On The Brink with Castle Island - Mike Cahill (Pyth) on Solving the Oracle Problem (EP.364)

Episode Date: October 24, 2022

Mike Cahill, Director of the Pyth Data Association joins the show. In this episode we discuss: The origin of Pyth and how the project is addressing the 'Oracle Problem'. How Pyth works and how companies and protocols are using the product today. Mike's views on how equities on chain will evolve, and how Pyth is positioned in this market. The recent Mango Markets exploit and the future of on-chain market manipulations in DeFi. Pyth's network of partners and how the network is growing in 2022. To learn more about Pyth visit Pyth.Network.    

Today on the podcast, I sat down with Mike Cahill, director of the Pith Data Association. Pith is a data Oracle protocol that's built an impressive coalition of partners. And I was excited to have Mike on the podcast today to discuss a range of topics. Talked about the institutionalization of defy, talked about the future of on-chain equity, talked about recent defy exploits and market manipulations, and a lot more. So I think you'll enjoy this one. Without further ado, here's my conversation with Mike Cahill. Matt Walsh and Nick Carter are partners at Castle Island,

All of these expressed by them or the guests on this podcast are solely their opinions and do not reflect the opinions of Castle Island Ventures. Guest and host may maintain positions in the assets discussed in this podcast. You should not treat any opinion expressed by anyone on this podcast as a specific inducement to make a particular investment or follow a particular strategy, but only as an expression of their personal opinion. This podcast is for informational purposes only. Brought down by bad mortgage investments, Lehman, which has 25,000 employees will be liquidated. The federal government loans American International Group, AIG, $85 billion. This is a different kind of mortgage. market and the Fed is asleep. The federal government is stepping it to stabilize Fannie Mae and Freddie

Mac, the two mortgage giants that have been threatened by the housing crisis. The Bank of England has pumped 75 billion pounds more to Britain's ailing economy with a new round of quantitative easy. You print a couple trillion dollars and all of a sudden people started to worry. So out of this worry, we have something called the Bitcoin. Bitcoin. Mike, well, thanks so much for joining us today on the podcast. Excited to talk about Pith. Maybe before we do, why don't you just give a little bit about your background, how you got into crypto and how you came over to Peth. I have had a traditional finance background, started at Morgan Stanley out of college, worked in FX, then moved to a couple of different trading firms. The first one was KCG,

which is now owned by Virtue. And then I moved over to jump the focus on crypto and on business development. You know, my draw into crypto was just kind of trying to stay ahead of the curve of innovation within finance. And so initially I was doing things on voice trading and then I moved into electronic trading, and it felt like, you know, crypto was sort of the next evolution and cutting edge for the financial industry. And for the last three years, I've been working through Jump at running some of our partnerships. And two years ago, we started becoming big contributors to the Pith Network. We found an opportunity to solve the Oracle problem in what we think was an innovative way that allows for scaling. And I've really been leading Jump's efforts

on the business side to contribute to the project. People would be probably shocked at how much money is spent on data in the traditional financial services space. I'd love to get your take on that comment and then maybe a broader framing of what Pith is doing now and where you see this going in the future. Yeah, I agree. And that was sort of the exploration that we had when trying to figure out whether or not the current solutions to the Oracle problem were going to scale. So, you know, unpacking. like what you said. There's like somewhere between five to ten billion dollars a year in revenue for financial market data just in the U.S. most of that ends up being in the regulated U.S.

equity markets as well as like the futures markets. And it's segmented. Some of it is for super, super fast in the Kolo market data like down to the nanosecond. And some of it's for like the human readable kind of fast stuff that you'd buy through like a Bloomberg subscription. So when we were thinking about it, we were doing the thought experiment of like, well, how do you get that expensive data on chain? Because it ends up being pretty consolidated to just a handful of large exchanges that make most of that revenue. They're probably not going to be incentivized to be putting it on the blockchains and certainly not going to be incentivized to do that in a period where things are being bootstrapped and do it for very low fees. So what we kind of explored was like let's get Amazon or like Tesla data on. chain and you can't use any of the large exchanges. Like, how would you do it? And so at jump trading,

