 This 10th year of Daily Tech News show is made possible by you, the listener. Thanks to every single one of you, including Philip Les, Daniel Dorado, Howard Jermisch and Stephen Radke. On this episode of DTNS, Facebook's going to start charging. Can that be true? A new way to protect your identity online and why your attitude towards AI changes how it responds to you. This is the Daily Tech News for Tuesday, October 3rd, 2023 in Los Angeles on top merit. And from Studio Slytherin, I'm Sarah Lane. And I'm the show's producer, Roger Chang. Wait, are you House Slytherin, Sarah? Today I am. What do I do about it, Tom? I'm just, well, I would not mess with the Slytherin. First of all, second of all, I just, I didn't know if that was like you took the house test thing and that was. Yeah, I just feel Slytherin today. It's just a vibe. Yeah, it's a vibe. Yeah, it's kind of talk of Tuesday Slytherin kind of stuff. I think I might know, I might know what causes that vibe. Or vibe, either one. The Sam Bankman Freed trial. That's probably what it is, yeah. SBF, given those Slytherin vibes. Yeah, the SBF trial over fraud accusations over FTX has started. Let's see what else you need to know in the quick hits. The U.S. Federal Communications Commission has begun cracking down on violations of orbital rules in an attempt to combat space junk. Yeah, that stuff up in space that is considered junk. DISH Network gets the dubious honor of receiving the first fine of $150,000 for failing to move an old satellite far enough away from other satellites. DISH admitted liability over its Echo Star 7 satellite and agreed to a compliant plan with the FCC. I just realized this is the first space parking ticket. You can't park your dead satellite here. It's like putting your dead satellite up on blocks. You can't do that. Yeah, you got a boot on your car. Okay. FCC is like the HOA. It's like, you can't do that. That's a fine. There you go. One ongoing theme in tech news these days is the diversification of manufacturing away from China. We've covered a lot of companies moving plants to places like India, Vietnam, very frequently. But here are a couple of less commonly mentioned locations. HMD Global, which makes phones under the Nokia brand, launched its first smartphone manufactured in Hungary. It's the 5G Nokia XR21. Now, metals and components were still shipped in from China, but making the phone, assembling it all and finishing it in Europe helped avoid problems with security laws. In fact, that's one of the reasons that this is the XR21 5G. The earlier XR21 is still made in China. Another location that has to ship in all its materials is Spaceforge's Forge Star 1 set for launch to orbit late this year or early 2024. The satellite contains an automated chemistry lab for the development of semiconductor alloys. Now, this version will only test the process, but they do have plans for future versions, if the tests go well, that would have the system able to return ships it makes to Earth for use. Apple now requires new apps to show proof of a Chinese government license before the app becomes available on the Chinese version of Apple's app store. Developers now have to submit an Internet content provided or ICP filing when they publish new apps. This change brings the company in line with its local rivals, which adopted the policy some years ago to meet Chinese state regulations. OnePlus has its own foldable phone coming out. Its parent company, Oppo, also has a foldable phone coming out. They're the same phone. At least that's what OnePlus told the Verge. Oppo Chief Product Officer Peter Lau, who co-founded OnePlus, oversaw the development of the phone which will be marketed under different brands in different markets. GSM Arena recently reported that they think the names will be the Oppo Find N3, which would be sold in China, and the OnePlus opened, which would be sold internationally, though that's just speculation. None of that is yet confirmed. It sounded silly to me, but then I thought, you know, think of all that, you know, apparel lines, where it's all the same thing. Yeah, yeah, or the house brands of something where you're like, oh, that's the same evaporated milk. It's just in a different can. Google will begin a new email policy starting in 2024 to Combat Spam. Any sender who delivers more than 5,000 messages per day to Gmail users must offer one click unsubscribe options and also configure their systems to authenticate that they are not spoofing their domain name. Fire letters will be blocked if they're found anyway. Google said Yahoo made the same changes for its email services well, so, you know, it's just kind of keeping up with the Joneses. Yeah, Yahoo's gonna take the filters from Google and they're gonna cooperate on this. So it's good for everybody. Let's spam for all of them. Now, for years, Sarah, we were able to say that if somebody was sending you spam saying Facebook is soon gonna charge playing on your fears that it was probably a scam email and you could safely ignore it, but apparently not anymore. What's going on? Yeah, boy, have I talked some folks off the ledge saying this is not true. This is not happening. And it still isn't in that sense, but here is what's happening. The Wall Street Journal sources say that last month, Metta presented a plan to European regulators to comply with European Union rules that require platforms to get user consent before targeting them with ads. If a