Niran Babalola ~ Slate Governance for Effective Token Votes

[Music] all right thank you everybody for coming my name is Neron babel Ola I work at consensus on the diligence team where we focus on making sure that smart contracts are safe enough for people to use in particular I work on a product called pen valo and what we're trying to do with pen bala is build a sustainable flow of funding for work that makes aetherium safer that work currently has a difficult time getting funded and we want to make sure that it's easy so how about that all right so basically I'm going to start by talking about the problems that face decentralized governance systems go into a couple of systems that have been deployed on the blockchain that people have governed and then I'll give a high-level overview of Housley governance works and dive into how we're using it in pan VAWA in particular so the first problem that a lot of systems end up having to tackle when they're trying to govern something on the blockchain is addressing 51% attacks whether you're dealing with hash power or token voting if you can get 51% of the economic power in a system you can effectively do whatever you want now in the real world if you are trying to govern a company and 51% of the shareholders tried to take a hundred percent of the assets of a company you would go to jail it's not an area where code is law it's an area where law is law and the directors of a company have a fiduciary duty to represent the interests of all the took all the shareholders not just 51% so there's a centralized backstop that prevents malicious proposals from going through but in smart contracts you don't have that centralized backstop code is law for better or for worse so a lot of the design decisions that people make are trying to accommodate that and make sure that to 1% attacks aren't successful another problem that they have is decision overloading if you're trying to decentralize the governance of a system that's making a lot of decisions a lot of the token holders are gonna have to be active all the time to make all the decisions that need to be made and sometimes the proposals that get made aren't even relevant they're effectively spam but to prevent them from going through you have to show up and vote to make sure that the spam doesn't go through so that's another thing that creates the pressure for a lot of design decisions that people make when they're governing these systems structuralist nests is a problem that they run into when you're not a whale in one of these ecosystems where you're an insider who's involved in every decision but if you're a small token holder and you have a proposal that you think is good for the system there's no clear path to make that proposal get through it's it's kind of hidden we've defined decentralization in a lot of circumstances to me and chaos so there's no defined process for you to know how to get the support of 51 percent of the token holder so you could get your good proposal approved so rather than distributing power to all the participants in the system power ends up being centralized with the active participants with the most tokens that they hold deadlock sure a problem that occur when you're not just dealing with simple binary decisions where we're trying to choose between path a or path beam but in the real world there's so many decisions that we can make together and so many options that we can choose from one example of this right now is brexit in the UK there's no majority to stay in the European Union there's no majority to leave the European Union with no deal but there's also no majority for the deal that's actually on the table so they're stuck in a deadlock and nobody really knows what the path forward is avoiding and building governance systems that avoid deadlocks helps cooperation continue over time and we need to design systems that can achieve that disunity becomes a problem when you're not sure that your rights as a minority token holder will be respected when 51 percent of the participants decide something else if you think that your rights won't be respected you're probably not going to join the system but the flip side of that is if it's so easy to leave the system when decisions don't go your way then it's easier to split than it is to stay together and keep on cooperating so systems need to strike a balance between making sure that you're not tied in to decisions that other people are making for you but also to be sticky enough that people keep cooperating over time and lastly cartel resistance is a big problem that I don't have enough time to go to in detail but basically if you were trying to use a token governance system to for instance allocate the block reward from a blockchain the optimal strategy for someone proposing how to use those tokens that they're getting as to bribe as many of the voters as possible basically just pay them off and you end up not achieving what you set out to do you were trying to get those tokens to go towards work that improve the network but instead it's either going back to the voters or going to the pockets of the cartel that's formed so the first system that I'll talk about is the down I'm sure a lot of people are familiar with it down in mid-2016 the etherium network had been live for less than a year and a lot of people were really excited about the potential to create new kinds of organizations that didn't have CEOs they didn't have officers at all they were governed from the bottom up so one of the first design decisions that the Dow made was proposals would be submitted to a set of curators who could veto any proposals before they got to the token holders so if you take the 51% attack problem if those curators see a proposal that's