you basically are just on the purchase side of that market, right? You're only ever buying market data, even though, like, if you cut it down to its first principles, market data consists of the bids and offers that traders want to trade at, and then the prices which they just executed at. Now, trading firms are the ones that are submitting all these orders to exchanges, but exchanges have only been the ones that have ever monetized them. And so you end up having these troughs of data that just really have had no outlet. There's been no marketplace for it before. We realized that you could combine that together with traders and you can come up with a very comprehensive market and very comprehensive view of the market. And so the way that Jump contributes

to Pith is for U.S. equities, it will take its trades that have executed and report those as a function of where they think the market is. Now, it jumps trading sometimes thousands of times a second in various instruments. And so it becomes a very meaningful contribution. The other trading firms like Jane Sri Susquehanna, DRW, Hudson River, who've all joined the Pith Network, do the exact same thing. And so you end up creating this network of contributors that didn't really exist before. So you're kind of using like a found resource for some of them. They've never monetized this. And this is a way for them to sort of bootstrap a new network. So in one way, it has a similarity to Airbnb. You know, you're creating a new inventory and allowing people to monetize an asset that largely wasn't being

used before and doing so in a standardized way, right? Like, there was a way that you could use Craigslist to like rent out a room, but not many people would do it because it came with all kinds of risks. So by Pith being like an absolute standard, it is institutionalized the way that large trading firms and exchanges can distribute their data and then monetize it on the blockchain. So that's how the Pith Network creates its data and then it's distributed all onto onto blockchains, but it's not limited to U.S. equity data. This model, of course, while it solves that corner case, it's really scheduled to solve the greater financial market data, including crypto effects and metals and so on. It's a fascinating construct. And when you think of it, it's something that probably is uniquely

enabled by blockchain. So you read it. reference the Oracle problem. Maybe talk a little bit about that with some of the prior attempts to do this in the context of Defi have looked like and just how Pith is addressing that Oracle problem. Yeah, sure. So the Oracle problem really underscores the fact that smart contracts will execute autonomously based on what the code has explicitly written out. There's no kind of third party or centralized person or real fact checkers or let's say someone with discretion to make a choice. So if you have created a lending protocol and you have a liquidation level at, let's say, price 10, and the data comes in at 10, you will take that liquidation.

It's not like, well, we'll hold off here because we don't necessarily know if this 10 is trustworthy. Like the way that a bank would do, they're not going to release all your funds because there's going to be someone that says, hey, Matt doesn't always do this. And so we're going to just double check with them before we allow that. You don't do that on blockchains. And, you know, that's one of the that we like about it, right? Blockchain don't have middlemen, and so there's actually a lot more scalability with them, and you can cut out some of the fees. But because the data that's coming in is going to create a lot of actions that can have

high value on them, it becomes really important. So a simple analog is to say, like, let's say we had a smart contract, we're betting against the Pats and the Jets, and you're betting for the Pats to win, I'm betting for them to lose, and we lock up something to smart contract. And, you know, at the time that the game is done, one person's going to walk away with money from the smart contract. If you give the update, you're going to say that they won no matter what, I give it. And, you know, I'm going to say that they lost no matter what.

And so that sort of underscores a problem. You need something that's going to be independent that will have resilience and it'll have enough distribution of sources. So the way that this was done initially, like the bootstrapped way, it was pretty smart. You look at the internet and you say, you know, there's a lot of data here on the internet. And then you look at blockchain. you're like, well, I can get some of that over to the blockchains. And so there's the chain link model, which has a notion of a node.

And that node goes out and fetches data from the internet. So it's usually like public APIs. So they can go to coin getco, coin market cab, job finance. And they're just fetched data. And there could be 10 of them that are nodes, but they could be fetching data from like one or two sources. The sources are never disclosed on that protocol, but like the nodes are. And there are certain problems with it, right? it reminds me of when music was being digitized and like the late 90s and you know you had napster

and lime wire and people could go take the CDs that they had rip them make it available and people could download it there was one huge problem with that the person who owned the licensing rights to the music wasn't getting paid and so those companies got sued out of submission and we ended up of course with a model that ended up working for digital music but it looked more like Spotify where the record labels were producing and placing the music available. And then people were buying subscriptions. And then as they got played, they got some revenue share from the platform. So we saw this in the same way.

