 Coming up on DTNS the robots can't help Apple, but the AI may Facebook lets you push a button to send your photos over to Google and what tech companies will do about the pressure to change their business with law. This is the Daily Tech News for Thursday, June 4th, 2020 in Los Angeles. I'm Tom Merritt. And from Studio Redwood. I'm Sarah Lane from Oakland, California. I'm Justin Robert Young. And I'm the shows producer Roger Chang. Hey, we're just having a discussion about who we think the quintessential interviewer was. Maybe you have an opinion? Go listen to our expanded show, Good Day Internet, by becoming a member at patreon.com slash DTNS. Let's start with a few tech things you should know. Fitbit obtained emergency use authorization from the US FDA for its Fitbit Flow Emergency Ventilator. The lower cost design is based on the MIT event system and is the automated resuscitator style design replicating the function of manual resuscitation bags used by EMTs. With the FDA authorization Fitbit can now use its existing manufacturing partners to build these in large volumes. Messaging app signal announced its next update on iOS and Android will offer a dedicated face blurring option in its image editor. The feature can automatically scan for faces using an on-device system library and users can also trace over faces if it fails to detect one. Signal says all image processing will be done on device. Huawei launched the 6.81 inch Honor Play 4 5G. A lot of numbers there, but it's the Play 4. The Honor Play 4 and this one is 5G. Also a 6.51 inch model called the Play 4 Pro 5G. These are launched in China. A version of the Play 4 Pro will be offered with an infrared temperature sensor. The Play 4 5G costs $1,799 yuan, that's about $150 US, and the Play 4 Pro costs $2,899 yuan, that's about $400 US. And the version with the temperature checking costs you an extra hundred yuan. Some musical chairs going on at Google. Google announced Ben Gomez will leave his role as head of search to go to oversee Google for Education, Google Scholar and Education Search, as well as Google Arts and Culture. Gomez will remain a technical advisor to search, however. Prabhakar Raghavan will replace Gomez as the head of search and assistant, which encompasses news, discover, podcasts, and also Google assistant. Jen Fitzpatrick will now run Google's central engineering team. That's not a hard job, which has 8,000 employees and is responsible for the company's core technical systems and infrastructure. Corporate IT, UX, design, Google accounts, and privacy. But that's it? Yeah. Yeah, it'll be fine. Small little job. Facebook is rolling out labels that identify media that is either state-owned or state-controlled. Those outlets will no longer be allowed to buy advertisements on Facebook in the US starting this summer, and advertisements they run outside of the US will contain labels as well. Facebook used outside experts to classify the state-controlled media based on multiple factors, and outlets can appeal that label if they feel it was applied by a mistake. Alright, let's talk a little bit about all the streaming. People are starting to come out of their houses again, Justin. Are they still watching the streaming? Oh, heck yeah, Tom. Nielsen estimates that streaming is staying higher than it was pre-lockdown despite lifting of restrictions across the country. Connected TV includes game consoles, smokers, and the like. March 2, Nielsen estimated 2.7 billion hours of viewing on connected devices. That peaked at 3.9 billion for two weeks from March 23 to April 5. As lockdowns were eased, the hours fell, but not by much. The week ending May 10 still had 3.5 billion hours viewed. Nielsen said traditional live TV usage in early May was only marginally higher than 2019. Yeah, it's still probably too early to draw too many conclusions. That 3.5 billion might settle back down closer to 2.7 billion by the time everybody's really back to going outside all the time. But it does show that people turned to streaming more than traditional and have kept using it. So I mean, I guess you could conclude if this holds that we have accelerated the adoption of streaming among a large number of people. Well, and as far as just sort of like, okay, people have been forced to adopt different, you know, I don't know, entertainment strategies and all sorts of other things. You know, when the whole sort of like, this is my routine got torn down and a few months ago, you know, at the beginning of March, you know, what's stuck? Certain things will continue to stick even after life goes back to quote normal, whatever that means. And I think for a lot of people, it's like, I like that show. I'm going to keep watching that show where I'm going to, you know, I'm going to use the streaming service that maybe I hadn't used before. And I had, you know, taken advantage of because I was stuck inside, you know, some of that stuff doesn't just go away because because life is normal again. Yeah, and not to mention, these streaming services are optimized for retention in a way that traditional television is not. Former Foxconn executive David Bourne tells the information that back in 2012, Apple executives, including Tim Cook, were reportedly in China to see the results of an experimental iPad production line that wanted to replace humans with robots. Terry Guo, Foxconn's chairman at the time, reportedly told Apple that Foxconn's assembly lines would contain one million robots within two years. Well, 2014 came along and that didn't happen. But last year, Foxconn was using 100,000 robots across all of its manufacturing and the information sources say that's based on dissatisfaction of the finished product of the robots by Apple. Apple launched its own stealth robotics lab in 2012 to try this out themselves, which housed a team of automation specialists and robotics engineers who initially tried mimicking that Foxconn