Solving Crypto's Oracle Problem | Mike Cahill

Transcript

we're focused on having the fifth product continuously expand more Publishers cover all Financial Market data and so like if done right then every month the number of symbols will go up the number of users will go up the number of Publishers will go up and um and the rewards is kind of one of the incentives and levers that you know that we have to to make sure that happens all right everyone so on Empire you obviously know that we talk a lot about the institutions coming into to crypto and that is why we are super excited to share that we are hosting the digital assets Summit we've hosted this since 2019 it's coming up in London March 18th to 20th don't miss your chance to get ahead of the curve you can get 20% off

with code Empire 20 we'll see in London this episode is brought to you by toku the first comprehensive Global solution for both token compensation and tax compliance toku makes implementing Global token compensation and incentive Awards simple with toku you get unmatched legal and tax support to both Grant and administer your Global team's tokens across the entire token life cycle make your token grants easy today with token all right everyone welcome back to Empire uh really excited we have Mike haill uh spent time at Morgan Stanley nura sibo jump was running I think special projects at Jump uh then uh

today is CEO of D labs and also contributor at pith so Mike welcome to the show man thanks so much excited to be here yeah uh so you guys obviously we're going to talk you guys just had this big uh token generation event we're going to talk about pith and pith token but I think first there's like two setups to this conversation that will make the pith conversation more fruitful one is the Oracle problem in crypto I would say the V1 of oracles has been dominated by chain link the last couple years you guys are coming into that same space and maybe doing what I might call like the Oracle V2 approach uh then there's this like problem that exists in traditional Capital markets um where all of the data is monopolized with these centralized exchanges like like the NASDAQ of the world so let's start there

can you just give us the setup for what actually data looks like in traditional Capital Market structure yeah absolutely so um 20 years ago Financial Market data was never really monetized basically the exchanges would monetize their executions and then it be became more and more of a component of their um their revenue mix and today it accounts for 20% of the revenue um across the top six exchanges at six and a half billion dollars that's for real-time streaming Financial Market data that doesn't cover things like benchmarks that's a that's another multi-billion Dollar business um so what does that model look like so

basically when you're buying access to trade on say NASDAQ or nisy you will basically ask for a market data feed and an order entry feed right and so one will be um over like a websocket or or an itch um or rest API connection and depending upon what type of trading you expect to do you're going to pay different amounts and you're gonna have different types of access so if you're like a let's say a retail brokerage you don't really need the most sophisticated Market data so like Robin Hood if they're going to Route orders to an exchange they would say all right we're

just going to get your most basic standard package and the exchange is gonna go okay right we're gonna give you a wholesale price for that that's based on how many users Robin Hood has and if it goes up you're gonna pay a little bit more then if you go up one rung and you have like a tiger or Appaloosa or more someone like that it's like a hedge fund there's probably some sort of a human Trader there and he runs their smart order router um and from time to time he'll have to intervene he's going to want to have access to like the full order book look at it Place orders watch them cancel them um amend them that sort of thing and also from a historical perspective they probably run some um

regression testing around performance of say like t- wops and so they'll want to have all that data um recorded and stored and so that would be a little bit more expensive and then the third package is what highfrequency trading firms have and so this is the most robust data it's down to the nanc um and also the access to it is typically gated such that you you pay for collocation so you're really close to their server box in caucus or in Aurora depending upon which exchange that we're talking about um and that's going to be the highest and sort of most optimized um and so over time this product set has matured and it has become you know now this top level big contributor to the revenues um now the thing is the trading firms are

only ever buying this and so there has been some Chagrin within those entities for a really long time because it's almost like they're giving something you know the tra the the exchanges are monetizing the trades that just occurred Traders are the only way that trades occur or the bids and offers where Traders want to trade Traders are the only ones that can send in bids and offers and so they're basically giving these exchanges all of this data and then they have to buy it back and for a long time they were really upset with this um and and many of them still are today um some are more vocal than others um and that's just been the the the state of the world

so all of the world's data has been kind of financial Market data has been gated monetized by big exchanges and then Traders are just sort of stuck on the bid side and can't really participate in this market and pretty unhappy what happened 20 years ago that let them that let them do this it was the evolution of financial markets to become more electronic previously it wasn't as it wasn't as necessary to have this segmentation of Market data and so I think as you and you basically used to have more human Traders so you'd be able to like have at a bank desk for instance you'd have like 50 human Traders and so you could say all right for each one of

those they're going to have access to you know this particular um order book or this um software that allows them to trade you'll pay for pro seip just like the way that like Bloomberg monetizes like and it's they're super expensive um and so because there were less and less Traders and most of the trade was coming through a single API they really needed to change their business model um and there was also consolidation amongst the different um exchanges um for instance CME has bought several other exchanges including EBS which was the the large FX order book for a long time um and so this this kind of lack of fragmentation um and you know these large conglomerates taking up such a a big percentage of the market share um

