Speaker 3
Guys, one other question I wanted to ask you, which I was one of our audience questions from Twitter is basically, let's assume everyone, all the best scalability solutions for defense are on this call. But out of the ones that aren't on this call right now, the ones that I think have seen the most differentiator or unique to me have been some of the deep and specific layer ones that are launching. I mean, outside of people on Transylana, which you already talked about and folks on China there's IoTeX, there's peak. And from what I, if I had to articulate the thesis behind why a deep and specific layer one should exist, it's that this doesn't happen today, but eventually we'll see deep ends actually accessing a lot of each other's state. Right now it makes, they don't type into each other's, but we see little signs of it with, I just saw like, I think demo maps like 15% of the helium-loral hotspots through a new macaroon device that they launched, a new OBD reader. And so I think more and more over time, just from a high level conceptual, I think, if DY is going to make telecom 20% better and decentralized energy is going to make energy 20% better and transportation is going to make that 20% better. Like when you start to compound these things on each other by bringing data up and down the stack, that you truly build something that traditional sort of like infrastructure writers just can't, you know, you can't compete with like three sequential 20% improvements compound, but on each other's just not possible. And so I'm curious, like, does that resonate with you guys at all? Do you see deep ends eventually touching each other's on chain, like interacting with each other based on network activity that's happening on chain? And how do you do need to build for that? Yeah, I guess I'll leave it there to hear what you guys, if you guys have any thoughts on it.
Speaker 1
Yeah, I think a composability is a critical property for crypto dapps. And we're starting to see an early stages with deep end apps, but I'd still argue that you're better off in an overall share chain because then even non-deepened apps can compose on you. You can have additional financialization through DeFi apps composing. You can have NFTs that get traded in marketplaces or liquid secondary markets. So it almost seems like they're caught in this awkward place where they're not quite getting the same benefits of having your own chain and being willing to pay that cost and interrupt and just actual money. But then they also are not on a fully shared chain, so they can't fully compose with everyone either. So I don't know if the deep end specific L1s make more sense than just a general purpose L1. And I think even in the Cosmos ecosystem, people often start with app chains or maybe even sector specific chains and they very quickly expand to basically accommodate everything because it's one of the most logical ways to expand. And there's typically some dapps that actually do want to compose outside of the specific category that the majority of apps exist in.
Speaker 2
Yeah, I don't have too much to add there. I think that you also, to some extent, a lot of the deep end projects we're talking to at least, they definitely don't conceive of themselves the same way like an NFT project or a DEX conceives themselves in terms of their position in the crypto space. But a lot of the times they do want to access some part of the crypto community. Maybe that's sort of like degenerators, general liquidity on an Ethereum or other chain. Maybe just kind of, yeah, even just like the ability to maybe make competitive NFTs and have people in the crypto community care. These are all things that I think are a lot easier if you're within some established ecosystem, right? Ethereum or Solana. Then if you're like trying to build your own kind of siloed chain somewhere else that only, you know, few people in this and a relatively niche category of the industry still are interested in. Like I'll say, truthfully, I think like the idea of sector specific chains is really, really, really interesting. And so I'm a little disappointed at myself that I'm not familiar with a lot of the offerings that are happening in the deep end world for this. But I wasn't familiar with the chains you mentioned. I think a lot of the folks, the crypto community, the ones who are going to probably be the early adopters for some of these protocols aren't super familiar with those chains either. And so just in terms of like access to liquidity, go to market, being able to kind of have some share and like the mind share of the crypto industry, it seems to make sense to try to build on a more established ecosystem.
Speaker 3
Yeah, that makes sense. One more sort of related question. This one for Unile on since I know where Matt stands on it. You guys have made a really big bet on SVM. There are people building other types of virtual machines too, like newer, newer, newer next gen virtual machines that claim to have certain benefits over SVM or over the state of the arts today. Curious if you, how you feel about some of these ecosystems and you think that's the truly the direction people are moving or if you think like SVM is it.
