 Coming up on DTNS, another announcement about Apple TV Plus, Google Go goes global and streaming subscriptions seriously stacking up. The news for August 20th, 2019 from Studio Feline, I'm Sarah Lane. And back on the show, I'm Patrick Beja from Finland. And I'm the show's producer, Roger Chang. Patrick and Roger and Amos and I were talking about why French children have fewer tantrums than American children. And some people are not believing me that that is the case. We might continue this discussion in the second part of Good Day Internet. Exactly. The wider conversation happens before and after DTNS. We call it Good Day Internet or GDI. And you can join in on the fund by becoming a member at patreon.com slash D T N S. For now, let's start with a few tech things you should know. Apple's no fee Apple card is now available to all United States users through the iOS wallet app. Apple announced it's extending its 3% daily cash offer to Uber and Uber Eats purchases with more third party merchants on the way. Customers get 3% cashback on Apple purchases as well, 2% on any Apple pay purchases and 1% cashback using the physical Apple card. The Wall Street Journal reports that a group of state attorneys general may launch a probe next month into whether large tech companies are abusing their power to stifle market competition. Last month, the Department of Justice announced its own antitrust review of how online platforms like Google, Facebook and Amazon achieve market power. And back in February, the Federal Trade Commission formed a task force to monitor competition among tech platforms. Regulation. Maybe. Yeah, seems like a lot of people are interested in exploring further. Ubico is releasing the Yuba Key 5Ci security key that can plug into an iPhone's Lightning Port or a USB-C port. For 70 bucks, the key will support one password, Bitwarden, Dashlane, Adaptive, LastPass and Okja. It can also securely log into Twitter, GitHub, OnePassword's web app and a few other services using the Brave browser for iOS. The 5Ci doesn't work with the newest iPad Pros. A few people will be disappointed by that. And services also have to individually add support for Lightning connectors on the 5Ci into their apps. Microsoft has acquired J-Clarity, which makes tools for Java developers. J-Clarity's sensome product identifies software inefficiencies and its eliminate product flags performance issues and uses machine learning to troubleshoot issues and solutions. Microsoft says it will work with J-Clarity engineers to make Azure a better platform for Java customers and internal teams. All right, let's talk a little bit more about the facelift on Facebook. We have some new tools anyway. Got a couple here. Facebook is planning a new section in its mobile app called NewsTab. That's not the news feed. It's a whole different thing. To be curated by human editors and journalists, not an algorithm. NewsTab ends to surface recent and relevant stories and says that jobs will start posting to its employment board on Tuesday. They're going to hire some folks. They say seasoned journalists. So I guess they'll be wanting to hire all of us. NewsTab will exist outside the main news feed, which Facebook has been tweaking in recent months to focus less on publishers and marketers and more on personal interaction. That's what they tell us anyway. Sources tell The New York Times that Facebook will pay at least some partners to license and display articles, not all of them. And another new tool that has seen months of delays. People will be a little bit happy to know that Facebook's clear history is now rolling out in certain areas. Ireland, South Korea, and Spain, with other countries to follow in the coming months. Clear history is part of a new section called Off Facebook Activity that shows which apps and websites are tracking your activity for ad targeting purposes. You can disassociate from your Facebook account using this new tool. You can also preemptively block companies from tracking you in the future if it's a company you don't like and you don't want them to know what's up. I find the human curated news thing an old trope, really. It certainly works out well for a lot of other services. But I just don't know that anybody equates Facebook with this kind of news gathering. I think it's something that they don't have a choice in at this point because they've really tried and them and everyone else has really, really tried to make automated algorithm based news selection work. And they still have parts of their news surfacing that works like that. But I think it's them not really throwing in the towel, but maybe thinking, all right, we'll give it a try and see what happens. Obviously, this is something that cannot really scale in the way that algorithms do. And if you think about all the different locales, they would have to do this in order for it to work globally. It's certainly something that would be a significant investment. But on the other hand, nothing else they've tried has really worked. So if they want to have some news elements, which I think they do need, this might be a solution. So they have to give it a try, right? Yeah. I think that, and again, you really hit the nail on the head that it's just, it's very late in the game to