 Coming up on DTNS, Samsung says the Galaxy Fold lives! Gaming adults say they're getting harassed and did the retweet set us all back? This is the Daily Tech News for Thursday, July 25th, 2019 in Los Angeles. I'm Tom Merritt. And from Studio Feline, I'm Sarah Lane. From the shores of Lake Merritt, I'm Justin Robert Young. And I'm the show's producer, Roger Chen. Now we just had an interesting conversation of when is it too late for you to respond to a text message? Like what's your cutoff time? If you're a good day internet listener, you just heard that. If you're not, and you're like, man, I want to hang out with you guys and talk about stuff like that, sign up, become a member on the good day internet tier at patreon.com slash DTNS. Let's start this show with a few tech things you should know. eBay is launching to manage delivery and to end fulfillment offering for eBay sellers in the US next year. The company will store seller merchandise and third-party warehouses and give sellers the option to provide free two to three day shipping. Around 1.5 million packages are currently sent daily in the US by eBay sellers. So managed delivery will better brand those packages for the company. NBC and Twitter are partnering to bring limited live Olympic coverage and highlights to the social network during the Tokyo 2020 games. NBC previously worked with Snapchat for the 2018 winner games. The Twitter partnership has an interactive twist. Users can vote on preferred events or athletes and live streams lasting around five minutes will be broadcast on the platform. Javelin, team handball, team handball. Stock Training Service Robin Hood is informing some customers it stored their passwords in clear text in an internal system. According to email seen by ZD net Robin Hood says it's fixed the issue and is now resetting passwords out of caution and says it did not find any evidence of abuse passwords are being recached using the b-crypt algorithm at Robin Hood now. Samsung announced it made improvements to the Galaxy Fold and plans to sell it in September for $1,980. The top protective layer of the display has been extended beyond the bezel which should help deter users from thinking that they should remove it like the typical screen protector which was a problem in the past. Both hinges of the device now feature protective caps. The space between the hinge and the body has been reduced and the screen has been reinforced and additional metal layers underneath. And Apple is buying the majority of Intel's smartphone modem business for $1 billion as was previously expected. That deal got done a little faster than people expected actually. Apple will get 2,200 Intel workers along with equipment, intellectual property and leases. Let's talk a little bit more about a lawsuit Justin. Indeed, Tom US presidential candidate Tulsi Gabbard is suing Google for $50 million after her campaign was suspended from buying ads and for six hours following a debate in Miami in June. Google told the New York Times that its automated system flags unusual activity like sudden increases in spending. The system triggered the suspension of Gabbard's campaign and the account was reinstated shortly afterward. The campaign also believes that its emails are being placed in Gmail spam folders at a disproportionately high rate. Well, maybe you can be a consultant to the campaign on how to defeat the Alderitos that puts your newsletters in the spam folder. But this lawsuit against Google is very interesting to me because this is basically the same thing that underlies a lot of the debates around moderation and whether things should be allowed to be posted, which is, hey, we have an automated system for that and the automated system screwed up. So it took us six hours to notice and get it fixed. The problem is, is that this was a very, very, very valuable six hours for Tulsi Gabbard. Here's the reason why I think this lawsuit is being filed and they should be very, very angry about it. The primary system this year is very unusual just in terms of how much we're paying attention to it. It normally doesn't have this kind of attention this early in the process, but more specifically, the Democratic National Committee has in an effort to thin out the field a little bit put in debate floors, debate qualifying floors for people to be able to get on stage during the, during these 10 debates, one of which we've already seen, and this is the one she's suing over. So she has a good debate in June and right afterward when attention on her is at its highest, she now can't buy ads to get people to fund her campaign. This is why it's even more important because one of those debate floors is individual donors. And that's a lot of these candidates that are polling around where she is are having a hard time getting to that number. So that six hours is prime time where people just saw her on television. They might have connected with her message any way that she can connect with them so they can give a dollar is disproportionately valuable to her. And that's why you're seeing such righteous anger in a $50 million lawsuit. All right. So based on all of that and the fact that she might have a really good case, is she going to get $50 million from Google? Likely not. I mean, I'm not a legal expert. I think a lot of these kinds of things are, I mean, to be honest, I think