[00:00:03] Harrison Hines: DePIN, it is a natural evolution of what Uber and Airbnb sort of started.
[00:00:08] Kyle Okamoto: Some people have called GPUs one of the most precious resources in the world.
[00:00:13] Harrison Hines: Over the next few years, you’re going to see cloud platforms that are built on clouds like AWS.
[00:00:19] Kyle Okamoto: There’s a very large social mission that we all have to take on here to improve how physical infrastructure networks have been consuming resources.
[00:00:27] Elisha Owusu Akyaw: Hello, everyone. Welcome to Hashing It Out. This would be Episode 62 of Cointelegraph’s flagship podcast. My name is Elisha, and today we are back to a topic that we have spoken about a couple of times in the last couple of episodes, simply because it’s the hottest, one of the hottest narratives in the Web3 space. We are talking decentralized physical infrastructure networks, also known as DePINs. I have two amazing guests here. I have Kyle and Harrison. Hello, Kyle. Welcome.
[00:00:57] Kyle Okamoto: Thanks, Elisha.
[00:00:58] Elisha Owusu Akyaw: And Harrison, welcome.
[00:00:59] Harrison Hines: Thanks. Excited to be here.
[00:01:01] Elisha Owusu Akyaw: Great. So, Harrison is the CEO and co-founder of Fleek Network, and Kyle is the CTO of Aethir. Both companies are doing some great stuff in the DePIN sector and space, and we’re going to be having a long conversation about what they’re working on. But before we do, I’m going to ask both the most boring question of the podcast, which is how did they end up in Web3, and how did they end up at the projects that they’re currently working at? I’m going to start with Kyle.
[00:01:30] Kyle Okamoto: Yeah, sure. I mean, I’ve always been in large-scale enterprise hyper-scale infrastructure. Back in the day, we would think about how to optimize and get more efficient and effective, and blockchain was a great technology to do that for things like billing and logging and mediation and operational control. Same thing for using artificial intelligence. Really started out from like a cost savings perspective of how do we use AI to comb through, you know, massive amounts of data and transactions and pull out information that can be acted upon out of that information or out of that raw data. One of the use cases that we actually started with was around streaming media, and I see that the Fleek Network does a good job on decentralized static and dynamic content delivery. That was one of the very first use cases that we implemented. I later then got hooked up with Theta Labs, who was a pioneer in the peer-to-peer streaming media space on a decentralized platform and was very blessed to be an early adviser and investor to that original DePIN. And then, you know, continued from there.
Been with Aethir for a while. Aethir has morphed quite a bit, but we really started out in gaming. Started out in both cloud rendering of online games for both mobile and PC, thinking about how to allow gamers to play the same experience that they would on a console, or downloading a game locally on their $4,000 computer without having a $4,000 rig, and doing all of that in the cloud. So, that’s where Aethir started, and then continued to add additional GPUs and additional use cases, whether that’s virtual phones or running a training model. That’s kind of how we’ve evolved over the time, and looking forward to the conversation.
[00:03:10] Elisha Owusu Akyaw: Sounds really good. Harrison?
[00:03:12] Harrison Hines: So, basically, I entered the space doing a project called Token Foundry, which was essentially like trying to bring in some standards and best practices for token sales and token sale infrastructure and just ways to go about launching networks. But what brought me there is prior to that, I had helped build a project called SeedInvest, which was doing equity crowdfunding, and it was an equity crowdfunding platform that was trying to take advantage of the JOBS Act and that new regulation around crowdfunding to make equity crowdfunding and sort of like democratizing venture capital and all that sort of stuff. But the regulations there really held back equity crowdfunding from ever taking off. And then, so when Ethereum launched, and the first use case that really showed promise was these sort of like decentralized crowdsales, that’s where I kind of had the idea that you could probably leverage these new rails to build a business like that, and I had a feeling it wasn’t going to be the Wild West forever.
