Shaping Digital Economies with HiFi Data

Transcript

[Music] all right with this we are ready for our next talk and for this talk I'd like to bring on Mike from Dual lapse to tell you about all the things that we can do to make defi a lot better so please give Mike a big round of applause let's talk about making D5 better and we'll start off with slide number one which is how I make friends at a crypto conference so D5 sucks obviously we're all here building in defi so um we need to put some context around that so defi

sucks when we compare it to the boss boss being trafi we're obviously trying to build new Financial rails now there's obviously problems with trafi or CI um think about trying to move a country onboarding is pretty terrible sending money always has fees getting a decent interest rate is all all very difficult um but when it comes to a lot of stuff that doesn't involve the once every couple months thing that really bothers people it's pretty good Robin Hood Charles Schwab nzq these are indicative of what I'm talking about you basically have super super low commission zero zero dollar commissions on Robin Hood um

and and on Schwab roundtrip times are incredibly fast 500 milliseconds or less for retail trading about 300 milliseconds or less for small institutional trading and if you're trading as a large institution in a data center for one of the top tier exchanges we're looking at sub one millisecond and Trust obviously is trusting the brands so if they screw up well you can sue them so there's no programmatic trust but you know they're not rugging you all the time they are obviously exploiting trust in certain ways such as um putting in some of these fees um now your average defi app is pretty bad relative to those three categories um for the most part the fees are going to be nowhere near a

sustainable zero commissions um that you would get on on some of these trading apps your time for round trips is usually around two seconds or more um on on most blockchains and the trust for the most part for applications is relatively centralized um and doesn't have the brand around it so there's no really someone you can go out we all heard about rugs now pith makes things a little bit better we're going to get into some of these topics today about these three categories um and how it helps improve them but um we we start to make progress towards um towards straty so what's competing defi from competing with trafi the sad reality of new technology is that there's the thing that the new technology is famous for

which is going to be kind of security and and and uh this new form of trust um and then there's everything else which is the incumbent and the reality is that the people that really care about that thing move on very early and they all are very frustrated with how long it takes everyone to come on the problem is everyone only comes on when the other elements are at parody with the incumbent we'll go through an example to show how this actually works so if you go back to the '90s you had Amazon as as as your first kind of online shopping um it was competing against a Sears catalog at the time right Sears catalog was like the Amazon of its day and had been around for like over a hundred years you turn a page and you can load 10 images

in about a second right you turn a page there's 10 pictures there on Amazon using a 50k 56k dialup modem it would take you about two and two and a half minutes to download 10 10 images um so obviously it's not nearly as good as the um the competitor here we are I go through this um not nearly as good as the competitor and so all of the things that you do when you start talking to somebody who doesn't necessarily understand the benefits of being connected to the store getting the inventory updated in real time not having to go through another means of of execution using a phone or doing a mail order to get access to something they don't understand any of this because all they see is that Sears catalog is going to be so much faster

now you cut to 10 years later and you get 100x Improvement on the speed you can start to download pictures at about one second 10 pictures and this is when things start getting interesting and people don't generally know about what's going on in the back end but it's because infrastructure keeps improving so we went from 56k dial modems 10 years later we started to get Broadband um those lines were just getting laid down so not everyone had them all at once um and then we got fiber and we started moving into mobile technologies and 5G now you can go on your phone and if you go to Neta Porter you can look at real streaming video or real live video um without any lad times and it feels very seamless and the things that people who

don't necessarily know about the technology say is just it feels better right it's that that innate ability to notice that there's no friction or it doesn't feel like it's a it's a janky tradeoff now on blockchains I mentioned the big innovation that I'm sure that we're all excited about is trust right so pre-web trust was personto person very in scale able um you can only interact with people that you knew or you can you can interface with um with the web you can now interact or communicate with anyone around the world um but it was difficult to have trust everyone's heard the story about buying an Xbox on eBay and getting a box full of rocks um so you have all kinds of

