Okay, we're rolling. Henry De Valence from Penumbra Labs. How are you? I'm great. It's great to be here and really excited to chat. awesome. Well, I've spoken with a number of privacy projects and Penumbra Labs is in that space and I wanted to begin with, before we begin our conversation, I wanted to read a post that you wrote recently. I guess this was back in August. I think this will set kind of the stage. What I like about this post that you wrote on X is that it's quite opinionated. And it shows a very strong point of view. And I think we need more of that in, in block, in, in crypto. You said, and I quote, Penumbra is the world's first useful private on-chain application. Everything else is either too small scope to be useful, Zcash, Monero, et cetera. Users cannot perform economically useful actions on their own or too large scope to be immediately practical, Alejo, Aztec protocol, et cetera. I'd love to hear you. Kind of explain what you're describing there because it really positions penumbra in an interesting space. Yeah, so it actually, sorry, it's just very mistimed with the water, but that... Yeah, so that really actually gets right to the core of what we set out to do with Penumbra. So if we back all the way up to say 2020, when the original DeFi Summer was happening, at the time I was working on Zcash and that was kind of the culmination of a kind cryptography engineering arc. where I got into doing cryptography research kind of post Snowden and worked on, you know, bigger and bigger stuff eventually culminating with Zcash at the time was like the most advanced cryptographic protocol ever made. And looking out at the DeFi summer, it's like, okay, for the first time people are really using blockchains. They're actually doing stuff on chain. There's on chain activity that's happening and that's really exciting. And None of that activity is happening on any of the kind of privacy chains that I'm spending my time and energy on building. And so that's kind of like, on the one hand, a bit of a crisis of confidence. Like, what are we doing here? But on the other hand, it's also an opportunity to reflect on why are things going the way they are? What's kind of the reason and the cause. And what that led me to... get to was this idea that we had got to the point where we had good enough cryptography to build privacy oriented blockchains, but we hadn't actually built something that was a kind of compelling product. You couldn't do these sort of end end flows. And one of the really cool things that was happening in DeFi at the time was that it kind of transitioned crypto from a mode where in order to use it, you had to go out and find some other person, right? Like if you want to make a Bitcoin transaction, you want to pay somebody with Bitcoin, you need to go out and find that person that you're going to pay. And with DeFi, the kind of innovation was, well, now we have these contracts and users can interact with those contracts and that can be valuable to them. And so that led to this thinking of how would we build private DeFi? but without going kind of all the way to let's reinvent the entire universe, right? Because when you look out at, and I think this was a somewhat contrary of you at the time, but I think it's definitely become consensus now. The reason that people do on-chain activity is not really about like the programming language or the smart contract system or whatever. It's about like what is the valuable kind of contract state? And that that comes as this kind of emergent thing It doesn't just come from like someone came up with like the best, you know fully general like the EVM was not like Perfectly designed. It's this like messy emergent thing. And so the idea of let's you know Rebuild the entire world from scratch and make it private which is sort of the mode of like Aztec or Alio, like I hope that that works as a privacy enthusiast, but it seemed like kind of too big of a thing. And so that thinking led to, what is the narrower thing? What's the minimum, the simplest possible transaction that I could make as a user that is actually gonna be beneficial to me, that'll have some economic reason to do? And that led to thinking about token swaps. And that had another really nice intersection, which is a second sort of line of thought of how do we break privacy out of this mode where private X means like, it's like X, but it's worse, but it's private. So you should use it because don't you care about privacy? Cause this is not like a compelling pitch. Instead, you have to find a use case where privacy actually makes it a better product. And that those two kind of trains of thought, one is like, how can you have a kind of useful single player mode? And also what is a case where privacy is providing product value, intersected with, let's try to build a private decks, and then let's use the lessons that we learn from the system design to figure out what should be a design of the kind of next gen like fully private on-chain future look like once we've got one thing that works. And so that's kind of, yeah, the big overarching arc of how we got to building a private dex, why we thought that was useful. And it turned out to be quite complicated. So we spent about three years doing protocol design, designing the software stack, the client stack. and went to main net last summer. And since then we've been continuing to refine, iterate and keep going. That's great. Help me understand, I, from my understanding, based on, you know, the little research I've done, penumbra is both a layer one as well as infrastructure. Is that correct? Yeah, I mean, so the term infrastructure is kinda nebulous, and a lot of people use it in different ways, but... I guess what I mean is like, it's like a chain agnostic infrastructure that could be used by apps and other chains. Is that, is that true or false? could be used by users of other chains, but less so by