A Conversation on Full Stack Decentralization

Listen in for an excellent discussion between three leading experts in Web3, including STORE co-founder and CEO Chris McCoy:

Link to the Twitter/X discussion

‍

Transcript of Conversation on Full Stack Decentralization

Host: David Attermann

‍Participants: Chris McCoy (STORE), Derek Anderson (Blockless), Andrew Smith (Versatus)

‍David Attermann:

I'm David Attermann, Senior Portfolio Manager at M31 Capital, and I lead liquid and illiquid Web 3.0 investments. So basically infrastructure and middleware. Today, wanted to round up the smartest people I know in Web 3.0 infrastructure, and talk about full stack decentralization and it's not something that is talked about a lot, it's not. I feel like there's a huge education gap as to how this infrastructure works, who's actually running what, and the limitations that exist today. And so I want to both educate people on what full stack decentralization means, but more importantly why they should care. And why I think, and I think everyone on this panel thinks that it is structurally superior to Web 2.0 and centralized systems, and beyond kind of ideologies, beyond cost savings, beyond all that, just focusing on incremental and novel functionality.

Stuff that you can't do on Web 2.0 rails that Web 3.0 will power. And so that's the point of the space is just context, the fundamental reason why I bought into Bitcoin in 2014 figuratively and literally, and the reason I joined Web 3.0 full time in early 2021 is because I think decentralized and trust-minimized networking infrastructure will power really cool novel utility. And I'm like a student of disruptive technology, spent 10 years in traditional finance covering Web 2.0 networking infrastructure companies looking for the next technology to disrupt incumbents. And so for me, this is a no-brainer and kind of where data networking is headed in these distributed systems, composability, all of that, but it's not until we have real tangible products that demonstrate this technology, that's when people will care and that's when the value will be more apparent. So today have three teams building full stack decentralized platforms and each actually doing it in a unique way, which is really interesting.

And so would love them to introduce themselves, tell everyone why you decided to build your project, what is the ultimate goal and vision? And what is the ultimate value and utility? Why should people care? So let's start with Derek Anderson, CTO of Blockless Network.

Derek Anderson:

Hey, hey, hey, thanks for having me on board today. I'm glad to be part of this conversation. So as David mentioned, my name is Derek. I am CTO of Blockless Network, and we view decentralization in today's kind of legacy decentralized apps as kind of a problem. So we do know that the back ends or the state layers have been decentralized for a long time, but what we really feel is missing is that the middleware to the end user has really been, for a lack of better word, really focused on. So today's end user is really maybe getting decentralization out of RV or IPFS or some other kind of CDN based decentralization, which is still pretty good. But there's a whole set of tooling and tasks that a developer needs to do. And what we've really seen is that that tends to be pigeonholed into a single server or maybe a few servers where the redundancy is then managed by the team itself.

And this is not the same kind of parallel trajectory that we've seen in the back end where you can launch a smart contract and then kind of set it and forget it. And your decentralization and distributed technology is all taken care of for you. And for Blockless, that's really what we want to focus on. It's allowing developers to write middleware and end user applications that are much like traditional web apps, and then taking care of the fault tolerance, the distribution and the task management for them, really recreating what a Web 2.0 Developer kind of does today, packaging that as a Wasm assembly that is architecture and platform-agnostic, and taking care of the distribution across several different types of topologies, whether that is in the data center as a node operator or even on the web browser as a background task.

And that's where our magic sauce is, is really the distribution mechanism that comes into play. We do novel distributed reputation keeping using an Elo Arpad score, if you've ever played chess or a PVP game. And we combine that with a NASCAR pole ranking position so that we can ensure that compute meets the facilities of the developer, whether that's by cost or performance.