If you're taking data from Yahoo Finance, fundamentally, actually, that data doesn't belong to Yahoo Finance. It usually belongs to like the NASDAQ, NISI, who's making all this money selling market data. And if you were to distribute it onto blockchains, there's an unlicensed or uncontrolled distribution. It's not like you go and have every blockchain user, fill out a form and say, hey, you're just going to use this for yourself. You put on the blockchain,

anybody can use it. It's fully transparent. And so that's where we realize that the model wasn't going to work when it came to scraping data from the internet or pulling data from the internet. So the Pith model is, well, let's look like Spotify, right? Let's get the data owners to publish directly. And you get some really nice benefits from doing this. You cut out these middlemen. So there's no fees. Fees come in kind of two forms. Number one, it's the actual fees. like the middleman needs to get paid something. And so they're always going to add some fee somehow. The other thing is they add in a latency fee because usually they're going to aggregate this together.

Coin Gecko is not going to be as fast to publish FTX data as FTX is. So if you were to go to them to get the updates, you're going to be at whatever interval that they're choosing to publish it at. You're always going to be limited with whatever data they get. They may not have full order book. They may just have like the midpoint, top of book. and those might be things that you want to include in the future. So when we developed this architecture, we did so in a way where the output could be a price. And then we can include

something that had never been done before was a confidence interval. And so what this is designed to do is come up with a reference point where the price is likely to be. So like the price of Bitcoin generally has some kind of agreement around where they expect to be. But the confidence interval is designed to show the fact that different exchanges will have different values and there'll be dispersion across them. So if you've got an exchange like upbit in Korea, they may be trading in a big premium due to like the kimchi premium versus like finance global finance.com. And it would end up being included in this confidence interval. And so you basically can provide all that information in a single benchmark. The confidence interval thing is a really clever way to hold people accountable here.

What is the actual operational burden to provide data into this system? And you're curious how this works if you are known to be submitting the wrong data. How do you prevent bad data from getting into the system? Yeah, so that's one of the biggest problems with basically an Oracle network in general. So, like, how do you attack the system and how do you defend against attacks? So there's a couple ways that Pitts does it. So the first one is by having sufficient number of diversified, putty sources or data sources. So today there's 75, and each one of them has been disclosed.

And they are all basically blue chip names. And so they sort of stake their reputation on the fact that they're going to be the ones that are providing the data. So the trading firms I mentioned already, but there's a lot of exchanges that are contributing, both in crypto as well as in traditional assets. So if you can increase that number, and so 75 is a pretty high number, you know, you're feeling pretty safe from the likelihood that you're going to get a lot of trading firms to collude on trying to move the price of PIP. Now, we want to get to a point where it's entirely permissionless. So in order to do that, there's a transition to move to a proof of stake model. And basically, you require each one of the data providers to post a stake, and then they're

rewarded for contributing correct data, and then they're penalized or slashed if they submit bad data. And you can do this by basically generally assuming that the aggregate is going to be correct, and then like kind of penalizing people if they're purposefully malicious. And then we also have got a process with which you can challenge it and it will go to sort of an adjudication or a review and you can have like human intervention. Yeah, you can see the proof of stick part there makes a ton of sense. Yeah. Yeah.

And so we've outlined this in the white paper. It's not fully live and that's why it's for the moment relying on the reputation of large firms who like most most of the large firms have another business. They're not in the business of like just trying to make money from Pith. Like they're big trading firms and like they have much more to lose from their reputation if they screwed up the price in the Oracle. And that's sort of the model. That's what we like for now.

But over time, you know, it could be that someone wants to just be in the data business and they can contribute to it. That model would work better for them. It's an interesting time to have you on the podcast. This Mango Markets incident happened at the end of last week. And there's a lot going on in Defi. I mean, it obviously goes way beyond just cyber security, which is obviously also a huge issue. But some of these attacks that we're seeing on these D5 projects look a lot more like market

manipulations or whack of appropriate mechanism design at a protocol level. What's your take on what we're seeing out there? Curious your take on Mango. Curious your take on how Pith fits into this overall story. What's crazy is that you know the person exploited this one. And it's very unusual for this to happen. And I think in traditional markets, as you said, this would very clearly fall under some sort of a market manipulation regime in crypto. I suppose it's not as well defined.

