Decentralize, or Die - Nick Almond

Transcript

Hello everyone. Nice to see you all, lovely Ethereum people. Um, it's good to be here. Um, I went for the enigmatic title this time, decentralize or die. Um, which I mean quite literally. Um, which I'll get to in a minute. Um, so I'm trying to tell a bit of a story here about trying remind us a little bit about why uh, decentralize is uh, decentralization is important cuz we've kind of forgotten it a little bit. Um, but also I think it's like even more important than ever.

Um, so the kind of story I'm starting with is to try and paint the context a little bit of um, a kind of world that I think is in trouble. Um, I call it largely the crisis age. I think we're likely going to get a lot more crises in the future. Some people call this like the polycrisis or whatever. Uh, generally means everything's a bit broken. And one of the reasons why I think it's a bit broken is because of uh, a governance crisis effectively. So, um, being a kind of governance nerd, which is uh, a curse I can promise you.

Um, you take a look at these systems, right? You've like your interest is governance, you look at how the governance is going. And uh, largely it's in a diabolical state. Um, and actually almost every large governing entity uh, are totally out of their depth and consistently make terrible decisions. Like it's got to a level where they only make bad decisions. I think there's a level of institutional decay that's got into the system at large. the the just is consistent bad decisions. So consequently, institutional trust is collapsing, legitimacy is evaporating, and the geopolitical environment is getting shifty.

Um, and so basically like my point here is that every sufficiently centralized system collapses under the weight of its own complexity. Um, it's not my point. This is just a well-known thing that anyone who kind of studies complex complexity science will be well aware of. Uh, and I think we're getting close. And one of the main reasons is is that I if you look at what causes complexity, it's largely like the these main drivers, like interdependency. So when you have a large-scaled system, lots and lots of people in it, and there's a lot of us, you know, there's many billions of people, and we've increasingly got more connected. So about, you know, 30 odd years ago, we like switched on the internet and connected everyone in the world, which has significantly ramped complexity. We

now interact with people on a global level way more than we ever have done like by orders of magnitude. Uh, and now like a Trump tweets can like annihilate trillions of dollars in value like in like the we we are hyper-centralized in some some ways. And what that does is create a very complex system, which creates non-linearity. So this all is like the child of chaos theory. Small things small changes can create massive changes. Massive changes might do nothing. You can't really predict it. Um, and consequently we're heading into a world of like massively ramping complexity.

And the kind of my argument here is that we're trying to govern these increasingly complex systems using institutional architectures built for a much simpler world. I mean, you just think about the your own governments or whatever. When did they last have an upgrade in the way that they made decisions? They haven't is the answer. Um, and to make things a little bit more spicy, we we're entering into the age of acceleration. So, uh, this little chart here looks like a straight line, right? Um, and what it is, it's the amount of like new inventions every 10,000 years.

Um, and this is a log scale. So the straight line means exponential. So, we've actually always been accelerating, right? You know, so we've been accelerating in technology. But, if you look back from like our first tools to when we got fire, millions of years. And around the Paleolithic era, we used to invent something like once every 100,000 years. Uh and now we invent like hundreds of things a day. Right? We we just at this uh apex moment. And um AI is an accelerant, right? It's uh take off time. Like it it's only going to get more complex from here. And the problem is we're not upgrading our governance.

And bit of cybernetics theory for you. Um there's a thing called Ashby's law, otherwise known as the law of requisite variety. And this is a guy who like mathematically was trying to understand how you can regulate systems. And he meant regulate in a kind of like more mechanical cybernetic sort of way. But, basically said like if you're going to try and regulate this system, then R needs to be at least as complex as the system that you're regulating. And if you don't, you get negative externalities. And so, we're basically using um governance systems that are too dumb.

And we did a lot of this in the DAO world, right? We we we said the DAO will do it. But, the DAO was just a dumb smart contract. Right? It just It's just a very simple voting system. Why would you expect complex decision-making out of it? And we didn't, right? We got a lot of bad decisions out of them. So, only uh complexity can absorb complexity. Um and I like this, every good regulator of a system must be a model of that system. And that shows you that the people who are regulating our systems are often very different from the system itself, right? We should do self-regulation.

Um and the thing is in a complex world, simple is a drug. We love simple. Sell me the simple thing, please, uh because life is much nicer if that's real. But, it isn't real. Um so, it just gets more of a drug. Um And that drug is fed to you by uh the totalizer. Um and this is like a imaginary force um I've I've dreamed up here. Um and you can see the totalizer wants to see everything, know everything, control everything.

Um and my sort of um argument here is it's kind of metaphorically, but the totalizer cosmic force of centralization and control. Uh and I I've been thinking like universally, right? That this feels like a reaction to complexity and complexity is everywhere in the universe. So, everyone's going through this on every civilization everywhere. It's just uh it's a thing. When complexity gets too much, um you start to try and control the public. Um this is Liz Kendall, who's the uh Minister of Science and Technology. Not scientist or a technologist.

