Futarchy: Your Next Treasury Manager is a Prediction Market

Transcript

[Music] and the clicker works it's going to be incredible I'm so happy all right let me get all the extra fluff out of the way want to see my uh um bookmarks I'm sure or my apps or any of the other things I'm going to try and click now uh and boom we're in all right here goes everyone I'm going to race through this so it's everybody dialed in we need to be locked in it's going to be great um part what I'm going to do is I'm going to go through this in kind of three stages the first time I'm just going to basically uh tell you why I'm short token voting then the second thing I'm going to do is tell you why I'm long uh prediction markets and then the last thing I'm going to do is present kind of an alternative vision for governance

how's that sound right remember the three the three huh you can't hear me okay all right I'm going to project a lot okay all right let's go so futaki your next treasury manager is a prediction Market it's a bold claim um so we're going to start out with a proposal that you may or may not be familiar with depending on how involved in Dows you are tally is uh you know one of the sponsors of this event um they're a huge governance um I guess protocol now but before like application uh they kind of sit on top of the governor contract so you can imagine the governor contract as a as a like a a system for making decisions and they sit on top they provide like a nice

interface application that kind of extends the functionality to things that you and I would like to use like a nice clean application a website we can go to um like buttons we can click and so on and they've built this incredible platform they've been building it for a while and they recently put in this proposal to ask the da if it would fund them so the question is there's a dow or protocol with some resources and they're asking for some of those resources and we have this system of decision-making called voting and the voting system has already been set up and it's a way of gathering information from unop token holders so question should uniop Dow distribute $250,000 a year to tally for its services fairly simple question I think everybody in this room even if you're not directly involved in Unis

swap might have an opinion on this do you have an opinion raise your hand okay cool great people have opinions so uh as you can see this passed very happy for tally I think they're actually awesome um and very necessary and um I'm glad this passed the question you have to ask yourself is why did it pass when people decided to vote with their tokens why did they pick four now I don't know about you but myself for me I'm dangerously dangerously long me right I hold a full bag of me tokens right just like everybody else and what I'm trying to communicate here is we're all dangerously long ourselves we're all very self-interested we only really care about the things that affect us in a positive way or a negative way right and

you can't escape those now if you think about a token vote a good way of thinking about a token vote is that it's just a token weighted average of everyone's interests everyone's own interests so you ask every single person based on how many tokens they hold what they think and they're just going to think well for or against is this good or bad for me right and and this is not entirely true right because people have loads of different reasons for voting but this is this is basically the only thing you can guarantee when you ask people to vote they're going to make decisions that are in their interests everyone following me so far right so again this is kind of leading up to why I'm short token voting um now I'm going to change uh gears a little bit we ran an event called GG in Edge City a couple

of weeks ago and each day we called it governance games and each day we ran a different mechanism a different way of making decisions and on the first day we ran this so we ran the the First Column which was we asked people to come up with governance problems and then we asked them to vote on them we gave two everybody two stickers and we said put the sticker on a thing you think is most valuable to solve and we produced that there were other questions but these are kind of like aggregated so um voter apathy five votes delegate incentive six votes lots of the people in the AUD with delegates uh use acquisition three votes accountability six and kpr based resource allocation two now the next day we ran the auctions day so this is where we were auctioning off proposals it's a weird mechanism and I won't go through

it but imagine that instead of voting on proposals you had to bid on them right we're going to have an auction running and when you get the proposal you get to execute it and so the bid technically should reflect the valuation that you have for the success of the protocol right or sorry for the success of the of the proposal and this time around I said imagine when you win this bid and you get the proposal you get to keep all of the return that's created for the Dow for the protocol right so you bid and if it's like voter apathy voter apathy improves the Dow by some margin and then you get to keep that and it produced that distribution of uh bids on the right so interesting Parts voter apathy people thought like voting wise five this is really important right one of

the highest one bid people don't think that's valuable another one is accountability six n uh the best one I mean delegate sent of 66 again we had delegates in the room uh and then kpi based resource allocation to bid seven so what I'm showing you is even though you may have the same people when people have their incentives slightly change they're going to give you different information right so voters have an unknown stake in the outcome of a proposal you don't know what their self-interest are you don't know what bags they hold you don't know bidders can be given an explicit stake in an out come that we choose right this is one of the core insights of of how prediction markets

