#191 Richard Craib: Numerai - A Revolutionary Hedge Fund Built on Blockchain and AI

Transcript

[Music] this episode of epicenter is brought to by Ledger and The Ledger Nano s half peace of mind and knowing your private keys are protected by industry standard physical security go to lal.com to learn more and by shapeshift.io the easiest fastest and most secure way to swap your digital assets don't run the risk of leaving your funds on a centralized exchange visit shapeshift.io to get [Music] started hello and welcome to epicenter

the show which talks about the Technologies projects and startups driving decentralization and the global blockchain Revolution my name is Brian Fabian Crane and I'm Roy today we'll talk to Richard crae who is the founder of numerai numerai is a new kind of hedge fund built by a network of data scientists that is using blockchains and cryptocurrency in an especially interesting way so uh let's get started great to have you on the show Richard thank you good to be here tell us a bit about your background how did you come to be involved how did you end up doing what you're doing right now well I um I originally

um was uh I was always interested in math and finance and uh I started working as a Quant after I graduated with a degree in math and um machine learning was just taking off at the time and so I started to uh work on these data sets that uh you know they were using Quant at this asset management firm I worked for but they didn't use machine learning and so I was got to basically create M their first machine learning algorithms and make a whole fund that was based on machine learning um around the same time I was also reading about what was happening in blockchain and encryption and I was you know invest in the in the ethereum crowd sale and

the Olga crowd sale and I was sort of kind of keeping an eye on that as well but mainly my focus was uh was on machine learning and um they ended up building this fund that uh worked very well but um we couldn't really uh collaborate with anyone because the way Finance is set up is like structur natur Al uh everything is secret and no one's working together and so I had this data set that was proprietary that I couldn't share with anyone um and I started to have people I knew who were better than me at machine learning but I couldn't collaborate with them and so that's where I started to think about the idea for a hedge fund that had a public open data set that anyone could model and uh

so that's where I started had the idea for numerai cool yeah I mean I guess guess if you R it from that from that position that you were in it's sort of almost an obvious idea right like you think like okay it should go in that direction right that would be better if it did but when you first had this idea how did did you think it was possible that you how did you think this was going to turn out it was uh it was a feeling of knowing that um everything made sense uh and I'd really thought through it because I'd been working on this data set for so long um the Big Challenge was how do you frame a data set a financial time series data set into a machine learning problem that's

actually quite fun and um a lot of the competitions machine learning competitions on kaggle um they had uh every time there was a finance one it was actually it was actually uh bad uh it ended up being very easy to cheat or uh overfit um and time series is somehow not really the the right kind of uh thing for machine learning and there's very low signal in these data sets so thinking about all that was the hard hard part and uh and the next step was to actually be able to share it without giving away the data and so that's where I started to read I started reading about hom homomorphic encryption and how

we could use that um to basically share the data while preserving all predictive structure but having it be that none of none of the people who are looking at the data had any idea what they were modeling and that was the big uh Insight um because if you give people raw data they end up uh cheating or overfitting uh or really they don't most machine learning people don't want to learn about Finance either so they they make a model that wouldn't be wouldn't make sense but by Framing it in this very specific way um we can get models that that actually Help The Help the hedge fund cool well I mean we're going to speak a lot more in detail about how

exactly this works but just to give some introduction for people who aren't as familiar with hedge fund and Quant trading when when you were managing this fund was this at the hedge fund as well or was this an asset management company so can you can you give some overview of this industry yeah it was at an asset management firm they uh they have a number of different fund products they had about managing about $1 billion um but a lot of it was long only and so that means that you only buy the stocks you don't go short um but when you have a really really good Quant model the most profitable way to run it is in a long short uh style because then

you get the alpha or the Returns on your Longs and your shorts and and are hedged against Market risk and other risks so um when I you know when I did build this it was to create a long a long ownly product to beat the all country World index which is like the S&P of the of the world um and uh it worked really well um and but it was really much more suited for a hedge fund strategy so that's why I quit my job moved to San Francisco and started uh numai um and I think there are definitely lots of hedge funds out there using statistics and mathematical models

but I don't think there are that many using machine learning kind of at their core um so even just having a great machine learning hedge fund would have been a good idea but I but I think um the crowdsourcing and the uh collecting intelligence from all around the world would took it to another level and and in general what kind of big trends are happening in the hedge fund industry how is that industry changing well they definitely have Quant investing On The Rise um I think it's I think it's close to $1 trillion or something in Quant and it was like 500 billion a few years ago um so it's definitely moving out of Human Fund

management into Quant fund management but it sounds a little bit cooler than it is Quant um it's really a lot of it is still people um but they've just made some mathematical model um so it's automated but it actually isn't really artificial intelligence um so I think what's going on in inside of uh Deep Mind and Google in terms of AI is you know way ahead of where hedge funds are at uh with it um and a lot of them are are really doing things that theyve they've been doing for many many years um certain kinds of statistical Arbitrage strategies and so a lot of them are actually extremely correlated

so when they go down they they all go down together even though they're all hedged even though they all claim to uh be uncorrelated um so it's it's a strange thing where you have um the shift from fundamental into Quant but now Quant itself is crowded and um and that's why new hedge fund Innovations uh I think it's a really good time for it um and there's certain hedge funds that are doing things in cryptocurrency and Ai and I think right now that it's not clear you know who's going to uh uh be be the leaders in this next wave of hedge fund but I think it is clear that there is a new wave and that um that traditional Quant can't really lost the