David Attermann:

Great, great intro. And one thing before I let Andrew introduce himself, I initially was very into still [inaudible 00:05:43], but decentralized compute and these decentralized compute marketplaces, permissionless access to compute. And the difference here is these platforms are taking it to a much more complicated level where they're orchestrating and coordinating a full stack. Think of it as AWS, fault tolerance, service level agreements, and getting all of these decentralized nodes to really act harmoniously and give at least the same Web 2.0 Experience. Obviously, I think there's more utility there, but it's a much harder engineering problem than just kind of these like decentralized CPU or GPU network. So I wanted to get that out there, but Andrew Smith, CEO, founder of Versatus.

Andrew Smith:

Yeah, thanks. I'm Andrew. We're building Versatus, which is a peer-to-peer cloud built on EigenLayer. Very similar I think philosophy to Blockless. Mike and I are buddies. I have a ton of respect for the Blockless team, I think what they're doing is fantastic and a step in the right direction. And what we're building is very similar. One of the big differences is that we're not Wasm focused, and we're not browser-based focused, and platform-agnostic. We take more of a server-based approach. Hence, the relationship with EigenLayer why that makes sense because they bring sort of this decentralized data center and trust bank to the market that allows us to build what it is that we're building on top of them and inherit some of the primitives that they provide. So what is it exactly that we're building? We're really trying to fill a gap of three core cloud services.

One is elastic compute, so scalable virtual machine provisioning and management. The other is functions as a service, so sort of your Lambda equivalent. And the third is web hosting, and basically web and mobile app. Dynamic content hosting is the way that I typically refer to it if we want to get technical. And so the way that we've built this, the entire approach is really to try to route around DNS. DNS at the end of the day is still going to be your ultimate choke point. DNS is centralized, it's regulated, it's managed by a monopoly. And if you don't get around DNS, nothing can truly be decentralized. The RPC provider that you use today to interact with your smart contract on Ethereum is still centralized for the most part. There are some decentralized RPC providers, but the funny thing is most of those decentralized RPC providers really just allow you to run an RPC node inside of an AWS virtual machine on EC2.

So whether it's infra layer, middleware layer or application layer, we think the entire stack needs to be decentralized. What our goal is, just to be completely frank and transparent, is we want to do for applications what Bitcoin did for payments. We want to make them unstoppable. We want to make them extra legal and extra judicial and make it to where your access to the internet and the applications and information that you want access to, is not at the whim of some petty bureaucrat. If you want access to information or applications, you're able to access that information and applications.

David Attermann:

Awesome. Great overview. And last but not least, Chris McCoy, founder, CEO of Store Cloud.

Chris McCoy:

Hey, morning, David. How are you man?

David Attermann:

I'm good. I'm good. Thanks for coming on.

Chris McCoy:

Yeah, I'm heads down writing, which is a tense activity, so I apologize for being a little delayed, but how about you let everyone know what we're doing and then I'll clean it up.

David Attermann:

So the way I would-

Andrew Smith:

Wait, why didn't you tell me that was an option, David?

David Attermann:

Well, it's not, but I'll let this one slide. And for everyone listening, Chris and I have been internet friends for years, talking through decentralized infrastructure. Three years ago when no one even knew what decentralized compute even meant outside of blockchains. We've been going back and forth and helping each other think through the longer term vision, but Store Cloud, I think the best way to think of it is it's an L-one blockchain. So it has a single settlement layer, shared state, but it connects directly with off-chain cloud resources, CPU, GPU, storage, memory, all of that. And it's kind of like, and I'm sure Chris is going to correct me here, but it's kind of like, I think of it as ICP, but with the cloud resources off-chain. And so arguably more efficient, there are pros and cons to everything, but that's how I think of it.

And so, it's like all of these projects. It's a one-stop shop for all of your application needs, front end, back end, any off-chain computing resources, and then also access to blockchains and the Web 3.0 part of all this. So that's how I would describe store. But please, Chris, tell me where I'm wrong.