So, that was kind of what got me into the space. And I met Joe Lubin and the Consensys folks. And so, I joined there to build that project. But then, after two years, a) ICOs were sort of falling out of fashion, maybe temporarily. It seems like maybe they’re making a bit of a comeback now after people, I think, are a bit fed up with the whole low-float, high-FDV reality of the current situation and the way these networks get funded. But so, in the end of 2018, I realized I wasn’t so interested in the decentralizing Wall Street narrative as I was in the decentralizing the web narrative. And I realized everyone was pretty much focused on the decentralizing Wall Street narrative, and not many people were focused on decentralizing the web. But back then, there wasn’t many technologies that looked like they had any staying power in the decentralizing the web side of things besides IPFS. So, that’s kind of where we started building products and use cases around what we saw people wanting to do with IPFS, one of them being hosting front ends on IPFS.
So, that’s kind of how we made our name. But then we were always curious, how can we actually move beyond just like static site hosting on IPFS and provide more full stack web services and infrastructure that could leverage decentralization or these other sort of benefits that Web3 sort of offers, but in a more meaningful way. And so, we had started building on different protocols like Filecoin and the Internet Computer, and tried to figure out where there was opportunity and value and real benefits that could be provided to developers. But then, after doing that for like two years, we realized that we thought there was an opportunity to take a slightly different approach and leverage all the new sort of approaches that were being taken to build high-throughput, scalable blockchains and apply it to a web infrastructure use case. And that’s kind of how we landed on Fleek Network and kind of the direction we’re heading in now.
[00:06:18] Elisha Owusu Akyaw: So, the key word of the discussion today is DePIN, because I’m sure we are all going to be making multiple references to that. Just to make sure that people who may be living under a rock and haven’t heard the word, and are able to catch up, what does DePIN mean, Harrison?
[00:06:37] Harrison Hines: The easiest way for me to explain it is by example, but it is a natural evolution of what Uber and Airbnb sort of started, where essentially, they realized that there were advantages and efficiencies to decentralizing the supply side of an infrastructure network. In the case of Airbnbs, it’s apartments or like hotel rooms, and with Uber, it’s cars or taxis. And so, what DePIN is, what those are, they’re essentially like centralized physical infrastructure networks. But not completely. They’re like sort of a hybrid where the supply side is decentralized, but they’re operated and owned by one corporation. And so, DePIN is essentially like a natural evolution where instead of that decentralized, let’s say, infrastructure network being owned and operated by one centralized authority or corporation, it can essentially be owned and operated, thanks to blockchains and sort of their ability to coordinate very well, it could be more community-owned and operated.
Where like in a centralized physical infrastructure network like Uber or Airbnb, the platform operator is kind of at odds with the, essentially, operators of the network, like the drivers or these people renting out their apartments. Where their kind of incentives are misaligned. The platforms trying to extract as much value from them as possible, and vice versa, the operators are kind of getting screwed because they’re at the mercy of this corporation. So, DePIN is essentially trying to correct those inefficiencies with the models like Uber and Airbnb, where you could then have more fair and transparent and decentralized ownership, where you can align the incentives and the interests of all the participants in the network much better. And so, you can actually have those operators in the network, also the owners of that network, and also just lead to further efficiencies by removing the corporate structure and then just allowing a blockchain to essentially handle those aspects that are typically handled by the corporation in a network like this.
So, that, to me, is sort of what DePIN means. It’s kind of building on this work that an Uber and Airbnb started as a better way to build these sort of physical infrastructure networks. And instead of having a corporation own and operate them, you essentially have a blockchain and sort of a token-powered network, which leads to what we feel is just further efficiencies and optimal outcomes.
[00:09:16] Elisha Owusu Akyaw: Kyle, same question. That’s if you have anything to add.
[00:09:20] Kyle Okamoto: Well, I mean, that was a fantastic answer. I would say that it’s really utilized to democratize access and to distribute ownership of physical infrastructure around the world to solve real-world problems, whether that’s finance or healthcare or energy or artificial intelligence. Like from Aethir’s perspective, we are an aggregation of enterprise-grade GPU compute clusters around the world. Those are actually owned by the community. They are staked by the community. They are then monitored and enforced by a community of 91,000 checker nodes around the world, owned by individuals like you and me. They are the ones that are ensuring that the quality of the network and the metrics associated with that performance are up to enterprise-grade. They’re the ones enforcing SLAs. They’re the ones administering the overall platform. And the rewards associated with those economics are staying within those same communities.
