The Next Chapter of Modular Blockchains: Decentralized RPC w/ Gil Binder (Lava Network)

[00:00:00] Hello and welcome back to the Strangewater podcast.

[00:00:21] Thank you for joining us for another episode. Best case scenario, this person is operating one of these real world computers. They would be able to pull up their local copy of the virtual machine, read the current state, and then send commands to change it. Any changes made locally would be synced to every other computer in the network. And likewise, any changes made on another computer

[00:01:43] in the network would be synced back to their local copy.

[00:02:42] to a lot of people, even if they're interested in being node operators.

[00:02:45] But without a node, you cannot access

[00:02:47] the blockchain computer.

[00:02:49] And so, centralized RPC services arose.

[00:02:54] Services like Infura and Alchemy

[00:02:57] are simply just node operators

[00:02:59] who offer to relay requests and data back and forth

[00:03:03] between the blockchain and its users,

[00:03:06] so that users can access the blockchain very much an episode of hope and solutions. Today we have Gil Binder, co-founder and CTO of Lava Network, a protocol looking to take all of the paradigms and technologies we've built to create a decentralized blockchain computer and apply it to the RPC level.

[00:04:24] Over the next hour, you'll not only gain an appreciation of how RPC works and is kind of the forgotten stepchild of blockchain computers and decentralization and all that. So super, super excited to get there. But I'm a huge believer that the most important part of every conversation are the people in it. So with that as a preface, will you tell us a little bit about yourself, how you found computer science and crypto? And I guess, why? What about it resonated with you so much that you

[00:05:46] decided to build a career in it? stuff like that and then like finding the flaws and finding the bugs and seeing you know how things work with each other in the system sometimes even completely without access to the code and you know after doing that you know I was looking into crypto from a friend and he was saying you know Gil you have to come and see you know what the theorem is and I was like and this was

[00:07:03] only like two and a half years three years okay, I'm in. I love this technology, what it brings to the world. And I think it's so early so that I can still really innovate and bring something to the table. So your background is really hardcore cybersecurity. I mean, you're doing military stuff like I can only imagine, but can you talk a little bit through this moment where your friends,

[00:08:22] you need to check out Ethereum. And I you're using Flashbots at the time, that transaction is not going to be included, so you never paid gas. So basically, you had a pure profit way to generate arbitrage. And I was like, he's bullshitting me.

[00:09:41] I don't believe that.

[00:09:42] It sounds, you know, whatever.

[00:09:44] But then I went home, I went on vacation,

[00:09:46] and I had some time, and I pick up, Now I understand the technology and what it brings. And so can you talk a little bit through, maybe as a get us to Lava Network, but can you talk a little bit through why you, so you entered this space like very much through kind of the MEV lens, or let me put it a different way from your story. I could totally imagine you starting a startup

[00:11:00] that is about MEV, whether it's block building

[00:11:03] or part of the supply chain or whatever.

[00:11:05] And so can you talk a little bit get like, to build actually more technology. And you know, seeing that blockchain is not just MEV, right, there's so much more, just like, you know, this primitive that gives you the ability to reason about logic in a way that's distributed

[00:12:21] and there's consensus.

[00:12:23] I thought that was really, and give them the ability to give service to users. Yeah, I hear you, and I guess like just to close off the MEV part of the conversation, look, like obviously

[00:13:41] a lot of people like continue down that road

[00:13:43] and are making a lot of money from it,

[00:13:45] and that's not like a dead Yeah, so I still do, you know, I still have these cool projects from time to time. I see something cool and I build a mod. For example, French Tech, I was able to bought it and make really good returns and really bought really cool people on Twitter and then dump it. So it was a lot of fun. So I still do it from time to time.

