 Coming up on D T N S Windows 11 is coming. Should you be excited selling NFTs with featured journalism and a programming language that puts accessibility first? This is the Daily Tech News for Thursday, June 24th, Windows Day 2021 in Los Angeles. I'm Tom Merritt. I'm from Studio Redwood. I'm Sarah Lane. I lost the deck except for Justin Robert Young. And I'm Roger Cheng. The show is pretty, sir. We were just talking about company culture and how you get the kitchen you deserve on Good Day Internet. If you want to know what that means, become a member. Patreon.com slash D T N S. Let's start with a few tech things you should know. Google delayed its plan to block third-party cookies in Chrome until late 2023, saying that the delay will give sufficient time for public discussion on the right solutions and for publishers and the advertising industry to migrate their services. The company had originally announced it would phase out third-party cookies by 2022 as part of its privacy sandbox initiative. Indian Telco, Geo and Google announced the Geophone Next, what they call an ultra affordable Android smartphone. It's aimed at getting 300 million users in India that still use 2G networks to upgrade, start using 4G. The phone will launch on September 10th, run a highly optimized version of Android and, of course, include an LTE modem. Although price and exact hardware specs were not announced, it's expected to come in well under $100 US. TCL showed off a wearable display called Next Wear G a few months ago at this year's CES and now plans to release it in July. Initially in Australia for $899 AUD, it's about 680 US and eventually coming to other markets as well. It's not an AR headset, it's not a VR headset, rather it includes two 1080p micro OLED panels to provide the effect of viewing a 140 inch screen. So just an R headset? Snap reached a deal with Universal Music Group which will let Snapchat users use Universal's tracks in messages and posts as well as share links to full songs from streaming services. Snap first launched this feature in October with Warner Music and previously had access to songwrites from Universal and Sony but not the music performed by Universal artists. Back in 2018, President Trump signed the FOSTA-SESTA Act which set financial penalties for hosting illegal sex work related content. It's an exemption to Section 230 which otherwise does not hold platforms liable for content posted by their users. The US Government Accountability Office issued a report on how often the law was invoked since its passage three years ago. It was invoked once on June 2020 in a criminal case that's still in court and once in a civil case that was dismissed with no damages awarded. It did have the effect for causing many commercial sex platforms to move overseas or shift to conventional social media making prosecution more difficult. Alright folks, let's talk about it. Microsoft introduced Windows 11 and they packed in some surprises even though we had a leak of it ahead of time. The Microsoft Store will accept pretty much any kind of app a developer makes. Good old fashioned Win32, universal Windows platform of course, or even a progressive web app. And any developer can bring its own commerce engine keeping 100% of its revenue. Microsoft won't make you use its payment platform. If you bring your own, Microsoft gets nothing. Adobe Creative Cloud will show it off as an example. Also the Amazon App Store, yes you heard me right, the Amazon App Store will be available in the Microsoft Store meaning you'll be able to get Android apps and run them in Windows thanks to Intel Bridge, a runtime post compiler. This looks like it'll work like it does in Chrome OS. And Windows updates will be 40% smaller and can happen in the background so you don't have to reboot. On the interface side, the start menu moves to the center, loses live tiles, sorry folks, and has an integrated search bar. You can now snap two to six apps into something they're calling a snap layout that adapts to the screen size even when you unplug your laptop. So let's say you have five things tiled in various sizes, you unplug, you go from your big display to your laptop, it'll still be there, it'll just adjust all the sizes of the tiles. You can save these layouts and snap groups and pin them to the taskbar for quicker access so you can have multiple groups of different kinds of apps, kind of like virtual desktops, but different since Windows actually does have virtual desktops as a separate thing. If you detach your keyboard in a convertible, Windows 11's tablet mode is gone. It will not be changing the layout. It'll just note that the keyboard's gone and make touch targets like icons bigger. And if you turn it vertical, side by side apps will just switch to a vertical stack. The on-screen keyboard will still show up. It'll add swipe typing, emojis, and voice commands, and haptic feedback will be available for your stylus. A widgets screen will be able to slide in from the left. That'll include a personalized feed of things like news, weather, calendar, to-do