try something like this. I think if some years ago, because of course we're news gathering all day, and I've got my tech memes, and I've got my feeds, and there are certain sites that I know and trust, and I've got a whole system. If Facebook could replicate some, probably never all, but some of what my daily routine is, I would be into that. But unfortunately, I don't need Facebook for that at this point. So it would take a lot for me to be like, oh, I'm going to hang out on Facebook more and get my news from there from trusted sources. The trusted sources part is great, if indeed that's how it's going to work. But I'm more excited about the clear history to be honest. I think that this is, and we don't have the tool in the US for now, coming months, don't know how many months away that'll be. But it does sound like after some delays, some of the world is at least going to get this very important tool. Yeah, it's both are important. Clear history is a privacy tool, which obviously is important for Facebook to improve in. And the curated news is society impacting. So that's important as well. All right, Google Go, Google's lightweight search app, is now available to all Android users worldwide. It first launched in 2017. It uses just over seven megabytes of space to help in emerging markets with less stable internet and lower and Android devices. It includes offline features, and the company says the search results are optimized to save up to 40% data. At IO, Google also announced it would be bringing Google Lens to Google, to Google Go, just Lens, actually. According to data from sensor tower, Google Go has been installed approximately 17.5 million times globally, with almost half of that number coming from India, then Indonesia, Brazil, Nigeria, and South Africa. I love Google Go. I don't have an Android device, but I love the sound of it. Hey, it takes up very little space. You got Lens capability, and clearly a lot of folks who have had access to Google Go like it. I mean, 17.5 million, quite a few in India alone, people who have not downloaded Google Go. But it's clearly a number that the company says, okay, this is actually something that people are taking advantage of, particularly slower connections or feature phones that don't have a lot of space on them, a lot of offline mode stuff that Google Go takes advantage of, for when you don't have any internet at all. So you can still access a lot of the things that you might otherwise need a connection for. I don't know. You're clearly sacrificing some features by using Google Go rather than my mobile Chrome app that I'm using now. But in general, I hope that mobile browsers can get more lightweight over time rather than bulkier. Yeah, I guess in this case, it's less feature-full, I suppose. I would have to try it out. I have to admit, the 17.5 million downloads seems low to me. Maybe they're not fully advertising it yet or something like that, but I would have expected that number to be a little bit higher. Sure. Well, again, in certain markets, not available at all. So yeah, over the next six months, that number may skyrocket, may not. We will revisit. According to a survey conducted by Motherboard, natural language processing artificial intelligence systems, often called automated essay scoring engines, are now either the first or second most used grading system on standardized tests and at least 21 U.S. states. Of those states, three said that every essay is also graded by a human, but in the other 18 states, between 5 and 20% of essays got randomly selected for a human grader to double-check the machine's work, meaning the lion's share was done by a machine and not checked by a human at all. Psychometricians, cool word, right, who study testing, and AI experts argue that these tools have bias flaws. They can also be fooled by essays that actually don't make any sense, but with the right sophisticated vocabulary might trick a machine. The engines are trained to recognize patterns that correlate with higher or lower human assigned grades. The machines can predict what score a human would probably assign an essay, but metrics like sentence length or spelling and subject verb agreement can't always pick up creativity and other nuances or account for English language learners. So, you know, and I know the testing system varies by country, Patrick, and so this is clearly U.S.-based, but when you're talking about machines that are grading, what may mark the difference, which may set the course for where a student's going to go to college or what kind of funding that school is going to get from the state in the next school year, these are pretty big issues. My reaction to this was a lot less measured than yours, Sarah. I was like, are you kidding me? Like, seriously? Between 80 and 95 percent of those essays were just graded by machines and no human laid eyes on them. Like, this is unbelievable to me. We know the state of AI. We know how well it can work and how flawed it can be. It can work well for some things, but an essay in school with those consequences and the importance of those things that you mentioned, I cannot believe that that many states have integrated, and I'm sure there's a question of costs that is involved, but it would be one thing if