Oracles: Does the Backbone of DeFi Need Fixing? - Ep. 537

Transcript

So, the vision that we have for Pyth is that all the world's financial data should go through Pyth to be brought onto blockchains in a way similar to all the world's music going through Spotify. Hi everyone. Welcome to Unchained, your no-hype resource for all things crypto. I'm your host, Laura Shin, author of The Cryptopians. I started covering crypto 8 years ago, and as a senior editor at Forbes was the first mainstream media reporter to cover cryptocurrency full-time. This is the August 29th, 2023 episode of Unchained.

Arbitrum's leading layer-2 scaling solutions can provide you with lightning-fast transactions at a fraction of the cost, all while ensuring security rooted on Ethereum. Arbitrum's newest addition, Orbit, enables you to build your own tailor-made layer-3. Visit arbitrum.io today. At Token 2049 Singapore on September 13th to 14th, Balaji Srinivasan, Tyler and Cameron Winklevoss, Arthur Hayes, and 200 others will hit the stage, joining over 10,000 attendees. Visit token2049.com for 65% off regular ticket prices with the code Unchained. Link in the description. Buy, trade, and spend crypto on the Crypto.com app. New users can enjoy zero credit card fees on

crypto purchases in the first 7 days. Download the Crypto.com app and get $25 with the code Laura. Link in the description. Today's guest is Mike Cahill, CEO at Dorahacks. Welcome, Mike. Thanks so much, Laura. Excited to be here. You're a long-time Pyth contributor, and today you have some news about new developments in that ecosystem. But why don't we just start with your background? Tell us about Pyth, what it is, and what problem it's been trying to solve. Yep, absolutely. So, um, Pyth is a data oracle network, and it is focused on bringing financial market data on chain in a trustworthy way with very low latency.

Um and my background is in traditional finance. Uh I started at Morgan Stanley on the FX sales and trading desk out of college at a very interesting time where the markets for equities and futures had already become electronic and FX was undergoing that transition. And so I could kind of sense out the paradigm shift that was underway and I tried to position myself to be in an an opportunity to take advantage of it. So the first thing I did was move over to electronic trading desk at Morgan Stanley.

But then I realized that the winners were not going to be big banks, um but instead trading firms. And so I made my way over to KCG who is the largest US equity trader and I um helped build out the the FX trading business there. Um and then in 2017, um it was being acquired by another trading firm called Virtu. And during that time of the due diligence, it was effectively pencils down. And so you had a bunch of basically hungry traders sat in a room. Um the ICO boom was well underway and you can imagine the scene where um we were we were very quickly learning about crypto. Um so we were all attracted to the kind of very obvious arbitrages that

were were available on on centralized exchanges. But what really ended up being a paradigm shift for me is as I started playing around and getting an understanding, the first trade that I did on EtherDelta was an uh absolute eye-opener. Um I had already understood that the the blockchain um innovation would allow for smoother remittance around the world. Um but this was the first time where I fully appreciated that something like wealth management um or decentralized interactive brokers was possible. Um and you'd be able to effectively give these financial planning tools to say a like a Kenyan farmer.

Yeah, and so that was my my journey down into uh into crypto. Uh I I got a job or I was recruited to jump trading when the crypto desk was emerging and I started there in 2019 and that's when I started getting involved in the in the Pyth network. And today I'm announcing the the formation of Duoro Labs of which I'm the CEO. We have got 20 people and we're a blockchain infrastructure company and we're really focused on accelerating the growth of the Pyth network. Yeah, so let's give listeners more background on Pyth.

You know, you're trying to solve this Oracle problem which is pretty sticky problem in crypto. Describe a little bit what it is the problems that Oracle providers face in the crypto ecosystem. Yeah. The Oracle problem kind of by definition is the shared state systems of of blockchains don't have access to exogenous data and so someone needs to bring it on chain. And then there's a whole bunch of trade-offs with who you're going to trust to bring it on chain. And we've lived through several cycles of them and each time there's a big issue, I think a new Oracle network kind of gets created. So if you were to just start with building a application

and you wanted to connect it to a an Oracle network or you wanted to connect it to an Oracle. Let's say that you were building a lending protocol. The most intuitive thing you that you would do if you were not that custom to dealing with blockchains is connect to say the Coinbase API. And then you'll run into some problems with that. For instance, there is downtime with the Coinbase exchange or downtimes with their Coinbase API. They may update it at some point. They may not have all the assets that you want to have covered or they may under represent the market.

