 Daily Tech News show is made possible by you listening right now. Thanks to all of you, including Jeffrey Zilks, Kriya Ardham, Tony Glass, and Bill Baggins. On this episode of DTNS Intel's Future is Becoming Arm. Apple has a free sports app and a defense against our quantum future. What don't they have? This is the Daily Tech News for Wednesday, February 21st, 2024. In Los Angeles, I'm Tom Merritt. And from Studio Animal House, I'm Sarah Lane. And beneath, finally, Sunny Skies on the show's producer, Roger Chang. Oh, is that why you have such a sunny mood? Because the skies are sunny? Yes, I'm getting my vitamin D indirectly through the window. Through the window? Yeah, one of those windows that just lets in the vitamin D rays. Yeah, hopefully. Sunscreen optional. Well, I'm excited because we have a great live audience. Dan Krafton, who's very active on our Patreon, threatened to make the live show on Patreon this morning and then showed up in the live show. Yay, Dan. R.W. Nash is going to show up live. Who knows what can happen? Let's get into the quick kids. Overnight, several users reported, let's call it unpredictable behavior from chat GPT. One example that the Verge pulled from a Reddit user was a request for a biography of the Jackson family of musicians that contained the line, Schwitendli, the sparkle of Tormor on the crest that has as much to do with the golfer of the moon paths as it shifts from follow. Now, if you said, I didn't didn't understand what that had to do with the Jackson family, well, you're not alone. A lot of other people said something seems wrong with chat GPT. At 10.40 p.m. Eastern time on Tuesday, OpenAI acknowledged it was investigating reports of unexpected responses from chat GPT. A few minutes later, announced it was working on effects. The problem seems to have gone away, but OpenAI says it's still monitoring the situation. Yeah. This is the thing that happens when you've got a tool that just tries to predict the next word. Sometimes it doesn't. I mean, sometimes I'm tired and I sound weird too. Yeah, that's true. Chat GPT. AI Vertex, Google's Ann Nemo. Oh, I'm sorry. I just had a little glitch. Google has released Gemma 2b and 7b. These are two open source AI models that let developers use closed AI model Gemini more freely. The lightweight Gemma models are supposed to be better for smaller tasks like making your own chatbot or doing summarizations of your own data. Google claims its Gemma models, quote, surpass significantly larger models on key benchmarks. Sort of an empty phrase that doesn't mean anything except we think ours are better. They are also capable of running directly on a developer laptop or desktop computer. Now, that's a quote that does mean something. These are local models. You can run them yourself. You don't need to put them in the cloud. You can try them out on the Kaggle laptop. You can get them at Huggingface, Nvidia's Nemo, and Google's Vertex AI. At Nintendo Direct earlier today, the company announced that grounded and sentiment are Xbox exclusive games that will be exclusive no more when they arrive for the Nintendo Switch. Sentiment launches Thursday, February 22nd, grounded on April 16th. So that's two of the four that Microsoft said would be coming. The Verge's sources think Hi-Fi Rush and Sea of Thieves are probably the other two. These could still launch on the Switch later, and any of the four could also come to the PS5. I appreciate the sentiment, but it's pentament. Oh, pentament. But just, but yeah, there we go. Now we know. Now we know, at least two of them. They still might preview that. Well, someone should make a game called Sentiment, okay? Yeah, please. For goodness sakes, Xbox. ZTE's Libero Flip for the Japanese market is going for $420 or $265 if you pre-order. That might help indications that the Flip market is back, if it ever went away, as the Libero Flip is priced well below the $1,000 Samsung Galaxy Z Flip. And darn near as low as a clearance sale refurbished Flip you could still find around the net if you wanted to look forward on a used market. Libero Flip has a Qualcomm Snapdragon 7 series chip inside and a 6.9-inch 2790 x 1188-foldable OLED panel, slightly larger than the Samsung Flip. It includes a 50-megapixel main camera, 2-megapixel depth camera, 16-megapixel selfie in the front, runs Android 13 with 6GB of RAM, 128GB storage, Wi-Fi 6 and Bluetooth 5.2, charges up to 33W over. It's a USB-C port, but I think most importantly, its front cover screen is circular, so it looks like a little compact. The Wall Street Journal sources say that Reddit may set aside a chunk of shares in its upcoming IPO for 75,000 of its most active users, sort of a thank you for being a friend. But as mastered on user, Carnage4Life points out, I was initially impressed because I thought they were gifting them shares. However, an opportunity to buy Reddit shares at a $5 billion valuation hoping that it goes higher is more of a gamble than a perk. That's just true of every IPO though, right? Every IPO, they don't give away the shares. That's not how IPOs work, but they limit who gets to buy them at the initial price, the idea being that IPOs generally go up. If you're a diehard Reddit user and you're part of the 75,000, this might be a cool perk. If you wanted to buy the IPO, it's a cool perk. If you're like, no, I never wanted to buy stock in Reddit, then no, it's probably not that cool perk. I get that. Well, back in 2021, Intel CEO Pat Gelsinger promised Intel's customers and investors five nodes in four years, but