allowed them to really increase their their their exposure the same thing happened for FX it just happened later so FX was like 10 years lag to the the equity markets um both on execution as well as you know Financial Market data um and so now it's become a big component of the of the FX markets it basically for futures it was pretty similar to to equities um and that's why I think all markets sort of go through this Evolution and like crypto today is in those like 20 years ago phase where most of the exchanges within crypto don't charge anything yet because they're thinking about it as well we actually probably need to grow the pie a bit first and it's a good marketing tool like having the binance price feed on

trading view is like a useful way for binance to get its name out and um the more people that are trading like you know you can extract the costs later but for now it's just like let's let's just think about this from a marketing perspective the one outlier is probably coinbase because they have such an Institutional coverage model within the US and a lot of DNA from Banks they already have a pretty robust business model around Financial Market data although I don't know that their enforcement is anywhere near the enforcement that like NASDAQ or n has so still people use it and um and there's some you know level of kind of you know quasi commercialism commercialization around it do you think they'll Al so is that the Natural Evolution of crypto

exchanges is right now they kind of treat it their data as like this top of funnel pull in more users get the the binance brand awareness out there but you know maybe fast forward three to five years they'll start monetizing their data this becomes a big business line for them like it does for you know NIS or NASDAQ yes I would I would heavily predict that to be the case yeah okay so that's the okay so one bucket here with with pith we have uh this problem in traditional Capital markets where all the data is monopolized talk to me about the uh the crypto Oracle problem and like how some of that financial information uh streams into the apps right like an A or compound today pre let's let's go prepi yeah Oracle 1.0 model was so or the model was um take the data from the internet and

put it under the blockchain um and do it in a way where you preserve trust the the the 1.0 model looks a little bit like a bridge um so it doesn't have any particularly proprietary or in network access to data it's just basically saying like look there'll be some endpoints on the internet and we can use those endpoints if we have enough nodes and those nodes will run a very simple aggregation where you get like say 10 nodes who are who are replicating the price of coin Gecko onchain and they'll do it every so often and and it'll be good enough every so often meaning like they'll push an update every hour or 50 basis points um now that works for for crypto data because of the phenomenon I

mentioned it's it's it's immature in its life cycle if crypto data was already protected then you know that business model wouldn't work you would have to go buy that data and then buying data means that you're subject to the terms and conditions of it so you don't get uncontrolled distribution rights to it so there's this like temporary time where you know you get this bootstrapping to create a solution to the Oracle problem the Oracle problem is how do you bring ex exogenous data on chain in a trustworthy way and and it involves nodes that are going grabbing it from the internet replicating it publishing it but they add all this latency takes a long time for them to do this and then because it's being

published as a push on chain you need to be kind of gas conservative and so you basically don't want to you want to publish it as infrequently as possible and so you basically end up with this very low resolution um feed uh that doesn't track super super well to kind of the the updated or high-speed markets what what does low resolution mean in this case so if you've predefined the the epoch with which you're going to publish data on chain to one hour or 50 basis points 50 basis points is the is basically the resolution you're basically saying we're going to have a heartbeat that comes every every 1 hour irrespective of the delta in that time

period or it has to move 50 basis points so um if you think about like just pixel resolution um as being like this 50 basis points requiring a print anything below that is going to you know not be printed and so that's you know that's where I think about low resolution financial markets you know go out to several decimal points um and you know with things like Leverage those can be incredibly meaningful um and so there's been like the general evolution of markets has to become electronic is that they update more frequently and that there's generally more spec specificity or granularity with the prices so Oracle model 1.0 chain link all this info is available on the

internet right so then you have nodes scraping the models online then they publish it to the blockchain why do we need to necessarily like improve the latency of Oracle data today like compound doesn't need low latency right they don't as is um so for the very first kind of um proof of concept lending protocols you don't really need to have very fast markets because you're not going to have really high loan to value um levels anyway so what you really need for what you really need highspeed oracles for is applications that are going to take advantage of a highspeed Oracle so some of those would be like onch chain Perpetual markets now

if you're running an onchain perpetuals Market using a slow Oracle then it can be arbitraged very easily like a like a synthetics for example yeah so synthetics is a great example so synthetics was on V1 using chain link they created a V2 um they launched the V2 at the beginning of of this year um and evaluated at that time the best oracles out there that would suit their needs um and the governance chose pth as the Oracle to use for as their primary Oracle for execution because it's the fastest and has the lowest latency and that was that enabled them to really sort of Lea frog um you know some of the competitors that had caught up during

the um the time where they they hadn't had that and became a very big force with regards to perpetuals on chain it's funny I feel like this low latency of data in in uh in like OG defi has actually been a it's sounds like a negative but sometimes it kind of saves defi uh right Oracle being slow to report volatility ends up sometimes on days like you know March of 20120 I'm sure Santi knows 10 times more about this than I do because he's actually experienced it in depth but sometimes it can actually help defi in a way well I would probably make it a more precise statement which is it's more so the way you calculate the price feed like it's some of the where mistakes have been made in defi where you always I think