Like they they may have one of the assets but it mostly say trades on Binance. And so your price on on Coinbase may not be reflective. And so you can start to build out a case for having to add more data sources and you effectively have to deal with certain trade-offs. And most of those trade-offs are around the speed with which you can update. And so the reason why we developed Pyth or I started working on Pyth was the prevailing solution at the time, um, a few years ago, which was Chainlink, had two issues we thought with it. The first one was that it was relatively slow. So it was

updating every say hour um or 50 basis point move. And then the second problem with it is that it only uses free data. And when you're working at a kind of a systematic trading firm, um there's two things that are drilled into your head constantly. The first one is that latency matters, and if you're slow, you will have adverse selection or you'll lose money. Um and the second is that market data is pretty expensive. Um so financial market data in the traditional asset uh world was $6.5 billion in revenue in 2022. So if your model is predicated on getting that data when it's free, you're going to have a very high

limitation to what is available. And so that's really where Pyth decided to focus on um in terms of the sector and um and and how we were going to to develop the network to provide that with the lowest latency possible. And so how does Pyth do that? So Pyth is inclusive, it's a first-party data network, and it's inclusive of sources. So kind of the nodes in the Pyth network are the ones that are actually publishing the data or generating the data themselves. And so those consist of trading firms and exchanges. So there are 85 data providers in the Pyth network today. Um we count pretty much every large trading firm, almost without exception, from Jump Trading, DRW,

Susquehanna, Jane Street. And then we have pretty much every large exchange um on the crypto side from Binance on down to the smaller ones um and also some traditional exchanges. So Mayax, IEX, Memex are US equity exchanges. And in December of last year uh Cboe Global Markets uh joined as a as a data provider. Um now to to us, this is an a very a very important component of it. And the analogy that I sometimes will make is that when we looked at the model that was not inclusive, so the reporter network which relies on nodes that effectively go scrape data from the internet and publish that on chain

felt and looked a lot more like digital music in the late '90s where you had Napster that enabled people to for a short period of time make all the world's music available for free. Um obviously that model is broken because the intellectual property creator is not rewarded for it. And so you have this kind of tragedy of the commons. Um so what model worked was the Spotify model where the content creator or the the um song creator um is rewarded. And that is what we sought to do uh on the on the Pyth network rather what the Pyth network seeks to do um which is to be a place where all the world's financial data can be distributed through in a similar way to all the

world's music being distributed through Spotify. And the crypto markets are notoriously volatile and there have even been flash crashes on exchanges as kind of um you know, professionalized as Coinbase. How do you distinguish either volatility or flash crashes from something like price manipulation? Ooh. So, typically what will happen with like an Oracle attack even, and we'll just use that as an example because that'll um give you someone who's deliberately doing this and it'll give us a a framework for thinking about how it could happen without the deliberate element to it, is that you will find a small exchange

that's being used as an Oracle and you will momentarily move the price and then take some action on chain. So, for instance, if you were able to take out a loan with something where you can make the value of it on the Oracle go really high, you take out this loan and it comes back down, the lending protocol typically ends up with bad debt and that's sort of what's happened with like the the Mango um Oracle exploit and it's happened on on Venus as well. Um so, those are obviously, you know, bad outcomes um and something that you want to defend against. There was a really cool blog post written by Sam Szyszka like a number of years ago that dove into the details of that and his

conclusion was the best thing to do is to slow things down, right? And so, he kind of advocated for some version of a time weighted average price. And in fact, you've had many people use the Uniswap TWAP as their Oracle. And the idea is like, well, we don't really need updates all that frequently. Um we just want to have something that's not going to have this exposure. Now, from a trading perspective as I mentioned before, latency makes you effectively lose money somehow. Um you can think about impermanent loss almost as latency arbitrage um with adverse selection being the the LP holders because they're dependent upon somebody who has faster information

to come and update the market price at their expense. So, when you kind of put this framework together, you know that, you know, it's not great to be slow, but you don't want to take all the risks of being fast and wrong. So, what we what Pyth has designed is very innovative. So, the first thing is each of the data sources are required to update at the kind of the the block time for our PythNet, which is every 300 milliseconds based on this one of technology. Um but, they're updating not just a price, a single price at any point.

