 Daily Tech News show is made possible by you, the listener. Thanks to all of you, including Kria Artem, Tony Glass and Peter Bohack. On this episode of DTNS, why it makes sense for Apple to take Google's AI? Sony pauses PSVR to production. And Scott talks Adobe's AI advancements from the Game Developers Conference. Well, the advancements are not stopped. This is the Daily Tech News for Monday, March 18th, twenty twenty four in Los Angeles. I'm Tom Merritt in Salt Lake City. I am Scott Johnson and I'm the show's producer, Roger Chang. Roger, you have no T's in your name. Oh, I'm sorry. Does that I just I never noticed that before because I made a big deal of saying Tom Merritt and then Scott Johnson. And I'm like, oh, but Roger, I could be Roger Roger Chang for today's show. He's Roger Chang just today. Microsoft just announced a surface and AI event will happen May 20th in Seattle. That's the day before their Build Developers Conference. And separate, if you're like, wait, don't they have a surface event coming up later this week? They do March 21st. So we're going to get two surface events. This one seems to be enterprise and the one in May. Sounds like it'll be consumer. And now to the rest of the quick hits. The Apex Legends Global Series Championship has postponed its North American finals after attackers forced cheats to be applied to the games of two professional players during a live stream Sunday. A cheat monitoring account on X speculates that the attackers may have found a remote code execution in Apex Legends, anti-cheat protections, which is ironic. Both players realized the cheat had been applied pretty fast and then abandoned gameplay immediately. EA and Respawn have not announced when the competition will resume. Google researchers have written a paper describing a project called Vlogger, which can generate videos of people speaking and moving based on a single still image and an audio clip. The team used a diffusion model, which is good at generating images from text. The model does not rely on training training data for the specific person or face detected or the cropping part of it. They used a data set called Mentor, which is larger with more diverse sets of data than previous data sets. The resulting video matches the audio clip with facial expressions and movements, though it has to be short and have a static background. By all accounts, it wouldn't fool anybody into thinking it was real, but it's still an interesting advance. Use cases could include matching lips in video to audio dubbed in other languages or creating compelling virtual assistants. Interesting and going to make a lot of people decide that it's scary. Qualcomm officially launched its Snapdragon 8S Gen 3 chipset Sunday, bringing on device AI capabilities to mid-range mobile devices. The 8S Gen 3 is similar to the main 8 Gen 3 chip, except that the S means it has one fewer performance core and runs at a lower frequency. It also uses the previous generation X75G modem, which still has Wi-Fi 7, but it's not the current modem. It supports hardware accelerated ray tracing, though, on device multimodal generative models, up to 10 billion parameters so it can handle Meta's Lama 2 or Google's Gemini Nano, for example. Honor, Aichu, Realme, Redmi and Xiaomi have all said phones with the new chip will come to market in the coming months. Watch for it. LinkedIn will bring some puzzle games to its social network called LinkedIn. LinkedIn told TechCrunch it is working on games called Queens, Interference and Cross Climb. App researcher Nimwa Auji found code referring to the new games that could rank scores based on place of work. A spokesman told TechCrunch, quote, we're playing with adding puzzle based games within the LinkedIn experience to unlock a bit of fun, deep in relationships and hopefully spark the opportunity for conversations on, quote, no word on when this might happen or when they may launch the games. All of you like Stoic Squirrel, who are like, why would they do this? Come on, work has to be fun, too. It'd be fine. I don't know. I don't know why they do it. It's a day full of predictions about Apple's future product releases. Tipster Tech Rev posted a screenshot of Apple's roadmap taken from Samsung Securities. That's a financial services company. So don't get too worked up about it being a Samsung company. Document starts in 2023 with accurate info on the iPhone 15 mentions of the Vision Pro and then a rumored OLED iPad coming in 2024, which we'll find out soon if that's true. Then the list gets interesting. An iPhone SE4 in 2025, a foldable six and eight inch iPhone and a low end XR OLED MacBook in 2026. AR glasses and a 20 inch foldable iPad supposedly on the roadmap for 2027. Also, Bloomberg's Mark Gurman has some thoughts of his own, saying sources tell him that Apple will have two new AirPod four models coming later this year and an Apple Watch series 10 with a blood pressure monitor. All of that coming this autumn. And that is a look at the quick hits. All right, Bloomberg all over the place today. Mark Gurman doing a lot of good work, as he always does. Finding sources that say Apple and Alphabet are in negotiations for Apple to use some of Google's Gemini to power AI services in iOS. Apple may also