 This 10th year of Daily Tech News show is made possible by you the listener. Thank you so much. That includes Tony Glass, Philip Less, and Daniel Dorado. Coming up on DTNS, Apple's new accessibility features include one that lets the phone speak in your voice, and why it's perfectly natural to treat AI as if it's a human, but don't. It's a bad idea. We need to stop that. This is the Daily Tech News for Tuesday, May 16th, 2023 in Los Angeles. I'm Tom Merritt. And from Studio Redwood, I'm Sarah Lane. From New York City, I'm I.S. Aktar. And I'm the show's producer, Amos. Yes, indeed. We have a fun-filled show full of philosophy and accessibility. Let's start with the quick hits. Good news on the Chinese tech front. Baidu beat revenue estimates for its first quarter, reporting revenue of $31.1 billion, which is about $4.5 billion US dollars, after launching its chatbot called Ernie in March. That's not all. PanDaily reported that AI program Mid Journey is launching a beta test for its official Chinese version. The deal will have Mid Journey come to QQ. That's 10 cents in Steam messaging platform and web portal. Seems to be a limited test. Mid Journey originally posted on QQ that China-based users who want to try it out can do so Monday and Friday at 6 p.m. Beijing time, but it'll close the platform for the day once a certain number of people are using it. That explanation has since been removed, so a little unclear there, but it seems like there's a test, so to speak. Maybe the word for Friday at 6 sounds like TNN. A lot of very small things to note today. I guess I would call these super quick hits or insta hits or something. Anyway, the company that owns Drobo filed for Chapter 7 bankruptcy. That's the bad one. That means liquidation, not restructuring. Warren Buffett sold all his TSMC shares, saying tensions in the region are too great to consider it a safe investment. That's a bad sign. Twitter spent $10 million to acquire a job recruiting company called Lasky. If you're waiting for one password to start letting you store pass keys, that begins June 6th. Google is rolling out a fix for battery drain issues with its pixel phones, and Google announced that starting in December it will delete accounts that have been inactive for two years or more. Now, I just want to give a round of applause to Joe on video for actually getting a screen up for every single one of those. That was impressive, man. WhatsApp is now rolling out a new chat lock feature that puts designated chats in a password or biometrically protected folder. When a user gets a notification that they've received one of these messages, they won't see the sender or message content. Not at first anyway, until they do the right thing. Right now, chat lock uses a phone's own password protection, but WhatsApp says it plans to add support for custom passwords over the next few months. Good news and other news for Microsoft users. Here's the good news. Microsoft released its updated phone link app for iOS. We told you that was coming. It supports sending and receiving iMessages from a PC. Also, let you make and receive calls and see phone notifications. There's the other news as well, though. Security researcher Andrew Brant posted on Mastodon that Microsoft began scanning password protected zip files on SharePoint for malware. Now, that wouldn't be so bad if they were just scanning them from the outside, but it does cause a problem since Brant uses that service to back up and share malware with other researchers. So they need to be able to do that. Researcher Kevin Beaumont, however, also noted that Microsoft does this on other cloud services, but it appears to be a new behavior on SharePoint and Microsoft apparently extracts possible passwords from the body of the email or the name of the file itself or uses a list of common passwords. But however it does it, it does crack in to your password protected zip file in order to do the scanning. A zip file password protection is not particularly secure anyway, but some people might take issue with that. Reuters sources say that LG display is set to start supplying 2 million large OLED TV panels to Samsung Electronics starting next year. Initial panels are said to be high end 77 and 83 inch models. You might say, well, hold on a second, Samsung has lots of large TVs, don't they? Yes, the company has its own display unit, but it mostly focuses on smaller OLED panels for mobile devices. Yeah, so this is Samsung giving up on making the big panels and having to pay somebody else to make them instead. There you go. Apple previewed new accessibility features Tuesday. Let's run through them here. The first one, at least a few of them, there's a lot of them. Personal voice is a feature for people who lose their voice. So you want to do this ahead of losing your voice. This is for something where you might know it's coming, but it can synthesize your voice to make the phone sound like you either out loud or maybe on a phone call or something like that. The on device machine learning feature needs you to record 15 minutes of audio with the person reading text prompts. It then integrates with live speech, which is a feature of iOS that lets you type words and then the voice reads them. It'll be available at launch in English for