problems with trust on on the web web 3.0 um crypto gives you the you know the the can't be evil rather than than don't be evil um and we have that programmatic trust um so if I use that same parallel to compare defi to his incumbent let's say Charles Schwab is 300 milliseconds round trip time um and zero doll commission payment for order flow let's see how defi has come in terms of users and adoption um through the lens of infrastructure so 2020 is defi summer and you had 12C blocks about 12 TPS um push oracles at that time updated every 24 hours or 50 basis points and so You' basically have to take the worst of when the Oracle price had updated or the

block time um you can't do a whole lot of computation with the limited space and so we ended up getting D5 1.0 apps which are delightfully retro and cool had great trade-offs um to make them work but they weren't what you'd expect to beat Robin Hood like the amm from generation one um or lending protocols you know those were really experimentation for people that really wanted to get dirty on um on on decentralization then a year later salana came out and really changed the um the operating system from 12 seconds all the way down to 300 milliseconds um this gave us a whole lot more expressivity with writing smart contracts uh applications we saw for the very first time um derivative uh

trading platforms on chain we saw some order books on chain as well um behind the scenes you were getting pith prices that were updating at every Salon block you synchron synchronized U block updates um with Oracle updates at 400 milliseconds which is quite fast um and then in 2023 I think the rest of the world realized that rather than convincing people that performance wasn't important as decentralization and performance had to improve so we saw evm take on the same performance it got up to around two seconds on like say the super chains on on on some of the l2s but some of them are actually a little bit lower so we saw that sort of salana style speeds everywhere the pith network is designed to Super charge defi infrastructure so

that daps can compete with apps um and it does so across three different verticals uh and these are ones that we kind of compared on on on the first metric so there's trust speed and cost um currently there are three product Lines within the pith Network um which we'll go through now so the first one is probably the most wellknown it's price feeds so the Oracle problem for blockchains is very well understood you've got a blockchain that has programmatic Trust built into it there's either proof of work or proof of stake and there's validators um that are maintaining that stake with high levels of trust now as soon as you bring in something exogenous from the state you have to come in from a trusted Source or you've broken that trust

completely right and that's effectively the Oracle problem now the first generation of solving that Oracle problem was all around building trust but didn't think of this from speed and that was what I showed you initially um to solve that Oracle problem at speed you need to understand some of the mechanics of how Market Market data Works um so the market data IND industry for financial products is very welldeveloped it's about a 7 billion dollar a year business um there's about six large exchanges that make the majority of that money um that data is all most valuable at a time of zero because after about 15 minutes you can get the data for free so there's a very fast Decay on it so closer you can get that data to time of zero to put onto the chain the much more valuable it is

now the business model that large exchanges have for selling their data is incompatible with blockchains they sell via subscriptions every user has their own subscription it's limited the number of users that that um that they have and a blockchain breaks this because it's a shared state which means that it could only ever be one subscription not only that these large institution large exchanges have no incentive to see blockchains grow they have no interest in it um and so the first problem with getting this data on chain is where you going to find the sources from it so the p network innovated this by going to large holders of data that have never participated in the market data economy

before and that was large trading firms so if you think about the the way that the financial markets work in traditional markets there's your average user or Trader and then behind the scenes there's really large trading firms um Jane Street susah Hanah um drw jump trading um and these firms generally take on a tremendous amount of the volume that gets executed on on on different asset classes they are only on the buying side of Market dat they actually have to buy that market data back and it's something that's bothered them for a long time so the Innovation number one for the network was go get data from them they're never monetizing it it' be the first time for them to monetize their data and they've got tros of it put it together and you'll be able to stitch

together a very very comprehensive view of all markets with a cost basis of effectively zero because they're not comparing it against opportunity cost of selling Market data elsewhere um and that's how pith creates this price there's over a hundred institutions large institutions including all the large crypto trade uh exchanges like binance um all the way down to the smaller ones um there's traditional exchanges us Equity exchanges like myx I MX um and sibo Global markets which has has the vix um and they contributing data all in real time to a place called PNET which is a public blockchain it gets aggregated with full transparency and from PNET it's then made available to be pulled onto 80 different