list, recent photos, and even have a built-in ability to tip creators. I'm interested to hear more about that. Oh, and Teams is now integrated into Windows instead of Skype. A Teams icon will show up on the taskbar, pop-up chat list, all of that, so Skype, not going to do that anymore. Teams is the new future. The Xbox Games app supports Xbox's Auto HDR for automatically showing a game in HDR and the DirectStory JPI that can load a game faster by sending it straight from an NVMe drive to the GPU. And of course, Xbox Game Pass. You probably expected that. It's in this Xbox app and includes cloud gaming. Windows 11-ready PCs are available right now, so you don't have to wait. If you're buying a PC, you can get one that will work with Windows 11, but you can't get Windows 11 itself yet. A Windows 11 test build is going to roll out early next week. With any new OS, of course, there are new minimum requirements. To run Windows 11, your machine is going to need at least the following. 4GB of RAM, 64-bit processor with at least two cores running a gigahertz, 64GB of storage, a 720p display, a DirectX 12 compatible GPU, and a TPM 2.0 module. We didn't get a date for a consumer release, just a promise that it will arrive by the holiday season, so probably November it should be my guess. When it does arrive, Windows 11 will be free to existing Windows 10 users, arriving through Windows Update, just like any update to Windows 10 did. However, if you're going to do a fresh install of Windows 11 Home, you'll have to have an internet connection and a Microsoft account to set it up. Pro and Enterprise versions won't need that, but Microsoft Windows 11 Home will. So there you go folks, Windows 11? I haven't been using Windows 10 on any daily basis for some time. My single Windows 10 system can be upgraded to Windows 11, so I'll do that for fun. I gotta say, snap layouts, just because this particular laptop I'm talking about, I did used to use it with this monitor that I'm using with another machine now. And the layout was always an issue going from one to another, so little thing, but that's nice. The biggest thing that I'm intrigued by is the idea that Windows 11 is going all out with apps, especially in what I think kind of returning to an elemental DNA of what Windows is. This is not necessarily while it's going to look prettier, while it will indeed continue the design progression that I think has made the interface a lot more pleasing to the eye. I love the idea that it is a total free-for-all when it comes to apps. And Android, the Amazon store, build your own e-commerce engine, whatever you want, you're allowed to do it here. It draws a stark line in the sand with the Mac ecosystem, and I think it's cool. I'm glad that it exists. Yeah, what's not traditionally Microsoft about this announcement is that instead of introducing 17 versions of Windows, they're simplifying things. They haven't simplified everything. There's still little Windows complexity under the hood. You can still find some of the old interface elements under there. We'll see how it works in practice. But instead of Windows 10x, they put Windows 10x into Windows 11, and now Windows 11 will be adaptable. And so you don't have to have the burden of which Windows do I need. Theoretically, it's just going to run on whatever you install it on. Now, I'm very curious how that's going to work on ARM and AMD, especially with running Android apps with an Intel technology. There's a lot of questions to be answered there. What is traditionally Microsoft about this is how Sachin Adela and Panos Panay really pushed personal agency. You are in control, not us. We're not telling you what to do. We're making a beautiful, simple operating system, but we're not going to stop you from doing something. You want to put that start menu back on the left, go for it. Personal agency, they hammered over and over, and they talked a lot about being for the creator. So they were positioning themselves against Apple. It's the old-fashioned debate. I kind of love that. It's back, baby! We also must say goodbye to a few things. Windows 11 Cortana, no longer part of the boot experience. Skype, OneNote, Paint 3D, and 3D Viewer, no longer part of clean installs. You can still get them in the Microsoft Store, though. And the ribbon interface is going away from File Explorer. You're not getting tabs, but you're losing the ribbon interface in favor of a more touch-friendly control system built into the top section. Well, artist Mike Winkelman, you might know him better as Beeple, became one of the faces of the rise of NFTs selling his, every day, the first 5,000 days with an NFT that sold for a cool $69.3 million. Now, he's co-founding an NFT platform called Winoo, spelled like Renew, but with a W, designed to sell limited edition NFTs representing iconic moments in the careers of athletes and also artists. Pitchfork founder Ryan Schreiber serves as the editor-of-chief of Winoo to curate the moments sold on the platform and then build stories around them. The idea is that each