user doesn't give consent, the platforms still have to offer the service. They can't just make you pay. Metta's plan would let users who consent to personalize ads continue to use the service for free. Those who don't consent would have to pay a fee. How much are we talking about though, Tom? Yeah, so the rate would be around 10 euros a month on the desktop on mobile because the app stores both Google and Apple take a cut. They would charge 13 euros. And then if you have a Facebook plan, you could add an additional account for six euros per account. At least that's what the sources are saying. None of this has been officially announced. Metta apparently called the plan SNA for subscription with no ads. And this all follows a July court decision that said social media platforms who were telling, were not getting consent for personalized ads could charge a reasonable fee to those people who wanted to continue to use the service but didn't want the personalized ads. Last month, Metta reported that European revenue was around $6 per user per month across all its apps and services. So that's below what they're charging. 10 euros a month is somewhere around $8 or $9. But the $6 per user per month includes all European countries in Facebook's market, not just EU countries. So it includes Russia, for example, where Facebook makes less money on ads. So we don't know exactly what the average is in Europe, but it might be closer to 10 euros a month. The questions here are whether this is legal, right? I know the court said it's a reasonable fee, but that wasn't a legal ruling. That was just kind of a commentary that they said. We still haven't tested that in court. And whether it's fair to require someone to pay in order not to be targeted when Metta could technically offer a free ad-supported service that just doesn't target you, doesn't personalize the ads. Oh, imagine a world where Metta did something as nice as that. So before the show, you know, I was, you know, my knee-jerk response was like, well, but lots of companies offer either an ad-supported free option or certainly a greatly reduced option or you pay to have the more sort of premium offering. Yeah, it goes Spotify does it. YouTube does it. Lots of new sites in Europe do it. Like it's fairly common. Exactly. And you rightly pointed out, Tom. Yeah, but in those cases or in many cases, maybe not everyone, but in those cases that content has to be licensed by the company. So the company is kind of eating the costs where most of anything that you see on Metta is just social network generated. It's just, you know, people who are certainly not getting paid to say, hey, I had a really great day today with my family. Here's some photos kind of thing. And I, you know, when I think of it that way, I go, hmm, interesting. I don't think that unless the hand was forced to the point where Metta simply had to legally offer a free option to users that was not ad-supported but also did not make them pay. And I don't really see that happening even though we're laying out something that floated by the EU, you know, for specific EU laws in this case. I don't, I don't know how that would ever work. I also, I just, Metta has such a, has such an optics problem right now for a lot of folks that were on the platform for a really long time that now have other options. They might be, you know, aging out of some of the content that's still there to say, well, hey, if you don't like ads and you don't want to be targeted $10 a month or the equivalent, you know, in euros, that's, that's insane to me. I would just stop using it. But that's also because I'm not like. Why is 10 euros a month insane? I know it's insane compared to zero euros a month but you pay that exact amount for lots of other services. 10 euros seems to be like, that's what you pay for Spotify. That's what you pay for YouTube premium. That's what you pay for all this stuff. Like, I guess what I'm arguing is if it's too much, then people will just not use Facebook. Like, if you really want to use Facebook that bad, right? To where people are saying, well, it's unfair to charge them cause they have to use it and they don't want to get, their only choice is to either get personalized ads or pay, there's also the choice of not using Facebook. Now the argument runs there is that, yes, but there aren't any other competitors to Facebook that are equivalent. At which point, I feel like the assumption here is like, well, obviously everyone has to use Facebook and it's a monopoly because there aren't anything, there isn't anything else like Facebook, but I question that. But is that, do we have to use Facebook? Well, yeah, that's kind of where I was going with this too is if it comes down to, well, okay, I'm not going to pay and so I can't use Facebook and how, you know, how am I adversely impacted by this? Well, totally depends on who you are. I'm not going to speak for anybody about myself. I don't use Facebook all that much. So it's like, I just wouldn't use it. I mean, I'm barely there anyway, but if you're listening to this saying, well, I'm there a lot, you know, that's where all my people are. That's, you know, where I, you know, buy and sell goods and, you know, and find, you know, cool stories from my social graph. Like all of that, I don't want to diminish that if this is the place that you want to be. I also think that there's some workarounds when it comes to being tracked online, but of course