EigenLayer Introduces EigenCloud: The First Crypto-Native Cloud Service | Sreeram Kannan & JT Rose

Transcript:

David:
[0:03] Bankless Nation, I'm here with Sri Ram Kanaan and JT Rose from EigenLayer. JT, your first time on Bankless. Welcome.

JT:
[0:10] Thank you for having me.

David:
[0:11] And then Sri Ram, good to have you back.

Sreeram:
[0:12] Thank you so much, David. Always a pleasure.

David:
[0:14] Big day in the EigenUniverse. EigenLayer is announcing EigenCloud. Now, there are a bunch of things that when you pop open the hood of EigenCloud, you'll see a bunch of different modules. And we're going to talk about EigenDA, EigenVerify, EigenCompute. I kind of want to just like speed run through the history of eigenlayer as it stands so far, because I think when people first hear the words eigenlayer, we think of restaking. That's the big primitive where you stake your ETH, you stake it again. And then there's this notion of shared security and AVSs, where AVSs are providing services using eigenlayer to package up and sell trust to end users. And I think the average crypto person on crypto Twitter or the average Bankless listener might actually kind of like lose, lose the idea of what Eigenlayer is around there. There's this notion of intersubjectivity. There's this notion of renting security. Polkadot has tried to do this. Cosmos has tried to do this. There's something there. And the precise nature of like what Eigenlayer is and how it actually reaches mainstream is, I think, still a little bit cloudy.

David:
[1:19] But today we are really, Eigenlayer is introducing Eigencloud. And I'm going to kind of call this as like Eigenlayer and EigenDA and AVSs. These are all kind of like components in a car, right? You got a transmission, you got an engine, you got a carburetor. And that's what I think when people really like zoom in on Eigenlayer and they see these different components and they're like, oh, what is it though? What is it doing? How do these things work?

David:
[1:48] Eigencloud, what you guys are introducing today, hey, I'm going to call the whole entire car. It is actually like kind of the packaged up modules. And, you know, in crypto, especially the crypto natives, we kind of like to like open up our cars. We don't really want things abstracted from us. We actually like to see the bare metal. We like to poke around. We like to, you know, do contract calls from Etherscan. But really, we know that like if we're going to take crypto mainstream, we actually need to package these things up. And so the way I see Eigencloud, which we're going to get into today, It's like the packaged, sellable vision of Eigenlayer. And Eigenlayer is just kind of this foundational layer of this whole entire vision. It's the thing on Ethereum. It's the layer one smart contracts. But Eigencloud...

David:
[2:30] Is like AWS, where it's a place where developers can go and buy access blockchain properties. And so Eigencloud is this end consumer product, the consumer being developers, that actually allows some of the blockchain properties that we have. Verifiability, trust, really, economic security, really, is being packaged up by this Eigenlayer system and then made accessible in the same way that a developer would go to AWS and rent out some hardware, they can go to Eigencloud and access properties that blockchains tend to give things like economic security, like trust.

David:
[3:08] That is my like kind of high level pitch for Eigencloud. Now I am a podcaster. I am directionally correct, but technically imprecise. So Sriram, maybe you can kind of like fill in the gaps a little bit and like assist the listener with understanding what Eigencloud is.

Sreeram:
[3:24] Amazing. Thank you so much, David. Really excited to be here to talk through this, as Azeen mentioned, we're announcing this new package, EigenCloud, where we want to make the whole greater than the sum of the parts, right? So the parts people have seen, and frankly, people are a little confused, right? So we're building Eigenlayer, which is this staking layer. I think if you go ask anybody about Eigen, to the extent people know it, they know it for staking, restaking. So that's what, you know, the single word that you would hear. But we've also been doing other things and people are like, what are these guys doing? They seem to be doing a lot of unrelated things. We have EigenDA, right? Like a data availability engine built on top of Eigen layer. So that's already there.

Sreeram:
[4:10] It secures the most TVL of all off-chain DAs on Ethereum today. It's the fastest, highest throughput DA. So we have these two parts that people have seen. We've talked a little bit about, oh, you know, you can use Eigen to power AI, and people have seen that. And so today what we're unveiling is this whole, which is greater than the sum of the parts, which gives coherence to all these things in a single clear message, as well as a packaged up product. So David talked a little bit through our history, so I'll just spend a minute on like how we got here. So we started with Eigenlayer, like this whole restaking protocol. You know, we went and launched that, I think, June, July 2023, the staking contracts. You know, we got stake. And, you know, today we have like $12 billion, something of stake. On the other side, we have operators, people who run the computations for these stakers. We went from, again, zero to now like 2,000 operators on mainnet, going from home operators to professional operators to things like Google Cloud running a first-party operator.

JT:
[5:23] And then we have services built on top of Eigenlayer.

Sreeram:
[5:26] We call them AVSs. AVSs are, hey, I'm renting the economic security and running a certain specific specialized service. Think of like data storage. I want to run a data storage service, right? So you need a group of nodes who store and make a promise that they're storing these things correctly. So that's an example of a data storage service that people run. So you have these specific primitives, many people building AVSs. It was one of the hardest categories to create because for somebody to come and build a service on top of a platform like us, they're taking a massive amount of platform risk because their entire protocol there is dependent on this eigenlayer set of interfaces. And so we have like 50 plus on mainnet and 100 plus in other stages of development showing up. So really excited to have created that category. So that's kind of history so far. And we separately also talked about the eigentoken in a white paper, right? So we had the eigentoken white paper, which we published, you know, around a year back.

David:
[6:34] The famous intersubjective paper.

Sreeram:
[6:36] The famous or infamous intersubjective paper where we introduced this idea of like, hey, what is the kind of like foundational value prop of blockchain? How do we abstract it into a token? And I think it was kind of like too abstract for, you know, most people to track. And, you know, rightly so. It was not meant for the general public. It was meant for technical people to understand like what sets of things you can unlock.

Sreeram:
[7:02] But I think what we're doing now, so that's history, right? And one of the things I'm very happy to say is we've pretty much delivered every single thing we have stated till like 18 months back. So if you think about like tracking the say-do ratio, like what we said to what we did, right? Everything we've stated till say November, December, 2023, we actually delivered. Like we have Eigenlare working, you know, it's slashing went live, redistribution, which is this new idea of like, When you're slashing in Ethereum, the stake is just burnt, right? On Eigenlayer, you can redistribute it to the harmed parties. So we introduced this in an academic paper called Stakeshare. With redistribution, that's going live. So essentially, like the whole Sets of ideas that we had introduced till like 18 months back are all live and in production. Same thing with EigenDA. We claim this big thing that, okay, you can run a DA with insane throughput, you know, live on mainnet with 15 megabytes per second. We did an experiment six months back where we dumped the data from every single blockchain. This is Bitcoin, Ethereum, Solana, you know, Celestia, Base, everything, and then put it roadblocks into EigenDA and still it couldn't saturate EigenDA's bandwidth. So that's how much bandwidth EigenDA has. Yet we continue to scale it.

Sreeram:
[8:26] That's at 15 megabytes per second. Right now on testnet, we're running at 50 megabytes per second, a 3.5x increase.

Sreeram:
[8:35] Internally, we're running at 100 megabytes per second. So we're just moving the throughput barrier out of the question. So that's like a say-do ratio of the history, right? So what we said versus what we do, because we're saying new things now. And, you know, obviously a question in people's mind is, oh, are these guys just saying things? Right. No, so we have a track record actually delivering these things. So that's the context.