that this lawsuit will probably do more to just get her name in the news. So people might reconnect with her. And again, if you are for Tulsi Gabbard, then donating a dollar to her does do a lot to keep her on the debate stage. So I think this does more to just get her name out there and is a little bit of a dog whistle to anti-technology folks. This has become certainly a campaign issue to say Tulsi Gabbard versus Google. However, simplified that version, that headline would be that is on brand for her. Yeah. She's not the only candidate on the Democratic side that is using anti-tech as a way to drum up support. So being one of the only ones that's actually suing a company could be a badge of honor, I suppose, because that $50 million isn't going to help her get on the debate stage because that doesn't qualify as a small donation, right? I mean, it certainly couldn't hurt, but also these court cases don't get resolved fast usually. So I wouldn't expect her to get that $50 million anytime soon. This is all done for being able to talk about it. It seems. Yeah. No. You'll just know if the settlement is that the entire head count of Google has to make individual donations with Tulsi Gabbard. I don't think that would be legal, would it? I mean, no, no, in their name or something like that. Let the lawyers figure it out. Yeah, right, right. That would be, that would be crazy. That'd be really interesting. The Anti-Defamation League published a new report that found 74% of adults who play games online have experienced some form of harassment. In a survey from April of 2019 of more than 1,000 adults, 65% reported severe harassment, including physical threats, stalking, and sustained harassment, while 29% reported being doxxed. The report found that Dota 2 had the most adults quit as playing a result of harassment. Other games found to have high instances of harassment were Counter-Strike, Global Offensive, Overwatch, Player Unknown's Battlegrounds, and League of Legends. 80% reported having positive social interactions within games, with World of Warcraft and Minecraft and NBA2K having the highest percentage. Fun games. 35% of respondents said that they engaged in negative social behavior themselves, from trolling to using offensive names. The report also found moderation policies outdated. With some large commercial spaces unmoderated, ADL suggested that the ESRB take into account online behavior and a publisher's moderation policy when assigning radians to games. Yeah, this is probably not going to shock anybody, right? These numbers are not surprising, but they hadn't been collected. And good science means, as I have said in lots of other arenas, going out and testing and finding out if your assumptions are true. The fact that it found out that a lot of assumptions are true is not bad or silly. It's saying, hey, now we've got some evidence that that's true. And now we can actually take some steps to address that. If we're seeing that this level of harassment is reported, maybe companies should do something about it. Maybe they should update their moderation policies. Maybe they should do moderation in some cases. And I think it's an interesting suggestion. I'm not sure how I feel about it for the ESRB to take that into account. I don't know if it should go in the ESRB rating, or if there should be some sort of level of being able to check how bad an online community seems to be. That starts to get weird, though, like how do you possibly rate that? But it's an interesting suggestion to say, like, hey, the game itself may not be that violent, but you may be exposing yourself to a lot of people talking trash at you. And maybe you don't want that. You should know that before you go in the game, buy the game. I think it's important that it be reported to the ESRB. And I do think that the more and more interaction with these game platforms results in online multiplayer situations where there are toxic communities, then especially because we're talking about a lot of these games are franchises, right? They are. They are people, they're games that make, you know, different versions of themselves. And if you have a persistent toxic community in one, then maybe on the next game you should just, the publisher should know, hey, look, we're going to ding you a little bit the next version of this game because you didn't take steps to make this better. Also, 1000 adults is not a huge sample size, but to have 65% of those adults saying, yes, I've experienced severe harassment, which include physical threats, that is not to be taken lightly. That is bad news for any online community, whether you're a gaming community or otherwise. I'm sure some people are like, yeah, that's just the way it is, right? Like, you know, people talk like that to each other in the game. I suppose maybe that is the way it is. I think gaming, ribbing, and I'm going to hurt you in real life are very different things. Well, certainly. But I'm just saying, like, I know there are people out there like, yeah, but you know, people say these things, you don't have to take them seriously. Okay, if that's the game you want to play, that's fine. But I think it's fair to say people should know like, oh, people are going to threaten to kill you if you do poorly on their team in this online game before you go in there. And also, heaven forbid that your voice