want to have a constant and as like constant real- time data feed but the way you calculate the price is sort of like in a smooth out manner which is not taking a specific data point but taking the average of the last five minutes 10 minutes 30 minutes an hour and every a lot of the mistakes that have been done uh and exploits have been because a lot of def5 protocols the way they they the price feeds that they're using are just relying on a single data point and not a combination to triangulate to the best price and that's where like compound was affected by this like even the the what I'm trying to say is some of the best D5 protocols have made this mistake and have learned from it so yeah um and and I think that the fact that there there are way better ways to

architect a lending protocol so that it's not subject to exploits rather than just using a slow Oracle I know what you're talking about need to evolve then we need then we need better faster data yeah of course yeah exactly like it's like the the point is if if the best thing we can come up with from a for our lending protocol in two years is compounds current instantaneous instance um you know that'd be pretty depressing so I'm I'm pretty optimistic that there's going to be a use case for higher Fidelity Market data within an or within a a lending protocol it's just they'll probably have different mechanisms that'll defend against Flash attacks or or volatility spikes like you

you should be conservative with your ltvs or you should should be conservative during periods of high volatility so one of the things that pith also innovated was the aggregation is a tricky part so as STI mentioned right if you if you're dealing with a single point um especially during High volatility that's going to be a problem and when we're building the you know the pith Network we kind of had this first principles discussion like what is the price of Bitcoin right now you know if we have 10 data sources is it the average it's it's probably not the best representation um the more data sources you get sure it's a better representation but it's still imperfect um for instance how do you represent the

fact that there is going to be anomalies based on the idiosyncrasies of um like funding rates or the access to a restricted currency market like kimchi premiums so you actually do want to represent that in real time so what we came up with was called confidence intervals and pth is the only Oracle that has these so in real time every price from pith is aggregated together um but along with the aggregate each price has a confidence band it says plus or minus price so if the price of Bitcoin is 37,000 it could see plus or minus $200 and this fluctuates at every single update and in periods of high volatility what you could do if you're

building you know a smart kind of collateral evalu ation you could yeah there you go um you could say all right the if the confidence interval is quite wide right now we're going to limit the amount of collateral you can contribute for a period of time um and those are smart ways to address this and then it becomes more dynamic because with a lending protocol you don't necessarily want to be the most conservative to all like the 99th percentile incidents like that would just give you a pretty low leverage product you want to assume that you know during 99% of the time you can give High leverage because that's what would make

your protocol more attractive but you have defense mechanisms in place such that you don't get blown up by black Black Swan events um and you know this is one way to do it and another way to do it is is using things like t- wops and you know all those stuff are available on on p as well um what's the confidence on your confidence intervals that is a good idea high confidence I'm sure you've had I mean I'm speculating here because I don't know but have you actually changed the way you think about and calculate confidence intervals we've changed the aggregation technique since the very beginning um so

now the way that um aggregation is constructed it filters out um outliers pretty well so each one of the Publishers basically gets three votes on what the price is and so they get their their price and then their price plus the confidence plus and then minus the confidence so every publisher within the pth network has to publish a price and a confidence um and then all of those votes are plotted and you end up with something called a weighted median and then the algorithm creates a llao distribution based on all of the inputs to come up with the aggregate

confidence interval and it tracks fairly well um it also has characteristics where if you have these outliers and you have a cluster of outliers then the confidence will represent that cluster um but it won't necessarily skew the average price or the aggregate price to the the the cluster of bad prices because you basically want the output to in plain English say look the real price seems to be or where all the trading happened seems to be 36,000 but there's some stuff that's trading plus or minus a thousand dollar which indicates High volatility and that's really good information you know that's that's

better information than saying you know it's skewed up a thousand and then having some average that shows you something that's a kind of a bad price in in both instances yeah so you you're normalizing but you're also paying attention to these anomalies that's right yeah yeah because that could indicate there's an exploit there's a flashline in a particular D5 protocol there's some sort of funky thing going on with your data providers that needs to be addressed exactly Mike can you actually can we just zoom back out for a second here can you go back like take us back to jump my understanding is uh pith came out of jump you were like we were so happy going really deep Jason I mean as much as i' like confidence intervals I'm like I got to understand pith before we go

into these I wonder how many people dropped off over this like last confidence if you're still here and listening Kudos you get a some you know Pat in the back but anyway let's go back to the origin story I guess yeah you guys you guys flip the model right now the now instead of the exchanges owning the data the the funds are actually saying hey we've got this data let's let's band together but take us back to the days of jump and how this kind of project came to be yeah so um in 20120 I was working at jump and we were involved in a lot of basically trading and partnering with a lot of projects there was a lot of investing as well and there was a big thesis conversation looking at some of the bottlenecks around the markets and

when we did deep Dives we we we basically realized that there was two areas that were going to really constraint the short-term growth of defi the first one was the solution to the Oracle problem which is Oracles 1.0 um we thought that with these with this low resolution data it was never going to be able to do anything interesting from an application perspective you could basically do lending protocols and that's about it and that's really all it did like with with um defy 1.0 um the other one that we thought was a big limitation was was around um bridging um but at the time we had basically we talked to other trading