[00:15:02] And I think there's still, I don't need to think, empirically there is like a very similar experience to what you had. It reminds me of Slack as well. Their story was very similar where they were building something else and they're like, well, we need a good tool to communicate inside the team and they've built Slack. Wait, this tool is really awesome. And that became like obviously a major success. So I guess before we get into lava and decentralized RPC,

[00:16:23] let's start just with like a real brief explainer

[00:16:26] of what RPC is and you donossips about it to other nodes and then it gets included in a block. Mm-hmm. And I guess just to be real basic here, I thought the point of Ethereum was that we're able to interact with it directly and trustlessly. Like why am I as a user

[00:17:43] have to go through this third party that is doing the trust. And like, well, I'm not even referencing the fact that if you don't have 100% uptime, and you're staking a crypto asset that has implications for your P&L, let weeks to sync? No. You want to trust somebody else to do it for you. So how do you do that without losing the values that are everything that we've come to love in this space, which is

[00:20:21] being permissionless, not having to trust send this back to the centralized RPC. And the centralized RPC, it says, this is a signed transaction. And so I can submit it on behalf of MetaMask to Ethereum. And then

[00:21:42] once Ethereum has it, hopefully we all know what chain. It's like a program. You set a world computer, it's like a program. Your swap has a program. And the ABI defines how to call, what is it called?

[00:23:03] A swap, let's call it swap.

[00:23:05] The swap function.

[00:23:06] So you call getting this list. This list has let's say 10 providers. We have something that basically talks to these providers on your end and finds the best one. The ones that's closest to you.

[00:24:20] The ones that is the fastest.

[00:24:22] The one that has the latest data.

[00:24:24] And then it takes that transaction and sends it of their service determined by you, the user. So over time, based on how much requests you've made, this will affect the quality of service of the provider,

[00:25:43] the reputation of the provider.

[00:25:45] So this is how they mechanism that we can talk about. But it's also really dependent on the actual blockchain that we're talking about. For example, on Ethereum, you don't have finality for like 15 minutes for your transactions, right? This is one of the things that Ethereum is working on to bring fast finality like Cosmos

[00:27:04] and I think some other chains like NIR have them. which is okay, you know, it's not their fault. They're just seeing a different partition of the blockchain. They're thinking this is the right blockchain. The other reason is someone's lying to me. They're telling me I have no money. What the hell? And the third reason is like the second one, only they made a mistake, okay? Their server didn't work properly,

[00:28:21] there was a corruption, whatever.

[00:28:23] They gave me the wrong number.

[00:28:25] So for this, the malicious option, Let's say but I think maybe it makes sense to just like let's quickly hop over the lava Blockchain and then we'll bring this all together So can you describe this a little bit just like from from almost like spec standpoint? Like what type of blockchain is it? Did it is it you know an EVM fork is it? Yeah, just just can you describe like at the base layer? What is lava blockchain? Yeah, of course

[00:29:42] So love blockchain is a cosmos based chain

[00:30:44] to be this slashing market. It sounds like all that's going on on the lava base chain is a slashing mechanism, which is the ability to take in reports from the individual participants,

[00:30:52] validate that against, actually, I'll ask you what are they validating that against? And then

[00:30:57] if there's a problem, then executing slashing. And these providers, they're joining to get paid, they're not doing it for free, right? So they submit proof of payments, proof of service on chain to get paid based on how much service they gave you.

[00:32:21] The chain also allows users to buy subscriptions on chain. that ecosystem lock-in and the data that you can collect to build analytics off of the requests that are being made are worth it to, you know, the company that owns both these products. And I guess eventually we're either going to realize like we're all just gonna be farmed for data and gonna be okay with

[00:33:42] that a la Facebook or we're gonna have to enter a new paradigm where it's you of blockchain. So Ethereum is a huge blockchain, right? It's the number one blockchain, the most used blockchain and the most robust and reliable and known blockchain, right? The wallet, all of the data loads. And in that case, the D app is leveraging grade almost every production grade dapp that you'll use uses a separate RPC. All you need to do to check is click right click inspect page go to console or network and you can see that it's making queries to an RPC server. It's almost every production grade dapp or service. So this is the separation. We've been working with a lot of ecosystems. For example, we just launched what we call IPRPC. It's a really, really cool concept. I would love to get into it. But we launched it with Evmos successfully. We're launching it with StarkNet.

[00:39:02] We've launched it with NIR and many others. last few months since we started IPRPC, the end of last year, I think the foundation and the different ecosystem have distributed close to maybe $50,000 plus in tokens. And I think there are hundreds of thousands more that will be distributed this year. Not making any promises, but as far as I know, there's a lot of contracts that are happening now.