It was initially reported as an Oracle exploit or attack. And, you know, those have always been the case. Like the synthetics one happened. And like synthetics was using kind of this very small retail brokerage for the price of Korea won. And it was on the weekend. And they were like, we're a 24 exchange. There's no liquidity there. And they ended up having the same thing and got drained.

It happened with Venus when they were using the Venus governance token. It tends to be when you put your governance token into the protocol, which usually makes sense when you're developing a protocol. You're like, well, we use our token at some point. But you end up coming against the fact that there's not usually as much liquidity on those. So in the Mango one, so full disclosure, Pith is the primary oracle for most of the markets that Mango offers. And there are two markets where they use a different oracle, they use switchboard. for Mango and Avalanche. And so Pith was not involved in this one at all. You know, the mechanics of this

where the trader noticed that there was an imbalance of what could be taken out in liquidity as a loan versus how much it would cost to move the underlying markets on FTX and Radium. And so they basically deployed capital. Now, the switchboard Oracle was only looking at FTCS and radium at the time. Pith at the time had 13 contributors. And Pith also has got that confidence interval. And that's designed to give you some sense of what's going on in the market. Usually it's for fast-moving markets to say, hey, this confidence interval is really wide. And some protocols on Salon in particular will take that into account and limit the amount of transaction you have to go through. Like if the Oracle, sorry, if the confidence interval gets too wide, they'll actually halt a lot of the features.

When we looked at like what happened with Pith versus the switchboard Oracle, switchboard went up to 40 cents, Pith went up to 15 cents, which is still very high considering it came from, I think, about four. But the PIF confidence interval was also like plus or minus about 30 cents. So if you were to take this in sort of the most conservative view, you'd say, all right, 15 cents minus 30 is actually minus 15, but you'd say actually the collateral now is probably worth zero. It's not really designed, though, in that way. You know, it's sort of a, this happened to kind of work. I think that the design of the protocol should be much more conservative when looking at liquidity. The fact that you don't have information on chain currently that determines or that gives you how much liquidity is available

is a problem. We have a plan and we're close to being able to be more public with it on how to bring liquidity oracles out and sort of bridge the lack of information and the asymmetry around that to help protocols make better decisions. And I think that will help. I think that's probably the best answer. Having more data sources is sort of like the kind of the best you can do for now. you can slow things down, but that's usually not like the right answer to like not do things fast enough. I think that you'd rather have like the right information and be a little bit more on the conservative side. I guess you can get at this from a technical perspective, but there's also just the regulatory perspective is that as you point out, this would be market manipulation if you did

this in a commodity market that was regulated by the CFTC, for instance. And so how do you see the industry grappling with some of these questions around just should you even be allowed to do something like this? Yeah. Ethically, no. And so I think that the, you know, it would be very cool if there was like a self-regulatory way for us to, as an industry, prohibit this from being an allowed thing. One of the things we get asked the most for Pith is to add more symbols.

And we usually will push back, given the lack of liquidity. Because it tends to be ones that are like just available on, say, radium or like a uniswap. And even if you have a bunch of trading firms that are also trading those, like you end up having a bit of a hall of mirrors if there's only a single exchange. If you've got a couple exchanges and some of the traders are doing OTC, then it becomes, you know, slightly more robust. When you've got a centralized exchange, you tend to trust it a little bit more because you know that they've done some KYC. In this case, because there's no rules against it, it sort of limits the amount that you can take that trust in the KYC. You know, I think it makes sense for some rules to be in place and for

this to not be necessarily allowed. I don't know that there's anything that FTX can do right now, assuming that this is the same person who's done the exploit. So it ends up being a pretty confusing thing because you can't get too heavy-handed with regulations in one jurisdiction because we've already seen this happen, right? Everything else just moves away. It's like the currency markets. Currency markets are global. If you end up putting in two heavy regulations, like they just move to liquidity just moves elsewhere. So the trick that the currency markets did is they regulate the underlying people who are doing it. And so like if you work at a bank, you tend to be registered, like a registered rep and you're trading FX or whatever,

like, which is an unregulated asset. If you do something that is incorrect, you'll actually be penalized for it. So that framework kind of makes sense. And I think it's probably the right one to use as an analog. That's probably where we end up going. But we don't even have really clarity in the United States on who oversees the spot market for cryptocurrency. So it could be a while before they figure out defy is my guess. Yeah, I think so. It's really interesting to look at Pith through the context of, look, it's a huge market to just focus on cryptocurrency.