firms and we we realized that the solution for the Oracle problem was to get you know data directly from the source and we combined these two problems together that we mentioned in the beginning right so the Oracle problem Oracle 1.0 wasn't that great and number two like most of the data is owned by the exchanges and we started talk to exchanges and said you know if if you guys were to publish your data on Shain you know that would be create a potentially very interesting product for solving this other Oracle problem and it would potenti give you a new Revenue source and um you know combat that Monopoly in the web 3 space um and very quickly the idea resonated so the first the first companies that joined into the

network that um were kind of founding members were virtue GTS uh myx um elmax and a few others and that really where the idea was born so jump was a a contributor started doing a lot of the development um but really it was a you know it was a collective um and then the collective grew quickly so today there's over a hundred Publishers all institutions and the development is mostly done from Dural Labs the company that I'm the CEO of now uh I remember talking to Colleen Sullivan from brevan about this project early on um when I think she was still at CMT and she said look the craziest thing about this is you have a bunch of

firms who are hyper uh confident like uh sensitive with their own data and they never and they they're hyper competitive as well and now they're all collaborating and working together how did you get like jump and virtue for example to work together here yeah you would never expect you think it's like cats and dogs and we actually thought this when when we were when we'd gone out with the idea um it it just so happens that there's two two things number one there's like a bigger cat which is the financial Market data is something that everyone deals with and is is incredibly frustrated by and number two the Innovations of blockchains allow distrusting people to have incentive

mechanisms to collaborate on something that gets shared as a common good for other people that wouldn't be possible there is no way that virtue jump Jane Street drw would enter into any sort of an LLC for any kind of profit sharing or information sharing in a traditional sense if there was a single shared it was a single database owned by by just one of those companies it would never happen it's all the Innovations around blockchains that allowed this it's not owned by anybody no one has preferential access to it the aggregation contract lives on the blockchain and everyone publishes to it um so it's Equitable access for all of the contributors as

well as all the users that was a that was an important innovation and then there was a model that was described at the time on how the Publishers were going to be incentivized for for providing their data and that was what really resonated with people but like okay let's let's unpack that for a bit because if you're jump you're at the peak of the bull market you're you think you have an edge I don't know but I'm speculating you think you have an edge you think you have better price feed and that gives you huge advantage over all the other cats and dogs out there and so no matter how like I I got to think like the economic incentive is kind of like not there there to provide the data and get paid to contribute to this like model of like a data feed

because if you fundamentally believe that you have an edge then sort of like an adverse selection problem all The Underdogs do have an incentive to provide the data and is then what you're saying that aggregation of that data was enough to convince the guy that thinks it has the most amount of Edge on the table to finally come forth and provide the data but it takes time or does that make sense because like if I if I feel that I have the edge no [ __ ] way I'm going to contribute I'm like no this is my data everyone else can like not have an edge yeah I think there's there's a small Nuance here as well which is the the time scale resolution that we need to focus on so the fastest time scale possible within blockchains today is

quite slow for the traditional Finance markets and so the traditional Finance markets you're dealing with sub milliseconds like u nanocs in some of the markets um in terms of the the the Alpha and the resolution that you you you need to to be be competitive the fastest times within blockchains right now pith updates every 400 milliseconds so the the scale of magnitude there is in the you know thousands times um category so from their perspective some of the data has already decayed or some of the alpha has already decayed for traditional markets um but now we're still at the fastest update times possible on

blockchains because of the binning of around 400 milliseconds so for now you know we have this Paradigm if blockchains were to become much faster and to have like feature parity then I think the the model would need to change a bit there would need to be some protection of the data of the visibility of the data for a period of time before it gets released um but that's not yet a problem okay um and so I just want to understand and go back to something that we've talked about earlier which is of course you see the opportunity the chain Link's not particularly addressing but like can you just crystallize like what is the real Advantage maybe from a if

you're like synthetics or some of these other protocols choosing their data provider their Oracle um are I've always felt that the Oracle space is not win or take all you want to have redundancy and so like I was involved like in the early days of synthetics they running their own oracles and making the transition to chain link seeing that happen gave me a appreciation of like maybe you want to have not just one data provider like one Oracle uh provider like chain link but many others um because the risk of messing it up is really high so do you have a sense of like the projects you're working with um how many of them are have chain link and are working with chain link are you replacing chain link are you complimenting chain link like

how do you think about that yeah in the case of synthetics because I think that it's a good example they they still have chain link as a backup um but the way that it works is it's a toggle um to determine whether or not pith should be used um or basically trading should be used right so so pith is the primary um Oracle if you trade on synthetics then you're you're leveraging a pith price um now if pith and chain link disagree over a time period like and it doesn't need to be super high resolution it could be like you know the the moving averages are have gone off out of out of whack um then they say you know everything off something's wrong here I