Um and this was her announcing digital ID in the UK and she said the word control eight times in 1 minute. I counted. and they love this stuff. They're horny for it, right? They want um because they think that's the solution to all our problems. Take control. It's not us that's the problem. It's them. Um and so we're starting to see a rise in this kind of like total totalizing thinking, right? It's like a drug. this was an actual Instagram reel from uh Shabana Mahmood, who's now the home secretary. She got promoted on the back of this campaign.

We'll break the law and we'll take away the things that you love. And it was a kind of social credit system that would be work brilliant with digital ID. They can yank the pints off people. You know, like, I'm going to take away your trip to Benidorm and and and you like football, don't you? We'll have that as well. And um this is another quote from her. Um I'm a fan of Bentham's idea of the Panopticon, where the eyes of the state will be on you all times. Direct quote. Um she obviously never got the message not to go full panopticon. Uh uh Um, but yeah, I would call this big totalizer energy. And there's a lot of it about at the moment. And that's my my kind of argument here is it's because we're the the the world is spiraling away from us, right? Everything's getting more complex

and the then the instinct is to take control. So, I call this totalizer tooling. There's loads of it. It's hot right now. Raise loads of money on it. Um, KYC, digital ID, permissioned blockchains, CBDCs, orbs, of course. Claude I count as a totalizing technology. Wants all your data. Um, they're just going to rug everyone of everything at some point. Telling you. Uh, Palantir, Eth Al. There's just like loads of stuff coming out. We're talking about doing like full tilt facial recognition rollout across the UK.

Cheap. Actually, they'll be looking at it. It's like just CCD CCD chips and Claude. Like, you know, it's uh cheap. Um, so yeah, we're getting there. Um, the problem is is that this kind of like brand of governance always fails. Like, you know, all you have to do is like look at any moment in history when things have got too centralized, they've collapsed. Um, no, it doesn't stop us though. We just keep going for it. Um, so it's guaranteed to fail, but we're doing it anyway. And that's kind of where we are. Back to crypto. Um, we've sort of let this kind of happen ourselves, right? We um we let the totalizers win too much. Um, it's crime season, et cetera. So, that means shut down the DAOs. Like, who

needs a DAO anyway, right? Just slows things down. It's very annoying. Um, you know, scale back the security councils. Uh, let's just put the team in control and ship faster, et cetera. Um, and consequently, because the DAOs are a bit of a mess, they've just gone, "Ah, shut it all down then. And we'll just go like max centralized cuz now we can for a bit." Um, but the thing is in crypto that this kind of uh uh centralization reveals itself uh pretty soon. Normally by this guy. Uh, who robs your coins cuz you use a one of one multi-sig, etc.

Um, so fortunately, there's a counterbalance to this force. Uh, if you're a bit bit more hopeful. and I call this the sovereign. It's the opposite, the the yin to the yang, if you like. Um, there's also a desire to be left alone. And, uh, people like peace and quiet. And, uh, people don't like to be dominated by centralized, uh, powers and all that sort of stuff. Um, and it just so happens that the tools to have a sovereign existence are baked into the fabric of reality.

Um, so again, there's aliens elsewhere in the universe today using elliptic curve cryptography to probably do some money transfers and things like that. These are kind of you universal technologies. Um, so, um, decentralization is this other component, and largely my sort of argument is we should be doing more of it generally everywhere, right? Um, and you can think of decentralization as a kind of continuum. Um, so if you think about what it is like at the system level, um, you've got people like Kim here, max totalizer, I would say, in in the world, you know, top these like one mind rules everything. We've got a nice example of

it. This I'm less sure on China now cuz I think the UK might do this one pretty soon. Um, then you've got like autocrats, the military force, very hierarchical, very centralized. And then you can get into like big tech, which are like these massive companies that people are very scared of. Uh, a lot of the battle that we are getting, you know, chat control and and and largely the the state just wants the same surveillance database that they have. And they're feeling like not as powerful because they don't have it, which is why they're trying to get it. Uh, and then you've got like the EU, All on the centralization spectrum.

Um then you can go to like transnational corporations, getting a little bit more decentralized, like the Leviathan state state, the Americas. Um then this was with the UK was meant to be like the sovereign will break away from the E.U., do our own thing, didn't do any of it. Um but theoretically we are more sovereign now. Uh and then you get into like the super decentralized orgs like the Taliban, for example. Can't stop those guys. They're very good. Um and they're very good because they're decentralized. They you just like we just picked a fight with two more of these decentralized armies. Not a good idea.

Um and then you've got kind of where I think the DAO zone is. Um so yeah, I am saying the DAO zone is effectively the Goldilocks zone between Vitalik and the Taliban. Um but effectively like the in there you can you can design decentralized systems in very sort of like elaborate ways. We can do anything. Um and then you've got your like smart contract blockchains out to Bitcoin. Uh and then you get in like super decentralized like a religion. If you start a religion, it takes a few thousand years, but uh very decentralized org. Um centralized in some aspects, but it's a good way to scale an idea.

