Danny Ryan - Turns Out It's Complicated

Transcript

All right. Hello. Hi guys. So, we have Denny Ryan and I have notes because I need to know what I'm asking him. Uh, from Etherealiz previously a researcher. Um, I think you're really deeply involved in getting Ethereum to institutions and what that means for well the centralizing pressures that institutions exert on Ethereum. And I the title of this talk is it's complicated [clears throat] as in we did proof of stake and we thought that's awesome. We're done. And we're really not even a little bit right >> done. Right. And I just I I gave a talk about some other stuff recently and I

was talking about how at the beginning of 2018 I talked to Ya Miguchi and she said so when do you think when do you think like proof of stake is going to launch? I was like I don't know like six months [laughter] and uh it took quite a bit longer and you know I think optimism and naivity but like not only did it take longer but the dynamics of the system the dynamics of the mechanism design the dynamics of the financialization of ETH and all sorts of implications on this thing that we've been working on. Uh we still have a lot of work to do.

>> Yeah. So let me ask you um a probing question about institutions that you're talking to and centralization. So what I see when I talk to some traffic clients is they have a risk team. The risk team needs to underwrite the node operators that run the stake and the risk team is like we can do four maybe max five of these right you know and we need to know that they're ISO compliant and what their financials look like etc. And so they're like, "We can't do 39 or hundreds or thousands, >> right?"

>> What are you seeing? >> I certainly see that and I see, you know, there's some gems. There's some that do want and plan on doing self-custody, uh, which is pretty cool. Um, but by and large, I have, I don't know, kind of like sad conversations where I I talk to someone the head of head of digital assets somewhere and they say, "Yeah, yeah, we get it." And we get that like by making these decisions we might lose more money but also I can't remember the the example they use but they're pretty much no one gets fired for choosing Microsoft.

>> Mhm. >> So it's like actually if we all go down together we're fine. [laughter] >> Oh boy. >> Uh which is terrifying. Right. And instead it's like what he's saying is if we make an irresponsible decision and we we lose all of our customers funds, we're good as long as everyone else loses their customers funds and we're all in this boat together. When it's like you can just make a different decision, but they're worried about when they choose an operator that other people aren't choosing >> and they go down and they look, you know, they lose >> probably some some forego some rewards for a few hours. That's they're worried they're going to lose their job for that, but they won't lose their job.

catastrophic because as long as everyone else is in a catastrophic >> state I I think that answers the next question which is are they thinking about network health and I think we just heard not necessarily >> right >> it's a bit it's a bit of both right like some of these people especially in the digital assets teams they get it >> they understand the risk they understand the network health but they're also you know the way that institutions think it's like this weird herd mentality um and quite frankly pretty broken incentives to the people that are making these decisions. Um I do we shall see we shall see. I do believe that uh you know they say some of these

things not entirely in justest but they're like these we do see these real problems. We do see um you know the risks associated and um I think there's an attempt at least in the forward thinking institutions uh to try to navigate these in a principled way. But it's a tough space. >> Have you found ways to educate, bridge the gap? Maybe. Do you have ideas on how Ethereum can incentivize institutions to stake responsibly? As in think about the network health when you're staking, you know, not just if we all go down together, it's fine.

>> We'll just orchestrate a catastrophe. >> Yeah. >> I mean, honestly, if somebody gets burned, they're all they're all going to figure it out. Yeah. Right. >> Um >> I'd like to avoid that. [laughter] I have bags too. Um, but I education is important. >> Yeah. >> You know, we uh we attempt to be in the room. Uh the the EF kind of enterprise team is going to be in the room. Um I've had some dialogues with the SEC about like if these are going to be regulated financial products, isn't like the whole point of the SEC to like communicate risk? Mhm.

>> So, if there's a big rubber stamp, uh rubber stamp's the wrong word, but uh you know, if you use this ETF, you can lose 50% of your money. Use this ETF, you can lose 5% of your money. I think it would go a long way, right? Like if if these things were elevated as >> uh part of like the risk profile of investing in these products. Um >> so part of the prospectus almost. >> Yeah. Yeah. I think that'd be great. I think it'd be great. And there's, you know, there's some people in the SEC are like, "Oh, that that makes a lot of sense." There's some that are like, "No, that's overstepping the bounds of like how we structure products and how we talk about things, you know." Um, so but I think if we beat the drum, I mean, at a certain point, it's about education.

It's about actually understanding the risks. >> And again, when you're talking to somebody who got served by the SEC a year and a half ago, now I was talking to them, but like they do have a purpose. Help educate consumers about risks. Flashing's a risk. Losing your money is a risk. Let's, you know, do your job. >> It absolutely is. Absolutely. and and you know I'm very happy that there is now an organization like Etherealize who's trying to bridge that gap which I think was truly lacking before. Um do you see that institutions are worried about capture of governance of Ethereum at all? Are they thinking about this capturing governance or worried somebody else will capture it?