way it has so uh let's let's go into what uh numerai does but before that um I think in in many of our discussions we are coming across this word which is a data model right so can you explain um what a data model is can you give us an example of a data model and how uh Quant fund would use a data model yeah well um you're just looking for a patent so one of the easiest ways to think about machine learning in terms of modeling a data set is just imagine you have a scatter plot and there's just a whole bunch of points in 2D uh you can fit a line to that uh data uh and um maybe

it's like uh corn growing over time or something like that and you could say after to six weeks how far would it have grown uh based on some historical data and um you can just use a simple linear model to predict that and and then you've Ed data created a model and now you can predict corn growth um now taking that into other domains it gets more complicated because you have many many different variables so I think on numeri right now we have 20 21 features um so that's like a 21 dimensional scatter plot and you're not really just trying to fit a line you're trying to

fit like a very complicated curve in 21 Dimensions um and so that's the kind of principle of it and what you're looking for is patents and uh finding the best best model that's possible on that data set uh based on the patn in it but as you can see like nothing I've described needed any domain knowledge so you don't need to know what the axes are in order to fit a line and uh you don't need to know what numer data is in order to to make a great model on it um so that's that's what uh what modeling is all about and um you can go from it's gone from being very simple uh having just a

a linear model to very very very complicated with the things you can do with uh deep learning and tense of flow neural Nets so applied applied to like Financial trading could could you give us like a toy example so you're talking about these 21 Dimensions could you give us a toy example of like maybe something like something that is like three dimensions or something and um is a relevant example to the financial industry well um you could uh a lot of people uh there's two two big Quant factors uh one is momentum one is value so how cheap is a stock on like a PE basis and momentum

is how well has it been doing lately in terms of the price um and so just fitting a model to those two features um you can say if a stock has good momentum does it go up if a stock has good value does it go up and if you take lots of different stocks over long periods you might find a patent that actually says yeah if momentum is high and value is low uh it's a good time to buy and uh and the stock uh stocks that have those properties will tend to tend to go up um and that's what you can learn from a data set so I downloaded this this data set before and I had a look at it because I was just you know kind of curious and yes as you pointed out right

there was uh you know you have this thing of of an ID right which is some number and then you have all these features like 21 features and then you have this thing called Target so so it would Target be the price of a of a stock or is that a way to think about it it's more yeah it's more complicated and um we don't really talk about the data that's why we we uh we encrypt it and the way we do but um it is it is it is worth thinking about it like that um that the a one is like up and then like a zero is down so if you have Target of one that means the stock is is good

Target of zero it's bad in some abstract way um typically you don't really just want to make money uh you want to actually have you know also hedge out risks so um we put a lot of things into the data that actually make the models uh very nervous about taking risk and so you take away all the p patters that uh that are too obvious um and actually risky but so it's not like cuz you know for me I I would think okay so you want to make money in the market right and you're going to say okay this Apple shares right so let's say that would be uh a variable and then you're going to take uh all these things maybe I don't know

Consumer Price Index GDP growth uh exchange rates or all kinds of stuff and you you try to understand okay what kind of Dynamics are there that affect uh this share of apples and then you know you give this data and then people figure out even though they don't exactly know actually what they're modeling is that wrong uh is that how it works or or is that different that's pretty much uh pretty much right we we do um we do Global Equity uh it's a global Equity long short hedge fund so we're actually not too interested in where is the market going to go be next month we have no idea and we don't need

to know um we care about where individual stocks will be um and we also don't really care about where interest rates will be or where um GDP will be so all those things are kind of like macro uh variables and we are really looking for idiosyncratic stock uh risk uh that we should take so um mispricings in stocks and uh it turns out you don't need you don't need to know you know everything uh about the macroeconomy to to find mispriced things in the market so some people may be aware uh which is is always you know you sometimes see it on

reddits and different places where you have this technical analysis you know people do these sort of charts and lines and you know as I I studied economics in college this always look extremely ridiculous to me uh but but I don't know maybe it's not so is is a lot of the data coming from that kind of thing too that you say okay you actually going to just look at this price feed right over time and then try to figure out things like momentum right because something like momentum you would see purely based on the changes in price of the asset itself yeah um you we're definitely not doing like technical analysis uh like like like the pictures you see they but obviously there like momentum is in a

sense a technical indicator in that it is a it is a uh based on price um based on total return so um that's a it's a variable based on price uh things like PE um it's more about you know whether the stock is cheap for its price versus just uh to De with its price um generally on shorter time Horizon you can have technical analysis style things um like relative strength index and things like that they actually M they actually work on on shorttime Horizons um but we actually hold things for a very long time so we don't really look at uh look at that um but certainly yeah

there's I don't think uh the chart chart theory has um has much uh credibility anymore this episode of epicenter is brought to you by Ledger makers of the Best Hardware Key security Solution on the planet but Ledger is more than just a hardware wallet it's your path to Eternal Bliss and happiness and peacefulness do I look like I'm losing sleep I am but it's not because I'm worried about my cryptocurrency my Bitcoin or my ether that's because I use a ledger Ledger devices support multiple cryptocurrencies like Bitcoin ether zcash and more and you can even secure your ERC ethereum tokens with them or you can add the security support from