not everybody either knows how to or even cares to use some of that stuff. So I guess it's an okay option. I just can't see it catching on that much. Again, I still stick on the idea that like, well, if it's worth that much to you, is it worth 10 euros? Like maybe it is worth 10 euros if you're like, no, but I have to be on Facebook. If 10 euros is too much, well, what about five euros? I feel like the argument is, no, I want it to remain free. I just don't want it to change. And what Facebook is saying is we won't make enough money on the ads that aren't targeted to make it worth our while. So we are going to give you two options. Either got personalized ads, which case we're making enough money to give it free or give us the equivalent. Now we can argue all day long about whether 10 euros is the equivalent or not, but I don't know if that's the government's job to set the price. I don't know if it's the court's job to set the price. It shouldn't be the market's job to set that price. And the only way to let that happen would be to let Facebook try to charge 10 euros and see if people quit. I think where my argument falls down. Your bites, you know, they might get a lot of signups. Is that a lot of people will just not pay and let themselves be tracked because they're not thinking that hard about it. And I think that's where a lot of critics go is like, yes, but for people's own good, we should force Facebook to give them another option. Well, and I think Facebook has the not great distinction of being a place where perhaps more than anywhere else, I don't know, maybe X is in the running at this point, you know, for a lot of people to say, well, you know, they're going to take your DNA unless you pay, you know, because that's, you know, that is sort of the, you know, the one side of the, you know, the conspiracy theory thing that's been going on on Facebook for years. This has always been like a weird strange urban legend that's been floating around. Oh, they're going to make you pay unless you like, you know, say this specific sentence and post it to your feed and then make 10 of your friends post the same thing. You know, it kind of feeds into something that I think is going to turn into a misinformation thing for a lot of people. Should it go forward? But yeah, honestly, on its surface, you want to pay for no ads, no tracking. Great. You don't. Okay. You have another option. Yeah. And it's not going to happen outside of the EU. That that's fairly clear from Wall Street Journal sources. Yeah. So, you know, even if you're listening to the saying shakes fist, it very well may not apply to you now. Rest easy, Australians. You're fine. There you go. Yesterday, we noted Android authorities report that code indicated that TikTok might offer also offer monthly subscriptions in the U.S. for $4.99 for ad free experiences. TikTok has clarified to TechCrunch. It's testing the feature in a single English language market outside the U.S. Didn't say which one. But yeah, it's not happening in the U.S. Yeah. That's right. Not right now. And not happening in the EU then because I don't think there's an English language market in the EU right now. TechCrunch notes that New York based company called cloaked, like a cloak, C-L-O-A-K, is out of private beta. Here's what they give you, Sarah, for $10 a month, close to the price of Facebook, for $10 a month or $100 a year, cloaked lets users create multiple identities of different email addresses, usernames, password and phone numbers that can be used at different websites. It works as a Chrome plugin, a web app, and on Android and iOS apps. It can not only relay emails to your real address. Lots of services like this do that where they're like, we'll give them a fake address and then anything to that address will relay to you. It can also act as your inbox. If you don't even want the stuff to from a particular source to even hit your Gmail or your Yahoo Mail or whatever, you can use cloaked as the inbox, as the endpoint. In the United States, you can even create virtual phone numbers since cloaked is a phone company. They created themselves as an MVNO basically, so they are working with the tier one carrier and have phone numbers that they will route for you and it acts as a password manager too. So if you're using that Chrome extension, it can enter your passwords. It can even enter your second factor codes. It can handle your multi-factor authentication and enter those codes for you. And they say customer data is stored in separate encrypted databases from everything else to try to keep what happened to last pass from happening to them. There's lots of other interesting things about how this works, Sarah, but just on the surface there, what do you think? Well, at first, I was sort of like, all right. Well, I do a fair amount of obfuscating my email address anyway, depending on who I'm working with. And then there's sort of like the kind of people who are used to Google Voice or Apple allowing you to use an email that isn't actually your email and Apple read it to you. And I asked a friend, what do you think? Does the price point seem good? And they said, is this not different from Pseudo? I use Pseudo for this already and I have for some time. And I looked at the kind of the price point and the feature list of both and I was like, I don't know because this is kind of new to me. I don't hate this idea. I think you probably, if you are a cloaked user, you