Futarchy on Ethereum: Practical Learnings from the Kleros DAO

Transcript

Cuz you have the talk after? >> Yeah, I'm next. >> Yeah, I'm in the second one. >> Okay, yeah. Sure. Can you go? >> Hi all. Yeah, I'm smoking. Perfect. So, I am Alex. I'm integration leads at Kleros. Kleros is a decentralized dispute resolution protocol. And today I'm going to talk about futarchy on Ethereum. Um I'm going to die deep dive into the concept of futarchy and I'm going also to share some practical learnings about what we have experimented with the Kleros DAO.

So, let's get into it. First of all, futarchy is an old concept. It was introduced more than 25 years ago in 2000 by Robin Hanson. Robin Hanson is a professor in George Mason University. And he introduced this concept of futarchy with this paper called basically "Futarchy: Vote Values, But Bet Beliefs". In this paper, Robin Hanson described a new form of government. In this government, basically, you have the democratic process that is like any big democracy with elected representatives.

This is the vote values part. It's still really important to keep in mind that futarchy is not prediction market on everything. The vote values part is really important to give the long-term vision and the big goals. But in order to implement, basically, uh this policy and achieve specific goals, the how part is done by market traders. And this is the bet beliefs that is leveraging prediction markets and conditional markets.

So, let's dive into it with a theoretical example of government futarchy as it was described by Robin Hanson in his paper. So, the vote on values part, you have as I said the the some representatives that were elected and they try to basically maximize the wealth of uh their nation and they choose a metric. They choose what metric? They choose GDP per capita. Why do they choose this metric? Because usually it's a good approximation of the wealth of a country and also one point that is really important to keep in mind when it comes to futarchy is you want to have a metric that is easily measurable and that is

really clear to measure. So, they took GDP per capita. But the And then I put this meme there because usually when you ask people what they want, it's not a big issue. Everybody wants the same thing. Basically wealth, health, happiness. However, when you ask people how do you get this? Basically, you have different answer. Some are less tax, some say more tax, some say sacrifice a goat. And this is where in this part where futarchy plays a role in this how to get those.

So, basically in this one uh you have one proposal. One proposal of policy that could be to create a UBI. Universal basic income. You know, some people are proponents, some people are opponents and there could be they have some debate. In order to basically run to do to do futarchy on this, you will basically ask these questions uh into a conditional market. What will be the impact on GDP per capita if we create a UBI?

And basically you create a conditional market and in this control conditional market some traders are going to trade and it's basically a way to fork the current reality in two. You have the really the reality where this proposal pass and in this example for instance, the GDP will be 48,000 per capita and you have the reality where this proposal fail and in this example for instance is 46k per capita. I took this this number as an average basically it's the current French GDP per capita.

And basically what happened in the paper described by Robin Hanson is you let the trading the markets trade and at the end you choose basically so you implement the proposal if the pass price is above the fail price. So in this case the markets are telling us that if you create a new UBI the basically GDP per capita will be higher so you implement the policy. So it's fully futarchic basically uh um process.

So this was the theory. This was tried in some companies basically never at a government level from from my from my knowledge and for a long time yes it was just experiment but with crypto basically and the power of basically open and trustless trustlessness we saw new experiments and more successful ones. And one very successful one when it comes to futarchy is about Meta Dao. Meta Dao basically is a process that is is a project built on Solana and it was a it's a project that got a quite a success leveraging futarchy in order to fund

raise. And it's interesting because even media like Bankless that is quite Ethereum aligned and pro Ethereum invites Meta Dao in order to talk about futarchy. And during this podcast one of the podcaster Ryan he basically asked yeah this is super cool. Do we have something similar on Ethereum?" And our treasurer, Juan, basically answered him and said, "Yes, we already have things going on on Ethereum. We have basically Futarchy for DAOs in Gnosis DAO, for instance, but we also have Futarchy in Velodrome DAO, that is the ParaSwap DAO that that has re-branded recently, and also in Kleros." And I will I will go a bit more in details about about Kleros

one. All of these DAOs are doing Futarchy thanks to a project that is called futarchy.fi, and Robin Hanson is the chief scientist of this project. It's using basically Seer as a infrastructure layer for prediction market, and Kleros as its as its oracle. So let's better understand what is it. So, I put it in the title, it might look a bit complicated, but at the end it's quite simple. So, for DAOs, we are talking about price-based advisory Futarchy. So, I will explain a bit more in details why. So, first of all, price-based. Why is it price-based? It's because the metrics we are taking is the token price. Why the token price? Because it's

again easily to measure. You can basically It's also reflecting most of the time the strength of a project. If you have a token of a project, you want the price to go up, but you also want the project to succeed. It's like the the the token has some specific tokenomics, and the more the project is used, the higher the price goes, usually. So, it's a good metric, it's a good proxy. And then, proposal. If you are a bit into DAOs, we know you know that we have some proposal. It's called improvement proposal. In the case of Kleros, it's Kleros improvement proposal.

For example, here it's the 86. I don't go in the details of this specific proposal, but just to to illustrate the thing. And we do exactly the same thing. We We run a conditional market. What will be the impact on PNK price if we pass this specific proposal. And in this case, basically, we are going to to to trade on the the traders are trading and you get a price yes PNK price if the proposal pass and you get a no PNK price if the proposal fail.