the plan to go all in on fabs and become a top to bottom foundry service for the whole world to use. Now that roadmap is seen some light with Intel's first EUV based node, Intel 4, available in the market today. And it's high volume counterpart Intel 3 also ready. Intel is also ready and it's gate all around or GAA FET, Ribbon FET for 2024 and next year 2025. Today, Intel's Foundry Group has formally become Foundry Group, used to be called Intel Foundry Services, and held its first conference, Direct Connect. Direct Connect is Intel's first chance to talk at length about the five nodes goal and where the company's going forward. Direct Connect is also Intel's chance to update folks on what comes after those first five nodes. Intel Foundry wants to expand capacity, customers, tooling. That's obvious. The company wants to make money. And now the group is looking toward a slate of even more advanced nodes, but also packaging technologies that would be necessary to back all that up. Now, Roger, you had mentioned in our pre-show meeting that Intel didn't exactly deliver on the promise from three years ago timing wise, but they're getting there. What stood out to you about how far Intel has come? I think two things stood out. One is that they really do have a plan to become a top to bottom Foundry. That means they just don't, they're not just going to be a fabricator or Foundry to make other people's designs. They will, they'll work with the client and develop their own chips designs, as well as testing them and then producing them. And it doesn't sound like much on the face of it, but you have to understand that Intel's Foundry's have historically just been used internally. That means they now need to have the process and the tooling to have a variety of potential clientele come through the door and say, we want this chip designed and made on this process node. Can you do it? So that is an entirely different set of requirements because now it's no like, hey, can I borrow this from this Foundry because we need it for this chip because it's in the same company. Now you have to be a little more well oiled as it were to be kind of a clientele facing operation. The other thing is that the EUV or the extreme UV lithography machine that they got is the new numerical aperture, which means it's a larger aperture. Effectively what it means is they can produce the smaller node process in a shorter amount of time. And it's a machine that TSMC itself has passed on because they're fine with the current process and they can effectively move to the next generation without too many tweaks. But for Intel, this could possibly be their way to leap frog TSMC into the next generation of chips. Still too early to say if that actually will happen, but at least based on their flow chart that they gave, it looks like they're aiming for that. Yeah, I think Intel's doing better than a lot of people expected, even though they're running behind their own timeline. Gelsinger is committed. I think that's the important point here. He's not doing this halfway. He's not saying, hey, we'd like to make chips for other people, but really we want to make all our money on our own chips. So we'll be grudgingly make chips for other people, which is I think what a lot of people thought. They thought the Intel Foundry would sort of compete with TSMC. And that's that. And what you're saying, Roger, is they're competing with ARM too. They're like, hey, we can design a chip for you. You want to come to us with the design? Great, we'll build it for you. We'll build it faster than TSMC is what they'll tell you, but they'll also design. They'll also do packaging. That is smart. Gelsinger is not doing this halfway. He's like, if we're going to be a Foundry, let's be a Foundry from top to bottom. And it's interesting because I think a very few companies have done that. I think IBM has done that in the past with their PowerPC and they developed Fab, the cell processor used in the PS3. But it is a very capital intensive goal. It's not cheap. The fact that they are sticking to their guns and Gelsinger is effectively emphatically saying this is where we're going to be 2025 is kind of the benchmark year where you're going to see if all these promises that they've put out come to fruition. And it's something they need to be patient with if they want to supplement and eventually surpass the in-house Intel chips. X86 isn't going to last forever. That's what Gelsinger is identifying here and he needs to get the Foundry business in place before that happens and hopefully come up with new designs that keep Intel chips as a going concern as well. I don't think they're going to abandon making their own chips. But they know that this is a much more competitive market than they used to be when the old wind-tell dominance just shoved everything else out the door. Bloomberg reporting Microsoft has contracted Intel to make its chips. Those are the kinds of big contracts that Intel's Foundry is going to need to have to make it successful. So if that ends up being as Bloomberg reported and Bloomberg's pretty good with this stuff, it's a good sign that Gelsinger is on the right track. A couple of Apple stories to talk about. First, sports. On Wednesday, Apple announced... Now it was sports. Now it was sports. A free standalone app called Apple Sports. Very clever. Thank you, Apple. Sharing scores and stats to your iOS device. This is not unlike a Yahoo! sports or some other sports app that you might have already been using for some time. It's also designed to