 This 10th year of Daily Tech News Show is made possible by you! Imagine that we're all just giving you a nice big platonic comfy hug. Thanks to all of you, including John and Becky Johnston, Chris Benito, and Steve Aya Darula. Coming up on DTNS, Netflix is not what it seems, folks. Subscribers are up. Live streaming problems not as bad as you think. However, passwords sharing crackdowns are coming for us all. Plus an update on using AI to better target weeds. This is the Daily Tech News for Wednesday, April 19th, 2023 in Los Angeles. I'm Tom Merritt. And from Studio Redwood, I'm Sarah Lane. In Salt Lake City, I'm Scott Johnson. And I'm the show's producer, Roger Chen. Hey, Scott. Did you hear Google Fi is now Google Fi wireless? I'm shocked that Google would have a name change for no good reason. I try to keep it together for the rest of the show. I know. I know. It is a shock. Seriously, but there are other quick hits than these. Indeed there are. CNBC saw internal marketing documents that show Google plans to announce a foldable pixel phone at its Google I.O. developer conference, which is taking place on May 10th. Photos show a 5.8-inch front screen that unfolds into a 7.6-inch tablet. It will supposedly include Google's Tensor G2 system on a chip, be water-resistant, and claims to have the most durable hinge on a foldable with a cost of around $1,700. Windows central sources say Microsoft is preparing a new Surface Go, featuring an ARM processor later this year. The Surface Go 4 reportedly will be codenamed Tanta. Isn't that from one of those Santa Claus rank and best things, Tanta? Anyway, it will ship with a Snapdragon 7C-based system on a chip in the base model, along with some design changes, possibly 5G support for the first time. Microsoft might also ship a new 11-inch Surface Pro codename Luxor, which would go along with the current 13-inch model. Samsung began rolling out an update to the Galaxy Watch 5 and 5 Pro to automatically check skin temperature to support tracking and predict menstrual cycles. Samsung partnered with the Natural Cycles app on the feature, and all data is encrypted on the watch. Stability AI, the folks who make stable diffusion, the text image generator, have released a suite of text-generating models, not text image, just like chat GPT sort of stuff. The suite is called Stable LM. The models can generate both code and text, and you can get one and create your own, spin up your own chatbot. Just go to GitHub or Hugging Space. You can get them under an open-source license. Models were trained on a modification of the data set called The Pile. If you're familiar, that's made up of some open-text samples taken from places on the internet like Stack Exchange, Wikipedia, PubMed, and more. An alpha release from Stable AI was fine-tuned with open-source data sets from Anthropik. The European Union officially launched the European Center for Algorithmic Transparency, or ECAT, as part of its existing Joint Research Center in Seville, Spain. ECAT will serve as a dedicated research unit tasked with identifying smoking guns related to violations of the Digital Services Act, or DSA, which doesn't come into full force until early next year. However, the European Commission will very soon designate entities as very large online platforms and very large online search engines. You can probably tell who they're going to be targeting there, who must immediately start proactively assessing system risks potentially posed by their algorithms. Yeah, if you were wondering how they're going to enforce the DSA, there you go. That's how they're going to do it. ECATs. We've talked about tech for farm management before on DTNS. In fact, we've talked about this company I'm about to mention before. So it caught our eye to see Bloomberg's Coco Liu write that a Canadian startup has developed an AI-enabled crop-spraying drone that reduces chemical use by up to 90%. Precision AI, as the company we've mentioned before, uses drones to identify and kill weeds so that you don't have to spray the entire crop, saves you money because you're not spraying as much stuff, is nicer for the plants and for the edible things that you're getting from the plants because they're not covered in the chemicals as well, and it kills the weeds. Precision's model is trained on images of 15,000 plants, and that is then used to teach it to tell crops from weeds with 96% accuracy. Precision claims it can reduce herbicide use by 90%. Currently it's supervised by a human, but if the regulators ever allow it, precision says it can be fully autonomous. So when can a farmer get this, Sarah? Well, Precision AI expects to launch an on-demand service in 2024 in the U.S. where it has approval to fly. So kind of like Uber, you book the spraying when you need it. But if you want to run the whole thing yourself, they can also sell you the whole system and charge for software maintenance. Other similar efforts that you might have heard of include Israel's GreenEye, that upgrades existing sprayers to identify and target individual weeds. A similar system from Bosch Base-F Smart Farming is another one. And you think about this and you think, okay, well, herbicide's not always good for the planet. You know, you've got drones in the air, you've got less on the ground. Seems like a good example of using data in a constrained way and then training it for a pretty specific purpose, don't you think? I