What's Wrong With Cloud Computing (And How to Fix It)

0:00Your favorite neighborhood spot grows with Square. Indeed, my favorite neighborhood spot has quickly become Todd Snyder in Williamsburg. Todd Snyder is one of my favorite menswear shops and has supplied me with all the clothes I have needed this quite hot summer. Every business has different goals, but Square is the business platform that supports them all. From opening a new location, selling something new, or just expanding their reach. Indeed, I've seen it with Todd Snyder. In Square, also, you can get real-time insights, so don't wait for end-of-day reports. Go to square.com forward slash go forward slash realvision to learn more about how your business can grow with Square.

0:43That's S-Q-U-A-R-E dot com slash G-O slash R-E-A-L-B-I-S-I-O-N.

0:57Hey everyone, if you like this podcast, go behind the paywall to get privileged access to the smartest minds in finance. Join the Real Vision community and learn how to become a better investor. Visit realvision.com slash rvpod and use the promo code podcast10, that's podcast10, to get 10 % off our essential membership for the first year. Now, to the top analysis of today's crypto markets.

1:30Paolo Frigameni, co-founder and CEO of Supply Socket Company. Welcome to Real Vision Crypto Daily Briefing. Hey, thanks for having me. Paolo, lots to talk about here. I'm excited for this conversation. You're doing some interesting work in the field of true decentralization. But just a second, I want to talk a little bit about what's happening in markets right now. Now, let's take a look at CPI. This came out this morning, as you can see, 4.9 % year-over-year CPI. I'm going to keep this short and sweet today. Basically, inflation accelerated less than expected, slightly less than expected. It's down slightly from last month.

2:05But the reality is it's still far too high, significantly above the Fed's 2 % target. By the way, we should say we're looking at CPI, that's Consumer Price Index, out today. PCE, personal consumption expenditures, is the number that the Fed looks at, but it's It's a pretty close proxy and it's closely watched. Let's take a look at what's happening in digital asset markets, Bitcoin and Ethereum specifically. Bitcoin bounced up a little bit on the news, trading above 28 ,000, 28 ,178 on my screen right now. Make of that what you will. The expectation perhaps of a worse print, a bit of a relief rally there.

2:39Tough to say, but that's the data. Up right now about almost 3 % on a trailing 24-hour basis. So seven days basically flats off about half of 1%. If we take a look at Ethereum, as you see on that chart, very similar pattern there. Probably no surprise reacting on the same news. Trailing 24 hours up about 2%. Seven days also roughly flat, but in the opposite direction, up about 1%. Not a whole lot going on there in terms of the longer term picture for the last seven days about what's been happening in these digital asset markets. With all that said, let's jump back into something that's a little bit more substantive, a little bit more on the infrastructure side.

3:16We love builders here at Real Vision. Paolo, welcome again. Tell us a little bit about Socket Supply and what you guys do. Yeah, sure. So we build an open source project that helps web developers to create applications for any OS, so native applications. And that means the average web developer who uses HTML, CSS, JavaScript, they can build an application, a native application for Android, iOS, Linux, Windows, Mac OS, basically anything. And there's nothing really new to learn. The skills that they already have are all the skills that they need. And so we create this framework for doing this. And I think that what's pretty interesting is that the thing that people generally want to do after they build an app is to connect their users together.

4:28And so we have kind of a novel way of doing this, which is making peer-to-peer capabilities available as native JavaScript APIs to the average web developer. So this is an important point, I think, for people to understand. I guess people with technical backgrounds already understand that. But let's talk a little bit and explain why exactly this is so important. So what you talk about, the basic tools, HTML, JavaScript, CSS, these are the basic core tools for doing web development wherever you're doing it. If it's going to appear rendered in a browser, those are basically the tools that you need to do that.

5:02What's interesting here about what you guys are doing conceptually is trying to figure out how to plug in this type of development architecture into a true peer-to-peer system. This, I think, is what confuses a lot of folks about the space because it's interesting. We talk a lot about decentralization in the digital asset cryptocurrency space, but most of what's happening out there runs as software as a service on these very large cloud forms. That means Microsoft Azure. It means Google. It means all of the big companies that have dominated Silicon Valley for many years now. And this transition to a peer-to-peer network in terms of the back end, I think, is probably a critical one.

5:39This may seem surprising to people because people think of Bitcoin as a decentralized network. They think of Ethereum as a decentralized network. But the compute power is, in fact, running on these very large server farms. Let's talk a little bit about the solution that you guys are proposing in terms of attempting attempting to make that a more true peer-to-peer experience? Yeah, so our team is made up of quite a few browser engineers, people who have been contributing and hacking on browsers, probably the ones most people are using right now, whether it's Chrome or whether it's WebKit, you know, based browser.