sound like a woman or somebody not white. Oh, yeah, because by the way, it's important to note this applies to both text and voice chat. But yeah, yeah, I mean, if somebody threatens me, you know, like I'm going to come to your house and hurt you. Just not fun anymore. Yeah, don't don't want to play that game. It's like, no, that's it. I'm back to my life. I've known that before I spent money on this game. Pew Research Center analyzed every video posted by 43,770 YouTube channels with more than quarter million subscribers in the first week of 2019. Among the findings, 10% of those channels created 70% of the videos posted. The top 10% of the most viewed videos made up to 79% of all the views in that period. Only 17% of the quarter million videos analyzed were fully in English. Videos intended for children were longer and attended to receive more views than other videos. Videos featuring featuring a child younger than 13 whether intended for children or not received nearly three times as many views on average as other types of videos. And videos with children targeted at children were the most popular kind of video of all. And finally, 16% of English language videos were related to current events or politics. The majority were international and did not mention the United States. Yeah, I think this is this is a good perspective on on YouTube when thinking about what YouTube should do and why they're behaving the way they are. Is that yes, a lot of children watching YouTube, a lot of children in YouTube videos and a lot of videos intended for children to watch and and and so whatever reason all of these conversations about children watching YouTube are like, Well, but it's not meant for anyone younger than 13. Well, it's being used that way in a dominant fashion and YouTube needs to address that. The other part of this I find very fascinating is the the amount of English language videos and the content of current events or politics implying that the large majority of viewers of video out here are not in the US. And so a lot of these things where the where the US has a concern and YouTube seems to be slow to accommodate it maybe because of that. I don't think that we should be surprised by this. America, although it is obviously the home of YouTube and very, very popular here, we talk a lot about the disruption that YouTube has brought to our entertainment landscape. We have a lot of children oriented programming. We have multiple cable networks. We have multiple over the top streaming platforms around the world. That's not always the case in native languages. You know, this is something that YouTube is a global force in a very different way in some other parts of the world than it is here where it's just, oh, wow, it's this cool thing that the kids are paying attention to, but they're in more restrictive media environments or less children centric media environments. It can be the only game in town. All right, we've got a couple of stories in a row with numbers with hard numbers. And let's get back to some theory, some qualitative analysis. Jonathan's a train. A big fan is out. Jonathan's a train. He's a really smart guy has a piece in the New Yorker about what he calls intellectual debt. No, that's not what you think it is. That is studying and discovering that something works, but without knowing how. So, you know something works, but you haven't figured out the reason. So, he calls that an intellectual debt. There's sort of a deficit in our understanding. For instance, aspirin existed in many forms as a folk remedy. And then aspirin itself, the modern day version was discovered in 1897. But a convincing explanation of how aspirin actually worked to dull pain and reduce inflammation was not determined until 1995. We were just using aspirin like, I don't know, takes a great care of my headache, don't know how it works. So, as a train suggests, that machine learning also incurs intellectual debt because we don't know how algorithms get good at identifying cats, et cetera. We know how to train an algorithm. We know how to set up an algorithm. But how it actually learns what is a cat and what isn't? We don't know. That's a black box. And that means we will not know when it fails unless we know already what the answer should be. We can tell how good it is at identifying cats because we know cats. But there's a lot of scenarios for machine learning, especially in public health, where they're saying, we want machine learning to discover something that we haven't seen yet. And that intellectual debt could be a problem if we can't tell if it's right or not. Not a big deal, as the train says, for creating new pizza recipes, right? But kind of a big deal for health recommendations. He points out a study from MIT's lab six in 2017 that altered the pixels throughout a cat photo so that the photo itself just looked like a regular cat to a human. But Google's image recognition algorithm inception, which does the classification for image search, identified it as guacamole. And that's a weird looking cat. It was a green cat or was sitting in a taco bowl or something. No, it was just a regular image of a cat. But pixels that we couldn't see were put in a pattern that fooled the machine learning because the machine learning doesn't actually know what a cat is. It knows this arrangement of pixels. And I say cat, I get