apps. kind of, the, the way that I would describe it is Penumbra is sort of a interchange privacy layer. Users, has a multi-asset shielded pool that can record any type of cross-chain asset. And it has native interop with other chains via IBC. So you can have whatever asset you want, record it privately in Penumbra. And then you can either, do trading within Penumbra or you can take those tokens, withdraw them to some account on some, some connected chain. That part, you know, is unlinked from all of your past activity and you can do some action and then kind of come back to Penumbra when you're done. so that there's sort of this near term product of private trading, and then there's a long-term development of private chain abstraction. but it's a kind of chain abstraction that really puts the user first. Got it. Now privacy is a large space. I've spoken with number of privacy projects. met with Oasis, Nillion recently, and a couple of other ones. And so there's privacy, private transactions, private compute, private from a regulatory perspective. Here we're talking about private transactions and the first application of which will be a private DEX. Is that correct? Got it. Okay. That's really, really cool. Well, you mentioned you launched the decks last year. Yeah, so the protocol itself went live last summer and it has sort of a natively integrated decks, shielded pool, staking system, intratransfers, et cetera. And since then, we've been working on building out a lot of the kind of client side parts of that stack to sort of improve usability and get kind of the like end-to-end product flows working. We can talk about it a little bit later on in the episode, I'm sure, but one of the interesting things that happens with privacy is that you really have to rethink the kind of whole data architecture of how blockchains work. And that means that sort of through the whole software stack, a lot of things need to be rethought. So yeah, we can get to that a bit later, but... really sort of the kind of arc is we designed and built the whole protocol and we did that in a try to be like sort of maximally decentralized from a sort of principles based approach. So it's not like, oh, it's launched with the training wheels, whatever it's like, it's out there, it's live. And the main focus right now is on user experience, smoothing out all of the sort of little needless frictions. and kind of taking all the powerful capabilities that are already there in the live protocol, but actually making them usable and useful to people. Makes sense. The initial target audience that you have for Penumbra, how would you describe that audience and the type of messaging that you are presenting to them? Yeah, so I'd say there's two different types of users who could really get something out of Penumbra today. So one is, if you just think of normal, everyday crypto traders who are doing stuff on chain, Sorry. it's all good. if we think about like normal, everyday users who are doing stuff on chain, what we want to present those users with is something that looks and feels, like a centralized exchange or a Dex with a great user experience. but actually behind the scenes it's private and it respects their autonomy. And the strategy there is kind of like with signal, right? So. one of the reasons that signal was successful versus other, encrypted messaging apps that came before was they did a lot of work to get all of the cryptography right. And to kind of line up the cryptography with the user experience so that, you know, if you're using it on, you know, a mobile network, that's kind of flaky and the messages go out of the order, like whatever, like the cryptography all still works. But as a user, you don't have to think about that. You don't have to know about that. It's that there's a sufficiently high kind of alignment between the technical properties of the system and like what your imagination of what it does is that you can just be like, it's like a private iMessage. And that's really the goal for that class of users that we're trying to achieve is looks and feels like a centralized exchange. The user doesn't have to care about any of the details of how the privacy is accomplished. they just have to know, okay, I could use, you know, this thing, or I could use Penumbra and they're kind of roughly similar in terms of the user experience. But with Penumbra, know that my data is safe. so that's, that's kind of one side. And then the other side is for, LPs more active traders. the Penumbra debt, the decks design that we built into Penumbra is really, really cool. And it is, think, perhaps like the most powerful, like spot dex engine that exists anywhere, like regardless of privacy. And part of that was thinking, well, it's not enough to just like be a private X. We have to be better. And what we did is we basically tried to holistically align the privacy properties of the system, the consensus properties of the system and the defy properties of the system. all into one thing. So with Penumbra, there's very little MEV compared to a traditional DEX because we're able to build the whole trading mechanism together with the consensus system. So on Penumbra, there's no transaction ordering. Everybody just like submits effectively on chain intent for this is the trade I want to do and built into the node software. There's actually this like parallel solver system that does like graph traversals of all the liquidity and optimally routes trades over different pair, different routes, different pairs, different fee tiers. And so there's really interesting games as an LP, like a active market maker that you could play where if you're smart about how you're providing liquidity, you can get a lot more volume. and a lot more fees, then you would be able to do by just like passive LP. And then for the kind of retail traders, that's actually better because they're getting better execution. That's