Sreeram:
[9:00] And then now I can switch into Eigencloud. Any questions on the context, David?

David:
[9:04] JT, you want to add anything? No, that all makes sense to me. Yeah. In the Wild West of DeFi, stability and innovation are everything, which is why you should check out Frax Finance, The protocol revolutionizing stablecoins, DeFi, and Rolex. The core of Frax Finance is FraxUSD, which is backed by BlackRock's institutional Biddle Fund.

David:
[9:21] Frax designed FraxUSD for best-in-class yields across DeFi, T-bills, and carry trade returns all in one. Just head to Frax.com, then stake it to earn some of the best yields in DeFi. Want even more? Bridge your FraxUSD over to the Fraxtile Layer 2 for the same yield plus Fraxtile points and explore Fraxtile's diverse Layer 2 ecosystem with protocols like Curve, Convex, and more, all rewarding early adopters. Frax isn't just a protocol. It's a digital nation powered by the FXS token and governed by its global community. Acquire FXS through Frax.com or your go-to DEX, stake it, and help shape Frax Nation's future. Ready to join the forefront of DeFi? Visit Frax.com now to start earning with Frax USD and staked Frax USD. And for bankless listeners, you can use Frax.com slash r slash bankless when bridging to Fraxel for exclusive Fraxel perks and boosted rewards. Uniswap is your gateway to a more efficient DeFi experience. With Uniswap swapping and bridging across 13 chains, it's simple, fast, and cost-effective, helping you move value wherever, whenever. Thanks to deep liquidity on the Uniswap protocol, you'll enjoy minimal price impact on every trade. And now Uniswap V4 takes it even further. Swappers benefit from gas savings on multi-hop swaps and ETH trading pairs, while liquidity providers can create new pools at 99% lower costs. The best part, you don't have to do anything extra.

David:
[10:37] Each trade is automatically routed through Uniswap X, V2, V3, and V4, so you get the most efficient swap without even thinking about it. Whether you're swapping, sending, on-ramping, off-ramping, or bridging, Uniswap's web app and wallet gives you the tools to unlock DeFi's full potential on Ethereum, base, arbitrum, unichain, and more. Use Uniswap's web app and wallet for a more efficient way to use DeFi. Celo, the blockchain engineered for everyday payments, has officially returned home to Ethereum as a blazing fast Layer 2. Built on the OP stack and EigenDA, Celo delivers one-second blocks.

David:
[11:07] Subcent fees, and fee abstraction so users can pay gas in USDT, USDC, or any whitelisted ERC-20 token. Add native bridging, Ethereum-level security, and ultra-green money, and you have a network purpose-built for global commerce. In just five years, Celo has processed 730 million transactions for 12 million addresses. It's launching with 100 partners and fresh DeFi firepower. That makes Celo the on-chain FX hub for borderless remittances and real-world spend. For builders, distribution is baked in. Programs like Presentee Grants, Retroactive Public Goods Funding, CeloCamp, and proof of ship will put your mini app on farcaster and opera minipay tokenomics leveled up too the great cello halving cuts inflation from two percent to one percent aligning the chain with its ambitious vision 2030. A trillion dollar on-chain economy where regions cypherpunks and dgens create prosperity for all follow at cello on x and start building at cello.org today and i just want to return to the metaphor of just like the car being built and i think like eigenlayer learned that i'll call.

David:
[12:01] Eigenlayer, the actual thing on Ethereum, the restaking. I'll call that the engine because that's this thing that powers trust. That's the thing that takes crypto economic security, the stake on the Ethereum layer one, and then turns it into this raw material that other modules can hook into. And those are the AVSs. You need data availability in order to actually have that be interoperable and verifiable. And so all these extra layers, because you can't just have an engine, right? The engine needs to connect to things, needs to have a transmission, it needs to have wheels. But eigenlayer being the engine and then these additional modules to really fill out the form of the car. And I think what I'm hearing you saying is like, okay, the other modules we've actually built and hooked together. And so now it's time to talk about the car, the car being eigencloud. That's how I'm going to carry that metaphor forward.

Sreeram:
[12:52] That's absolutely right. In fact, it's funny, like, you know, if you look at some of our earliest papers on like what is Eigenlayer, like, you know, this is even before we knew about restaking, you know, the ideas you would see are like a cloud. And like, hey, you know, here's a cloud. It connects to Ethereum. It connects to, it then lets many people build applications on top. You know, some of the early visuals are identical to like what we're trying to build. So this has always been in our mind that this is the orienting direction, the North Star. But it needed all these perimeters and boxes to, you know, it needed, okay, how do you build an engine? How do you build a drivetrain?

JT:
[13:28] How do you build,

Sreeram:
[13:29] You know, fast tires? How do you build all these things? And then you can slam

Sreeram:
[13:32] them all together to build this cloud. Okay, so what is this Eigencloud and what is this in-service of? So the high-level concept is we want cloud-scale programmability, cloud-scale programmability. Anything you can program on the cloud, you should be able to program on Eigencloud. Okay, so what's new with that? That's like AWS or Google Cloud. Like what is new about that? Those already exist, right? What is new is you want that. You want cloud-trial programmability, but you want crypto-grade verifiability. You want to verify things, just like is normal course of business in crypto. You don't want to trust that Sriram or Eigenlabs or somebody else is going to run their stuff correctly. You want to make sure that these things are all running correctly. So that is the kind of goal, the North Star. The promise is we want...

JT:
[14:25] Cloud-scale programmability.

Sreeram:
[14:26] Anything you can program on the cloud, you should be able to program on Eigencloud with crypto-grade verifiability. If you do it on this, you don't need to trust anybody. It's verifiable. It's secure. It has a measurable unit of economic security or other kinds of security attached to it. Economic security is our starting engine. That's like David said, that's our foundation on top of which we're building a lot of this stuff. But as we expand out this cloud, you can have different grades of security models and stuff that can be built on top.

Sreeram:
[14:57] But core promise is cloud scale programmability and crypto grade verifiability. So if you take this premise and then deconstruct it, you know, I was talking to David like a week back and he asked like, hey, aren't blockchains already Turing complete? Like what is this new cloud, you know, programmability you're talking about? Aren't blockchains already like Universal and Turing complete? David, if you want to.

David:
[15:19] Yeah, there's just like a bunch of blockchains out there that are doing the IBRL thing, right? Like Solana has, you know, to its credit, incredible scale. Base has incredible scale. We have MegaEth coming down the line, incredible scale. And so we already have Turing complete blockchains. You know, Ethereum invented that with smart contracts. Now with this like next generation of blockchains, we have, you know, seemingly so much scale that we can't actually figure out how to congest these blockchains other than in very like narrow paths, like very specific meme coins being contested. But other than that, scale is, I kind of consider scale to be solved. And so my question to you, Sriwama, is at the time is like, okay, what does it mean to have even more scalability and verifiability under that context? Because I kind of thought that we had like already solved both programmability with smart contracts and scalability with like things like Solana, Base, MegaEth, Monad, like all of these things. And so like, where does that fall through?