They're also including confidence interval. And so, you have a band with which you can use to determine whether or not that price is very trustworthy. And this gives us this innovation allows us to update prices very frequently and give the protocols who are downstream the ability to make more informed decisions. So, for instance, if something did flash crash on a single exchange, and all the other data publishers on the network did not represent that, then it would be represented through the confidence interval because it would show you that the confidence is very low and the confidence band is wide.

So, the idea there is that you can basically interpret the price as saying, "This is generally where the price is, but it could be somewhere in this range as well." And so, that is what allows Pyth to be able to be very fast, but also very accurate. And you mentioned some data providers that definitely come more from the traditional financial world rather than crypto. And I wondered how those conversations went. What does it take to get that kind of institution to want to partner with a company that is essentially trying to bridge, um you know, from the traditional financial world to crypto.

Are they open to it? Um or is it something pretty out of the box for them? Is there a lot of persuasion that needs to happen? Tell us a little bit about that. So, there there's basically, uh, two types of parties that are interested in providing data on, um, on PIF. So, the first one would be sort of, um, like trading firms. Um, and their motivation tends to be that they've never monetized their market data in the past, right? And so, uh, I'd mentioned that six and a half billion dollars of revenue, um, is earned through market data. It's mostly earned by the large exchanges, and the trading firms

are not participants on that. They basically use the the market data to be able to make trading decisions, and that that's about it. So, one way you can think about that is, you know, PIF in this example is sort of like Airbnb, and you know, the trading firms have these found assets where they can now monetize them for the very first time. And then the second group, so, if we think about someone like Cboe who does monetize their market data, the way they think about it is it's about 20% of the revenues from all these large centralized exchanges.

And what they want to do is be a part of the future, which they assume could potentially be 20% of the, say, the the the size of DeFi, um, as a part of the actual data component of it. Um, so, they want to be early on the data. So, those are really the two things that um, I think motivate most participants within the network. And then also, some of them are all already creating their own, like, for instance, reference rate. I don't know how that differs from this. Is that, um, something where those products would be competitive with what, you know, it it like if they were like if let's say cuz I know CME, so you have Cboe on your network, but I know CME, I think, has these Bitcoin reference rates. And so, is, you know, would working with you be

competitive with their own product or is that complimentary or I don't even know how those really compare. I think they're quite different. So, a single sourced reference point is going especially one that's going to be mostly off chain is going to be a certain type of a market um and um what Pyth does is it creates the references that are the Pyth prices um through the combination of all these different data sources. Um now, there's a couple of ways that you can classify pretty much every Oracle network or um or solution.

And I can go through them now kind of very quickly. There's basically four categories. The first is um push or pull. Um the second is fast or slow. The third is uh first party versus third party. And then the final is sort of single source versus uh a blended source. Okay, so in the first one push versus pull in a push model basically the price gets updated to the shared state to the the ecosystem on a predetermined schedule.

Right? So, this is the way that Chainlink mostly operates and they are for Ethereum the um the schedule tends to be around every hour or 50 basis points. And there's a sponsorship where the gas is going to be paid for by the nodes, but usually it gets subsidized by someone else. Um there's some conflicts with this. So, if you're like a profit maximizing uh mindset what you would try and do is kind of degrade the product to the point where it's just above the switching cost in order to kind of keep the sponsorship and update as infrequently as possible.

So, there there's some some some issues with this. Um and then the other element of it is that if you wanted if you as a user wanted something that was faster, you're sort of limited. The pull model, which is what Pyth has created, um, means that all the prices are updated to a single repository at first. So, Pyth uses, um, something called PythNet, which is, as I mentioned, a version of the Solana. It's validated, um, by the data providers. There's 85 data providers that are validating it, and it has all the sorts of transparency requirements you'd expect from a from sort of a layer one, and so you can kind of check the um, the fidelity and the, um, providence of the aggregation. And any one of those

updates is able to be delivered to any of the chains that Pyth is connected to. Pyth is currently connected to 30 blockchains, and all 320 symbols, um, that are currently live are distributed to any one of those three blockchains. Um, so that's the kind of the push versus pull. Fast or slow, we sort of went over before, um, and in order to be sort of the fastest, um, you need to be kind of inclusive. You have to have this first-party network. For slow, you know, this TWAP idea, um, was, I think, kind of, um, one of the things that people thought about as being a potential solution for a little while, um, but we sort of kind of moved on from that.