Uh and then you're into the mushrooms, very very decentralized. Can't beat the mushrooms at that. And then you basically just like there's nothing else, right? The at the end game of decentralization is basically death as well, right? Because nothing interacts with anyone. It's the heat death of the universe. So anyway, key point, every complex system that fails uh to meaningfully decentra- de- decentralize itself will die. Uh you just need to look at history, right? So basically all of these centralized uh systems need to be offloading um uh the power that they accrue to themselves out to a wider distributed set or they will collapse. It's just like it's a fact. And um we're showing the signs of of getting there.

So just to kind of finish off, like, um if decentralized ultimately and decentralization ultimately ends in death cuz nothing gets done, um it implies that there's a sweet spot. Um and I think this is uh we need to get a little bit more uh analytical about what we're actually trying to do with decentralization. So, um decentralization is actually super useful, right? Um it creates antifragility. It stops your system collapsing. So, uh system collapse is the thing that we should be really worried about. Um and if it's if you decentralized authority out to a large number of people and there's localized control, it'll last for a long time.

So, DAOs and decentralized systems are a way to make things happen indefinitely. Like, Ethereum will run indefinitely as far as we know. And that's the key goal in the new mandate, right? Walk away test. Um it's better for security. There was a few hacks recently. If they had their keys uh governed by a DAO, wouldn't have happened, right? They just over-centralized, left the keys too exposed. Uh it's better for security. Provides paths to meaningful control to more people. Uh and paths to zero control. I can incredibly say that I'm not in control of this, and consequently, it opens up permissionlessness. Can't have permissionlessness and central control.

Uh and we'll find out this, you know, with the whole security council incident on Arbitrum. This is the the equilibrium has been destabilized on on these things, and it's about to get interesting on that front. Um so, localization, context sensing, participatory governance, all good things that you get out of decentralization, but centralization's also quite useful. It gets done. Um a lot of the kind of uh burn down the DAOs is largely conducive to the kind of dysfunction of decentralization. And people just flip back to the other way around. So, I think we've got a, you know, elements of centralization make sense. Uh you can get you get control, predictability, allows you to compete.

So, one of the things I'm working at the moment, we have an LST which underperforms in the market uh at Jito because it's decentralized. We have 400 validators instead of one. So, we're competing with single validator LSTs. So, what's the balance? Um And this is kind of where I'm at at the moment and uh will be doing some work on over the next month or two. I call it pragmatic decentralization, which is to actually like look at the authorities and the real control. You have to boil it down to what do the token holders have control of? Uh what does the Dow Dow actually have control of? We can characterize these things cuz we're talking about literal functions on smart contracts. Now, the question is that some of those things now Solana is a

kind of move fast break things kind of blockchain, and it does. They'll They'll just throw in upgrades and then downstream that's just broke our protocol. Um so, we have to like go in and modify things cuz they've changed a digit somewhere or something like that. Um so, we have to actively manage the protocol. So, what I'm doing is is splitting this out into a dev council. Super centralized. We're trusting the labs team to control a certain set of authorities. Um but what I'm doing is like characterizing them in tiers. Tier one is never going to leave the Dow, and that's the authorities that are in custody of user funds.

things that are ossified. We don't plan on touching these things anymore. They're They're decentralized. Um and things that need stable maintenance. They're just there on the road to ossification. And things that need active management. Um and these will stick with the dev council. Now, uh so, what we're doing is taking power away from the Dow, which makes me very uncomfortable, but it's kind of necessary be- because we need to actively manage the protocol. Um so, we're building this kind of veto trigger, which means that any point the Dow can vote, and we build a pre-execution pathway for it to just yoink all the powers back.

So, this is the kind of like pragmatic decentralization. It's let's credibly think about the ways to safely trans transfer protocol powers to a trusted team. But make sure the token holders have hard powers to take over if if there's ever a defection. Um so basically in summary, the world needs decentralization to balance out the growing forces of totalization. Like it's getting a bit tasty out there. And generally my dream for crypto was that it's like osmosis distributing decentrali- decentralization out into the universe around the world.

So I I'm very convinced crypto will be the place we rebuild governance from first principles. It's just going to happen. Uh so don't give up on it. Please keep keep being passionate about it if any of you are. Uh but generally sort of broad guidance is like decentralize meaningful amounts of power. Uh make it meaningful. Make it credible. Um but you've got to compete in the market cuz there's no point in decentralizing a dead product. Like you have to win. So you're going to have to find the right pragmatic path. Uh unless you're kind of basically a protocol in which case you can do the kind of things that Ethereum's doing and and and go hardcore on it. But as a protocol above that you can you can play with it a bit. So but ultimately you've got to protect

hard rights for the token holders. Provide hard veto rights. Allow them if they turn up and decide that the that they can do it better than you, they've got to be able to do it. Uh and kind of resist this idea of just transfer everything to the devs. Okay. Thank you very much. >> Thank you, Nick. Thanks. I don't think we have time for questions. Next is Evo Matsec.