JT:
[16:14] Amazing. I think if you look at,

Sreeram:
[16:17] You know, the ambition of crypto, like let's say in 2017, 2018, when I kind of got in, my opening lectures would be, we're going to bring the entire, disintermediate the entire world, right? Like, we're going to bring Uber on chain, we're going to bring insurance on chain, we're going to bring everything that is intermediated through trust parties, we're going to remove it. And then we're going to use blockchain to disintermediate trust, make it verifiable, make it permissionless, all the core value propositions that we know and love in the blockchain space. That was what our goal and ambition as a crypto space was in 2017, 2018. And I think what happened is that as we traverse that path, we hit a bunch of like foundational infrastructure bottlenecks. Say, oh, you know, I want to build Uber on chain. But OK, you know, this really can't support any of the throughput. So I say, okay, let's just do token transfers and payments for Uber on-chain, right? So I want to just point that out. Even today, if you ask today people, like, what is blockchain? For example, what is Solana's, like, North Star?

Sreeram:
[17:20] Decentralized Nasdaq, right? Like, trades. I want to trade. I want to have assets on-chain, and I want to do trades. This is a great, like, really high-value proposition, but it's not the entirety of crypto. Our thesis for the entirety of crypto is how do we bring the entire cloud, how do we bring the entire economy even, like forget cloud, entire economy on chain, because the economy is built on trust. And how do we upgrade the trust rails for all of that? So specifically, how does this, what are the specific bottlenecks in blockchains that actually like make this not possible? Let's say I want to build X Twitter on chain, right? And you say, okay, now like base or mega ETH or somebody has enough scale. So I'm going to just like take the X, you know, take some open source code base and then slam it on top. Is it going to work? No, you can't even use any open source software library directly on a blockchain. You have to rewrite it in the specific languages and formats that are okay to be built on top of blockchains. You'd have to

David:
[18:24] Rebuild Twitter with Solidity or with Rust to get on Solana, right? Is that what you're saying?

Sreeram:
[18:31] That's absolutely right. And so the programmability and the software limitation is massive. JT, you want to add something to it?

David:
[18:38] Well, yeah, one of the things.

JT:
[18:39] When we first started talking before I even joined Eigen, I think the belief that you had was that many, many more things could be built on blockchains that just weren't. We have sort of applied the superpower of crypto, the verifiability of crypto, very narrowly at this point. DeFi, stable coins, real world assets, things that you can tokenize, things that you can coin and you can bring on chain. But your ambitions was more in line with, like you said, 2017, 2018 Ethereum, when I think it's all of our particular vintage. And your approach was instead of saying like, how do we remake apps to fit on chain? It was, can we remake the app architecture so

Sreeram:
[19:16] That we can

JT:
[19:17] Apply crypto guarantees to anything? And so restaking and eigenlayer was, David, as you pointed out, it was a means to an end. it was a way to aggregate capital and willingness to monitor transactions and apply it. And that kickstarted a bunch of AVS development. But ultimately, the vision, Seram, as I understand it from you, is that we want to apply that power to many, many more types of applications. But there's sort of a fundamental constraint that remains that we're trying to address with Icon Cloud.

David:
[19:48] Yeah. And the reason why we're talking about 2017, 2018 is like we're kind of talking about actually the ICO meta, where there were these incredibly imaginative ICO projects that were so incredibly pie in the sky that I think really struck at everyone's imaginations, which is why the ICO meta became the ICO meta.

David:
[20:06] Everyone realized, oh, blockchains, trust, verifiability, we can rethink things. We can rethink the internet. And then there were a lot of people who were like, yeah, and I can package this up in a big narrative and sell this as an ICO, but that's aside from the fact. I think what SreeRam is really calling out is there was this era in crypto where people's imaginations were just completely unbridled and were running rampant because they saw these core primitives of trust, verifiability.

David:
[20:32] You know, security and kind of just like applied it and then kind of forgot about how those things were connected, how the application and like there was some missing links there. But nonetheless, the ideas were there. And we also tried out like things like Peepeth. Peepeth, this is like an old example that I keep on bringing up on Bankless over and over and over again. It was Twitter on the blockchain. And you would write a tweet, you would pay ETH and gas, and then Peepeth would actually write that tweet into the Ethereum layer one. And this is back in the days of zero GUE gas fees. So it actually was like free back at the time because no one was using Ethereum. And this was an example of a Web2 app. We're coming on chain. We're going to make Twitter be on chain. We're going to have like, you know, the immutability and verifiability of Twitter be on Ethereum. And that was back in like the experimental days of 2017. It's like, sure, why not? Like, no, we don't really know what blockchains are good for. Why not put literally Twitter on the layer one? Now, if you if anyone like wrote, like tried to like start up a Twitter on the blockchain project in 2025, no one, everyone would understand that that's not how you do things. And I think Eigenlayer understands this, too. It's like, no, we're not going to put tweets on chain. We're not putting tweets on chain. We're actually packaging up properties of blockchains with Eigenlayer, with restaking, and we're turning those into modules that can be exported and like sold as a SaaS business offering by AVSs using Eigencloud to things like Twitter or things like Web2.

David:
[21:58] And so we're actually exporting, it's going the other way. We're taking economic stake, we're turning that into a product that can be sold to Web2, to other things. And Shreemam, I'll hand it back to you in a sec, but there was one of my favorite like deep cut bankless episodes was our first episode, I think, with Tarun Chitra. And we went through the evolution of just digital marketplaces with like starting with how the stock market traded in like the 70s, moving into where like the NASDAQ came in and like different technologies. And like the main takeaway from that episode is like the markets that we trade on all got improved in this like step function way. Every time some technological innovation happened, that found a way to remove lawyers out of out of like a core product of the product. And then the markets just got 10X larger. And like there was some like transition in like the late 90s where like markets went digital. And it was because somebody found a way to like just completely remove an entire layer of lawyers. And that's actually what I see Eigenlayer doing is like you don't need an SLA. You don't need a service license agreement. You don't need to like get a lawyer to draft up a contract between these two companies. You could just go to Eigenlayer, pay some like fee for renting security from an AVS that has provided you this service.

David:
[23:17] And instead of having the legal system, the nation state legal system, ultimately backed by an army, sponsored by Coinbase, having that be back up the contract, you actually just can have an economic weight back up that contract. And that's actually far more interoperable and like verifiable because of Ike and DA. I'm going on a little bit of a rant here, but I think that's some of the context that I think is worth bringing into this conversation. Sirom, I'll throw it back to you.

Sreeram:
[23:41] That's right. Right. So there is a famous saying that says civilization progresses as we find one more thing to automate. Right. So that's like lawyers is one thing, but like, you know, it's lawyers, it's audit firms, it's, you know, trust is the fabric on which the entire society is built. So as we try to automate trust, like how much bigger can our economy size be? Like that's, I think like much, much bigger than what, think right now, the ambition inside crypto that we're seeing. And it goes to like, how do we disrupt digital marketplaces? How do we build trust in an era where there is like, you know, AI as a kind of like a dominant force in society? It goes to how do we reimagine societal scale incentives at a global level where like all these contracts can be self-enforcing?

Sreeram:
[24:33] So that's like our big goal and vision. And then like, I think it's shared, like a lot of us here came to crypto because of these ideals and ideas.