drive viewers to the Apple TV app to watch those live sports, which does include now major league soccer matches. Apple has a 10-year deal through Apple TV's MLS Season Pass subscription add-on. Apple Sports includes a watch on Apple TV button for each game. Now it's not always going to apply, but if it does, then you can open up your Apple TV app to the service that's live streaming the event and where you can subscribe if you haven't subscribed already. It's a portal to sports. Apple Sports is in the U.S., Canada, and the U.K. app stores for now with support for English and French and Spanish where it is available. It's a sport. It's a sport. Because it's a sports portal. I love this and hate it. What do you hate it? Well, should I start? Yeah, would you like me to start with the love or start with the hate? Bad news first, always. Bad news first. Why isn't this cross-platform? Well, I know why because it's Apple, but I want this cross-platform. I don't want to have to only have this on my iOS device. I want it on the web. I want it on Android. I'll probably be able to get it on the web at some point, maybe, or at least on an app on my Mac, but I doubt I'll ever be able to get it on Android. I also don't like that it's just Apple TV, although I want to see how this works in practice because Apple TV has add-on channels, so maybe it'll also have Paramount Plus Sports, maybe it'll even have Peacock or ESPN Plus or eventually ESPN Sports, because that would be, I think, a better deal for Apple is to say, watch all your sports through Apple TV's app, whether it's Apple-provided sports or not, and use the Apple Sports app to launch them. I think that would be super compelling. So those are the things I don't like is that it seems more limited and very Apple-y. The thing I love about it, though, is I've wanted a simple sports score app forever, ESPN complicates things. It tries to give me things I don't want, or at least things I don't want when I'm just looking for the scores. It tries to guess what I want, and it doesn't guess very well. Yahoo Sports is old, and it's fine, but it is, again, it's trying to show me things that it wants to show me. Apple is like, no, we know you just want to follow your teams or your leagues and your sports here. We made it simple and made it easy, and that's the thing I love about it. Yeah, I installed the app this morning, works as advertised. I was like, here are my favorite teams. You tell me about those favorite teams. Well, in theory, maybe not in execution at this point, we're still crawling toward a world where I'd be like, I would like to watch this game. I've heard it starts in 15 minutes. No idea what platform it's on, what channel. Don't know. Maybe Apple Sports will just help me get there. Yeah, let me just tap the button. Tap the button, and then I can watch it. Take me there. Maybe I have to pay. I don't know. I mean, just help me do that instead of googling, like, on what channel are the wires playing? Right, because Apple has all of the apps that play the sports, right? They have the apps for all the streaming services, even apps for all the cable services out there. If you could just tap the button that says, oh, this is playing here, we geolocated you. So we know this is the service that has it in your market. And if you subscribe, we'll just launch you into that app. And if you don't, then we'll give you the option of like, oh, if you want to watch it, you could subscribe to these different services. That would be great. I would love it if they did that. I mean, there are Apple apps that actually do this quite well already, not for sports, but for example, I talk about my love for Plex. If something's on the media server that I have access to, great. But if I don't have access, it goes, okay, Apple TV Plus has this, and here's what you would pay. Well, I have Apple TV Plus, so that doesn't really matter. But you might not tell you where to get it. On Hulu or on Netflix or Apple TV does the same thing. Like the actual device, the Apple TV does the same thing. And it even has the sports. The teams that I follow on the Apple TV app were pre-populated in the Apple Sports app when I launched it. So it does a pretty good job of telling me like, oh, the Illinois basketball team is playing in Apple TV. So it's halfway there. Just put it in the sports app, right? Which I don't know, maybe they will. I haven't really had it long enough to be able to test if that's the case or not. But they're not showing me a way to watch the Illinois game. The album art for the app is a soccer field. I think Apple is going hard in on MLS right now, seeing where it sticks and hopefully opening up after that. Yeah. I think that is probably true. Well, the other thing we didn't talk about in regard to sports is AI deciding for you what sports you want to watch. If you want to stay up to date in the fast-moving world of artificial intelligence, you cannot miss AI named this show. Each week, Tristan Jutra and Teja Kastodi cut through the hype. They cut through the doom saying and just talk about the things that are important to know about artificial intelligence. And generative models and deep learning and machine learning and LLMs and all of that. Catch it at ainamedtheshow.com. In the next OS software updates for phones, tablets, watches, and Macs. Another Apple story. Yes, we are doing it. Apple will add a new cryptographic protocol to its messaging service iMessage called PQ3. The new protocol is designed to be resistant to future attacks from quantum