First party versus third party. So, in the third-party model, it's kind of easier to set things up. You you can go to websites such as CoinGecko and CoinMarketCap, um, but you're limited to whatever is free. And so, crypto market data has for the longest time been mostly free. And if you looked at the traditional market data, it was largely free for a long time, and then it became 20% of the value of of the revenues from from large exchanges. Um, I would imagine something like that could happen in the future.

Coinbase has started charging market data, um, a little over a year ago, um, and and we'll start we can expect that other exchanges do the same. Um and then first party is kind of inclusive. And then finally, single versus blended source is if you use a single source, there's going to be some of those constraints that I mentioned with um using Coinbase um versus a blended source, which you can just have contributors that add and create just a more robust price. Um so there's another example of someone who also does an Oracle, which is Binance. They have the Binance Oracle, but they're also a contributor to Pith. And so there's some users that may only want Binance, and then there's others that would want to use, you know, kind of Binance as a contributor to to

Pith. And so when you're vetting data providers to add to the network, what are the qualifications that you're looking for? We are looking for institutional grade companies that have a strong reputation um in so much where there is more at risk for them to lose in another business than there is for them to gain by trying to manipulate the Pith price. And so they're very very much household names um all of these 85 data providers.

Um and they're all public. So you can go on the the website of pith.network/publishers um and you can see all of the um all the publishers are part of the network. At this stage of the network, it's very important for people to derive the trust from these names, right? So it would be a crazy thing for like Jane Street to try and manipulate the price of um something on Pith because the amount that they would end up losing in terms of goodwill and on their name from kind of publicly doing this um is going to be much more than they could potentially gain from this.

And so that is really the litmus test for whether or not um someone should be a data provider into Pith at this time. In the future, when this move when when things move to a decentralized permissionless setup, um this will be determined by the kind of the governance token holders of the network, and they'll be able to do that based on the quality um of the proposed data provider that gets added. And there'll be some staking and other kind of incentive mechanisms in place to protect anyone from coming in that doesn't have the best of intentions.

So, something that I find so impressive is that you've integrated 30 blockchains, you know, as far as I understand. That's that's a technical feat. Um, just talk a little bit about what it is to try to provide for so many different blockchains, just yeah, on a integration on a technical level. It's it's it is a lot of work, and um my co-founder and CTO, Jan, um has lots of opinions on um kind of the best practices that he's learned by dealing with so many blockchains.

Um, some of his insights that um I think he would he would mention is that um dealing with EVM, you get lots of benefits of ancillary tooling. Um, so you get good great wallets, great analytics, and um you know, kind of sample code that you can use to develop applications. Um, and so that's that's actually a really nice thing. Um, within Rust, so that'd be Cosmos and um Solana, there are great attributes, and those languages are not specific to blockchains. There is also, you know, kind of a a a rich history of other application development there, and so you also get very good libraries and um um and

resources. On Move languages, you end up having some really cool developed things specific for blockchains, um but you don't yet have the full robust suite of toolings. That sort of needs to be built out. And so each time we expand to a different chain, um we have to go through the process of learning how to operate on that chain and and it is very difficult and humbling to do. And um, but it is a core tenant to what we at Pyth or at um, Duoro would like to kind of lead with. We want to make sure that we scale very quickly to as many blockchains um, and that as many projects as possible have the

opportunity to use this data. Um, and um, and so far it's been the case. So we're on 30 blockchains and on 10 of them we have a market share of over 90%. The way that we think about scaling Pyth to different blockchains has similarities to the way that Facebook grew. So initially it was just the social network for Harvard and then it was the social network for Harvard and the rest of the Ivy Leagues. And then it was US um, universities and so on. Um, and so Pyth started out at on Solana.