6:13So, you know, we've taken the core of the browser engine, which is an OS component, and then we've added native peer-to-peer capabilities to that. What that means is that, like I said, the average JavaScript developer can use a really simple API that should be familiar, shouldn't be very much to learn. And what they can do is they can connect with other peers in the network and they can completely circumvent the cloud. They don't need the cloud at all. There's no servers required for the solution. And, you know, along with that, we've made, you know, put in a lot of effort to make sure that there's that it's secure and that it works in terms of people needing to communicate, even when other peers are offline.

6:59right so one of the big misnomers people have misconceptions people have uh about peer-to-peer is they think that if other people who that are trying to do peer-to-peer with aren't online then it doesn't work but that's no longer the case so uh i can build let's explain a little bit what peer-to-peer is for folks who may not know i mean i think you know for myself and people of my generation, we first heard this terminology around Napster in the late 90s, early 2000s. And what's interesting about these networks, I suppose at the time, what got most of the attention was the idea of people circumventing copyright law.

7:39This was, of course, a very big problem for the music industry. But the core technology of peer-to-peer, I think, was really interesting to people conceptually, particularly in the field of computer science, because essentially you had a network where the larger it grew, the faster and more robust it became. This being very much the opposite of the traditional centralized model, where when more people try and pile in to download something, for example, the network becomes strained. But with peer-to-peer, you essentially have a series of nodes connected as, well, peers on the network. And as people join, it becomes more robust and it becomes faster.

8:15This is a huge leap, I think, in terms of the mechanisms for distributing information on the Internet. And also, in many ways, many see it as something that's potentially freeing from these very large Silicon Valley behemoths that now control the cloud market. Yeah, right. So the cloud is a set of servers that's owned generally by a handful of incumbents. And it's a landlord-tenant relationship, right? You want to build a product, you go to the cloud and you lease the services that you need and you end up building yourself into that lease for basically the lifetime of your product. So as the demand for your product grows, so does the cost of running it.

9:01And so this is prohibitive in a lot of cases for entrants into competing in certain spaces. right? I mean, if you look at the cost of running something like Twitter, I mean, it's absolutely massive. I don't even think Elon was able to pay his cloud bill at one point. And there have been a bunch of stories written here in the last couple of days about the cost of running chat GPT, for example, which is also massively processor intensive, as you might imagine. Right. So a lot of these things, and especially the things that are, you know, the basic kind of day-to-day things like chats or things like Twitter, these are incredibly good candidates for peer-to-peer because, you know, they don't need servers.

9:45They don't need the cloud to work. And so what we've done is we've put together this open source, which has been highly peer-reviewed and we're starting to see early adopters start to build things with it. And we're starting to see people realize that they don't need the cloud to build a lot of the things that they're building today. And I think that this is, in particular, this is really well suited for DeFi types of things. So as you said, we're seeing a lot of - Let's double click on that because that's really a huge paradigm shift, this idea that people do not need to distribute software in the cloud.

10:23That is the overwhelming dominant paradigm of the way Silicon Valley runs here in 2023. And this notion that individuals running peer-to-peer browser-based networks can supplant that really is a huge shift in the way business gets done. Yeah, it's pretty transformative for a lot of people. So, you know, you think about like the largest cost for most people running any kind of software business, it's cloud. And it's not just the cloud bill itself, but it's all the SREs, all the DevOps engineers, all the time spent, all the time putting and debugging the glue that puts all of these services together.

11:00And so there's a mountain of hidden costs associated with that. And then the risk, certainly, of operating a business where a third party kind of owns everything that you've built and in such a way where you really can't practically move it after it's been built. People talk about multi-cloud, but multi-cloud isn't really that practical. And so, you know, at the end of the day, the biggest problem really is the man in the middle, right? So having this kind of central figure who is really in between you and your product and your users, it's a big liability in terms of security and privacy. And so, you know, there's that.

11:44And I think for us, really, what's interesting is the reduction of the cost, making it possible for people to enter new as entrance into a new space. And then just the complexity. So, you know, building software these days just means this enormous amount of effort and hours spent. And I think when we're talking about something like DeFi, the D in DeFi is really important. We want to achieve decentralization. So we kind of already understand why decentralization is important. But this is a broadly applicable thing. Pretty much anybody in any space can benefit. But I think maybe it's pretty interesting to talk about why DeFi in particular could benefit from building peer-to-peer software.

12:34Hey, everyone. We're going to take a quick pause and hear a word from our partners. We'll be right back.

12:43Well, let's talk about DeFi in just one second. But obviously, this is an important goal, and it's a very big one and audacious one at that. But let's talk about some of the risks here before we talk about the specific use case of DeFi. I'm curious in terms of a couple of things. First, where are you in this being actually deployable? deployable. And second, what are the risks that you see in doing this? For example, security, scalability, load balancing, all the traditional cloud-based parameters that people think about. How does this actually work in terms of the security model, first and foremost?