iPad automatic production line. Apple reportedly wanted to cut 15,000 workers from the production line, but machinery couldn't handle Apple's precision methods. Apple also tried an automated production line for the 2015 MacBook, which was not only abandoned, but that MacBook itself was postponed by some months because the experiment didn't work. The information article also goes on to talk about how the adaptability of human run production lines is better because humans can be just told to do something a different way. Whereas machines have to be reprogrammed at this point. So if you want to change something midstream or increase capacity to lines that didn't have the robots, it's not as flexible as the humans. Not that the world needs any darker thought that isn't already populated, but the first thing I thought of was specifically with Foxconn, if the idea was to move to automated lines to remember their suicide epidemic that happened. And indeed, that was through 2010. So if they were introducing and touring Tim Cook through that, I wonder whether or not there was an element of like, hey, we know that this has become a scandal worldwide. It certainly was here in the United States, the uncomfortable nature of how happy these Foxconn employees were. Maybe this is a way that we can go forward and continue to keep your business. I wonder whether or not that had anything to do with it. You know, also, just the idea of the, okay, so there was an experimental, let's see how iPads get assembled, you know, using robots instead of humans. Come on over Tim Cook and various Apple executives at the time. And this is a few years ago, back in 2012, so eight years ago now. And, you know, the Apple kind of going like, not precise enough, you know, there's, you know, like, I would, I would love to know exactly. I understand that the, the concept of needing human hands to be more precise when you're talking about tiny little parts of, you know, of very expensive gadgets. But, but the fact that it didn't really work out that well for Apple, you know, Apple kind of going, well, maybe we can do this ourselves and it'll be closer to home. You know, that the, the experimental lab they had was reportedly just a few miles from Apple's main campus in Silicon Valley. And that didn't really work out either, to the point where we're still, we're still kind of using the assembly line with all of its problems, as you mentioned, Justin, as we have for many decades. Yeah, this reminds me of 1830 and the, the race between the Tom Thumb steam engine and the horse drawn rail cart over a 13 mile stretch of track. The horse was supposed to lose. This was going to demonstrate the superiority of the steam engine, but the horse kept taking an early lead. It could adapt to the steam engine. Steam engine could just hit its top speed and go. The horse could like give itself a rest. And when the steam engine caught up, it could put out a burst of speed and eventually Tom Thumb, the steam engine broke down and the horse won. But less than a year after that race on July 31st, 1831, all horse drawn carts were replaced by steam engines on the rail line, BNO, that sponsored this race. So I just throw that out there as a sort of like, yeah, stuff doesn't work until it works. And this is how you get there. You do a bunch of experiments that don't work. And then you go back to the drawing board and you try again. So I think this, I would look at this story and say, huh, we're probably getting closer to robots being able to do this because we've figured out where they fail and people are working on fixing that problem. Well, sort of pretty related, but sort of on the flip side of this whole story, a California based startup called Instrumental Inc., perhaps you've heard of it, perhaps you haven't created by formal Apple engineers, has created a system that could help current Apple engineers during the coronavirus crisis, which has been rumored to be contributing to the delay of the upcoming iPhone 12, possibly anyway, particularly because travel to China has been restricted. Engineers who have to go and make sure everything's going well and make sure everything's up to speed can't go. They can't do it. They can't do their jobs. Instrumental system equips manufacturing lines with cameras and then analyzes those images using AI software so that the engineers can have more flexibility and not always have to be at any physical location to make sure that things are running smoothly. Cameras and artificial intelligence are designed to spot things like missing screws or bent springs or damaged batteries all in real time. So yeah, this is an example of something that people, you know, the idea of putting cameras on the assembly line so you could look and see what was happening. That's not new. But this company finally made it work. This is the one that said, oh, we can make this so that it is a replacement for you having to be here in person. To the to the point that Motorola and FLIR were using this before the virus to monitor lines like this isn't like, hey, maybe it'll work now. This is again kind of like the streaming thing and acceleration of an existing trend. I was once on a flight with one of these Apple quality assurance engineers and she told me a great story about how she was very, very aware of her favorite hotel bar in Shanghai and how she would take her little day trip and tear down a few products that she was set there to tear down and make sure everything was okay and then make sure she got right back to that hotel bar because she was going to be on a flight out as soon as her tour of duty had ended. And that could last as long as a few months. Sometimes she was there for a significant period of time, nor at least at this retelling on this flight was she particularly happy about it. So I could imagine that this is something that Apple could use. And you're right, Tom, this is just forcing the