ledger to some of the wallets you already love and use like electrum copay my ether wallet and others all your keys and segregated accounts are derived from one unique seed seeds are generated on the device and are never exposed to the host computer so when you make a transaction your Ledger will present you with the details and kindly ask you for your confirmation before signing how polite is that so the best choice right now for for anyone looking to invest in security is The Ledger Nano s it's a keychain siiz device that fits in your pocket it has a screen and buttons and connects your computer or Android phone using USB look if you're holding crypto and you're storing your keys on your computer on

your phone or Wars an exchange you know that's a disaster waiting to happen don't be the person that loses their keys because they were careless with them so don't wait any longer secure your Bitcoin secure your zcash secure ether go to Ledger wallet.com and get your Ledger Nano today we'd like to thank Ledger for their support of epicenter can you explain why why are these data sets so valuable for hedge funds why is there such a you know keeping the data private well it is just like it's sort of the core so um if we gave it gave it out um for us our users could just go and start their own hedge fund uh with our data taking everything we've bought

and learned and uh and then just running away um so we wanted to make a system for for us in particular where you would actually have no incentive to to do that and actually have more of incentives to work together um and to build value together um but generally um the data sets are very expensive and um so hedge funds don't want to part with them and then um they also have uh there are lots of um different kinds of data sets too so there's some hedge funds that talk about buying satellite images to look at uh Park the parking lot uh see how many cars are at Walmart

so there's all these different weird weird data sets and some of those are very expensive because it's like the hedge funds have have to pay for the satellites and stuff so um it gets kind of crazy but basically you can't give up your Edge and that's why you it's so secretive but in the tech community and in other Industries actually you can both benefit um and it Finance is just so so competitive that uh you can't you can't have a situation where you you're helping each other and so the best hedge fund managers they can be friends for 30 years and they have no idea what each other other's hedge funds do it's kind of weird so it seems to be that the key here is to have private

data but to enable collaborations like computations collaborative computations on that private data without disclosing what that data itself is yeah so what uh what kind of Technology does numerai use for this purpose well yeah so this is the the key thing um when you give away a data set one thing you could do is anonymize it right so you could just say okay well this stock is Apple I'm just going to call it um ID number 7 this stock is Google I'm call ID number eight and so you take away all the names in of the stocks um but that actually isn't isn't enough because the patterns

uh inside that data are still so strong like everybody will know oh I know Apple's p is 10.7 because I can I can I can cross reference it with Yahoo finance and um so you can have kind of weird things um where uh people can use use do comparative attacks to basically figure out your anonymization scheme um so you really want to go much further than than that and into encryption um uh and so I started looking at what's called fully homomorphic encryption and it's this really cool futuristic thing uh when it was invented it was about a billion times too slow and then uh a couple

years later it was like a million times too slow and actually hasn't got much better than that so it's this very very slow thing and what it does it takes one megabyte of data and turns it into 16 gigabytes of data and that so that's a huge uh weight plus it also um makes it much harder to work with with machine learning so what we ended up doing was figuring out ways to to do it uh differently and um uh we're using something right now called neuroc cryptography in our in encrytion so our VP of engineering this guy called Jeff Bradway from Deep Mind and uh he he came up with this way of uh encrypting the data that was uh using

two neural Nets so it's really interesting you have one one that's saying um uh F like take this data find uh find some features in it that preserve the signal of the data and then you have another net that's saying okay I'm going to try to I'm the aders serial Network I'm going to try to decrypt what you just did um and then you make them kind of fight with each other and uh the one is saying I've preserved the signal the other one is saying I can't decrypt it anymore then uh then you know you have something that uh that neuronet can't uh can't uh decipher what it is um so we do things like that to make sure

that we have all the signal preserved but also it's very difficult to kind of invert the transformation that we made um and so far you know we hadn't had anyone say what the data is or reveal uh reveal it so most of them don't really care about it they would just want to solve the problem uh they don't really mind they don't really care what problem they're solving um but that is kind of a critical technology that we've had to develop and I think we're pretty you know have done some really interesting things I don't think there's any company in the world um with the problem that we have which is like we want to work with everyone but we have we don't want anyone to know anything and um it's an

interesting uh creates an interesting Dynamic I I just wanted to uh sort of add one point here because I guess even if people decrypt the data okay they would have the data right so that would be you know certain loss but they wouldn't really be able to replicate what you guys are doing right because you also get uh the predictions from all these different people and presumably those aren't public so so you are the only one that can then combine all those is is that correct yeah exactly um and that's that is an important point we are not trying we're not looking for um the best model and then we're just going to shut the whole thing down and hire that guy

give him a job at numerai um they're we're really looking for lots of different uncorrelated models built on our data um so what would happen actually if someone were to decrypt it and say they were going to go start their own fund with their model sure we would lose them but our combined model will be way better than his um so it's like even if you're the best user and you and you quit numeri the remaining users are still better than you when you put them together um and that's uh that's this principle is called Ensemble Theory um and it's kind of a big part of machine learning it's like if you have

lots of little models um that are different then you can actually make a make a much better model um than any one model so with these with these models there's like there's a way by which you Source these models which is like through your crowd mechanism and then you also need a way once you have sourced these models to combine these models so let's first talk about how how how how do you decide how to combine these models is the are the combinations also sourced from the crowd or that's totally know yeah that's that's where we stop uh we we have to people have asked us why don't we let the crowd you know combine the crowd models and then it's like well who's going to combine that