And it grew to over 90% of the TVL um, TVS there that used an oracle. Um, and it really has never left. Um, and it was an advantageous time because this is where, you know, high throughput DeFi was really being innovated. And so high throughput DeFi was first only possible on Solana and that was the first place that we saw on-chain order books or on-chain perp um, trading platforms. And now it's sort of table stakes. Um, high throughput DeFi is a trend that is sort of unmissable. You've got high throughput specialized blockchains like Aptos and Mou and and Sui. Um, and then you also have the modularity of

layer twos um, for the EVM world. And so uh, Arbitrum or Optimism. And when we think about where we can kind of build up the most kind of uh, critical mass of users, we tend to find areas where this is a priority um, because Solana has done such a good job in showing what was possible and now other chains are are sort of running with that baton and that's kind of how we think about scaling. As you mentioned, Pyth has grown very rapidly in recent months. So, why is that and how has it grown?

Yeah, it's it's been really remarkable. It's exceeded my expectations for sure. So, basically Pyth has gone cross-chain in January of of this year and on the Pythnet side we create a supply or Pyth creates a supply of 320 symbols that are updating call it two times a second. So, it's roughly speaking 50 million aggregate prices. Now, in January 2,000 of those were being requested per day to be delivered. And then in May we saw a spike to about 150.

In June we saw about double that and then in July it was around 1.3 million per day and this month so far we've been seeing update requests for over 2 million. So, why is this? The the the first kind of power user of Pyth was Synthetix on Optimism when they upgraded to V2 with their new perpetual markets. And what they realized was that having a low latency poll model oracle would enable them to create an unprecedented trading experience. So, perpetual markets a year ago were not all that common within DeFi and they've

grown very very quickly. Part of the reason for the growth has been the lack of FTX and the kind of desire for people to now have self-custody in their trading. And so we've seen perpetuals on chain explode in terms of their growth. Now, because Pyth has been able to help synthetics in creating a kind of best-in-class trading experience, I think that most of the the competitors or people that would like to build something similar have found Pyth in similar capacities. Today, if you were to do a derivative or take a derivative trade on on on a DEX, there's a 50%

chance that it's powered by Pyth, which is really remarkable growth given the kind of the short time frame with which we've kind of gone multi-chain. And how is Pyth like what what are the range of ways that Pyth is being used? So, like most oracles or most protocols that require oracles, the two largest use cases are lending protocols as well as kind of trading ones. But we also see things like structured products and other kind of um kind of collateral management protocols that that that use Pyth. But generally speaking, those are sort of the broad two categories. It's kind of trading and then and creating collateral for

for lending. All right, so in a moment we're going to talk about your news, but first a quick word from the sponsors who make the show possible. Join over 80 million people using crypto.com, one of the easiest places to buy, trade, and spend over 250 cryptocurrencies. With the crypto.com Visa card, you can spend your crypto anywhere and get rewarded at every step. Up to 5% cash back instantly, plus 100% rebates for your Netflix and Spotify subscriptions, and zero annual fees. New users enjoy zero credit card fees on crypto purchases in their first 7 days.

Download the crypto.com app and get $25 with the code Laura. Link in the description. Join over 10,000 attendees for this year's biggest crypto event, at Token2049 Singapore, on September 13th to 14th. Sandeep from Polygon, Eric Wall, Chris Burniske, and over 200 others will hit the stage, joining the industry's most influential for an unforgettable experience ahead of the Formula 1 Grand Prix race weekend. Singapore will transform into a crypto hub for a week, from September 11th to 17th, with over 300 side events that will make for unparalleled networking opportunities.

Builders and investors at the bleeding edge of innovation will drive an agenda that covers the ever-evolving regulatory landscape, the convergence of crypto and AI, Web3 gaming, NFTs, and the metaverse DeFi scalability interoperability, and many more. Visit token2049.com for 65% off regular ticket prices with the code Unchained. Link in the description. Arbitrum stands at the forefront of innovation as the premier suite of layer-2 scaling solutions, bringing you lightning-fast transactions at a fraction of the cost, all with security rooted on Ethereum.