think that's a pretty good way to hold on to the side of the pool um the the bad approach would be to just like average two two oracles together um that that would just both may be wrong yeah yeah well then you just end up with a bad product because you you remember when search engines were coming out um they at first it was like there was a bunch of search engines and then there was like the aggregator like dog pile would be like the search engine of other search engines um and then people realize actually the thing that you want is just the best search results um you don't you don't want to have like a bunch of you know bad search engines included and so I think that's probably the the right way to think about like an oracle or primary Oracle no I think a redundancy makes sense um

and so pith today covers about 25% of the applications that use an oracle um the other 75% use different Oracles um chain links obviously the largest there P has only gone cross chain 11 months ago um from it first chain of salana and now it's on 40 different blockchain so most of that growth has happened in in the last 11 months and basically it's been three new applications per week have integrated pith um so I I would say that there's there's probably some redundancy um but if you do if you do redundancy really well within a single protocol I don't think that there's need for a separate

one I think that like pith could could be a redundant solution there there are sort of backups that um that are available as well and like pith has a if you click on a pith price you'll see like the the EMA um and then there's some stuff that are being rolled out regards to like putting together any kind of arbitrary teawap and so you can use those as good good kind of checks as well um although I I do think that competition within the Oracle space is smart um and so you know if chain link does introduce new products and those are things that require the the the pth protocol to react like that's probably good for the entire ecosystem yeah so chain link uh we're looking at charts here for people seeing video but like chain link uh this is defi Lama data by

the way so chain link secures or 366 protocols you guys do little under half of that 125 in terms of total value served I guess uh um or secured chain link is like roughly 15 billion what is this Mike wink link oh that's the Justin son um for wait Jason you can't you can't switch it up you're Miss uh I was I was I was toggling between so if you I mean this is interesting like chain Link's 366 pith is 125 protocol secured but chain link has 14 times more yeah total think that the real metric I guess is value secured maybe pith is a fraction you know chain link is close to 15 billion

you guys are 1.6 billion so walk us through like what do you need to believe for you guys to if if to your earlier Point redundancy may not be necessary and you could just have one really good Oracle solution if if you guys believe that then what would need to happen for you guys to flip chain link yeah so there they the the majority of that total value secured is Consolidated around aan compound so like it's it's mostly most to those two apps and that's just sort of the nature of of how value locked on chain is right now I don't think that's the best metric for an oracle um I think it's a good one it shows you which lending Protocols are

using um are using an oracle it definitely shows you something that's at at risk so you know securing is is an important component of it um the other component I think that's important to to add um and you know hopefully defi llama adds this at some point but it's the total value Tred that's secured and so every time someone's trading something that is using an oracle like that volume kind of should be counted up um and in this model in particular it makes more sense for for basically two reasons number one it's it's the primary users like the synthetics of the world um you know they're the ones that are going to be most focused on low latency the other element of it is that's actually the way

that the protocol collects fees from the P protocol collects fees so each time there's a trade there's a fee that gets paid through the network today the fees are set to to next to zero but not zero um and governance can choose to to update that now governance is live um sorry Mike can you say that one more so like when you do a trade on synthetics what what happens is I I'll give you the whole workflow just so that you're just so I can articulate that there's no latency as well um but basically you you do a trade you see a price on the and that price is actually reading off of PNET and you click to execute and a message is sent

from optimism to pathnet saying hey we'd like to bring the price from this time St time stamp um that Jason just tried to exe or just executed and so it goes back you know whatever like 400 milliseconds and sends that that price over and then you execute the trade and you pay a fee and that fee has three components to it there's the gas on optimism there's the fee that synthetics charges you if you're a maker or a taker and then there's a small fee that pith charges as well and so every time a pith price is delivered to one of these 40 different blockchains a fee is is paid to pith and right now the fees are all set to effectively nothing um but

they're set such that people are paying in in P's accumulating fees there are I think over two million fee paying transactions per day um and so as governance chooses to um figure out what fees are kind of appropriate we'll see how that number of transactions changes like it may go down a bit it may go down a lot may not go down that much um depending upon you know how those those dials are changed yeah yeah so it really is around utilization no different than like you know tvl is kind of like a very crude metric it's really you know if you look at un SW it's the best decks because it has a high solarization on the pools because everyone's trading there um and so you care less about you

know you care less about how much value is being secured care more about the type of protocols I guess that work that are working with you if you're like a high frequency kind of type of D5 application then you're going to earn way more fees relative to a very kind of uh low like activity protocol exactly yeah given this business model that I've just described right the a lending protocol is not a very Prof Prof client now there are other business models that will make more sense for Lending protocols um that I can Envision and could potentially be rolled out in the future and would actually work for helping them kind of secure and perhaps monetize their um their kind of protocol a little bit better like what's the when

you think about like Revenue projections how do you how do you think about that like is this say what's the market opportunity like what's your guys's and I'm sure you can't give too specific of numbers here like when you think about Revenue like what is the scale how many zeros are behind the one uh you know fast forward a couple years here yeah I mean it's it's it's so hard to predict um because there are so many assumptions particularly around like you know what is what is the market opportunity for defi um so like onchain derivatives today are like sub one% of binance futures um I think that in a year they'll be over

