Proof of Humanity DAO Governance Lessons from the Sybil War

Perfect. Perfect. Okay. Perfect. Hello, how are you?

Hope you had a good lunch. My name is Federico. I am founder at Kleros. Do you know Kleros, guys? Oh, that's amazing.

Good job for the marketing team. So Kleros is a solution protocol that can be used for solving lots of different disputes happening in the Web3 ecosystem. But I'm here to speak about something different and one of our projects that was called Proof of Humanity. So 2014, Vitalik publishes this post where he says that there are some problems to be solved in blockchain and one of them is the anti-civil you know attacks registry and the question is how to give a unique identity to all of the humans in the in the world. And this has lots of different use cases which are super interesting, like for example, enabling one person, one vote in governance for DAOs.

So we launched this app in 2021 in March. And it's a system that I guess any of you registered to Proof of Humanity? Maybe some of you. Yeah, for sure you did register. So it was a community tool.

I mean, people could submit their profile and then community people would verify if the profile complied with some conditions. And for making your submission, you had to send a picture of you with a little video and a address that you wanted your profile to be linked to. And for launching this, we partnered with Democracy Earth, a project that had a UBI token. And one of the use cases is UBI airdropping because you want the token to be airdropped to each person one time and not having people with many accounts. airdrop to each person one time and not having people with many accounts.

So we partnered with this organization so we could have more reach and also start to this use case. And when you got registered, you had this in your interface. This is the tokens that were dropped to you in your account. And then we launched the DAO. you in your account.

And then we launched the DAO. And for this, we did a very simple bylaws or governance system that was based on the idea of democracy. Because one of the use cases was one person, one vote. So we kind of pitched this as the first democratic DAO ever. And one of the main topics was all of the users who registered would get one vote in the DAO.

So it was the first truly democratic DAO. And 19 months later, let me tell you something, it really didn't work as we expected. It became a very big drama. And so what were the governance dynamics that led to this crisis and collapse of the proof of humanity? People accusing each other of lots of different stuff.

And someone from Blockchain.gov wrote this paper about ethnography. They made lots of interviews with different people who were involved and they analyzed lots of interviews with different people who were involved and they analyzed tons of conversations that were on Telegram, on forums and stuff like that. And then they analyzed and they came up with a number of situations that, well, happened in that DAO. And my talk is a bit going to explain some of that comes from that report and some of this is my interpretation.

The first thing is that the community did not have a shared values or vision because there were like two main groups. One of them coming from like people, more like libertarian, more technical people, and they really saw the potential of proof of humanity as an anti-civil system. And the other side of the community came from the democracy earth, more like political-minded people who were seeing this more as a tool to deliver UBI income. It was basically for them a tool to fight poverty. So we started with two different motivations from people in the community.

And we said, okay, maybe this is okay. Maybe this will work because Aristotle said, if you have more diversity in your political community, then you can have also better decision-making. And it didn't turn out that way because at some point, problems started. And the trigger was when some people made a submission for the profile to be registered and they were challenged and then because they made a mistake in their address and stuff like that. And they were pissed.

But this started to break the community in the sense that the people who came from the Claro side, they said, well, but this is how it's supposed to work. This is how the crypto economic system needs to work for it to be secure. But people who got the profile challenged, that didn't come from the blockchain ecosystem, so this is an unfair situation where they were just wanting to have a UBI token and they were losing money just because they made a mistake in their registry. So the community, instead of this diversity, instead of becoming better for decision making, ended up in two separate groups that became factions. And at some point, this started to same factions.

At some point this started to radicalize in the sense that each group started to see each other as a demon. The Kleros group saw the democracy group as idiots that were trying to do something like universal basic income to fight poverty. And I guess the democracy group saw the Kleros group, as people who were not, I mean, they didn't mind about people losing their money when they just wanted to have a UBI. So this started to radicalize. And there is this very good book called Against Democracy by Jason Brennan that says that in every political community, there are basically two types of people.

One are the hobbits. They are mostly apathetic, and they just go about with their life without caring too much about politics. And the other ones are what we call hooligans. They are like sport fans of politics, and they really see their side as a football team. I'm Argentinian, so I know there are lots of fanatics in football.