interesting. I remember back in DeFi summer, there were conversations around dark pools, which was at the time that I was like, what if you could actually trade on a DEX and no one knows what you're doing and these bots aren't able to sandwich any of your transactions? And it sounds like you guys have achieved that. How is that going so far with users, how are users enjoying that? So it's been interesting. So one of the things, and I'll get to this in a bit, I think when we talk about kind of the mechanics of how do you build a private chain, when the protocol went live, there wasn't actually any like gooey front end for the decks. So like exists in the protocol, but because there's not really a front end for it, that experience of like feels like a centralized exchange. but is actually a dex, right? That didn't exist when the protocol went live. It's just like, here's the decentralized chain. And so in the meantime, what we've actually seen is sort of this like really interesting experiment where the dex is live, but there's very little liquidity because people aren't really, like there's not really the kind of tooling to interact with it. But. you can see really interesting sort of emergent behaviors that I think are, that show how it's going to be really cool once the, the DEX front end goes live in a few weeks. We saw, for instance, someone made a stable swap liquidity position between USDC and USDY, which is like a yield bearing stable. And They were actually able to get roughly $150 in fees on a $100 liquidity position and run like $50,000 of volume. And the reason is because like that position kind of happened to be placed kind of as a bridge between different parts of the liquidity graph where like there's some assets over here, there's some assets over there. And this sort of stable coin, stable swap allows liquidity on USDC pairs to serve USDY trades or USDY pair liquidity to serve USDC trades. And so we've learned a lot of really interesting lessons about kind of the market structure that make me feel very optimistic about the kind of like launch of the DEX as like a product and not just as sort of a protocol capability. Now the liquidity provider, do you think that was accidental or purposeful? they absolutely like, I'm sure had some strategy, but the sort of cool thing about Penumbra is when you go and look at it on the Dex Explorer, all you see is like, here is a concentrated liquidity position. Was this part of somebody's like bigger strategy? I don't know. Did they have other positions? I don't know, because all you can see on Penumbra is like someone made this like position in the order book and you know, who are like, how is that connected to their other activity, know, overall, you don't know. All you can see is like, here's the aggregate market structure. And actually that is kind of an answer to a question you asked earlier about dark pools and sort of what's the privacy model for the decks. For Penumbra, the model is we wanna have transparency for aggregates, but privacy for individuals. Right? So one question that you might have for privacy projects is like, well, like I thought that transparency was one of the whole like core values of blockchains. like, isn't that what we're here for? And the answer that we have to this question is like, you get this like best of both worlds situation, right? So with Penumbra, you get to keep all of the verifiability properties. So everybody gets to see here's the aggregate state of the market. Here's what all the liquidity is. Here's where it is. Here's like the total kind of volume that's flowing between different pairs. I get to see sort of the tickers of all the different assets and how they're moving around. But what I don't get to see is like your specific trade history. I don't get to like log onto your account and see your P &L because that's your private information. All I can see is kind of the aggregate of the system and your specific actions are known just to you and to whoever you choose to do that disclosure to. That takes us more to, you know, one criticism that traditional finance folks have made against, you know, decentralized finance is that, yeah, that's great, everything's on chain, but now I can see what everyone else is doing. You know, what about, you know, protection, privacy protections for the retail customer? And it sounds like Penumbra takes us closer to that. Yeah, exactly. I would say it's like, you know, the way that in some ideal world, right? This is the way that blockchain should have been designed from the first like in the first place. But they weren't just because the ZK technology didn't exist. But now that it does, we built it. is it differential in the fact that, or in the sense that I can choose what others can see or is it just private by default? So Penumbra is private by default and all transactions on Penumbra are shielded. However, Penumbra also has disclosure capabilities, right? So if you want the basically the way that you would choose to make something public, if you wanted to do that is you do it on chain privately and then you can publish the disclosure of here's the information. But more likely you would probably just want to disclose to specific parties, Like, you know, I'll give my viewing key to my accountant and then they can like scan through all my activity, compute my P &Ls and all the tax obligations and so on. And then I get my report. Another use case for that actually is for front ends. We wanted to sort of... build a similar kind of connect to wallet type of experience, but for private apps. And in order to do that, we actually had to build a system to allow your wallet to kind of selectively disclose activity to the front end that you're working with. So this idea of selective disclosure isn't just like, abstractly, we could do this at the protocol, but it's like literally how all the apps work. You mentioned viewing key. So it