 I think you're taking sort of intense literal logic and applying it to a colloquial situation that has everything to do with tradition and nothing to do with literal truth, right? Who are you and why? Nice. Do you want to go first? I'd rather you did. Oh, thank you. My name is Dionne Almer. The why is an interesting one, you know. So I've worked with this guy for a long time, lots of different companies. Way back in the day we did this thing called Ajaxian. There's this thing called Ajax. You were too young to know it. I remember Ajax. I'm not young enough. And that was, you know, really fun to kind of see the web kind of come back and go from docs to apps and the like. So I've been living in the web for a long time, used to be at Google. I'm now back at Google running DevRel for a few different teams across the web, Android, Assistant and other things. What about you? So I'm Ben Galbraith. I am on the Chrome team. I lead product for the folks who are working on the web platform, so the folks who are thinking about how the web works, when you're under a web page and things. We've been doing these interview things all day. So we never done interviews before. Yeah, and we missed your talk. So we opened the talk just by showing that PWAs are everywhere and PWAs it's fair to ask what the heck that is. Basically the web has been evolving dramatically over the past few years and it's been getting a ton of new capabilities. And it's a whole new platform that it used to be just four or five years ago. And so we shared a lot of data around this. And there's kind of two categories. One is, well, what's actually changed about the web? And the other is what happens when people take advantage of the new capabilities that are in the web? And we shared a bunch of case studies. Getting to the capability side, the big one is this thing called Service Worker. I've heard of this. Have you? Go on. Have you? Let me explain it to you in very simple terms so that you can understand. Oh, but it's like forever everyone's understood the web is pages. Pages that the browser requests and renders. And in Service Worker is the first major new primitive on the web that changes the game because with Service Worker, you now have this code that's resident in the browser. It can receive events from the site. Even when the user has left the page, you can use it to override the network. I love how you guys are critiquing my explanation in real time. It's actually really good. But yeah, it lets you as a developer take control of the networking stock, which seems like a geek. Why would you want to do that? But it's super important because there are a lot of people using the web who have intermittent connectivity where it just drops out or it's really slow. And so as a developer, you're in the driver's seat. It doesn't show you the dino screen. If there's a problem, you can create these really smooth experiences. I feel like this time too, we actually talked about the Google properties are actually using this. Oh, that's right. Let me wrap up Service Worker. Because the big news is that it's finally in Safari. And it hasn't been for a long time. And it's finally an edge. But yeah, you were talking about some other stuff. Yeah, like Google search, actually using the Service Worker. Got to talk about that today. 50% less JavaScript. Yeah, that's right. All of the intermittent things. You do a search, you're offline. When you come back online, do the search, pop back. Hey, you want the results? So we shared a bunch of these examples. And then we had, I think, some pretty exciting new stuff to share. I think stuff that's new for a lot of people. The problem with the web is that it's out in the open. So for insiders, there's really not a lot of new things to say. And if there is, that's a problem, right? Because we don't control the web at Google. So we're not going to have this brand new thing. But we are highlighting a lot of the recent developments, one of which is this thing called desktop PWAs. This is cool. We showed how Spotify is using desktop PWAs. So you go to their site. And you can install it to your launcher, in this case, in Chrome OS. And it gets its own top level window. So Alt-Tab now works with websites. Oh, so good. It's fantastic, right? I'm glad we did that on mobile first. But it felt like it was a missing piece that we couldn't do on desktop. Productivity on the web is a big deal. Like, we talk about mobile, mobile, mobile. And mobile is super important. But there's billions of people on the desktop. But the other thing is using a lot of these popular productivity apps as top level windows on Chrome OS. Because that's where it is now. We're bringing it to Windows and Mac, but it's not there yet. And I felt like lots of companies doing their apps in Electron show that there was a desire to do that. There's so much to that. So to write web and just have it as if it was a normal app on the computer. Let me talk to you about AutoCAD. Because AutoCAD, they created a web version of their application, which is kind of mind blowing. Like, you go to a lot of analyst speeches and the intelligent analysts are up there going like, well, a lot of stuff's coming to the web. But old applications like AutoCAD will never come to the web or it'll take them over like, it's here. They actually shipped it in March. That's right. And I said, old, AutoCAD's been innovating a ton since it shipped. So it's actually a pretty cutting edge. But they brought their C++ code base to the web. So it's sounding great, right? So kind of like, job done. We can relax for the next few years. Can I take the negative side? Okay, I think we're done here now. Wrap it up. So I think the interesting thing about the state of the web, like the talk is about like everything's going on all the good things. But that's not the entire web that you actually experience every day. And we want to keep adding capabilities. We want to keep bringing Electron back to the web and pulled over back to the web and all these things. But how do we get all of the long tail in there? And that's why we do a lot of work with the word presses and all these different CMSs to how can we affect that long tail? Yeah, what does the new capabilities matter if nobody uses them, right? And I feel like we're kind of unlocking both at the moment. Well, that's where AMP comes in too, because a lot of the keynote was talking about AMP because the web platform has a ton of capabilities, but it's also become a