And this is kind of what the hooligans do. So when this radicalization started, what happened is that the hobbits started to leave the community and only the hooligans stayed. And the hooligans were those who wanted to win this war. And at that point, remember, we had decided to make this one-person, one-vote system because that was the point of proof of personhood. But since the system of voting was not good enough, one-person, one-vote, if you combine this with simple majority, just to pass proposals and no current requirements, this results in what we know as the tyranny of the majority, where if you have a majority in some moment, you can just do whatever you want in the DAO.

And the majority started to crush the minority. And then full civil war started with accusations And the majority started to crush the minority. And then full civil war started with accusations and escalation. And in the end, there was a fork and the community just split. So this is a failure of Adao.

And this is the post-mortem of Adao. And this was a bit predictable. and you know this was a bit predictable James Madison when they did the Federalist when they built the US system of governance they had seen how this type of direct democracy had failed in ancient Athens and well that was what Madison wrote in this Federalist had we read that before we had read that before but we maybe thought this time would be different we didn't so some lessons that we learned and that we can implement in the future for better DAO governance so you need a constitution that explicitly states the vision and core values of the DAO, of the community so what was proof of humanity? an anti-civil system or a tool to fight poverty? There was a big disagreement at the core of the project.

And you need clear rules for the governance system, including how to solve disputes when they arise. But this is something that is sometimes easier said than done because when you want to launch a new DAO, you want just to launch it and put it in the hands of people and get users, and do you want to spend lots of time arguing about the constitution and the rules? Well, maybe not. So that's why, you know, you just sometimes launch the project without, you know, taking good care of the governance and this can, you know, backfire. Second, DAO needs a mechanism to prevent the majority.

Of course, I'm not going to explain to you what separation of powers is, but I'm going to explain to you one way in which this can be implemented in a decentralized way. Imagine some people make a governance proposal to the DAO, then there is a vote proposal to the DAO, then there is a vote and the proposal passes. And then now, well, there is a period where people can challenge this proposal if they think that it breaks the constitution or breaks the main values of the DAO. People could challenge it. Brexit main values of the DAO, people could challenge it.

And then there's going to be a jury that's going to analyse if this proposal is or isn't constitutional, and then they're going to vote. And if they deem the proposal to be not constitutional, they can, I mean, people who made the proposal or people who voted for that proposal can be fined or can lose some stake or lose some reputation. And this makes sure that every proposal that is passed will, well, be according to the Constitution. And this is probably the only thing I will try to sell to you. This is how CREDO SafeNAP system works.

This is a system that you can use for your DAO to organize this constitutional panel to make sure that if there is something not great, it will be shut down by the Constitutional Supreme Court. Took your picture? Yeah. Okay. So, third, a mechanism to fight factions.

And one of the main topics of the ancient republics was sortition, random selection of citizens. You know, democracy for ancient Greek was not about elections. It was more about random selection. 99% of the people who occupied public position in ancient Athens were random selection of citizens. And just 1% were elected.

And let me see if you know what flag this is. Because this is a flag of a country, a republic, that actually was very important in the use of sortition as well. Nobody? I will give you some maybe tip. I mean, the city of love.

Not Paris, this other one. Very well, Venice, the Republic of Venice. Before being known because of love, it was known as the most serene republic of Venice, the Republica Serenissima, the stablest and longest running political system in Western history, 1,100 years, and that's a lot compared, well, even to Rome. And it was not that they really liked each other. The Venetians was a society that was fractured in different factions, different families that were armed to their teeth, and they were ready to start a civil war.

But they didn't because the system worked extremely well. Because most of the governing positions were chosen by lot and they had really elaborate procedures to verify. So everyone could verify that the system was working as intended and nobody was tampering with it. So each family had a chance to be randomly selected to rule Venice. And then when that family's time ends, the next one will come.

And this worked for 1,000 years. And this made also politics an attractive career because there was nothing you could do if you wanted to be the political leader. You could not game the system because there is no way to be the political leader, you could not game the system because there is no way to be elected by populism or demagogy, but you cannot fool the random selection system. So that's why there is a very famous Venice proverb from that time, that the leader of Venice, you cannot get there. The office will pick you through randomness.