mean, yeah, like, you know what this makes me think of? It makes me think that why stop there? I mean, we're going to get into all the reasons why AI is great at things that we are not thinking of right now. But this made me think, well, what about wildfires? Got something crazy going on in the mountainside somewhere and it's just insane. We need accuracy. We don't need to just blanket everything with the same chemicals or the same fire retardants. We need to go in there and be really surgical about it to curb the growth of this particular blade. I guess if you're, I mean, I want to say, a lot of times you're thrown as much at a wildfire as possible because you need to, but maybe early on, if it catches it early on, I can see what you're talking about. It just makes me think of all the things we do where we kind of do a more blanket approach and less precision. And, you know, like things like, I don't know, very careful surgery. It's not a very good example, but surgeries need a very tightly knit close focus. There's a lot of things in the world that need that. And I think crop, you know, spraying crops and trying to reduce the amount of chemicals being used but also trying to help farmers be productive, all this sort of stuff all kind of works if you apply something like this to it. I know where you're going with that because it reduces waste, right? Like if you can be precise. Well, and I don't think any farmer says, you know what I love doing? Trying to kill those weeds that keep coming back to the crops every season. I mean, that is a perfect example of here's where, you know, we want, let's say it's a strawberry field, you know, I'm just using that as an example. We want the most strawberries possible. Right now, we're fighting, you know, various insects and weeds that end up crippling the crops, et cetera, et cetera. I mean, strawberries are like famously very herbicide sprayed because they're hard to keep going otherwise. And if you can do that and you can sell more of your fruit, I think you're a pretty happy farmer. Yeah. Well, and what I like about this story is it addresses, it's a good example to use when talking about the fears, right? One fear is that it's going to replace jobs. This doesn't replace jobs. Farmers do not have enough people to do this right now. And this is going to help them increase their crop yields because it's going to save the money that they can use for other things. The other fear is like, oh, it's, these things are full of bias and they're just going to be used irresponsibly. And yes, that is still a concern, but here's an example of why we shouldn't just stop developing them because this is a very good use that will not be misused. I mean, yes, you could probably misuse anything, but the intention of this use is not something that would easily be misused. It's, you know, I'm going to go across the farm, see all the good stuff and then see the weed and kill it. I guess it could be misused if, I don't know, your neighbor has just like a cool garden and you're like, I'm just going to screw with them and take down the road bushes. I guess you could retrain it. Yeah, try to retrain it or something. Yeah, I don't know. But it's kind of hard to figure out a misuse for it. So I think it's a good example of that. I just remember the last time we had a conversation with this was when everybody was like, can you believe they're putting chips in everything? Your coffee machine's got a chip in it. All these chips are in all these devices now that we usually think of sort of not smart devices, our refrigerator or refrigerators, our heating and cooling systems. So this feels a bit like that step in the sense that we're going to take a traditional method and we're going to add in this very smart technology and now we're going to make it more useful, more efficient, more all of those things. And to me, that's just, that is the ultimate goal with all of this. It's not to create a super AI that can be better than humans. The idea is to use these things as tools and make them work for us. And we just have to learn to think more complex about it. Yes, this also doesn't excuse bad behavior with deep learning and machine learning. We also need to guard against that. But sometimes I hear people say, well, why are we even developing this at all? And it's like, well, because there are good uses that like this one that can be very beneficial for humanity. Yeah, I'm into it. Good on your precision AI. I wouldn't mind some of this kind of technology in the home and I realize this is a far flung idea from that, but more precise vacuuming, more precise cleaning of certain hard to get spaces. I'm trying to think of something where I feel like it's wasteful. I mean, actually one of the things that is part of this trend of precision farming is sensor controlled irrigation. And imagine if you had good enough sensors that your lawn only got watered when the plants needed it. So part of your lawn that maybe a little that has thirstier plants get watered a little more often, other parts that can go a long time, the sensors like, yeah, you don't have to water that part, saves a lot of water, which is good for everybody. Yeah, the irrigation system at my house now, there are like five different stations, you know, and you can program everything, but it doesn't know if a plant needs water or not. It's just a timer. Yeah. It's just a schedule. It's not precise. Yeah. Well, farmers