blockchains and this is where the economics starts to work so as I mentioned economics for pushing onto a singular blockchain is broken because then it's available to everyone nothing gets passed back to the data owners by having everything put onto a singular application chain called PNET and then having each one when it's being used or called uh paid for it creates this model so every time there's an application that needs a price price on ethereum optimism arbitrum salana they pull that price directly from PNET and then they pay a small fee those economics are able to get passed back to the owners um and there's been three or there's been two iterations of uh of pith so far price feeds so the first one was pushed

onto salana um and then the second one was the migration over to pith net salana obviously only was was only available to salana and now it's on as I mentioned 80 80 different blockchains there's currently about 400 users um application users um of of fifth price feeds you can see what we get out of that from those three categories so number one the trust as I mentioned is a public blockchain it's all verifiable there's 110 actually um different data sources who are are validating the chain as well um the cost is very cheap because as I mentioned we're we're sourcing this from brand new data sources we're not sourcing it from somebody who's looking to extract billions of dollars for the cost of it um and and it's fast you know 400

milliseconds is is pretty close to the 300 milliseconds that we were talking about from Schwab the second product line for the network is onchain random randomness so it's called entropy um and this has a very similar problem set to that of price feeds it's an oracle problem so if you're running a game of chance how does somebody know that this has not been manipulated um how do I know that there's trust that's been built into this and um entropy does this with a commit reveal um mechanism this is a old cryptographic um uh technology that's been around for a long time two parties effectively commit Secrets um and and then they hash those secrets and those

hashes then get revealed um and then the the end result is hashed on chain um so that allows applications to show to their users that this has been um done by a program that cannot be manipulated um from here it's very cost about 5 seconds per update um for for entropy um and it's quite fast it's around um it's around uh uh 400 milliseconds uh response all right and the third product for the network is called Express relay so Express relay focuses on this meev problem so maximum extractable value or minor extractable value um is basically the term for all the um all the Revenue

that gets leaked out from a protocol to validators or miners um and when you think about the you know the best example being a a liquidation that would happen on a lending protocol generally what happens is the validators will either reorder the block so that it's more favorable to them or they'll create an auction where they're getting the the ability to put something in um in first and then they're keeping the proceeds of it express relay is different so Express there's a new primitive that connects protocols directly with an established set of Searchers um and it allows them to run the auction themselves and then be able to keep the majority of the

proceeds um that's number one so a lot of these costs get not leaked out to validators but maintained within the application protocol itself um and can be passed on to users as well so this gets us down to some of the the lower fees charging um similar to to what we've seen in in in FY um and uh um yeah so here we are um this other element of it that you can think about from a speed perspective is kind of time to Market um so if you can plug into this established set of Searchers you know one of the big problems that a new protocol will have is I have liquidations but I need somebody to be monitoring it so that when the

liquidation is triggered um there's somebody to go and execute it um new protocols actually struggle with this quite a bit B an established set of sear Searchers gives you the ability to plug into that day one and have Liquidators who are watching for opportunities right away so kind of time to Market is the way I would think about speed as being the fastest here and then obviously on the cost side those get passed on and not and not leaked out so there's a large number of applications that are are um using pith products already today as I mentioned um there's over 4 and uh 10 on 80 different blockchains actually 60% of Pro derivative protocols on any blockchain are using pith according to defi llama so that's about 100 applications in fact

it accounts for more derivative trading that has been enabled on blockchains than any other single protocol blockchain or Oracle um and that number has grown by about um about 10x um in the last year um and you can see the growth and the market share that onchain derivatives are currently eating up against um binance um and you know that's it that's the the the main um the main point of the network is to make sure that there is a possibility to keep making infrastructure better so that these applications compete not just on

decentralization compete on portability beat on those things that make somebody Wonder Isn't there a better system but it makes it so that everything is better as well as all those elements and the pith Network won't stop improving until we've beenen um trafy on all the other metrics cool thanks so much [Applause] [Music]