But can you imagine a world where every asset is tokenized and represented on a blockchain and U.S. equities trade on blockchains? Do you envision us getting to that level? What do you think stands in the way of Pith being able to service a broader kind of asset class? Yeah, I think we have to. You and I have both been in crypto for a little while now, and we tend to get very probably focused on solving the incremental problem. But the things that excited us when we were first thinking about this were like, well,

when I go by a house, it's a real pain in the ass to do all the title transfer on paper. Like this stuff all needs to happen in a database. So in order to get to those imaginative, like, this would clearly be a better solution. You need to be able to do this with real world assets. So I don't know when things will happen on exchange, but I do think that something like Pith will enable them to happen much faster than they would otherwise. So when we were thinking about like how does somebody in China buy exposure to Tesla today, it's really tough.

It's going to be a really challenging time for them to open up a U.S. brokerage account. And so, you know, there's that kind of blockchain fixes this idea. And it sort of does, right? Like, because as long as you have something to tie back a small paramutual market to the greater markets, then, you know, you can give them a reasonable experience. And then, you know, because this happens today with perpetuals. The perpetual markets are not fungible with one another, but they tend to be very liquid. And also, they don't trade at huge discounts or premiums that often to the rest of the spot market

because they have this idea of a basis or a funding rate. So I could create a Bitcoin perpetual market with just a handful of people in it. And so long as I have a way to incentivize the longs or the shorts to try and bring that into line with what the Pith price is, like kind of the global price for Bitcoin is, then it will be a decent market for people to trade at. And so I think that's what we'll start to see.

We'll start to see experimentation on smaller markets that were basically empowered by the market. the fact that Pith is representative of the greater market. And then I think those will grow where there are needs. And it is very similar to the idea of Bitcoin solving payments in countries where people don't have bank accounts. Like, you know, to me, that's very similar to like in Scandinavia, Nokia was the way that people can communicate with one another because they weren't able to build landlines.

In the U.S., they didn't need Nokia early on. They didn't need a wireless phone early on. they had great landlines. But then eventually, when you had it shipped over as a final product, you're like, oh, actually, this is better than our landlines. I think that all end up happening as well. This will get used as an access point for people that are really kind of financially repressed. And eventually people that are not so will be able to realize the benefits of it as well. It's fascinating. So you have some contributing members of Pith that are very deep in the crypto space, crypto-native, but you also have some firms that either aren't in crypto yet or haven't been

public about what they're doing. So I'd imagine that that longer-term vision of, hey, equity is on chain, different types of assets represented on blockchain, that has to be a unifying sort of thesis area for some of your members. Absolutely. Yeah. So one of the most recent additions to the Pith network of data providers was CBO. And so they own, most people will know them as owning the VIX, but they actually own the third largest U.S. equity exchange bats. And they have a material market. Data's business that they sell this high-quality data to. And they view joining Pith as a way to be an early participant in this ecosystem. And that's really exciting because they bought RXS, so they're getting involved in the kind of the crypto order book side of things. This is very

clearly a DFI play. They understand that this could be transformational to their business and they want to be involved at the ground level and ahead of the other ones that haven't yet joined. So I find that to be really cool. The one other element to like early participants within Pith is a lot of times they'll join because they're looking for guidance on how to get involved in DFI and they're looking from guidance from a project that has largely been reputational to be risked. So like if you've already heard of all the other names and something were to go wrong, like we're all in this boat together. So they can benefit from joining where you've got really professional integration people and engineering talent that can kind of handhold somebody through the process to onboard. Because

each one of these data providers, like CBO is publishing data to the blockchain on their own, right? They're running this and it's being published. That is not a small feat. When a lot of projects will list like, here are our investors or advisors, usually they just cut a check of some sort. With the 75 people that have joined PIP, they're literally contributing data on chain to both Solana and to now PithNet. And that's like, that's a pretty big commitment. That's a huge commitment. One of the things that comes to mind here is which underlying blockchain are these assets going to be traded on? Obviously, Pith has a very strong presence in the Salana ecosystem, but I know you guys are moving to cross-chain functionality. How are you thinking about which base layer to prioritize over time? So Pith was deployed in Salana Mainnet September of 2021. The thesis was always to be cross-chain. And when you do a quick analysis of block times, you realize that you have to be at the fastest one. was the fastest one, cheapest gas.

