybe for people who are a bit younger who weren't aware of how the web used to be, this might not be a well-understood story where back in the 1990s, when you have the internet in its first incarnation, Web 1.0, largely static websites, limited bandwidth, very low throughput because everybody was using dial-up connections and you and remarkably, all internet traffic was fully transparent.

You did not have well-established encryption primitives to transmit information privately. More importantly, such encryption algorithms were still kind of considered classified from that Cold War era mindset and were considered kind of munitions, the weapon that couldn't be used by mere civilians. Um, you know, and so so the way that we use the internet today for e-commerce, for finance, for banking wasn't possible. You know, you had people like the VP of Visa saying, you know, "Don't send anything over the internet um, you don't want to be read by 40 million people. Does that sound familiar to anybody today working in blockchain?

And people could glimpse the future of what the internet would become, um, but they didn't have the capabilities to realize it. Um, this led to an enormous amount of investment, an enormous amount of attention, and then the dot-com crash in 2000 because all these ideas this this grand vision of what the internet was supposed to be largely failed to materialize at that point in time. Um, and I would like to draw a direct parallel to that to that and perhaps what we're going through today in blockchain where if we think about the cypherpunk visions and ideals of Bitcoin, the Ethereum white paper, what people wanted Ethereum to be, um, you know, this decentralized banking institution where you could bank the unbanked and everybody would be

transacting on chain. Bitcoin imagined itself as a payments network where people would transact in Bitcoin, not as a store of value. We're in a similar situation, I believe, and I believe that the two conditions to resolve blockchain's current issues are the same as the conditions that require allowed web the web two to succeed. Um, web two came about when you when you had broadband connections and you had high transaction throughput, high information throughput, and privacy by default, HTTPS, uh, you know, where every single, um, request that you're that you that you're sending and every response you're getting from a server is end-to-end encrypted. You have a trust like a chain of attestations through digital

certificates so that you understand the data where your data is coming from, but you can't be snooped on by anybody. Uh, you know, your payment information, your sensitive details, they're not visible to observers. Um, it's the power to selectively reveal oneself to the world. We've had the scaling equation down in blockchain for some time, uh, and now privacy is starting to arrive. And I've been working in this space on privacy for eight eight years and it's very good to see finally um but this is going to be completely transformational. We already just heard um Paul uh earlier talk about how institutions are ready and eager to adopt blockchain-based technology. Once

you have privacy uh technology embedded in it so that you can enable selective disclosure, you can perform KYC, you can hide information from observers and competitors. But it extends far beyond enterprise and institutions. It extends to the people in this room. It extends to the consumer applications that we want to build. It extends to people who wants to create financial products and services and applications that compete on a level playing field with Web 2 and TradFi um that can compete with them and not necessarily are those products.

So, how do we fix Web 3's privacy problem? Because privacy is a hot narrative. But could I get a show of hands here for how many people regularly use blockchain-based privacy technology? It's about four of us. Um it's just a narrative. Uh the technology has matured in the in the in in people's eyes in this industry to the point where you can see the future. You can't grasp it yet. Yet. Uh and I believe this is about to change. I believe it perhaps changed yesterday and I will explain why.

Blockchain needs privacy. Uh you can't build institutions, whether they're companies or decentralized organizations or something in between. Um you can't run those completely transparently. You can't do your your payroll transparently. You can't have contentious votes transparent and open to manipulation and lobbying. You can't govern in the open. And real-world transactions, the interactions that we have with each other day-to-day, they require information asymmetries. Um you know, where I know something you don't. You know something I don't. This could be as simple as voting.

I know how I voted in an election. You don't know how I voted, but you know my vote counted. Um or it could be, you know, um uh performing a sanctions check on somebody where they need to give you certain amounts of information. You you have all the other information required to perform KYC compliance checks. You don't want to share it. Maybe you can't share it. And privacy enables information asymmetries. It enables identity access management. It enables information from the real world and from Web 2 to be put put to be piped into Web 3 smart contracts, distributed smart contracts, decentralized autonomous code running on a blockchain. It allows that sensitive information to be placed into the smart contracts

fully encrypted, yet those smart contracts, those autonomous programs can reason about that data. And this is relatively radical. It allows private payments at scale. It allows um high-quality understanding of who somebody is. To give an example, you know, if you look at your phone and you look at the apps on your phone, because we're in a blockchain conference, I imagine people here, many of them, will have blockchain applications. They will have crypto apps. And they will have Web 2 apps. And these things are like oil and water. They don't mix. It's very clear if you have a blockchain app um or a Web 2 app.

