Fireside Chat: Balaji Srinivasan & Shayne Coplan - TOKEN2049 Singapore 2024

all right cool well s Shane um you know I uh I actually I both a disclosure and a I guess a small Humber brag I I had the fortune to back you in 2020 um back when poly Market was just uh a twinkle just barely a twinkle in someone's eyes at seed stage right and now it's actually something that is cited all around World in fact uh we were just posting on this other day poly Market is the news and maybe you can talk about that and explain what prediction markets are what you built why it's the news yeah I mean simply put when you think about a prediction Market it's the idea that if you let people bet or trade with one another on the outcome of an event what people are doing is they're pricing the likelihood that something's going to happen and you can look at the price Discovery the price of yes of it happening as effectively the likelihood or the probability that it that uh that it'll occur and you can track it with a chart over time and um if you take all the the Academia and like the the nerd stuff uh about prediction markets you just think of it like hey it's this percentage likely to happen um and this is how it changes over time it becomes this really digestible Mass Market concept that uh even now we're just starting to begin to see see the see the potential of so it's exciting that what was once confined to sci-fi books and Academia and and is is now something that's widely applicable and accepted a show of hands here who here uh has a poly market account or knows a poly Market okay more people should that's pretty good it's actually not I'll take it yeah yeah uh you guys should just check it out because um so one of the big wins recently I think breakout moment was the Biden prediction right maybe you could talk about that yeah I you know there was always on P Market a considerable chance that Biden was going to drop out and people said that there is an anti-biden tilt but in reality the murmur in in the the group chats was that hey there's a significant chance that he goes on there there's this early debate there's chance he goes up there he botches it um and then there's calls for him to drop out and that's why you know it's nonzero conventional wisdom would have said it's 0% chance as as that debate that first debate happened and it was almost like the perfect storm for poly market and it it seemed like every group chat every person I'd known was reaching out and and there was this big Wall Street Journal article about the market that everyone's refreshing um and people had said to me like hey there's two ways this goes this plays out one is Biden stays in and your site has lost all credibility and two is you've changed like the course of history um because in these like damn chats like people were sharing it and being like hey you Biden has no chance according to poly Market um and you know I I'm surrounded as being born and raised New Yorker like a lot of people who are are really staunchly behind a guy like Biden um in terms of their political beliefs and even until the final day you know when they when they see their Guy saying hey uh I'm staying in like everyone's going and taking that at face value whereas polymer was like hey don't be too sure about that it it provided this this measured and and unbiased uh public good of being able to just say hey when when it comes down to people putting their money with their mouth is it's far from a certainty if anything it's it's more than likely that he does drop out and that just helped people be more measured in in it's it's kind of like diving into water though because there is resistance and it's now completely gone nobody's like yelling at you oh my God you called for cares like the same people who would have been like really mad like two months ago are completely over it and don't care and actually they now are like yeah you know you guys called it right and they give you credibility there they have no emotion about it anymore right absolutely there there is that you know the the truff of disillusionment right like there's the elements of cope and usually if you just keep your head down for a few days it kind of it dissipates because truth prevails and you know the thing about the that graph in particular if you if you guys Google poly Market as the news you might see my tweet on this or I paste the the graph and it looks like you know um if the if the y axis is 0% Biden you know runs for reelection 100% uh he runs for reelection it was at like 90 something percent which was already like 10 points off and then it kind of went like this and then one fine day just went jaggedly down like this with a bounce up and down and then it had zero right and to my mind that's almost the inverse of the typ startup exponential right because you're flat for a while and it's finally recognized and you go vertical like this right so you're kind of flattish at like you know a few percent for a while and it's recognized and then it goes down like right right go ahead well the cool thing now is with pmer you can almost quantify like an implosion you can look at it uh you visually like that graph can go and embody something like that and what's funny is if you know you're saying it's the inverse of startup exponential curves or like an s curve um you know now that now that poly Market exists now that it's part of the news now that it's part of the cycle of something happening and and playing out it'll be interesting to see if that almost like the gardener hype cycle too where it's like if there's I I if if the if the trend ends up being something that's almost like reusable where where like the sentiment and how it how people cope and then come to terms that follows the same the same Arc I I think that's I mean because the thing is there's uh there's a similar senent with lots of startup things where they're ignored then there's a period where they're going vertical where they're resisted and everybody's yelling like remember how oh my God Uber oh you know they were all so mad about Uber like six seven years ago lift ride sharing whatever and now it's just a part of life and there's zero emotion about it or what have you so it goes from nobody cares because it doesn't work then everybody's mad because it's working and nobody cares because it's working right that's that's spot on that's spot right so um you know I'm interested in prediction markets for a few things first is you know one of things we've talked about is um I'm not so sure prediction markets predict the future but at a minimum they verify the past which they're like verification markets right because they give a strong incentive to actually uh it's it's like um let's take a horse race right if people are betting on a horse race you know if number seven or number three or whatever is going to win but you definitely have a record of which horse won the race and it's digital and it's written down because money's at stake right so there's an incentive to keep good