the algorithm would pre-study the essay and then give recommendations and then a human teacher would look at it and decide what is appropriate and what isn't, but just giving it to the machine and then it grades it and that's it. It is, I cannot, like if the people watching me on video, my brain is exploding a little bit. Like, I cannot comprehend that this is acceptable in any academic institution. I will add these are for standardized tests, so these aren't like for, these are for tests, but it's not like every essay you write. For example, a book report or a book of Louisiana purchase or something like that. Right, but it's still an essay. It's not a multiple choice question, right? It's something that is written in English with sentences and- That's right. And it's also, it's obviously individual grades, but it also tends to make researchers be like, okay, in this region, the students did more poorly than the students in this region, or you can make assumptions about certain racial groups and economic factors and a lot goes into this and Motherboards study, which was extremely detailed, they did say in many cases, if there are some tests that are being randomly looked at twice by a human and there's enough of a discrepancy between what a machine came up to and what a human might have come to, then sometimes these things do get looked at again. I mean, it's not as if they're just like, well, we just, we just, just look at this, no human cares. There are people who care, who are trying to figure out where the system might need to be in better shape. But yeah, I mean, this is just, you know, this is just, it's not like, oh, the robot delivered my Uber Eats order. Yeah, it's a lot more at stake. To me, it's in the same category as voting machines. You need sometimes the technology, even if it is all ready, which it isn't in this case. First, we need to wait a little bit longer to implement it on a large scale. And second, for some of the things, you just don't mess with it. IBM is moving its open power foundation underneath the Linux Foundation, an umbrella, if you will. The move means Power ISA will be open source, removing the expensive licensing that open power foundation previously required to implement the technology. Along with the ISA, IBM will be releasing reference guides for embedding the instruction set on non-open power chips, processors, and memory interface protocols. The underlying technical architecture of open power chips, however, will still remain under the open power foundation license. Roger, before the show, you said this is kind of a big deal. Yes, so for people who aren't familiar with the ISA, the instruction set architecture is essentially the language or the kind of set of instructions that a processor uses. If you use an Intel chip, it's some variation of an X86. So when someone says instruction set architecture, well, what does that mean? That's what it means. So it means that whatever code base you have to compile to run on that machine, you need it to run on that process and you need to know what the instruction set architecture is. Now, what's so important about the power, open power foundation move is essentially the power ISA, which is based off the old power PC, actually the power instruction set, which was also used on the power PC if you had one of the old PowerMax. If you had a GameCube, if you had some variation, or if you had a Sony PlayStation 3, or if you had a Xbox 360, I'll use some variant of the power ISA. And what this means is that there is now a viable third alternative to ARM, which licenses its instruction set architecture and its specs to whatever company, whether it's Samsung or Apple to implement that technology, or an X86, which you need to either go to Intel or AMD to implement and run. It just means that ideally there will be a third player, especially in the hotly contested server market, which ARM is trying to reach into. But as we've seen with Qualcomm, not everyone is very successful with implementing that particular set. Was power still being developed though? It seemed to me like it died with, okay. Oh no, at this stage it's more of a back end stuff. So very few consumer products that you use will be using it. But definitely on the back end. So if you're running a cloud service or something, or a database, some or an academic arena, you'll be using some of those machines will be running a power processor. And again, this is just the instruction set architecture. That's basically how everything should run and interface with each other. It does not mean the actual physical architecture of the open power chips, which is like how we built the thing. And you still need a license for that. Interesting. Well, nine to five Mac posted their posted their hands on impressions of an early preview of Apple Arcade, the $5 a month gaming service featuring 100 games or more on demand operating across iOS Mac OS and Apple TV. The site notes Apple charges 49 cents a month to test Apple Arcade until iOS 13 launches. Nine to five Mac notes that these currently isn't a unified experience across the different games they played with each having their own settings. Other observations, certain games like Hot Lava struggled on the author's 2019 MacBook Pro at high settings. The