DAO Governance Sucks

Transcript

[Music] okay so uh this talk is uh on the basis of an ongoing research article I'm producing on uh on governance um it's been quite interesting but also extremely broad broader than I thought it would be let's get started so governance is everything crypto was sold is the magic pill right Dows specifically um as a disruptive no way to create Democratic uh communities in

crypto uh that can self organize through code and need no Central entity you know there's a lot of uh writing in governance uh comparing it to my Cal networks I don't know all the all these people are very green pilled so uh well yes uh governance in Dallas is supposed to be a positive sum game you know with my serial networks which uh if you're not aware they uh they're composed by of multiple nodes that uh feed each other other uh there are

multiple organisms that uh are independent but they Fe feed each other nutrients or whatever they need in order to all you know form a a larger network but in reality and that a is is on purpose like a in Spanish uh in reality Dows are extremely complex by nature uh and they go ahead and amplify uh many common issues in organizational structures Dows by definition have multiple different uh actors and that leads to large

overheads uh a lack of accountability and especially poor information flows so in reality they end up being like the honey fungus which goes on uh it's uh it's a parasitic fungi that goes ahead and uh clenches itself into a tree and ends up rooting uh rotting sorry the the root of the tree and ends up killing the whole organism um I mean decentralizing decentralization is gradual two crazy cats don't really need a dow uh those are two cats at Wonderland

two however many existing DS in crypto ends up getting stuck in poor practices and being unable to uh well to increase their decentralization over time uh for example one uh one pretty poor uh decision the decision made was to go for very monolithic and opinionated uh governance in early OD Dows which has led them now to uh be unable to to grow on chain and by consequentially you end up getting a lot of dows that end up moving

their processes both offchain uh yeah just offchain or or just use uh poor systems uh a quick note on voting systems not all voting systems are made equal uh we in crypto have ended up for generally simple voting systems where one token equals one vote uh and uh where for B and also for just binary decisions that being approve or or deny but uh there's a lot of uh let's say

unexplored uh voting systems in crypto that could be could be uh built uh that's a that there there's a lot of writing on this uh the one that's most well known is Aros theorem every voting system is flawed and uh prioritizes one prioritizes only some uh optimizes only some of the uh for some of the voters and not all so but all of these decisions you know haven't really been exploding in cryp we ended up with just

simple voting systems in many cases now the article that we'll be producing um I did explore quite a lot of uh governance in more mathematical terms uh generally this is the setup to then go on and explain that uh Dows actually end up being centralized in crypto uh you kind of I started out with minimum Quorum which is a um it's essentially the percentage uh the number of tokens it's a threshold of

the number of tokens that need to decide in favor of a of a proposal uh in order to for it to go through in a simple in these D simple voting systems and that's does this does the no okay I was seeing if the like the laser point of work that doesn't not um yeah that can be defined as the sum of the individual votes from 1 to K where K is the number of uh token holders and well I def kind of the the whales over the total uh or the the minimum number of

Voters that need to uh decide in favor as the number of whales as in in this uh cool but the problem with these voting systems is that most most uh token distributions end up following a parallel uh this is one such example uh uh you have some exponentials that's why it's a parallel and Alpha determines like the level of of a of inequality in a system where and and this uh this parameter is pretty high in in most of the Dow tokens and well parallels end up looking kind

of like this on or the cumulative so on both charts these are two uh two two instances of of of actual Dows in in in crypto and the x-axis just the largest number of token holders and the y-axis being the percentage uh that these holders uh have over the total number of tokens and most thresholds are set up either 10% or 20% and well in this one it's just one guy that decides so it's not really a dow is it and don't I'm not going to name them but there's some some food item uh

yeah and this one is just two so actually Dows end up being just oligopolies where just a few handful of people decide and the rest just hold the token and don't really have a say yeah you actually matter in in D and that's sad um cool now it's not just like it's it's it doesn't just have to do with you know one or two people holding 10 or 20% of the votes and Supply it also has to do with the cost of information um many teams that build Dows uh don't really pass through the

information to the community so they can make uh you know good decisions it's it's not really a doubt if if you know I have some some contracts on chain but I only know what going on and added to this the cost of voting is extremely high for many of these Dows so you can I wrote here a regressive voting system now the thing is if if you only have $20 worth of governance tokens and it cost $5 to vote why would you and so in the end only whales end up