trying to take a hundred percent of the assets of the organization and give it to 51% of the token holders they can just veto it and never present it for a vote so that doesn't become a problem when you have that spam problem with decision overload the curators can do the same thing they can decide we're gonna hold off on these things that you're proposing because it's not worth the attention of the token holders so that design decision solves two of those problems the other design decision that they made is the losing side of a vote can always exit the DAO and split everyone else who wants to split ends up forming a new Dao that can go on and continue cooperating with the subset people who are on the same side of the split it's you so that makes it very easy to decide to participate in a system like this if you know that if a decision doesn't go your way you can always take your ball and go home then there's no reason to not join in the first place except for the fact that the dow had a bug it was a small problem but if you're not familiar with the backstory of the dow i highly recommend reading the wired article that takes you through the story it was kind of like living through a science fiction novel you should really be familiar with that story so after the dow failed Eragon kind of picked up the ball of the enthusiasm behind these decentralized organizations and built a framework for creating new kinds of organizations it's effectively a toolkit for building new organizations so the most interesting aragón dow right now is the dow that they're using to govern themselves through what they call the Eragon governance process or a GPS so the core of Eragon has a feature called access control lists and that's the method that they delegate different rights and responsibilities within their organizations so to decide the decision that they made for the AGP process is that they would have a GP editors which are basically two people that are the only ones that have a right to create a proposal for the token holders to vote on so again if you have a 51% attack proposal that's trying to take all the assets and give it to 51% of the voters the AGP editors can just say we're not gonna ever present that to the voters because it's a bad proposal and if there's spam they can also say I'm not gonna include that it solves those two problems but again there's no one Aragon way to solve the various governance problems it's really a toolkit for creating your own organizations that solve these problems so if you're trying to experiment with decentralized organizations you should really start with building on top of Aragon they've solved a lot of problems for you and it's structured in a way where you could plug in which you're trying to innovate on without having to resolve all those problems from scratch use Erika so Moloch Dow is one of the interesting experiments happening in right now it launched on the maintenance two days ago and what they're trying to do is build a grant making Dow basically a Dow that raises funds and starts spending that those funds on things that improve the etherium ecosystem so the way that it works is the existing members can make proposals to add new members and adding new members is basically just giving them voting weight in the system new members are either doing work so they're getting voting weight for free or they are putting in money and they're getting voting weight because they're contributing to the organization so to make one of those proposals you have to put down to ten ether deposit to prevent spam and if again just like the Dow if you're on the losing side of a vote there's a waiting period before those proposals will take effect so if you don't want your assets to be taken into that decision you can take your assets and go home it doesn't create a new down like the doubted so hopefully it'll be a little more secure but since you're never a baited to abide by the decisions of the majority it really makes total sense to participate from day one so I highly encourage people to participate in this molecule experiment there's really no reason not to except for the fact that you might lose your funds in a smart contract bug but that's a risk that we all take when we lose these systems put in however much you're willing to risk two smart contracts and have fun with the experiment that's going on lastly maker Dow is the Dow that operates the dice table coin that hopefully a lot of you have played with here at the conference they have a couple of different types of governance proposals but the one that's most interesting to me is their executive proposals that actually take effect on chain to govern these they use a system that they call continuous approval voting so you can make as many proposals as you want during a quarter but at the end of the quarter the proposal with the most votes is the one that's going to take effect so if there are seven different proposals and you think three of them are okay and four of them are total crap then you vote for all three of the ones that you approve of and whichever proposal ends up having the most votes is the one that's going to take effect so this is interesting to me because it's very similar to what we're doing with slate governance so a slate governance the the hypothesis behind it the reason that we built it in the first place is because our community strongly believes in decentralization and decentralization is a good thing I'm not anti decentralization but a lot of people don't have the same definition of decentralization if you ask two different people they're going to give you two different answers the only thing that we agree on is if we see a system where chaos reigns then that's decentralized but it doesn't