So, AWS and GCP and Azure, like those are centralized Web2 models of providing compute. They set the rules, they enforce the rules, and they reap the rewards. And ultimately, the users are the ones paying for that. Whereas in Aethir’s DePIN network of GPUs around the world, that’s all completely community-based. And I think that’s a major seismic shift that enables scalability, enables a completely different cost structure, and ultimately actually improves the efficiency by giving a massive amount of transparency and visibility. So, anybody in the world can see exactly what everybody is doing on this decentralized physical infrastructure network at any time.
[00:10:56] Elisha Owusu Akyaw: Great. Now that our listeners get what DePIN is about, the next question would be where is the traction coming from? Why is this such a hot narrative in the Web3 space right now? Kyle?
[00:11:08] Kyle Okamoto: Specifically for the use cases that we’re serving, I mentioned in the introduction, like, gamers want to game. They want to be able to play games and experience new types of games and experiment with different models, and they don’t necessarily want to go out to the store and buy a $1,000 console or a $3,000 computer with a nice GPU in it. They want to be able to go around the world and still play the same games and interact and buy things and trade things and sell things and monetize the hard work that they’ve put into those games. And they don’t necessarily want to have to lug around a specific piece of hardware to do that. So, they can now get open access to wonderful different types of gaming experiences and new experiences by leveraging our physical infrastructure network.
On the AI side, same exact thing, right? There’s a massive... I wouldn’t necessarily call it a shortage in GPUs, but there’s a massive inefficiency in how the current environment on the Web2 side is leveraging compute. It’s why Nvidia is a $3 trillion company and why there’s a six-month backlog of orders of GPUs. If you look at actually how efficient those are, on average, enterprises are only using 20% of that compute. So, they’re wasting a massive amount of space and energy and power and hardware and capital that can be leveraged by the community. And there’s a lot of companies out there and innovators and developers that just simply can’t get access to that hardware, nor can they get access at an affordable rate. And I think that by democratizing access and improving the efficiency of those cloud networks like Aethir, it opens up a whole new world of possibility for innovation and ideation and invention, ultimately. Because now very, very smart people can get access to the tools that they need to go build things. And I think that’s revolutionary.
Some people have called GPUs one of the most precious resources in the world. I know that’s hyperbole. Obviously, clean air and clean water and things like that are very important. But if you think about how much electricity and energy that these clusters are taking up around the world, if you can improve efficiency by 20% or 40%, a small number, and we’ve seen 80% improvements. But even if you just did a 20% improvement, that’s putting a massive amount of energy back into the system, which ultimately helps all resources around the world. So, I think that there’s a very large social mission that we all have to take on here to improve how physical infrastructure networks have been consuming resources. And that ultimately leads to how do you improve the efficiency.
And then secondarily, I think it’s about giving more access. I mentioned democratizing access. Regardless of your socio-economic background or your geographic location, you should be able to play games and enjoy yourself in this world, and you should be able to innovate and get the necessary tools that you need to build. And I think those are very valid reasons as to why this is such a hot space today.
[00:14:06] Elisha Owusu Akyaw: A really good answer. It’s one of my favorite answers to this question that I’ve been asking multiple times in the last couple of months. Harrison, anything to add?
[00:14:15] Harrison Hines: I think on the demand side, what I think is interesting about DePIN is that for the majority of use cases, a lot of protocols are building infrastructure and selling products into just the crypto ecosystem. But with DePIN, it’s kind of the first time you’re seeing protocols and projects sell into more traditional markets. And so, I think that’s why DePIN is sort of having a moment right now, is there are real benefits and real efficiencies you can get by building infrastructure networks in this way, and those benefits resonate with traditional customers in whichever the industry is. So, whether it’s performance or cost savings, open-source alignment or whatever the reason might be, I think that’s where you’re starting to see DePINs find success, is having real, tangible benefits over existing incumbents in large traditional markets where there hasn’t been a lot of innovation due to the fact of how hard it is to build these infrastructure networks.