really voting or engaging with the [Music] community now that's called rational regr and really it's what you end up seeing in in many any of these systems um you know it makes more sense to just not do anything rather than to become involved and you end up getting a very small number of Agents actually you know interacting in these governance systems and so that's where're in a in a dow you you you want to have uh engagement by the whole of the community in order to then you know understand what's happening and

be able to grow so limit limiting this to just a coring number of number of very opun people can you know have it have severe consequences with its growth um now we do have delegation which you know where token voters can delegate their their vote to uh a delegate um and that person can then you know make decisions uh and represent their their delegates however tooling isn't really quite there to

understand uh what deleg what delegates are doing uh the number of delegates is really small and they not only that but many delegates end up working over multiple different projects and there isn't really a tool to kind of understand whether there are any conflicts of interest in the in their in their delegation practices uh in order to see you know their whole voting history in an easy way and this is something that really needs to be built and this is just a note I added uh towards the end I spoke with multiple

uh with with multiple you know larger uh projects at and they kind of they well they said uh that the the um that was not only a way to kind of get the vote in cor for the people but to offload much of the legal troubles in whatever they're doing uh I should have said that in a different way but the problem really is that if you have centralization in the in the Dow governance token um it in the same way that shares in a company need to be uh disclosed if they if they are held by a single entity

with more than 10% of the vote uh in many countries uh encrypt it's a matter of time until you know you more more uh until Dows are recognized in a in a more uh in a more holistic way in legal system and really you end up just you end up just creating a security if a single person holds much of the vote there isn't really a dow there at all it's just some useless contracts that cost money on chain um cool uh

so yeah some takeaways so um for once I would say um that let's say existing contracts that are used by Dows really aren't enough uh to allow for gradual decentralization uh the open Zeppelin governor for example kind of sucks and even open Zeppelin has agreed to this trying to make it better and that's what most of most STS are using and the the uh Aller versions of argant or also

extremely monolithic and many Dows are struggling to then adapt and and grow that system out further it's extremely opinionated so really we need some sort of govern we need a better uh governance framework on chain that allows for you know less opin Unity code to be written and maintained and and and uh upgraded such that the core can also be upgraded without everything breaking yeah I don't know if there's any questions there will be there's there's definitely more to do and I will have an article soon TM okay first question

here hi I'm an open Zeppelin When you say the opens up and governor sucks I'm very curious to hear your specific feedback and who it opens up and already agreed with that cuz I need to have a word with them so and I agree we should be improving it but that doesn't necessarily mean it sucks what I mean to say it is quite limited and it's it's it's quite simple and like it makes sense for for uh DS to begin with but if you want to build you know over for example multiple uh in in a in a cross chain uh solution that opens UPL government relas an in off um the you guys do have a you have started a a group in order to

then like uh feedback I'll I'll give it a one to one it's so the well in the stuff I worked with the open Zepp link Governor the data was uh like it essentially just collects the number of the the voting weight over over time uh on upon every transaction and some there's some extensions to do

snapshotting as well but there there isn't like for example as said like interfaces to then uh do uh to have governance be natively cross chained so that you can have an L2 uh voting system that just relay the the um the voting decisions of the token holders back to L1 so that the decision can then be followed through I think that's one I I so that's a good point and it's something that we've worked with teams to like kind of have custom uh solutions

to it's something we've considered adding to the library but something we're we're trying to get into the future road map so that's good to hear for something that's more natively supported okay thank you and we got a another question there testing testing okay this works hi Rex thanks for the pr um my main takeaway from this was governance sucks because it's not decentralized enough right I mean that was my main takeaway from this because when I think of like successful dos I think of ens optimism I mean I'm thinking of it as like from a profitable stand of view right like ens is profitable they're making more money than they're spending right and so if in your eyes if or from what you've said

and I agree with by the way is Dows are like oligopolies right so you know only a handful of people make decisions so if that's a problem then I guess that means it's a problem on the decentralized point point of view right rather than like the profitability or the success of that Dow is that is that correct and if you if you do agree with that statement what kind of like would you mind expanding on like those kind of solutions like around that that was a lot of information yes um so I mean profitability is just one metric of whether a DA is successful or not I mean we do want to make money but