but um so we stop there and and we that's the kind of one model we make is the model that that uh is the meta model that combines all of them um we first have a few different measures uh one thing we don't want to do is pay is pay you for giving us something we already have so if you submit a linear model we can we can run a linear model ourselves believe me uh and but we'll pay you but if if you if someone else submits a linear model then he doesn't get anything because he's very correlated with something that we already have so um on numeri we're really trying to look for original

models and so if you pass this test called originality then uh then uh you kind of you can kind of earn earn Bitcoin um and and if you don't pass then you kind of don't earn anything and the other one is uh concordance which is like it's sort of to do with how much we think your model is kind of like Shifty uh like because there's lots of different ways that users might try to gain the system and so we have another check for that um and then we we have we rank them all based on logarithmic loss which is like a machine learning metric um and after that we have basically about 300 or 400 really good models that

pass all these tests um and then we do a layer of machine learning on top of that so we say Okay um let's let's try to combine these and uh you can take a simple average that does very well and you can go even further um and do machine learning and then that does even better um so that's that's basically how we do it and so the other part of your of your of your company is the sourcing of these models right so so explain to us how how you Source these models currently well all we do is um I just made this website and uh put put a data set up uh in 2015 and just people came uh it's it's it was a very interesting

uh problem it's very difficult to do well it's extremely hard to do well on the data set uh there never been a a data science tournament on encrypted data there never been a hedge fund that was actually going to trade the predictions um we also paid in Bitcoin um and we also one of the key things is we don't actually get the users models um all they do send us predictions so it's like we're licensing predictions from their models not actually taking their models um so they're submitting a bunch of predictions to us um and that means that we don't know what their model is either so they have no idea what the data is we have no idea what their model is um and and that that

is a sort of kind of yeah makes it symmetric in a way uh users could say you're not paying me enough I'm not going to do this anymore and we'll we'll have to uh they'd have a position a barain in position uh and that's why I think a lot of the blockchain community likes numer like this sort of trust trustless part of it where we don't trust you with our data uh but you they don't trust us with their models but so if if you don't have the models themselves how are you able to combine these models and apply machine learning on the on their combinations cuz like you can't you you can't generate infinite like you can't play the model on different data sets

indefinitely Or or can you yeah what we do is we make them submit on a big holdout set I think it's like 100,000 rows um and they're submitting probabilities on all of those so having it be um on this big holdout set having it be like that they're submitting predictions uh for this over this very long period we can actually train the model on on like almost like on their back test so um so it's almost like they're giving us probabilities uh that we can model without actually giving us their model so but that's the importantly that means we can only get new um new insights

every time we do a new competition so right now we do a new machine learning competition every week and that's to get updates from our users it's actually to get get probability updates on what they think is going to happen cool uh yeah this is super super crazy so so that means the the they go like a year in the future or can you say anything about like how long cuz cuz you run it every week but presumably the predictions that people give are actually beyond that time Horizon yeah we end up holding uh things for 6 to n months um which is quite long in Quant space um and so but but we do like

getting weekly updat dates on okay maybe we should rebalance uh or sell um things if if suddenly the all the all the numeri data scientists think we should and and why do you hold them so long um I think it's hard to compete in the shorter time Horizon um and uh there are very few people doing uh machine learning on longer time Horizon I think there's basically lots of mathematical modeling happening in these short time Horizons one week uh one minute or less um but going further it's actually it's actually harder to do and so fewer

people do it uh it's harder to do from a machine learning standpoint but it's actually easier to implement so to start a new hedge fund that do does high frequency trading and it would cost you like a $100 million just to get set up um but if you can do longer term predictions then you don't have as many trades you don't have to worry too much about execution you actually don't even have to worry about being ripped off because um you're going to hold long enough to cover those costs of of being ripped off in the by high frequency Traders so uh it's like a sort of a special uh time Horizon I think uh for AI and also you know sort of philosophically morally

maybe you want uh to be actually investing right you don't want to be um just just finding a picking up nickels uh in the market you want to actually be holding the stocks that you like and um don't want to have an AI That's just like fighting to compete in short short time Horizon you can actually have an AI That's like a real investor that's like making decisions about buying stocks that actually people would want to hold that that's very interesting how you sort of you know uh almost make these moral judgment and emotions about these like predictions that the AI is going to create it's not a

judgment The High Frequency people are doing uh doing doing something good for the market um but I do think it's it's over overdone and you never want to be doing the thing that's kind of already happened uh so if there's a book about it you don't want to be doing it there's no good no good book about blockchain so you probably still want to be doing it cool so so so tell us about tell us how this tournament works so let's say this week right so so you you run a weekly tournament and I'm somebody who builds a data model can I use my model that same model week after week after week and keep earning payouts yeah you

can you need to develop something new yeah you can actually because the training Set uh Remains the Same week to week um we're just asking you to predict on the latest live data basically so because the training stat stays the same um you can actually set up a server uh can automatically send predictions as soon as new numeri data uh comes comes about um and so some people have actually done that and so it's not actually them uh spending hours and hours working on it they've just put a server up and and and we're pinging that server effectively so so so basically like you you're asking all of these different models so let's imagine

there's like 100 data scientists having 100 different models and here model is basically like a function right there's some incoming data and there's a function and it's going to compute an output and that output is a prediction of some form on how a certain segment of the stock market is is is going to move right and so you have so for example I have a model here I'm giving you a prediction about how a certain part of the stock market is going to move and there's somebody else around the world that's doing that and there's hundreds of us basically you're pinging all of our servers and getting all of these predictions and then you have some method of like combining these