quality of many games is high. And if only 25% of titles Apple Arcade were of similar quality, the service would still be worth it. Well, Patrick, do you agree? So we go before the show that I what the report was stating and what I was seeing on screen in the hands on video didn't really match. To me, this really seems like a beta version of essentially what I'm saying is as a lifelong gamer, the games I saw were not impressive at all. And the report was kind of gushing over these experiences, which I had a hard time understanding. But beyond that, the system seems to be working. It's kind of what we expected. It is, as we said, a cross platform, cross Apple platforms. And the initial games are just for the purposes of testing. So I think if we have more games and more interesting games, it could be an interesting service, especially at the lower price of $5 a month. It's kind of what we're expecting. And I'm hoping that we see more games that convince the big question is, is this going to be for, quote unquote, traditional gamers or for mobile gamers? Mobile gamers already, they have their fill with the existing games and the free to play games. So if this is a play for other types of gamers to get into that ecosystem, we need a different kind of games than what is currently present in the beta, which, again, is just a beta. So it doesn't mean much. Yeah, the $5 a month, you go, OK, all right. That's going to be a sweet spot of a price for some folks. More than a hundred games. That sounds like a lot. That also sounds like all of the channels I don't watch when I used to subscribe to Cable. So how many games am I really going to play, even if I'm, I don't know, really into it, 20? That would be a lot. I think that there's a lot of that where people go, quite a bit of variety. That's great. But you have to think about what you're actually paying for and what you're actually using. If you get, I would say, a handful of games that are interesting every two months, it's already kind of worth it if you look at. And again, I'm not talking about mobile gamers who seem to prefer free to play games. And so that might not be for them. But for more traditional gamers, this could be an interesting price for if the, even if you get only a handful of good games. Well, folks, to get all the tech headlines each day in about five minutes, subscribe to dailytechheadlines.com. Well, you might subscribe to Apple Arcade. You might not. But it kind of brings us to another question, Patrick. And we circle back to this question often here on DTNS. And that is, how many subscriptions can we take? There's only, you know, money is finite, right? Time, time as well. So how are we, and I want to throw the question at you first. And I don't know how many things you subscribe to, whether it's television content, movie content, gaming content, or some combination of all three. How do you decide what you pay for? And are there services that you really want that you've decided aren't worth it? I realize I'm subscribing to quite a few things, because in the way I view content, all of them are incredible deals. And I kind of get into a situation where it's too much for me to actually enjoy, so the limit might be time more than money, because all things considered, of course, everyone has different means. But even if you subscribe to, you know, two or three video streaming services, a music service, and one or two gaming services, the value of it, it's definitely less than 100 bucks. I would say 50, 60, 70 bucks a month, but you have the entirety of your entertainment needs covered. So personally, I have, of course, Netflix, Amazon Prime, but that's included in, yeah, I mean, Amazon Prime Video included in Prime. I have HBO Nordic, because I'm in Nordic lands. That's just for video. I also have one music service, YouTube Premium, because I don't want the ads, Audible, because I have. So I have a tolerance, I think, that is quite high for stacking up subscriptions, but I'm hearing a lot of people who are saying, I will never subscribe to more than two video services, for example, or especially in the gaming space, oh, I don't want to pay for more than one gaming subscription service, because that's ridiculous, and there seems to be this allergic reaction to stacking up subscriptions. I wonder if I'm alone in that weird state of, I don't mind, it's great value. Yeah, I also am, when people complain about, oh, you know, I have to manage all my subscriptions, I kind of go, I'm sort of auto paying every month, I know what I'm paying for, that doesn't bother me so much. The solution to have everything all in one place is sort of beside the point to me, I just don't want to pay too much. And that actually kind of leads us into Bloomberg article today that Apple is planning to roll out its Apple TV Plus movie and subscription service by November, and sources to Bloomberg, that service may go for $9.99 per month after a free trial. So you can check it out and then see if you like the shows enough to want to stay. And a big criticism of this is 10 bucks a month is a lot for five original programs, you know, out of the gate. Of course, Apple is going to expand on that. The same thing with iTunes back in the day. But I also, you know, I know people