have to be that way you can to me what decentralization means is that you have an emergent order rather than a top-down order and if we can build these centralized systems that facilitate that emergent order we could probably get a lot more work done together than we're currently getting done right now so at a high level the way that slate governance works is that there's one slate of proposals that can execute each quarter slate can have as many proposals as you want to include on it and there can be as many slates as you want competing with each other to be that one slate that's going to execute that quarter but at the end of the day the one with the most votes is the one that's going to execute secondly if there's only one slate then there's no vote at all if there was no competition to be that slate that's going to execute this quarter just execute the one that seems to be the consensus of the community and move forward without a vote third the identity of the recommender of each slate is emphasized in our voting process instead of it just being a purely technical decision we care about who's making the recommendations for these slates and our systems and last anytime you're making a proposal you're recommending a slate of proposals you have to stake tokens on that slate and any of the losing slates for that quarter are gonna lose the tokens that they staked on that slate so that's how it works at a high level and the outcomes that we desire from that is that really consensus is built before votes ever happened if you look at a lot of the existing systems where decentralized governance is happening the conversation really starts when the proposals hit the chain that's one view trying to figure out what they're gonna vote for and how things should proceed we don't think that's the right way to do things we think it should be more like the philosophy behind plasma where you set up a set of rules where you can coordinating and you only go back to the chain when something has gone terribly wrong that's when lots of transactions get sent and pen Bhalla if there's just one slate that everybody agrees on that's the normal case scenario one transaction got sent you had one slate for that quarter it gets executed but when something goes wrong when someone submits a slate that wasn't actually the consensus of the community that's when more transactions are going to be sent to recommend more slates and that's when hundreds or thousands of transactions are going to be sent to vote in the system but that only happens when stuff goes wrong not in the normal case we won't consensus to be formed off chain not on so in the system we rely heavily on incumbency so when we're deciding which slate to vote on in the app we highlight who the incumbent recommender is if you recommended the last slate that was successful we the next time you proposal slate we're gonna make sure that the community knows that you're the same person that recommended the last slate and that's important to us because it establishes a different status quo from what usually prevails and decentralized systems and most systems the status quo is stasis it's hard to make changes in most decentralized systems because by default nothing happens but in Slate governance there's a different status quo the status quo is continued progress along the same trajectory that the community has been moving on because when voters are comfortable with the status quo they vote for the incumbent not for nothing we think we can get more sustained to progress when the status quo is progress not stasis secondly we rely on a top to run off when there's more than three slates present for a quarter so and then by default if you're just going to vote on slates and there are seven slates and a quarter you could get a slate that wins with 30% of the vote and that might not actually be an accurate representation of the consensus of the community so one way to address that is doing something like Instant Runoff voting where you just rank all the slates and have the algorithm go through a run off one at a time but then you get into that same scenario where you're reaching that consensus on chain of which late was actually the consensus of the community and that's not what we want to happen we want people to build consensus off chain and then publish it to the chain so that's why we use a top two runoff so when you're voting you select your first choice of a slate and your second choice of a slate and we count up all of the first choice votes and if one of those slates gets 50% or more then that's cool that one wins and there's no runoff but if not if none of the slates get 50% of the vote that's when the top to runoff kicks off we eliminate all of the slates that didn't get either the top amount of first choice votes or the second top amount of first choice votes we take all the ballots from the people whose first choice was eliminated and then we count up their second choice and that's how we determine who wins out of the remaining two slates so we think that kind of coerces people into reaching consensus off chain rather than relying on the outcome of a vote to determine what the consensus actually was this the solution isn't a cure-all it doesn't solve all the problems when it comes to splintering in a decentralized system it's really similar to how the president is elected in France they use a top tier enough and in their last election there is a candidate on the Left candidate on the right and one right down the middle and the voting system did its job like it's dear to people towards the middle ground candidate but they also have riots in the street every weekend the voting system doesn't guarantee that you're going to be able to build unity over time but