From DeFi to gaming, Arbitrum One plus Nova is home to over 500 projects. And with the recent launch of Orbit, Arbitrum welcomes you to build your very own tailor-made layer-3, or as the Arbitrum ecosystem calls it, an Orbit chain, directly on the Arbitrum tech stack. Designed with you in mind, Arbitrum empowers you to explore and build without compromise. Propel your project and community forward by visiting arbitrum.io today. Back to my conversation with Mike. So, as we mentioned earlier in the show, you are now launching this entity, Duro Labs. Tell us um how it will be working with the Pyth network and I believe

there's um, a task uh, coming up very soon which uh, you you might want to talk about this Perseus upgrade. Absolutely. So, so yeah, we're very excited today to to announce the launch of Dural Labs. So, Dural Labs is a company that has 20 people ranging from contributors that were at Jump previously um, to alumni from Goldman Sachs or AWS, Chorus One and many others. And what we're in looking to do is build out tooling and development for the network and really try and accelerate the growth.

Um, now the Pyth network has had a few stages of of you know, big technology upgrades. And so, as I mentioned before, Pyth started its life cycle on Solana. And that was in September of 2021. And the idea was we needed to be on a layer one blockchain where there was going to be the characteristics that could give people the necessary assumptions for trust. And um, we wanted to make sure that you could use the Solana block explorer and look at the aggregation back into the um, the math of it. Um, and you know, kind of really understand how it was

working. And so, a layer one blockchain was important. What we always viewed that Pyth needed to be on multiple chains. And so, because latency is such an important component characteristic of valuable data and the type of data that Pyth wants to provide, we needed to be on the blockchain with the fastest um, intervals. And so, that ended up being Solana. So, the you know, decision criteria was was quite straightforward to begin with given the those constraints. And we knew that we'd go cross chain.

There were times where Solana became so popular and it became very congested. Um there were NFT mints and landing Pyth data on a congested Solana was very difficult. Um and there was also downtimes from time to time. And so we realized that we needed to create an application-specific chain. Um and if we can do that and we can convert to this pull model, um we would be able to scale very quickly to the number of symbols and number of data providers um without having to have a new variable cost with how much we grow. We can have the variable cost sort of be downstream. And so this is our V2 and the emergence of PythNet.

Um and then we plugged in PythNet to Wormhole um and started growing cross-chain about 8 months ago and now we've grown to um to being on uh 30 different blockchains. The Perseus upgrade is a continuation of the development in the PythNet architecture that's get sent over Wormhole in such a way where now using Merkle trees for proofs and that allows for way more customizable uh delivery. In the past, we were relying on basically preset batches to be delivered to various blockchains, but now they can be fully customizable um with a single proof that anyone can use to validate um the the the messages

that were sent. So this allows us to see a decrease of gas from anywhere from 40% up to 90% depending upon how many symbols you're you're um you're using. We can increase the publishing frequency to other chains um by a factor of about two and it allows us to really continue to scale and grow the network to make sure that it's, you know, constantly the fastest and cheapest it could possibly be. And one other aspect that I think is coming is a permissionless state with token led governance. What will that look like?

So right now becoming a a member or a node in the Pith network is managed by the Pith Data Association. Um and that eventually will get turned over to governance holders. And so um that and other key features. This is where Pith as an application will have critical areas such as which symbols to add and which data providers do what um managed by the token holders. So this is a important transition in kind of the life cycle of Pith as it continues to decentralize.

In a in a decentralized state, if you're going to add new data providers, how does that work? Um maybe there's a proposal to you know, try to talk to this other centralized party and then are like certain people delegated to to represent the network that way or I'm just wondering how a decentralized entity interfaces with all these centralized entities. Yeah, so um it it is an interesting point given that the Pith network has probably more contributors than most decentralized protocols today. It is going to be a little bit of a challenge, but I think this is something that will get proposed in governance and a

framework that will allow basically general participation. There is a couple of good models out there um where we see, you know, highly participatory governance structures. Synthetics is a is a good example of one. And then there's kind of other models that are more like say Aave or Compound. Synthetics uses sort of committees. Uh Aave and Compound are sort of just like token holder votes. Um and so we'll allow the community to make proposals on um how this will work, but um I think that uh it's a it's a it's an exciting place to kind of experiment.