13:14I assume that this somehow uses the PKI, the public key infrastructure, to do encryption. But talk a little bit about what makes developers feel secure that their applications are running in secure sockets in secure spaces on the computers that are being scaled throughout this peer-to-peer network. Yeah, well, there's quite a bit to unpack there. But in terms of the individual applications and the network itself being secure, all of the packets that are distributed into the network are obviously encrypted. And there's a couple of different types of encryption schemes that you can opt into. There is no one size fits all solution for encryption.

13:56If you need, for example, point to point, or if you need, I mean, there's different schemes that you'll potentially need to opt into. So, yeah, it's not one size fits all. And so the packets as they're deployed into the network, they're encrypted and the peers that relay those packets, if they relay them, then they're not able to see what's in them without the public and private keys. And so it makes it possible to create this network, this global network of sort of shared responsibility so that if some users in the network are offline, then other users in the network who may not be directly related to them can continue to support the delivery of those packets.

14:41And this isn't really something... This is a kind of dynamic load balancing that you guys are able to do because... Yeah, it's a distributed load balance thing. Yeah, and we didn't necessarily invent this particular aspect of this. This is a well-researched, really kind of widely known academic research space called disruption-tolerant networks. And a disruption-tolerant network, or sometimes it's called a delay-tolerant network, really has been researched for the last 20 years. And so what I'm talking about in particular is network buffer optimizations. So being able to support data peer-to-peer, even when some of the peers in the network are offline, and make sure that that data is always encrypted until it gets to the party that was intended to be reached.

15:30And let me ask you this. You mentioned this idea of not necessarily building all this technology. To what extent were you guys able to leverage existing technology for the network distribution layer and also for the security model? I think it's probably something that people think about in terms of the existing security. Is this something that you had to build sort of out of the box, or were you able to leverage different layers that have already been put together? Yeah, so this is a, unfortunately, you know, we surveyed the peer-to-peer landscape in terms of libraries, and we really didn't find anything that was comprehensive in terms of NAT traversal, which means being able to discover and then connect to other peers.

16:08we really weren't able to find anything that existed that was comprehensive and really worked really well in that space. And then in terms of also having a well-defined specification that deeply detailed how it worked. We just didn't see that. We didn't find that. And we know really kind of everybody in this space who's working on it. And so we started from a first principles kind of ground up mission to go out and read all of the prior work that's been published. A lot of really interesting stuff has come out of various universities, really nicely peer reviewed papers that have been published that we were able to read and to source or to cite as a material that we based our solution off of.

16:56So, you know, not a ton of it is novel. We were able to use a well-known corpus of research work to base this off of. But it certainly wasn't like going out and building a SaaS product where you say, okay, we'll use a little of this and some of this and some of that, and we'll put it together and we'll have a product. But we had to really think about this from the ground up. And there are some areas where it's slightly novel because, like I said, a solution like this hasn't really existed before. So that's where we're at with this. It's coming out of the alpha stage into the beta stage. There's people building products with it.

17:37There's people who are using it, who are the sort of brave early adopters. But it's working pretty well for them. And so we're also working with everybody from the public sector to the private sector. public sector obviously is interested in this because this allows them to get away from the dependency of the gov cloud, which is nice that it's air-gapped and it's nice that it's fairly secure. But the problem with it is that it still represents a lot of complexity. And the time to go from the idea that you have to something that's deployed and working and doesn't require any additional infrastructure, server infrastructure is really appealing.

18:25And so, yeah. Really interesting stuff. I just want to invite our viewers to join us in this conversation. Please, if you're watching or listening, put down your questions in the chat, wherever you're watching. We'll make sure to ask the best ones on air. Remember, Real Vision members take priority. But the good news is that Real Vision crypto is always free. You can go to realvision.com forward slash crypto to sign up. That's realvision.com forward slash crypto to sign up. And of course, if you're watching on YouTube, please tweet this link out so we can share this conversation. We're committed to keeping this conversation free, and we want to get it out to as many people as we possibly can.

18:57Of course, you can follow me at Ash Bennington on Twitter, and please follow Real Vision at Real Vision. Okay, with all that said, Paolo, let's talk a little bit about the use case, the application that you mentioned at the top of the show, which, of course, is DeFi. Why are peer-to-peer networks especially well-suited to DeFi applications in your view? Yeah. Well, so I think that one of the biggest problems that people face when they're building products in DeFi is that they end up having to build SaaS. And it's really not the dream. The dream is really to build a piece of software that doesn't rely on these centralized systems like DNS or like web hosting, like services like AWS, where they end up in a landlord-tenant kind of situation.