NFT sale will be a series of items that tell a story, accompanied by feature-length stories, of why these events mattered. The first series of NFTs will commemorate Andy Murray's journey to winning Wimbledon back in 2013, ranging in price from $49 to $4,999 each. Some of the more expensive NFTs come with physical items. You see that sometimes these days, like signed posters, trophy replicas. The buyer of Murray's game winning points gets 30 minutes of tennis time with Murray himself, so that's a pretty cool IRL perk if you're a Murray fan. Wee news, first sale begins on July 2nd. I think this is smart. I have long said that NFTs are not quite the gold rush that people thought they were a few months ago, but rather just a proof of concept on whether or not people were ready to buy digital collectibles. This is taking that idea and stretching it out a little bit further by making it more exciting and informative for the consumer before they buy digital collectibles. I think the idea of partnering with the athletes themselves to make a big event out of this is a good idea, and now you're going to curate some long-form journalism along with it. But ultimately, I think this is what we've always understood NFTs to be a buying of a file, and that might be something that's foreign to some people, but to others, it's just another collectible that they can own. Yeah, it's way too early to say what NFTs will become if they become anything. But what I take out of this is that the direction they're going in now is an NFT alone, not enough, packaging it with a physical item. Interesting. Packaging it with an experience because we know that millennials particularly love buying experiences. Okay, now you're talking about something. You get 30 minutes to play tennis with Andy Murray, and you get to read this amazing story from the Shriver-led editorial team. I think that's interesting. I'm not saying it'll work, but I think it's an interesting approach to say let's bring people in with featured journalism. This could be a way to pay for journalism and then sell them memories. It'll be worth it because people buy experiences now for charity, right? The fact that it'll come with an NFT will be its own thing. If anything, that's like the least of their worries. You're always going to be able to have somebody spend a lot of money to play tennis with Andy Murray for 30 minutes. It's them moving the $49 stuff, and I think what they hope is that you do this great feature story and somebody gets the end of it and they're really excited, and then it's like, would you like an NFT for 50 bucks? Man, that moment in 2012 when he apologized for losing. I could own that. I could own that moment. That's what they're going to try to sell you. Alphabet's DeepMind announced a new partnership with the Geneva-based Drugs for Neglected Diseases Initiative, and it uses AlphaFold AI to discover new treatments for chongus disease. Oh, geez. Leishmaniasis. Both deadly diseases that impact millions of people, AlphaFold was used on January 2020 to map out a number of the SARS-CoV-2 virus proteins that were later confirmed to be accurate, and in November 2020, it topped the protein predicting challenge called CASP, or Critical Assessment of Protein Structure Prediction. So it is unprecedented at figuring out how proteins fold. For chongus disease, researchers already had identified a molecule that can bind to a protein on the parasite that causes chongus disease and kill it, but they used AlphaFold to generate a prediction of the protein structure, something that would have taken years otherwise. That prediction will hopefully let them design drugs that can bind to the protein in different ways, making sure they kill all of the parasites in a patient's bloodstream, which is required to cure the disease. A similar approach will be taken to other diseases in the coming years. AlphaFold hopes to expand structure prediction, making it faster and more accurate. If it proves to work, it can help doctors who lack resources to study infections work on new treatments, no matter where in the world they are. Although some researchers caution that algorithms are never perfect, and there will still be some instances where it doesn't work. Yeah, so accessibility is an interesting part of this. If this works, you just need the computing power and a connection, which is much easier to get than a fully equipped lab that you can spend years investigating. So the accessibility, I think, is incredibly important here. Obviously, it's early days and it's important to caution like, hey, it may not work that often. We've only tried it a couple of times. So one or two hits doesn't mean that we'll keep getting more, but obviously worth pursuing. It's helped with COVID. It seems to have helped with chongus disease, and it'll be interesting to see what else it might be able to help with. Also, I mean, just taking into account that there is an initiative called drugs for neglected diseases. Meaning, okay, this is a disease that affects a lot of people, often is a fatal disease. It's