have had talks with open AI and might even consider anthropic as well. Bloomberg's Mark Gurman wrote, the two parties haven't decided the terms or branding of an AI agreement or finalized how it would be implemented. So they haven't signed this yet. They're in talks and there might be some questions about do we say it's Gemini or is it just, you know, working behind the scenes to provide answers? I like and gadget's DaVindra hardwars take on this the most. Apple would partner with Google for cloud based AI services the way it partners with Google for search. All of Apple's own AI tech will be on device. Now, Bloomberg implies the Gemini would power cloud based generative models for Siri and some of the other iPhone apps. So that's consistent. Apple has been publishing a steady drumbeat of its own research into generative models. And often the research that it's publishing is on device stuff, not in the cloud as a way to protect private data. Scott, what do you think? Are you on the side of like, oh, this shows Apple's no good at AI? Or is it more like what Engadget's Hardware says? I think it actually fits a piece of the puzzle that I hadn't considered, which is, well, great, you're going to do all your AI advancements on device. That sounds like an Apple thing to do, given their strict adherence to security things. And that's great. I actually kind of really like that. But there is a ton of stuff that you wouldn't be able to do that you will need cloud enabled services for. And so what is their solution for that? Well, in a weird way, it's like a search engine choice in that regard. It's like, well, we got to do it outside. Who are we going to use? I made a quip earlier today and I don't think it'll hold, maybe it will. But maybe one day we get to choose. We get to go choose our default AI provider like we do our default search engine. That would be kind of cool, I would think. The only thing that would cool me on this if I was on the Google side of it would be that little niggling memory that Apple went and yanked the Maps rug out from under you and started their own Maps thing. And caused a bit of a kerfuffle there, which also didn't go real well with fans of the Google Maps because we didn't really like Apple Maps when it first launched. It's fine now, but those memories are still kind of hot in my brain. And so I worry about that kind of stuff a little bit. But as far as why they would choose them, say over Open AI or Microsoft or anyone else for sort of a default cloud service choice. I mean, maybe it's apples and apples in the sense that I I'm not going to care as a user that much. I just need to get what I get and Gemma and I can do it. So it's fine, I guess. Yeah, I think the narrative often out there amongst folks on social media and elsewhere is, aha, Apple would like to lock everything into their own ecosystem. And so when they don't, it means they're incompetent at that thing, right? That's that's the extreme way of of having that that conversation. And I don't think it's 100 percent wrong. Honestly, I think Apple would like to keep everything under their control as much as possible, but I think they're a little more clear-eyed about it. I think they understand that they won't ever be good at everything. And when they are in when they determine that something is necessary to have a good user experience in their opinion and they have a competitor or not even a competitor, they have a chance to to purchase that service under certain circumstances, they will purchase that service. And that's exactly what they did with search. There have been so many rumors of Apple developing its own search over the years. And I think Apple probably is always investigating search and always realizing, yep, so far, still better to outsource web search to Google. We can do on device search really well. And that would be similar to what's going on here. Running a cloud based AI gives you advantages over on device and it's expensive. So if Apple can take advantage of the what is it called? The economy of scale of hiring Google, who's spending all the money running it. Then that makes sense for them to do it. And they'll focus on the on device stuff, which is, by definition, more private, which is fits right into Apple's marketing. Yeah, you just said it. Focusing on the on device stuff seems like what they would want to do. Because what does that do? That ultimately sells phones and devices and computers. That's that's a huge selling point of Apple products, or at least at least they feel very strongly that it is. And so for them to say, we're not really interested in competing in the cloud based stuff, we're not really trying to make another service that we now have to compete on a whole another level on that has nothing to do with hardware, really. Yeah, they're saying, well, let's do the things with AI. Let's acquire the companies, R&D, the stuff we want to work on that will make on device best in class. Yeah, and then we'll hire out for the other. That makes perfect sense to me. They could license some of their technology to people to use to for on device stuff, like why duplicate efforts when they're anthropic and stable diffusion and mid journey and Google and meta and like everybody else out there