Security Token Show Interviews: Yuval Rooz on Etherealize, On-Chain Privacy, & Canton Developments

I'm good Jason, how are you doing?

Yeah, so first of all, kudos to VC for being focused. You look at his background and clearly he understands capital markets and what the driving forces behind what Wall Street does, and that's a good thing. I always say that having different views in this ecosystem, people coming with no context of what Wall Street is, what capital markets are, and having this vision we can do it better without even being in it, is a good thing. But having the contrast of people that came from the industry and understand some of the inefficiencies, why they were created, why they are there, sometimes for good reasons, is good because you get diversity of opinions and create better products over time. So first of all, I'm happy that he's doing it, it's good to see the energy behind it. From the day we created Digital Asset and started building Canton, what VC is talking about was our one singular motion and vision: at the end of the day, Bitcoin is a unique asset class, it doesn't have an issuer, it's censorship-resistant store of value, and there might be a few of them, but most assets in the world are going to be RWAs. So it was always the vision of how do you bring $300 trillion plus worth of assets on chain. I'm very aligned with what VC is trying to do. Our approaches might be slightly different, but I'm supportive of what he's doing. Financial institutions have a certain set of requirements to put higher risk assets on chain.

Sure. First of all, if you subscribe to my assessment that most assets on chain are going to be RWAs, it's very different from Bitcoin. Bitcoin doesn't have an issuer, you don't have counterparty risk, you trust the code. It's decentralized and doesn't have real-world issues where you have different types of risks factored as part of the issuer. For a stablecoin provider, how do you manage deposited cash? Do you spread it across custodians? Do you invest in the instruments you say? For private credit, you have credit risk. All these assets have different risks. By putting them on blockchain, we're reducing friction in the financial system, but we're not getting rid of the trust needed through people bringing these assets to life. There are mechanisms to reduce risks, but it doesn't go to zero. So my point is, there's always need for trust with RWAs. When you think about institutional-grade use cases, privacy is going to be a must-have. Even today, crypto players don't like whale watchers displaying what they do, and people front-run them. I talked to a company doing cross-border payments using stablecoins, and FX traders front-run them because they see massive FX movements on chain. Privacy is to some extent a human right, and practically, large-scale institutions must have it. A publicly traded company doesn't want all their activity visible in real time. Privacy is a must-have to bring these assets on chain. When we started building our layer one called Canton, privacy was an anchor tenant. We wouldn't launch until we solved for privacy. For public permissionless chains like Ethereum, privacy was not a design requirement, which is why companies like Elliptic, TRM, Chainalysis can track what's happening on chain. There's excitement about zero knowledge today because the hope is you can add privacy as an afterthought to systems not designed with it. This is where we have a strong disagreement. My co-founder was deep in the technology, one of the guys who wrote libsnarks, a C++ library highly used. We have appreciation for it; for scaling, it's good to package transactions and speed up. But for financial services, not only is it not suitable, it's actually risky in many cases.

I'm sure ZK experts will grill me after this, but the way I think about ZK is it gives you the ability to say, as a consumer, I only care about the output being valid, and I don't need to know the inputs. A rudimentary example: you go to a bar, give your ID; the bartender doesn't need your date of birth or address, just that you're over 21. Imagine something scans your birth date, an algorithm checks if you're over 21, and gives a thumbs up. That's a great use case. But this industry started with 'code is law,' then we had the DAO hack, and code is not so much law. We've had other examples where code has mistakes. What if the algorithm didn't get your birth date correctly and says you're over 21 when you're 18? That's a bad outcome but won't melt the world. Now imagine a ZK algorithm used to deposit collateral; you don't want to see who's depositing, just that the collateral is there. We'll call that box 'SBF.' Every time you give SBF collateral, it says it's there. Then a market event happens, you need that collateral, but you find there was a bug and you could have minted collateral out of thin air, and the box still says it's there. That would be a very bad outcome. In capital markets, relying on a black box telling you everything is good can have bad consequences. This isn't theoretical; Zcash had exactly that issue. You could mint coins out of thin air, but because it's zero knowledge, the chain said everything is okay. When discovered, how do you know if someone exploited the bug? In financial services, there are points where privacy is not anonymity; it's not a black box. If you're a clearing house or lending protocol and need to know collateral is there, you can't rely on a black box. Privacy is not 'trust the black box'; it's for you to verify it's there, but for the rest of the world not to see it. That's the big difference: we disclose information on a need-to-know basis, whereas ZK anonymizes information and just trusts the algorithm.