10% um you know and so that gives you a sense of like a first kind of really scaling and growing um now what what are people going to be paying on those and like what percentage will go to market data for that one it's it's difficult because you'll start to basically find an equilibrium by like starting pretty pretty small but if you zoom out and you just say well you know you ask me in the beginning is this how the crypto markets develop and you know that's my prediction is yes so at maturity you could say the revenue model for defi will perhaps look in such a way where 20% of the revenue goes to the

market data and if the market data is all done via oracles then you know perhaps it's 20% of the revenues of defi um so I think that you know is is is decent enough as like a bad jistic um because all the other eristics are going to be highly speculative based on assumptions that have very low conviction yeah what about uh I want to get into the token in a second but um what about non-financial data yeah I I I like non-financial data conceptually um I the the thing that pith has done really well has been take highly valuable data and incentivize users to make it available to everyone with a business model that will work

work um things that are just like the public in the public domain I think are a little bit less interesting um there's perhaps some requirements to just get them on um and if there is like a a big use case for I don't know weather data or things like that then you know I can see that you know pith would eventually expand to it but for the moment there's they're just not highly valuable data sources and and there's also no clear Avenue of application sector that is really building something unique there you know defi is really one of the biggest potential use cases for blockchains and so Financial Market data is a Lynch pin therefore if you can be

specialized and you can look like Bloomberg you know you can kind of own that segment as opposed to Google which you know Google Google Finance is not used on Wall Street anywhere um so I think that there's there's a lot to it but that said there's no reason why pth data couldn't be expanded or P data sources couldn't be expanded to other types there are other types of data though that I think could be could be pretty interesting to to add in in the future stuff like um credit scores um if done in a way that preserves a bit of privacy would be pretty interesting um and then there's a couple of other things that we we've kind of just ex like thrown against the wall that I think would be would be pretty cool to add it at some

point all right everyone so we talk a lot about the institutions coming into to crypto on Empire Santi and I are both headed out to London March 18th to 20th for blockworks is eighth ever digital asset Summit Das this is an Institutional buttoned up conference that we've hosted since 2019 I like to joke that it is probably the last remaining kind of suit and tie event in crypto people are still wearing suit and tie it's pretty funny but you'll actually hear from a lot of the largest institutions in the world coming from Standard Charter FIS JP Morgan framework folks coming out winner mute vanc golden sacks there are a couple big themes of this conference one Bitcoin Catalyst the having and the spot ETF two A View From the buy side three rw's tokenization and

stable coins four Global regulatory Frameworks five institutional infrastructure including Banking and payments and six the macro case for crypto if you have anything to do with the institutional side of crypto you have to be there Santi and I got you back we hooked you up with a 20% off code it is Empire 20 there is a little competition running internally at blockworks to see who can drive the most number of tickets so help Santi and I out register with our code and you get 20% off that is Empire 20 this episode is brought to you by toku the first comprehensive Global solution for token compensation and tax compliance if you say yes to any of the following four questions toku is a no-brainer solution

for you number one are you planning to launch a token number two is your token already live number three are you currently granting your employees or contractors vesting token Awards and number four are you trying to figure out how to take care of taxable token events for your team if yes you have to get in touch with the toku team toku to high level makes implementing Global token compensation and incentive Awards simple you get unmatched legal and tax support to both Grant and administer your Global team's tokens toku navigates this across the entire token life cycle from easy to use token Grant award templates through tracking vesting to managing tax withholdings toku makes it simple for

leading companies in the space including protocol Labs dydx Foundation Min Foundation hadera nosis safe getcoin and many more reach out to toku that is tok.com Empire T ku.com Empire click the link in the description or DM me on Twitter and I'll get you connected to the team what's the role of the token the token is a governance token and so exactly mean yeah so the um the the pith network is a decentralized Oracle um and it's designed to be run by the community um and so the token is the the way to

govern the network um and the token has now been distributed to over 95,000 um wallets and users um historically it was just a you know middleware B2B infrastructure Network um and it was a successful one you know as we've sort of talked about we've you know amassed over 100 Publishers um there are 225 applications that are using or 100 I guess on the defi llama numbers it's a it's it's a little bit lower but 125 applications that are using pith um and that has been done in a series of checks to you know confirm that this is the

right idea right there was this hypothesis that defi needs low latency high high frequency oracles um in order to be successful and so the first way you're going to do that is well if you can get enough users within the network or Publishers within the network to create that data so that you don't need to go and ask NASDAQ or I so that was done the second one was well if you've created this data do people really care about it or do they want just slow data so that was step number two was when pith launched on salana and stood out there as the the primary Oracle over 90% of the applications use use Fifth and number three was can you build up a cross-chain Oracle model um that requires people to pay some fee have

prices be delivered across chain and that happened earlier this year so we basically passed these three proof points and now it's a time to you know we've gotten out of this experimentation mode and we can move on to letting the community decide how to govern things going forward so all the updates will get turned over multisig will be turned over to governance and then governance will control things like which symbols to be added um which new Publishers to be added how should rewards work do we add new segments of data such as weather um credit scores things like that um does the aggregation logic need to change should we have a