issue like now. Okay, now you can't go to China, even if it's just, you know, a company that has infinite money, and it's just, all right, go literally put this woman in a hotel so she can drink and look at tear torn down iPads for three months. Now you can't do that. What's the best way to do it? And this seems like a an effective and more drill down version of quality control. Well, and she was probably really good at her job. She was like you said, you know, you're over there for months at a time sometimes I mean that's that is extremely disruptive to anybody's life. You know, maybe some people enjoy it probably a lot of people don't but who else is going to do it. This is your job. So, to be an engineer who has a lot of skills that can be put to different uses. Because some of that just kind of like, how's it going on the assembly line type stuff can be sent to you remotely just it seems like such a no brainer. I know there's more to it than just like why didn't I think of it before. Obviously, people have but but it's a good solution. Yeah, because before the angles weren't right. You didn't see the things you needed the system couldn't spot the stuff you wanted to and it looks like maybe this instrumental ink has figured it out. The Facebook photo transfer tool that launched in Ireland at the end of 2019 is now available globally. The feature lets users directly copy Facebook photos to Google photos with an encrypted transfer rather than having to download and re upload them yourself. The option transfer a copy of your photos and videos is found in the settings menu under your Facebook information. Facebook support for other services beyond Google photos is indeed being built now. The feature is part of the open source data transfer project. Yeah, and that's that's the big thing you should be aware of if you weren't. I mean, we've had this in the US, Canada, Europe for a while. I think it came along around April earlier this year. But now it's everywhere. It's super simple. I did it. You press a button, it tells you okay, it's in process and your photos show up over there. It's so much easier than having to download. Where is it? What folder? And the data transfer project is an effort by these companies to stay ahead of antitrust legislation, at least in one sector that might have forced them to do stuff like this and say, hey, we're making it open so you can take your data with you. Granted, right now it's just Facebook and Google. It's just photos. But this is the beginning of the ability to move your data around between services. And that's something I want to keep an eye on. Yeah, you know, I'm curious to see obviously this happened because Facebook was under pressure, but I think it's a fairly, I mean, it obviously took a lot to build, but it's a low calorie give for them. And it shows you the evolving market of what these walled gardens really required to thrive. Google's threat analysis group announced it had recently identified phishing attacks against personal email accounts from staffers on both President Trump's and Vice President Biden's campaigns. Google sent a warning to the targeted campaign staffers on both campaigns, as well as alerting federal law enforcement officials about the attempted attack. The attacks against the Biden campaign staffers appear to have originated in China and the attacks against the Trump campaign staffers appear to have originated in Iran. Those are best estimates. It's easy for someone in another part of the world to make it appear like their attacks coming from somewhere else. But the threat analysis group is pretty good at those kinds of forensics. What's the significance of this host of politics, politics, politics? Well, it certainly was something that became a major issue after the DNC got hacked and the Hillary Clinton campaign got hacked specifically after one of the higher ups there, John Podesta, clicked on a spear phishing link. So if this is something that was being attempted on both the Biden and the Trump campaigns, it is certainly no surprise. To me, this is the new normal. Everybody has to deal with spear phishing attacks. You can only imagine that the prize is that much richer when you are looking at a heightened media environment. And remember how much of a chaos bomb those Podesta emails were and those DNC emails were the DNC emails were released the day or sorry, two days before the Democratic National Convention last year. So you could only imagine even if these were successful, I mean, these have been detected. If they are successful from somebody else that clicked on some of these links, then you never know when these kind of treasure troves of information will pop up. Yeah. All right, folks, if you want to get all the tech headlines each day in about five minutes, be sure to subscribe to DailyTechHeadlines.com. Tech companies are under fire for the tech they sell to law enforcement right now. I don't think you need me to explain to you why there's a lot of opinions about law enforcement right now. Amazon sells hundreds of law enforcement agencies a facial recognition tool called recognition with a K. We've talked about that a lot on DTNS. Among the interesting things about recognition, AI researchers at the University of Toronto and MIT did a study that found that recognition misidentified people with darker skin more often than those with lighter skin, particularly females, but also males. Amazon says that that research is flawed. They denied that claim, said you don't have the most current version of the algorithm. Amazon has also been criticized for its Ring Doorbell video sharing partnerships with police departments. They've also modified those to change it to be under the control of the ring person very clearly whether you would hand over video to the police department rather than just on an ongoing basis being able to improve it on a case-by-case basis. That's not just