devices running on Apple Silicon. So you could do it on a Mac as long as you've got an M1, M2, or potentially in the future M3 Mac. One of the questions I had initially was like, because I lose my voice like once a year. So for a week, I sound either crazy or I can't really talk at all. I guess that could be helpful, but that's not really what this is for. This would be for some sort of degenerative situation. Yeah, if you have like ALS and you're like, I know my voice is going. Let me record it now while I've still got it. Yeah. Yeah, to preserve it in a way that probably couldn't be replicated easily. Otherwise, unless you have, I don't know, just a treasure trove of, you know, recorded stuff of yourself already. We have to have the exact phrases that the machine learning wants because it does need particular phrases. What I'm looking at is the fact that it's on machine and Apple's got a pretty good history generally of privacy because if you're thinking about voice samples being up in a cloud and then it's going to be analyzed, brought back down, all that stuff, that shouldn't be happening on these devices, which is pretty neat. Assuming all stuff stays secure as for just 15 minutes of audio, that's not a lot of time. If you think about generating out all kinds of different words, things is pretty cool. And like a voice is an incredibly powerful thing to hear. Being able to hear somebody else's voice again who's lost that ability, it means a lot. Yeah. This sort of thing is only going to get better and better to the point where I wonder at some point in the future, will I be able to do my show sounding like this even though I'm 95 or something? Well, and yes, some people get a little bristly when you talk about accessibility features who are for the people who really need them but can be used by everyone. If there were situations where I could train personal voice to use my voice and for whatever reason, I don't need to use my actual voice at the moment, there are some use cases for that too. Next one here is assistive access. So this streamlines your core apps on iPhone and iPad to make them easier to use. This is meant for people with cognitive disabilities. So for example, when my dad was alive after he had a stroke, this would have been great for him. There are modified versions of messages, camera, photos and music, and a combined phone and FaceTime app. They put them together, make it simpler to use, basically strip it down to its essential features. They also have high contrast buttons, large text labels, and it's available in multiple languages, English, French, Italian, German, Spanish, Portuguese, Chinese and Cantonese. So I assume that means Mandarin and Cantonese, Korean, Japanese and Ukrainian. This reminds me a lot of Samsung's Easy Mode. They used to have that on their Galaxy devices. I don't know if they still support that. I know my Note 20 still has it. Basically, all the icons get really large. You got to hold and press buttons or icons so that way they'll be activated, not by mistake, but by intention. I think when I took a look at the streamlined version of the apps, I saw the camera app and I was like, oh, that looks great. So it's got a button to say, take photo. I'm not knocking iOS at all. But there's a lot of features that creep in and to a point where a lot of people that have been using iPhones for a long time, they forget that all this feature creep can be somewhat confusing to somebody who's just looking at, how do I take a picture exactly? How exactly do I switch what camera I want? So some of these app streamlining I think could be useful for more than just this feature. And it's not an Apple's best business interest to take away those features because a lot of those features are meant to get you to do things that Apple would like you to do, but they're doing it anyway for accessibility to just make it easier for a certain class of users to be able to use their stuff. Hey, I mean, half the time I'm like, where is this feature? I know it's in here somewhere. And you could argue that I already have some cognitive disabilities, but for people like you mentioned, Tom, somebody who is recovering after a stroke and learning how to do some things a little bit differently, this makes a lot of sense. The next one is a new detection mode called point and speak. This is in the existing magnifier tool. So it provides assistance for low vision users interacting with physical objects with text labels. One of the examples they gave is a microwave. So it can read out the labels, like say for the buttons on the microwave, as you move your finger over them, it'll be like popcorn button, time cook, stuff like that. That's going to work with iPhones and iPads that have the LiDAR scanner. So you're going to need one of their newer ones. You've got like iPhone 6 or something. It's going to be working in a lot of different languages. When I saw this little demo in action from Apple site, it seems incredibly useful that you just basically put your camera over something. The camera will read what's on the microwave in this example. It'll say popcorn and you'll be able to move your hand to the right so you can hit something else. It seems like a really useful and amazingly powerful use of technology. Every