Press "Enter" to skip to content

Vlad Zamfir 'Blockchain Governance'



you do you hear me okay okay sweet well thanks for having me I'm super super super happy to have made it so I was gonna give a talk about Casper and sharding and it was gonna be a very technical talk and I realized maybe that's not the best use of our time because I usually don't have enough time to pack to tag to give technical talks and and I thought maybe instead we could talk about blockchain governance and specifically like aetherium governance and challenges in aetherium governance I think it's gonna be a little bit more digestible it'll be something that you won't can't find on YouTube already and so hopefully you don't mind the change in schedule so here is the outline and all my bullet bullet points got ruined in the trend like translating formats here I'm gonna talk a bit about governance in general bit about blockchain governance a bit about aetherium governance some challenges in the theorem governance then we will have like I'll have a open-ended discussion or series of kind of prompt questions about what a theorem should be like and then and then I'll give a conclusion and maybe I'll have time for Q&A so governance is about decisions and also about you know the kind of how they're made and what happens after like the decision is many people go with it people fight against it we're gonna go through a series of interpret like perspectives on governance the first one is governance as a decision-making process it's kind of a very inputs produces outputs kind of view of the world where there are two sit there are things like voting protocols where people send them votes and then there's outcome and it's very mechanical it's a process and it usually like for example if you describe like the proceedings for you know like a full formal proceedings for like a legislature for example it would be in terms of this description of a process that produces some output and this is understanding governance is a decision-making process it's kind of the most very basic way to understand governance and and if you understand governance like this you try to come up with ways to make decisions as your tackle as you as your attempt to tackle governance problems there's gonna be people and process processes upstream from decision making process from decision making processes and also downstream and the ones that are the agents that are upstream are called participants and the agents that are downstream are called stakeholders participants can have of influence they have a say whereas stakeholders are just affected by the outcomes you could be both though a stakeholder and a participant so we can also see governance is coordination between participants which is quite a different perspective than governance just as the decision-making process hey can someone bring me a bottle of water I'm gonna be here for a while Oh be nice thanks be metallic just whipping a bottle at me great all right so so so so governance can also be understood as a chord coordination between participants to produce these outcomes so rather than just think about the process you think about governance being oh the way that a whole bunch of agents coordinate around the decision-making process in order to produce outcomes perhaps they talk to each other before going to vote to all agree on what outcome to vote to vote on and that would be an important thing to note on governance we're not just the voting rules but also like how do they how do people coordinator round the around the around the game around the decision-making process right so governance is not just the decision-making process but also like how people coordinator on them and coordination problems have been studied in game theory and another and economics and sociology you know all sorts of places this is the most basic setup of a coordination game in game theory where people could choose between two two outcomes that are kind of favorable but only if they coordinate but somehow my valise and Bob choose to choose different strategies or choosing an uncoordinated way they don't get the outcomes they and Ted and so Gordon coordinated games core coordination involves like this property that only if you coordinate you get the outcomes you intended so how do we coordinate around decision-making processes well there's lots of things that we do we have used focal points we use expectations that we built about of each other's behaviors we use norms which are socially enforceable expectations we just use communication communication is very useful for for for coordination common knowledge is very useful for coordination common knowledge is like this distributed state of knowledge where we know something to the point where we can act like it's a fact where we know that each other know it and we don't have to question that each other know it and we don't question that we know that each other know it and so we can really behave as if as if as a distributed in fact that everyone knows so all these things are examples of the kinds of things that people use to solve coordination games or tests or to or to to coordinate around decision-making processes so the the the other thing to focus on here is there are that there's gonna be decision-making processes upstream of the of any governance process and but the players the participants in the game could to choose what upstream processes they use they can come up and use a process and or bend in it and there coordination around what processes to use the coordinate just to to involve themselves in governance which is kind of understood as coordinating around a decision-making process you know will itself involve the act of establishing processes and of establishing their use so this is a really important concept in governance called legitimacy legitimacy basically is this sudha state of knowledge is common knowledge that we're gonna use a process