discussing governance & reputation w/ jack du rose from colony

Welcome to the Armada Podcast, where we are focusing in on DAO governance. Let's hop into the show. If you could start with who you are and what you're working on.

My name is Jack de Roos, and I'm working on a project called Colony, which is a DAO framework. We've been working on it for quite some time. We first presented our initial prototype at DEF CON 1. So it's been a labor of love over many years, and we're just about to launch what we call Colony V2. And we think it's basically the biggest, baddest DAO framework out there, and hopefully a lot of people are going to enjoy it.

Yeah, I mean, I, I'm certainly bullish on that. It's going to be awesome. It's quite the, the journey. If you could kind of take us back to what inspired you to specifically work on that so long ago.

Yeah, sure. So, back before falling down the Bitcoin rabbit hole, I was a jeweler. I used to make really high-end fine jewelry for really fancy people. And at some point, I was at a gala dinner, I think it was, for some art charity or something in Moscow. And a client that I'd got that was there was the daughter of this Russian oligarch. And she was asking me to help her with starting a company that would have very similar characteristics to my own. And I had a surprising reaction to it, a sort of really visceral negative reaction, which caused me to inspect why that was. And I realized because I really hated what I did. I hated it in a number of different dimensions, but the most salient one in that case was, I think, the— well, no, it was— I really didn't like what I was doing, and this was sending me down a path where I'd be locked into it forever. So I found myself back at my hotel trying to figure out how I could both not turn down the blank check that I was being offered— Right. On the one hand, and on the other hand, extricate myself from the mess that I was in the process of making of my life. So, I wanted to see whether it was possible for me to overcome the challenges of the way my company ran, which was basically as a very distributed organization. We were kind of competing with the high-end fine jewelry houses of the world, which are, you know, under the aegis of massive conglomerates like LVMH. And we were kind of competing with them on quality. And in order to do that, we had to work with a distributed team of elite craftsmen and suppliers all around the world. And that was just a massive sort of supply chain pain in the ass. And I was in the middle of all of it. And then also dealing with competing incentives, you know, people basically always looking to make the maximum money in the short term. Which ultimately could have detrimental effects to the success of the company and therefore their long-term profitability. And so I wanted to figure out whether I could make my own company become more self-organizing or indeed entirely self-organizing through software. And that turned out to be both an interesting challenge and much too big a challenge for me to tackle for my own little jewelry company. But But it was, it was the thing that kicked off my, my whole interest in it. And when I read Vitalik's white paper in December 2013, it kind of made all of the different things that were whizzing through my mind at the time just click into place of, ah, this is, this is the fundamental underlying technology that I need to make this work.

And just because I'm curious, like, how far did you get with doing that to the jewelry business?

Not at all. No, no, it was—

It was just sort of like a passion and thinking and like, how could I do this? And it didn't get anywhere?

Yeah, I realized that it was much too big a challenge to do for my jewelry business. And I naively initially thought that it was going to be just a case of knocking together, you know, like, a bit of Facebook and LinkedIn, and somehow I'd be able to make it work. But, you know, it's, it's a

Yeah, certainly. And so what do you feel like you've watched the most change since kind of embarking on this journey as it relates to DAO governance and governance more broadly?

Well, the biggest change is a very obvious one. It's the fact that we're having a conversation now which is about decentralized governance and decentralized organizations. And the first thing I'm having to do is not to explain what is meant by any of these ideas because nobody had got these ideas at that time. This was an incredibly difficult conversation to have. And I'd first have to just try and make people understand that the technological trends that we were seeing from Bitcoin and colored coins and the way that worked as a distributed incentivized network of work, of proof of work, like computational work, and social media and all of these different things meant that inevitably we'd get to a point where it was possible to have organizations that were entirely mediated over the internet, where it was pseudonymous, where it had got all of the attributes that you would expect from Web2, but where there's real money, there's value, people are coming together online to get shit done rather than just talk shit. And that was— That was so hard to get through to people. And now I just don't have that problem at all.