told I'm right. So I keep doing that. Now, as a train is concerned, the trade secrets, which keep us from being able to see data sets, proprietary code from businesses, could hinder how the analysis of ML works, because we can't work on figuring out how a data set ends up with a particular result. And that reliance on well-performing AI could reduce human oversight and prevent us from catching errors or bias. So something to be thinking of as we use machine learning. I think an obvious question that someone might ask is, well, okay, aren't humans the ones that are putting the machine learning trials into place? If lab six altered the pixels of a cat photo in order to trick the machine, couldn't they teach the machine to understand it was being tricked? Sort of. Yes. Like doing things like that help to figure out, okay, how do we make sure that it doesn't get tricked by that? But the principle might be so broad that, well, we could teach it how to recognize that is a cat, not guacamole, but is that lesson going to be widely applicable? And the point is bad actors could also go in and figure out other ways to rearrange things, other ways to fool the algorithm to get away with stuff. So it's more of a general situation of like, how do we make sure that we improve machine learning so that it doesn't have these errors as often? Well, yeah. And look, all systems can be hacked, including the ones between our ears, right? Like there are many, many ways that we are fooled and machine learning is no different. But I do think that this is a good, everybody let's pay attention to the reality of what machine learning is that this is not AI machine learning gets tossed around so much these days that we kind of tend to think that it's like, oh, no, it's at a certain point it becomes a magic spell. And it's not. It's a program and it has all the fail abilities of a program and we shouldn't over rely on it or just assume that it knows best going forward. Because as Sarah pointed out, it is just a system that we're going to have to bit by bit show, you know, we're going to have to patch when people try to fool it in the same way that we patch any system. Yeah. And it gets to the point where saying and machine learning can recognize a cat is an easy way to express something. So we understand what it's doing like, oh, when I give it a picture of a cat, it'll tell me whether it's a cat or not. But that's not actually what it's doing. It's a misrepresentation because the machine learning doesn't know what a cat is. It has no conception of cat. It doesn't have a way to check like, wait a minute, that cat looks a little weird. It we don't know how it's determining what a cat is and what it isn't. And that's that's kind of the point of the trains column. Hey, folks, if you want to get all the tech headlines each day in about five minutes, be sure to subscribe to Daily Tech headlines.com. Chris Weatherall has had a conversation with the folks at BuzzFeed saying he regrets building the retweet button for Twitter. He led the team that made the retweet button in 2009. Initially, when he was working on it, he said he thought it would help elevate underrepresented voices and to a large extent it has. But after he launched the button and saw how people were using it, he says he thought we might have just handed a four year old eluded weapon. Now he's not alone. Jason Goldman, the head of product for Twitter in 2009, is quoted in this BuzzFeed article as saying the biggest problem is the quote retweet. Quote, retweet allows for the dunk. It's the dunk mechanism because you can retweet it and then put your little add a little extra something. Yeah, Twitter CEO Jack Dorsey also told BuzzFeed definitely thinking about the incentives and ramifications of all actions, including retweet, retweet with comment, for instance, might encourage more consideration before spread. So he's like, well, maybe it's not so much dunking. Maybe it slows you down. And the concern seems to be around how the retweet button reduced friction in spreading. So before you had to copy and paste to make a retweet, and you had to type RT, that slowed you down, made you think about it. And maybe all of that was too much trouble for you like, you know, and it's not worth it. And you might not do it. Whereas pressing a button, boom, you didn't have to think about it. Yeah, I'm going to spread that boom, there it goes. That kind of virality attracted publications and journalists and politicians to the platform it caused Facebook to add a retweet like function at the time called mobile share. And in 2014, Weatherall says he started noticing how Gamergate posters were using retweets. There was no real defense for targets against reputational harm when they were targeting retweets at a particular person. And he says it dawned on me that this was not some small subset of people acting aberrantly. This might be how people behave and that scared me to death. Now, he doesn't say we should get rid of the retweet button. He has some ideas about what we could do about it. But Justin, I know just the hand wringing has an effect on you. Tom, this bothers me. It just it is it is a trend that we are now seeing and be they ex founders, ex board members, ex engineers. This is just this new trend in stories where Dr. Belieger, Dr. Frankenstein comes out and very dramatically says, yes, it