and go with it as opposed to replace or abandon it the legitimacy of a norm or a process of a decision of a principle means that we know that we're gonna continue using it at least for as long as it's legitimate which you know and because legitimacy is like oh the agreement it's kind of like you can think of it as agreement that we'll use something but it's more like distributed knowledge that we're going to use something so we might have not agreed to this but still we all know that we're gonna going to use it so you can understand governance as legitimacy management basically as a process by which participants agreed to use or come to a point where they know that they will use the process in governance where they will come to know that they will abandon using a process or a norm or a principle or or just to respect a decision or not and and in this view of governance it's kind of coordinate it's it's a it's the coordination view of governance with a specific focus on the process by which we decide on decide what the processes are that we coordinated so that we started off with a very we started off with sorry I'm getting ahead of myself so a couple of slides here so so so so basically governance is legitimacy management is is the is understanding governance as participants legitimizing and D legitimizing different kind of processes institutions and norms that are understood as being mechanisms for coordination so we need to coordinate so we need to create norms we need to create institutions when you create processes in order to solve that coordination and governance in this frame is understood as the process by which we legitimize and D legitimize the ways in which we make decisions and the norms by which we that which we follow while we're making decisions there is this quote from metalic here that the in the coordination model of governance we look at governance as something that exists in layers whether the bottom layer is basically the real world or laws of physics so somehow the the decision-making processes at the at the bottom are the ones that that pulled the triggers that are that are that are really making the decisions and then a lot of the in terms of like actually operating on the governed resource and then a lot of and then what governance is is all of the processes by which we establish revoke and maintain the legitimacy of the decision-making processes that we use to govern these are the the underlying reality that we're trying to govern so that was all kind of kind of abstract but here's the recap so we started thinking about governance as a decision-making process and then as a coordination game and then as a this kind of coordination game what we're doing is we're managing the legitimacy of different processes norms and institutions basically managing our agreement to use something into the future and so you can kind of see how this evolved from a from a very rural based kind of interpretation to one that is more about coordination to one that's more about politics really one that's about like what are the kind of principles and norms and and and you know that drive the legitimization and the D legitimization of the of the of the processes in north and like practices that we have for making decisions so I think it kind of broadens our understanding of governance from only making decisions to the coordination game to something that's a more political and about the legitimacy of our institutions so let's talk about blockchain governance so the blockchain is something and so it can be and there's something in the world so it can kind of be governed right it's basically it's a distributed system it's a bunch of software running on a bunch of nodes and so blocks in governance at least includes the governance of nodes and the governance of software repositories and because it's a consensus protocol the changes that are made to the nodes and and the changes that are made to the software have to be coordinated because if the knows run incompatible versions of the protocol they won't synchronize they will get their utility out of it and if the if different software implementations of the protocol implement the protocol differently they can fall out of sync and so the software that runs on nodes and has changes to the software repositories half-hour kind of very have to be done in a very coordinated way because the consensus protocol that requires that nodes are completely synchronized on the state of the consensus so Boston governance is kind of at least the governance of the nodes and of software resource repositories someone less fundamentally though block chains are also identified by trademark and watching governance also involves the governance of a trademark so you know watching governance at least is how community will manage changes to software changes to what software runs on nodes and how the trademark is used so basically anyone who's making decisions about these things right when you're deciding what software to run on your node when deciding what mmm what software to merge into a repository you're participating in blockchain governance when you decide to use a trademark or to or to exercise your ownership of a trademark you're also participating in blockchain governance when you refer to Bitcoin caches as Bitcoin that's participating in blockchain governance when you change your your node and you stop running Bitcoin you start running block Bitcoin cash that's blockchain governance when you commit a change to the Bitcoin cash rivo that's block team governance now that isn't necessarily participating in the same process that governs the Bitcoin core repository which is governed by maybe a different process and so you know you have to understand that the nature of the repository plays into the nature of decisions that are made in it and the nature of the of the node and distribute network plays into the nature of the decisions that can be made