Like you look at some of these use cases in telco or internet infrastructure or cloud infrastructure, it’s like these are pretty hard industries to break into because of how much capital you need upfront to even set up an initial minimum viable network. But now, with DePIN, you kind of flip that equation on its head, and so, it allows new entrants into these sort of large traditional markets that have been without competition. And without competition, it gives them very little reason to innovate. And now you have these new projects taking this new DePIN approach, able to enter these markets with real products that have real, tangible benefits over existing alternatives. And I think that’s starting to resonate, and that’s where the demand is coming from. And I think that’s why DePIN is sort of standing out now, because the usage is more real and it’s more tangible, as opposed to, you know, these purely digital or crypto-native sort of products and solutions and infrastructure.
[00:16:27] Elisha Owusu Akyaw: So, Kyle, I think your answer to the question on why the narrative is hot did mention a lot of was really centered around the need for specific use cases, and especially in places where there’s a need for computation power. What is the difference between traditional cloud computing and decentralized cloud computing?
[00:16:49] Kyle Okamoto: Harrison touched on several good points there. I think the industry as a whole is just simply not okay with an oligopoly, with the few controlling the resources for the many. And I know that sounds very philosophical, but what that actually means practically is that it’s hard to get resources, it’s expensive to get resources, and those resources are only consumable in a limited number of ways. It’s literally controlled by the few, and that access is hard to get. So, we see clients coming to Aethir that are paying five or six times more for the same GPU resources, the same enterprise-grade sitting in a Tier 3 or Tier 4 data center, backed by a 100% SLA that they can get from Aethir. But now they have more flexibility, to Harrison’s point. More control, more adaptability, and no lock-in. I think that’s what’s really been prohibiting and completely stopping innovation and ideation. When you have to spend so much time managing your costs and resource allocation.
We also see that clients that are just simply aren’t ready to commit to very large capital expenditures, specifically on the Web2 side. They just don’t want to be locked into one specific hardware technology. They don’t want to overcommit and be left with like an upside-down return on investment. And they certainly don’t want to wait in the supply chain lines for six months when they have an idea that needs to be executed now. And I think that’s ultimately why cloud, centralized cloud computing was born in the first place, is because they didn’t have to go out and buy a bunch of gear, find a bunch of space and power, rack and stack and cable and label, and then figure out how to operate that. And they could just simply go to AWS and spin up some EC2 instances. Well, you can do that on the GPU side, but you’ll quickly run out of your small amount of cloud credits, and your costs will start to balloon and scale out of control. And you’ll be left with, what do I do next?
And I think that’s really the challenge there that companies like ourselves are looking to solve, is how to give a scalable, flexible solution that allows for hardware evolution as the newest and newest and newest constantly come out without getting that vendor lock-in, so to speak. I think that’s something that the centralized clouds have done a really good job at over the last 15 years of figuring out how to create a business model that is very hard to move away from. Originally, it was all about cloud transformation, and now you’re seeing people that want to un-cloud, and they want to go back to a different model. And I think that’s what the Web3 decentralized physical infrastructure networks of today are really trying to solve, and help companies move away from that oligopoly structure.
[00:19:28] Elisha Owusu Akyaw: Since we are talking about decentralized cloud computing, let’s talk about the DePIN projects that we have in the space currently. Are they truly decentralized, or do they still have certain points of centralization? And how do we evolve from that as an industry?
[00:19:46] Kyle Okamoto: Yeah, it’s a good question. I would say that there’s a lot of different projects out there. Speaking for Aethir itself, we have been evolving. It’s not a revolution, it’s an evolution. And we’re constantly decentralizing more and more aspects of the platform. So, for example, in February, we launched our node sale and handed out the policing and the monitoring and the enforcement of the system to 91,000 different nodes around the world. We just launched our New Horizons program, where providers can apply to join our platform. But those providers are, of course, KYC’d and KYB’d to ensure enterprise-grade compute and SLAs can be met for our clients. And so, only certain providers are accepted. So, it’s not an open free-for-all to where anybody can just add compute to our platform. You know, we don’t take on MacBook Minis, for example, or a consumer’s laptop that might be affected by them watching Netflix or playing a game. So, in that sense, we are onboarding GPUs that we expect our customers to utilize in the ways that our customers want them to utilize. That’s where we focus.