And this is because crypto applications don't actually know who you are. Um if I open up my wallet, I am a string of hexadecimal characters. I might be an ENS domain, but what that relates to in the real world is completely unnable by that application. Whereas if I open up my banking application or if I open up even my social media, that application has a much richer understanding of who I am as a real person. Um and that allows for more complexity. And the fact that there is no implicit intuitive way of communicating that information into a smart contract um without doxing yourself, without revealing information that cannot be revealed, that prevents these crypto apps from

becoming infrastructure that is used by mainstream applications. It it prevents the boundary between crypto apps and real world apps from dissolving away. Um you know, I think we all understand that when crypto succeeds, if we use blockchain technology as infrastructure for things like payment rails um or banking services you won't have crypto apps anymore anymore. You'll just have apps that use the blockchain in the same way that apps use IPv6 and they don't talk about it because they don't care. It just works. And if we don't have privacy you know, from the mouth of the Zcash Foundation itself, what all of that we're building is just a information panopticon. Um it is it is surveillance technology. Uh that's that

does the exact opposite of what everybody here wants it to achieve. So this is what I've been doing for the last 8 years. Uh I would argue that I Zcash built the modern privacy stack. Um and we we built it um you know, whilst uh we we we built a lot of it from from pretty much from scratch, you know, from from ideas and mathematics into papers into code into products. First, you need a uh extraordinarily faster non-loss proving system. If you want to privacy, that means you're constructing your proofs like like locally on your clients, on your phone, on your laptop. And you need something to handle towards that that can handle that.

Which is what we produced. Um you then need a programming language that actually understands privacy semantics, that understands the difference between public information and hidden information, that understands uh private state and how to manipulate it and the semantics around private state. And finally, you need an execution environment and a state machine um that's uh can uh execute privacy-preserving transactions a global distributed ledger that's the Aztec network and we've been building this for many years now in various incarnations and I'm very happy to announce today that Aztec is definitively landed and launched.

We launched our alpha version of our of our network on the ethereum mainnet yesterday. And the yesterday the first Aztec transaction came through. So this is now the combination of something that we've been produced working for for a very long time. You can now write privacy preserving smart contracts, deploy them to a fully permissionless decentralized network and execute this transactions. You can bridge into and out of them from ethereum and and it all works. And it's opening up a new design space in web 3. So this is just a smattering of the applications and that and projects that are being built on Aztec from bridges and private decentralized exchanges to on stream games to stable

coins to to compliance portals. It's the start of something very radical and very new that's not been possible before because all these applications in some form or another have much more of an understanding of who the user is than a regular application because that user has private state. And you know ECC is always been a more developer focused conference. So I want to talk a bit more about the actual technology behind it, how it works, how it's built, how how we how we build with it. You know it's a as I said it's a fully decentralized permissionless L2 you know from day one from launch. It doesn't have back doors. It doesn't have

you know the ability for for selective selected individuals to pause or upgrade or change the network. It is all governed by our sequences and anybody can be a sequencer. Perhaps there are some in this room. Our block times are relatively competitive with the serum. Privacy adds a lot of extra complications there. Um. And we have a very feature rich smart contract environment written entirely in Noir, which is a programming language we've been working on for many years to that makes the the creates abstraction layers around handling private information so that one can actually define more idiomatic patterns for working with this with this material that that are intuitive to

engineers. So you don't need to be a cryptographer. You don't need to understand this the stuff from a from a low level. Um and the goal is really to make blockchains blockchains like Aztec in particular a canonical source of encrypted credentials so that you can perform the compliance checks you need to understand if somebody can own a real world asset or not, but to do it via a smart contract so that then that real world asset can be used in other smart contracts as real liquidity. And I think this is alluded to partially to what Paul was saying earlier about just the revenue the liquidity universe of traditional financial markets is orders of magnitude greater than anything on blockchains. And if you can tap into that, that changes everything.

And so um I want to explore the anatomy of an Aztec contract here. So this is a a a card game I've a card game I've been developing myself to to stress test the network and its capabilities. It's uh um relies on hidden information. It utilizes a bunch of the components of the protocol um uh to stress stress test them out. You might be able to see some of the players' cards are hidden. Some of the opponent's cards are hidden. This is all mediated on chain. And um the semantics of this are pretty straightforward. You know, you you we have the ability to define functions as private. And what that little keyword private there means is that when you when that transaction gets executed on chain, Um, all the information in that

function, all of it's hidden. Um, so this contract has private NFTs in its storage uh that are being manipulated, they're being removed, they're being added. Um, events are being emitted and um none of this is visible to observers. They don't know who's called They don't even know So, if you see an asset transaction that consists of private functions, you don't know who the transaction sender is, you don't know what contract is being called, you don't know what state has been changed. You genuinely have no idea that something's happened, but that zero-knowledge proof that comes with that transaction proves that whatever is happening it has followed the rules of the smart contracts that are being called. Um We also have public functions. So, you

can build hybrid models where you can have um strong privacy guarantees for users, but you can still build protocols that have public guarantees when they're necessary. Um, uh so here I'm linking this NFT contract to a token contract. Pretty straightforward stuff. You want that to be public. You don't want that to be private, probably. Um, and we have a fairly idiomatic way of defining storage and some of these variables might be these types might be might be uh un- like un- uh novel uh they might be novel to to a regular smart contract developer, things like private sets. Um, where that's basically it's a collection of private information that's owned by an individual. But you don't need to worry about how that's handled under the hood. You don't need