sounds like there's different levels of permission in a transaction. Can you describe what those are? Yeah, so what I'm gonna do actually is just pause and then back up and I'm gonna give a kind high level explainer of like, how do shielded transactions work differently than a traditional blockchain? So if you think of something like Ethereum, fundamentally, the model for data on Ethereum or some other transparent chain is global mutable state. So like, here's the chain state, it has all the contracts, all the accounts, and I'll make my transaction. And then the node is going to execute it and it's going to change the state somehow. And then everybody has the new state and then, you know, you do a next transaction and so on and so forth. So, right. So each transaction goes and it can like edit the state of the chain, which is accessible to everybody. For a shielded chain, you have to structure the data very differently. And that has a bunch of implications that I'll get walked through. And the fundamental change is rather than one big sort of global mutable state, you have all these little immutable composable state fragments. So if people are familiar with Bitcoin, they could think of UTXOs, but that term has like a lot of Bitcoin specific baggage. Just think of it as like, here's like a little fragment of state. Like here is, you know, 0.1 ETH that I control. That's some piece of state. It's going to be linked with my account, but that can exist kind of independently. And you want to kind of fragment all of the state as finely as possible. Why do you want to do that? So that you can take all of the pieces of state and move them off chain onto the end user device. So we're building end to end encryption for blockchains. And the way it's going to work is I'm going to track all of my state on my device. And then when I do an update or a change, I'll make a proof that I did it honestly and send that to the chain. And now everybody else can verify like, this update was legit, but they don't get to see these specific pieces of data that are relevant to me, right? So I'm tracking my account locally on my device. You're tracking yours locally on yours. And if I send you funds, I'm basically going to send a message that only you are able to read. And so it's kind of this really interesting, I gave a talk about this at modular summit, I think two years ago, it was called like shielded transactions or rollups. And a really interesting way to think about this, I think is like in a shielded chain, every user, like every transaction that they make, is like their own like little micro roll-up. And what this means for the overall system is that you have to be a lot more kind of client side focused, right? So if your wallet is tracking all of your information, that actually has to be on your device. You can't just like go and query some RPC and say, hey, what's my account balance? Because It's a private chain. The RPC isn't supposed to know that, but that means that in order to have a good user experience, you have to build the whole protocol so that every wallet can have its own little like ultra light node that maintains all your data locally and is so efficient that you don't even notice that that's happening. it also means that when you're building, you know, a front end, the data about what the user's on-chain activity is, isn't going to be coming from some server. It's going to be coming locally from that user's wallet. So you have to build a way to make front ends that can interact with this private data. And now kind of coming full circle to your question about like different levels of visibility, how does that work? The way that it works is every piece of information inside of a penumbra transaction Um, it's kind of individually encrypted with some kind of different piece of, of key material. And when your, uh, front end says to the wallet, like, Hey, what's this transaction? The, your wallet that's running locally on your computer is going to decrypt whatever parts of the transaction it's able to see. And then kind of hand a pre-decrypted view to the front end. And that means that, um, the front end is compromised or hacked, like it can't leak any kind of long-term information. It can at best like see stuff that it already had access to. It also makes it lot simpler for devs. Cause like if you are trying to write a front end like, and the first step is like, Hey, here's all this like shielded photography. That just sucks. But basically, different pieces of the transaction are going to be visible to different parties. So for example, if I send you funds on a traditional blockchain, like if I sent you a hundred USDC on Ethereum, right? The information you get is, okay, there's a TX hash. I can look at the transaction. I can see, you know, my account and you know, all my data is public and I can see their account. And then I can go and look at like their whole history of every other transaction they've ever made. So if you're trying to like, you know, pay a contractor, like, oops, you're actually like leaking all of the information about like your contracting rates on chain to the entire world. on Penumbra, the model is, you know, if we take that same scenario, I send a hundred USDC to you. you're going to see a transaction. get detected by your wallet, so you'll see the transaction. And then when you open it up, what you're gonna see is first a memo that has my return address and some texts that describes what the transaction is for. You have to kind of add that in because otherwise it's pretty hard to keep track. And you're also gonna see an output of 100 USDC that goes to your address. And that's it. So you don't get to see what my account state is. you don't get to see every other transaction I've made. it's actually basically the same model as if we were doing like a TradFi like bank transfer. Like if I send a hundred dollars to your bank account from my bank