can be used for decision making and finally we can give bit of skin in the game in games that give us the information we need to make a decision when we want to decide whether to give money from a treasury to some project any investment what do we actually care about what's the thing we're trying to Target is it like growth of the token is it like growth of the number of people who actually use the protocol and we can actually craft these really specific games or markets to try and it like pull that information from the actual participants and then make a decision and this is why we think that we can use prediction markets for decision- making so now we're going to talk about markets in as a general mechanism for information aggregation anybody who's been following the US election which I assume is pretty

much every single person in the room um knows that we recently elected a candidate and um that there was a prediction Market running the whole time and the question is where the prediction Market is more accurate than the polls polls are also kind of could be seen as like a vote right um so let's say that I take a dollar and I put it in a contract and I say that in the future we're going to end up in one of two worlds they're going to end up in a world where we've elected one president or one candidate and we're going to end up in another world we've elected the other right this is a way of thinking about like the probabilistic representation like from prediction markets and then I say that anybody who holds the the a token the candidate a token or candidate a

contract will get to like redeem that dollar and I can sell as many of these as I want and this is that prediction markets are and the way you have to think about this is if someone's looking at these two options say they're looking at the first one they're going to be in this world the only thing they have to ask themselves is if I buy this contract what's the chance I end up in this world because that's going to be the expected value I have for that dollar if I think there's a 50% chance that we could end up in that world I'm only willing to buy that contract if it's less than 50 cents does that make sense right because you can only only going to make a dollar if you end up in that world only believe that's going to happen 50% of the time

so that dollar is worth 50 cents to you like if you understand that and that makes perfect sense to you you understand pretty much everything about how prediction markets work and you can take this and you can dial it up into conditional markets like the what's the probability that that you get end up in this world if conditional on something else happening and if you take that again you can you can take the same probability and figure out how to make decisions right so so let's go through that process so again this is the other the other world you could end up in and so as I said you might ascribe like 50 cents to each of these and if let's say you believe that candidate B has an edge like you're the French wh you're like that candidate has an edge you're willing to bid up to the probability you

expect to that you expect you will end up in that world so again if I buy that 50 Cent token or 50 Cent contract it's going to push the price up and push down the price of the other one right so sorry for the uh the slow loading um but yeah so then that would mean that you're at 1 and then obviously if the candidate wins everybody who holds that contract is going to get a dollar and everybody holds the other contract is going to get zero make sense because there's only1 it's going into one of these worlds right cool so let's talk about information aggregation for decisions so let's say now we go back to our original question right which is if we should fund tally and we ask the question in a slightly different way we say what's the what's the thing that what's the outcome

we want it's not that we end up in the world let's say we split the world into we fund tally or we don't fund tally and then we go into those words and we say what if we fun tally and we get the outcome we want or we fund tally and we don't get the outcome we want similarly what if we don't fund tally and we get the outcome we want or we don't fund tally and we don't get and we get and we don't get the outcome we want and so you can craft a question or a contract that pays out only if you fund tally and you get the outcome you want right this Mak sense right so the $1 is only going to be $1 in the world where those two things have happened right there's all these other branching worlds that could

exist right um and so this this now gives you the probability that the thing you want happens and that's the only time when you should actually issue the funds right if if we're trying to make a good decision here that's the reason why we would make the decision we want to fund things that produce the outcomes we want is everyone following me any anyone disagree all right so so this is an interesting um uh structure because we can now start to ladder this up into what we call conditional funding markets so this contract on the on the top will pay $1 if uh funding tally will increase voter retention by 20% assuming that's what we care about like voter retention gets increased and then on the bottom we'll pay out $1 if funding tally increases voter retention by less than

20% so really we're saying we only really want to fund tally if it's going to produce this thing which is like growing voter retention by over 20% and so the markets will price the top one say they say like 70% of the time we end up in that world and 30% of the time we end up in this other world and and a good way of thinking about this is like let's say there are 10 versions of the future seven of those worlds tally does this right it works out and in like three of those worlds it doesn't it just doesn't actually produce the outcome we want right and so you kind of say it's probably quite probable that we do end up in this good world so we should fun tally right okay I'm going to keep like hammering this I think it's like once