predictions and converting them into actionable trades on on your side exactly yeah so how does your tournament identify uh so let's say like 10,000 people submitted models and in the end you maybe want only 100 that are that are unique so how do you kind of incentivize people to um what to give give new models how they exactly paid for it and how like for example my question is like so I make a model that's unique and somebody else makes a model that that's unique for a different part of the stock market how is the magnitude of how much I make versus how much he or she makes

determined it's based on um based on your life performance so what we do is if you submit a model uh we don't give you any any no matter how good it is we don't give you any money until uh a month later when we can see how well it worked on live data um and uh a month is a decent amount of time uh to to see to see whether the model is behaving the way we thought it would um and so everyone is ranked uh uh based on their logarithmic loss uh which is just like how many times you are right effectively um and so what we do is uh

we rank them based on that and then pay them based on like a schedule so um the person who's coming first uh makes about um $500 and the second $400 um and that goes on all the way down um but what so that's and but we pay those we've paid those dollars in Bitcoin um but the other piece that's new is this is our own cryptocurrency numer rare where actually people are earning hundreds of times more money based on that than they are from from uh dollars so uh that's the key uh Innovation I think as well where it's

like people wondered how you know you're not paying these people enough they could just go work uh somewhere else and I think a lot of people weren't doing it for money they were doing it because it was interesting um but now um by having our own cryptocurrency uh it's it's like we we can actually afford to pay them uh a lot of money and the top users are making a lot of money right now so tell us how this cryptocurrency works the key thing that the smart contract can do so it's an ethereum erc20 token uh it's called numer rare um it's uh it's it's live it's came out about 2 weeks ago um and what you can do with it is you know we always try to

figure out which models are good uh that's our big thing if we can tell which models are good we can make a better meta model and we can make more more money in the hedge fund um so the a one way you know a model is good is if someone's willing to back back it with their own money and so actually when I was um at my old job I was like we have to do this machine learning fund I I I promise it's going to it's going to work really well and I actually and they said okay well put your own money into it and so I put money in with the firm's money and uh and my boss's money and we all put money in because we we were we were prepared to uh stake uh our our money on it and so by having that you actually

work much harder because you like you have a lot to lose um you don't just you going to say okay well let's see what happens I'll give you five models and one of them will work and we'll just see what happens um you have to really commit and say well this is the one I want to put my own money on and so that Dynamic uh we weren't capturing at numeri before numer rare because no one had anything to lose um so we created this cryptocurrency we gave about 1 million to our users and um we said okay you can use this uh to stake your models which means a little bit like taking a side bet that your model is going to work and um so by doing that you're expressing your confidence as to how how much um uh

you think uh the model is going to work and uh if you're right we actually pay you dollars so if you stake numer rare you earn dollars and so because of that Dynamic that uh the numer rare uh is connected to the payouts uh if you if you bet a lot and you have a good model you'll get this payout um that actually gives numer rare value as well and I think that's also what uh what's uh so so it's sort of like necessarily people value it above zero because it's it's a way to actually claim dollars in this in the payouts um so and it and it has worked extremely

well I mean we've seen the stake models um be very uh you know much better than the than the regular models because because people are are staking on them so it's like this final piece that we needed and um in a regular hedge fund they have the situation where all all the employees have money in the fund and and we didn't couldn't really capture that with our user base and that's why we started to think about blockchain and ways you could do that where everyone actually is in it together and uh uh has a lot to a lot to gain and a lot to lose if numerai wins so so just to kind of clarify how numer Works does that mean you you

created have all the num rare been created or are new ones being created continuously or how does that work yeah there will be 21 million Max uh that's the cap and uh it'll get there um I think in four or five years um so every week we Mint new ones and it's in the contract so we can never change we can never change it um and right now we've we've uh haven't Meed yeah we only gave away about a million to our users so there's still plenty more to be to be one as time goes on yeah the the idea though is as well is um if your model does well and you staked it you win money right but if your model does badly

we actually destroy your numer rare uh provably on the blockchain it's not like we're taking it for ourselves or something uh we just destroy it and um so these dynamics of the minting of of new Coins Plus the sort of destruction of old coins um and also the coins going tending to be held by the better data scientists over time I think it's going to play out a very interesting way where the people who it becomes more and more valuable to people uh to have it and to hold it so so does that mean um because if if I don't have num rare I can still participate right and if my model's really good I can still get some

money dollars and numer rare or do I only get numer rare in that case you get dollars and numer rare for the main tournament you can just do anyone can do that and then the staking tournament you risk numer rare to win dollars okay okay and so let's say I'm I'm putting in uh I'm I'm risking $10,000 worth of numer that that means I can then earn double my money or how is that kind of mechanic it's actually in a kind of complicated multi-unit Dutch auction me mechanism um but it really is uh the whole the kind of outcome of it is that um the more you

stake the more you stand to win of the payouts and also um because of the way the auction set up there's a sort of property of self-revelation which means um you it's rational to to bid the true the true the true amount um so people are uh participating um and we can because of the auction mechanism we can actually say oh given that you did that and given that you're rational um we can say oh okay we know what you're we know what you've got we know you have a very very good model we know that you believe your model is going to work with uh 90% accuracy so