doing the cloud based stuff. It makes sense to work where not as many people are working. And I'm telling you, like, I bet WWDC is going to be a litany of here's all the AI research we published in the last year, because it's every month, there's at least one, if not more Apple scientists published a research paper about a development that improves something usually on device, not always. This month, nine to five Mac reporting on one that is a method of training large language models more efficient and more efficiently. And these are multimodal models using text and visual information. The models that Apple worked on are good at context learning, multi image reasoning, allowing for more open ended questions. So you could have a made a difference in choosing the right image. They figured out that resolution and encoding made a difference in how well these models worked. This is all public information that other people can use, but it just shows that they got the chops, right? Yep, gonna be fun to watch that event. Because I feel like you're right, it's going to be a lot of talk about this stuff. It'll be good. This is not as good news. Bloomberg sources say Sony has paused production of the PlayStation VR two headset, because it's built up too much inventory. So reportedly, Sony made 2 million units. And if you look at IDC estimates IDC estimates, they have sold 1.7 million. So there's 300,000 PS VR twos just sitting around. IDC also estimates that during the holiday shopping season in Q4, PS VR two sold its slowest. Usually during the holiday shopping season, it should sell its fastest, but they sold just 325,000 units. And there's a lot of speculation about this. Sony hasn't released that many titles there's only been four VR experiences for the PS VR to they closed the London studio that was one of the few studios in house working on VR games. You can't play PS VR games on the PS VR two. So if you upgrade you lose access we don't lose access you have to use the old headset to access the old games. Sony did say that it's testing PC compatibility. That would open it up to a whole new market. But the other reason that it's slow seller would probably keep it from being adopted massively in the PC area because it's $550. That's more expensive than the PS five. Scott, is this thing toast? Ah, kind of. I mean, I'm getting flashbacks of the Vita the Vita, which I believe was released in oh eight or so had so much potential and was such an awesome device. Yeah, the Vita was great. And we a lot of people bought in and loved it. We liked Sony's effort at competing in the mobile market, mobile gaming market anyway and giving Nintendo some competition. Well, in this case, you know, new frontier, it's a whole new bag in the in the VR one device did pretty well for Sony on the PlayStation four at a time where maybe the hype for what then was cutting edge VR was a little higher. People were more excited to bring this into their homes. You know, we were just sort of getting used to that stuff. And so it did pretty well. I think they banked on that for a long time. I think this price has always been kind of wrong. It's just too much. They also just released a brand new PlayStation five last winter, right for the holiday season, the slim and the version of that device sold pretty well. But when you're already out buying that for $500 plus, and then you see this next to it, you're like, Well, wait, I what do I get with this? Well, we used to have PS VR. Can I play all those games? No, I'm sorry, all your old games don't play there. Well, some of them do. Yeah, but you got to pay an upgrade price some are 10 15 bucks to to upgrade or their entirely are entirely new purchases. Oh, like there is a lot of reasons I could talk people out of picking up a PS VR that are reasonable reasons not to grab it. And I think that's that's to its detriment. I also just get the feeling that they are like, Well, we had to follow through we had to launch it. It's out there. 1.7 million isn't terrible. But then there are games like Hell Divers 2 on their very own platform, a console exclusive right now that has sold upwards to like eight to 10 million copies. Yeah, just a video game. And so I don't think they've they've made the case. And marketing around it has been kind of abysmal since day one. It's never been very strong. They haven't really pushed it very hard. Again, it felt like the tail end of the Vita to me. And that price point's never been right. So I don't know what this means. Ultimately, the future of VR on Sony platforms, specifically console platforms, I think is in real question now. Like, I don't think they're going to do it again. I think I'm a little surprised. Yeah, they blew the lead. And they not only blew the lead, but they they got I wouldn't say they get cocky. Like it's a generational leap in quality. The PSP are two is without question, a better device than the PSP are one in every respect. The problem is you I think you assumed too much about that loyalty because a lot of people got the PSP are one played a little bit and then just got busy with regular gaming again, because that's what we do. We kind of fall off this stuff until it becomes smaller, cheaper, more readily available, more title, all those