new um fee model for Lending protocols and things like that so the the governance is gonna be managed by users and what has been really cool about this airdrop is that this retroactive airdrop is that the the end users really just got told about pith perhaps For the First Time by seeing that they qualified for this airdrop because pith is this you know this middleware B2B company it's almost like armm or you know Nvidia everyone knew about Nvidia because they had a stock and you know people could speculate on it and armm people didn't necessarily know about until they had a stock um now you can assume that

there's people that want to trade different symbols on say a synthetics but they don't have the ability to ask for the Oracle to you know make those symbols available they can they can participate in the governance and so this is the the way to invite them to the community and say yeah like you have a bag then okay I want pit to service the protocol that I'm trading with but what does the average DJ know about oracles and the way like like confidence intervals like they have no business in understanding that nor do you want them to be governing that candidly like this should just be a sass business as much as I'm a token maximalist like I just don't see it like like what needs to be governed that like the wisdom of the

crowds which is not so wise or has been and DOW participation is terrible why even introduce that and why not just keep executing as a team run an incredibly profitable business model like Bloomberg has and call it a DAT yeah I think that the ethos of it is to have it be decentralized now I agree that if you if you do the the decentralization or the governance wrong or poorly then you aify a you know a protocol and and you and you don't get it to develop any further it's like if United Airlines all of a sudden gave everyone points in the ability to to determine if they should be using Rolls-Royce as some other

provider for jet engines and and and the community somehow like decides to use like a shitty jet andion manufacturer well guess what folks I'm not gonna [ __ ] fly United after that I'm exaggerating here but like you know my point like I'm just uh yeah exactly but I mean it goes to the like what the utility of governance token um and you know if theoretically if you're buying a a share of of ofm you you get a a say on the proxy and um you get a vote for somebody I think good governance will evolve into things like this like synthetics does a good job they've got these these councils and so you you you end up creating an agile environment where people actually get to make

decisions over the period of their tenure and then they go and run for counsel again um and and they they ship like they have new versions um and there's a lot of you know it's a dynamic it's a dynamic protocol versus something like say a or compound where you know the the governance model is completely different and it doesn't move nearly as fast are you subsidizing part of the fees or incentivizing projects to work with you guys and has that become easier because you have a token and you can like subsidize or incentivize them to work with you in a kind of referral manner um that's not something we're actively doing but part of the the retroactive airdrop um did include protocols that were um that were using pth and so

during the snapshot window um there was an allocation that went to the end users and then there was allocation that went to the protocols Dows or treasuries um themselves yeah so that's a very big number I think you said 90,000 wallets or so yeah 90,000 wallets and it was it involved 20 seven blockchains in terms of the the usage so largest cross chain airdrop of all time yeah so so basically if you used an application on 27 different blockchains um you could have been eligible for pith um for the pith airdrop um and what I think people re like what I started to see over the last week since

the tokens come out um number one so many more people heard of heard heard of pith it felt a little bit like the the overnight success that was three years in the making um no one really appreciated the fact that there were so many protocols that were integrating or that there were so many Publishers in the network the business model wasn't quite clear to people the fact that there was these two dilemmas wasn't necessarily clear to people um and so so that was that was amazing the other element of it is that because there are so many participants within the network a lot of this stuff can be made to look quite easily pretty quickly so the token distribution mechanism was okay well

let's just look at who was using these protocols there's a lot of users there wasn't a okay we're gonna have this thing where you know people are in our test net and playing around and we're going to try and figure out which ones are sibles and which ones are you know airdrop farming you know it it was quite simple to be able to make it relatively um like Cil resistant because we had a long history of of users um and then the other thing was the token was then listed on a number of exchanges well majority of those exchanges are Publishers within the fifth Network there are 10 of them that are actually Publishers from okex bybit um you know there there's there's many that are already part of this protocol and then there's Market making firms who

are trading their token and you know all all those are in the network so when you amass these these participants a whole lot is possible very quickly um if you've done the leg work and so I think that's what's been kind of most eyee openening to to me is that you know all the work that was put in especially during the bare Market has really now paid off because there's so much possibility and and and it won't necessarily slow down there are you know there's a good argument to be made where new segments of Publishers may join the network from asset managers to Banks and it won't just happen like the one off it's similar to the trading firms when they get the model it happens really really quickly yeah yeah um so I think a large part of what

you're describing in the Inception story was like this like credibly neutral Switzerland model to provide dump your data and get some sort of compensation if you're within you know if you're providing accurate data that was working and it doesn't sound like I mean you were successful in convincing a lot of people right that wouldn't work otherwise and you solved that coordination and incentive problem and you created all this value um so oftentimes you need a token to do that but in this case you were successful pre- token because you had all the major sounds like you had a lot of the best data providers out there for from their perspective does a token matter is it a distraction from them do they get compensated what is now a more

volatile unit of account or do they like the idea of having this token and maybe participating in governance in some way shape or form yeah I mean I'd like to think that they're they'd be interested in participating governance um many of them I think are but the you know the reality is that there's 22% of the supply that's reserved for rewards uh for Publishers um that is has kind of been set already in the in the white paper um and that's really been the carrot you know that's inflation and so if you know fees don't pick up for a while then you know there's there's these rewards that will be able to be used to incentiv them and a lot of a lot

of the Publishers are motivated by this now if you just paid them in cash then what what they would try and do to optimize for maximizing their profit would be to go and connect to as many marketplaces or oracles as possible and so having them incentivized you know with something that's like the you know the pi Network token or governance token they're a little bit more aligned with the long-term success of the project um and so you know there's backfire if the pro if the token value collapsed 90% doeses that backfire um I mean I mean I guess it also like economically like if if the value of all the thing that you're securing and the swaps just goes down