And do you have a sense yet of how tokens will be distributed? No, we haven't had any announcements around that yet. Okay, yeah, I'm sure probably the regulatory environment in the US is perhaps going to complicate that discussion. Um so tell us a little bit more about the the plans for Douro Labs. Um will you contribute exclusively to Pyth or do you think in the future you might extend services to other projects? It's totally possible that we could extend services to other projects. Um but for the moment we are laser focused on um really developing Pyth network. Um we have a whole lot of embedded DNA um with contributing to Pyth from um myself

and co-founders Kieran and and Jan. So, you know, that's where we think that we can add the most value, but you know, who knows over time I think that there's certainly within scope for us to be able to do um and work on other projects. And so the focus for Douro Labs will be contributing tooling for it just tell us a little like you're servicing developers, is that it? Exactly. So, it'll be continuing to advance the development of the network. Um and so that could include additional upgrades like the Perseus one. It could also be helping build out the network by um working directly with data providers to go through the um kind of governance process um and so those sorts of things.

The crypto markets are in a really interesting place right now. The volumes are down. Um we have, you know, market makers that have been withdrawing. It looks like Binance is in a precarious state. Um there I mean there's just so many things that make it feel right now like like the crypto market is just changing generally. And I was curious for your thoughts on kind of where it's going. And and I feel like also you you'll have an interesting perspective because you've seen so much growth during a period when everything else has been kind of um you know, in this transitional state. So, yeah, where where do you see the markets going?

You're absolutely right. It is such a weird time. Um we've now um seen Bitcoin kind of make new lows at a time where I think people felt like we'd kind of bottomed out for a bit. Um and I think that's sucked a lot of the air out of the room. Um so, the you know, the the interesting thing is we have not seen a kind of general retreat in terms of the number of builders. So, Pyth has had three new applications per week integrate. Um and that to us tells us that people are still building things.

Um and and that's where the optimist in me thinks that we've got the potential for you know, some some real long-term value and for if the markets become exciting for kind of the general public again, um way better kind of toys than we used to have or better tooling. You know, the the meme is, you know, builders build in these bear markets. We think it's the case, and I think that looking at Pyth's growth in terms of all these metrics that around usage underscores that. Um and so, um it's maybe a happy story for us to try and take a rosy view on things but if we were to get things to turn around, I think um in in terms of just the the market sentiment, um there's just way

better tooling and and um applications possibilities um than we had sort of in the last couple cycles. And what vision do you have for once things kind of work in a sort of ideal state or in a in a fully mature state? What will that world look like? So, the vision that we have for Pyth is that all the world's financial data should go through Pyth to be brought onto blockchains in a way similar to all the world's music going through Spotify. Um and what we hope to have happen is for um the vision of DeFi being what I understood as kind of that paradigm shift um when I first did the trade on on EtherDelta. It's um a decentralized way for people to

gain access to capital markets. In all the other trends that like major growth trends that have been enabled by technology, things become cheaper, um but they become much more broad broadly distributed. And so I think that's one of the big benefits that we have within crypto and and moreover within the potential of DeFi, um we would expect to see sort of the the kind of the general usage costs go way down. Um but in order to compensate for that, the number of users go up incredibly, which will increase the um the benefits for people who just haven't had access to these tools before. Um and that's what excites me.

And you know, if if you thought about places that didn't have taxi drivers before and now you can go get an Uber anywhere, um it's just a way better like way better experience for you as a user of the technology. Um I think that that is what's going to be enabled by crypto um when it's at its full um kind of potential. All right. This has been a fascinating discussion. Where can people learn more about you and your work? So, you can learn more about the Pyth network at pyth.network. You can learn about Dora Labs at doralabs.xyz. And um you can find me on Twitter at M Domkeho.

Perfect. Well, it's been a pleasure having you on Unchained. Thanks so much, Laura. Thanks so much for joining us today. To learn more about Mike, the Pyth network, and Dora Labs, check out the show notes for this episode. Unchained is produced by me, Laura Shin, with help from Kevin Fuchs, Matt Pillchard, Zack Seward, Vanna Vannavich, Sam Shirazi, Megan Gabus, Jenny Hogan, Lena Chimino, Shashank, and Margaret Kuria. Thanks for listening.