to worry about complicated things like Merkle trees, nullifier sets, notes, encryption, note secrets. Don't need to worry about that. We have idiomatic ways of handling private state and private information. And this is one of the values of Noir and the Aztec network. Uh you can like this little function here, which is just a private transfer of an NFT, um five I can't count five lines of code. And what it's doing is, you know, it's accessing a uh the protocol has Merkle trees and nullifier sets that contain um inf- like information that's linked to specific storage slots. It's nullifying some UTXOs. It's create um, it's creating some new ones. It's um, accessing the user's UTXO store. Um, so

and and iterating over it. It's doing a lot of work, but it's all being abstracted away. So that's from the developer's perspective, it's just like, yes, you just you access your storage slots, you pop on your assets, you insert your notes, and then you deliver it. And that's and that and that deliver function, what it's doing is under the hood, at the protocol level, um, uh, an uh, an encrypted message is being sent to your recipient. So that when your recipient, when they open when they when they run that that um, they connect to the Aztec network, uh, via something called a private execution environment, uh, which is embedded in every single Aztec app, that pixie will automatically sync to their note state. Um, so I can just do a private transfer between you and me,

and it just works, and it's just automatic, and nobody other than us understands what's happened. Um, and one of the reasons why uh, it's been a challenge to build this is finding the semantics and the patterns to make this stuff make sense, to make it intuitive. This doesn't exist anywhere else other than Aztec. I would make that claim. Um, where we've built this smart contract architecture where it looks like it it is it the semantics are familiar to regular smart contracts. They're similar to regular smart contracts. You don't have to think, twist your your your brain in mental knots, uh, to to understand transactions that are on Aztec. It's just contracts, they call contracts, they call contracts.

It's pretty straightforward. And then you have this this issue of state management. Um, so anybody who's built a Bitcoin wallet might understand that if you're working with um, UTXOs, which is how private state has to be man- managed, that could be quite challenging. It's challenging for Bitcoin because a user's balance is made up of multiple state variables within the Bitcoin ledger. If you have If now you have a a blockchain where that state ledger represents abstract state instead of token balances, you might think it gets more complicated.

But again, this is all very contained um because you just write getter functions in your smart contracts, the same way that you would in Ethereum. Um uh so that if I'm an application, I need to understand what state a user has for a given smart contract, I just call the getter function. So you just have an interface. You have an API as one would expect. Um but what's actually happening under the hood here? Well, uh um that that would take me That would take many hours to explain. Um uh And you know, Aztec contracts even have state shadows built in by default. Um something I discovered when when dog feeding Aztec myself. Um this this little snippet here basically just takes in nine more proofs and verifies them recursively using our advanced ZK tech

uh to perform this step very, very efficiently. Um so this this happens very, very quickly uh on the order of of uh seconds in a web browser, this part. Um And Aztec and Aztec use the same tool chains. They use the same language stack. And so this this this all uses the same unified tooling and allows for basically every Aztec contract that it wants to be its own roll-up. Um which just massively expands what you what you can efficiently and easily do within the smart contract. Um uh And so I guess to to to close on this, um privacy is has been the missing link all along, I think, in blockchain. You know, I've been around since 2017 when, you know, the the the first blossomings of DeFi were happening, when

um al- like uh when uh collateralized stable coins were first becoming a thing, where decentralized exchanges were first becoming a thing, automated market makers, and I remember going to hackathons at the time, learning how Ethereum works, how blockchain works, and the the things that people were building or trying to build were anchored in the real world. You know, people were trying to build insurance marketplaces. They were trying to do um uh like store shipping receipts on chain to handle supply chain complexities. They were trying to do real-world complicated financial flows. I myself was doing a startup to try to do um uh mid-market corporate debt. Uh in Paul's talk, it would fall under the illiquid private assets category that is going to come on chain with this new

privacy technology. But all these people either pivoted or they fell away because because they couldn't realize their ambitions because everything on on chain is so synthetic because it cannot anchor itself to the real world because of the privacy gap. Um because you cannot you cannot put real-world events on chain if they require you sensitive information in a transparent world because you cannot utilize information asymmetries. You just can't do most of the most of the If you open up your phone and you look at all the financial applications you have that are not blockchain on transparent networks, you just can't you can't realize that kind of functionality.

Uh but with privacy, you can. And uh I think this is going to be the start of a new renaissance for blockchain. Uh and uh I I think Aztec is going to be at the core of it. Um our network is now live and you can build on it and develop on it, and so I would encourage people in this room if if any of what I said has resonated with you to to have a look. Um you know, it is extraordinarily easy to get started, particularly these days with with AI. Um so this is me. This is uh this is Aztec, and uh thank you very much for listening. >> Mhm.