Some might say, well, you’re not truly decentralized if only specific providers can add their compute. Eventually, it will become more and more, but really, it’s about how do we check and verify that all of this, the compute that’s being added to the network, can actually be consumed? Can it be verified, can it be validated, can it be enforced to be able to offer an enterprise-grade experience to customers and clients? So, yes, I could argue both ways. There is some sense of control. Take Harrison’s previous example of Uber. Uber makes sure that drivers have a driver’s license. They make sure that their cars are inspected and insured. Airbnb makes sure that the hosts that are providing rooms or villas or apartments or flats, doing constant quality controls and cleaning things, and if that doesn’t happen, then there are penalties associated. In Web3, we would call that slashing.
So, I think that there’s pros and cons to decentralizing things that make sense and then evolving those things over time. You know, we just launched our Aethir Catalyst program, which is our grant program. And ultimately, that’s governed by the community. It’s governed by the tokenholders as to how we invest in the community itself that is serving this, right? So, I think over time, you’ll see more and more, but I think it has to make sense in your approach and in your scope.
[00:22:15] Elisha Owusu Akyaw: Harrison, same question. Is DePIN really decentralized, or they are still points of centralization?
[00:22:20] Harrison Hines: Today, every single protocol or project in Web3 has points of centralization. And the fact remains, as long as these projects or protocols are accessed via DNS, there will always be some level of centralization because DNS is by far the hardest piece of the stack to decentralize. There’s two parts, I would say. There’s the core protocols themselves. That’s on a protocol-by-protocol-specific basis with how decentralized or centralized they are. And I’d say decentralization is very much a spectrum. So, it’s dependent on the use case in terms of what parts of the stack really do need to be decentralized versus not. Like if you look in the Ethereum ecosystem, like the smart contracts themselves being decentralized and running on a platform that is decentralized is important. The RPC endpoints that people access and interact with these contracts through, maybe not as important. Especially if you have multiple options and there’s no vendor lock-in, then maybe centralization at certain parts of the stack isn’t as critical.
But what I would say, and what’s very relevant to Fleek and Fleek Network, is if you look above protocol, every project is pretty much using Cloudflare CDN in front of their gateway or RPC or whatever else they’re using as kind of like the access point to the network. Most are using like AWS Lambda functions or other infrastructure like that on Web2 clouds within their stack, and most of them are hosted on Vercel or just another Web2 hosting platform. That’s really where we try to help bring more decentralization. Not just DePINs, but the whole stack is, if you look at the infrastructure that actually like encompasses these DePINs, a lot of it is just traditional web services on AWS that packaging into a node software and then having people run. And so, that’s where I would say we’re trying to make a lot of progress. And I think one area of focus where we think we can make a lot of progress, especially on some of these like above protocol and more like front end and access layers of infrastructure to these underlying protocols and networks is where we’re building TEE- or SGX-powered, trustless or verifiable web services.
So, one which we’re putting out content about this week, for example, is a trustless name server that leverages the new Edge SGX capabilities of Fleek Network. So, what’s cool about that is, essentially, where does decentralization matter? One point it matters is if you look at a lot of the ways these projects, and especially like DeFi front ends, get hacked, it’s because their GoDaddy account gets compromised, and then some hacker is just redirecting their app to some phishing version of their app and draining all their users wallets. And so, with the more of that infrastructure you can move on, chain... ENS is doing a good job. There’s a project called 3DNS doing a good job trying to get more DNS domains onchain. There’s a project called BOX Domains that’s like an onchain-native registry for .box, the TLD. And so, with this trustless name server, you could essentially like eliminate another point of failure where you essentially can allow these DNS records to be controlled by, let’s say, like a safe multisig or something like that. That’s cool, because then you can prevent hacks and you can ensure the front end and the name server is just doing what it’s supposed to do and you don’t have any third-party trust assumptions or centralized points of failure where someone could hack your GoDaddy account and change things. So, like that’s a very relevant piece that is pretty much applicable to all protocols as another step you could use to help decentralize or trust minimize your stack because now, instead of those records being in the control of some developer on the team’s account and control, you could kind of now put it in the hands of your DAO contract or your tokenholders, or some multisig with trusted people within the protocol or the foundation, things like that.