little complex. I mean, I don't know if you've noticed. You don't say. It can be hard to figure out how to do it. And so AMP is all about how can we make the web really easy for a set of really targeted use cases with an opinionated approach. And it's been maybe a little controversial. And this year we did a couple of things that I'm excited about. One is that we announced that we're fixing the problem with AMP URLs. And if you don't know what that issue is, but it's this double-edged server where AMP does this cool thing where like it hosts content on this server that optimizes it and like shrinks images and all this stuff, but it's hosted on that server. And so I always understood why they were doing it, but still the artifact of the URL was just so annoying. And it's annoying to everybody. Some people look at the URL and they think we're trying to like move the whole internet onto our origin, onto our domain, which is like not at all what we want. So there's this new emerging standard called web packaging, which on itself could be like the subject of a conversation. I'm super excited about it. But from a high level, what it does, it lets you as a developer take your site and effectively wrap it in a secure envelope and just say a package, maybe a better metaphor. And then like it doesn't matter where it comes from, it's advertised as though it came from your origin. And so you can imagine like a browser of the future that's more like peer-to-peer, right? So like if you're in this area where you can't actually get to, let's say the New York Times, if any peer has a version of the New York Times that's like fairly recent, you can get it from them. Creating this really decentralized networking thing, which is at the heart of the original spirit of the internet. Anyway, so AMP's able to use web packaging and it's an emerging standard. We'll see. The other thing that I'm excited about is that we've announced that AMP is embracing web standards and things like the Chrome User Experience Report so that we'll be able to use the actual performance of web pages to decide whether or not they're fast. Oh, so that's what gives it the little icon, right? Well, today you see that icon if something's AMP. And moving forward, we've shared our intent to use things like the Chrome User Experience Report and other things as the canonical indicator for whether or not something's going to perform really well. We always talk about performance with AMP, but there's the privacy preservation piece to this too. Yeah, that's right. If I go in, I'm in an aggregator, if I pre-fetch or I start pinging over to this other server, then all of a sudden, D on the user can be on search and jcarchable.com knows something and has set a cookie on me and that's kind of weird. Without you ever clicking on that link. That on me ever knowing, yeah, exactly. So by having it somewhere else on Google Servers or whatever AMP cache it's on, we can hide that and now with Web Package in, it's baked into the platform. So that's really exciting, I think. So what do you think the web needs to do within the next 10 years to survive? I mean, there's the foundations and there's the real future future stuff, right? Yeah. And like foundationally, we just need to fix the performance problem on the web. You don't say it. Just focus in on getting people using the web. I mean, the web is this fantastic treasure. I feel like if you look at it, what do we have in the web? We have this open ecosystem where the standards can be implemented on any platform. I'm thinking about that for a second because most other platforms use their ecosystem as a walled garden to block other people and to sort of chain people in that ecosystem. The web is totally open. We've got this inherent indexability so you can go through and discover it. You don't have like a single gatekeeper or toll collector. These are just fantastic attributes of the web. One of my first talk as a Googler I was using a slide where I opened a website on a Nintendo DS. And I was like, that's really a unique feature of the platform. Like you write a website and suddenly it is in your car or your phone on a Nintendo DS. But I think Jake, the web is worth having in society. That's an understatement. But I think when you talk about 10 years, what I'm really focused on is how can we make sure that the web remains the platform where developers can bring their best experiences because I think there's a lot of competing platforms. Some of them can do a better job at animations and visual effects in the web. Maybe. Maybe. I don't know if that's a fair statement. Not 10 years from now, though. Yeah, two years from now. One year from now? But so we've got to make sure developers have the capabilities they need. We've got to make sure that users actually prefer the web. And we've got to make sure that browsers are where computing is moving. And when you think about that frame, we have a lot of work to do. I feel like developers, we've got to do a lot. Houdini is one of my favorite projects. And second service workers. There's a lot of additional stuff we need to do. So when you talk about 10 years, I want to see the web platform evolve to have more capabilities. I don't want to see us chasing app platforms. I don't want to see us advertising the web as kind of like, it's almost as good as this app platform. You should develop websites that are like apps. That maybe is a controversial statement. You were going to say the web has its own unique flavors, right? It has its own unique capabilities and use cases. That's right. Like there is still use cases. I say no, that's a use case for a native platform. That's not necessarily something you want to solve on the web. I couldn't agree more. I think if what user developer want to build is something that feels like it's an extension of some specific operating system, that's a native app. And you'd be kind of crazy to use the web for that. But if you want to engage with the most users, you can. If you want to do it in a way that gives the user the best experience in terms of getting something done and being able to have sessions that span devices and things like that, that's the web. That's what the web should be good for. Someone came up to me and asked me for a selfie before. I was over the moon. But then he said, thank you very much for Facebook. He thought I was Mark Zuckerberg. Yeah, I think it is the blue. It's the blue. It's the face. And the face. It's probably also the face.