uh it's more that uh it's it's also a definition question in that um like what is a what is a DA isn't really black and white there's uh many of the ones you mentioned they have some aspects of decentralization but they end up working like corporations without saying their corpse uh okay H we have room for two questions more test test okay so I got a question

because you just said that they end up working as like corporations but not saying that they are cops why is it bad like why necessarily is it bad what is bad with corporations it's not bad but it's not a doubt so it's it's kind of a it's a false promise then more than anything else why well it's called a decentralized autonomous organization if it fails on decentralization or autonomy it's not a doubt I mean it is a spectrum yeah so like aren't public corporations

decentralized they like have thousands of shareholders usually and you know they are kind of like a dow simply they operate differently but in terms of decentralization I would say that I know many corporations that are more decentralized in many ds that's some i' love to look into and speak to you more about I don't have a direct answer right now time you have sure I mean you have to look at the decision making right like in the end in most of the corporations you have the system where the liability falls back to directors um so not the shareholders even so that's why from this this perspective it's

quite centralized and in da you don't have directors right like as an idea so it's it's not the dial is decentralized right yeah I think we can continue this discussion later like not taking time here on the on the lunch break guys we we have a question here hi uh so you are saying that if uh like initially uh it's not a good idea to launch a do then according to you like when is the right time to launch a do like if I'm protocol uh initially I don't have much users or uh maybe you know it uh it will make more centralized

but then what is the right time or what what what are the metrics which can uh led me uh lead me to uh decide what is the right time to launch a dow I would say my my argument is not on when is like the right time to launch a dow but that there needs to be a um a a process by which Dows can be gradually decentralized and do not stay stuck in somewhere along the way um it's more so that you know with Dows much much ends up happening uh just uh as as a dow expands for examp as many Dows expand

for example they end up creating uh just large Core Group core groups that are just governed by a multi um and and they enact their their decisions over the forums of chain and well all of this process I then cannot cannot uh be you know follow through easily and uh that that information flow rather that information flow rather than being offchain ends up you know only being held by a select number of people so you're saying like uh expecting some people who are public figures no it's that Dow's

um like when when is the right time to launch a token uh is a is a tough question and it depends on Project by project in in with regards to to uh for the decentralization um it it really should be done in a in a in a fairer way than selling all the tokens to the select VCS because then you you can end up having a a a more holistic uh voting Corpus that can help you uh develop the project further because other otherwise the

the the power of governance isn't really wielded but you know what icus are anyways dead so without VCS protocol can cannot scale right of of course of course it's it's not a binary of you know sell everything to VCS or sell nothing there's a there's uh there's levers so can we say like semi Dows are solution can be a solution where ini level of centralization at least we gave a start we uh once we have launched a semid then gradually we start we keep decentralizing with some key parameters exactly yeah okay thank you last one hey Rex over here uh you mentioned

uh conflicts with delegates um I'll be interested to hear more of your insights on that like what what sort of um conflicts are you seeing with delegates maybe maybe naming some um but then but then also what Solutions do you see uh like how do we resolve these conflicts with delegates conflicts of interest conflict of interests are plenty not even in DS but even in more centralized systems I mean in the UK government uh Mo many of the MPS don't really need to say that whatever they're uh or they do need to say but they don't on whatever their secondary jobs are uh which you know ends up deciding uh

What uh kind of laws they try to pass through Parliament um but that's going on a on a more of a on a tangent um some conflicts of interest I see for example are uh some um delegates that uh have power in uh pretty much every L2 now there is of course a positive sum game with l2s where you interoperability for example is is is great so to have a delegate that works over all of them is useful however there's also a lot of uh uh eating into each other's Market

because the number of users in crypto is still small and some delegates either don't don't take this into account or or don't aren't uh uh forward with uh with their decisions I know there's been a number of conflicts I just don't want to name I don't know if I'm to name people uh there was a there was some conflicts in uh in uh in retr pgf where the delegates indeed voted their own projects in to get more funds uh or voted a number of different projects in to get more funds I ended up

gaming the whole system and those conflicts of interest weren't uh raised initially okay thank you very much a big round of applause for Rex thank you Rex and thank you Wonderland for sponsoring this event and if you are looking for a job the Wonderland team is hiring so reach out to them and connect thank you thank you Rex