>> Either one. [laughter] >> Uh they're not thinking about it. Um >> they don't. There is, for better or worse, the people at institutions that understand Ethereum, they're like, "It's decentralized. It's great. It's neutral infrastructure. That's great." And there's not a lot of nuance to like what could go wrong with how those statements could could become untrue. >> Right. >> Right. >> And there's certainly, you know, I want them to be confident and to use and and begin to adopt this technology. um while we at the same time kind of harden that layer.

>> What I've been worried about for the past decade is is capture of the institutions actually being like, "Oh, let's >> oh wait, I know how to I know how to get a seat at the table. Let's build a client >> or let's buy Sigma Prime. >> Yes, >> they have a client, right?" Um and I think that, you know, part of the title of this discussion is turns out it's complicated. And one of the reasons it's complicated is we're not done. If if by and large we were done architecting the system and it was sound and stable and maybe that's a dream, maybe that doesn't actually exist. Maybe the world continues to evolve and continues to be complicated and the protocol must evolve to meet its needs.

We shall see. >> Well, maybe maybe Justin gets his wish and Lin Ethereum solves everything, right? Yeah, [laughter] >> that's a that's a different 30-minute conversation. Um I But as long as the protocol is malleable, uh we it is it is capturable, you know, and and as long as um Ethereum is valuable, people will want to capture it. So, I think we need to be incredibly worried. You know, I I think that we're >> I don't think JP Morgan's going to have

a client tomorrow. >> No. >> Um It's not out of the question that they would though >> if they're if this is critical infrastructure and they're relying on it >> and they're not only like have staking ETFs or whatever but they actually have you know they're powering some of their back office with this technology like of course it's going to make sense for them to potentially support the technology even even in a positive way like if we if you look at how Linux is built >> um >> it only does so much but like so many of these large corporations have people that just like help work on the >> kernel Yeah, like Intel and Microsoft, etc.

>> And so we might we might see that in the in the financial institution world. It's like oh you >> you want the expertise, you want to make sure the thing's running well, those allocates [clears throat] and and like that's all that's like all good. That's great. I I think it's a good thing uh that you know people hopefully will help help build and support this infrastructure but the fact that the infrastructure is still being the protocol is still being built puts it a precarious position once the vested interests these massive vested interests potentially enter the game again I don't I don't think this is a tomorrow thing >> I I think it's probably likely >> does the social layer need to evolve to

deal with these changing realities and If so, how? >> I mean, the social layer needs a backbone. >> Okay. >> Like, for sure. Um, honestly, I joked about like uh once they get burned, they'll they'll realize, but it, you know, we haven't actually had a good fire drill. Mhm. >> There's been some y'all done your job too well.

The networks have been more stable than anyone expected, but we haven't had to make these hard decisions of like what happens when there's a contentious work, what happens when uh the network splits, what happens when people might lose a lot of money. Um and I think you know after the DAO we had a number of these instances for forking the the execution layer right like I think parody lost god knows how much money um there were a couple of these instances and the social layer said no >> n >> absolutely not practiced and and and at this point like >> the answer is no like we're not going to fork to like fix your smart contract >> right

>> and um I kind of wish we have had would have had a fire drill on the consensus layer so that you know the precedent was set. It wasn't just taught. >> So a heski style meltdown but on mainet [laughter] >> someone's going to quote me on this. Yes, let's have a style. Perry, you ready? [laughter] >> Oh. >> Oh god. >> Yeah. [laughter] I don't know. I mean, but definitely I think that the social area needs a backbone. I think the social needs to be global. >> Okay, that's >> when we're talking about institutional finance. uh for better or worse, at least from what I'm working on and where a lot of the the movement is, you know, it's in the US.

>> Yeah. >> It's very concerning. >> Chicago. >> Um I think that the rest of the globe should step up and and institutions, you know, people should beat the drums and and get these get we want like disperate, not totally overlapping financial interest in the network at every single layer. And I think that that you know, not just the social layer of like enthusiasts and the home stakers and and and you know the more crypton native but like you know I we want good stable coins across the globe. We want good uh you know staking products like ETFs across the globe. We want um you know disperate interests

to be at odds with each other so that like actually the system just has to work like it is and it's much more hard to capture. Um, you know, we need we need clients in Australia. >> We need clients. Do we have a client in South America? Is FedE building a client? >> Oh, [snorts] um, Ethics is in Argentina. >> Yeah. >> Yeah. Namechecking Ethics. You most of you have not heard of Ethi. It's an execution layer client right out of here Argentina. >> I hear good things. I hear I hear he has a wine bar, too. >> Yeah.