that's a very interesting um part of this uh and the idea that they you know the the the bet the stake uh is actually communicating information to us let me understand if I get this correctly so let's say in the future uh two years or three years and you know this is become a big thing so do you think that most of the money being paid out is actually going to be paid out through these side bets and so that if somebody comes in and they say I have really good models right I I know I'm I'm doing really well here uh so I'm going to go and buy num rare and sort of bet it on my model and then I can make

much more money and that this is kind of how the demand for Num rare would be driven is that how you see it play out yeah exactly um and we've seeing we're seeing things like this now like there's a there's a a sort of hedge fund crypto hedge fund uh well just a fund I guess um that bought a bunch of numer rare in our slack from from our users um like OTC trades before it even existed um and I was like why are they doing this are they just speculating on the price and it was actually that they they wanted to um use it in the staking competition they wanted to basically hire data

scientists to work with them uh and they would be like kind of lending them numer rare for them to stake um if they didn't have enough but they had a good model and they so it's very interesting to see these kinds of things and I've I heard other people say since the launch of numer they've quit their jobs and now they can actually do numerai full-time um and they're working with you know maybe speculators who have a bunch who can't use it because it's kind of useless to them uh because they can't stake um and then they work together uh and uh swap it and buy it and so it's very it's very interesting to see and I

think it's yeah it's only the beginning I mean we've just been going for two weeks and you have these outcomes already it's amazing and I guess what what sort of is interesting here right because you guys are switching to paying out in in ether so you know I could and since neware is an rc20 token you know you could even imagine that somebody creates some smallart contract there where you know people can put in neware there and maybe that somehow gets lent at an interest rate to people submitting models like do you see that kind of application being built here yeah um that is uh yeah also a recent announcement we're not going to use Bitcoin anymore just pay everything in ether um and it really does ow for it's big it sounds s of silly but it's a

much bigger change than you would think because we can actually do the payouts uh kind of on chain uh in the same same place as uh as where the numerar stakes are happening um so you could have you could be a Speculator I could say I don't have any numerar but you have a thousand um let's make a Let's Make a Deal uh if my model works we'll share the profits you'll say well I have no idea uh whether you're going to honor that contract because numerai is paying you uh and uh how do how do I know you're going to share it with me and you can say no I'll build a smart contract and the smart contract is actually the

entity that is uh playing on numerai and the smart contract automatically whenever it wins money it automatically shares it between you um and so that concept of making um Dows that are connected to numerai which is connected to the real stock market is extremely uh compelling um and I I do like that there's a lot of innovation in blockchain a lot of uh things like prism and things but a lot of them are on on the blockchain um and I think it's pretty cool to have like this one company that's actually connected to the stock market but like pulling out all the wealth from The Real World into

crypto this episode is brought to you by shapeshift the world's leading trustless digital asset exchange quickly swap between dozens of leading cryptocurrencies including Bitcoin ether zcash gnosis Monero Golem Aur and so many more when you go to shapeshift.io you simply select your currency pair give them your receiving address send the coins and boom shape shift is not your traditional cryptocurrency change you don't need to create an account you don't need to give them your personal information and they don't hold your coins so you are never at risk from a hacker or other malicious actor shapeshift has competitive rates and is even integrated in some of your favorite

wallet apps like Jacks so you can swap your digital assets directly within your wallet just as easily as putting on your slippers whenever you see that good-looking Fox you know that's where shapeshift is so to get started visit shapeshift.io and start trading and we'd like to thank shapeshift for their support of epicenter can you can you use your data model to trade cryptocurrencies themselves rather than stocks we could but uh there's not as much data in crypto so we have for stocks we have data going back 30 years um but for for crypto it's not as much and uh they're also fewer cryptos so it's hard to uh harder to decide uh well you

always want more data so you have 10,000 stocks per month for 30 years but with crypto there's like 50 cryptos that you could trade uh and they only go back like one year so um it's not really quite right yet I think it would be great if all the stocks in the world were on the blockchain um then we would then we'd really uh be in at a huge Advantage um but I I'm not sure when that's going to happen but I'm sure it will why would stocks being on the blockchain be an advantage to you the hedge fund industry is like very structured for for the bigger hedge funds so it's not really like structured

in a very fair way um like the biggest hedge funds who need the mo who need them who need it the least get these huge discounts from like the prime Brokers like UBS and Goldman Sachs and um and and so it's just very difficult to even start trading and you need lots of capital so I think if it were to be on the blockchain it would be much more uh open and like everything would be visible and uh you would actually have a much more fair market I mean speaking about that to me it feels it feels like there's kind of a piece of a decentralization and crowdsourcing missing here right which

is you you crowdsource the the data you know BAS decisions on the trades but then on the other side when it comes the money that you're actually investing you know you're structured as a traditional hedge fund so I think your uh you know investors are institutional investors or I guess the kind of people that would invest in hedge funds currently so do you also see that happening that you say okay you're going to kind of crowdsource the pool of capital yeah we are a a proper us uh hedge fund um and there are rules around accredited investor status so you need to be an accredited investor to invest in a hedge fund um and you

also need to yeah uh and so and then they're also very high minimum so it's really not worth our while to manage less less than 50,000 from someone we'd be making a few $100 per year from that person so you want to have these much bigger check sizes um and a hedge fund can really only be profitable at like 300 million uh AUM so you really want to talk to institutions not individuals um that being said right now it looks like you can get $300 million in a crowd sale uh if you if you if you have good branding or something but uh I I just think it's probably the wrong yeah probably the wrong thing for