for it and the nature of the trademark also right changes that like the nature of the decisions that need to be made which changes the nature the coordination that needs to happen right the coordination around is changes to a software developer repository is very different from coordination to changes around what software runs in a note so who are the participants and who are the stakeholders in watching governance well the easy answer to the Hoopoe the participants are as I go if you're a node operator if you are contributing to software repositories or if you use or own the trademark then you're participating in blockchain governance but also if you help these people coordinate if you influence the agenda of developers if you in if you indirectly affect how trademarks are used perhaps by lobbying companies you're participating in watching governance anyone who's really engaged in the process of legitimizing and D legitimizing though decisions and decision making processes and norms that affect the way that the that software is upgraded the way that software has changed like in the repository as a first I meant the way that nodes upgrade their software the way that and then or the way that that software has changed in upgrade inside repositories or the way that the trademark is used then you're participating in watching governance so the per so somehow anyone who makes decisions or participates in coordination around decisions even indirectly through just legitimizing or D legitimizing a process is participating in blockchain governance so basically you know at least this at least these people participate in and blockchain governance but everyone is affected by the decisions made by made by watching governance everyone in the community everyone who is affected by the blockchain later on at least affected by the blockchains watching governance the decision not to do anything so watching governance at the at the end of the day is about the governance of software that runs on a distributed network so it's about the governance of nodes who run software and about the governance of software inside a repository and because the blockchain is accessible through this kind of marketplace and that block law chains also have trademarks especially public law chains and the governance of these resources are you know the the way that the community like the governance of these resources are is governance if you understand it governance is a decision making process if you understand governance is coordination or that then the the way that the community coordinates to make these changes is governance you understand the governance has this process of legitimizing and do the deal agenda maizing decisions in decision-making processes then the kind of public discourse by which we create the legitimacy of governance decisions is is is the kind of is where blockchain governance is right and and important governance really does kind of matter because it it it really influences a lot of outcomes so but first let's talk about aetherium so theory em has some legitimate governance decisions processes and norms already at least as far as participants in the that are kind of close to the software repository and you know that like you know at least as far as I can tell we do have some things that are kind of are kind of established and we all kind of agree well that we know that we'll each other well that what that will go with so we have this in the in this diagram thick arrows being kind of a lot of influence and thin arrows mean a little bit of a little bit of influence I think the role there's the role of full nodes to decide what software to run from different implementations of etherium I think that's something that we like have today in fearing governance and will continue to have as long as we have this norm of having multiple implementations of the consensus protocol the way that these are kept in sync is through this great process or institution or custom called the all devs call which is every two weeks on Friday a call where the core developers are the different implementations get together and and decide unanimously on the short-term release schedule or like features for the short-term release of the at least four of the consensus critical variety between cats and death and parity so this call guarantees that the that these teams are when they're independently implementing the protocol independently as to as much as possible you know they'd like to be independent they implement they try to implement the spec that they agreed on and the auto devs and they'll talk to each other to make sure that it stays synchronized and that like none of them are falling out of consensus the eeap's process is it a theory improvement proposals process and it is very much like the Bitcoin in program proposal process and the Python improvement proposal process where basically you you you make a request for comment or you proposed a draft of an improvement proposal and after sufficient review it will be considered by the devs for implementation if it gets to a certain level of like readiness and attention and the and the roadmap is kind of like our kind of common understanding of what the future of aetherium is and this is something that's negotiated and governed in a very different way than for example these software repositories this this is this this is what governed by Assyrian foundation developers for the most part this is governed by parity developers or eath core developers the the efz process is very is convened there's some editors there who are in charge of like editing the eeap's maintaining and making and making sure that there's a certain level of quality before they get accepted and the roadmap is kind of this you know evolving a picture that we have that we share about a theorems future and is it's it's governed in a kind of much more arguments based manner where there's there's there's not