was me. I built the monster that ruined society. And we all think about the flaws in our own humanity and what we have wrought with this technology. It just is tiresome to me. And it's not that he doesn't have good points. It's not that there's not a good discussion to be had around the virality of retweets, although I personally find that if we are this just is a very engineer's way of looking at the world that the world would have been different if I would have programmed these things a little differently because I am the true engineer of all humanity. And this would not have happened. Gamergate would not have existed on the level that it was. If I had not built the retweet button, which I don't subscribe to, but I can get where they're coming from. There's a lot of stuff that is now the norm within Twitter that was created by the community. Hashtags, another example. Twitter didn't do that. People just started doing that. And so it became something that has become part of the social network vernacular over a period of time. I personally, the retweet of Justin says, hey, had a great time on DTNS. Here's the link. I might just go ahead and retweet that. It's easier. But I might, if you were to say, had a great time on DTNS, but something weird happened, then I might do the quote retweet and add a little context. And I actually think that that's a great tool. And that's actually what I use more than anything. The idea that, well, bad actors are using this button to spread misinformation is not actually changing the behavior that existed before. It's making it slightly easier. But I'm not, yeah, I'm not sure that the retweet button has actually brought down our society all that much. Well, yeah, it's overstating it. And I get what Justin's saying where a lot of these protestations tend to feel like, I don't want you to hate me for being the engineer that worked on the thing you now hate. On the other hand, but also inherently give me credit for building a thing you love that you it was me. I do think though that it is legitimate and a good thing for an engineer to say, man, I wish I would have thought more about this. And now knowing what I know, let me come up with solution. And that's where I'll give weather all some points, where some other of these sorts of things I've read or just end up in. And so it needs to be shut down, essentially is kind of the implication where weather all says, no, we can't take it away. But there is a problem in that frictionlessness. In your example of retweeting Daily Tech News Show, with the retweet button, I do that all the time. Before the retweet button, I wouldn't have copy and pasted your post and put it in mine with an RT. I would have been like, well, if I'm going to do that, I might as well just write my own, right? It definitely has that effect of making you hit that button in a way you wouldn't have before. So whether else solutions that he proposes are if a group, not not an individual, but an audience he calls it, is regularly amplifying what he calls awful posts, devils in the details, how you to turn what an awful post is certainly. But if we all agreed on what an awful post was, and a group was retweeting that, and you could say like, okay, these 1000 people are all doing this, you suspend the ability of members of that audience to retweet, to slow it down to say like, okay, you're, you're breaking the rules by doing this. And so we're going to, we're going to pull that, but we're not pulling it from everybody. He also says maybe a better way where you don't have to determine content would just be the limit, the number of times any tweet can be retweeted, which is similar to what WhatsApp is doing with forwards. You can only forward a message five times. Well, but the thing, the thing with WhatsApp is those are private groups that you are then creating, you know, a mass hysteria. But often public groups, but, but yeah, yeah, but, but they are they are self-defined kind of groups in a world where you're only able to retweet something a certain amount of time, then the new measure of what is a really successful retweet will be how many versions of the same tweet you made. So you can retweet it, but that would slow it down because you'd have to make multiple versions. Yes. And yeah, it would, my point is this was a community behavior before they had the button. And so if we're saying that things got worse when you eliminated whatever that moment of time was between copying and writing RT before that tweet, then theoretically, it would be better if we could, we had to wait even longer. Maybe we have to wait an hour. Maybe we have to wait 12 hours, maybe a three day waiting period, like buying a gun before you, you okay now. Don't worry about it. I know. I'm just like, okay. So it's like conversation. It's the kind of thing like if somebody that you consider sort of like, ooh, social media villain, you might follow them on Twitter. Does that mean you like them? Not necessarily. We've all seen the bios that say retweets are not endorsements. Does that mean that every time you retweet something that someone said, it means that you're endorsing what they said? Well, no, but there's been a backlash from people saying, okay, well, you can't do that. You have to provide context or else people will think that you are endorsing it. And so I think that there's some