 Facebook, Amazon, Apple, Netflix and Google, we're good. So those are the killer apps right now. It wasn't like this, it wasn't like this in the past. The first killer app that I used was Prince of Persia. I didn't play it in Apple II, I played it in Microsoft, I can't remember which one, I was too small. And that made me dream about what you could do with computers. You could be a prince, you could do magical things. So immediately when this came out, I was the greatest fan. And let me tell you, one of the killer apps still today, for me, is the Zelda Ocarina of Time. But there was a problem, there was a problem with this machine. So one day I was playing outside, I was in my village and there was a big storm, I was on my bike. I was like 20 minutes away from home. I got soaking wet, but I was really hoping that I could go home and play Ocarina of Time with my friends. My boy, that didn't happen because I had left the window open and my Nintendo 64 was next to the window. So the big storm wetted everything and I lost my Nintendo 64. So the enemy for those killer apps was just a simple glass of water. A little bit of water could destroy those killer apps, this Zelda Ocarina of Time. And you're laughing at me because I'm an idiot and I left the Nintendo 64 next to the window. But you're going to laugh a bit more to these guys in the Metro of London who let the door open when they had some cement going on. So that was terrible for them, but nothing of this would have happened if they had what we have now. So the cloud is extremely convenient, right? The cloud is amazing. The cloud allows us to use everything we want from wherever we want. And this is me using apps as an app. I use Spotify for my music, I use GitHub to check the code of my peers and I use Google Docs to do this presentation, for example. And if I lose my laptop, what happens? Absolutely nothing. I can use it from here. So I know the story is that the cloud is amazing. The cloud is really good, it frees us from the burden of soaring, of backing up. You can access it from wherever, you can do the computation of extremely complex apps somewhere else and use it with a device that is not so powerful. It is very, very inexpensive, as opposed to buying the hardware every time that you want to play a game like my original time, and it's tremendously scalable. So in summary, the cloud is just very, very convenient. Let me prove a point here. Who here wants a theory of note? Alright, be a reviewer. Oh, no, no. That's pretty good. Who runs it in a VPS? Why? It's maintenance and security. Who else is running it in a VPS? There are more people. Why? I don't have a home. It's fantastic, exactly. So basically you run it because it's very convenient, right? It's very convenient. So that sort of proves the point of how convenient, especially if you don't have a home. To run any application that you want, including an ether note. We've looked at the apps that I use and apps that I used to use, but let's look at the apps that the world is using. This is, when I say fang, more broadly, what I mean is this. Those are the apps that people are using the most on their phones right now. And I took the liberty of investigating a little bit where those apps are hosted. Because you have Facebook on your phone, but that doesn't necessarily mean that all the magic happens on your phone. So what I found out is that four of the top five are in Facebook. Facebook has massive data centers. We can call it the Facebook Clouds, even if they don't sell it there. It's our cloud services. Four of them are in Google Clouds. Two of them are in WS. Two of them use a multi-cloud approach, meaning Google and WS, and probably some others like DigitalOcean, who knows. Five of them, they use cloud services, but they're mostly Chinese companies. They do not disclose who those infrastructure providers are, but I've checked on their annual reports and they use cloud services. And three of them do not use the cloud. Why? Because they're media players. And arguably, the distribution of these apps, if it happens through Google Play, happens through the cloud as well. But the app itself does not need the cloud, right? How much does it cost to have this app running? $13 billion on network infrastructure for Google. Last year. This year they have one. Snap, who was on that list, pays $600 million a year to Google and to AWS. Lift $100 million, slap $50 million, Pinterest $125 million. Who the hell uses Pinterest? Apple $360 million per year. Did you know that Apple Cloud is actually Amazon Web Services Cloud? So just five clients bring us to one and a quarter billion US dollar a year. If you look at some of the forecasts of the industry, it's $83 billion in three years. That's a lot of money. And those apps are free. Mind-blowing. There's two apps that are how subscription level, Netflix and Spotify. And two of them, Uber and Amazon, it's a big review. So the app is free, but actually if you want to use the services, you have to pay. Those apps are free. And they're spending billions a year supporting the infrastructure. You guys are in this room, I know I'm preaching for the choir here, but you know what happens when apps are free, right? But there's somethings you're doing. And that's cool, nobody cares. It's even fun. Everybody knows that. Everybody knows it. And we laugh at it when we read. Apple