or others, please join our Discord by becoming a patron at patreon.com. They're having great conversations in there every day, kicking things around. I talk about when I go live on Twitch for office hours when I'm prepping for the show. There's a lot of fun to be had. We chat with you in there as well. So join that conversation in our Discord by linking your Patreon account at patreon.com slash DTNS. Four main stories came out of Netflix's Earnings Report Tuesday, which I'll summarize as subscriber growth, details on the crackdown on password sharing, good news for its ad-supported plans, and four, why the love is blindstream failed, and I don't think that matters as much as you think it does. Let's start with this subscriber growth, though. Up 1.75 million subscribers. That's up 4.9% on the year to 232.5 million. I feel like that bears repeating because I hear so many people just off the cuff, like, well, Netflix is on the decline, and Netflix is doing poorly. 232.5 million is not doing poorly, and that grew its lead on Disney+. Disney+, if you recall, in February, announced it fell a couple of million to 161.8 million subscribers. So again, despite the narrative you might hear, Netflix back to growing, Disney+, recently declining, they're actually both doing well. And Netflix grew in every market as well, except Latin America, where it was down by a half a million. Now that was not because of password crackdowns. Those launched in Q1 outside of Latin America, Canada, New Zealand, Portugal, and Spain. Alright, let's move on to that password sharing though. Yeah, so the quick explanation, if you don't already know, is that Netflix will try to detect if someone outside your household is regularly accessing your account, and then Netflix will ask them to pay an extra six or seven dollars per month. You might say, okay, well, how's that going to work? Well, Netflix promises that travel won't be an issue. If people want to just move to their own account, they can take their profiles with them. The company admits that launching the programs raises cancellations in the short term, but as borrowers start to activate their own accounts and existing members add extra member accounts, we see increased acquisition and revenue. The company says, this is what we wanted to do in the first place. So an example, the Canada paid member number is larger now than it was before password sharing crackdowns started. Revenue is growing faster there than in the US. So, yes, people do get upset about a crackdown on password sharing, of course, but no, it's not causing havoc while traveling, and no, people aren't really staying canceled. Scott, does this blow you away? It doesn't blow me away. What it says to me is that the company is addressing this stuff so that it, of course they're not going to say, well, we're seeing dips in some parts of the world, and yeah, we think that's a real bummer. It's not going to recover. Of course they're not going to say that. And I think they're probably right, by the way. I don't think they're actually making this up or anything. I think that is the trend. But we do have this weird tendency, and we'll get more into it here in a minute, but we have this tendency to just dog pile on Netflix. And we don't do the same thing with other streaming services, although I guess the Warner Discovery HBO mess has gotten close to that. But it feels like Netflix is just always in the target, probably because they're the biggest dog around and you got to compete with that and it's harder. But to me, this all sounds like business is normal. You've made some changes. How are those affecting the bottom line? Here's how we think it will. It doesn't feel like too much stuff that's on fire to me. It's also, it sounds to me like Netflix is a company that's confident about what you can watch on Netflix. If I have been, if Tom or you, Scott, have been sharing your password with me, for the last couple of years, and I love Netflix, and I've got my favorite shows, and all of a sudden that's not going to work out anymore, or it's going to cause a hassle or make us feel like we're doing something that's against their terms of service, I might say, well, that was fun, but I still want Netflix. So seven bucks a month isn't going to break the bank. Not everybody's going to feel that way, and yes, you have other options, but it's Netflix saying, listen, you know what you were doing, and here's where we're going to make it as easy as possible for you to just become a Netflix member. Yeah. I'm a little skeptical of Netflix's math here. I would like an independent analysis of this because my feeling is that the password sharing crackdown doesn't have any effect. Yes, it probably causes a few people to cancel out of anger, but it also, I don't think, really causes new signups. They're probably getting new signups from elsewhere unless they're saying, no, we're seeing the paid add-ons, right, because you can add someone to your account for an extra six or seven dollars. If they're seeing a lot of that happen, then okay, maybe, but my guess is most people that were password sharing aren't going to take that option, and the growth is probably just would have happened anyway. So I feel like we're going to have more growth from other things they're doing. In fact, Anacracketow was asking, hey, if Latin America is down, do they have the ad-supported plan? The cheaper plan? And the