now and then I look at phones and what it was all this stuff for. It's so powerful. I wonder if this would be good on like a vending machine, like to tell you the options. Danny, especially something that you're, maybe I'm so familiar with my microwave that I could, you know, in the dark. I know where the button is, but yeah, if you're out and about. Because I'm thinking like ticket kiosks at subways and stuff usually have a lot of assistive capabilities built into them. Maybe they're not that great because I never use them, but they're there. But I'm trying to think of cases where they wouldn't be built in that this could be this or this is just easier. Yeah, there are several more updates that Apple announced. If you go to apple.com, these tools and the others arrive on their respective platforms later this year. If you're excited about using one of these, I apologize. Sorry, no firm date from Apple yet. Well, Vancouver's Sanctuary Cognitive Systems Corporate just announced a bipedal robot called Phoenix. Phoenix is five foot seven inches, weighs 155 pounds and can lift up to 55 pounds at a time and move it up to three miles per hour. Phoenix's hands have five fingers with 20 degrees of freedom that come pretty close to human dexterity. The hands also have haptics to deliver feedback to the computer as the robot touches and lifts things. Phoenix was tested at a Mark's retail store where it completed 110 retail related tasks, including front and back of store activities like picking up and packing merchandise, cleaning, tagging, labeling, folding and more. It runs on the company's carbon machine learning platform to execute these tasks. Sanctuary Cognitive Systems makes a point that its videos aren't renders, but actual video of the robot in action. So, you know, it's pretty promising. That said, there's no talk of selling Phoenix or who it's actually marketed for as we assume this is more for that retail environment and not for your home. But if you could buy Phoenix, what would you pay? I probably would realize there's no way I could ever pay for something like this. It would be super expensive. I'm sure it would be on like the Sanctuary Cognitive Systems Corporate subscription plan plus premium ultra, whatever the heck it's going to be and it'll cost you $20,000 a year plus overhead. And to take care of maintenance and maybe there's like a free electrical charger when it's walking with me to the supermarket or something. But what I was looking at this because in regular human life, not so much like in factories and things, we have so many single purpose robots like Roombas and mops and lawnmowers. And if you go into the automotive space, you've got a ton of those kinds of robots. These robots, if they're not that tall, not that short, five foot seven, 155. I mean, to sit in a human chair, you don't need to especially reinforce grounds or floors or anything to make sure that this robot doesn't like crash through. So there's a lot of things about it that seem interesting. The dexterity to me, the only thing I keep thinking about and this is this is completely narrow minded, stupid, and silly. I just wanted to fold the laundry. I know that there's machines out there to do that. But this is what I would use it for because I hate folding laundry. I know that makes no sense. But if it's so perfect, I don't know what it's fingers. I'm like, just that way, you don't need to have a special robot washing machine and robot dryer or any of that stuff. Let's just go ahead and give it to Phoenix because Phoenix can do it. Although Phoenix looks a lot like Chappy, I just wanted to get that in. It does look a lot like Chappy. This company is obviously looking for investment. They are making a point of like Sarah said, this is not a render. It's a real product. It's able to do actual things. If they are able to make this work, which it sounds like they can, it's impressive. I'm not sure if it can reach a price point that makes it work for the situations it's needed for because at a scale where you can afford hundreds of thousands of dollars for a machine like this, you probably want a more specialized machine. Maybe something more like a Boston Dynamics. This is a more humanoid general purpose robot. It feels like retail is a great use case for it. But can you bring it down to a point of price that makes sense for a mid-level retail store in a mall to be able to have this thing stocking stuff and going through inventory in the back? Because I feel like it could do that. I feel like it's possible it could do that. I just don't know if it could do it cost effectively or not. Anybody who's been to, I don't know. I know a lot of us aren't going into physical retail stores as much as we used to, although we're coming back. Actually, physical retail is back to booming now. Well, sure. I think a lot of companies that are maybe in that mid-level range are trying to think of, at the end of the night, if anyone's worked at retail, it's like store closes, half the lights go down, and then the employees work for another couple hours making everything look nice again. In that situation, you don't care if a robot is humanoid or not. You're just trying to get things done. That's one use case. If this was more of a, well, Phoenix is walking around