But I’d say, yeah, like today, the bottom line is everyone is centralized in some way, shape or form. But now you’re starting to see a lot of projects who are addressing, piece by piece, these different parts of the stack that could help lead to better decentralization or trust minimization. But as Kyle pointed out, it’s a spectrum. And for different use cases, there might be layers in the stack or parts of the stack where there could be advantages to centralizing. And so, I think it’s just project-dependent. And I think the focus on everything needing to be decentralized is probably misguided. And as we’ve seen in smart contract land, we’ll probably find some areas where decentralization is critically important in DePINs and others where it’s maybe less important because, as Kyle mentioned, there’s less vendor lock-in and there’s more optionality for people to escape what traditionally are issues with centralized and like vendor lock-in situations, where now it’s not so much of an issue.
[00:27:40] Elisha Owusu Akyaw: Let’s discuss some stakeholders. A lot of what Fleek does involves getting Web3 developers outside of the Web3 space to get into the Web3 space, and developers not building for DePIN yet to start building for DePIN. How is that going? And what’s the game plan?
[00:27:57] Harrison Hines: So, our goal is not necessarily to bring developers into Web3 so much as we’re just trying to leverage decentralization to build and provide better web services for developers. And by leveraging decentralization and taking this more DePIN approach, we think it does give us a lot of tangible benefits and advantages for developers across performance, cost, open-source alignment, and which kind of also, I’d say, is coupled with like what we like to call no corporate nonsense, which is where you can bucket all the like permissionless, trustlessness, censorship-resistant, vendor lock-in type of benefits. And then, with the Fleek platform, we really try to focus on seamless developer experience. So, that way, if you look at any other cloud platform and how they’ve really been able to scale or gain market share, it’s kind of by both providing the infrastructure and the experience to make it seamless, to leverage that infrastructure for the various use cases developers want to leverage it for.
And so, if you look at a lot of other protocols today, it would be as if they built like the Uber decentralized driver network and then expected third parties to build the ride-sharing app and the food delivery app. It could work in certain circumstances, but I think like what you’ve seen with a lot of the more recent breakout protocols, like Base, they have Coinbase as a distribution sort of channel. TON has Telegram as a distribution channel. Helium has Helium Mobile as a distribution channel. Sort of similar to Helium, where like Fleek Network is this layer 1 blockchain focused on highly performant, edge-optimized cloud infrastructure and web services similar to like Helium Network. And then, similar to Helium Mobile, we have the Fleek platform, and that’s just a very modern, developer-seamless, devex-focused cloud platform on certain use cases. So similar to Helium Mobile, I don’t think they care about bringing cellphone users into Web3. They just care about them using Helium Mobile as a better alternative to AT&T or other cell phone providers, and leveraging the benefits they have from Helium network to hopefully provide a better product at a better price point to those customers. So, we think about things very similarly.
The idea is just to leverage DePIN and decentralization to build better web services. And those benefits, hopefully, are not Web3- or Web2-developer-specific. Those benefits should hopefully resonate with everybody. That’s the approach we’re taking. It is going quite well. I think one of the advantages we have is we could take a much more like content marketing, SEO, more normal cloud platform growth approach, as opposed to like the hackathon sponsorship, conference sponsorship route that a lot of other protocols are kind of forced to take because they don’t take a product-driven distribution approach. They rely more on, like a third-party ecosystem of developers to kind of build out that product and distribution layer. And so, I think we’re starting to see the results and benefits of that approach. And so, really, now with Fleek Network, it just puts us in a position where we could offer a much more full-stack suite of products and tools and use cases that we know developers want to use.