answer is no. They don't have it widespread down there. Let's talk about these last two bits. So we mentioned the first two bits of Netflix earnings. The other one is the ad-supported plan is working where it exists. It's making more money per subscriber than the basic ad-free plan. So the one that costs a little more but doesn't have ads doesn't make as much money per subscriber as the one that costs less but has ads. So Netflix is beefing that up. Starting in Canada and Spain, but then eventually coming to the 12 markets that do have the ad plan, the resolution will go from 720p to 1080p. So you're improving the resolution, and you'll be able to watch two simultaneous streams. It launched where you could only watch one at a time. So you could have two people in two parts of the house watching different things on Netflix on the same plan. Plus they say 95% of the content on the paid stream is now on the ad-supported stream. That was much lower when it launched, so they're getting closer to 100% as they work through those weird rights issues with the contracts they had. And they updated us on the Love is Blind livestream failure. That also came to light during this call. Here's what co-CEO Greg Peters had to say about that situation. It says we've got the infrastructure. We had a big bug that we introduced, actually, when we implemented some of the changes to try to improve live streaming performance after the last live broadcast, which was Chris Rock in March. We just didn't see this bug in internal testing because it only became apparent once we put multiple systems interacting with each other under the load of millions of people trying to watch Love is Blind. But that may not end up mattering, which is kind of the funny part. 6.5 million people watched it on demand. It's a number, but it's a number. And Peters revealed that 90% of people watch Chris Rock special on demand, not live. So I think a lot of people just still aren't used to this idea that Netflix has a thing that isn't on demand. Well, and the majority of people weren't even trying to watch the live stream. I mean, sure, they kind of brought down that whole operation, and there were certainly people who were very excited about the live stream happening on Sunday, which didn't happen, but it's like, I almost no, I don't want to say Netflix is just sort of like hot, hot, laughing behind the scenes, but it's like most of you didn't even watch the Chris Rock live stream. And I'm one of those people. I was excited about his special because he's a comedian and I wanted to watch it. I didn't even know it was streamed live. I just found it on Netflix once I started to see people talk about it online. But it wasn't about the fact that it was live. It was about the fact that there was a joke or two that either somebody liked or didn't like kind of thing. Well, that brings up the question and maybe one of you has an answer to this, but it feels like Netflix has a built-in marketing problem. They're Netflix and you go there for Netflix, but they're not very good at saying to you, by the way, we have these games. Oh, by the way, we have live streams and they happen. They're terrible at it. I'll stop you right there. They don't care. The games thing, they're still developing. Once they get those like AAA titles out, oh, you will hear about it. They are not newbies to marketing. They're not marketing them because they're not ready to push it yet. They're putting all their marketing against other things and I think the same is going for live. They did market the Chris Rock thing pretty well and to the people who cared, I feel like they probably marketed Love is Blind well enough because I did see people talking about it, but I think it was smart to do these live because they are things that more people would watch on demand anyway. You get the splash of saying we're doing something live, you get to test it out and when it did fail, it wasn't fatal. Trying to launch the DVD services Quikster probably did more damage to the brand than this Love is Blind failure. Yeah, it was a big deal on Twitter for a day. Nobody's really going to remember that and they have now figured out that configuration change that caused the problem when it got pushed out to the CDN see Monday's episode where we kind of called that ahead of time. I don't think this is bad for Netflix. What it does is helps them get ready and learn how to do live streaming, which is very hard in advance of the SAG Awards, which they're going to do next year live and people don't watch award shows on demand usually they watch them live if they watch them at all and any other stuff they might do like sports and news. Yeah, I think you hit the nail on the head. I am a Love is Blind watcher. I like the series for better or for worse. The live stream option had I known about it, I probably would have tuned in. You know, I just hadn't heard about it ahead of time, but so much of Netflix is just we're programmed, especially on reality shows to watch things on demand even if I had been like, okay, it's, you know, 5 p.m. Pacific time because I think that's when the live stream was set to go live 8 p.m. on the east coast and oh, it's not working. I'll watch Netflix and then come back, you know, they'll figure it out. That's what some people did and some people were just like, Netflix can't do anything right, but it doesn't, what does it matter? Well, not only does it