contractors regularly hear confidential details. They hear people having sex. They're doing funny, right? Unless you're homosexual and you live in Saudi Arabia where you get killed, where somebody reports you. So that's a little bit less fun. There was another example of data that was misused when it fell into the wrong hands. So we're in March 1940. Germany invades the Netherlands. And the Netherlands at that point was one of the most progressive countries that we have in the world. They were a secular country. And they had one of the most just ways of dealing with religion, which means that they had a religious budget on their annual accounts. And they split it accordingly to the percentage of the population than was of one religion or another. Which means that in the census they knew what religion was. So the Germans come in, they just go to the census office, they look and they have the name and the address of every single Jew in the Netherlands. That's where the Netherlands has the biggest interns in percentage and in absolute terms of Jews deceased during the Second World War. Right, so we don't have to go so far. We don't have to go to Germany in the 1940s to see examples of how diet data can be misused. Everybody knows about Cambridge and Leadinger. Now they've boomed. We probably... Probably, if it wasn't for them, you guys would have a different outcome. Probably, the what-if scenarios have never been my favorite. But it is for sure that they used data to target people that were on the fence and take the balance to one side. There's another person that used Cambridge and Leadinger services, another president, and as you know, you know how much he paid for Cambridge and Leadinger services. Make a guess. How much does it cost to win an election? Ten million. Low. Lower. Oh, five. Five million. Five million. A cool five million. You get data of pretty much everybody that has a Facebook account and you can go really deep. This year, this is the former CEO of Cambridge and Leadinger, and they based their work on... I can look it up for you guys. So there's a researcher that's created back, like a long time ago, created a series of algorithms that with 150 likes on Facebook, they could know you better than your best friends. They knew things about you. There are more things about you than your best friends. If you are a closeted gay, they're fucking mean. Right, so this data is dangerous and can be used for the wrong things and this is happening right now. Does that make people not use these apps? Is privacy really a value proposition? Look at how many users do we have on WhatsApp which is owned by Facebook. One billion daily users. Daily users. I am one of them. Messenger, almost one billion. TikTok, 200 million. You put everything. You're almost half naked there. They know everything about you. Facebook, 1.6 billion. That's a lot of people. Daily users. And just for comparison, let's look at this. This weekend I look at the set of the apps and that's the 3,000 results. But what are we doing great? I mean, we're fine. I think we need to improve. And I suggest that we take another model. We take another approach. We take another approach that takes the power out of that, out of all these cloud services. How do we get our decentralized apps to get close, even close, to the traditional apps level of adoption? Well, first of all, they need to be... Free. Free, okay, yeah. What else? What was the magic word convenient for? Convenient. Convenient. They need to be convenient. They need to be convenient. There's this unswoken rule written that says that more or less the technology, just because it's bare, will not take over the previous technology. It needs to be 10 times better. So we need to be... We need to make things that are 10 times better than the things that are just the killer apps right now, which are not Ethereum at this point. And what's better than free? Pay for it. Okay, pay for it. Okay, pay for it. And we're paying a lot of money to this cloud provider. So I suggest a personally owned student cloud that sheets the transfer of volume from the cloud to the user. So this is the current model right now where the data is paid... Well, the money comes from sending the data and goes to the cloud provider. There's also other flows, but I want to focus specifically on that flow of money because we have one thing that they don't have, and it's a native way to transfer value. So I suggest that if I want to do a decentralized Spotify, we do it like this. We pay the users to host this task for us. For us. I'll use the Spotify that pays me. I'm sorry for you. You couldn't be here and have a house. You can't have it in your house, but maybe you can put it in WPS as well. It should be fine. So I suggest that here, when we have Spotify, somebody pays... Yeah, okay, most of us are free. We're going to touch into that. They'll be here, and we transfer the value to the user. We're far away from that. We're far away from that, and I'm going to try to make a list of what pieces need to come together to create a model where this would work. Number one, it needs to be as convenient, as accessible from anywhere and all the time like the cloud is. This is not a 10X. This is a bare game. We need this, and we don't have it right now. It's a pain in the ass. But we're working on this. We're working on the infrastructure side