who subscribe to HBO for one show. They just want to, you know, they're not really watching the back catalog of all the movies that the HBO Go app has to offer, say. And they're not alone. That's kind of the way it goes. You need one hit and a lot of subscription services do quite well as a result. Just look at Hulu. Hulu's, you know, on the up and up these days. And you could say that about, you know, Netflix and Amazon Prime and many other, even regional offerings. I know France is launching its upcoming Salto service, which is something that at least it tells is, you know, it's going to be a big hit with the French community. I know you mentioned that it's been a bit delayed thus far. But the UK has that as well. So I wonder how much it really matters the amount of titles and back catalog you get with a service to make you feel better about paying a certain price for that service. When realistically, you're never going to take advantage of all that. I think it's a comfort thing. People are going to be a little bit, you're going to see the memes of, oh, 10 bucks a month for five shows. If the five are hits, I think it would be worth it. But it's kind of difficult to rationalize this into actually paying for it, for people who are, you know, already against it to begin with. And if we get back to the stacking thing, 10 bucks a month, an additional 10 bucks a month on top of, you know, Disney is coming. A lot of people are going to be subscribing to that. Netflix is not the only player in town anymore. A lot of people are drawing the comparison to the olden days of cable and saying, oh, you know, we're back to paying 150 bucks a month for cable. That was, it's just the same situation. We've just changed it into something that's organized differently. There again, I don't really see it that way, if only because you don't have things bundled yet. But, you know, you don't have bundles you choose. You can swap whenever you want very easily on each of those services. You pay for the ones you want from one month to the next. It's vastly improved in my opinion. Yeah, I've actually signed up and, and, and canceled YouTube TV several times. Just because, because it mimics cable as much as any service I've seen so far. And there will be months where I go like, you know what, all this is doing, all I'm doing is watching junk. There's nothing to belly. I'm going to cancel it. And then there'll be something where I want it back. And it actually is extremely convenient to go back and forth. And the bundles is interesting too. Disney Plus is doing that $13 per month bundle thrown in Hulu and ESPN Plus, where they're all cheaper because you're getting three of them together. If you did all a cart separately, you'd be paying more. So companies are getting creative about this. But, but yeah, it does start to turn into something where you're like, well, it was a little bit cheaper, but am I actually using it? Did I really want it? Would I have bought it on its own otherwise? If I can jump in here a little bit. My philosophy, and I'm a little bit older on this, is that I used to pay $150 a month for cable. And with now with gaming service through Humble monthly, my Apple Music subscription, Netflix, HBO, Disney, when it comes out, because that's guaranteed I'm going to get that, I'm still paying way less than that $150 a month. And I'm getting more variety, more mediums, you know, more different types of media. And I, as you said, Sarah, I have the choice of whether or not I want to cancel right away or let it go. And I can still tailor and still stay way below that $150 a month for not having a cable subscription. Yeah. Well, one size does not fit all. Certainly we have more choice than ever. Well, that's a good thing. And yeah, it's kind of fascinating how, how, how people have figured out what works for them and what doesn't and what still needs to get better. Hey, thanks to everybody who participates in our subreddit. A lot of our stories come from you. You can submit stories and also vote on others at dailytechnewshow.reddit.com. We're also on Facebook, facebook.com slash groups slash daily tech news show. Patrick Beja, it's so good to have you back. Thanks for being with us today. What's new? What's been going on? Well, I'm back from vacation. So not a huge amount, but if you do speak French, I already have new episodes for Laurent Des Voutecs and Laurent Des Vouchers. So this is back up and running. So go check it out either on your podcast app or at Frenchspin.fr. And English language shows are coming fairly soon. So there's that. Excellent. Also, thanks to our patrons. Patreon makes us what we are and we could not do it without all of you. If you become a DTNS member and become one of our patrons, you get all sorts of perks. Add free RSS feeds, special episodes behind the scenes stuff, weekly newsletters, exclusive interviews with tech luminaries, smart people with really fascinating information, and lots more. You can sign up at patreon.com slash DTNS. If you got feedback for us, our email address is feedback at dailytechnewshow.com. We're live Monday through Friday, 4 30 p.m. Eastern. That's 2030 UTC. And you can find out more at dailytechnewshow.com slash live. Back tomorrow with Scott Johnson. Talk to you then.