us uh uh at the moment and will be for a long time because I don't think the Regulators really want individuals to be invested in hedge funds it's a little bit like they're protecting protecting investors from risks that they don't understand yeah although I guess we are also starting to see uh now with coin list a first sort of Ico platform that's uh limited to accredited investors you know so uh maybe we'll see that kind of crowd sale too yeah yeah it's interesting that's uh I like what they've done there and that that could be a way a way in the future where you could actually raise

money so you did this fantastic interview with uh Chason kakanis on on this weekend startup so we'll we'll definitely link to that in the show notes but one of the things you said in there was kind of you know striking to me and was like that that's a crazy statement but you said in the long run you see numerai invested uh managing all the money in the world can you explain like what what you mean with that yeah this uh this that statement I actually I made that to a journalist in um a few months after starting numerai so I hadn't even like really had barely started so and I just said that I want to manage all the money in the world and I don't remember saying it but then it ends up

sticking uh and um it is true though uh I do think that's what we that's the goal here um I don't think okay think about how many people are working in finance um some in the best universities about half of American the best universities go go to uh go to finance and and they're basically uh playing a zero sum game um and uh with each other and we are not having nearly as much uh investment in basic science and technology and stuff as we used to um because all these people have gone to

Wall Streets and that would be fine if they were all working on something but they're basically working on the same thing and just in different offices uh competing with each other so where it's like so that's a really big problem if NE I can make it be possible for those people to do other things um and uh and actually a system like numerai could do that um there's many things in the world that have been automated like think about the phone switching like there used to be people switching the phones to make calls connect calls and now that happens automatically um and I think we could have something like that like if you're a public company there's this thing called numeri it'll give you

it'll tell you what the price is um and uh and it give you liquidity and it'll just be like this thing um and you don't really need many hedge funds in a world where you have an open hedge fund and uh I I think think numerai is the most likely candidate uh for for this where we can leverage blockchain and AI to actually uh be the last hedge fund and um and I and so I think that other hedge funds um it will be very difficult to compete even the best the biggest hedge fund right now has about 150 phds working at it uh that's the biggest one and we have 20,000 uh data scientists on

numer right now so that's already like you know a lot more um and it'll be hard I mean we'll definitely have 100,000 in a year so it's just like I don't really see how it will work um for the for other date for other hedge funds to compete with us your answers were were really great but with this one my feeling is if imagine there were only there was only one hedphone in the world right and like it was managing all the money of the World by definition its returns would be the same as the returns of the market it couldn't beat the market if it were managing all the money in the world so I I just I just in intuitively feel

that the bigger a hedge fund gets the lower the margin by which it beats the market until it gets to such a big size that it becomes the market and do you really think that in the hedge fund industry they can any day be the single hedge fund that wins it all because why do you why do you need to beat the market if you have all the money in the world it's like but then if if if I'm not if I'm not beating the market can I just not own index funds why do I need a hedge fund it is but I mean this sort of happened to you know happening to Warren Buffett and uh and other hedge funds you know he's got so much money it's very difficult for him to return anything

different than the S&P 500 return um but I I think it's it won't yeah it won't really be framed in this way I guess think about like Bitcoin um that's how I like Bitcoin is a sort of like a way to do money transfer it's like a bank uh sort of like a sort of bank but it's like got this weird crypto feel um and you can keep money there and you can transfer money and I think that's what we're trying to do is is actually not really make um another bank but make like the bit or like another HED fund but make like the Bitcoin of hedge funds where it's like this protocol thing and it's not really it doesn't really play

by the rules um and uh it actually is a is a sort of utility for the world that no one really owns uh unless they own the the cryptocurrency or something um so I think that's kind of where things are going that like this idea of making a protocol something open um I can see yeah can see something like that working this Century I mean I can also see I I can see that claim okay we're going to or that aspiration we're going to manage all the money in the world I I can see that direction being possible if you change something fundamental about uh how numerai Works

which would be you'd have to open up so that I other people other funds uh could go there with their own pools of capital and their own data sets uh and basically you know have the same data scientist the same Community around the thing right because you know you're us hedge fund right well but there will be other ones first of all in other markets that you know they can't put money into a US legal entity right so you'll have that factor and of course the other factor is also going to be uh you know we've already talked about how data sets are this proprietary thing you know some people are going to have these satellite datas other people are going to have all kinds of other data so it it also makes sense there that you know there will be

competitive differences in in the type of data you have and and that's not going to go away right so it would be make sense if those people also can kind of use numeri plug in there uh you know maybe they'll pay some kind of fees via numer or you know somehow uh Drive the value of the platform uh but can compete in some way too and maybe even compete then uh in terms of as a as a data scientist maybe I I will be able to choose okay you know this one pays me better than that one so that's where I'm going to um try to do the best model for yeah I definitely see that um I think first

there is one thing you said there which is uh what if you have you have your own data set and you want to bring it to numeri like we don't have any way for you to be able to use your own data set or even your own analysis or you have to use our encryptor data um and that's the only way we we do it but I do think they could and we have been thinking about things around what if we allowed people to submit data sets all you know and um and I have been thinking about that but we don't have anything to comment about that yet but the on the other hand you know letting other funds beond the numeri platform um it would be great for them to give us their data uh but I