mentioned LastPass, Tom, you need to really trust that the company has its wits together as far as keeping all of your information safe. Now, not to say that LastPass doesn't, but things happen. We know that things happen. So that would kind of be my first question is like, all right, well, how much do you trust this company that you may not have worked with before? Yeah, I would like to see some vetting on them before I signed up for it. I'm not familiar with Pseudo. Is it S-U-D-O or P-S-E-U-D-O? S-U-D-O, yeah. Because from what I've seen cloaked is taking a lot of things that other folks do and wrapping it up in one service across multiple places in a way that I hadn't seen before. It looks like Pseudo, you get a phone handle that does phone, email browser and virtual credit card for 99 cents a month. So it's much cheaper. I'm going to guess what cloaked is trying to say is we're a phone company. So we're doing the phone numbers ourselves. Pseudo may or may not be doing that. I don't know. And I think they're offering more ways to use it with the web app and the Chrome plugin, et cetera. But yeah, Pseudo looks pretty interesting as well. Yeah, I got to say it all comes down to the price point. If this is something that you can part with $10 a month for and it makes you feel more secure, well, not just feel more secure, be more secure and private. I think it's great. I'm totally into this. I don't really feel like I have a need for this personally and that's probably just because I'm used to my data being collected and getting lots of spam email. And I just sort of try to circumvent that the best way I can, but this would be, I guess, probably the best way to kind of say like I declare data bankruptcy and we're going to start over and build something, you know, based on something that cloaked is offering $100 a year. Yeah, I don't know. It sounds like a lot to me. It does. Yeah, especially when you compare it to Pseudo, which I'm just learning about now as we speak. Apologies for that audience. But this is something that's much more competitively priced and offers things that cloaked is not offering yet. You know, things like credit cards that they want to offer. One of the interesting things cloaked is doing to protect your privacy is using one phone number for multiple peoples. So because they are a phone company, they can do that. They can route the calls and say, well, when the origin of the message is associated with this account will send that phone numbers call to that that person's phone. It's like a party line basically. The idea is to prevent aggregators from deducing who you are by aggregating data around a particular number, even if they don't have all of your data, they have all the data associated with that number. But this kind of so-called data poisoning will mean that people who aren't you are related to that number. And so that data becomes worthless because it's multiple people. I would wonder what would happen the first time that goes wrong and they route a call to the wrong person though, because that's going to be pretty deadly for a privacy-protecting organization. So I'm not sure how bulletproof that kind of system is or isn't. Well, for all you Android users out there, we got a show for you called Android Faithful hosted by Ron Richards and Huan Tui Dao. Android Faithful is a podcast devoted exclusively to Android news and information. You can catch it Tuesday live 8 p.m. Eastern Time, 5 p.m. Pacific. Watch it live on our YouTube channel, youtube.com slash Daily Tech News Show. Or subscribe to it right now at www.androidfaithful.com. Well, your expectations and beliefs about a chat bot might affect how you talk to it, thus affecting its responses back to you. Scientists at MIT and Arizona State University told users that a language model was either empathetic or neutral or manipulative. So Tom, what did they find? Yeah, so if they told the bot, if they told people that the bot was empathetic, those people gave it a higher rating than the people they told the bot was manipulative. And remember, they got the exact same bot. This is all their perception. 44% of those told it was manipulative though, rated it as malicious, whereas 88% of those told it was empathetic, rated it as empathetic, and 79% of those who were told it's neutral, rated it as neutral. So in other words, being told it was manipulative didn't affect your opinion into thinking it was malicious as much as being told it was empathetic or neutral affected your opinions of those. And those who believed it was empathetic became more positive in their interactions with it over time. And remember with these large language models, how you say things to it affects how it says things back to you. The opposite was true as well. If you were told it was manipulative, you started to be more upset with it faster and therefore changed how it respond to you. These results are published. They are peer reviewed. They're in the Nature Machine Intelligence Journal if you want to take a look at them. But what do we do with this? The author has suggested some ideas, but I don't know, Sarah, what do you think? I was trying to, you know, my first comparison is always like, well, if I'm at a dinner party and someone's being manipulative, but it's like, okay, but that's not how you're interacting with a chatbot for the most part, right? Also this is, the chatbots are acting the same way to everybody, right? You're just