sorts of factors kick in, then they then we're talking a very different market. But they're in a market that is already difficult to convince people to do. It's hard to advertise without putting it on your head and showing people. And you have real competition from meta and others where those are better solutions because they're just a little bit more versatile. I can use a meta quest two or three on a PC. I can tether it. I can untether it. I can use it with its own ecosystem. There's a lot of reasons why I should do that. On the pro version of it. Like, yeah, there's other apps besides games, which not really with the PSP are to know not really. And there are a couple of games that are exclusive to the PSP are to that are good, but they're not enough to make the difference. Like they really don't have any killer apps. So at the end of the day, I just feel like, you know, if they've got another pro coming as rumors are talking about, and you guys brought up on the show and it happened, but they've got a pro coming out a mid a mid generation upgrade coming out this winter. They're asking for another $500 console purchase from the faithful. Where does VR fit into any of this? So I don't know. A dance here original question. It kind of feels dead. Yeah, it may not be toast, but it's browning. Yeah, it's in the it's in there. And if we don't get it out, it's going to smell the house up butter. There's not enough butter to pull the butter out or get it out of the oven. This metaphor is perfect. It turns out I'll throw I'll throw one last little piece of speculation out here. What if the move to adapt it for the PC isn't just a way to maybe sell a few more, right? That's not going to save the thing. But a few people will Oh, if I can use it on my PC, then maybe I'll that justifies that you know pushes them over the edge. What if it's because they're working on a PC headset that they will put out later that would work for both the P like the PS VR three is meant to work on both. So they build they build it with that in mind. See, that's interesting. Maybe because they've got the tech they have the quality they have the game studios. And if they suddenly made it compatible with the vast library of say steam VR or anything else on PC. Now you're talking, but I don't know if they're willing to go that far. We'll have to see. Yeah, I don't think I can see a Sony PlayStation VR headset that runs Windows either. So you know, it's not going to be standalone no matter what. Well, folks, I know a lot of you are Apple fans and you're like, hold on, you just went like more than five minutes not talking about Apple. We got an entire show for you. It's called Apple Vision show. Now, they do talk about the Apple Vision Pro quite a bit on that show. But they also talk about other things from Apple. And it's not just a rumor show. This is going to help you figure out how does Apple fit into your life? Maybe you don't like Apple, but you're wondering that what's the big deal? Why? And why does everybody always want Apple? This is the show for you. Sarah Lane and Eileen Rivera, Apple Vision show available wherever fine podcasts are vended. And you can find out more about it at AppleVisionShow.com The annual Game Developers Conference is on in San Francisco. Obviously the conference covers game development, hence the name. But they also cover a lot of things related to game development, game design, development frameworks, and Adobe, you may have not thought of as someone who had participated at GDC, but you know, in your mid designing games, you need to design the art. So Adobe is there showing off two new integrations for its design software suite. One of them is called text to texture, which can generate photorealistic textures from a text prompt. So scaled skin or woven hemp fabric, and then the texture could be applied to a 3D model. So the texture itself is 2D, but you know, you need to make something look a certain way in some fabric or somebody's skin that you're working on in a game, you can apply it to that 3D model. The other one's called generative background kind of does the same thing for backgrounds, you type in some text, and it makes the image in a way that can be used for a 3D scene. So one of the examples they gave in their demonstration was abstract wave swirl with magenta blurred ribbon. There was it was a longer than that actually, but that gives you the idea. Then it creates a 2D background that can be adapted to appear 3D in the background of your image. Text to texture is available in the beta version of substance 3D sampler 4.4 and generative background in the beta for Stager 3.0. Scott, it's not I know you don't work on game design art necessarily, but you you work in art and I know you work with a lot of Adobe stuff. So I'm sure you got an insight on this. Yeah, so for the last couple of decades, we really haven't changed much or altered too much of how texture mapping is done in video games really since the polygonal era of games. There have been the need for texture artists to create textures for walls and floors and character models and whatever it may be. And for a while they're back in cheese early 2000s late 90s. I did do some skin work for an Unreal Tournament mod where I took the character