don't think you you kind of get a lot of benefits from having the It Be One Fund um and I think what would Mo most likely happen is we have very very low fees because um we can charge way lower fees than any other hedge fund because uh just the way we do things uh we have everyone outside uh the company we only pay the people who are good not the people who are bad um so I think we could have extremely low fees in our hedge fund and actually that will be an incentive to have other hedge funds give us their money or other their LPS switch uh switch from that fund to Oz if they don't um and I think the fee

the fee in how we could have lower fees would actually be a very big uh Innovation too uh and I I'm optimistic that and and you know the hedge fund industry is very efficient once something starts working people really go for it and so it can actually happen very quickly you have a hedge fund with managing $5 million the second year they're managing half a billion the third year five billion so it's kind of amazing how quickly it can it can scale and um so by if we can have higher returns because we have more talent and and uh and lower fees it'll be a really good hedge fund product so uh what's one thing that's

really interesting about numerai is um this is one of the first startups SL projects that is I think successfully merging the blockchains and the AI neural network space right like these two these two technology areas seems seem to be like silos they're like experts in this area and experts in that area very rarely do you get U get a mixture of it uh is is the particular technology You're Building applicable to other kinds of uh data fitting problems like computer vision or something like that the machine learning we using no not the particular machine learning you're using

but like this way of building uh a data model like so essentially you coming up with this function right like your company is this like this building this giant mathematical function that takes all of this data as input and gives out predictions right now I see like a a lot of a lot of problems in like neural networks and artificial intelligence are like that right like it's it's given all of this let's say visual data what's a function that can result in the identification of objects inside that particular particular piece of data so is your general approach of like crowd sourcing data models and combining them

in order to have a meta model and that meta model like predicting things is is it applic is it applicable just to this particular Financial Field of like stock trading or Stock Investing or can it be really scaled across different kinds of um other kinds of problems yeah some people have said what about um healthc care data because um it's interesting you don't want to actually share Healthcare data because it's sensitive so numeri seems like the perfect thing you can have uh you encrypt the healthcare data let people model and uh and then find the best model I actually think um numerai might be one of the only applications for crowd

sourcing machine learning um I don't actually think the healthc care one would work uh or really any other one and the big reason for this is um the in the stock market a very small Edge uh really matters but in other Industries it's actually many other kinds of frictions that matter so for example there are already brilliant um neural nets for Diagnostics so you can have a neuronet diagnose lung cancer or whatever based on an X-ray better than a human um but it's really uh regulation and um and uh and just general you need

the doctors to believe it you need the patients to believe it you need some new laws you need all these different things and actually you don't what what you don't need is just a 1% Improvement like that's actually not very it's not going to move the dial on the friction so I think many other Industries have these frictions and um Finance it doesn't you can actually no one can stop you from Trading if you find find an edge it's kind of amazing to me when I was a kid I you think how how come I can buy stock in a company without asking their permission how can you buy Apple stock without asking Tim Cook's permission and it is this amazing thing you you it's it's public you it's your it's uh it's totally totally up to you uh what what

you buy and um and every other industry doesn't really have that you can't uh it's always quite a lot of churn and uh to to get to get to move move things forward but the stock market's kind of like instant so we're almost at the end of our episode but be curious what what's on the road map what's coming up for numerai in the next you know two years and then where where do you see uh the project in like let's say two or three years from now how what stage is it going to be at we were working a lot on numer rare and it was you know we had to get it audit smart contract audited and things

like that and um and now it's just very interesting to see what people are doing with it and how much it's uh traded and and uh Stak and it's it's it's amazing um but I do want to build out that part I think that's the key thing is like how do we have um a whole ecosystem and that's kind of plugged into this to the numer rare smart contract and uh staking models and and just generally develop it I I love this idea that people would actually speculators would actually team up with with data scientists I had no idea that would happen when I was coming up with it so um and and then definitely user growth is something we've never focused

on um it just happens um but I I do think we we could do we could do more we could also with more data I think we our data set um you know there new there are new data sets coming that we haven't announced yeah I mean if we do if we do have uh good returns and our fund over the next year like that will be really a huge um kind of proof of everything working because we will have had the numer rare out and lots of data uh and we only recently did this neural cryptography thing only recently did the

numer rare So proving these things are working uh is going to be very important um and then the question is how C like can we manage a lot of money and a lot of like you were saying earlier it's actually difficult to manage more money as as it goes up so um figuring out models that haven't ever been invented and that can do huge capacity um is something we have to do and and take uh take billions of dollars and can you share anything about the return so far is it working you know beating the stock market or hedge fund averages um we actually contact talk about the returns or the AUM um because of regulations and so we

you know we're not trying to pitch uh when we do podcasts like this it's to you know get people to learn about the company not not to pitch the hedge fund well Richard thanks so much for coming on that was super fascinating learning about numerai it's an incredibly interesting project I think it's probably uh the first of many that we're going to see at this intersection of blockchain and Ai and it's just uh incredible the surprising strange things that are uh coming out of this so yeah so thanks so much for coming on and of course uh thanks for listener for once again uh tuning in it was uh great being here this week so we'll be back uh with another episode next week

and of course we are part of the list of Bitcoin Network so you can find this show and other shows on list of bitcoin.com and if you want to support the show the best thing you can do is you can leave us an iTunes review which helps new listeners find the show so thanks so much and look forward to being back next week [Music] [Music]