Yeah, I love that quote. It sounds like the foundation, the abstraction, I feel like when I first got into web development, my cousin taught me web development. He's about 10 years older than me. And, you know, the explanation that he sort of gave me of understanding exactly how a server worked and how the files were transferred and all this stuff, I grew up in the Microsoft FrontPage world and I thought about a server as just being basically set up for me through a hosting provider. Sign up for a Bluehost and drop some files and I have a website live. It was amazing as I got older and understood the journey he went through that I got to skip because all of the abstraction had happened for me. It sounds like there's a little bit of a parallel in this world where it's like, okay, now you don't even have to explain all that because now we can just start building on top of the abstraction. I think that it's a game changer. Because it opens up the degrees, the amount of people that can contribute to something like that. That's really cool to hear.

Absolutely. And there's actually a couple of layers of abstraction there as well, which I think was what probably made it so hard. The first was that if you're trying to explain to somebody what an organization is, or more simply, a company, what a company is, and you're having to do that outside of the context of a company, It's kind of almost like going into the, you know, Amazon and finding a tribe who've got absolutely no concept of what a company is, and explaining to them all of the complexity that constitutes the activity of a company, why people would be doing that, what the incentives are, even what are incentives in some cases. So you've got to explain all of this stuff without ever really using the word company, because they don't know what that means. And then the second abstraction which even if I could get that across to people was what a blockchain is and why this is important. And so, I would often find myself, even before I'd got to the challenge of explaining to them what a company is without using the word company, having to give them a primer on a blockchain and how blockchains work and why this is a relevant technology.

Yeah, absolutely. Not, not to get too Joe Rogan-y all of a sudden, but I imagine if aliens are interacting with us in any way, they got to feel a little bit similar. Like, how do I even begin to explain to you the layers of abstraction that society has to go through in order for this message to even be correctly understood?

Absolutely.

Yeah. What a, what a mission. One question that I feel like you're uniquely suited to answer that I've been dying to ask is I've been thinking a lot about DAOs and their decentralization. And obviously, there's a lot of excitement in exactly that future and what that vision and those principles and those values unlock. And then I feel like there's this other side of the world that I've sort of been poking and prodding around that I've been calling, calling like digital organizations. And what I mean by that is maybe not decentralized, maybe it's not particularly autonomous in certain regards. It's like this new way of using code to actually replace the way we use laws currently to just run kind of classic companies. Do you see a world where there are both sort of this DAO world and then there's these like digitally native organizations? Or do you suspect that DAOs are the thing that will consume, you know, that kind of open space and be the foundation?

That's a really good question. And we actually We've kind of been strong-armed into using the term DAO. It's not actually a term I enjoy because I think it's both too nebulous and too prescriptive at the same time, which is a weird position to find itself in. Right. But yeah, absolutely. I think that, you know, what we're dealing with is the emergence of internet organizations or organizations for the internet, which is a digitally native organizational form that may have many different control modalities within it.

Right. Yeah, that's the piece that I've been— probably my eyes have opened up the most in the last month or so of doing the podcast is that I just imagine that it's much more of a spectrum. And that I totally agree with you that the word DAO, I've even considered just replacing it with digital organization. That's sort of my pet word at the moment. But like, I totally— it strikes me that we will see all organizations being existing on some type of governance world something like Colony that is supportive for the full spectrum. And you might be 100% sort of decentralized and very protocol and currency-based and all the way to just, hey, this is a straight-up organization that is just like any old business that we've ever seen in the last 50 years. Is that kind of how you see it?

Pretty much, yeah. I think that there's enormous benefit that can come about from just being able to automate things like the accounting and HR functions that you would find within a traditional company. Exactly. That can, you know, really create enormous efficiencies that a lot of people would find attractive. I think that one of the interesting challenges of a digital organization is that in most cases, you encounter the challenges of scale and being able to coordinate people effectively over distance. And so, you know, obviously going back to the Coasean theory of the firm, the reason why firms exist at all is to decrease the transaction costs of the market mechanism for production. And I think that one of the challenges that uniquely kind of come through introducing the whole contracting side of things, um, in a very distributed setting is that like you can't avoid the market mechanism to some extent. And, um, so I think that any platform has really got to work to reduce those transaction costs of, of coordinating whatever is being produced. And, and I actually think that that's something that can be done in either a centralized or a decentralized way. But I kind of think that it's easier in a decentralized way, in a weird way, because— well, because I think that if you can sort of get an organization into a place where it can self-organize and people can be self-serving the work, and basically authority has been distributed to the edges of the organization rather than the information having to flow up and down. the organization all the time via layers of management, then I think that those organizations will ultimately prove more agile and more able to adapt to this sort of digitally mediated organizational form.