extremely complex. There's folding of proteins that it's just a huge amount of resource and time suck, even for people who dedicate their careers to this. Anything that helps the process along is a great idea. It's a great use of AI. Yeah, that's a really good point because it's a neglected disease because we only have so much resource, right? And if you can make it less resource intensive to investigate, then you can devote more resources to it because you don't have to devote as many. And it also takes out a layer of specialization. I've said it here before that AI is a very, very, very good guessing machine. That's a better way to think of AI than to think of it as some kind of magical oracle or a way that technology can create reality, right? It is good at guessing, so it will never be perfect. But also when you think about it, so is research. The best researchers in the world are taking their best, most educated guesses and trying to prove whether or not it is real or not. It is all trial and error at the end of the day, and this can take a huge element of specialization and just give you some good guesses. And whether or not this is going to be better or worse, I think is the wrong way to think about it. The best way to think about it is it's a replacement on some level that you can just move forward in areas that probably wouldn't have the same kind of ability. And it replaces tedium. The guessing we're talking about is we think the protein folds like this. Let's see if we assume that it does, whether stuff we do works. That takes years to figure out, like, okay, we're pretty sure it folds like this. This takes the tedium out of that, where it's not like being smart in guessing. It's like working through a bunch of possibilities. And that's what machine learning can short circuit really well in this particular case. Hey, folks, if you're listening to this and you're like, gosh, I bet Jerry had the best face just now when he was reading that. Well, you could see that in our video podcast. If you want DTNS as a video podcast, get the video RSS feed at dailytechnewshow.com slash subscribe. Often we talk about accessibility. And a lot of times we discuss the way applications work for people who are maybe vision impaired, deaf, limited dexterity. But what about the tools that create those applications like programming languages? Andreas Stefik has developed Quorum, a programming language that takes blind users into account. At least it started that way. He's a PhD and associate professor of computer science at the University of Nevada, Las Vegas. Andreas, thanks for joining us. Thank you, Tom. Well, thanks for having me. How did Quorum get its start? So Quorum started out when I was in graduate school. I was scouring for a PhD project and I was reading a number of mailing lists of problems people had. And I came across these mailing lists of blind computer programmers. And the problems that they were discussing were really fascinating. So as you can imagine, like how do you get access to someone tactically through Braille? Or how do you listen to something through audio to try to represent? And the interesting, the thing I found really compelling about it was is that if you have a disability and you want to invent some kind of technology to make your life better. Well, how to use the programming environments is exactly what you would need to use. So if those aren't accessible, it's like a chicken and egg problem. So we started by looking at that and then we got some support from the National Science Foundation to start working with schools for the blind and visually impaired. And we just sat around basically talking to children and watching them and seeing what problems they had and trying to get a better sense of the situation. That's where it sort of started, I guess you would say. What did you reevaluate in developing a programming language for vision impaired programmers? Yeah, that's a good question. So there's a couple of things that we noticed sort of pretty quickly. One of them is when we had children use them. You see all sorts of neat things because kids, unlike professional programmers where they've been using technology for a long time, they're used to it. You might even say they're biased toward whatever they're used to. They're a little set in their ways, sure. Yeah, there's actually some strong evidence of this. You can actually calculate, there's an equation. I swear this is true. There's an equation to calculate how much bias someone has gotten toward their programming environments. Anyway, it's not important, but for blind children, we basically sat and we'd watch them and we'd let them use their screen reading, their sort of talking devices as they programmed. And we basically like, hey, how hard was that? Or did you understand what that meant? And when we did that, we observed that a lot of the things that we take for granted in programming were actually really hard for kids. So for example, if you want to tell the computer to do something over and over again, that's traditionally a