Amazon. Google has been criticized for working with geofencing warrants. These are warrants that permit the police to request anonymous information for phones in an area surrounding a crime. So you say, give me this boundary line, any information, anonymous information on activity there. And then if the police with that information can narrow down to suspects, they can get a warrant to get user names and location data for specific devices. Geofencing warrants are criticized for sometimes being overly broad in the area allowed. Then again, Google says we just comply with warrants. So if there's a problem with geofencing warrants, then we should work on reforming geofencing warrants. Google also pulled out of a contract to help the Department of Defense develop machine vision technology for drones. So Google has bowed to pressure in the past. Who else? Microsoft partners with the New York Police Department on a system that aggregates data from a network of cameras delivering license plates and other database-driven devices. IBM sells machine learning tools for predictive policing in an attempt to predict future crime spots and sometimes actual future criminals. But again, what kind of biases are in these data sets that these machine learning algorithms are trained on? IBM has also partnered with the NYPD to use images of New Yorkers to train AI tools. Well, not necessarily getting the permission of everybody in those images. And Nextdoor has been criticized for tolerating racial profiling by its users. Nextdoor has taken action, told Wired that it has seen profiling decline since it began requiring more detailed information before reporting someone as suspicious. So if you just try to type something on Nextdoor, sometimes it'll pop up and be like, hey, it sounds like you're trying to say this. Have you thought about that? Can you add these more details? There's also problems with Salesforce and other companies out there. Microsoft and Amazon are sponsoring the International Association of Chiefs of Police Conference in New Orleans later this year. Microsoft also has been an advocate for having legislation about facial recognition and how it should be used. They have leverage. They have lobbying. They have relationships. Should they all just stop selling this technology? Jeffrey Albert, a criminology professor at the University of South Carolina who serves as a member of the International Association of Chiefs of Police Research Council, told Protocol that if these companies step back from providing technology, law enforcement agencies would find other vendors. He said, quote, someone's going to sell it to the police to make money. So it might be principled for Microsoft or Google or Amazon to stop providing their technology, but that would not necessarily stop law enforcement from using that technology because they might be able to find and probably would be able to find another vendor. So the question is, what do you want them to do? You know, I think the biggest thing here is transparency and specifically pushing for that when you are dealing with law enforcement or governmental contracts. Let's also remember that Google sold Boston Dynamics and part of at least the rumored reason why was because Boston Dynamics was inherently going to move more into a defense contractor model. And Google did not or Alphabet did not want that to be necessarily such a gigantic part of their portfolio. The reason why people do this is because the money is big and it's good and it never goes away. You can you can continue to cash that check if a system is built on. And this is government in general law enforcement is no exception. Do not change systems all that often. They don't nimbly go from one thing to another. So if you can get a law enforcement contract that is a lot of long term money and you know specifically looking at the people who are sponsoring the association of the chief of police chiefs of police Microsoft and Amazon both sell AWS and Azure. And they want people to be on those platforms and they build products specifically for law enforcement so they can secure those contracts including some of the AI stuff. So all I would ask and all I've ever asked and by the way this is only going to become more of a problem up front as up front and detailed as you possibly can within the agreements and should the agreements become too opaque. Then I do believe it is our consumer right to hold them accountable for that. Well and in a situation like recognition and Amazon saying well our software is not recognizing certain people over other people. That's just not happening. But there's a study that says it is happening. It's on Amazon to say OK well maybe the people who are making the software you know we have to rethink that. Maybe we need you know another group of folks who are making sure that the people making the software are not introducing biases that are going to become problematic later on. I don't think a lot of this technology is inherently bad. I mean it's all about you know get the bad guys protect the good people. But it is rife with issues and so it kind of just falls on the people at the end of the day of like who's making this why are they making this and what are they thinking. But Amazon would answer you Sarah to say no we've done all that. It's these researchers we're looking at an old version of the algorithm. Well OK I mean that's so so I mean but but instead OK I think I think my point is instead of Amazon being like well that's wrong. We are not in a day and age right now of people being like well Amazon just said that they the researchers are stupid so we'll just we'll just carry on with life. It's it's more of like no here's why you might have thought this here's why we really care that that wouldn't happen. And here's what we're doing to make sure that that doesn't happen in the future. I mean one answer to this could be don't worry about it