Chris McCoy:

I appreciate that. No, it's been a lot of fun working through the problem space with you. And obviously we've been working on an experiment for some time, attempting to decentralize and really democratize cloud computing infrastructure, such that it's owned and funded by monetized, and governed by the world, not by big tech. And we would classify ourselves as an L-zero where L-one is integrated. You could sort of use the [inaudible 00:12:09], but we've always been marching toward B, looking at where decentralization is headed, why it's needed, what does that unlock for the world, what does that mean for the world, and how to scale that out in the lowest cost and most efficient way. And so we've really almost atomized computing resources, the cloud layer, and have transformed them from a dollar-based economy into a token-based economy, creating our own virtual machine and enabling ultimately where we're headed with stores.

If you're a developer, whether you're building a full performance application or an AI agent, or eventually a machine or an FT, you can build on store where 95% of your workload is likely off-chain in our trust minimized cloud that is governed by store democracy and incentivized by store's monetary system. And from there, you can issue tokens onto any chain, including store chain seamlessly bridging together cloud and block space computing into a single protocol, bringing in the $550 billion plus cloud industry into the potential revenues of decentralization. But if the experiment plays out the way we're designing it for and building it for, then we are giving the world the economics of infrastructure, not just five companies. And from here, we can almost set data free. And what happens with AI gets really interesting when the infrastructure layers truly trust minimize, and there's a formal and efficient governance structure around that that's affect democracy. So our experiment is interesting, and yeah, it's been fun. Like I said, discussing this with David throughout our journey.

We're getting closer, and for the most part feel fairly confident in our implementation. So you caught me the middle of a... I feel like Ernest Hemingway or Thomas Jefferson on some level. Been writing basically for the first time, finally specifying a research in more formal structure, a bit of a white paper meets a constitution, and...

Andrew Smith:

It sounds like it would be well classified as a manifesto?

Chris McCoy:

Maybe, it's quite philosophical. I walk to the Golden Gate Bridge every day and think through, how our founding team sees the future and how we get there. And ultimately, store is an incentive. It's a series of incentives. Our economic consensus mechanism is what we call dynamic proof of stake. And what it does is really unique to the space of decentralization. It first enables what we call leaderless economics to emerge, which they were tried by Algorand, they were tried by Silvio there and then tried them. They did not get them right within the block space layer. We have, but what we recognize is that as a consensus mechanism is applicable, not just the blockchain layer, but also to the infrastructure layer. And so it can incentivize infrastructure to emerge in a trust minimized equitable way. And blockspace vertically integrates in as VMs, which ultimately that's all it is technically.

With that said, it's almost like Uber surge pricing meets an economy that incentivize and governs infrastructure, and ensures that the work that you put in you get paid for. Whereas in leader-based systems, it's essentially contest and a raffle, a slot machine. Different models of, we'll call it compensation, but it's variable. And what we're working on is a predictable economic system where the security you provide to the protocol determines the proportional revenues you receive from the protocol. Whether that's fee-based or inflation-based. That also means you're responsible for that proportion of costs. So we call it leaderless so every participant gets paid for work, every work session, more or less maps, more so to the real world economy. And then the second breakthrough there is just the ability to not just incentivize blockspace computing, but incentivize infrastructure layer zero. And so it's the first consensus mechanism from an economics perspective that enables us to do that.

So we're bringing these together in a seamless way, and it is a very interesting experiment and done correctly. It results in a fairly efficient cloud, a hyperscale cloud for developers to build on. And where it gets really interesting from a virtual machine perspective, from economics perspective is, we're headed with it is if you're building on-store and not on AWS or Azure, is you can utilize your own token economy, your own token that you may have issued on Solana, you may have issued on Ethereum, maybe even on store chain itself when store chain comes online. But that token can be used to pay for your computing infrastructure, for your cloud resources. And so we're going solving that economic abstraction problem, that if Vitalik, in our view, made a colossal error on in his internal initial design, which we've always believed from the very beginning, and I got started in a space in 13, building on Bitcoin, and I met the Ethereum team really early.

They recruited me to build, and I couldn't understand if I issued a token and the miners only want to get paid in gas and ETH, why would my token ever prove real value? There's no monetary premium, no demand for it as a money. It's effectively a loyalty point. And so...

‍

‍