[00:40:24] So it's really, really interesting and cool to see. Just like Google doesn't charge you a subscription to use their servers, so does, doesn't, you know, Lava or anybody else to charge the end user for service. There's different models that allow for this. And you know, one of the things that I'm thinking about in the future of blockchain is complete chain abstraction. I think, you know, if you're bringing your grandma or, you know, your uncle or whatever,

[00:41:42] and they want to use the blockchains, they don't care.

[00:41:44] They shouldn't care. This is a little bit of a tangent, but I'll ask it anyway. I just want to cue in on what you said about end-gain state of chain abstraction. Because in a certain sense, I hear you. Obviously, UX for us is just a mess. But taking a step back, if grandma was like,

[00:43:01] Rex, I want to buy ETH. We lived in a way that allows users to extract more value from the network RPC systems is never expose the direct cost to the users. That's just going to kill user experience. Yeah, but at the very minimum, let's agree that right now it's a fragmented mess and

[00:45:40] that the more time that passes, we're getting, listen, my data is accurate, you can trust me, because if not, you can slash me on chain, right? There's so many things you can do with this primitive to prove to other people, other chains, that the data that these providers are bringing on chain, if you want to, is correct. So this is how we see the vision

[00:47:00] and how we see Lava, part of the modular stack,

[00:47:03] whether it's on Cosmos or Celestia or Eigenlayer,

[00:48:07] champions. Champions are members of the of the market that basically build and maintain these specs for these chains. They basically champion the chain

[00:48:12] into lava. And these specs allow us to dynamically observe what's happening

[00:48:20] inside of the actual queries. And this is the superpower I was talking about. one more thing it might be a bit more technical is what we call archive nodes so everybody knows that or I think a lot of people know that blockchain takes a lot of storage to store right so the state every block every block in a blockchain the state changes right there's transaction transactions change

[00:49:43] the state but to store all the blocks the state at each and every single one So this incentive can be built in into the spec system, incentivizing archive data. So essentially what you're saying is the spec system not only allows you to define the specific data or requests that you're willing to serve, but it also allows you to add the economic lever onto it and say, okay, we want to incentivize this behavior, so we're going to pay more

[00:51:04] for it.

[00:51:05] Exactly, yeah.

[00:51:06] And we bring it on chain, and then you get splashed. There's a consensus, there's a vote.

[00:52:21] Everybody provides proof

[00:52:23] that they're able to capture the same result.

[00:52:25] And if the dozen and whatever side

[00:52:27] has the most votes based on stake, with the response on chain and the response from the provider. And this is like a Swiss knife, so anybody can build whatever spec they want and whatever validations they want. So these are all the options. I want to get to just modularity in the modular blockchain revolution in a moment,

[00:53:40] but while we're just still very much on lava state and claim that the data is correct. And there's a mechanism to dispute it. And then others could potentially in the future

[00:55:01] rely on this data to build an next obvious one is RPCs. And so I'd love to hear just your reflections on what the modular roadmap means, what it means to you.

[00:56:20] And if you're feeling really ambitious, I you have one sequencer you have to trust, and sometimes the proof is not even yet, I mean the optimistic proof or the fraud proof

[00:57:42] is not yet implemented.

[00:57:44] For example, OPStack,

[00:57:45] I'm not sure how many people are familiar with that,

[00:57:47] it's on the roadmap, they're developing it,

[00:57:48] you know, it data is the capital, you can end up losing money, which is really really bad. we have distributed sequencers and then distributed DA layers, and that's on top of our distributed main net, and now we have distributed RPCs. And the next layer of these giants? I would say it's probably close to nothing. Yeah, 0.01% maybe. Yeah, so this is just, and remember, this is just, it's a nascent field. It's like, it's still like, it's very, very new. So very, very early.

[01:01:41] And again, my view is that what we live? What are the next steps? Yeah, of course. Thank you. Yeah, we just announced our $50 million round, sit around from super OGs in the space like Tribe, Hashkey, Jump, and many, many, many other supporters.

[01:03:01] It's been very exciting.

[01:03:03] We've been developing Lava for one and a half years now.

[01:03:06] We've been on Testnet for the last year. what this means for lava, but really, in a bigger sense, congratulations on really understanding one of the true centralization and risk points that we have in this space, and doing something to solve it. And I think that you could have very easily stuck in MEV, or you could have very easily become a shitcoin trader,

[01:04:20] or anything that is really about value extraction,

[01:04:24] and instead you found something that's about positive sum,