necessarily any clear final decisions but at any point but but the things kind of we get swayed by argument so basically that's all I really had to say about a theorem governance as it works today but I'm gonna talk about hearing governance challenges including some past challenges and that will kind of also bring to light some I claim that blockchain governance is important so let's talk about the DAO maybe you guys remember the Dow the Dow was this great smart contract that raised 120 to 150 million dollars depending on how you count to invest in in in in projects as as a choice of what the talk with the contributors to the Dow voted on but unfortunately it had a bug and it dropped all the funds into an attackers Dow which so it's so it's it's a long story but basically everyone 40% of etherium counts invested in this thing called the Dow and everyone who invests in the Dow kind of expected it not to blow up and a bunch of people expected the etherium theorem smart contracts – and I'm always execute according to the rules of the virtual machine and so something had to give right either like oh everyone was going to lose their money and the Bitcoin was gonna have a really good time or we could do like a hard fork and maybe just like forget this happened so the hard fork solution to the Dow crisis was basically okay let's let's let's have all the node operators agree you know like meaning like oh if if all the noter operators were to install a software update which had some previously determined block heights were to move all the money from the Dow's say to a smart contract that was designed to recover to give give the funds back to token holders then we could maybe act like this never happened mm-hmm so and and and this is something that actually happened that like lots of the etherion community vast majority of it I think did this upgrade and enabled the this in this flag or just left the default to to undo the effects of the Dow hack some portion of the etherium community didn't and some people some of the observers who didn't didn't you know like this decision who didn't believe that it was a legitimate decision when enjoying the etherium classic community which exists to this day and you know is still somehow defined by the rejection of the Dow hard fork so the down hard fork it was it was a really important case study in governance and the Dow crisis itself was really important case study and blockchain governance it was a was a huge crisis that made and pay attention to what we were gonna do to try to you know recover the funds we had it we had a number of ideas like to use the white hat hackers to because the the funds were in a compromised AO also because all the Dow's are compromised and so we had hope of recovering the funds by a counter-attack we considered a soft fork which was the miners could use to shut down the Dow's and to allow the counter attackers to drain the funds without changing the rules of the EPM but that was discovered to have a das moaner ability and so the hard fork and the white hat attack were kind of the only two real options and and we had we we engaged in a tremendous amount of discussion and soul-searching in the etherion community for a month and a half maybe maybe more maybe a bit more it was a tremendous amount of conversation and pretty much everyone who was part of the community participated we had the ability to sit down not not really in public channels but in private channels with all the stakeholders that we knew about and to talk and to try and try to see where everyone's where everyone stood and because we all knew that by some deadline we had to coordinate to decide whether or not to do this hard fork to recover the funds or not and through through this through a process of informal communication between many many actors in the space and you know one of the one of the big misconceptions about the doubt hack really is that the ethereum foundation you know did the Dow hard fork somehow the etherium Foundation actually did very little to almost nothing during the Dow crisis I was quite upset about it at the time but you know I'm over it since and you know the the the the Foundation's lack of communication during that time maked made it seem like it was in the background doing the Dow hard work but really the data hard fork was a really community oriented community driven thing where basically we used our private channels to communicate and to detect between our peers where everyone was on this decision of whether or not to upgrade their nodes to virt the doubt thou hack or not it was quite hard to observe that in public channels because the amount of attention that we were getting from Bitcoin trolls and and and and and the the broader media was so much was like overwhelming all of our all of our channels but we had we got would be we did get to have is basically like town-hall moment for a month and a half where we like discussed with everyone you know and everyone shared everyone's a bit like our opinions about all the different risks that we might be silly might be taking in order to try to decide whether or not we were playing we were we were going to install this hard fork and at the end of the day aetherium today you know has in its in its history this block number don't remember it right now unfortunately where the where the funds moved from the Dow's – idiot – to the recovery contractor and today we have a couple of more if we hope we have lots of other we have lots of other cases where we're interested in talking about governance there's this recent stock funds discussion around the parity hack which we're where library used in a multi-state contract self-destructed and everyone who used that that multi-state contract lost access to their funds the conversation around governance around this has gone in a much different way we had were having a conversation kind of now about bitman Asics and about whether or not