loop. And in computer programming, you would often use a phrase visually and it would say something. It would look like for left, right, and into I equals zero, semi-colon, I less than 10, semi-colon, I plus plus right, right, left, right. Which is not easy to understand in audio. If you're talking, if you're making a tool for people that are blind, you want the most important content first, and that must be short. Right? So like, if you have an error, it might say name missing as the very first thing. And then it will give you some other information on top of that, that sort of idea. So when you say it's evidence-based, I think we kind of get a sense of what that means. But explain that a little more. What do you look for? How do you change a programming language based on evidence? Oh, that's a really good question. So other communities outside of computer science have really strict standards about what they mean for evidence. So the one that I'm probably the most familiar with is the medical community. So here's an example. If you want to sell a drug, you have to go through a series of phases of evidence-gathering. And those are imperfect. They're flawed. They have their issues and stuff like that. But the idea is you first test with animals. It doesn't apply directly to programming, but you'll get the idea. And then you sort of scale those up, testing certain other things along the way. And we do the same thing with programming languages. For example, in quorum, if you have a loop, this is one of the simplest things we've gathered evidence on years ago. Instead of saying for left, for and blah, blah, blah, we say repeat 10 times. And why is that? Because we've done surveys on the word choices. We've done formal randomized controlled trials on comparing them each together. You know, like one group gets this, the other group gets that, and we see who wins. And we do that across hundreds and hundreds of studies and other academics do it too. And we see if we get the same answer. I feel like even if someone's very excited about this, even if they understand the ability to make it more accessible, makes it more widely applicable, there's probably people wondering like, yeah, but I can't get a job programming in quorum. Can I? How does quorum fit into the practical level of like, I need to get out there and do some work? Yeah, that's true. I think that like today, one of the downsides of focusing on the evidence base and the K through 12 space first is that you aren't going to be able to get jobs in quorum right now because we haven't even tried to like push it up to professional practice. But there's two questions that I think are interesting. One is, which language do you use when you want to get to professional practice? It turns out universities don't really have standards, right? And we also know that these languages change. Like 20 years ago, people weren't programming in Python. And that's the truth. But today that's very common. So we actually just got funded by the National Science Foundation like a month or so ago, our first attempt to try to push up to communities. So we started gathering evidence across different professional groups. We found one that was the most amenable to change was actually data scientists. And it makes sense, right? Because they have to learn not only is it for a left brand type junk, but they name everything after the mathematician's name or by single letter names. And they don't like that, right? Like t-test, f-test, pairwise t-test with a Bonferroni correction. Those words won't even mean anything to people unless they're already deeply embedded in the community. So what we thought to do was to run a lot of the same types of studies to make data science easier. And that is for professional practice as our first attempt. We have about four years to solve that problem. So we'll see if we start to make inroads on the professional practice over the next couple of years. If people want to learn more about quorum, get involved, where should they go? We have a website, quorumlanguage.com. And you're welcome to tweet at me at Andres, if that's your wish. Andres, thank you so much for chatting with us. This was fascinating. No problem, Tom. Thank you very much. I really appreciate the invitation. We also want to give special thanks to listener Daniel Hurtwig for putting us in touch with Andres. And if you want to catch the full 18-minute interview, and I think you probably do, we'll be posting it up this weekend for our patrons. All right. So Rembrandt von Rhein, a pretty famous artist, may have heard of Rembrandt, also has a very famous painting called The Night Watch. It was originally completed in 1642. And then it was trimmed down in the early 1700s. Yeah, the sides were just trimmed off because it was supposed to fit on a new wall where it was being hung and the wall couldn't fit the original painting. Now, although the trimmed portion hasn't ever been found, the Associated Press reports that researchers have been able to reconstruct the missing pieces. It's very interesting