So it made the most logical sense. It also was a really vibrant defy community. So it was quite good to be based there for the first year. And over time, we realized that we needed to have some redundancy as well. And so we created a private instance or an application-specific chain for Pith built on the Salana technology. That's called PithNet. And then we use Wormhole to basically deploy on lots of chains almost all at once.

So the typical way that cross-chain was done, in the past is you would deploy in a kind of linear way. Like you would take two or three chains per year and you would have a deployment. Like Ave would deploy locally and they would have some sort of liquidity locally, but it wasn't really meshed together too well. In this model where like it's becoming really exciting and interesting on the interrupt interoperability side, we get a deploy really fast. And so Pith just today, actually, during the Aptos launch.

So I don't know that there's there, but it's. happening today. So Pith is now gone live with 10 launch partners on day one with Apto. So they've been working on Pith for the last couple of weeks in DevNet. And so we've got pretty strong penetration there. Piff had a launch partner on BNB, which is Venus. So they're the largest lending protocol there. And then Wombat, a few others have also joined. We expect to see Pith basically pop up on every EVM chain over the next couple of weeks with really strong launch partners. So today on Silana, Pith is like massively comprehensive in terms of its coverage. So it's over 98% of the TVL is secured by Pith that used it as an Oracle. And, you know, Optos, it seems like it's going to

have similar characteristics. It's almost like they come for the speed and then they stay for the confidence intervals and the ability to add symbols very quickly. And so Pith is going to go cross-chain. We're not going to necessarily make a bet on one or, you know, one or more chains, but we have got the advantage of being able to fully distribute 100% of the symbols in a really scale. scalable way. I'll just describe it in slightly more nuanced. So of all of the symbols, there's about 100 symbols today on PithNet. They're basically published through Wormhole. And then from Wormhole to any of the chains where there is a target integration or target contract has been created, you can take a signed message and you can use it in your protocol. So it's like you're watching

the Pith update in real time and it's public, right? It's not private. You're seeing this update every 400 milliseconds. And so if you're like a liquidator running the role of a liquidator for Venus, you could be watching the PIF price update on Wormhole. And then you can say, okay, I'm going to take this sign message and I'm going to bring it over to BNB, just pay the gas. And I'm going to use it in my liquidation. I'm going to earn the rewards for it. And, you know, it's been programmed into the Venus smart contract to allow you to do so. That is really cool because it's very skill. It's got this hub and spoke model where anyone can join in and grab one of these symbols and then bring it anywhere, as opposed to saying, right, we've got 200 symbols on this chain, but 10 symbols over

here or something like that. That's fascinating. I mean, it has to be hard to keep up on all of these new launches. You have L2 launches on Ethereum happening at a pretty regular interval at this point. You have privacy layers that are coming to some of these chains. How do you guys prioritize? Is it just follow where the customers want to be building? Yeah, we have some, we have some, like, view on where there's the most activity and where, like, Pith can help the most. So Pith is very much in the defy ecosystem today. And so if there's a vibrant defy community, we look to try and secure that. I would say that there are two primary use cases for Pith are lending and trading. On the lending side, it's quite straightforward. It's used for kind of the marking of collateral and the liquidations.

On the trading side, there's three different primary flavors. There's like decentralized options, vaults or fixed settlement date contracts that could be like a future or a forward, where they just look at PIF when they're going to settle. They're the ones that I described earlier, which is like the perpetuals or paramutual markets that need to get tied back into the benchmark and they use PIF for mostly like an hourly funding rate. And then there's a third category, which is cool. We're seeing a lot of this grow. And I actually think it's probably the right model for AMM's to take.