records of the past and that's actually less trivial than people think to actually write down and and get a decision you've probably seen some contentious who won type bets around maybe you could talk about that well ultimately the Market's pricing ability is only as good as its as its resolution criteria um because you can have a question but at the end of the day the way it's going to resolve whether that means it's you know if there's the potential of a faulty Oracle or something like that like that is what's being priced not just the the headline question which I think we you know at least to date the the people who are working that are always continuing to get better at and um they're they're the best it gets when it comes to writing unambiguous resolution criteria what are what are good examples what are good and bad examples of resolution criteria for prediction Market it's funny because the the the new llms actually do an incredible job of auditing it as well as well as drafting it um I think with 01 you probably get even better with the advanced reasoning capabilities but interesting um because they think of all these edge cases and exactly uh I I mean ultimately you know there's this idea of the arrow de bro complete markets um ideally when you read the resolution criteria should be very explicit why it happens and then if it if it isn't that it's the other side it's a no uh but but the thing that I find most interesting about poly market and prediction markets is people are desperate to to just consume information about a current event that they're that they're following and you know there there may be push back that like the price at that moment is inaccurate Etc first of all if you think it's inaccurate you can profit and the more and more these markets become a mainstate the more and more that's going to be popularized as as a way to participate in what's going on in the world um and the way to consume news but but the second thing is just that the market unlike a stock or a cryptocurrency where it's like the market can stay irrational longer than you can stay solvent like these markets are not levered and they very explicitly resolve and if you're right you're going to win if if you're wrong you're not so the intrinsic value of an outcome token is the probability and maybe people will you know dispute that but that that's that's the the rough mass of it and uh what's what's exciting about that is it it puts you in a position where at least the Market's best guess for is is something that anyone can go and reference and if you think about what is that how do you Benchmark that against some other way of predicting and it's like okay well either you're going to go look on your news feed if you're on your news feed you're going to be S there's an infinite user generated content and they're going to just serve you whatever it is that's going to keep you engaged if you lean left they're going to keep serving you left off left until you're all the way on that and vice versa for if you if you lean if you lean right and um or you look at polling and it's like it's this antiquate technology that again there's a ton of different flavors and it's like what what plays into your confirmation bias so at least with poly Market there is something that is real time that people can engage in and site for free um that at least gives someone like a guiding a guiding principle or a guiding measure of of the right way to think about it and I think just the amount of adoption of poly Market as an information Source tens of millions of people if not more um It seems impossible that we don't one day live in a world where every time anyone's going to make a decision they're looking at what's what's the PO you know what's poly market pricing this at or what does it say about this and if you know that something's very likely to happen all of a sudden you can go better planned for the future and I I think that flywheel of people being able to uh get more information about the future make better decisions have a better outcome so on so forth is is awesome so okay so now let's talk about like um you know let's call criticisms or things to tackle in prediction markets so one big one is being the question of liquidity right like you know political markets H are so large because they're National or International in scale everybody's got some familiarity with what's being wagered on and some strong opinions on it and they want to prove it and so on so those are the things sort of get liquidity first they're like I don't know the BTC USD of the space right but what are the other order books that you know is it just going to be all politics for a while what are what are things you think because I've got some thesis but I'd like to hear yours yeah if you go down the rabbit hole of poly market and you check all the different markets people again and again are surprised that like the the tightness of of of what people would say longtail markets um as well as just liquidity and the volume and I think when we think about building this because people were always like oh it's extremely difficult it's impossible to provide liquidity for these markets that's not right it's very difficult but it's not impossible and you know if you if you tone it back it's like in ancient Rome gladiatorial matches people were making markets on it so if there is an outcome and there's disagreement there's a market to be made so the question for us in order to go and bootstrap this and you know we didn't have endless ammunition to go and just flood incentives to to go and boost D quity we have to figure out how to do it in a sustainable way which in hindsight has has been really beneficial for us and and kind of been the reason why poly Market is the one thing that that works um and it's it's basically going creating the supplies of tooling that helps people price these things uh it's interesting it's like once you have price Discovery and again there can be a margin of error maybe it's 10cent wide um in terms of like the the Market's confidence interval and and and you know what the market prices um and then once you have price then you can start to focus on growing liquidity and once you have price and liquidity um all of a sudden now you can let people go and participate in markets trade bet on things that they never would have been able to otherwise which I think is really exciting uh but we had to just get into the nitty-gritty of um how can you make it just so that it's in the early days you can solve the chicken and Ne in the early stages of a stand loone Market you can solve the chicken neck problem and there's a ton of technicalities and specific spefic for it but I think the secret sauce of what we've been able to do is we we we got that got that right where if there's an outcome if there's a probability there are people who almost like the Philip tetlock super forecasters just think probabilistically um and making it easy for those people to start pricing the market and posting orders is