account, I don't get to see like all of your transactions ever. You just get to see or, And you don't get to see mine. Like we just get to see, here's what the transfer was. And that's exactly the way that Penumbra works. It's actually in some ways, although people think of privacy as being, here's this like, you know, totally radical, on-chain thing in a lot of ways, the mission for on-chain privacy is just how do we take the existing visibility models of TradFi that work and just put them on chain. Like this isn't some like, in some sense, it's like not even that weird. It's like, yeah, that's just how a bank transfer should work. And I think that kind of positioning, we're already familiar with that, right? Everyone that has a bank account, we're familiar with that world and how that works. And I think positioning it in that familiar way, it takes kind of the mysticism out of privacy and also the of the ideological baggage that I think sometimes comes with privacy. You mentioned that, so the decks, you're... You're planning on the front end will launch in a month, you said or so next few weeks. alpha version. If people want to check it out, they can go on the Penumbra Discord and there's an alpha version. you know, it still has some, some rough edges, but, it should go live, imminently or it's technically live. just are sort of surfacing it to the core community, first. so if you're excited, go join the Penumbra Discord, come check it out and tell us everything that sucks about it. you know, feedback like that is actually always great. Like you learn so much from negative feedback. And, or, you know, also the things that are great. We love to hear that too. But there's sort of the kind of end-to-end product experience of not just like, this Dex engine theoretically exists as part of the protocol, but... There's a whole UI for it. It looks and feels like a centralized exchange. And then the second part of that, that is gonna hopefully arrive a bit later is a proposal that we actually just put forward today for an on-chain liquidity tournament that allows liquidity providers to basically compete to provide liquidity on Penumbra at the best prices. Let's talk about the product launch. There's a lot of things that go into a product launch that's important. Tell us about the planning that is going on behind the scenes and in marketing a DEX when there's already many, many DEXs available, limited liquidity or finite liquidity and competition for that liquidity and How are you marketing to liquidity providers who are important and also traders? Tell us about the work behind the scenes of what you guys are doing. Yeah, so I think there's a few points there. So one is about the main behind the scenes thing is just like really thinking through the sort of end to end product experience. Like how do we basically make like the signal of DEXIS? And how do we make sure that every piece of friction along the user journey for someone to test out this thing is reduced as much as possible. So that most of those kind of core sort of technological pieces are already there at the protocol level. It's really just a matter of making sure that every piece of that user journey is kind of lined up. The second point about, you know, the sort of liquidity levels and so on. This is an example of where the mechanism design and the capital efficiency of the Penumbra decks, I think are really gonna play an important role. So the decks design is built around kind of all concentrated liquidity all the time. And it also is able to, because it's doing this sort of more sophisticated solving and routing, it's able to synthesize liquidity from different parts of the system. So in other DEXs, you have this setup where there's like these pools, right? Like, Hey, there's this liquidity pool, people can like deposit into the pool. And then once they've deposited liquidity into that pool, it's just like passively sitting there, right? And it's not really able to do much elsewhere on Penumbra. the overall DEX design is a kind of very different strategy. It's like, instead of trying to say, do you, it's almost like the answer to how do you stop liquidity fragmentation is you fragment the liquidity as much as possible. And that's maybe counterintuitive, but on Penumbra, every single liquidity position is its own little micro AMM. And the overall DEX kind of, indexes and arranges all of these individual positions so that it can route through any of them. And what that means is instead of having liquidity that's limited to here's a specific pool and now it's just useful for that and if you want to do something else then you know too bad, all of the different liquidity positions on Penumbra are basically synthesized into one kind of virtual giant pool for the whole system. And on the Dex interface, one of the things that's really cool is when you look at, you know, where on a traditional say centralized exchange interface, you'd have like, here's the order book on penumbra. We call it a route book because rather than just showing here's like specific, you know, orders on this specific pair. It's if you were going to trade, what would be the optimal routes? And so you can get much greater effective liquidity for any particular pair by routing through all of these other positions that might be on other pairs. And you don't have, for instance, on other DEXs as a liquidity provider, you kind of have to pick like a fee tier, right? Like, deposit into this pool with the... 30 bips fee or maybe I should do the 100 bips or whatever. On Penumbra, there's no fee tiers. It's just every position sets its own fee and that kind of gets wrapped into the price to see where things should get routed. So the overall strategy there is for users, this mechanism design, they don't have to know about it. but it operates behind the scenes and allows the decks to much, much more efficiently use the capital that is there. So