The Case for Hierarchy in DAOs - Joseph Delong

so my name's joseph de long joe delong and i am founder of a platform called astaria anastari is a like super capital efficient lending platform we've just gotten started so um i was formerly cto of sushi up until december and um we are um today i want to discuss about um uh why there should be hierarchy endows in kind of like this idea that structurelessness um isn't really a viable methodology here we go all right and um i am keenly aware this is deeply unpopular the idea of having a hierarchy in daos i think that like most people who have been in the corporate world have kind of experienced like what they consider middle management and have had a pretty bad experience but then um the um opposite side of that is kind of like um this like uh total um flat organization uh results in you know a lack of productivity overall for the organization and so um i also want to talk about the exception cases here um the ones that i'm talking about really are like highly performant dows that are building products that are shipping i'm not talking about like flamingo or people pleaser who are more like investment clubs ones that are um really um hyper in a hyper competitive landscape building products and shipping them like mega dow ave sushi uni etc um and the other exception to this is like teams that are early stage that are about five to six right we tends to be at that stage um you're able to have a completely flat organization where decisions are are made um uh by the participants but um as you kind of scale the organization you start to hit um you start to hit um barriers to scaling um just mostly around the idea that you can't round trip communicate with that many people right scales at like factorial for communication to message to everyone and get their input on an independent decision um cool cool yeah so these yeah these flat organizations um i i think um i kind of like blame our early contemporaries um in ethereum they wanted to they had this idea of decentralization where decentralization for protocols is really good um from a game theory perspective and they wanted to kind of communicate those same principles into organizations that these flat organizations are decentralized they're kind of like leaderless um and uh i have a feeling so far from like the major organizations that were early they've all kind of pivoted into a more structured hierarchy and that's the nature of flat organizations is that they eventually um they eventually go there and so there are two like key like i'd say reason categories for why flat organizations can't work and the first one is like psychological those are um ideas like uh how how you know how how to lead an organization you can't really um lead by committee things like architecture you can't really do architecture by committee either and the second are practical practical reasons like maybe legal or financial liability those sorts of things cool and so i think that what you'll find in the history of human organization is that there have always been hierarchies and places where they say that there is no hierarchy hierarchy tends to form and it really is about how uh transparent that hierarchy is versus how opaque um that hierarchy is and that's something that's covered in this um uh this uh great article um from a 1960s um a member of the women's liberation movement named joe freeman and she wrote a um essay called the tyranny of structurelessness and um she's got um some thoughts on structurelessness uh that i think are oh wait no that's not her that's her okay um i think that's probably more me at the top structurelessness is basically um this whole idea of having a structuralist organization tends to just form a an opaque structure inside of the organization because people are naturally organized that way access to resources um social hierarchy those things form naturally and wouldn't it be better if those hierarchies were transparent and not opaque uh-huh so this is um this is one that i would say is like um a hybrid of psychological um and practical decision-making it's really difficult to make decisions as a group especially once you scale past five or six and that's really has to do with communication your capability to stop and halt um all operations to make a decision and have kind of like all the voices heard of participants um this is where you see like organizations break down you need to move fast you can't really do that um when you're uh when you're kind of having to stop and make um uh when you're having to stop and make um uh considerations for everybody's voice to be heard um a another one is like uh hiring firing um evaluations access to resources these sorts of things that are um uh in that like decision-making realm are important things that need to be done correctly and being done as a group doesn't necessarily mean that um you'll arrive at the correct decision um widely like um i i see it all the time people are like um so and so is a great dev i you go and check their github never made a commit ever in their life and um and so uh it's just like kind of a becomes a popularity contest and the individual contributors are not graded based on their merit of their contribution but basically on their popularity and this becomes a problem when you're making decisions um the other thing i would say is like setting like a roadmap right you need some sort of leadership to help set a road map guide vision um and and direction ultimately there's going to be some authority that has to be the final say in a decision when there's um a disagreement ah yes so this was um me before uh i have since definitely changed my mind about this i think like working out in the open like this is really interesting and fun and it works early on and then fails to scale later on you can definitely build an org this way but you are going to have problems down the road um necessary secrecy i would say that this is a practical reason if i'm endow and i'm building like let's say a really um if i'm if i'm a dao and i'm building a really high-end product i want to have some competitive intelligence i um i want to not share with my competitors necessarily what we're building or and and that requires um uh we can't necessarily share every single bit of information with