Sreeram:
[24:43] And, but what happened, I think like you mentioned the ICO era, David, like you can go back to like 2010, 2011 Bitcoin talk, right? Like people were like, wow, there's this Bitcoin thing. Now, you know, what other thing can you build on top or build other things like this? You see that thing. And similarly, I think what happened is 2014, 2015, 2016, Ethereum came on the scene and showed that this is not just theoretical ideas. This is actually like right now possible. You can write a smart contract. You can build that Twitter-like app and deploy it, right? And that kicked off a set of imagination that was pretty contagious because now there's a developer tool along with a kind of like an idealism and those two meet and create something really interesting.

Sreeram:
[25:30] But then what happened was CryptoKitties happened. We ran into the kind of throughput bottleneck, like the ZeroGue era ended, right? CryptoKitties was this NFT kind of game project. And it took up pretty much all of the Ethereum mainnet bandwidth. And then everybody realized, hey, we need to scale. We went through the whole L2's roadmap, I think, very successfully. It's really amazing to see, David, that I was sitting as an academic, right? And I think in long timescales, looking at the ZK roll-up roadmap and then thinking, what are these guys smoking? Like, why is this a solution? Yeah.

JT:
[26:09] It's not a long-term solution,

Sreeram:
[26:11] I know that, but it's crazy. It's Ethereum and the culture around Ethereum that made this kind of absolutely crazy imaginative roadmap of ZK rollups become real in this timescale beyond the wild bets. I think it's parallel to something that happened in AI that, you know, even the most bullish of people underestimated the timelines. There's something similar to that. But those solve the core scaling bottlenecks of Ethereum. They solve the core scaling bottlenecks of Ethereum and like the blockchain

Sreeram:
[26:44] in the blockchain architecture. But they do not solve the programmability bottlenecks, which is really what we're speaking to is I as a developer should be able to take any code anywhere and then slam it on top of a blockchain. I should be able to take use any hardware that I want. I should be able to use like H100 GPUs to run my AI. I should be able to use the latest, you know, GPUs to run my gaming engine and still have it be verifiable. I should be able to run trusted execution environments and have it be private so that other people don't have to see every single transaction for it to be verifiable. So I want as a developer, much, much higher control than what blockchains offer today because they're built on this idea of every node comes to consensus on a deterministic state transition. This is the whole premise.

Sreeram:
[27:38] In fact, you know, David mentioned earlier that if you look at today, like what's the biggest things going on, things like Solana, like MegaEat, like Base, like Sui, trying to accelerate like the throughput bottleneck. And if you look at what they're primarily trying to do, they're trying to primarily do parallelism, right, inside the blockchain in architecture.

Sreeram:
[27:59] And if you ask, like, hey, isn't parallelism a 1980s computer architecture thing, you know, multi-threaded, multi-core, like, you know, why are we like talking about it like it's a new thing today in crypto? It's because the underlying assumption of consensus on deterministic state. And so you want basically like blockchains to arrive at consensus on a common deterministic state. And that basically means I'm going to force everything to be very rigidly deterministic. So I build a rigid construct, like the EVM is a rigid construct that says you can only do certain things and anything else you have to like say that, oh, somebody is signing off on it. It's an Oracle problem. Somebody signs off on it and then brings that data in. And so you build this very rigid wall of determinism, which gives you a very strong grade of security, right? But it restricts massively the developer surface area of what can be built on top. So that's what our goal is, cloud-scale programmability, crypto-grade verifiability. Where blockchains are is they took this original decision, which is to bring this tight consensus on highly deterministic state. You have to fix a VM.

Sreeram:
[29:15] Because you have to fix that VM to be very rigidly deterministic, you can't have parallelism. Because when you do things in parallel, maybe the threads don't meet back at the same time. There is asynchrony, there is mistiming, there's all kinds of complexity that go with it. And so that has led down to like a path where you have like specific VMs. And then now like the next generation of like updates to these blockchains are like parallelized EVM, Let's build other VMs like Move or, you know, C-Level and so on, which actually solve some of these problems. But there are many, many, many problems like this. There's multi-threaded, there's multi-core, there's multi-computer, there's cluster programming. Are we going to invent new and new like systems and blockchains for each one of these? Or are we going to have like a foundational framework that says I can throw pretty much anything on top of a common substrate, anything that I can do on cloud on top of a common substrate and have it be programmable and verifiable. So the big question mark is, how do you get any of these things to be verifiable? And speak to it a little bit, but JT, you want to add something there?

JT:
[30:22] No, I think probably David, for a lot of your listeners who don't have PhDs in computer science like Sriram, the thing that we're looking to solve here is, like you said, there were two kind of primary constraints hindering, hampering, blocking the creativity of kind of early Ethereum days. One was scalability, which we've thrown the kitchen sink of VC money and talent at and is largely solved. We've got, we've solved the throughput problem, but I think the

JT:
[30:49] The constraint that remains is the programmability or the expressivity of the on-chain VM. It's constructed in a certain way to maintain, as Sriram says, the determinism that makes it verifiable. And so we can either accept that and just accept that there's always going to be limits of what we can put on-chain. And thus, we'll just try to find new creative ways to build DeFi applications. Or we can say, what if we could take the application logic that doesn't need to be sort of adjudicated on chain up into an off-chain container? Are there ways that we could guarantee an L1-like verifiability of that execution so that you could offer developers cloud-like programmability, cloud-like developer experience, just run application logic in a container? And then underneath the hood, a product like Eigencloud can abstract all of the underlying blockchain verifiability. And that's what the original thesis of Eigenlayer was, is that we could do that. AVSs have been building those models for a while. And so we're taking this a step further with Eigencloud.

David:
[31:54] Yeah. And that container word, I want to double tap into because that's something that if you're familiar with computer science, it probably just made instant sense to you. But if you're not, then you're like, okay, off chain container, like what does that mean to me? And so like, I think we've pretty established this like notion of just like building on chain is rigid because we need it to be deterministic. And so what it means to build on chain is like fundamentally constrained. Now it's Turing complete. Yes. And now we have scale also. Yes. But nonetheless, the rest of the internet is not built in a way that fits on chain. And sure, you can rebuild Twitter to go on chain if you really want to, but does that really make sense? And so there's this notion that instead of making the world come on chain, literally, let's put Facebook.com as a smart contract on Ethereum. We know that that doesn't work.

David:
[32:42] And so there's this notion of a container in computer science. Maybe JT, you can kind of unpack that a little bit more. But really how I think about this is just like, there are parts of a website, parts of any sort of like code that we are going to write that we want to relate to some important property. And so we're going to write that in a container and that can go in like your Amazon S1 bucket. I think I'm getting this computer science right. Or it could go on chain. But like what Eigenlayer does is it provides a container for you to write your code logic to for very important information. Again, things that would otherwise go to a lawyer, things that would otherwise be drafted up in a legal contract. And now you can write it inside of a container that would go, again, like previously a container would go on Amazon AWS. Eigenlayer via Eigencloud is offering something very similar where logic about state transitions can be put into this container. And then using a truly Eigenlayer, like the L1 security that will be adjudicated in a particular way that will enforce a certain outcome. That's my interpretation of this, but also I'm quite constrained on my ability to talk about computer science. JT, talk about that container thing in the off-chain world context.