And the big as you can see here, the yes price is above the fail price and the big difference here is that's why we talk about advisory futarchy, it's the fact that it's not passing directly based on the market price, but it's based it's the proposal pass or fail based on the snapshot vote. And that's a way to basically, as it say advisory, it's just here to give extra information to the voter and at the end, the one that keep the power are the token holder uh and they still can vote in the snapshot.

And here is a So, two screenshots illustrating this. So, you on the right you can see basically the screenshot from futarchy.fi where you have the yes price trading above the no price and you can see the snapshot from Kleros where the vote basically was casted by the holders of the token. So, let's go now a bit into more details about the learnings from Kleros DAO. So, the three learnings that we that we got So, the first one is about providing governance signals.

So, in Kleros we saw different kind of proposal. We saw a good proposal, let's say according to futarchy. In this proposal, it was about to change basically some rules about the jurors incentive program. Um yeah, this proposal was a bit complex. There were a lot of debate. As you can see, the community was split. 2/3 of the community was for, 1/3 was against. And for some basically token holder that don't have the time to go in all the debates uh yeah, and all the complexity, what they could do, they could just have gone into futarchy.fi, look at the impact price, the estimated price impact on PNK, and see that in this case, for instance, it was 14% positive.

On the opposite side, we saw also some example of bad proposal. We had this proposal, for instance, to ask for funding in order to have like to do some research. The topic was a bit vague. It was not very clear. And in this case, for instance, we saw some futarchy market showing a negative price impact for PNK. The interesting point here is we can start to see in the snapshot some voters and using using, basically, futarchy markets as a rationale to justify their votes.

Then, we went further than this. That's the first level of doing governance signal. We went further by, basically, implementing some safeguard for sensitive proposals. This proposal was passed in the governance, basically, of Kleros. And this proposal, futarchy-based governance rule for PNK minting. So, minting the new token of a project usually is quite sensitive. Why? Because you are diluting, basically, the the the the value of the existing token. So, for any proposal that is asking the DAO to mint extra token, um there is an extra rule. You need to win the snapshot base space

vote, sorry. So, snapshot vote, but you will also need to win the futarchy market. So, it's two conditions. The only way, basically, to win it is the two condition. If you only, that's what he's showing here, if you only win one condition, your proposal is blocked, is rejected. And we saw this with a proposal of a minting a minting proposal that happened in Kleros. Uh basically, we saw that uh there was a lot of trading on those markets, as you can see here. Uh because Why? Because you have you had some people that were basically uh conditionally selling their PNK. They were saying, "If this proposal pass, I want to be out." And by doing this, basically, they were maintaining the no PNK price above

the yes PNK price. As you can see here. Uh so, yeah, this is an extra mechanism in order to protect, basically, uh for specific sensitive proposals. The last point is about integrating with governance tool. We learned that, basically, the voters are used to some specific tools, like Snapshot, for instance. That's why we integrate with Snapshot in order to have directly the price uh of those markets integrated in the plugin. So, most of the voter, they don't go in details in the forum. They just look at the Snapshot uh proposal. And here, they have directly this extra information from the

um futarchy.eth market that are there. And also, recently, there was this delegate.futarchy.eth that is a delegate where you can, basically uh delegate your your PNK. Uh and what you know is this delegate is going to vote according to futarchy.eth signals. So, it's a good way, basically, to deal with one of the big problem in DAOs that is voter apathy. Most of the time, you know, we we we we you delegate your token to a specific delegate, and the guy sometimes doesn't vote, or there are some problems. Uh in this case, you're sure that this delegate is going to vote uh for maximizing the price of the

underlying token. So, there are these three main concrete learning we learned from the past, basically. And now, if we look a bit more at the future, where we are going and what we want to do, uh we launched uh recently a new platform that is close Kleros for sites, basically. And it's a platform that is leveraging an ID from Vitalik that is the distilled human judgment concept. Uh so, it's an ID from 2025. You can have a look in his article AI as a engine, human as a steering wheel. I don't have the time here to go too much into details, uh but thanks to this platform, you will be able to do uh distilled human judgment. Uh that's why

uh Vitalik recently retweeted one of our post introducing this uh platform. Two things to keep in mind is this platform could be used for a lot of different use cases. One of the use cases that you can leverage this platform for, it's KPI-based futarchy. If you remember from the beginning, you need a specific metric. The specific metric was yeah, the token price. But, when you think about it, when you are a DAO, for instance, you want to grow an ecosystem. In order to grow this ecosystem, you want uh for instance, if you are a DeFi DAO, you want to maximize the TVL uh that uh one specific project uh will achieve if you give him a grant. And

that's perfectly what we need for a for a basically a prediction um sorry, uh for for futarchy and for KPI KPI-based futarchy. Uh so, basically, you ask a question, uh if this project get a grant, what will be its TVL? And thanks to this platform, we will be able to do things like this. And if you stay tuned, uh you will see some example uh coming from us in the future. So, thank you for your attentions. Uh yeah, I I hope it's clear a bit more about uh what is futarchy and how it can be used for your DAO uh with like uh some concrete examples of what we can can now and also what we could do in the future. So, thank you all.

And if you have questions, let me know.