computers. Now, Tom, people might say, well, I mean, is encryption just resistant to all sorts of attacks? Why would quantum computers be particularly problematic? Yeah. And we don't need to worry about quantum computers this minute. They're making slow progress, but they are making progress. And one of the things that quantum computers are good at is breaking encryption. Right now, they're not powerful enough to practically break current encryption, but they're on their way. If you want to break a key, write an encryption key, a classical computer has to run through all the possible combinations. Now classical computers can break through combinations very fast. But if you make that key long enough, if you make the factors of the encryption long enough, the time it would take to crack a key can become impractically long. So it could be like, yeah, you can run through all the possibilities in 100,000 years. Yeah, exactly. And then that's considered secure. Like, oh, 100,000 years from now, if someone finally does crack your key, you probably won't care. Quantum computers, on the other hand, work a different way than classical computers. One of the things they can do is make multiple attempts at once. But there's other ways that they can speed up the process and look at current encryption and break it fast. They could break it in a month. They could break it in a week. Eventually they might be able to break it in minutes. Here's why you need to worry about that now. Bad actors are gathering your encrypted data and storage is cheap. So they can keep your encrypted data around until the time that quantum computers become practical to use to break encryption. Now for most of this, maybe not that big of a deal, but people in sensitive situations, people who are very concerned about privacy are going to be vulnerable to someone saying, well, I got your encrypted data and eventually I'll be able to crack it. So to get ahead of the game, Apple is updating its encryption so that even if they steal your encrypted messaging data, it will be resistant to quantum computing. It will have strong enough encryption that even a quantum computer shouldn't be able to break it. You might say, okay, well, Apple is really, really worried about quantum computers, but not the only company doing this, right? No. Again, these are things that the very security-minded, people in governments, people in militaries, people in journalism, people in sensitive situations worry about. But those are the kinds of people that use apps like Signal. Signal recently added quantum resistance to its encryption protocol too. Now Apple created its own way of ranking their encryption and it turns out theirs is the best. What a surprise, Sarah. But there's actually a backing to that. Apple ranks its encryption at level one. It ranks Signal's quantum resistance at level two and Apple's PQ3 at level three. So the current encryption is level one. That's one that's probably not going to be resistant. Signal's is at level two. It's pretty quantum resistant, but Apple's is at level three, the most quantum resistant. And if you're into this, Apple has a really well done post that goes into how they're using, you know, difficult cues and elliptic curves and everything to make PQ3. But the short version is level two adds quantum resistance to the creation of the key. But if that key were to be compromised, then they could get your messages. What Apple's doing is like going above and beyond and changing up the keys so that even if they crack the key creation process, the ongoing messages wouldn't then be automatically decrypted. So that keeps the key from getting cracked. It protects your messages, even if somehow the key creation got cracked. Okay. So that was going to be my question is how is Apple testing this to feel confident in saying, we're the best messages or I message is where you send those messages? Yeah, normally you'd you'd run it through some crackers, right? If it was classical computing. And I don't mean Ritz. I mean, like you would you would try to brute force it and say like, ah, see, it can't be brute forced. We don't have quantum computers that can do this yet. So there isn't a computer you can try to break it with. So we really don't know how good this encryption is at resisting quantum computers. But Apple says its protocol was assessed by an unnamed third party security company feel better if they named it, but okay. And this one I feel pretty good about two groups of academics who have written papers about the theory and the math behind it saying, you know, to the best of our knowledge, this is really secure. Obviously, there's no way of knowing until you actually get quantum computers that contested. But given everything we know, this this is your best chance at quantum resistance. So that makes me feel pretty good about it too. I think there's probably a couple of people out there being like, okay, so a bad actor who's trying to break into, you know, any platform system. Are they all going to have access to quantum computers? I mean, how easy will these be to come by in the future? Well, the idea is that you don't know. So you might and what Tom was saying since storage ahead of it, it's a hedge. It's it's it's it's hedging their bed. Yeah. It's the security game of like, it's not how it's not how likely the attack is. It's how how much how close to zero do you want to get the possibility of your data being uncovered? And remember, like most crime is a crime of opportunity. So if you can save if you if you have an opportunity, you're