every person in the organization i touch on hiring and firing and performance earlier but like salary negotiations how are you performing these are things that are embarrassing for people those are not um things that we want to be shared out in the open publicly and it's perfectly capable or we're perfectly capable of designating an individual to uh evaluate and keep that as a private conversation another one might be secrecy is a legal reason or yeah necessary secret cities legal reasons maybe you are served with a subpoena and is sealed you can't share that with anybody i think this is a rosy idea that we have about daos is that just because we've created a new organizational tool that we don't suffer from any of the old world problems uh ownership how does a dow own anything i mean i can think of a few examples of dao's owning things but they're not owned by a dao i think that we'll get there um the laws in in wyoming etc are are improving for daos but um a dow is not a legal entity in most places and daos can't own things so if uh let's say that they own a domain name for their um application who owns that okay so now you have some legal entity and some people are owners of that legal entity and some people aren't um this becomes problematic and was required for there to be some sort of leadership to be the holders of that um i think of like numerous things domains discords hosting um you name it and uh all of that will get held hostage by um bad actors who are uh in the dow if there is not some sort of like hierarchical structure and a transition plan liability this is another practical reason without a legal entity individual contributors are liable for the actions of the dow and the things that they're building and that's going to keep you from getting the best contributors and if we're working in a super competitive space where we're all fighting for some of the best developers the best developers aren't going to come i experienced this a few times trying to hire developers and they're not being kind of like an org to put them under right they're worried about their personal liability for their actions within the dow so um and this is my proposal um for a way forward is to allow or to treat dowels as a organizational structure but people participate in dallas people aren't daos the dao is the architecture in which we cooperate with one another and inside that architecture we should make room for legal entities legal entities with their own hierarchy and this will allow us to develop excellent products that are able to compete in this um very hyper competitive space um that's all i have [Music] is there any questions yeah go for it i got what you meant about why uh uh a committee can't do leadership but could you say more about why a committee can't do architecture oh yeah totally um so this is like uh in engineering we talk about this a lot as they call it designed by committee and um kind of when you are having a group of uh individuals designing and engineering are a series of trade-offs and so you can never optimize on all individual axes but you'll have different stakeholders on the committee who will be trying to optimize on their preferred axes i think there are some instances when you can get like um a good version of this but generally this is like why um you know fortune 500 software is like pretty bad um is because they they kind of like participate in this design by committee um most like agile processes name a product owner and that product owner is the person who's making determinations on what um [Music] on on what uh parts of the product need to be built anyone else cool five minutes okay we got lots of time yeah i can rap no so yeah yeah so um yeah no this is just um kind of my my thought is that we use um that we use legal entities as a kind of a rapper and each of those legal entities um interacts inside of the dao and we use um i know token voting is just not a great design so far but it's all we have right now and each of those entities maybe they have some token allocation or maybe they are um uh you know uh like respected by the collective dow to give their input um and each of those are just like participants and we work kind of like inside the dao right it's okay to have a hierarchy um because it will allow us to ship products faster and like more competitive products does that cover that cool yeah that's a great question are there any dowels that i think are doing a great job of this today um and actually i have totally gone 180 on this um just based on my experiences um i think unit swap labs is doing a great job i think the way they structure themselves and set up to build kind of like products as a labs model and then kind of like deploy and have a um community who's managing that deployment use what labs is you know they're not beholden to the token holders right the token holders are doing um dow governance that's kind of overarching um the unit swap protocol and they are just product developers they're developing new products and uh asking the dao if they can launch them and accept them um the other one i would say maker before they dissolved their organization i think maker was doing a great job sure yeah that's a great question i actually meant to touch on it but i just like didn't read my notes um so uh this um this idea of like if you have to have a legal entity um that requires disclosure of who the individuals are participating in that legal entity so um i actually think anonymity is a knock-on effect of not having a legal entity right so like participants go okay i don't want to assume liability for this i'm going to participate under like an anonymous account or like a pseudo-anonymous account and this uh fails in so many ways because being anonymous is very difficult to do in public anything that you say um like the way that you spell gray or what time you start tweeting in the middle of the day right tells us a little bit about who you are right and we can kind of like narrow and narrow and narrow until um you know zack xbt is like you know doxing you um so uh yeah so so i i think that that's kind of the reasoning okay cool i gotta wrap it up but yeah that's my thought i think um uh anons are doing it because they don't have some sort of like legal protection cool thank you [Applause]