ambiguity there that does need to be addressed. I think you may be discounting how much of an effect that delay has. I don't think it has to be days or hours, but the idea that I had to cut and paste kept me from retweeting things that just clicking a button kind of advocates my responsibility. And I think on a massive scale of hundreds of millions of users, that doesn't make a difference. So even if you just say you can't retweet unless you've clicked the link, or you say you can retweet this, but maybe it'll be delayed for three minutes and you have to confirm. I don't know. I don't know what the best answer is, but I think there's something to what he's saying. I don't know. Two-factor authentication, a seven-day wait period. That's what I'm running my campaign on. Excellent. Well, just be careful that you don't get your account suspended by Google. Everything will be fine. You know what won't be suspended? Our subreddit, you can submit stories and vote on others, whether they be retweet stories or other ones, anything that you want us to put our eyes on, dailytechnewshow.reddit.com. We're also on Facebook, facebook.com, slash groups, slash dailytechnewshow. Let's check out the mailbag. So you made it pointed out that while Netflix mobile only plan, which we talked about in India at 199 rupees per month is cheaper than Hotstar's premium, 299 rupees per month, you can pay for a year of Hotstar premium for 999 rupees, which actually makes that cheaper. And Netflix's full premium plan is 799 rupees per month. SumiDip adds, I've added myself. I have myself unsubscribed to Netflix because of its cost and even this 199 plan of rupees not tempting enough. India is a price sensitive market and for Netflix to be able to have deep in roads, they'd have to price it more competitively. In my opinion, 199 plan isn't going to make much of a difference. Also, Netflix has the disadvantage of fewer local content offerings as compared to Hotstar or even Amazon Prime Video. More new Bollywood movies make their way to Amazon Prime Video than to Netflix and they don't have the television content that Hotstar does. So the operative principle in the mind of the Indian consumer is summarized in the popular Hindi phrase, Sundar Sasta Tikau, which means attractive, cheap and durable. Yeah, I think it's legitimate to compare the 199 plan to the 299 Hotstar plan because people don't look at things that closely sometimes and will think, oh, I know Hotstar's 299, 199 is cheaper. That's attractive to me, right? That's cheap. I don't know if it's durable or not. But your point is well taken that when you look at it closer, Hotstar's plan is actually more comparable to Netflix's plan because it's not limited to mobile only. And Netflix's plan is a lot more expensive. So whether Netflix wins that marketing more or not is interesting. I think the content offering is the bigger hurdle that Netflix has to overcome. And they have done it before where they come into a market and just don't have the content to compete with the incumbent but end up getting that content eventually. It takes them a while but they get it. So that will be the thing to keep your eye on to see if Netflix is really going to survive or not. That said, I don't think it's a bad idea for them to have the 199 mobile plan. But I think it's well taken by Sumi Dip to point out like this isn't the silver bullet. They don't win just because of that. Thanks everybody who contributes to our mailbag every day. We love you. We also love Justin Robert Young who contributes to our show at least once a week sometimes if we're lucky even more. Justin, where can people find out what other stuff you're doing? Oh, well you can find me talking about politics at politics politics politics.com. It's also a little bit of a blog now these days. So if you want, I'm doing old fashioned writing with these fingers. I'm dusting them off and writing things on a fairly regular basis. So that's pretty fun. But go ahead and check out the podcast. Had a really fun interview actually this week with a night chair at the Annenberg School at the University of Southern California all about religion and politics and how that has evolved over the years and specifically how all the 2020 candidates are using their own religious backgrounds to shape some of their message. Fascinating no matter where on the spectrum you find it. So go ahead and check that out politics politics politics.com Hey, a big thank you to John F. Walzer and Alan Moore who both just started supporting Daily Tech News Show. Welcome into the tent. You're in the membership alongside years-long members like Alan Shepard and others. So I hope you enjoy the ad-free RSS feed, the special episodes. I've got one coming up this weekend sort of talking about why I'm always for but also against the thing you love. If that sounds interesting, that's my editor's best topic coming this Saturday. If you're a member at the $5 level and up at patreon.com slash DTNS. We've got an email address and that email address is feedback at dailytechnewshow.com. We are also live. If you can join us live, we'd love to have you Monday through Friday at 4 30 p.m. Eastern. That's 2030 UTC. Find out more at dailytechnewshow.com slash live. Back tomorrow with Aaron Carson from CNET. Talk to you then. This show is part of the Frogpants Network. Get more at frogpants.com.