Yeah, I totally agree with you. It reminds me a little bit of— this is not the right analogy, but sort of like poisoning the well, right? Where it's like, there's this element of once you put poison in the water, it's distributed, and then it feels like this is the opposite of that, which is like if you aren't distributing that poison— this is a terrible analogy— but it needs to go through the entire body of water in order for it to really have the maximum impact. Again, terrible analogy, but I think that it's in reverse of saying you have to have that decentralization across the entire organization in order to really tap into these compounding benefits is what I'm hearing you say.

Yeah, I think so, sort of. But then I also think that Within any organization, you just probably need even within one organization to have different forms of decision-making that are able to take place. And that may be from it being actually very centralized in the authority, in fact unilateral authority to act in some cases, most likely at the periphery of the organization if we're talking about a primarily decentralized organization, so that the individuals who are perhaps within teams have the authority to act and make payments and get stuff done without going through any bureaucratic voting process. But higher up in the organization where you've got more people or, you know, just bigger amounts of money, more things to worry about, things going out of control, there you would probably want to have the greatest oversight. You would want to have the slowest processes, the most secure processes. So rather like in a traditional firm, things tend to get slower as you go up the hierarchy because you're going to fewer people with more things to look at and less bandwidth. In a decentralized organization, I think as you go up the organizational hierarchy, you'll likely have— more people who it just takes longer to coordinate around a particular decision.

I see what you're saying. Yeah. So my analogy, as bad as it was, was sort of assuming this sort of horizontal existence and sort of all-encompassing. And what I hear you saying a little bit, there's this vertical element of it that is important. And I'm thinking a little bit of traditional organizations around a board of directors is thinking probably much more on a quarterly or annual or even 5 years, 10 years type timeframe. CEOs, somewhere in there. As you're moving down, there might be an engineer who's literally making a decision on which bug to fix that day. And then there really shouldn't be a process to have to go all the way up and down. And so there's a vertical element that I'm hearing in your kind of mental model.

Absolutely. Yeah. Because I mean, I think that's just a fundamental requirement of companies, really. I think it's been really interesting to me to note that in the various DAO paradigms that are emerging, by and large, there's really a homogeneous set of decision makers within a single organizational blob. And there's, you know, these are deemed to be fungible, well, really by virtue of them using a fungible token to determine voting weight. But it seems patently obvious to me that that's not a scalable structure, because it is so homogeneous, and you're expecting everything to go through, in and out of the same kind of process. And, you know, that's just not how companies work. Companies are structured in departments and teams, or circles, or groups, or whatever it is, you know, you want to use in your organizational terminology. And it seems just obvious that decentralized organizations have to obey the same principles. And I think it's doubly interesting when we have got this, you know, real obsession with scalability of blockchains and how we've got, you know, sharding as being the solution to so many ills and indeed other solutions that sit on top and kind of check back into a main chain through one consensus mechanism or another, which is, you know, creating this really interesting and heterogeneous system. Because a blockchain, a single blockchain, will find it very difficult to reach the kind of scale and speed that it needs to. And yet we expect human organizations to do so without having similar structure if it's on a blockchain.

Yeah, I love that. I love that thinking. I, uh, it's probably been my probably number one or number two observation coming into this space of sort of not fully understanding the DAO space and realizing, to your point, of like how in certain regards it's quite centralized. And, you know, having a handful of token holders that predominantly own a vast majority, like the structure is not as, as DAO-y as I thought it would be. And so watching organizations like Colony and these other tools out there that are that are thinking about adding some height into it, on one hand feels like, oh God, like we're recreating the centralized mess that we created before. But on the other hand, I think there is just a fundamental efficiency and entropy in being able to have that organization to deliver their broadest vision as quickly as possible. And that's how I see and hear that.