just going to save it for that day when it does come around. What's interesting is on the testing, I mean, most testing with with quantum computing, unless you actually have a quantum machine, but even if you did, you couldn't really do this, you test it against an idealized quantum computer model, and you would go through a bunch of a theorized hypothetical, you know, scenarios to see whether or not it stands up. My question, though, is this really a game changer that Apple's presenting? Or is this attempt to kind of garner mind share ahead of its competitors, saying like, hey, we're taking security to the next level. So when you think security, it's going to have an Apple embossed on front of it. I think I'd answer yes. The answer to both of those is yes. This is a real attempt at real security from from everything I can tell. And I will bow to people who know more about security than I do if they find something weak about it. But it seems like it's an honest defense against quantum resistance. It's not like Apple's the only one doing it signals doing it. There are other security minded organizations doing it. So it's not it's not fluff. But also they want to use it to get mind share like they don't need to do this. This is not a practical thing for the majority of their audience. But I think they want, like you said, people to think that Apple logo means security. I don't have a problem with that when they back it up with actual security, which appears to be what they're doing here. And just for clarification, you don't actually need a quantum computer to use any of this. This is all done on classical computers. It's it's rather that the encryption is done on classical computers. It would be the attacker that would have to use the quantum computer. Good point. Exactly. Awesome. I can't tell it comes to Android. And also you don't need to do anything. This is just coming in an update, right? You're just going to get it. And it's going to happen. You don't need to turn it on. I mean, I will be like for $5 a month. Oh, we'll protect your privacy just for an extra five minutes. Well, you know, I say that and let's just hope not. All right. Let's check out the mailbag. So we got a bunch of good responses to our conversation with Lamar Wilson, who was our guest yesterday on the show, about him liking the form factor of the iPad mini. Rusty, who lives in Huntsville, Alabama. Hi, Rusty says, I've been a happy Android user since my first gen iPhone would not allow me to download podcasts over my required AT&T unlimited data plan. But then I decided to get my pilot's license. I decided for flight was the best electronic flight bag, EFB for me. For flight is only available on Apple products. So I succumbed back to Apple for a single iPad. The iPad mini is the perfect form factor for flying small aircraft. You can clamp it onto a mount on the yoke, just small enough to read ages around it while being large enough to read misapproach procedures. I've seen other pilots fly with larger iPads, but they're usually placed in inconvenient places. Yeah, I would say that Lamar's discussion of why he went back to the iPad mini has engendered the most response of any discussion we've had in months. Brian wrote in and said mini fan here, same as other comments as to the size. I love the size of the Kindle, which the mini is similar fits into pockets. So easy to take with I use it a lot for reading and PDFs. Also card games work very well on that size. I also have an iPhone SE because I like my phones smaller. So the mini works well as the bigger screen. And David wrote in and said I became an iPad mini fan when recording my son's high school basketball games from the bleachers. I didn't enjoy viewing the games by splinting at my iPhone pro max screen. And I had a hard time following the ball in action, but sizing up to the mini was just right. Fit on my tripod wasn't overly intrusive or obnoxious to those sitting behind or around me. And the screen was big enough that I could get immersed in the action, not have to constantly glance over to see where the ball went. I also found it to be an excellent fit for seatback movie watching on planes. In its folding case, it'll slide into the brochure slot be at eye level. I don't have to look down like I do with a laptop and leaving the tray top space available for other things. That's one of the more interesting uses of this is the basketball game. Thank you, David. Thank you, everybody. These are just a few of the many, many responses we got. And every one of them was like, yep, me too. Like the many. Here's why. So I thought that was really fun. Yeah, absolutely. Yeah, it's always good to strike a chord and hear what everyone else is thinking. Indeed. Patrons, stick around for the extended show. The fun doesn't end here. Good day internet. Yesterday on Good Day Internet, we talked about flying to Tokyo for lunch. Could you go to Tokyo for lunch? Well, one DTNs listener says don't laugh too quick. The answer is yes. Plus our theories on why chat GBT lost its mind. We'll talk about all that and more. And that person doesn't actually live in Tokyo, by the way. It's a great story. You got to hang around for this. But just a reminder, you can catch our show live Monday through Friday at 4 p.m. Eastern, 2100 UTC. And you can find out more at dailytechnewshow.com slash live. We're back again doing it all tomorrow. Talking about Sora. That's the text to video generating AI for moving AI with Charlotte Henry joining us. Don't do that. That's why it doesn't roll in.