JT:
[33:56] Yeah. I mean, I came from Azure and a container is really just an alternative deployment tool similar to a virtual machine that you might get in AWS or Azure. And so it's a place to house your how's your application code. And so in an AWS environment, you would deploy that container against AWS services. And here we can deploy it against nodes that operators have offered to run with the security behind it that will ensure the correctness of execution. That's the diff here is that AWS is just a public cloud where you can run the container, but there are no penalties or rewards for correctness. And in Eigencloud, we have an additional actor. The raw material is not just a piece of hardware that runs the code. The raw material is state capital that forms sort of a penalty and operators that are taking an active role in ensuring the correctness of execution. And so there's a slight difference in the raw material of the cloud and the sort of abstraction that Eigencloud is performing versus an AWS.

David:
[35:04] I think that's really helpful. I think at this point in the conversation, we run up against a new constraint, which is the listeners and mine imagination as to like what happens next. Because I think we all understand like, okay, we need to have our blockchains verifiable because asset ownership, all of these things we need to run on trust. But what does it mean for these same properties to be extended to like Twitter or Facebook or like anything else like Amazon. I don't really get that part of this notion of like, okay, we have trust, we have verifiability, we have crypto economic weight. I can make a promise about something. But like, what is Twitter doing for me? And maybe I'm like over fixating on Twitter. Maybe it's a different use case. But how would you like illustrate the notion of we're taking blockchain properties. Eigenlayer is taking blockchain properties through Eigencloud. It's making this developer-friendly cloud service that can allow them to access blockchain properties to build their applications. But what applications? What does it mean to imbue Web2 with trust and verifiability?

David:
[36:09] Can you just make that a little bit more real for us?

JT:
[36:11] Yeah, so I don't know that it's Twitter or Amazon or Uber that made it hit for me. The one that's really resonated over the last six months for me is a chat GPT. There was a story in the New York Times just over the weekend about users that were interacting with models and sort of uncovering that the model was optimizing for engagement over necessarily telling the truth to the user. And so... In very harmless ways that can manifest as like, that's a great idea, Dave. Let's explore that a little bit more, right? Like how they respond to your prompts. Perhaps in more pernicious ways, they could reveal certain parts of an answer, but not reveal certain parts of an answer in an attempt to get you to ask another question. And so we don't know how these models are built. And perhaps that's not really important for like the way that most of us are using ChatGPT. But as we offload more and more reasoning to agents, as we offload more decision-making and perhaps in the future, like sovereign action to agents, if that's a world that we believe is coming, then there's always going to be a cap on how far we can let those models go without us fully trusting them and having some sort of accountability guardrail built in. And so the question is, how do you build that? You can't really build it on a public cloud, but if you have the ability to slash the results

JT:
[37:28] Of a GPT model, a hallucination, for instance, It says, you know, for a given output, I'm going to commit that this is correct. And if it's provably incorrect, then it gets penalized. And you have just a different relationship with AI agents. That one stuck out to me. Sure, I know we've talked a lot about that in the office, but I'm wondering if there's other examples that you've got that maybe resonate.

Sreeram:
[37:54] Yeah, no, just adding a little bit to this kind of like high level picture that David asked, why do we need something to be verifiable? I think that's maybe a big kind of question. Think of what needs to be verifiable is things whose consequences are material.

Sreeram:
[38:13] I'm going to load, for example, here is an AI agent. If this AI agent says, go buy this coin, I'm just going to go buy it. I'm not going to intermediate it. I'm just going to make it autonomous that the AI agent says something and then it just does it, right? Like moves those tokens on a blockchain, let's say. That's the most material thing you could kind of like put downstream of like, of an action. So for those actions, they would absolutely want it to be verifiable because otherwise, why am I giving my money to an agent that is going to go do some actions and those actions are going to kind of like cost me either loss or harm unless I know it is verifiable. This is the kind of basic grade of why we need smart contracts, right? We need smart contracts because I know that this smart contract always executes using this piece of code in this way. What's the equivalent of that for something like AI? I would say, oh, here is a, you know, model, a specific open source model, like a Lama 4.2 or something like, you know, take an open source code, say, here is the model, here are all the model weights, here are the inputs that it will read, you know, it's either on the blockchain or comes from some vetted oracles. And then it is going to run this piece of code at this time and then give you an output. And you pre-commit to all of this stuff. it's this model, these inputs,

Sreeram:
[39:38] On-chain, and bring the results back. Like that's the covenants or the condition under which you want to run this AI. And that is going to immediately cause some downstream material token movements. That's very understandable for a crypto audience. You know, as we were thinking about this whole like eigencloud idea, we were thinking like, should we start from the outside in, more of the point of view that David brought in? Are we going to go to Twitter and like, you know, Microsoft and bring them on chain? Or are we coming to crypto developers and saying, hey, you know, you can up-level, you know, whatever chain you're building your applications on, you can up-level what you could be building today. And so we took the latter approach, which is to go to like, our approach is tailored towards anyway working with tokens, right? So that's what like crypto developers do, like build things, which are conditions upon which tokens move around, right? And I'll give some concrete examples to make this very clear. And so all of those applications, we absolutely want these upstream actions to be verifiable. So what are some examples, right? So let's start with the most kind of degen or fun one, meme coins, right? So, you know, we say like in crypto, there's only three use cases today. It's Bitcoin, stablecoin, meme coin, right? So let's take meme coins, the most fun one of these.

Sreeram:
[41:00] I know one of the things you want to do in a, the word says meme coin, right? So meme, somebody should be propagating the meme, right? A meme is an idea that needs to go viral. And so if you look at like the evolution of meme coins, one of the things that PumpFun did, what did PumpFun do? PumpFun just made it possible for somebody to verify that here's a meme coin with this much supply. So they made the supply verifiable.

David:
[41:25] Every token that comes out of PumpFun, if you make a PumpFun token, it comes out with the same supply that's hard coded. Everyone knows that all other interfaces, front ends know that that's true because that's just what a PumpFun token is.

Sreeram:
[41:38] That's right. So this is a very important mental model. So when you make something verifiable, you can make it permissionless. So let me just state it again. When you make something verifiable, you can make it permissionless. Now, because the token supply is made verifiable, now permissionlessly, 30 million tokens have been created. It opens up the aperture that I don't need to be a trusted person with a vetted, audited smart contract to actually go launch a token with a fixed supply. I can just press a button and do it. So that's what Thumb Fund did. But what is missing there is what is the incentive to continue growing the meme? So you see this typical curve like sugar high and then like, you know, goes up and then bonk, everything crashes after that. I mean, it's fun. And like, we all love playing. We're all DJs somewhat here. Otherwise we won't be here. But that idea is the superpower is like the verifiability and the permissionlessness. Now, what can you do if you had Eigencloud in a meme coin world? Can you create a meme coin which says not just a fixed supply, but an inflating supply where the inflated portion of the supply is going to people who are promoting your meme on Twitter or Forecaster or TikTok, whatever, right? How do you do that?

Sreeram:
[42:58] That part needs to have, because I'm tying it directly into the heart of the token, the token supply will go bonkers. If I said, you know, David's going to kind of decide who the tokens go to or ChatGPT is going to decide who the tokens go to, So the entire token economy is dependent on this person who's controlling how those new tokens are going to be allotted. And you can plug in Eigencloud into it and say, go verify Twitter, go use AI inference to double check that, you know, the people who are talking about it are talking about your token. They're talking about it positively. They're injecting new ideas into the discourse. They're getting new people on board it, whatever your criteria is.