it might be that, you know, right at the beginning of the launch, if you wanted to do, you know, some huge trade, like maybe, you know, you want to split it up into pieces or chunks, but for small trades, you can actually do very well. like the capital efficiency really shines through. And on the other side for liquidity providers, there's a lot of, like if I would say that like passive LPEing for Penumbra is like probably not as good of a strategy because there's this ongoing trend of actively managed liquidity is gonna outcompete passive LPs. So we did sort of skate to where the puck is going on that. And we built a system that if you're kind of smart about running bots or providing liquidity in some programmatic way, you can get a lot more order flow to your positions because the DEX is actually routing across all these different pairs. So that example of that, like, you know, one position that I mentioned where someone had just put in you know, even like a hundred dollars to start and run 50 K in volume. If you go on the decks, explore, and you look at that position, you can see a chart of like, where did all that order flow come from? And it comes from all over the decks. So as an LP, there's this different game of it's not just about like what pairs do I think people are trading, but it's like, what is the optimal place for my liquidity to be in this overall liquidity graph? in order to get the most order flow fees, et cetera. So it's a really interesting kind of challenge. And to bootstrap, like once you've got that kind of problem framing, you could write some software that kind of programmatically figures this out. Like what's the incentive for someone to do that? That's where this idea of this liquidity tournament comes in. Basically, we have a proposal for how the chain can, you know, every epoch, so every day or two, basically issue rewards to the liquidity providers that were the best at actually providing best priced liquidity based on seeing like what actual volume did those positions get executed. And I think that's gonna help kind of solve the cold start problem of how do you bootstrap activity. But overall, I would say one of the things that I feel pretty strongly about for Penumbra is optimizing for decentralization and resilience. This happened with the launch of the chain where when the protocol kind of turned on, there's not like, here's this official RPC or whatever. It's kind of this like, okay, we as collectively as a decentralized network have to figure out, you know, how's, how this is all happening. And I think one of the problems that, protocols can run into is when there's reliance on particular centralized entities, then if that entity kind of goes away suddenly, then the whole thing just kind of falls apart. and so for Canumbra, or a kind of overall general approach is like, let's try to create an incentive mechanism that makes it so that there's like some incentive to start participating, you know, as the first few users, but where there's always kind of multiple participants in every different role of the protocol. And so I'd rather have like a slightly slower kind of growth period of different entities kind of coming into the system and starting to use it, then saying like, there's gonna be like this one entity that will like do all the market making, you know, but it's still a decentralized project. And then like, you know, basically they kind of have this like big, you know, power influence. What we're trying to do is basically create an environment where different participants can kind of start. interacting with the system and allow it kind of emergent ecosystem rather than just sort of one centrally planned thing. In the cold start problem almost always evolved some kind of mercenary activity from liquidity providers, just chasing either yield and or some kind of airdrop. How are you viewing that kind of, I guess how do you view that type of participation since it's almost a necessary evil in at least the current deck structure? Yeah, so the way I think about it is it's basically kind of a mechanism design problem. And in some sense, that's sort of all of what we're doing with blockchains, right? What you want to do is create a mechanism where even if someone is trying to subvert the system, like their attempt at subversion ends up getting kind of like jujitsu'd into some behavior that the system actually wants. And so an example of this is in this liquidity tournament mechanism that we just put up on the penumbra forum, the way that the chain would decide which positions are actually providing useful liquidity to the chain is by looking at like, okay, what positions did the most volume. And if you think of, you know, a different type of decks design, there's a really obvious kind of attack, which is like, well, someone could just like, you know, try to like, targeted trade at their own position, and then, you know, pump up their numbers or whatever. But on Penumbra, because we have this kind of holistic decks design where the decks runs kind of in a batch at the end of every block, and it does all the routing on its own, as a, uh, as someone trying to sort of game the system in that way, the only way that you can guarantee that your, your liquidity position will get executed against is if it is actually the like overall best priced liquidity and you're actually offering that to the entire like set of user base. So There's no way to like try to fake out the system without actually providing the liquidity to everyone. And, and so I think that I wouldn't think of it as unnecessary evil. I think I would have maybe a. The way that I would frame it is like, first of all, it's definitely a challenge to figure out like, what is the incentive mechanism that is kind of compatible with the system's goals? And that doesn't have these kinds of spiraling takeoff properties, right? Another sort of aspect of that is like, you'll