models, flayed them out and then did texture map creations, different costumes, different faces, that sort of thing on top of these things. And it's really stayed sort of the same. That is not that different than what we do today. This marks potentially a massive shift in how that work would be done. And once again, we're seeing an application of AI and machine learning. And in this case, art generation that would really cut down on the pipeline of the work. And so I know there are people excited about this who run art departments or who run game development in general, small teams to big teams, who are thinking about the hours and hours and hours of time this will save. Now I know that this is one of those deals where it affects artists directly. This will affect artists maybe in a different way. We're always talking about artists work being taken and then used in generative art. This is different than that. This is saying, we used to have Bill sit there and create texture maps for walls and it would take him this many hours, hundreds of hours to handle this one level or whatever to build all those textures out. It's not always that much, but sometimes now we're saying, we don't need that anymore. Or we need Bill just to say these prompts and to whittle them down. We've already got the 3d models from the 3d team. Now all we're having to do is apply these awesome new textures that took us no time to make. So this is one of those major time savings corner cutters that to the player should be transparent. They won't notice any difference. They'll just see textures and they'll say, oh, that's a wall and they want to think about it. Obviously the games are a lot more capable now. The engines are capable of showing much more and there's more to the games than just flat textures. However, this does seem like maybe a way to just hurry up that process. I'm a little torn on it though, because I do worry there are a lot of texture artists in this business like tons. Every company has many of them. It's never just one guy. This indicates to me, based on what I've seen, that this could be one guy who handles all of this. So I'm trying to keep that separate in my head of like, well, what does this do for that particular job title? It's kind of hard to do that to have that conversation. But from a pure technical standpoint, it's impressive as crap. It's I got to make a whole castle level and you're telling me all I have to do is say, I need this brick on the walls, I need this clothing on the people, I need this kind of texture on the hats, whatever it may be. And it's done. And all you got to do is just basically proof it and say, well, all this works up this. Okay, change that one little. Okay, now we're good. Like that. That is a massive shave of time for these guys. And they're not going to be able to ignore it. Like Adobe's right. The concern is always to jump to the conclusion that a company will obviously fire all of the texture artists if they don't need more than one, right? And I'm sure some companies will do that. There's there's no end of ill advised decisions out there. But my my personal opinion is if a company is smart, they don't just fire all the texture artists. They keep as many of them as are good at other things. And they look at the fact that they can now put a game out faster, because they saved all this time. And they can afford to pay people to do other things that they couldn't afford to pay those people to do. So what are those things? And let's have a competitive advantage now, because we're using this, because we will be the ones who decide the right way to use these artists to make our games better and stun the world and get our games out faster. That that's what you should do as a company when you get this kind of tool is not just immediately jump to cost savings. Because if you do that, you're not improving your product and some other competitor will and then you're going to lose. Yeah, you don't want to double dip on this. If you say, well, this will save us time. Cool. This will also save us money. Well, that's true, too. But if you're going to do both of those things, I think you are at a competitive disadvantage. The fact that you could bring a tool like this into your current workflow, and you've got a whole team of artists, and they now have access to this. Their jobs are quicker. They also still have the expert eye you want for this sort of thing. We don't want robots just making textures. Everything's going to work out. Yeah, yeah. You still need some humans. So you still need those text props. Yeah. And no, no matter how intense this gets, you're still going to need them there to, in a way, proof, it's not proof reading. It's proof viewing, I suppose. But they need to be able to sculpt it and and bend it to its will to its to the human's will. Yeah. And until we don't need that and I don't see a time where we don't for creative objects or creative projects like this, then I think you're right. You can you can gain your competitive advantage, advantage, save money because of time, but also put out more products more often, which means more value, more people stay, less layoffs. Like that's what people really want in the end anyway. So yeah, it's interesting. And it won't