David:
[43:42] High engagement, yeah. Or high engagement over low engagement.

JT:
[43:47] But all of that application logic is not something that a blockchain could reason with today.

David:
[43:51] Right. You just fundamentally can't hack that into an on-chain. Right.

Sreeram:
[43:56] And so this is one of the places where it's the most trust-sensitive. Why do you have a blockchain to know? What is the big deal with Bitcoin? 21 million Bitcoin, right? Controlling the supply. And it only goes to miners who are mining the blockchain. That's the core part of Bitcoin's logic. Same thing in a meme coin. I want to mine a meme coin by actually spreading the meme. And the new tokens are getting printed because of that. This is the core logic and you can't express it on chain, but you can slam it on top of Eigencloud. All the complexity that we just said, use AI to adjudicate like, okay, you know, it's being run in this way. It's only giving the tokens to people who are passing this AI check. It's only given to people who are having this engagement done. All of that run verifiably, but brought on chain with a level of guarantee, right? So this is a very degen, fun example of this premise that using an AI off-chain, but that brought the results verifiably on-chain, can actually massively change the surface area of what can be built today. Okay, so imagine in six months, meme coin launch pads, which basically have all these configuration features which say quality, quantity, like token inflation, like control all of that. And it's perceptible and verifiable to a user says, yes, okay, I can just go launch these meme coins.

Sreeram:
[45:22] But why only meme coins, right? Why should an eigen have a similar thing? Like, why wouldn't we want to have a portion of the supply be allocated to, like, people who propagate the meme of eigen? Maybe we do. Maybe we want to do it. Maybe Optimism wants to do it. Maybe Solana itself wants to do it. So this is not, because once you make something programmable and verifiable, it can go deeper and deeper into the core logic of what you can touch. What is this, the holy cow? what is the sacred thing that you wouldn't touch otherwise? It's token supply, right? And so you would touch it if it is verifiable. And so that's an example. And so here, I want to point out one more thing about like the trust model. People think of Eigen as economic security, right? So, okay, you're running something off chain. How is it back? It's back because people are putting up money behind it and promising that this thing is run correctly, right? And that result is brought back in, and then that's what's going to now control the inflation of these tokens.

Sreeram:
[46:23] But when you do this, you only get economic security or how much of it is being staked. You don't get more than what is staked. So the first thing is for what is staked, we not only have slashing that you can slash the bad parties, you can redistribute it to the harmed parties. So an app can say, hey, I'm slashing the guy and I'm going to distribute it to the token holders because they got harmed because the token got inflated and went to random guys.

Sreeram:
[46:50] That's the first thing. The second thing is you can not only do this with economic security, but like an optimistic roll-up, you can wait for a settlement period. So in an optimistic roll-up, what happens is you make a claim, and that claim has some stake behind it, and then you wait for a lag for anybody to challenge it. You know, because in a permissionless blockchain, censorship-resistant blockchain like Ethereum, which is what we are rooted on as Eigenlayer, anybody can come and slam a thing saying, hey, you know, this claim is fraudulent. And so if it is fraudulent, it gets adjudicated and then like, you know, and the results come up. So if you wait for that adjudication period or the challenge period, you can actually get much higher security than the amount of stake. Because if nobody is challenging or if it's challenging, the results are already, the correct execution is brought in, you can actually, it's not just economic security. It goes far beyond economic security. So that's, it goes to roll up like security. So you can build not just economic security models, but roll up great security models. And there's a bunch of technicality underneath it I'm not going to get into. But this meme coin example is a great use case for thinking about, because I'm touching the root of trust and supply, you want to be thoughtful about what you can do there.

JT:
[48:09] It goes back to 2018, 2019, when all the big public clouds were looking at blockchain. They saw the potential of what smart contracts could be used to adjudicate. And there was Super Bowl ads on, you know, trusted track and trace supply chains from IBM and all sorts of stuff, if you remember that kind of deal. Yeah. I was at Microsoft at that time, and I remember Satya Nadella and Takeshi Nomoto, who's now the CMO, they were like, yeah, it's great. Like, this is a really interesting technology that could open up a new application pattern, like a whole new category of sort of applications. However, one, these things are too slow, scalability to your point, David. And two, the point of failure of any of these applications is when you have to trust something that happens off-chain. Right. So when you have to write something on chain that comes from an IOT sensor or an ERP software system or something that you are taking valuable action on

JT:
[49:04] or you're triggering a smart contract on that you don't, it doesn't have that verifiability. And so I think what Suram is and the Eigencloud tries to do is extend the sort of barrier of what can be verifiable on chain. So I think the MemeCoin platform is a great example, a great DGN example. But I think it hints at something that like, in order to do what you were just laying out there, Sri Ram, the blockchain would have to reason with a whole host of inputs that it just can't reason with today.

Sreeram:
[49:33] It can't reason with the inputs, like how do you know what's going on in Twitter? How do you know what's going on on TikTok, right? So you need the interfaces to be verifiable, number one. Number two, not just the interfaces, the computation, like we were talking about programmability constraints, right? How do I write like the latest LAMA agent inside a blockchain, right? Like they're not meant for it. I don't want to write LAMA in solidity or, you know, to write it at a C-level program. I want to use the latest GPUs and slam the thing on top and say like, okay, you know, I don't want to touch like TensorFlow or PyTorch or whatever programming environments are used to write this AI. But EigenCloud will let you do that while maintaining verifiability. So that's a big unlock. JD was talking about Satya. So an episode comes to mind that I want to hint at here. So there's a recent podcast episode in the Dwarakish podcast. Satya Narala, CEO of Microsoft, is talking to him. And Dwarakish says, hey, Satya, do you think AI is like a new kind of species?

Sreeram:
[50:41] Is AI a new kind of species? And then Satya pauses and then he says, yeah, it's a fascinating idea. Our Microsoft's AI CEO, Mustafa Suleiman, Satya says, he thinks the same. He thinks it's like something like that. But I think it's not yet there. And the reason he gives is not because it's not intelligent or powerful enough. The reason he gives is that AI agents today cannot own property. The legal infrastructure to own property, to take liability, to start companies, to to engage in contracts. This stuff doesn't exist. So it's always an agent is acting as a delegate of human agency, right? It's Sriram having an AI agent, right? So it's, you know, I can give my wallet ID to like an AI agent, or I can give my Stripe API to an AI agent. And this is right now, like the AI agent meta in crypto is like, oh, AI agents are going to trade on chain. Yes, that's through, but AI agents are also going to use a Stripe API to go and trade like other places. What is unique that blockchains provide that simply doesn't exist is it's a substitute to this whole legal infrastructure. It's a new form of property rights, right? And when you think about it like that, what is a smart contract, right? A smart contract is a program owning property, right? So let me just state it again. A smart contract is a program owning property.

JT:
[52:10] Right?

Sreeram:
[52:11] This is a program that can hold digital assets under its custody. That's what a smart contract is. So if you think about, if I can put an AI agent as a program, it can hold property, it can start DAOs, it can do all kinds of things. The missing link is how do you put an AI agent into a blockchain?