see things where there's like inflationary rewards, right? And you can see these like incredible inflationary death spirals where it's like, oops, we got the parameter wrong. And then the compounding inflation just kind of took off and now there's like 11,000 bajillion tokens and the whole thing is totally screwed. Right. Well, if you want to avoid that, you can just say like, okay, there's a fixed reward pool and people can kind of PVP each other sort of within that pool. And that also means that you get this dynamic where if nobody's playing the game, there's a really strong incentive for the first person to play. And then there's a slightly less strong incentive for the second person and it like gradually declines, which is exactly the kind of graph that you want for this cold start problem. right. So there's a lot of mechanism design subtlety. That's like, how do you ensure that the incentive mechanism is actually good? And then the second part is making sure that you're iterating on the product before you have the, giant incentive machine running. And the analogy that I would use like for that is, you know, if you imagine having like a campfire or a bonfire and you have this like can of gasoline, right? Like the incentives are the can of gasoline. And if you use them well, like you can really get things going. But if you like accidentally, like, if you're not really looking at what you're doing, the fact of having these incentives means it's kind of blinding you to does the product work? Cause you can't really tell. Like are people using this just because they're trying to make money in this mercenary way, or is this actually, you know, providing something useful? and so you, there's like a risk of like, oops, I like poured gasoline all over myself and I like set myself on fire and oops. and so I think there's, there's another challenge there where it's like, you need to do some product iteration and make sure that it like actually works prior to. kind of turning on an incentive structure. And I think that's sort of an opposite meta to what people commonly do. I think a lot of the time there are projects that have a very short term set kind of thinking where it's like, we want to, you know, pump a token as high as possible because that's all that matters, right? And that leads to these kinds of dynamics where people are like, there's like incentives and just like the incentives make the number go up and then invariably it kind of collapses and explodes. And then, you know, it's kind of like, well, what did we actually build here? And it's, I think a much better approach to really refine the product first, make sure it works and then have a kind of onboarding, structure that that works once there's some way to get signal about like, is this good or is this working? And I think that's how you have a much more durable environment. It sounds like you and the team have taken a really thoughtful approach to building the Penumbra protocol in the dex. And it reminds me a lot of Hyperliquid and how they've built their perps dex. It's just very thoughtful and very user-centered and doesn't have kind of like a runaway train on the incentives. How did you... become that way where the focus is not on short-term thinking, but long-term thinking, the user, and making sure that the product works. And getting mechanism design right is so hard to do. And you sometimes don't know if things are actually right until you launch. how, I guess, yeah, how did you get that way and what led you to think and build so thoughtfully? Yeah, I think, I mean, a big part of it actually is that I, I feel like I kind of walked backwards into DeFi and into even like blockchains generally. Right. So I originally got into cryptography because of like seeing, Hey, there's like all this Snowden stuff happening and there's like no effective political response to it. And so like, You know, we've got to build this like cypher punk stuff, right? And, I didn't actually, you know, I, I don't think I ever thought of like, I'm going to, you know, be a founder or like, I'm going to start a company or a blockchain project or something. It's almost like I had been working on all these things and I could kind of see all these different pieces, coming together. And a phrasing that I've used is it's like, I was like cursed with this like desire to eat a blockchain out into the world. And I think that, you know, there's ups and downs to that, but definitely, you know, I just, I don't have like, The original reason to build Penumbra was really this question of, I want to have a world that has more individual autonomy than it currently does. And I have a theory that if you have technologies that increase individual autonomy, then that kind of shifts the fabric of society. in the direction that you want. And so how do you build some kind of self-sustaining technical infrastructure that will kind of help other people to live with the values you think are important? And in some ways, you know, building a decks, building all this stuff, like, it's like this. very kind of long term project already of like, how do you reshape the kind of second, third order incentives that are determining how things are kind of downstream. And I'm very excited about what we built and about the kind of continued evolution of making that like a really usable product. Cause I think that's how we get as an ecosystem kind of escape velocity for these sorts of like freedom properties of the systems that we've built. Like I think we're definitely at a point where crypto is like mainstream or like going mainstream. And there's a question of like, how much of the original cypher punk values are we going to keep? And so I feel motivation to try and, know, It's not just like, let's make a useful product because it's going to be useful for like these people. also the second third order effects of that are like, kind of entrenches these