Um, how are we for time? >> 10 minutes. >> 10 minutes. Great. We're 10 minutes in. >> We have 10 minutes left. >> Damn. I >> Okay, wonderful. Um, if you could redesign the staking model right now with everything you know, what would you touch? This is more of a researcher thing, right? Like which you used to be. >> Yeah. Well, I've been a researcher. Um I know what the hell I was doing. Um that is hard. That is very hard. Um I I think that there are some very excellent components of the system. Um I think

that the fact that there's like this dynamic availability thing uh is a strange design decision, but I I stand by it. Um, I think we wouldn't have done the everyone 32 e kind of thing like continuous validators patch that up. Hopefully that'll that'll kind of fix in the network as people adopt it. Um, I think and this is more forward looking but with foresight maybe it would have been a better decision but like the separation of roles like the specialization of potential validator roles I think is actually probably very important. Um because I do think >> testers, proposers, builders, >> like if we would have if if we would have had a crystal ball and seen you know the centralizing impacts of me,

specialization of me, block proposing building >> coming in from outside. >> Yeah. then I think that uh the specialization of the roles at the outset might have been a very valuable component and that's because you know the long tail even if there are centralizing pieces of the system having many disparate actors that actually attest to and validate the consensus and do that relatively lightweight component compared to the rest of the system. Um that's what that's what gives us kind of the longtail protection. Um, and that's why I think, you know, solo staking or homestaking or like at least small company staking like is valuable is

because of this like long tale of that particular activity of validating the network and attesting to the network and testing the validity of it. Um and even in the sense of like u you know if there are is massive centralization a component of of the network and they go rogue or they go down or they go whatever like that allows um the kind of the long tail to fill the gaps. And so I do think again the specialization of roles at the outset might have protected us from certain centralizing things that are happening now but >> are we getting there with the changes that are coming?

Um, it's funny. I was looking at like the fork list to make sure I refresh it. I don't go to I don't go to the consensus layer call anymore, >> although Stokes lives like two blocks away from me, so we do talk. >> Um, [laughter] uh, but yeah, I think the the EPBS stuff that's coming in um is is beginning kind of thinking in the right direction in terms of the separation, the roles and different things. Um, I do think that there's still work to do. It's hard. You know, the priority list stuff the priority to this is long. >> Yes, it is. Um what so we talked about all the challenges with institutions and what they do and do not think. What gives you optimism that Ethereum can remain uh decentralized and credibly

neutral u under this centralizing pressure and we don't really care about the protocol pressure from institutions, >> right? Um I mean ultimately even Even with centralizing pressures, we need to design protocols that are resilient in in the event of disaster, right? Like protocols that slash the capital that was decentralizing and doing something evil. Protocols that >> can recover in the event of uh major downtime or forks and things like that.

And like if you look at even Vitalix, he has like on like github.comthereum/ressearch, there's like a a folder in there called papers. And this is him like writing stuff in 2016 2017. If you go look at his papers, it's all about handling tail and edge case scenarios. >> Um, and being able to recover from disaster and a lot of the core of the mechanism, even the components, you know, there are many components that could be better. Um, but I do believe with a resilient social layer that, you know, the protocol will be fine. That said, some of these scenarios might not be

pretty >> very true. >> Try not to get there. >> Yeah. Thank you, Danny. Really, really appreciate it. Um, we have time for questions. [applause] >> Oh, I meant to give a disclaimer. I I'm an observer of proof of stake at this point. Um, you know, I I have a deep relationship with it, but I also um there are people in this room that are carrying this thing on their back, so they have they probably know better than I some of these answers.

>> You um you're talking with all of these institutions who've been doing risk managements for years, decades, and centuries. And then when you you tell them that we have decentralization and credible neutrality, just take it as a given. Um doesn't that mean that we are ahead [snorts] of the curve in terms of like risk management and and seeing all these potentials issues in the features? >> Yeah, it's a different paradigm, right? Like there's financial risk management is is far less like mechanism design. It's far less like how do I orchestrate and architect the incentive of of actors to do

particular things. Certainly the financial markets push and promote and do different types of behaviors, but like they're not they're not mechanism design risk experts, whereas I think the people in this room are. And I think that's that's the problem. And I think that's that's the paradigm shift that financial institutions need to have to adequately understand these systems. >> Thanks. >> Hey, Danny. [snorts] >> Welcome back. First of all, um, as we go forward, um, in which, you know, the Ethereum Foundation has, uh, limited treasury and as we go through

with time, um, it becomes harder and harder for us to fund things that are incredibly neutral and stuff like that, whether it's critical parts of the ecosystem like client teams. Do you does it worry you at all that at some point in the future that would be capturable by say institutions with a lot of money who would end up actually you know buying out a team like Loadar or something like that and and how would you mitigate uh or continue to keep the ethos of Ethereum if say JP Morgan bought us?