90% as well because everything's so correlated in crypto then they're not making as much money and so they may detract they may leave they may be still incentivized to work with as many providers to make up for the loss kind of like mon yeah sure I think I think there's like the getting the rewards right is is a good idea is basically um my response to that is if you have inflation structured poorly then you're going to have a problem with the you know the emissions of the supply coming out too fast um if it's done too slowly then you've got other problems where you know people aren't necessarily getting compensated for even their kind of fixed

costs um and they're be kind of incentivized to walk away um so I don't have like a great answer for how how it will work I you know this will be something that kind of gets proposed governance um will'll probably have a couple of different models that Doral Labs will propose um but I you know I think there's a way for this to work and be Equitable and have a long-term economic model that survives for you know many years yeah that's one of the things that uh you know p is all the rage these days but it's been around for such a long time and I think one of the biggest criticisms that I have and I've heard as well shared by others is that the rewards are very simplistically modeled

and should be more Dynamic um based on the contribution it becomes hard it's like Milestone kind of fundraising it's just like how do you how do you like Express that in a smart contract or like in a vesting like you know in a reward um mechanism but I'm curious like in a perfect world like maybe share with us kind of how you think about this um reward mechanism and perhaps changes that could be done or that you may be proposing yeah I mean I I'm I'm I don't have like um really wellth thought out answers on exactly how it would work but I'd say it's working fine now it is working fine now yes um it's it's kind of early days for for the

rewards um but but basically I think the rewards should be used to incentivize markets that don't yet have a tremendous amount of demand but eventually will um and if you can do that sparingly so that you're not overpaying for things to get on um you know on chain and you can do it so that ex is constantly expanding the network then the product set becomes better and that's really what you know what we're focused on we're focused on having the pth product continuously expand more Publishers cover all Financial Market data and so like if done right then every month the number of symbols will

go up the number of users will go up the number of Publishers will go up and um and the rewards is kind of one of the incentives and levers that you know that we have to to make sure that happens why did you guys Mike you you guys could have done the token drop a different way so when I hear about pith and this has been a really interesting conversation it's a very B2B business right it's like it's B to B Toc in a in a sense but really it's a B2B business I think um why focus on the on the C part of this instead of just going direct like instead of really just focusing on the applications that that have built on pith yeah I think it's within the ethos of of crypto um you know you have to take a view that there are going to be

people that want to participate in open source projects um that's my view and I think that you know it's the reason why jump virtue and Jane Street are all doing this is because it's open sourced um that there's an incentive mechanism that they can contribute they can step in and out of participation um and we wanted to open up the door to as many participants as possible um there are smart people who will begin to participate within pith governance that aren't necessarily Publishers today um and you know I think that it's important to bring in the best minds and invite the best Minds to be able to do this so um you know in short it's a cast The Net

as wide as possible see who shows up and can you know kind of join and and and really wants to contribute because we've been blown away with some of the people that have just learned about the pith Network and now are looking for ways that they can contribute um so our awareness has gone up dramatically um and you know we're seeing people that are basically becoming tribal around pith like they're they're they're they're very bought into kind of the vi the vision and the mission um we're getting lots of applications for um for any positions that we have open on Dural Labs um I think that's really the idea around it it's you know make this so that it can be

distributed and it doesn't have to always be you know kind of building up a huge treasury War chest that can become like a huge company and then becomes extractive you know the ethos is like build the best Oracle and if that's the end goal of the mission and you don't need to build anything further then great then pith lives as an oracle product and you can kind of leave it as opposed to like you know the the exchanges which is well what are you gonna do this year how are you gonna increase your revenues by 5% so that wall Street's happy you know that's not really the ethos of of of crypto yeah Mike man appreciate the conversation I have a feeling this will be the the first of many that we have about pith in the future and uh yeah excited to see what you guys do so thanks for coming on

thanks so much for having you guys yeah thanks Mike congrats hey everyone thank you so much for watching today's episode really hope you enjoyed it we wanted to take a second to just remind you about our upcoming digital assets Summit in London March 18th to 20th Santi and I got your back seats are limited and we hooked you up with a 20% off discount code it is Empire 20 if you heard it earlier in the podcast there's a little competition running at blockworks to see who can drive the most number of tickets so when you register for the to lasted Summit make sure you use our code Empire 20 see you in [Music] London