companies sell your products law enforcement needs them and and and don't listen to the folks complaining. I don't know that that's a tenable situation for companies to do right now. They are on immense pressure inside and out to take this seriously and pay attention so they will possibly lose customers and employees. You know we're seeing petitions and walkouts. So companies are probably wise to pay attention to this and not just ignore it. The other side would be to say just stop selling all together. Don't sell the law enforcement. But again it's not a flimsy excuse to say well but does that solve the problem if someone else just steps in and sells them. Is it better if these companies take their privileged position and use it to have better rules. That's what I would like to see is Microsoft and Amazon and Google and Salesforce saying you know what we're going to improve. Amazon is going to say we're not just going to tell you it's better. We're going to do what Justin said we're going to affect transparency. We're going to tell you how it's better. We're going to admit where it's not better. We want to push for legislation that will put limits on how it can be used which is what Microsoft is doing. It's pushing for legislation because private companies just not selling isn't really an effective tool in a marketplace as open as this where law enforcement can just go buy it somewhere else. You you need to come up with sensible rules that say hey the technology isn't good enough to do these things yet but it is good enough to do these things. Let's come to a reasonable accommodation that helps law enforcement do their jobs properly and accountability. And look they're going to get they're going to get a vendor and that vendor if Amazon came out and said we are absolutely not doing this. Then the team that would develop it could spin out create their own company hook into AWS because they already know exactly how to take best advantage of it. And AWS would still get the money for whatever their server costs but it would be another company that would come out and do it. Like this is the genies out of the bottle on this particular issue. It's just a matter of how you can do it the most effectively and the most ethically. And I think a really good a really good statement from Thrumwalled in our chat is this tech is so new. How do we even know all the things that could be wrong with it and that's important to take into account which is there are things you know it can do. And those are the things you should allow there are things that it might be able to do and those are the things you have to be really careful about. You need to try them and test them responsibly to improve them. And then there's things you know aren't ready yet and those should be kept in the lab and not used in public. Well thanks everybody who participates in our subreddit. A lot of security stories end up there every day and you help us learn what matters to you submit your own stories. Vote on others at dailytechnewshow.reddit.com. All right let's check out the mailbag. Oh let's so on GDI yesterday we were talking about this this sort of like fun fantasy idea that some of us have on the show of having like a coffee shop but also where the podcast happens. So it's like it's part you know place to go hang and get your coffee or your wine or a sandwich or something and also partly a place to do your show. And somebody wrote in and unfortunately I did not get his name. But he said we had something like that in Pittsburgh in fact real real place. It was a restaurant that served waffles but was also a talk show. It's called the waffle shop. That's what we're going to do. Pack your bags everybody we're moving into the waffle shop was Brian. Thank you Brian. Yes the waffle shop a reality show was a performance art project and restaurant in the East Liberty neighborhood of Pittsburgh operated by Carnegie Mellon University students. That's awesome. That's so great. That is now the long term plan of the Daily Tech Show. Right. Yeah it turns out we all have to live close by but we can hang out in our coffee shop. Like everybody on France. It's going to be great. Hey shout out to patrons at our master and grand master levels including Reed Fishler Paul Reese and Mark Gibson. Also thanks to Justin Robert Young host of politics politics politics what's been going on in your world nothing I assume. Oh man you know a very serious episode yesterday kind of went out of our format and did kind of more of a breakdown of our relationship with the police specifically in a reaction to everything that's going on right now in the murder of George Floyd. But we are back to our regular format tomorrow we got a great interview about the the epidemics of 1957 and 1968 why they're different. And we got some great little tidbits in there including where social distancing our modern idea of social distancing came into the public health playbook. Hey thank you everybody for continuing to stick by us and support us. Patreon it means the world. I put up an editor's desk today for everyone. Usually the editor's desk just goes to folks at the associate producer level and above. But this one went out to the whole batch if you're wondering why you got it. It's because it's me talking about some of my feelings and the situations around that Monday show that we did this week. And thanks again to those guys and lady who participated in that show as well. And thank you all for being willing to listen and you can continue to support us daily tech news show dot com slash Patreon. You can also continue to give us feedback which we love our email addresses feedback at daily tech news show dot com. We're also live Monday through Friday for 30 p.m. Eastern 2030 UTC and you can find out more at daily tech news show dot com slash live back tomorrow where Patrick Norton will tell us where to put our speakers and Len Peralta will draw.