I would actually disagree there. I think that There has— there's a sense in which, uh, distributing tokens results in a decentralized organization. And I would argue that that's actually practically untrue, and that for those who are really doing that, it amounts to little more than decentralization or regulatory theater. Interesting. Because there's, um, there's a requirement for that to be— or it would seem there's an emerging requirement for, um, the decentralized governance of a tokenized crypto system To be in place for it not to be considered a security in the US, which is where most people care about. But I actually think that the, you know, whilst you might get a growing token holder base in the long tail, the bulk of the token holdings really end up in the hands of those who were earliest and the largest investors. And they're the ones who are really in control, and it's going to tend towards sort of ossification of that situation over time. And you know, I think that that runs precisely contrary to way to the way real organizations work, real companies get started and grow, which is that the the founding team are really just the people who happen to be paying attention at the beginning. And it's not until companies really start to see some success and they grow and they, you know, raise a whole bunch of money or, you know, just become very profitable that they're able to attract the very best talent. And that very best talent ends up in the highest positions of authority within the organization. And, you know, a token-based governance system really makes it very hard for those latecomers who are probably more competent to get a look in in on the governance of it. So you're kind of stuck with the people who were there at the beginning. And to my mind, that's not very decentralized. And so, and in Colony, I think it's precisely the opposite. When I'm talking about there being a hierarchical structure, I don't mean a hierarchy of managers in whom authority vests in fewer and fewer people the higher up the organization you go. I think that that vests in more and more people. the higher up the organization you go. And because we generally promote the use of our reputation system as the way to make decentralized decisions, that doubly increases the decentralization that is possible to have, because when reputation is earned within a colony, it's also subject to decay. So Everybody's reputation is decaying according to a half-life. And that means that the contributions that people make and earn them reputation, it means that the reputations of people normalize over time to represent their contemporaneous contributions to the organization. So, if you are, I don't know, an ex-VP of Google or something like that, and you decide that you've had enough of the world that that kind of technology is taking us into, and you want to start contributing to a more decentralized future, then you will have the opportunity to leverage your skills and get an appropriate degree of compensation and influence over the organization according to the skills that you bring to the table. Because of that reputation decay, because everybody's reputation is decaying, And you're earning yours. And so everybody's is going to be normalizing.

I love what you just said there. You did a couple of things there that I really enjoyed that I just want to go back on a little bit. You redefined/defined a couple of those key phrases, sort of like what hierarchy actually means and what governance is. So I totally buy the way that you described it with those kind of definitions. And I think going back to DAOs, it's like a lot of these words we have a lot of connotations to and definitions that are probably all slightly changing and sliding. So I appreciate you kind of walking us through that framework. To me, it sounds like the reputation piece of what you guys have built is really a key component of being able to do any of this. Is that true?

I think so, yeah. I think reputation is really key. And I think teams are key, or domains, as we call them in our whitepaper, as something of a very generic term. I think that if you can compose an organization out of a bunch of different units of team structure or departments or projects or whatever you want it to be, and you can allow people to be earning reputation specifically within those domains, then you can create something which, you know, resembles the heterogeneous skill sets that are required within a normal organization that has, you know, they've done pretty well for us over the course of human history in a decentralized context, which is trust minimized and which is pseudonymous.

Yeah.

I think that their reputation, reputation is kind of inspired by credit scoring in a way. And it's kind of the way we also think about identity in Colony. Like we really don't care who people are. All we really care about is what their behavior has been in the past. very much like a credit score. Your lenders don't care what your name is or where— well, perhaps they do care where you're from to some extent, but primarily they care about how you've behaved with money over time.

Yeah, I love that analogy. Certainly an area that seems like a big problem across the broader Web3 space. And my intuition is like, oh, you need a you know, this credit score or this reputation layer across the entire Web3 space. And then obviously you have Domain over Colony and be able to build that in. Do you see this as a thing that is deployed in sort of each, you know, world, or do you think that someone and somehow there will be some layer that sort of connects your entire broader Web3 at your discretion?

Yeah, I think that the idea of A universal reputation system is sort of a simultaneously terrifying and improbable prospect, because I think that being able to quantify and evaluate behavior is an incredibly contextually sensitive process. So even if we're thinking about reputation as it exists professionally and And between organizations, I think it's very hard to build equivalence in different contexts. So I think that trying to, or even the desire to build a reputation which equivalently captured both your financial behavior and your professional behavior would probably be rather counterproductive and undesirable. Because I can imagine many people who are extraordinarily competent in whatever work they do, but disastrously incompetent financially, and they're kind of being forced into lumping all of these different aspects of their personality into this effectively single number that evaluates them as worthy or unworthy. And I think we need to be able to separate these things out. And so in Colony, really, reputations pertain only to a single organization. So you'd have different reputations in all the different organizations. You wouldn't be able to carry one reputation from one to another.