we should have a hard fork to or wish we should even consider having a hard fork that would obsolete bit-bit mains Asics in order to have something that's more fair to miners with GPU miners you know we did always intend for a theory to be basic resistance but unlike the Dow for both of these for both these decisions we don't have an immediate deadline and we don't have a critical threshold of participation participation is much lower today than it was during the Dow crisis on these two discussions partly because there's no deadline but partly because the community is so much bigger now that these things seem smaller stakes and they are smaller stakes relative to ever to the stakes that are in the community but they are bigger stakes from the point of view of the dollar amounts so you know and as the community grows our ability to have legitimate decisions from the high levels of participation will decline our ability to actually convene and communicate and talk to our peers in order to identify where the community is will get worse over time as the community gets bigger and that means that there's a bigger need for legitimate processes and for legitimate process for establishing legitimacy of decision making processes because we can't just do it with a big Town Hall it needs we need we need to have more process if there's less attention so we might and we might lose our ability to have Lygia to make decision making processes to establish them through a debate because we can't really have community-wide debate as the community gets bigger and bigger and so you know already existing norms and processes will become more ingrained but may not be enough to lend legitimacy to all of the types of norms and processes that we have that we might have need for so unless we have a process for establishing legitimacy of governance process proceeds then if we don't establish the legitimate process then we were then we're probably going to lose our ability to do that except maybe in a crisis situation where the threshold of attention is high enough that we can do it through a public debate so now I'm going to ask some questions which I think or will will kind of illuminate the situation a little further about you know the extent to which we have a lot of big gaps in our understanding of how governance works or if how or what parts of governments are responsible for what but first of a couple of quotes so and this is this section is Oh what should governs be like for aetherium but also it's very applicable in general so this is a quote from Rick deadly Bitcoin core is like a boxing ring if you think you have an idea you have to be prepared for a fight and this and this kind of idea is there like oh maybe maybe for technical development we should have idea meritocracy where basically you argue with all of the Nerds in the relevant channels and if your arguments are seen to be strong enough then then the then you kind of have a chance of going through the development process it's not very accessible it's my main comment about this it's a neat idea meritocracy is a neat idea but there's a lot of tacit norm tacit informal norms and background knowledge that's kind of required at least to participate in Bitcoin core today and that can make it quite prohibitive you have to spend very many months learning about their norms if you want to actually participate in that discussion if you want to get in the boxing ring well it's a very particular kind of rank you see and only certain kind of moves are loud but if you work within those framework within that framework like you can be ok Eric Lombroso who I think is like famous for his participation in Bitcoin governance said that code followed the consensus and basically and he meant both consensus of protocol in terms of the protocol rules editors of the community consensus and you know I mean this part I think it's pretty crazy because I mean it says no to all hard Forks and any changes to the consensus rules and I think like as a researcher I think we could do way better on the sense of protocol stuff but this thing the idea that like with the code should follow the community consensus I think it's a really nice idea that we that we should entertain you know as opposed to the community following the developers the developers should maybe follow the community consensus of the community and another quote from Eric repo management is a minor part of watching governance and I think that's I think that's probably very true right I think most important governance is probably this process of forming consensus of a style around what to do right establishing legitimacy of processes and of decisions so most of the work is actually more on the politics side more on the convincing side then actually on the tech development side so here are some questions for aetherium I know how should the trademark be governed I don't know that's a good question should it follow the community consensus I don't know just an idea I'm not sure and don't really know what that means or how to measure the community consensus what should the view what should be the responsibilities of eeap's we didn't do the down hard fork through the eeap's process but we're trying to do the recovery of stock funds to the eeap's process there's the eeap's for technical changes or is it only for is it's only for technical changes or is it also for political changes or politically motivated or socially motivated changes one of the what is the role of e-edit errs exactly what is the relationship precisely between the all devs cause and the eeap's you know and what should be these relationships right what what is it today is one question but what should be the relationship should the should the all devs call the you know determining whether an EP is accessible or not acceptable or not should there be a separate process that determines whether an F is acceptable and that compels the old devs to