If you want to learn more about this, a good paper from the Institute of Mechanism Design in the Venetian Republic. So one of the questions we might ask ourselves is can we use lotteries to break factions and these collusion networks and make sure that all of the factions in the DAO can have a rotation in power and not become like in fighting. So then the franchise. Who will have the right to vote and how should votes be weighted? When we have token voting, it's easy.

You have a token, you vote. And one person won't vote is more complicated. So let me tell you a bit about John Stuart Mill, 19th century philosopher. He was a promoter of universal suffrage, but with a caveat. He thought that citizens with more education should actually get more votes.

And this kind of points to a very famous and very traditional discussion happening in political history between epistocracy, where the rule of the wise, where only people with knowledge should be able to vote, and democracy, where everyone has the right to vote. This was a tension that happened all the time since the ancient Athenians. And of course, I mean, if we were building a person who is because we want democracy, right? But there are ways in which we can combine one person, one vote with some idea of epistocracy, some knowledge to be able to vote. And one of the tools I have been studying recently is called the enfranchisement lottery.

And so this is how it works. You have all of the community, and then you pick randomly, like a small amount of them, like 1% or 5%, or depends how big your community is. And those people that you pick, now they will go through a learning process. Maybe it will take a few weeks. And then they will be able to vote on that particular proposal.

So they were picked randomly, they are trained, and then they have to pass an exam. And if they pass the exam, they get the right to vote. And you do this for each proposal. And you can pay them for doing this. So they can make money for that week they will spend.

And this should result in better decisions. Want to learn more about this? This is Democrats, Epistocrats, and the Enfranchisement Lottery by Claudio López Guerra, a great paper that explains how this system works and why it complies with our ideas of democracy. And complementary to this idea of enfranchisement lottery, try to create instances of deliberation in the sense that there is this author called James Fishkin that is a proponent of deliberative democracy that is a way to make decisions where people, instead of trying to look smarter than others on Twitter and gather followers, they try to get together and have a process of arguing and deliberating under good conditions. This is what these mini-publics look like in real life.

They get together and they discuss the topic, the decision to be made, with the guide of an expert, and then they get access to information, etc. And this could be done in the context of a DAO. So you could have groups of people who were picked randomly and then you make them work together on Zoom with some guidance of the workshop. And then if this had been used at the Proof of Humanity DAO, I think we could have broken this factionalism of two groups thinking that the other was an idiot and it could have resulted in just two groups trying to work together and understand the other group position and why they thought what they thought and then maybe have a better decision making process after deliberating together. So, to conclude, we have new systems of identity coming to blockchain and we will have more and more implementations of one person, one vote.

Now we have mostly token voting, but governance will go to one person, one vote because this is more aligned with our understanding of democracy and how a governance system should work. We can learn a lot from history of political ideas because humans have tried this type of topics a long time ago. And I will just give you two last things for you to read if you want to learn more. One is Governance Archaeology by Nathan Schneider and Federica Carugati. Follow them because they are great.

And the other one is this recently launched Blockchain Governance Toolkit by Blockchain.gov. That is like a guidebook that you can use to have better governance systems for your DAO. And remember, as Churchill said, study history because in history lies all of the secrets of state threat. Thank you very much.

Time for questions. Okay, thank you. Thank you. Two questions. Okay, two questions.

Yes. Okay, in the system when you randomly choose like 5% and then the 5% goes through the exam, who prepares the education material? when you randomly choose 5% and then the 5% goes through the exam? Who prepares the education material? That's a very good question.

It could be a commission of the DAO. DAOs could have commissions, what they call it, core units or whatever. So there could be a commission that is focused on preparing the educational materials and also choosing what experts will guide the discussions because you want this to be seen also as a system where parties are put I mean under an expert guidance that is not biased so that should not be that difficult because this has been done for a long time in the traditional mini-publics research done by... So we can just bring what was done in the non-blockchain world to the blockchain world because it should not be very different. What is your fear?

That people, I mean, corporations will provide only the evidence that favors their view? Yeah, I mean, but that happens traditionally. Also, it happens, you know, the good thing is that we don't have to reinvent the wheel because this was already invented by people who are doing... France did a convention citoyenne. I don't know if French people here remember convention citoyenne of Climat.

And they did something like this with this type of dynamics. They picked random French citizens from different parts of the country to provide some suggestions of legislation for France to pass on climate change. And this was done with this methodology. Okay, everything is clear? Thank you.