among us doing some stuff, doing some light stuff during the day, I could see certain retail chains saying, this is perfect. This is something that'll be fun for the consumer and not weird or otherwise make a mess of store or otherwise. That's a really good point. Because the advantage is that it's general purpose. It can do anything a human can do because it's got the fingers, it's got the mobility. If you build a special-purpose robot, even if it's got multiple purposes, like a Boston Dynamics one, there are going to be things that it can't do. Also, to your point, it'll scare the customers. This one could be made where it's like, oh, it's a fun robot. It becomes a perk while it's also doing its job. I was just thinking about this. If any super premium luxury apartments would have them as their doorman slash concierge because they'd be there all the time. The building itself, I would imagine Las Vegas hotels would have this at no time. The building itself is actually the owner of the robot. It just interacts with you like, hey, we're in packages here, Chappy, and it's like, no, go away. No package for you today. I don't know why the robot has an accent. I'm not even sure what accent it has. If you would like to hear me put on an accent, tell us in our subreddit. You can submit stories and vote on them at dailytechnewshow.reddit.com. Senior lecturer in mathematics at the University of Greenwich, Neil Saunders, wrote an article on the conversation called AI. Evolution is making us treat it like a human, and we need to kick the habit. He notes and agrees with other folks who say we need to stop treating AI agents as conscious and moral with interests, hopes, and desires. But he adds that's actually tough for us. It's easy to say stop doing it. It's hard for humans to do it because we evolved to treat things as if they're human. Cognitive scientist Daniel Dennett coined a phrase called intentional stance, which kind of helps describe the tendency of humans to treat any object as a rational agent in order to fully explain its behavior. Our ability to do that helped us survive as a species. It's a refinement on anthropomorphizing things. Saunders gives the example of chess. So when we compete with a computer playing chess, we treat it as if the computer wants to win, right? That helps us drive our competitive edge, but the computer is just executing a program. It has no awareness that it's even playing chess, much less that it's playing against you. We do this sort of thing with a lot of things. We do it with pets. We even do it with stuffed animals and rocks. But in Dennett's words, we treat things as human even if they have competence without comprehension. So knowing this and this is a little philosophical, I get it. Do you think this can help us deal with the disruption of AI IS? Can it help us deal with the disruption? I don't know. Like just knowing, like, okay, I shouldn't be treating this thing like a human, that's going to get me in trouble. I think it's going to take some time for this. Well, this is like the first generation kind of growing up with this, because we've had lousy chatbots before. We've seen lousy scamming text messages before. There was a great quote in the Substack article that we have linked. It says, what we need is for the public to learn that human sounding speech isn't actually necessarily human anymore, caveat and tour. That to me, I'm like that actually because to us, like, you know, used to be like there was phone etiquette. You could call anybody anytime. It'd be normal. Now you text before you call. Like all these etiquette changes that happen over time. And I think a lot of this is because we do have a tendency to humanize things all the time. We need essentially marketing. I was thinking about this in terms of animals and like farmed animals versus wild animals. You don't go to the supermarket and buy a half a pound of cow. You buy half a pound of beef. You don't get a half a you don't get pig. You get pork or you get bacon. If you call this something else, if you call it rational agent, or you call it human hating machine or something, maybe people will then go, wait a second, I have to be a little bit more cold to this because I still say thank you to Google Home. I say thank you to these people on time because I think it's rude if I don't do it. But I have to learn that this is a device that is probably fueling all kinds of advertisements for me. But it's called Google Home. It's my friend. It's my buddy. But it's called ChatGPT. That's not a particularly friendly human sounding name. And we still have people doing it. I catch it. Well, I mean, I mean, remember it wasn't it was just a couple of months ago where ChatGPT was the hot new thing that everybody was trying out. And it's like, look, ChatGPT said that I was mean. Yeah. And you know, it was said that I was, you know, not not a very nice person and has some issues with me. And you know, that is you could argue, well, okay, it's a human at fault from the beginning, right? ChatGPT didn't do that on their own. But it's still the humans being like, well, that was very human like and not very nice. Although I as your point about, you know, this is getting a phone call, right? Like, let's say it's a robo call, we're all pretty