But basically it's the idea of letting people participate in AMM that crosses at the Pith price and then has some slippage built around it in a bonding curve. And that's cool because when you do things that way, instead of being an LP that gets armed back into line and has to deal with the impermanent loss, the LPs in this system actually get very few losses. And it mostly trades at top of books so you don't have quite as much depth through the liquidity or any need for it. The TVLs tend to be a little bit smaller.

But it's a really efficient model to sort of run an AMM in kind of a lightweight way instead of doing a full order book. That's really cool. So maybe just stepping back from Pith itself and just talking about the market here for a second. Before we started recording, we were talking a little bit about just the lending landscape and some of the faces and names that have changed over the past year. It's obviously been a crazy year with Three Arrows and Luna and just the various lending firms that have blown up.

What's your take on just the overall health of the ecosystem now? Has this scared away people from joining Pith? Just seeing the carnage in the market. And your overall just macro view on crypto lending markets. I feel like we've sort of bottomed because it's tough for me to picture where there's a lot more leverage in the system. I feel like we'd have to go down a lot more. And I feel like the leverage would be a lot less. So it's funny because I talked to people that I worked with in the past that are really active traders.

And I asked them what they think of the crypto market. They're like, oh, no, man, I've moved over to trading FX. I don't mess around with crypto anymore, which is, I think, a good sign. You want the markets to have gotten boring here while the rest of the rest of the, the markets kind of do other stuff. I don't think it's scared away, people. It's kind of funny because, you know, and you guys talk about it on your roundups,

like there's tons of VC activity going on right now. And so that hasn't slowed down at all. And so we're seeing much more interest. Like the CBO, the fact that CBO joined Pith has really opened the eyes of a lot of other traditional financial firms to want to be a part of Pith. And so we're seeing a lot more interest there because they're a behemones. And like when they have chosen Pith as their like sole partner for Oracle, like growth. That means a lot. General, like, views on the market, I would say that

we're going to probably be in a sideways period for a little while, although I would say that the Aptos launch, which just happened, you know, today did seem pretty good. It felt like, felt like a launch that would happen like a year ago. Let's hope that it, you know, kind of holds up. But, you know, that was pretty exciting. People are very excited about that ecosystem. And so, you know, maybe that's something that can kind of kick off things again. I guess you never really know what the next catalyst is in the crypto markets. It always seems like it catches me by surprise. NFTs burst on the scene in a pretty meaningful way. Like Solana, you know, kind of seemed to come out of nowhere. There's so many narratives, right, that you could play as well. Like, you know, the portability

of money would be a great one. Like the, you know, wouldn't it be great if people were to sell all their assets and put into Bitcoin and move away from places like Ukraine? Like, that would help out Bitcoin. I mean, that story still could play out. Like, it hasn't yet. I don't know. There's, there's tons of different ways. I totally agree. I think it feels very similar to the other cycles that we've had. It feels like 2019. It's definitely not over. Like, you know, people are still building. That's fine. The prices are just lower for now. So I don't know what the impetus will be to kick it up. But it could happen at any moment. Well, you're absolutely right on the new funding. I just can't get over how many new early stage projects and companies are

getting funded every week. It's just it feels very different from previous cycles in the sense that the early stage market is just on fire still. What do you feel about valuations? I think there's still a little bit of a bid ask mismatch on the valuation front at the series A and beyond. I think the seeds and the precedes are still getting done pretty easily. But I think we're maybe have a little bit more bottoming to go on the valuation front. I don't know what you're saying. No, I think that makes sense. I would tend to agree with that. Well, Mike, this has been great. Where can we send people that want to learn more about Pith and potentially get involved, where can we send them? Yeah, send it at Pith.network, and then we're on all the socials under Pith.

Twitter is a big one for us. We've got a very active Discord as well. So any questions, reach out. Community managers are very active and engaged and would love to talk to you. Awesome. Well, thanks so much for joining us to podcast. Thanks a lot, Matt. Thanks for listening to another episode of On the Brink with Castle Island. To find out more about Castle Island, visit castle island.V.C. To listen to all of our podcast episodes, please go to On the Brink-Podcast.com or just click on the tab in our website. Thanks for listening.