Will ZK be bigger than blockchain?

At DevCon 2024’s ZK Hub, Sylve took part in a panel named: « Will ZK be bigger than blockchain? ». He spoke alongside Aayush Gupta at zkEmail, Neil Han at Reddio, and Ming Guo at ZKM, moderated by Sebastian Rodriguez from Privado. You can watch the video or read Sylve’s highlights below!

What’s the killer use case for ZK technology?

We first used ZK for privacy reasons, for example, with Zcash, where you want to hide your transaction. Second, and that’s the paradigm we’re in right now, we used ZK for scalability: to make blockchains go fast.

What I’m really interested in is using zero-knowledge technology for flexibility. To give an example, one of my favorite ZK applications is, after zkEmail, obviously, Proof of Passport or ZK Passport.

It is currently the best option for proving that you are of a certain age or nationality. DevCon, for instance, did this. If you come from Southeast Asia and you can produce a zero-knowledge proof of it, you get a discount.

In any other setting, without ZK, you would need full-on KYC. That means that it is more cumbersome, more expensive, and less private. So, it's the flexibility aspect that I really love.

ZK & blockchain synergies

Blockchains are 16 years old. And there's a 10-year-old forum post by Satoshi Nakamoto, in which they said there's this weird area of math called zero-knowledge proofs. If that thing worked, it would be really cool to have it in Bitcoin. But hey, it's just a paper at this point, so we can't have it.

Now, we're trying to cram ZK into the blockchain. If you look at the EVM, it's like it was purposefully designed to be ZK-unfriendly! Hashes are very difficult to compute, which creates a weird situation where basically all the ZK research right now is funneled into making ZK manageable within blockchain constraints.

Specifically on Ethereum, the constraint you’re working with is the 300k gas cost for proof verification, which means that you need to use groth16. You could use one of many other systems, but these verification costs constrict the design space.

Can ZK improve trust in government?

The end goal here is to establish trust. 

The two main ways we've done it are either zero trust or full trust.

Full trust is: I trust my bank. I trust that they're not mismanaging my funds because I have a legal contract with them. But in reality, it’s because these laws exist that I'm able to trust what the bank is doing. So that one's entirely trustful.

And then you have blockchain, where we decided to remove trust. We invented (or popularized) the term of trustlessness. I do not need to trust anyone.

What interests me about validity proofs is that you open up a spectrum between 0 and 1.

And this idea that we were talking about right before about the proof of passport is extending the trust that you have that the government is managing their keys. It's extending it for the application that you want to build, but you do not need to ask for permission. That's really the lock and key of the expansion of cryptographic keys.

Regulators and ZK

Something has to become customary before it becomes law.

The Homestead Acts of 1862 in the U.S. put into law something that people had been doing naturally. You were the owner of a piece of land if you had lived on it for at least three years. It was a fully decentralized law, and only when it was widely accepted did it become law.

Before lobbying and educating legislators, we need to learn ZK ourselves and invest in that.

Making ZK a first-class citizen

With the advent of computers and the Internet, we talked about digitizing: bringing offline data online. At the same time, the Internet had to work with the existing infrastructure, mainly landlines and phone lines. It was a second-class citizen.

Afterward, we created so much more digital-native data that digitizing doesn't really make sense anymore. We even flipped the phone so that even calling is done on online apps now, and the phone lines are wholly obsolete.

We sometimes say that Bitcoin is going mainstream. That is simply wrong: we're not mainstream by any margin. Any top 500 internet company has 10 times more users than all the Web3 users combined, and 90 percent of those are bots. We need to do a lot better than this.

I think it's going to be the same thing for attested data − that’s data on which you can make truth statements for legitimacy − as for the Internet. Right now, there are so few pieces of data on which we can sign and build, with the notable exception of emails. The cursive team has a really cool report called stuff with signatures, and it's depressingly sparse.

You have emails and passports. We realized in France that there's a standard called 2D-DOC that signs a gazillion documents, such as utility bills and identity cards… So that's cool! We can extend this. 

We're going to be able to onboard people extremely fast, starting with this. The cool thing with zkEmail is that, theoretically, every single person on the planet has a wallet. So imagine a worldwide airdrop. In France, we could use 2D-DOC and airdrop 65 million people just like that.

The second thing that's going to happen is we're going to flip signed data. We're going to create much more data that is perfectly amenable to this, the same way that we had fiber networks and then put the phones on them. I'm willing to bet on that.