Like, for example, we have Next.js app hosting that’s launching in probably next week. And so, what’s cool about that is it highlights the uniqueness of Fleek Network as being like a protocol with the capabilities and performance required to handle a use case like that. Yes, for the Web3 community, it’s cool that there’s now decentralized Next.js app hosting, and that’s actually a big unlock for Fleek, because 80% of developers, Web2 and Web3, are probably using Next.js today as a framework. And on IPFS, you can only do static hosting, so you lose all the Next.js benefits, which are like more dynamic server-side rendering, image optimization, things like that. So, now, with Fleek Network, we can support a use case like that. So, not only does it allow us to provide value and hosting to more Web3 projects who probably are using Vercel today because they don’t want to give up those Next.js benefits, but also that Next.js app hosting is applicable and useful to just Web2 developers who maybe are using Vercel and don’t want to pay Vercel pricing because it’s quite high and are just looking for an open-source, lower-cost alternative to host their Next.js app.
How we’re sort of going about things is just build useful infrastructure that solves real developer problems, and package it into seamless developer workflows and experiences that are familiar. So, you eliminate any unnecessary barrier to entry or any unnecessary cognitive overhead, that developers could just use it like they would a normal cloud platform, have all the features and tools and functionalities that they would expect from a Vercel or a Cloudflare or an AWS for Lambda-type product, and just win on merit. Performance, cost, open-source, seamless dev. Because a lot of protocols are focused on this like small 25,000 Web3 developer pool. But that’s what I love about DePIN and also what Aethir is doing. It’s like there’s a 25 million developer pool out there that is largely ignored by most Web3 protocols and internet usage and software, and where the real opportunity is. And so, I think that’s where the big opportunity is for these decentralized clouds and DePIN projects focused on web infrastructure is really cracking that nut and entering that ecosystem of the 25 million devs today, and also the incoming, probably bigger number of people who will become developers or builders now thanks to AI and just AI itself. So, that’s kind of how we think about it.
[00:34:10] Elisha Owusu Akyaw: Great answer. So, Harrison, I think now that DePIN is getting real popular and is getting a lot of traction, do we see a situation where traditional cloud computing companies pivot into like partnerships with decentralized cloud computing companies? What does the future look like? Are we heading towards a merger, or do we still have the traditional companies avoiding or pretending they can’t see what’s happening in the decentralized world?
[00:34:39] Harrison Hines: Yeah, I think you’re already starting to see it happen. And I think it’s going to play out very similarly to what we’re seeing happen in smart contract land, where first they sort of ignore it, and then it shows that it really is a better infrastructure to build financial infrastructure or financial services on. And then, you start to have these big companies leverage it and start to offer new products and services leveraging these new smart contract platforms to sell into their existing user bases. And you’re already starting to see it, like with Helium, where you’re seeing telcos offload data to the Helium Network. Because within Web3, there’s an example I like, if you look at Infura, one of the biggest RPC sort of providers, they send their free-tier traffic to POKT Network because it’s cheaper for POKT Network to serve that traffic than it is their own infrastructure on AWS. And so, that, to me, is a very telling sign.
And with Helium, you’re seeing it to where telcos are realizing, hey, it’s cheaper, faster, a lot less risk to use Helium to augment our own networks and offload data. Cheaper than it would be for us to try to scale and build out more of our own infrastructure. And that is very powerful. And I think cloud platforms are going to realize the same thing with these new DePIN networks focused on cloud infrastructure, because if you’re a CDN and you’ve got 300 POP locations and you want to get to a thousand, it’s way easier to just augment yourself with something like Fleek Network than it is to go build out another 700 POP locations. Because if you look at specific to edge computing, edge computing today in Web2 is really just sitting on like 30 or 50 different AWS or cloud platform data center locations because it’s economically unfeasible to do it any other way. And the reason a Cloudflare or these big CDNs only have 300 POP locations is because if they want to go to 600, now they’ve got to spend double the money to go set up another 300 POP locations, secure it, manage it, deal with all the jurisdictional headaches of having infrastructure and all these different countries. And they can’t just double their prices. They’re already in price competition with these other clouds. So, they’re kind of at this good enough state.