told to expect something. Like it's going to be manipulative or oh, no, it's going to be really nice to you. Exactly. I guess if I was, you know, if I'm going towards some treasure trove of data, like a library, for example, you know, if I had some sort of library bot that for some reason I thought was being manipulative, it's like, is it because you want me to buy a book or well, it's a library. So not even a good comparison. I can't really think of a real case scenario because it's a chatbot where this is to your dinner, to your dinner party example. It's not you're at a dinner party and somebody starts acting manipulative. It's you go to a dinner party and your friend goes, that guy over there is really manipulative, right? And then you start to expect it. Versus if your friend said, oh, that guy over there is so nice. The guy could act the exact same. Well, and manipulative is a funny term, right? I mean, by the way, that's the term I decided to use to describe this. It's not actually the way they described it to the users here too. So keep that in mind. Yeah, yeah. But if I knew, okay, so guy had dinner party over at the other side of the room is running a horse rescue organization and is manipulating me into like giving him money. It's like, that's not necessarily something I would consider malicious. It might be something that influences me. Manipulation isn't always you're a bad person type thing or you're a bad bot. I think, you know, a lot of this. For the study's sake, I'm just using the word manipulative and maybe I should have picked a different word because for the study's sake, it was like expect this chatbot to be mean to you and expect this chatbot to be nice to you is kind of the sense of it, right? Well, yeah. I mean, I think if you were to say, hey, Sarah, this person C, you know, we're a B person C is not going to be all that great to you. You know, just know that going in, you know, I might have my guard up, I guess. And I think that, you know, whether it's a bot or a human, we sort of if we have information like that going into any scenario, we might feel the same way. Doesn't necessarily mean that what you get out of it is going to be negative though. And I think that's kind of where the interesting part of this is is if people say, yeah, well, you know, it's a mean bot. So, you know, anything they say, grain of salt type thing. I mean, you can say that about so many situations where you don't necessarily come out of it any worse off. Yeah, what I found fascinating about the study is that it showed that we are willing to believe the best because when we were told, hey, this is going to be real nice to you. 88% rated it as nice. Same exact bot, by the way, let me just reemphasize that doing the same exact thing. Answering the same exact questions. When we were told it was malicious or mean, 44% rated it as malicious. So people were less likely to want to believe that the thing was malicious. And another thing that I want to point out is half of the groups across all three categories, right? So within each category that they divided up, they had 310 people. They divided them up into groups of around 100 or so. Half of each of the groups got a GPT-3 model. Half of them got ELISA, that's 60s era chatbot that is very mechanical. And they basically said that there was a lot more reaction to the GPT-3 models here, which I think you might expect because the GPT-3 model can react in a way that ELISA can't. The GPT-3 model, when you started, well, I expect this thing to be mean. So I'm going to be defensive. It changed the way the GPT-3 model reacted to you. So the upshot here that the... That's real manipulation. But the upshot here that the authors are saying is based on what they want to see and they want to do longer studies to see if it lasts or if it was just temporary. The way we think about a large language model when we go into interacting with it is going to change how it responds to us. And so how a company pitches it when you log on, what you've heard about them in the media is going to affect what kind of responses you get from these chatbots. I mean, honestly, I'm not talking to a chatbot currently on a daily basis, besides research purposes, certainly, but I still think it's sort of like, I was about to say her name, Amazon's assistant type thing, who I do interact with on a daily basis throughout the day, every single day of my life. Just imagine if I thought that she was not very nice to me, how things would change when I asked her to look at something on Safari. It's a great example. Those voice assistants, even though they're not large language models, they're a little bit less sophisticated, we ascribe tone to them when they frustrate us even though their tone never changes. It's always the same tone. No, it's always like, was that helpful? No, okay. Patrons, stick around for the extended show Good Day Internet. There are a few more AI stories kicking around that we're going to talk about, including the fact that watermarks just don't work yet. Oh boy. Just a reminder though, you can catch the show live. Monday through Friday, DTNS is live. 4 p.m. Eastern, 200 UTC. You can find out more at dailytechnewshow.com slash live. We're back tomorrow, talking about the Pixel 8 announcement with Android Faithful's Michel Rahman. Talk to you later. This show is part of the Frog Pants Network. Get more at frogpants.com. Ironman Club hopes you've enjoyed this bro.