familiar with robo calls, right? Especially the one where it's like, hi, oh, sorry, there's something wrong with my headset. Can you hear me? Dot, dot, dot. And then they go into whatever, you know, trying to get money out of you. You know, that has become this extremely annoying thing that we can now pick up on because it's a scam. But it was, it was good for a while. And probably is still fooling lots of people because it sounds like a human. And we're used to talking to something that sounds like a human no matter what it is. Yeah, there's gonna have to be some kind of, I guess, like, I wouldn't say human evolution, but sort of something like that. It's gonna be a learning experience for a lot of us to figure out, what am I reading? And if there are actual rules that all things need to be disclaimed, like let's say you're reading an article or you're watching a video that was AI generated or images that are AI generated, if there is a watermark of some kind that literally we all agree to, we all agreed this has to be done, then it'd be helpful. If we don't agree to that and there's no standards on this, this could be a real problem. You both are starting to talk about combating fraud. That's not actually what Saunders, Saunders wasn't even getting there. He acknowledged that in the article. He's like, yeah, there's things about fraud. He's like, but even before that, to regulate this properly, to treat it properly, to use it properly, to think about it properly, we have to remember that it is just a tool, that it has no awareness. Before we even get to like, yeah, of course, bad actors are going to make use of it. And that's actually harder to deal with because we won't know it's chat GPT behind it. They'll try to hide that. But even when we know, oh, this is Project Bard, this is Bing, this is chat GPT. We have to remember that it isn't thinking. It isn't insulting you when it says you're mean. It's just predicting words and spewing them out. And what Saunders pointed out is, if that feels hard when you're trying to do it, that's right. You're not, there's nothing wrong with you. That's human evolution. We have evolved to think of things as human that seem human. Even pets, which, you know, yeah, they have their own desires and stuff, but they're different. We anthropomorphize them. We treat them human. But there are times when an animal, even a pet, a dog or a cat, has behaviors that are not human. And it's important as a pet owner to not treat them human in those cases, right? There's great examples of that. And we needed, maybe we needed to treat text generators as pets. You know, I, as you saying that you sometimes say thank you to Google Assistant, I do the same to Amazon Assistant. Yeah, yeah. That's an example of this. Yeah. And you know, I know it doesn't matter. I mean, maybe I don't know what thank you gets recorded or, you know, it gets chalked up to like, you know, maybe I get like, cool person points somewhere in the, you know, in the ether somewhere. No, but that's not, I don't have to say anything nice to this thing that is not a person. But it sounds so much like a person. And that's how a lot of these products have been marketed to people. Like, it's not a person, but it's going to be real nice and conversational. And it's going to sound very human-esque. So you're not freaked out. And you, you know, can more or less speak, I mean, you still have to use, you know, certain keywords and whatever, but you can treat it as if it is a human, very much not though. It sounds like you're talking about like a public relations person, like confidence without comprehension. Like this is what people do. That's the weird thing. That's why I think so hard. I've met a lot of people that don't seem to like understand what's going on. People have comprehension. You're just saying, maybe they're comprehending a different thing than you want them to comprehend. Perhaps. They're humans. The difference is the AI doesn't have comprehension. It doesn't even know it's doing it wrong. It doesn't even know it's fooling you. It doesn't know that's the key. It doesn't know at all. There's no, there's no, no consciousness. Yeah. Yeah. Tough one. Well, I know what we do know. And that is what NBC universal and the NFL are announcing. I'll tell you now, they signed a one year deal to have peacock exclusively stream one NFL playoff game next season. Just one. The game will be the wild card game on January 13th, 2024 and won't be available on national TV or cable. Local markets will have access to the game through local TV channels, but otherwise got to go through peacock. Yeah, presumably bars will get it through direct TV. That's something a lot of people don't realize is direct TV commercial service for bars carries these games that are on Amazon Prime and Apple TV and stuff. That's why the bar always has the game. Yeah. Yeah. The bar still has the game. I kind of get it at home. So this, this reminds me of when they first put some playoff games for Major League Baseball on the MLB channel. And there were people who didn't get the MLB channel on their cable system. Like they, they couldn't get it. That wasn't an option. And people got really upset. Only difference being with peacock, as long as you have the bandwidth. So if