people understand it it becomes quite simple to like reason about prediction markets and conditional markets and also fuy so yeah so we select this so again what we're doing now is we're taking those probabilities and we're using them to produce a decision right we're not waiting for some outside event to happen then resolving them markets we're going to make the decision and then once we've made the decision we're in that world right we we've selected it and now we're hoping that the actual event occurs and if the event occurs anybody who holds that contract what happens to them they get a dollar okay this contract will resolv to a dollar this is a conditional funding Market this is basically what we're working on we're in the process of um of building the application I'll show you a screenshot in a second might be the next

slide it's the next slide um and um this is us using it for a hiring market so imagine you wanted to hire someone you could actually say what's the outcome that we want which is for example we're trying to hire someone to do some writing for us so we could run a market where we say what would be the number of let's say we wanted to measure engagement what would be the number of replies that person might get to their article right and then we could actually run a prediction Market to see who gets the most replies and use that as a as a method by which we distribute funds okay so this is a conditional funding Market um so if I was going to ask any one of you to go out of this room and then try and explain futar to someone else I would say the best way of

describing it is if markets believe that an outcome and this notation is just this uh bar is like conditional on so an outcome conditional on us making decision a so if we pick that we're going to go into this world we make that decision that world is more valuable than the outcome given we make the other decision right decision B right which might be decision a is do the thing decision B is not do the thing and if the first one is more valuable than the second choose decision a simple confusing anyone confused confused enough all right maybe comprehension has gone up a little bit okay good okay so um I'm going to like again change piece so now we're going to

talk about the like um the kind of like Theory or vision of governance and um what I'm going to do is kind of go back to like the theory of a firm we have Dows today right we have governance we're governing Dows and um that's kind of become the kind of zeitgeist of governance like we have these dows we have to govern them we have to make decisions and um when you ask yourself what is a DA like why do we have an organization you have to go back to like why do we have actual organizations already and if you go back to uh someone actually decided to ask this question like why do firms exist why don't we just have free markets of people running around doing transactions like entering into their own contracts doing their own marketing why don't we just have a

nation of Freelancers and um so someone tried to answer his name is Ronald Co this was 1937 he published a paper essay paper um called uh the nature of the firm it should have been called the the theory of the firm but it's the nature of the firm and he asked a question why the firms exist and his conclusion was transaction costs and a good way I think about transaction costs is there any cost associated with time triangulation we have to triangulate different things like people to work with and like different things to do and like different ideas triang triangulation and then trust right this is kind of the components of transaction costs and so theoretically if you're a human being which I assume all of you are um you have two options and you have more than these but let's just make it simple you

can sell your labor to the market directly you'd be a freelancer or you can sell your labor to a firm to an organization who who are organizing your labor and um and the reason why you might do that is because you're trying to again you're self-interested you're dangerously long you uh try and maximize your return right so why you would only really sell your label to a firm if the cost the transaction cost of being directly in the market is too great and the other way and obviously the other thing is true right the reverse is true um so in markets imagine every single person running their own supply chain that means they're going and sourcing all of the things the like raw ingredients that they can like pull together in order to actually like produce a product or a service that people want to buy imagine them doing

their own marketing and enforcing their own contracts you can see like most of the free like there are lots of Freelancers but they're normally in Industries where they can do this quite easily right they can find all their supplies maybe because all their supplies are digital right so there's lots of digital designers and digital developers um uh sorry freelance of freelance developers so if you had this world you basic basically end up in you're doing this because there are such like lower transaction cost for you to do this directly that you don't NE necessarily need a firm to organize the labor but some things like producing Burgers globally you really need someone to come in and do the job of like collapsing down those transaction costs by like entering into all the ingredients to secure all the beef globally so you don't have to do this

yourself and then they could basically they have all the contracts around them and then you just have to go and work for them and then you can be involved in the production of burgers and and because I think this is quite weird to like understand or think about I also did these stupid diagrams so um when transaction costs are low you just get all your contracts yourself when transaction costs are really high you just go and let someone else do that work and then you get to like they kind of amortize the cost over like all the people who come and work for them okay so you have to ask yourself if the nature of the firm Su just firms exist because they centralize in lower transaction costs why did our exist now this a question to you does anyone have a does anyone know

no no no no no answers regulatory ARR regulatory Arbitrage why are we here guys we don't know why Dows exist we're all here because Dows exist okay um I don't know I think I'm trying to find out the answer to this question right why did they exist um so I I was like looking through the like anals of crypto and I was trying to find like why did people come up with this idea there is like a lot of um information on some of these ideas from Nick Sabo which I definitely recommend reading on his blog or his I think it's not even on the blog I think it's on the uh Satoshi Nakamoto Institute website now and the podcast ohab actually never seen it I would