let me let me give some examples of things I think are interesting as prediction markets uh and then you tell me if you think there' be good bad or whatever right yeah so um so kind of the of people who are on poly market now are crypto te or cryptotech politics people let's say disproportionately right so for example you could have a prediction Market on um will Apple ship a three-fold phone right uh or will this GitHub new open source project that everybody thinks is awesome uh like what was that uh npm replacement bun right okay well like just some seemingly obscure thing will it get to 20,000 stars on GitHub or whatever even that's actually a pretty important you know library but and of course there's ways you can game that people can do fake stars to try to get on top of it and maybe that's where the 01 you know the AI can can uh can anticipate that it has to be like 20,000 real stars and real humans and we run some proof of human over it and so on right so a like kind of predicting when products will ship B predicting how many Star likes replies all of these social media stats are unambiguous and you can get time series data out of them and you can say you know you know the meme economy stuff have you seen that I like memes yeah yeah well so there's there's like a subreddit called meme economy and sort oh the OG meme economy like the Pepe stuff yeah and it's basically saying like uh when someone sees a new meme that they think is is going to do numbers right they should immediately go to poly Market and put down a bet on that yeah that it's going to do numbers that that essentially this will get millions of views or whatever right the problem is right now is uh often a meme like that getting the analytics across the entire web is hard right but you can imagine Sil versions of it which is like this YouTube video is going to do numbers right exactly right so go ahead I mean ultimately I think every data feed you know there'll be there'll be an opportunity to create markets on it and especially if there's a dispute in terms of opinion or um you know some people have optimistic views of it and some people have pessimistic views I think you know the all those things make sense and obviously you think about YouTube video it's like okay someone buys views right I think when when poly Market matures and people can ask their own questions you'll start to see the use case of bounties become a thing right because if it's one-sided that people can collectively or one person can go in and uh and go and like yeah people want to self fulfilling prop like there is a Continuum between a prediction market and a bounty because usually people who care enough to put down a prediction Market want one outcome or the other to happen and often either they want to make it a reality or they want to hedge against the undesired eventuality that becomes reality that's frequent not maybe not all way but I I think that's the those are the trading use cases like you're either going to hedge or give an opinion um and the opinion could either be through analytics and research or it could be through a hunch right um I think one of the coolest parts of poly Market that we've seen is like there is a lot of uh criticism about derivatives where it's like oh it's zero sum and no wi so you're not really creating value but when you have a fairly small subset of people trading these trading on poly market and then you have you know maybe even a 100 million people who've gotten value from the odds that's positive sum oh yeah no yeah well so it just goes and takes the idea of harnessing uh people's discrepancy opinions and and and derivatives markets where people are are are swapping risk um to go and answer really difficult really important questions in a way that that everyone can benefit from so that's this that's the exciting element of the the informational component well you know there's that saying price is a signal and in this case the price is in the most literal sense a signal because uh poly Market uh subsites for or um even invents data that polls wouldn't capture right like with a poll you know some launch poll as we were talking about it's very it's quite expensive to go and do a Pew or Gallop pole over time points that's representative across you know whereas you will get a you know crypto like 247 time series of people's confidence in this matter exactly and and the audience is basically or the market is subsidizing it right and so that means that you can get much higher resolution data and how much are they really paying basically it's like a few people you know essentially having a zero some game with each other fine but that that seems like the societal benefit of having those kinds of price graphs on many things that were unpriced could be exactly orders of magnitude Beyond however much the Zero Sum gain or losses like for example the Biden Dropout one is in a sense trillions of dollars in global markets could flip on you know what regulatory decisions are made one way or the other for example right totally I I mean a lot of these it also when you think about it like that it's like if you quantify the like the economic impact of some of these big decisions or big geopolitical events all of a sudden these markets that may seem kind of toyish all of a sudden seem a lot more important um yeah yeah cool it's it's uh what comes next it's exciting I think that uh ultimately like you were you know you have all the you have all these ideas a lot of people have ideas they want to see markets from um poly Market as it stands right now you know the the terminal state is definitely people asking their own questions and we're excited to to figure out how to bring that bring that to Market cool awesome well um one thing that um you know oh gosh there's one other thing I wanted to say which is um the there is we mentioned a relationship between uh prediction markets and bounties but also between prediction markets and oracles right because that stream of Oracle data is often something that's the input to prediction market so but you haven't actually tackled like crypto prices directly you do more of the political markets and so am I wrong about that I think that I think that's yeah at the end of the day right like if you think about it from Oracle is a funny word right in crypto it it means how you go and get real world information external information on train um but you know an oracle is also just symbolic of being able to tell the future so it's an it's an important word in the in the poly Market world but uh I think ultimately if there if there's a world where people can go to poly market and search up anything that's relevant to them or something that's going on in their in their life and they can just see the odds and they can see how it's trended over over time um and go see the people who have positions what they're saying like that's just going to it's just going to lead to a more more fruitful civilization that's what we're building cool awesome great thanks everyone