the what we're using it for is to make sure that consensus is built off chain rather than being built to on chain so of those governance problems I outlined to begin with 51 percent attacks and decision overload are very commonly solved almost every one of these systems solves that so does say governance for structuralist instead of not knowing where to go if you're a small token holder with a good proposal you know you can go to the incumbent recommender and figure out how to get your proposal into their slate for deadlocks we try to address that with incumbency we went there to always be someone who has a mandate to continue to make recommendations but that doesn't mean that's gonna solve all the problems again going back to the bridge that drags the example there's an incumbent there's a prime minister with a mandate to lead but there's still a deadlock they're still stuck so slate governance also does not completely solve the deadlock problem but I think we can't solver for disunity we still have that problem where it might you might be reluctant to enter a system governed by state governance if you think your rights won't be respected if you put your assets in and then 51% of the voters decide to do something you didn't like then you're kind of stuck so it doesn't solve that problem directly and cartel resistance it also doesn't solve that if you try to use slay governance to govern the inflation of a currency for a blockchain you're probably gonna end up with the cartel they're gonna put that money in their pocket and you're not gonna achieve the goal so we don't solve those last three problems in Slate governance itself we solve those problems by choosing what to govern we believe that instead of governing assets on chain you should govern loyalty we're not pooling some funds and a smart contract and determining how to govern it we're governing how we interact as a community together and to explain how that works I'll dive deeper into how we're using Slate governance in pandava in particular so what pan bola does is it uses tokens to organize people who want to make aetherium safer right now there's work that needs to be done to make sure that smart contracts do what people say that they do and we also have a vision that we've telling people do we have a vision that we've been telling people about what aetherium is and for it to fulfill that vision the network has to scale so what we're trying to do is organize people who want to make sure that smart contracts are secure and people who want to make sure that the network scales and get them to support the people who are doing that work so the core problem that we're facing is that the ecosystem funding is unsustainable a lot of the common infrastructure that we all rely on as being funded by one team or another that raised funds in a one-shot deal and they're spending down those funds and eventually they're gonna run out the second problem that we're facing is that the incentive system that allowed us to build aetherium in the first place it doesn't really work for improving the system over time it's easy to get people to start a new network but it's not easy to get people to improve an existing one so people with ideas about how to scale the network instead of cooperating with the existing community sometimes they splinter off create their own new incentive systems and build something separate that competes if we want to be able to make more progress together instead of splitting apart we have to solve this problem the way that Penn vollis else's problem is by wrapping the existing ecosystem of grant funders corporate open source projects and volunteers with the token that gives them a shared incentive to find sustainable funding together so we start off the system with just a plain old er C 20s token and we take half of the supply the fixed supplier of that token and we put it into a smart contractor that we call the token capacitor and what the token capacitor does is it releases tokens at a decaying rate over time for token holders to distribute the way that they want it every four years the amount that you have available cuts in half just like the Bitcoin block reward so every quarter the token capacitor releases some tokens and the token holders decide what they're going to issue tokens to reward they're gonna find people who are doing work that aligns with the mission and they're gonna issue those tokens to the teams that are doing the best work but those workers with those tokens in their hands that doesn't really give them anything if those tokens aren't useful for something else and what the what those tokens are for is for donors to buy tokens from them and deposit back into the system so Penn vallah isn't a system that controls your assets you don't put dollars in Panvel you don't put ether and Hanbali you don't put diet instead when you want to donate to the system you buy tokens from someone who did work and earn tokens for that work you give them dollars and you deposit the tokens that you purchased back into the system to keep the system issuing grants over time so that's one way to get tokens to go back into the system and complete that cycle but pants Allah doesn't need your money what it really needs is your power we want you to use your loyalty to make donations happen to pan Bala and the way that you do that is being loyal to the merchants the companies and the individuals that make donations to pan Bala there's a lot of money that companies currently set aside whether it's your local coffee shop your cloud services or your favorite luxury car brand they have a marketing budget that they spend to get you in the door and they're giving that money to Facebook and Google to make that happen with pen valo what