recommend checking that out if you can if you if you're inter Rogan um and then um there's a um there's also a um some really interesting blog posts from Dan Lara and Stan Lar on the DAC decentralized autonomous corporation which technically predates like the first time the ethereum spoke about Dows um again really really great would recommend reading that but one of the things that vitalic quotes is he said he read this book and this is really what gave him the idea for DS and the thing that's interesting about this book if you read it and I definitely recommend reading it um is that there's a it's like the idea of a distributed program a set of rules and it defines I don't want to give away too much but it kind of defines this entire New World Order it's incredible I definitely recommend reading it but the the new world is

based on meritocracy and decentralized control um one of the things that that this demon is not is it's not an organization the demon functions as a distributed intelligence it it uses the intelligence of the people in the network it kind of like finds a way to capitalize that which helps they have this like peer-to-peer like reward-based system they're ranking each other and so on it's really interesting it's kind of runs like a game it's actually a it's actually a game AI that breaks out that's kind of the idea um and um it kind of guides humans towards like these better Collective outcomes which is which which is really really cool I think it's it's really worth reading it's really fun um but the Insight is a demon is not an organization I that's the important takeway so why do we have organizations well I think it's mostly

based on this that like Dow's really the development is based on like us trying to avoid some of the things listed in this document right so this is based on the hoe test which is the um Securities test right the test by how by how they'll uh figure out that an asset is is actually an investment contract and I think lots of dows are really just trying to make sure that they don't look like they're issuing assets right the people who the organization issuing the assets so they distribute the control to a bunch of people and it helps them avoid this reasonable expectation of profits der from efforts of others because you're not the expecting there to be efforts of others if you're actually the person making the decision hence organizations so we call this a dow token voter assembly lot people get together they have ideas they have

tokens they vote I would like to suggest that we should start thinking about Dows as this the collection of organizations and people entities that orbit something they orbit a protocol they orbit like some collective like goal or something like that and I think that could be a really like actually a really refreshing way of thinking about what Dows should be though maybe we don't want to call them Dows H I'm going to try and come up with a new name um but I do think this much this much better reflects what I think Dows are like lots of people who are in the Dow part of Unis swap are confused about uni chain they say why didn't we know about uni chain why weren't we involved in uni chain and it's like well you didn't know but the Dow knew right Labs knew right because

they worked on it and I think it doesn't necessarily mean that you need to have all these things in one entity they can be distributed um and another really interesting part is if you think about like Labs foundation and kind of the token token voters Labs is kind of like the developers right the the the Avatar of the developers the foundation is kind of like the like Avatar like the representation of like the the the protocols users right um and then the you have the token as well which kind of represents the users as well and it's kind of maps to like Bitcoin where Bitcoin has like the developers that team of developers who kind of have some governance power then the actual like users who hold Bitcoin who have some governance power and then obviously the miners who have some governance power and so I think that we I think a good

model for Dows is that right it's like looking at Bitcoin and actually it's just a collection of those things called Bitcoin there's no one entity you point at and say that's Bitcoin um so potentially we could start thinking about Dows as instead of being this like uh like token or a token voter assembly we can think about them as being these like multilateral treaties between these organizations a contract that just helps to enforce the rules by which all of these organizations interact right the Nexus of contracts for that org a hyper structure I don't like that term but a hyper structure yeah it's not a Jacob term oh okay it's not I thought you meant the the Jacob hyper structure okay okay all right maybe I should go and check out the the

real meaning of hyper structure but yeah so this idea of instead of being Dows decentralized autonomous organizations decentralized autonomous money printers though someone did point out that the acronym is damp which is horrible so we're not going to say but we're also not going to say distributed resource allocation mechanisms dram they're both quite strange so I think I prefer dram so the idea is just a system a contract that itself coordinates the actions of people and corporations and entities and service providers and all of that group and it kind of like guides them towards better group outcomes so um I know I'm way over on time so I'm going to try and speed up actually I'm not doing too bad um so uh another way of thinking about this which maybe is quite relevant now is this idea of like industrial policy so um what