19:44um and so you know obviously the the autonomy is really important for these kinds of products um the the sort of sovereignty but also i think you know when we look at what are the potential margins from products like this they're you know they're they're already you know um uh you know not necessarily super high so what what we would love to be able to do is to uh make sure that we're not spending on operational costs like the cloud and that we're focused on top of line or that we're focused on the features that really attract users that make them, you know, that make the product irresistible, you know, that kind of, that's where we should be spending the time and energy and the capital, right?

20:29And so, yeah, so I think that the autonomy and the cost, I mean, especially, you know, we call it DeFi, right? It should be decentralized. So this is really kind of helping to deliver on that and make that come to fruition. Hey, everyone, we're going to take another quick break and hear a word from our partners. We'll be right back to the Real Vision Crypto Daily Briefing.

21:00Let me ask you this. Let's talk about something that we haven't covered yet, which is the model for monetization for this? How is this going to be supported in terms of the processing power? How are you going to incentivize people to do this? I know that you guys don't have a token. You said it earlier in the show. This is really a first principles-based product, really something that you're thinking about from a raw computer science perspective. But what's the model to incentivize behavior users to run these services? Right. So we don't necessarily, as a company, try to incentivize anybody. But what we can do is we can offer the protocol, the runtime, which web developers can use to build the software and then connect the users.

21:41And everybody who joins the network using this runtime keeps a small cache of packets. And that small cache of packets adds up to a lot. So if you compare, say, for example, Amazon has roughly 3 million servers, but there's about 15 billion mobile devices in the world. And obviously those are a lot less powerful than those 3 million servers that Amazon has, but they add up. They also have slower connections, significantly slower. Oh yeah, yeah. And when you think about mobile devices, I mean, they're absolutely unreliable. They're on for very short amounts of time. They have varying degrees of capabilities.

22:19They may have less storage or less battery at any particular time. But all of these statistically add up to a pretty powerful network at the size of the volume. And so if your application, like say, for instance, you build an Instagram clone or if you build some kind of social media and you attract only maybe half a million users, well, it's not very many users. But with that many users, they'll all be able to support each other to deploy packets into the network and keep the information alive for as long as it needs to be alive. And so it's sort of a network of shared responsibility. And so there isn't really a need to have a token in this.

23:06Now, there's certainly the potential for people to build products that incentivize people to store more data. So, for instance, if you had something like a Filecoin or if you had something like an Arweave, you may want to build with that on top of Socket. And if you did, you may find that your users are more likely to accept larger, to cache more data on their devices. And so there's a really big opportunity for people to build with protocols like that. But obviously, when it comes to DeFi, us thinking about the different types of UIs that can be built with this, user interfaces that can be built with this without needing websites or without needing URLs that can be taken down or somebody can say that you're not allowed to be running this service anymore.

24:06That's just something that we don't really face with Socket. So do you, in essence, then see this as a potential alternative to DNS to domain name servers that currently resolve the names of websites to IP addresses? Essentially, does it replace some of the architecture of the existing web itself? Yeah, definitely. I mean, this isn't really necessarily new. I mean, you know, we've been seeing, you know, competition between the web and native apps for a very long time, right? And what we're really doing is we're saying, okay, we're inviting web developers to build native apps. But then we're also giving them some of the existing, well, pretty much all of the characteristics of the web as it is today.

24:51The ability to link between applications, that's something that you can do with Socket. The ability to build with HTML, CSS, and JavaScript and not have to learn anything new. But then, you know, the ability to do that all without DNS or without any kind of web hosting, that's really the big change. But all of the other things should really remain pretty familiar to everybody. And how far away do you see this from being implemented in terms of when folks can actually expect, if they're listening to this, to deploy on sockets like this so that you can truly start running applications in peer-to-peer?

25:29Yeah, you can start building with this today. Today it works. And how long do you project it will be? You mentioned earlier, you called them brave alpha users, before this would be something that would be ready for prime time in terms of deploying industrial strength applications in a peer-to-peer methodology. Yeah, so the reason why we are not, the reason why we're in beta at the moment is because we like to be cautiously optimistic. We think that everything is working out pretty well, but it takes rigor to make something like this a very viable solution for replacing traditional infrastructure.

26:11Traditional infrastructure has a lot of boxes to tick to be able to say, yes, we can replace this reliably. And so that takes rigor to measure the types of latencies that people have expectations about. And all of that's looking really good. And when we can create some quantitative data for people to look at, that's when decision makers at some of the larger fortune companies will start to take notice and say, okay, maybe this isn't just for startups and adventurous DeFi folks anymore. Maybe we can start to take some of this on and replace some of our traditional infrastructure in larger legacy or fortune companies.