And so, a lot of people say like, oh, Fleek’s competing with Vercel. But the truth is, Vercel is a dream customer for Fleek Network because all these other developer platforms are built on AWS, and so, their baseline costs are already inflated AWS costs, because if you look at the statistics, 75 to 80% of IT costs are human-attributed things. And so, if you could run a system like this algorithmically and reputationally, just like Uber does, you could get the cost of providing these services way closer to the hardware and electricity and internet cost it actually costs to provide these things. And so, yeah, I think 100%, over the next few years, you’re going to see cloud platforms or developer platforms that are built on clouds like AWS realize that they’d be way better off switching out their infrastructure or augmenting their infrastructure with some of these DePIN cloud infrastructure protocols, because same reasons telcos are looking at Helium. It’s going to be cheaper, faster, probably just, if not more, performant, and you could get to market way quicker than trying to go build out your own infrastructure. So, that’s, I think 100% going to happen this decade.
[00:38:14] Elisha Owusu Akyaw: Okay. So, Kyle, quickly, what are the biggest challenges, and what does the future hold for Aethir?
[00:38:19] Kyle Okamoto: It’s a good question. I think that the biggest challenges in the space are getting folks comfortable. So, there’s a lot of education and a lot of evangelism. There’s definitely a stigma and a reputation associated with some bad actors in the space, and you have to constantly maintain your security footprint and your integrity in this space, especially when you’re dealing with millions of dollars’ worth of hardware and customers that are running their production workflows that support their core business on your platform. So, I think constantly staying not one step ahead, but three or four steps ahead and making sure that you’re completely transparent with the highest level of integrity and security possible, I think will always be a challenge in this space.
[00:39:02] Kyle Okamoto: What I’m excited about is that it’s already happening. Companies are collaborating and uniting and forming alliances and consortiums and tackling some of these big, hairy, scary challenges together. And I think that’s really, really exciting. So, thanks for bringing us together here and having this conversation, because I think the more we can talk about it and shine lights on some of these things, I think the better the overall industry will be. And not just the Web3 industry, but the industry will be. So, thanks again for having me.
[00:39:31] Elisha Owusu Akyaw: Thank you very much, Kyle, for jumping on this podcast. Harrison, what does the future hold for Fleek, and what are the biggest challenges building in the DePIN sector?
[00:39:40] Harrison Hines: For us, we still have a mainnet coming up, probably in Q1. So, I’d say the challenges are more so just on the execution side with just finishing out the builds of the v1 of the network and bringing it to life and proving out just the infrastructure and the reliability and the performance. But we’ve got a lot of exciting things on the roadmap, because now that the core protocol is finished, that’s where we get to start opening up the services layer of the network, which is the smart contract layer for a smart contract platform. But instead of building open or decentralized financial services on a smart contract platform, on Fleek Network, you build open and decentralized web services. So, CDN, serverless functions, databases, things like that. And so, that’s really what’s on the horizon is really opening that up, building more services ourselves just to get the first few out there, and then opening it up for third parties to then start building their own services for whatever use cases they want. Could be ZK, could be rollups, could be fully homomorphic encryption, could be Web2 use cases that have nothing to do with Web3. That’s really our focus is just building out the infrastructure, adding more services and functionality to the network. And then Fleek platform will continue to look at that new functionality coming to the network and package it into products and use cases that are useful for developers to just continue to create a better cloud platform, better products, better developer experience, and continue to eat into that big cloud market share in this decade.
[00:41:15] Elisha Owusu Akyaw: Sounds like we have some fun times and innovative times ahead. Thank you very much, Karl and Harrison, for jumping on this podcast. Really appreciate it, and I look forward to talking to both of you soon.
[00:41:25] Harrison Hines: Thanks a lot.
.png)