what how can we change the way that we allocate our funding to create better outcomes for us as a collective right especially if there are collective of different entities doing different things what could we do at the top level so that it kind of guides the actions of everyone at the bottom level without having to have explicit Arrangements between those organizations so you could argue that the token does this already right just the token existing kind of does this but um yeah you could you kind of see this already this is what nations do right so I think there like tracks with where we're kind of developing these uh these Dows um and it kind of chips sacks for anyone who's not like super familiar with uh American Trade policy which could be loads of people chips act for example is just a bunch of subsidies for

anybody building chips inside the US right so you could say like similarly in in like the ethereum ecosystem we want to have like a bunch of money for people who are doing X and actually we kind of already have this and I'll I'll kind of explain why in the future so the word we're going to coin and maybe this is like a working title so give me some feedback on this is autonoma so you may still have Dows the kind of meme of dows is fairly set but we might want to have this thing that sits above it we're calling that autonoma so kind of taking the A out of dows and just saying like what if we just Target that um and so an autonoma is one of these like conditional funding markets it's just this it has some metrics it has some goals U maybe it has a basket of metrics

maybe there's multiple of them and they're just Distributing Capital based on these metrics right right actually if you think about this is kind of what retro public goods funding is trying to do um and it's capture resistant right it's markets who run these things it's uh autonomous it's transparent and it's rule-based right like a a rule-based order which we quite like and you could imagine this being the system that's Distributing funding to the lab's organization it's Distributing funding to the foundation it's Distributing funding to the Token holders to the users like it's it's just it's it's a way of Distributing Capital where it needs to in order to kind of like keep the organization intact and the great thing about these things is they're just pieces of software they're just contracts they could be composed they could be like mapped to each other and so you could end up with this structure

that looks like protocol autonoma which sit at protocols and they distribute funding directly to the labs organization the foundations then you could have these Supra like buckets of capital that distribute funds across um ecosystems across protocols and then you could have this kind of like Su like super super uh versions of autonomo which are like right at the top end and they're just ecosystems they just distribute funds to ecosystems and when I did this I me it's pretty messy but when I did this I realized that we kind of have achieved this with the like L2 road map because if you think about the ecosystem autonomy you can kind of say that's the EF like grants right and then if you come down you can say that super autonomo is kind of like maybe getcoin grants right um or you could say that this is like l2s right L2 um grants like

retr funding and then that comes down into protocols and they distribute further funding so again this is what we believe is like a really exciting vision for what um like Dows could be if we just take the autonomous part out and we start doing that properly and we can add all the other stuff in later um so for me this is about on distributed intelligence right embedding the intelligence in the network and and what I did have here uh was instead instead of saying this is a uh a bet on distributed intelligence I just said the the implication is that we just beat salana um and what what I'm saying here is that um we already have a system with lots of distributed intelligence we just don't have the mechanisms to harness it if we can get those mechanisms correct we can have this distributed

intelligence with lots of these like automous like groups who are all working in the same direction without having to centralize control and this is about distributed intelligence over centralized intelligence AKA salana um and so this is a great example of prot protol Guild who's familiar with protocol Guild um really great system it's a list of people right these people are doing really important work technically offering a public good to ethereum by working on ethereum protocols or the the ethereum protocol itself um and this just distributes money to them money goes in distributes the money to these people in the list it's really simple but it works really well there's some rules by how you end up on this list which are considered like fairly capture resistant or credibly neutral um and it

is it it manages to attract and retain some of the best talent in ethereum and so we could have loads of these conditional funding markets Distributing Capital under certain conditions that we all agree on in our Collective benefit so if unop fund tally will this create more value for the unop ecosystem that's the question we should ask ourselves when we distribute capital from any of these protocols right if it was an autonom running it this is how this is the question the autonom would be would be asking but also we should ask if Unis fontal will this create more value for the ethereum ecosystem right because if it does then we should be Distributing more capital and that's basically where I'd like to end conditional funding markets could become like one of the most important Innovations if we get the capital right and if we get the metrics

right there's a big if and I think we still need to do some work on that but yeah this is conditional funding markets um so yeah if it works out being long me is being the same as being long the protocol and absolutely being the same as being long the entire ecosystem oh and also there's I keep saying this the end this is the end this is a an example of that same question but as a conditional funding Market um and that's it um thank you for listening to me talk about conditional funding markets we're really excited them about them obviously um and if if you're interested in learning more please come and ask we're actually going to be rolling these out in the next couple of months months with again two protocols um one we can't announce this yet but the other one is unis Swap and

um yeah we should have these conditional funding markets out soon and I hope you will jump in and participate thank you [Applause]