values and principles into the world. and that I think is the, the big long-term motivator. And, although I'm like very excited about, you know, how that's I'm like really, really excited about some of the stuff that we have in the pipeline for the first half of this year. But we've also kind of already shifted like the. You know, where the center of gravity is right. And like, I have seen other projects sort of like. Look at stuff that we're doing and then like start doing other stuff. That's kind of similar. And, you know, I think ours is like a little bit better put together, but even in that context, it's like, you know, there's like both winning as a product and there's also winning through the values diffusion. And I think that mindset really makes it. a lot easier to play long-term positive sum games. The space as it is today has a lot of attention paid to people who are playing short-term zero sum games. And I think that's kind of a shame. But I think that rather than just spending all of your time in some frame of like, I'm call out all these like bad things. there's more people who are doing bullshit than you could ever possibly hope to, you know, call out. It's like this ethos of like focus on what you want to see more of, focus on what are your values and what is your plan to kind of bake your values into some kind of self-sustaining technical infrastructure. I think you have that rare combination of ideology and pragmatism. The ideology comes from the cyberpunk values, but seeing those values implemented in a product where you get to live those values every day as you use the product is pretty unique. And I think we lack a lot of that. We used to have a lot of that in Web3. but have lost a lot of it, think, or at least some of it anyway, with a lot of the player versus player meme coin casino. And so it's refreshing to hear you say that, in that you're building something that kind of takes us back to what's important. Yeah. I mean, I guess it's, I keep coming back to this like frame, you know, like, what are we doing here? Like, let's, let's focus on what we want to see more of. And also, I think a question, a good corollary to what are we doing here is how did we get here? And it sounds like asking that question kind of led you to the design decisions that led you to build Penumbra, which is really, really unique. Now, looking forward to 2025, Q1, Q2, what are some... some things that you could share with the audience on what penumbra is up to and how they can get involved. Yeah, so the main development focus for 2025 is just product experience, user experience. An analogy that I have is if you imagine a skyscraper that's been under construction and they've finished the concrete core and built all the floors and everything, and now it's time to actually install all the furnishings so that people can live in it, that's the stage that the protocol is in. So right now, the big focus is on the Dex product experience. And that has two parts. There's first of all, this alpha front end interface that shows kind of all the power of the Dex. People can check that out in the Discord. Hopefully we'll have this kind of liquidity tournament to incentivize participation and get some development for the liquidity provision side. So testing out that's, if you're, if you'd like to test out things that are cool, still have some rough edges, definitely hop in the discord and ask around. And later on in the year, so we have plans to have a mobile app that has some really cool features that are kind of uniquely enabled by the protocol. So I don't want to say too much before it's ready. I think that's also been a big thing for Penumbra is under promise over deliver. And I think that's going to be a really cool way to show off that it's like, although we've built the protocol around the decks as this kind of initial use case. It also is a great rail just for doing private payments. And I think a lot of pieces in the ecosystem are kind of coming into place to make that really interesting, right? Private on chain payments. Like what I was saying earlier about you can just like do a payment and it has the same properties as a bank transfer. That's pretty cool. but also kind of showing that it's like the DEX is this initial kind of beachhead use case. And we're gonna transition into Penumbra as an interchain privacy layer, right? So if you wanna go do some action on some other chain, you can use your shielded funds on Penumbra. You can use the fact that there is the DEX, To like, you know, I did some NFT mint claim. on, on Celestia and I didn't already happen to have shielded TIA to pay gas, but that's not a problem because I just, you know, do the swap on Penumbra to get the token I want and then I withdraw it and I'm completely anonymous. we'll see a lot of the pieces kind of start fitting together and have like a really cohesive experience of like, I use Penumbra. Penumbra is my shielded home base. It's like. where I keep my, my funds, my activity private. And then when I want to go out and interact with kind of the rest of the world, like when I go out of my apartment, I leave, I go do stuff, I come back. Like the inside of my apartment is, you know, my private space and I can go in, but I'm still connected to this wider world. And that's really the vision that we're trying to, work toward and have a really seamless experience. Awesome. Well, Henry, this has been a really enlightening conversation and it sounds like you guys are doing some really unique things and we'll share all of the links in the episode notes. Any final words you'd like to leave the audience with before we get off? I don't know. Stay shielded. Yeah. that. Are you going to brand that as part of the branding for Penumbra? Stay Shielded. I dig it. yeah, it was a lot of fun. Thanks for giving me the opportunity to chat and asking a lot of great questions. Awesome. Thank you, Henry. All right.