Speaker 1
Couple of thoughts on it. So my first is that Solana from the ground up was literally built for performance, whether it's the Solana virtual machine with sea level or whether it's the fact they removed the mempool, the state accesses are specified up front. So the transactions are essentially optimized for parallelism. So anything that's built from day one now is likely going to land in a place that's very similar and you could make design decisions regarding do you specify state accesses up front or to compute optimistically and then rerun if there's a conflict and there's all sorts of different design decisions. But ultimately this is one point on the efficient frontier. So the second part is about developer experience and overall network effects. If I wanted to do like a matrix inversion and move or in field VM, I don't even know if it's possible right now because there just isn't that surrounding tooling that exists in the same way as what exists for rust or for the SVM more broadly. So like Matt was saying with the EVM, it took years versus the Solana virtual machine to get to this point where it actually made sense for it to expand and where they've actually reached that critical mass of adoption. So there's a tough battle ahead for any kinds of new VMs that are popping up. So those are the main points of resonate with me. But the last one is that even if there's an innovation that occurs in a layer one blockchain, I'd argue that it's fundamentally easier of a problem to solve like making a highly performant layer one blockchain as opposed to making a highly performant layer two. And the reason for that is you might make optimizations such as with Solana did, they removed the Merkle tree and we actually retain that optimization. But what that means is that settlement or doing those proof based mechanisms that I described earlier becomes much more difficult and sometimes impossible. So the Merkle tree is the that's usually how a state commitment is made for EVM chains. And that's important because when you do a fraud proof or a ZK proof, you have to prove one, here's the inputs to the transaction and two, here's the output if you run the transaction given those inputs. So that first part of what I'm describing, proving the inputs becomes impossible or it becomes much more difficult if you lack any sort of state commitment such as a Merkle tree commitment. So removing that is great for performance and layer one blockchains do that. But the downside is your settlement is almost impossible. So you have to get very creative around how you
Speaker 3
do that. Yeah, so close it out guys. Just one last question when asked a lot of the one of the things that I found in D-PAN is that a lot of the founders have a lot of experience and they're usually like older than crypto founders. You guys, I've had definitely my fair share of experience trying to convince founders to like take money from someone much younger than them. You guys are doing something even harder, which is like trying to convince them to build their company on top of a blockchain infrastructure from, you know, to really like couple themselves to an infrastructure that's a big it's a big mode of confidence I think for founders to trust you guys. But just curious like what do you guys what would you guys say to a D-PAN founder who's on defense about jumping into crypto at all, you know, they may not even be certain whether this all this pain of dealing with infrastructure and dealing with regulators and dealing with liquid prices and where fit at this point. Or whether they should just wait a couple more years for things to like sort of settle down before launching their project. Curious how you guys feel about the state of the waters for D-PAN founders to jump in.
Speaker 1
So one way that I think about it is we stand on the shoulders of giants. So I'm sure Matt has some great other folks who have his team plus advisors. On my side, our chief business officer Vijay has been in crypto for several years. He's seen it all was previously a head of BDA Uniswap, head of BDA DYDX and our head of engineering David Lynn was previously building other L1s. So any question that D-PAN founders have about crypto, we can likely answer it to sufficient depth and talk to great length about it. So that's one part of it. And the second is having a younger perspective also means that we can often approach things from a fresh lens. And especially giving crypto is at this great intersection of computer science, ECON, there's also a regulatory element to it and so many other things. So given the D-PAN founders typically come from the physical side, I think it's a great opportunity for us to add value and help them understand like what are the unique parts of crypto that are relevant for them to be thinking about as they further build out their product. So that's generally how I think about it, but y'all I'll hand it off to Matt in case anything to add.
Speaker 2
Yeah, yeah, I concur with all of that. I actually haven't found anything major with regards to kind of age. I think the really fun thing about working with teams that are building novel stuff like D-PAN is that often they'll come as Neil saying from hardware background or a telecom background. We come from a software background, cloud distributed systems, cryptography, blockchain stuff. And so it's actually very, I think, easy in some sense when you're having conversations with these teams, like on the technical level for both teams to impress each other and realize both teams can realize that it's much better to work together than for either of us to try to go out of the loan just because the technical expertise on both sides are so different. And for us generally we are like an open book when it comes to architecture, how our setup looks like, what we've done for security, what we've done for reliability. And so if anyone has any kind of lingering doubts, we can typically just kind of show people what our setup is. And if they have experience in cloud or blockchain or anything like that, they tend to find no problem with it. One of the things we found with Caldera is that because we're more positioned as like a cloud services business to some extent, where we're hosting chains for people, a lot of what we do pattern matches a little bit more closely into Web2 Cloud. So for the teams that are interested in things like SLAs, the poor, more kind of enterprisey features, the ticketing system, etc. We're able to provide that in a way that I think like a fully decentralized distributed network can't really do. And so there are advantages to that, there are disadvantages to that. But from the perspective of kind of working business to business with a lot of these more traditional teams, it's definitely quite cool. Yeah, makes
Speaker 3
sense. Well, thinking both guys for making time of chat has been awesome. I think the folks are listening will learn a lot and should definitely check out at the URLs for anyone to see. So the clip stop builders for Eclipse and caldera.xyz we often find Matt and team and their docs and info. So yeah guys, anything else before you drop up?
Speaker 2
No, thanks so much for having me. Yeah, thanks for having us. Yeah,