we can do is instead of spending that money with Facebook and Google and making them richer we can get companies to put money towards a goal that we care about to fund work that we care about and go into the door for that reason we want to give the same return on investment that you get for paying for ads from Facebook and Google but for putting donations towards a cause that we care about so the more demand there is to make contributions to the system the easier it is to fund the work that we're trying to do together workers are getting tokens from the token capacitor donors are buying tokens from the workers and getting those tokens from them and then they put those tokens back into the token capacitor to complete the cycle our goal is to reach equilibrium for every token that's going out of the system as a grant we want one token to be going back into the system as a donation and when we reach equilibrium that's a level of cooperation that we can sustain indefinitely there it's not that people are speculating on this token hoping that it'll increase in value it's that we've organized a community of donors organize a community of workers and brought them together so they can fund the work that they want to do together so when you have a system like this that can fund work how do you govern it that's what we designed slate governance to do and it's already in progress we released the first batch of token grants on February 1st and the way that we did it was we started accepting Apple patience from teams in the community we put together a packet of all those applications and we created the pen valo awards committee to make the first recommendation for a slate of grants to release so we sat down and we had a meeting which we recorded it's on our YouTube channel the pandava YouTube channel so you can watch how it played out at the end of the day they recommend Addis late of grymps we we had applications we had 12 applications from 10 teams and those applications requested 2.6 million tokens but the token capacitor for the first batch for restricted us to two point 1 million tokens so the applicants had two options they could decide to reduce the amount of tokens that they're awarding that they're requesting to make sure that every application could fit in under the limit or they could hold the line and say I'm not going to reduce the number of tokens and I'm requesting I'm willing to see some of the people be bumped out even if it's me but what they all decided is that they would reduce the amount that they requested and we were able to issue grants to all of the people that applied and batch 2 we won't have 2 point 1 million tokens we'll have 2 million tokens that's what the capacitor is gonna release in batch 2 so more applications will come in by April 5th we'll have another meeting where the pinball awards committee decide to which applications to recommend and we'll release the second batch of tokens that way so that's one side of how the system works that's how we get tokens to people who are doing the work but what about how we get donors to actually put tokens back into the system that's where you come in we want to demonstrate here this weekend at East Denver that even without putting your money directly into the system to make donations happen that you can use your loyalty to make donations happen we want you to use the ally guide as the way that you decide which businesses to patronize here at each Denver and again our goal is to convince businesses that instead of giving money to Facebook in Google to make people come in the door that they could give money to a cause that we care about and get our community to come into the door we want to give them the same return on for every dollar that they're currently spending on advertising but instead of spending it on advertising they're gonna spend it on making etherium safer that's our goal that's the problem that faces us but that's a problem that we're gonna solve and there's a lot more problems in this world that are more important than making aetherium safer they're huge but we're not helpless in the face of these problems where mankind solving problems is what we do and if we can figure out how to cooperate and then show the world there's no problem that we can't tackle together so let's start here today let's start here this weekend join us become a pinball Ally let's make it there iam safer then find out what we're capable of thank you [Applause] do we have time for questions any questions yes the convergence model emergence so the idea is that people like as an individual it's hard to get your single proposal through the entire system the idea is that you have to organize with people who've already been getting proposals through the system and if people if the community doesn't like the current process that the incumbent recommender is making then they can compete with them start a separate recommender with a separate process and convince the token holders to go behind them instead instead of just being isolated individual actors in the system we want to we want people to organize together and decide what the consensus of the community is I'm not sure if it's exactly that but the idea is just in general if you have these constraints the optimal strategy to play is to find all the other people who can put together a process that can maintain the trust of the community and work with them to get a recommendation through to get 51% of the votes and to get those proposals enacted all right thanks for coming everybody become a pinball ally this is the ally guide there's businesses on it that you should support and you'll get a badge every time you shop there and send us a receipt so get this from me or go to allied and Viacom