how they did it. Using a smaller copy of their original that was painted at the time by Garrett Lundens, then they created scans. They used scans, X-rays, and 528 digital exposures taken of Rembrandt's original painting to train an AI model to imitate Rembrandt's style and fill in the blanks based on Lundens' copies. So, yeah, you want how the original artist would have painted the missing pieces, but we kind of know what the missing pieces look like because we already had a copy. The work was conducted as part of the Operation Night Watch project. It is being exhibited at the Honor Gallery in Amsterdam's Rijksmuseum. There are two new faces that you can now see present on the left side. A small child was previously running on a frame, now seen leaning on a railing. The whole human is. So, if you're a fan of this kind of artwork, it is a pretty cool project. Yeah, I learned today that back in the 1600s, Rembrandt was not as respected as it is today, and you just lice it up if you had a small frame. That's crazy. I mean, it's a look. You got to fit it on the wall, man. It brings the whole right. That's incredible. It's good stuff. Yeah, check it out, folks. And the Verge has a link to the version that exists where you can like zoom in and see that the kid who looks like they're running out of frame. And then you can compare it to a smaller version of the restored version where you see like, oh, right, they weren't running at all. It's pretty good. Let's check out the mailbag. So we had a conversation yesterday with our guests about how we all back up our photos. You know, Google Photos is not going to let you do unlimited anymore. So we just sort of went around the horn and talked about how we all do it. Chris says he uses Google Drive to store photos. After the photos are uploaded to Google Drive, Chris says he can organize them into folders, rename them, have folders within folders, download those folders to his MacBook and copy them to USB for off-site storage. Or Chris says give to my daughter or another relative. Also, once the folder is established on Google Drive, you can directly upload new pictures of the same ilk to that folder. The upload process can be both push and pull. Chris says in many cases I've shared folders on Google Drive with family and friends so they can see our events, Easter, Christmas, et cetera. And in some cases, share technical pieces with other folks who share the same interest. Oh, that's so cool. I like that we're getting these. Daniel also sent us what Daniel does for organizing photos. I love getting these different perspectives on this. Chris's is kind of like half DIY, but also taking advantage of cloud. It's really cool. Totally. I also realize how many photos I clearly do not share with family and friends. These are all really good solutions that everybody on the show except me yesterday was like, yeah, you know, albums to share with family and friends. So yeah, this is a great one. A lot of people use Google Drive. Maybe you could use it for photos if you're not already doing so. If you have some photo tips, life hacks, tricks, all the things, send those to feedback at dailytechnewshow.com. Anything really that's on your mind, questions, comments about the show, we'll take it. Thank you in advance. Also shout out to patrons at our master and grandmaster levels today. They include Hi Tech Oki, Martin James and David Mosher. And guess what, everybody? We have a brand new boss and his name is Roy Riemann. Just started backing us on Patreon. Thank you a million Roy. Roy, Roy, Roy, Roy, Roy, Roy. Yeah, I'm staying. If you become a brand new boss of DTNS, we celebrate. We're going to shine. Yeah. Yeah. Exactly. We're going to have pie. We're going to have cakes. It's going to be the best day ever. And Roy. I'm going to Casa Vega. Also want to give a big thanks to Andreas Stefanik, who you heard earlier in the show. And also Justin Rubber Young for being with us today. Justin, you're a busy man. How do we keep up with you? Oh boy. Well, of course on PX3, I just came back from New York covering their primaries. Big, interesting race there. You can hear my little documentary of the final day of the trail. But really the big noise is all about world's greatest con. Thank you to all the DTNS listeners that downloaded it last week when it first debuted. We became the number three podcast in all of the history category of Apple podcasts that is gigantic and huge. You guys are amazing. And now episode three is out and live. If you've ever wanted to know how you get a corpse from London to the shores of Spain, while making it seem like a fallen war hero, every grimy, dirty detail is explained in this episode called In the Flash of World's Greatest Con. That is quite a teaser. We're live Monday through Friday on this show. Here's your teaser for 30 p.m. Eastern 2030 UTC. Be there. Be square. Find out more at dailytechnyshow.com. We'll be back tomorrow with Kristoff, Zajik, Denik and his personal experience with accessibility. Also, you'll be able to find some stuff. Talk to you then.