Sreeram:
[52:30] That's the missing link, not anything else. Everything else already exists. That's what smart contracts are. That's what Turing completeness is. What is missing is the programmability to put the AI agent as a program, slam it into a blockchain. That doesn't exist really. Imagine like magic box, Eigencloud, you throw that AI agent onto the Eigencloud. It's a program that's verifiable and it holds property on chain. Now you can create fully sovereign agents, kind of like the end game of the AI agent meta is fully sovereign, fully verifiable agents whose entire supply chain, the input data is verifiable, the program is verifiable, the inference is verifiable, the tool calls are verifiable, everything is verifiable. And you can do that with the cloud. And so part of like why we went with this cloud, eigen cloud branding, is if you look at what is the biggest thing going on in cloud, in the same interview, Satya says, the biggest investment Microsoft's making this year, Azure is making this year as a cloud company, is $80 billion into AI, right? This is the biggest thing in the cloud. What is the biggest demand for computation is AI. It's the same thing for us as Eigencloud. You know, it's a general computation platform that you can offload all kinds of logic into. What is the biggest demand for that? It's probably going to be AI. And so like, that's why we're using that as a routing example.

Sreeram:
[53:58] But you can offload all kinds of computation, and we'll talk about some of the other examples.

David:
[54:02] Certainly, yeah. It's probably worth just emphasizing the notion of blockchains as asset ledgers in this context, where like, what are blockchains really good for? You know, moving tokens from person A to person B, from Alice to Bob, under conditions, right? And like, you know, Uniswap and AMM pool is like a condition. And that's good for DeFi. You know, DeFi makes sense on chain. But really, all these things are the means for swapping assets between each other, between parties, plus conditions upon those assets being transferred.

David:
[54:33] And really, I thought that's really all they're for, is for just settling assets between people. And this whole idea of an off-chain container where Sri Rambo is going to go slam an AI into this off-chain container allows for just more expressive conditions. Right. Like what more expressive conditions can we create rules for for the logic, the settlement logic between A and B? And like, you know, it's worth noting that, like, you know, Polly Market is a good example of those. Like, let's talk about some some conditions upon the ways that like assets will be will be transferred from party A to party B. You know, will, you know, Donald Trump beat Joe Biden in the election? Well, if if this, then that. And that is actually all written off chain. And then really all the Polymarket does is like encode a few things on chain in order to be verifiable and then assets gets handed off. And so like the idea here is like, well, AI and AI thought and the ways that AI thinks and what AI's think about can all be something that can be put into a container and made verifiable, verify that we know that the code is the code and that's what EigenDA is for. That's the substrate that we are allowing to be able to like It attests to the verifiability of the code being run.

David:
[55:51] And then the output of some AI and an LLM can be turned into logic that results in assets being transferred from in a particular condition. So it's like a pretty, it's a pretty broad, I mean, boundless, I'll even say, like category of like what can be put into a container. Imagine if your checking account and DeFi wallet finally spoke the same language. That's Mantle Banking, an all-in-one fiat and crypto account. It lets you save, spend and invest all from one dashboard. Swipe for coffee, stake ME3 yield, or even use virtual cards for payments through Apple Pay. So it feels Web2 simple, yet stays Web3 sovereign. For allocators, meet Mantle Index 4, the S&P 500 of crypto. A tokenized, institutional-grade fund, seeded with $400 million from the Mantle treasury, and balanced across Bitcoin, Ether, Sol, and yield-enhanced stables. One asset, broad exposure, pure DeFi composability. The momentum is real. ME faults, FBTC bridges, and a $2.4 billion community treasury are all powering the next phase of on-chain finance. Mantle brings real-world access, yield, and utility to digital assets. Ready for the next era of on-chain finance that actually belongs in 2025?

David:
[56:56] Explore Mantle at mantle.xyz or follow mantle underscore official. Mantle, bridging TradFi and DeFi so you don't have to. Binance is the world's number one crypto exchange. Over 275 million users already trust their world-class security. Binance makes starting crypto as simple as it should be, whether it's learning about crypto on Binance Academy or browsing hundreds of assets and viewing your newly created portfolio in a clear, easy-to-track dashboard, Binance helps you go at your own pace. For hardcore traders, Binance Pro opens up industry-leading services for trading professionals with fully bespoke trading products, along with a suite of white-gloved services for VIP and institutional clients. Need support? 24-7 customer service is on hand whenever you need it. And with some of the lowest fees and deepest liquidity in the market, it's no surprise why over 275 million users trust Binance for everything crypto. Download Binance today and get started in minutes. Binance is not available in certain countries, including the United States. Check its terms for more information.

David:
[57:51] JT, the big reason why I wanted you to come into this episode is like you're head of marketing, which is a pretty big job to be head of marketing for Eigenlayer because it's a very complicated, you know, pop open the hood and there's a bunch of components. So your work is cut out for you. But, you know, today Eigenlayer goes live. It's announced to the world. What is the next like week, month, quarter, year look like for you as head of marketing? What does that look like? How do you take this and to whom? how do you take eigencloud and how do you just like tell the world about it? Who, like, Sri Ram talked about AIs first, but I'm sure there's so many other categories. Like, what's your job look like?

JT:
[58:24] Yeah, it's a good question. The difficulty of it was why I joined the first time I met with Sherem. I realized what he's attempting to do is build the first kind of crypto native cloud, right? He wants to build a cloud that's backed by restaking. That's the... The stick that under underpins it and also the carrot for the stakers and the operators, but a, a crypto native cloud that has slashable services that can build these, these new types of applications. And that's, that's, that's not going to be something that is understood overnight. And so I think the first week, month, quarter for, for us is just highlighting what can be built and highlighting what is being built. And so you highlighted, polymarket and prediction markets.

JT:
[59:14] There's a lot of innovation on top of eigenlayer, polymarket included, that we think these new eigenlayer primitives can open up new products for someone like polymarket. I think AI is going to continue to be a really useful pattern for us. When I went back to, in the Azure days, we were talking about this on our exec meeting this morning, that the internet of things was a really complex topic to sort of explain. And what you needed was sort of one pattern that could, people could say, oh, I understand why you would need to get sensor data off of a thing and then react to it. Remote monitoring of an elevator or a jet engine or whatever, and then predictive maintenance on that thing because it's an expensive thing to replace. So I want to make sure I know what's going on with that machine all the time. And then when it breaks, I go fix it. And so I think as we start to engage with developers, there's going to be new sources of inspiration for what can be made verifiable on chain. I think we're going to, we need to recatalyze some of that creativity that we talked about earlier on the call, that channel, that energy of builders circa 2017, 18, 19, where we were thinking about what could be disrupted on chain, what could be remade and re-imagined on chain.

David:
[1:00:29] I mean, we were thinking about what couldn't be disrupted on chain.

JT:
[1:00:33] In that area, you didn't know, right? In that area, you hadn't reckoned with the sort of scalability challenges of the underlying chain, nor the fact that the on-chain VM was as restrictive as it was. It was sort of unbounded creativity. Right. And so a lot of the marketing campaigns, the stuff that we've talked about doing is going back with the developer communities at base and Arbitrum and Optimism and Linea and all these. There are brilliant, talented developers all over crypto that have great ideas, but they're just capped. They can't do what they want to do. And we want to go sort of redefine the design space for those developers. You talked about, you asked where we're going to focus first. We could probably go to Web2 and go to just traditional developers and say like, hey, there's this new cloud and there's these new primitives that you can build verifiable apps for. And that would require sort of a two-hop, a two-step sale, right? A two-step pitch where we have to pitch them on verifiability and then why this thing looks like the clouds that they're used to building.