to make this infrastructure better, more accessible from everywhere. So how? We're working with Archival. I'm going to use some two DevOps concepts here. A high availability cluster is... Everybody knows what it is. It's a series of nodes that are connected, and when the first one goes down, the second one takes over. Archival is developing a system specifically targeted to eth2.0, so anybody can start staking from their own homes. You're going to Federation with your friends, and your node goes down, the second node takes over, and then you avoid being slashed because of downtime on your node. So good. We're getting closer to high availability here. There's another thing called load balancing in DevOps, which is basically distributing the load through the nodes for the servers that are there. And there's fucking network over there who are doing an excellent job on creating this orchestrator right here that redirects the traffic or the transactions to the nodes that are online. So we've got high availability. We've got load balancing. We're getting somewhere there. But this is just the bare minimum. This is not a 10X. Another thing that's not a 10X is throughput and scalability. There's things that are going to help us get better with scalability. I'm not an expert on that. I'm not going to touch on it. But the roll-up with zero-nodes groups, Ethereum 2.0, I'm sure you guys are absolute experts on that. This is what we need. It's a bare minimum. But we're starting to get somewhere with privacy. We said that privacy wasn't a killer proposition. It wasn't about your proposition, but it is. It's just not on your zone. It's not 10X there. It is 4X there. Why do I say that? Because given the chance of using a private app that sells your data, and they work exactly the same way that they're exactly the same level of availability and the same level of transactions, et cetera, et cetera, you will use them on the spread. But we're at the bare minimum right here. On the infrastructure side, what we can do to ensure privacy, we can include, in that hardware, we can include environment-structured environments like the one on SGX, we can create and then with zero-nodes groups, we can create them within those enclaves and that's available for the app users to do whatever they want in a secure, environment way. Also, we've seen that the model means it worked for apps that had the Spotify worked for somebody that was paying for subscription, then you pay the user to host it, right? There's other models. There's DataFun, for example, which is a way more ethical way of using the data. So we don't have to just throw away all the data, but we should be able to let people opt in on what data they want to give and to whom. Another thing that needs to happen is instantivization. Pocket Network, ETH 2.0, Storage, Rain and Vidno, there are many projects out there and one of them will not give you enough money to run a project, but if you stack them on top of the other, on the same machine, you start having a money box in your house that gives you enough money to sustain the box itself. So that's... That's not 10x yet. There's one more component. There need to be apps that people use that people can't use and they can be easy to use. Give value to user. It's not only UX and UI, I'm sorry, I should have changed that. It's basically value proposition. Give value to the user. Most of the apps right now use the hoop. I don't know if it's familiar with the hoop model here, but basically what they do is they trick you into using it all the time. They make you addicted, they make you hooked, because the most killer app in this universe is an app that buzzes every time somebody from the other side of the world gives you a like on your picture and that buzzes on your phone and you look at it and then you like another thing and then you put it back into the pocket and it buzzes again, because some other person has clicked on something, something as stupid as clicking is designed to make you addicted to it. So when you develop apps, please make sure that you're not a dealer. Make sure that you actually want to use it and make sure that it materially improves the user's life. This is the way we're going to have apps that can compete at least with the hoop model. So what we're doing at that note is taking care of the infrastructure side. We are working on the first part of the infrastructure, having this infrastructure being available from everywhere. Through the scalability of the Ethereum foundation, everybody is very, very much focused on that. Privacy, we're including SGX and we're including systems inside our boxes that you can leverage for your apps and make them private and secure. Instantivization, with e2.0, that will help us a lot, but you need to do the last part. You need to create apps that give value. And I suggest that the next time you're doing a human-centered design system to create an app, think about this model. Think about instead of giving money to the cloud providers who are going to take this data and they're going to use it for God knows what, think about this. Think about making people host, incentivize people to host your apps. If you want to know how to do it, please get in touch. We can help you design something on that note that will be hopefully almost as good as using BPS or even more convenient. Thank you very much.