I love your thinking on it, and it's awesome to hear how you think through that. And the part that I'm slightly struggling with, and I'm just curious on your challenging thoughts on this, is we do that naturally as humans, right? You say, hey, I did X, Y, and Z, and now I want to come over and do Project A. Humans, I think, naturally sort of assign like that project went well, that project didn't go well. Like, here's what I heard about you in that project. They basically are mentally doing that as it applies to, you know, writing a venture check or hiring someone, whatever that process might be. So we do transfer reputation from one area to another. And I totally agree with you that the context is really, really important. But how do you think about that process? Because it does allow us to move really quickly. And you see that with signaling, you know, from Ivy League schools and the Googles of the world. Like, do you think that's an outdated thing? Or do you think that it's just caught on the edge of this transition to this new reputation world.

Certainly not. I think that actually the fact that we need to explore this question just demonstrates the insufficiency of language or the vagueness of language that means that we don't have exactly the right word for this notion of reputation. Which is just, you know, an extremely broad one. And I, I kind of think it's, you know, we see a lot of conversation in, um, in DAO space now about the difference between soft consensus and hard consensus. And so you're, you're building soft consensus within an organization by talking to people, and you can get a sense of, of what people think about any particular proposal. Um, and that really informs whether or not it's likely to get to the position of being offered for hard consensus through some on-chain process. And I think of reputation very similarly. And when I'm referring to reputation within Colony, I'm referring specifically to that measure which is consensus relevant, that which gives you direct consensus relevant influence over decisions. I'm not talking about the fact that you, you know, spent 5 years at McKinsey and then 3 years at Google and whatever, you know, and you went to Stanford and all of those things, which are of course always still going to be valuable, necessary, components that make all the difference, just like an interview and references and all of this stuff, as to whether somebody is likely to want you to contribute within their DAO. And that may also influence the amount of compensation the DAO ends up agreeing to give you, which will also influence the amount of influence that you've got within that DAO. Right. And so all of that reputational stuff is just a different conceptual bucket than the hard reputational stuff of the points on the blockchain that give you the ability to opine with authority.

Yeah, love it. You added another definition to my vocabulary of soft reputation, hard reputation. I think it's a great mental model to apply.

Yeah. I think that in general, there's There's kind of almost a— like, it's somehow hard to bridge the gap between thinking about pseudonymous agents on the blockchain doing stuff and squishy people having conversations and liking things or not liking things and having bias and all of these kind of things that go on off-chain. And I think we just always need to remember that all the rules, all of the— everything that takes place in ordinary organizations, Zoom calls, they're like— everything still applies when we're talking about the DAO. It's just that basically the paperwork's more digital,

right? Yeah, love that, love that. Um, so I was reading through the roadmap, and obviously you guys are on the verge of this big release, but, um, what are you most excited

I'm most excited, I think, about getting governance out finally, because we've been noodling about with that for such a long time and getting all of the prerequisites in place that— well, they're prerequisites, they need to be there in order to enable us to do decentralized governance of the sort that we think is necessary. And that's actually only just a few days away. I think we're probably a week out from deploying that to our beta testers. And I think of all, that is the biggest step change for us. And I think it will be the thing that opens people's eyes to what a new, better way of doing decentralized governance can look like, rather than putting everything I'm

excited to see that, and I think a lot of listeners will be as well. I do appreciate you for taking the time to join the podcast. If you could remind people where, where to check out your work, and, um, thanks again. Yeah,

it's been a real pleasure. Um, I would suggest for people to, um, come along to our website at colony.io, follow us on Twitter @JoinColony. They should also join a colony if they can. And, um, they should also join our Telegram or Discord. So I think our Discord is clny.io/discord and Telegram is t.me/joincolony.

Awesome. Thanks again.

It's a pleasure.

Thank you for listening to the Armada Podcast. We would greatly appreciate a review in your favorite podcast player. And if you want to stay in the loop even further, head to armada.fm. Where we have a repo. You can check out what we're doing and what we're learning along the way. Love to see you there. Thank you.