to to implement it you know I mean I think the answer to both of those questions is maybe it's probably no but exactly what the relationship is and what it should be between the all devs and heaps it's kind of unknown today what should the core devs do when there's no clear consensus from the community about what what to do with a quarry or any protocol change especially in the in contexts of these more contentious changes like the Dow hard fork or the recovery of saw stuck funds or an aura change to obviously ASIC Hardware should they do nothing this is kind of the Bitcoin core answer there's no consensus basically do nothing or do we offer options and let the community kind of sort it out by by different node operators installing one version and software or the other should watching this is this is a huge one should lakhtin governance be active like meaning should we have governance that interferes with the execution of our contracts that you know at certain Heights maybe change has an irregular state transition should there be a legitimate process for doing regular state transitions you know like and and and and and and if so like what is how what's this relationship with the law you know there's there's the they're one of the big kind of concerns that people have and quite rightly around having a process for doing a regular state transitions is that it might get captured and that it's a security problem and that it might end up being inside US jurisdiction for example and I think that that is a you know super relevant question you know what is the relationship of our governance processes and the law and I'm being vague here the law isn't really necessarily Universal it's different jurisdictions it's a bit complicated but these are kind of deep questions that we have to answer as a blockchain space in general in terms of the actual governance of the blockchain itself in terms of the decisions that are made because sometimes like in the data hard fork blockchain governance is the only way to deal with certain outcomes the only way to recover the stock funds is stuck funds for example it's the only way to do the Dow hard to recover the Dow funds in a way that that that was that you know one counter so these are these are big questions and I don't think that anyone can really answer them I think that we need to participate in a public discussion about these questions and try to talk about them oops Oh No old version of my slides oh well I'll just I'll just I'll just talk about my conclusion from memory so so I think it's important that we that we that we talk about these these things because the decisions that we make now and the discussions that we have now will form and impact the decisions and discussions that we have in the future the the we will look tomorrow at the precedents we set today the norms we have now are likely to exist in the future the if we if we act now with no regard for you know social questions we will be more unprepared for them when they arise we have to talk about boxing governance and talk about like all of these difficult questions about the relationship watching governance in the law and whether we should do watching governance that interferes with the execution of smart contracts because if we don't talk about it then we're going to have less control of the situation and the probability that we'll have an unintended outcome or one where where we don't like that where we just don't like the outcome even though we might have intended it is much higher so I you know I basically I would encourage us all to think about you know these tough questions about what governance should be like to think about what are the processes by which we can legitimize blockchain governance processes today but also not to rush and to legitimize government because if we say for example choose a solution just because it's convenient just because it's easy to chair just because it's easy to understand how it works but we empower some group of stakeholders over all others that that's gonna have you know consequences down the line for for a long time so we need to be careful not just to quickly legitimize anything because we're anxious that we don't have govern and we don't have clear governance processes but we also need to be mindful that if we don't legitimize anything then we could be in a situation where we don't have any governance processes one them so you know I hope we walk this fine line have a public conversation and don't recklessly legitimize governance proceeds but still you know build legitimacy in our blockchain governance because if we don't if we don't have if we fail to have legitimate botching governance processes then someone else might govern the blockchain for us so with that thanks thanks a lot

One Comment

  1. Davinder Arora
    Davinder Arora June 5, 2019

    Blockchain is reality now and will replace most of legacy banking system in 2019 and social/governance systems soon. Please check if below social moral principles can help in the core fabric of this technology revolution

    1. One should always be ready to accept truth and to renounce untruth.
    2. All acts should be performed after deliberating what is right and wrong.
    3. The prime objective of new Society is to do well to the world, that is, to promote physical, spiritual and social good of everyone.
    4. Our/Every one conduct towards all should be guided by love, righteousness and justice.
    5. We should dispel ignorance) and promote (knowledge).
    6. No one should be content with promoting his/her good only; on the contrary, one should look for his/her good in promoting the good of all.
    7. One should regard oneself under restriction to follow the rules of society calculated to promote the well-being of all, while in following the rules of individual welfare all should be free.

Leave a Reply

Your email address will not be published. Required fields are marked *