just be Adobe. There's going to be other solutions like this popping up left and right. You watch for sure. And it can lead to better games. The smart companies will use it to make better games. And then there will be other companies who don't. Yeah. And they will rightly be reviled by people. That's right. Yeah. Alright, let's check out the mail bag. Jacob wanted to add context to our discussion on Friday about Apple MacBook Air being sold at Walmart. If you missed that story, Apple MacBook Air being sold at Walmart for $699. It's a discontinued item. Apple doesn't sell it new anymore. Jacob said this new deal is a recreation of a sales channel Apple deprecated many years ago in favor of its own retail stories. I worked for 14 years in an authorized Apple reseller computer stores and W which ran the Mac store retail chain and sold online as PowerMax. It began as an Apple to dealer had a strong relationship with Apple for decades. Apple would even contribute funds and guidance to our store remodels. I'll skip the part where Jacob talks about the sad severance of that relationship and how disappointed everyone was in that. But he says in my experience, Apple has never known how to sell anything other than the latest and greatest. They relied on third party stores that knew how to fit customers with the right size tools for their needs and budget. Walmart is an easy partner to move some old stock computers. But I also suspect they will not serve the brand as well as the old authorized reseller network did. I agree. I also missed that. Do you miss that? I remember going to like comp USA that have a whole yeah. And it was a specialized little zone and that people knew their stuff because Apple demanded they still have that a best buy. That's true. That's true. You know, but it's not the same. It's definitely not the same. And along those same lines regarding that same story, a shocked McVangelist wrote in and said, well, Apple did discontinue the M1 MacBook Airs for individual sale. They have kept it around for the K through 12 educational market. They did the same thing with the 21.5 inch iMac. So it's not uncommon. It gives schools a less expensive one to one or learning lab option. But what's interesting is that Apple offers schools the 128 gigabyte model at the same price that Walmart sells the 256 gigabyte model and Apple makes schools pay $100 more for the 256 gigabyte model. Well, look, you know, the the future of our children isn't cheap, Tom. So yeah, yeah, schools get on down to Walmart if you want to buy a MacBook Air with 256 gigabytes is what that told me. And then finally, John wrote in and said, I listened to know a little more. If you missed it, know a little more know a little more.com had an episode about bite dance, going through like, here's who actually owns bite dance. Here's how they operate. Here are the actual risks of the algorithm and private data and all of that stuff. And John said, I listened to that episode driving to work this morning. It was very helpful to have the background details you provided. I've also been listening to Justin Robert Young on We're Not Wrong. And they talked about the political aspects of the bill. It was nice to connect the two viewpoints. I appreciate the hard work that you and others put into researching these topics while I work away on other matters. It helps me be prepared to discuss these matters with friends and family. John, that makes me very happy patrons. You got a version of that episode that doesn't have all, you know, all of the production elements. And if you're not a patron, you can go get it for free at know a little more dot com. Nice. Scott Johnson, thank you for being here on a Monday. Oh, I don't mind at all. In fact, you know what, people say they don't like Mondays. I think Mondays are just fine. You're like, Bob Gildoff on this, I disagree with you. I do like Monday. Just this one thing, though, just that one. Well, thank you for having me. I don't know if people know this, but if you're trying to track down more of what I cover, and you may have noticed a pattern when I'm on the show, a lot of the big gaming tech stories come up. We talk about them here. And guess what? We tear those things apart even further on my show called core. I do this with two co hosts. I do it on Thursday nights. And it's a great live watch if you're if you're not busy. But if you go to frogpants.com slash core and finally get the podcast version anywhere. And if you want to deeper dive into the world of gaming, the industry all the way down to the little stuff that came out this week and what we think, check it out. That's frogpants.com slash core. Patrons, stick around for the extended show Good Day Internet. We have more discussion of GDC. In fact, the game developer conference is adapting to the fact that a lot of its attendees don't have jobs. Been a lot of layoffs in the game industry. Smart move of GDC help them get jobs. We're going to talk about it. You could also catch the show live Monday through Friday, 4 p.m. Eastern twenty hundred UTC. Find out more at daily tech news show dot com slash live back tomorrow. Talk to you then. The D T and S family podcasts helping each other understand Simon Club hopes you have enjoyed this program.