>> Yeah. Um, you know, people that run client teams are individuals. Uh, I it's funny, this, this is like child's play, but I remember when Prism was bought by Arbitum, there was concern. I was like, "Oh, Arbitum's buying a seat at the table." Um, oh, oh, no. And I I remember talking to the some of the guys over there and they're just like, I mean, you think I'm going to get on this call and like shill shill some like random that they're pushing behind the scenes like we're we're a bunch of individuals and we're just going to walk. Like this is an open source project. And so I think individuals can stand for what they want to um and and stand for what they what they

believe. But also I think the other mitigating factor is Man, I feel like I'm the only oificationist left in the world. I like this system's not done. And I I know it's not done, but like portions of it like when it gets to good enough, we should kind of signal like we're done. We don't mess with that anymore. And and it should be harder and harder to mess with the thing. And and I think that the protocol should be sufficient. Should be sufficient for the task. It doesn't need to be the best at every possible thing. we don't need to tune it and tweak it forever. And if there's a critical security problem, like anything's on the table, but like portions of the system as they become sufficient should be harder and harder

to change. And I think they have to be if JP Morgan is going to be buying a client, which is not out of the question. I'm not picking you. Someone go I think u we're all aware like the advantages of crypto and everything and like the slow settlement. It seems like a bug, but it's a um a lot of advantages in terms of like mistake correction and like legal things. So like I assume like a major thing that you deal with is the the fact that the irreversibility of like you know like ETH um is is like a major should be a major problem for institutions like how often does that come up? I assume it comes up a lot and like what's is there any answer you have for that or

>> um no I mean that that's that should be solved in the application layer right like you can program whatever you want into Ethereum if I'm moving a land deed to you and we have a problem like let's deal with >> oh on ETH um yeah I mean they need to handle that in their security and risk profiles it's not recoverable I don't I don't have a great answer and they're super super paranoid Um, I think that institutional ETH will not move a lot and they will have incredibly high kind of security practice around moving around, but it's it's a foreign concept.

It's not like sending a wire and well, that's the closest thing it is too. Sometimes you send wires into the void. But >> so we have hundreds if not thousands of EIPs proposed every year and you're saying that we should just oify [laughter] >> which like [gasps] which which part of the protocol do you think we should oify and should not touch? >> Uh it'd be nice to never touch issuance again. I [applause] but it also would be nice to not have 100% of ETH staked. So again, I'm practical. I'm practical. Um, you know, if if there's a if we need to

mess with that thing, let's do it fast and let's never do it again. [clears throat] Um, I think that uh, you know, we should make sure the pricing on the EVM is right, but we probably shouldn't make tons of breaking changes to op codes. adding op codes is a bit less potential for nefariousness than adjusting op codes. But even then like get the EVM to sufficient like and sorry I do believe layer 2's can and should diverge from EVM mainet and EVM mainet should be sufficient. Um and I don't know I was just looking at all those

EIPs. There's a lot of work to do. So I [laughter] have a bit of odds. >> Hi. Uh I'm not super aware of uh what Etherealize is actually doing, but as far as I got from a few talks from you that I heard that you're talking to institutions, which is good. At the same time, there was a big concern that institutional capital flowing into staking might ruin all this So, how do you find the balance here? And when institutions will ask you, okay, we want to go staking, how should we do that? What would you answer

them? >> Yeah. So, Etherealize is a champion for Ethereum in this this environment. We also spent a lot of our time not focused on ETH and not focused on staking and focused far more on how do we take uh institutional markets and assets and capital flows and actually rearchitect them on top of Ethereum and and bring the world's capital markets onto Ethereum. when uh we do talk about the ETH stuff as we as we talked about earlier in staking I beat the drum heavy that their risk models are wrong and they will lose a bunch of money and like I said

people understand but they don't have the proper worldview to incorporate it. I do think that um if Ethereum is going to win as like to talk about this year. We do need to onboard the world. Institutions manage anywhere from like 45 to 66% of the world's investable capital. Um depending on how you estimate that that's something like a hundred trillion dollars under their management. Um if we're actually going to make an impact on the world, that capital needs to begin to flow on chain. I think that if we are going to make an impact on the world, we have to do it now and not in

many years. And so I think we're going to be at odds with the very real time consideration of on boarding the world and attempting to protect these systems. I don't have all the answers and it's going to be hard. >> Well, thank you Danny and Thorston for this excellent conversation. [applause] >> Thank you. >> Thank you Danny. Thank you, sir. >> [applause and cheering]