JT:
[1:01:36] Or you can go to the developers that are frustrated by the limitations of the VM today that know it intimately, that know that it has limitations that, and if they could just offload the logic and be able to interact with normal software libraries or normal API calls, they could build some really cool things. Those are the ones, those are the devs that we're going to go talk to first. And so we want to go talk to, you know, Jesse at base and the EF and all these great developer communities, not just in Ethereum, across across chains and and kind of recatalyze that that ambition and that creativity around what crypto can do. Because I think Eigencloud extends the reach of crypto far beyond DeFi to a whole host of new applications. And that's that's the most exciting thing for me.

David:
[1:02:20] Sree Ram, when it comes to just like the size of the opportunity here,

David:
[1:02:24] I think we all know like the notion of the cloud is large, big, big market over there. How do we know what the TAM, the market size is for Eigencloud, like a cloud, but with blockchain properties, like, because it could be, you know, Eigenlayer has this big like story that it like fell into when it got launched. And now the actual like existence of the product is actually here. But like, is this, how do you think about the TAM here? Is this like a $1 billion TAM, a $10 billion TAM, a trillion dollar TAM? How do we, what do we know about like the potential of what this market even looks like?

Sreeram:
[1:03:00] Yeah, if you think about like what the cloud did, basically it made, you know, most of the economy programmable, right? Pretty much anybody doing anything, like if you're running any business, you're using the cloud in some way or the other. Either you're directly like using some application of the cloud or like you're using an application that is built on the cloud. So every man, woman, child has, you know, is using the cloud in some way or the other. So that's what the cloud did by making it very programmable, modular. People can build like lots of things one on top of the other. And what we want to do and what the crypto project is, and we just see ourselves as like an accelerant to the crypto project is to make the economy verifiable, right? So if you, if the more of the economy that is made verifiable, the more permissionless it becomes, because once it's verifiable, I don't need to know. oh, it's David or Sriram or JT going and doing this. It's anybody can do this. And so once that expands, actually what you see is

JT:
[1:03:55] The verifiability is the input permissionlessness is the output, right?

Sreeram:
[1:03:59] So many more people can build all of these things without them having to be trusted. Because once you need to trust them, you constrict it, right? So why is JP Morgan the only one that can offer derivatives or, you know, whatever narrow class of Wall Street firms? It's because it needs to be verifiable. And like what DeFi did is say, no, like anybody can write a piece of code and that code is going to be verifiable. So you don't need JP Morgan. Like a kid from, you know, Goa in India can actually write a piece of code and like people can slam things on top. And that made the DeFi economy big. Now doing that for any piece of code means basically you can rebuild the trust layer of society and like think of, you know, what the, what percentage of the GDP goes into, the GDP goes into like the trust services like legal, accounting,

Sreeram:
[1:04:51] You know, compliance, all of this stuff, if it's programmable on chain, how much frictions can it reduce and how much market opportunity it can create? I think we as crypto should be envisioning a portion of the entire GDP rather than like starting from inside out, like, okay, you know, what is the addressable market? So that's our, it's always our vision. Like we say, humanities coordination engine inside eigen. And like part of the vision is to see how we can move more and more of the economy, make it more of it verifiable. And so, yeah, if the cloud unlocked tens of trillions of dollars, like that would be our ambition for like crypto to unlock, you know, tens of trillions of dollars with this kind of, you know, verify everything approach.

JT:
[1:05:38] Yeah, another way I've heard you talk about it, Suram, is that the first verifiable app effectively was Bitcoin. And that's a $2 trillion asset class that disrupted fiat money. And then Ethereum extended the programmability of Bitcoin and opened up other types of programmable money, programmable finance. And that's another, call it, trillion-dollar industry that's been built on mostly Ethereum, but now on Solana. And so what happens when you extend the programmability? Again, David, to your point about step function improvements to technology, if you extend the envelope of what can be verified beyond the sort of narrow application of finance and token-based money to anything, then it's probably somewhere between what DeFi has done today and what the cloud has captured over the last kind of 20 years. That's the sort of long-term market opportunity.

David:
[1:06:30] JT, Sriram, I know it's been a slog getting Eigenlayer over the finish line. So congratulations on it actually finally getting released. Who can go play on Eigencloud today? Is this for developers? Who is this product for right now? How do people get their hands on it? What does that look like?

JT:
[1:06:48] So this is for any developer building on any chain in crypto today that wants to build more mainstream applications or work with AI on chain. And so EigenCloud is available today. DA is a mainnet service. EigenLayer is obviously the foundation on which all this is built. But we've got some exciting new services, I can verify and I can compute, that are rolling out in private preview this morning. So you can go and apply for access to private preview and start building with those today. And we lay it out on a white paper where you can dig into in a lot more detail, but those services are going to be available on mainnet towards the end of the summer. And we've got a lot more third-party AVS value, really, really innovative teams building on Eigenlayer, as well as some first-party AVSs that we're building

JT:
[1:07:33] in-house that are going to be super exciting coming out later this fall. So more to come, but you can go to eigencloud.xyz and get access to all that.

David:
[1:07:41] You know, I think your guys' biggest challenge is actually teaching developers how to think bigger or think in different terms. This notion, like with Sriram's like MemeCoin example, like why did MemeCoin become so big? Well, like PumpFun really like processed and like turn it into like a roof people structure via the verifiability of the total supply for every MemeCoin. But then with Eigenlayer, with Eigencloud, you can have this verifiable way of inflating the token in a way that the inflation only goes to like productive output of whoever was like contributing to growing the beam. That's like a non-intuitive way of like using eigenlayer. But nonetheless, it is like, obviously it was a great example. And so I think your guys' biggest challenge is like, how do you teach people to think in these new terms, to think with these new tools that you guys have created, where it's like, it's so boundless, right? Like it's just, it's a container. It's an open container you deposit code into. And then using eigencloud, using like all the modules, the AVSs, whatnot, all of a sudden there's like imbued with properties there. And so it's so boundless that I think your guys' biggest challenge How do you get people to think in these new terms, think in these new ways to open up their

David:
[1:08:52] imaginations or what they can build using Eigenlayer? I don't have the answers for you there, but I will kind of leave the listener with it's probably thinking bigger is probably going to be the biggest challenge to opening up the possibilities of what Eigenlayer can create.

JT:
[1:09:08] Yeah, I agree. I think we talk a lot about in our meetings about making crypto ambitious again. How can we help developers see verifiability as a tool which they can be, David, to incumbent Goliaths? How can they go compete in new ways? I think we've spent a lot of time in the last probably 18 to 24 months talking about the mechanism of Eigenlayer and how it technically works underneath the covers with these long white papers. And the challenge of probably the next 12 months in particular for me is indexing a bit more on the magic of Eigenlayer and what you can build with it.

David:
[1:09:45] Well, that's an exciting place to be. JT, Sriram, thanks again for joining me and sharing everything about EigenCloud. Thank you so much for having us. Cheers. Bankless Nation, you guys know the deal. Crypto is risky. You can lose what you put in, but nonetheless, we are headed west. This frontier is not for everyone, but we are glad you're with us on the Bankless journey. Thanks a lot.

Music:
[1:10:11] Music