you don't have the bandwidth, you can't get it. But if you have the bandwidth, then you just, it's whether you can afford the price. If you want to pay the five to 10 bucks to get peacock, then you can get it. It's nice that you can get it from anywhere that has peacock. That's great. But like I'm just thinking, yeah, the, the annoyance of trying to figure out where is what game on at what given time this has become such a pain in the neck. Like I have to like, I have like special routines that I do to find out where the Yankees game is going to be on, you know, NBA basketball is like on four different networks. There's like, it's getting so, so split up. It's going to be a real pain in the neck to find this stuff. And I still don't think search is very good for this personally. Not yet. Not yet. That's, that's the, the, the hope for the future would be the platform would solve it. Like your Roku or your Google TV would be able to tell you, hey, the game's on, click here. And it wouldn't matter if you'd hit the paywall, if you don't subscribe and then you decide whether you want to get it or not. And I won't thank that search. I won't do it. Yeah. Don't do it. Don't do it. It doesn't know. All right. Let's check out the mail bag. We got one from someone who would like to remain anonymous, who's hoping that AI could cut down on friends and family getting hit with spam and phishing attacks on social media and then passing them along to everybody else, which makes the problem worse use, used Facebook as an example. And on says, it seems like an obvious formula. The user logs in from a new location, user spams all their friends with the same posts, continuing to link 30 people, message the user saying, you've been hacked or hey, I think you've been hacked or my favorite. That link didn't work. Who died? Anon says, I wasted 30 minutes trying to delete all copies of the spam recently before anybody else spread it on a PC. I had to open every messenger chat, click on the hamburger menu, remove the post, click, remove Facebook scrolls to the top of the list. I scroll down, find where I left off, click the message 150 times. It seems like the company who's reading my messages to advertise to me would be able to recognize some pattern to this madness or even provide a way to mass delete the message that was mass sent. Yeah, I feel you anonymous. However, that always sounds right when you're dealing with the actual problem like you are. These companies, even Facebook run into huge problems if they try to implement a solution to prevent this when it works in the wrong way because nothing works perfectly. And suddenly somebody was trying to send a message to 100 people for their wedding and it got blocked by Facebook. And then there's a whole privacy kerfuffle and Facebook backpedals. That's part of the reason. The other reason is that this is an arms race. And so every time Facebook puts in a method to stop this kind of stuff, the bad actors figure out a way around it. And it isn't until those start happening that the engineers can see what they're doing and figure out a way to counter it. Indeed. Well, well, I as I hope that you haven't had any mass spam emails or Facebook messages lately, but you know, if you have, you wouldn't be alone. Let folks know where they can keep up with. What are you spamming people? When did you stop spamming people? I as act are and where can people find out where your good work is happening these days? If you would like to be voluntarily spammed as anyone to subscribe to go to this old nerd.com. It's a show where I show people how to have the most tech forward life and home as possible. Last episode was about mounting a sound bar on a really old TV and how that went. Tell you what, there were some hiccups, but I went through the pain. So you don't have to and you'll know how to do this stuff a lot quicker. So check it out at this old nerd.com. Go to the YouTube page and subscribe. All that fun stuff. Indeed. And good to have you on the show today. Thank you so much. Always a good day when I as is with us. Also want to extend a special thanks to Christopher Nelson, one of our top lifetime supporters. Christopher, we want to thank you right here right now for all the years of support. Yay. Thank you, Christopher. In fact, we have 1,400 new patrons that are free. Since we announced the new free tier last week, we have had 1,400 people sign up to get Patreon perks for free. You just scroll down past the paid options at Patreon.com slash DTNS. You get monthly updates, Rogers column and Friday's good day internet. So welcome. Well, yes, hopefully you can hear me like you're listening to the public feed. Anyway, and if you want to get more perks, of course, they are available whenever the money loosens up for you. But thank you for becoming a patron, even at the free level, Patreon.com slash DTNS. Patrons stick around for the extended show, Good Day Internet, CNET's writers have decided to form a union. And since I as and I worked there, we have thoughts. We're going to tell you about it. You can also catch DTNS live at 4 p.m. Eastern Monday through Friday. That's 200 UTC. And you can find out more at dailytechnewshow.com slash live. We're back tomorrow with Alison Sheridan joining us. Talk to you then.