matter, it doesn't feel like it matters much, but I think Tom's absolutely right and he made this comment about Netflix before when it came to games and stuff and I think he's right about their strategy, but also I just think people make a lot out of nothing when it comes to Netflix. Again, they're just the way they do it and I'm not, you know, a stand for Netflix or anything. I think they've made errors in the past, but I just think this all sounds kind of normal. You have a relatively new thing you're doing. It didn't work out great. Most people didn't even know you did it. Like it just doesn't feel like, it feels like we're all being a little, I said we all, the royal. A lot of people are. A lot of people are being really hard on the brand for nothing, right? And I feel like given time, this is Netflix is a bonus. This is how they do it. Well, and if you're one of those people who's like, well, it's not for nothing. I expected more. I pay for this. I want, you know, some sort of a refund. You ruined my livestream night. You know, there are some people who feel that way. That's real anger. I get that. I respect that. But, you know, I don't know. I got to give Netflix kudos for trying new things. Doesn't always work. We've seen that. We've talked about that for the last decade, bless. To be cold-hearted, the number of people who are angry about that is probably not enough to damage Netflix in the long run. And the learning they're going to get out of it is going to make it worth it, which doesn't help you feel better if your Sunday night was ruined. I get that. But, you know, yeah, maybe Netflix could issue some credits to folks, but I don't know how they determined who intended to watch Love is Blind live and who doesn't. Imagine you're like, I'm password sharing and you ruined my, oh, never mind. Oh, yeah, I hear that all the time, right? I'm going to cancel because you won't let me get a free, but yeah, okay, good luck. Well, we have some great gaming news if you live in a Petri dish. Yeah, so Australia's Cortical Labs uses synthetic biology and human neurons for its OI, meaning Organoid Intelligence System. You might say, what did you just say, Sarah? Okay, here's how it works. It grows neurons from human stem cells in a lab and then connects them to silicon chips for what it calls a biological intelligence operating system. Neurons have advantages over silicon alone because they self-program and they use small amounts of energy, smaller than, you know, what you would have from a machine. Cortical Labs published a paper in the scientific journal, Neuron, back in October demonstrating how neurons in a Petri dish learned to control the computer game Pong. The game itself is not super significant here. It's really just a demo of possibilities for healthcare because the neurons can learn and then do things. Specifically, could be used for testing drugs, for example. If you dripped a new drug onto the cells and those cells can't play Pong anymore, drug's not working. I don't know. That doesn't strike me as a very precise piece of information, but sure, yeah. The neurons are bad Pong. Crap. Okay. That one's not going to cure dementia. This is very early days for something like dementia, right? It's like, you know, let's figure out how stem cells, and this is a very personalized thing, right? My neurons are going to fire differently from anybody else's because we're all very different. And figure out, okay, well what might be going wrong with Sarah that then could be solved in some way because we understand how the neurons act differently with every individual. The Economist had a great article about medic-tailored medication in large and that is a very fertile area for research that this could tie into. Kudos to Cortical Labs for making it sound like they have more than they have at this point, I think, which is good for their fundraising, but that doesn't mean I think this is bereft of any value. I think it is. It's just like you said, Sarah, it's very early days of like, oh, we can do this and we think we can figure out other uses for it. It's quite figured out the other uses. It's not about how, you know, yeah, we get our stem-celled neurons to play video games, but they are smarter than we realize. This is what I love. I love that lately Pong has become this litmus test for how do you test your weird thing. It's doomed for neurons. It really is. It's that for neurons. Previous to this, somebody made chat beach GBT make a version of Pong in like 10 minutes or something and I don't know why, but suddenly Pong is like our baseline and I love it. It's a good touring test, yeah. All right, let's check out the mail bag. Let's do it. So we mentioned in yesterday's show that Netflix is going to shut down the DVD mail service in September and we kind of chuckled about who is even using that anymore. Comey wrote in and said, I have been using Netflix's disc rental since April 3rd, 2001. Comey writes, it will be sad to see the disc service gone. I think I still have a much better chance of finding a obscure title on Netflix than on streaming on Netflix, especially documentary and educational titles. Comey wants to know and let's throw it out to the audience here. Is there a website that looks into how many titles are available on Netflix disc rental versus streaming at the moment and also the number of subscribers for each? Comey said looked around could only find old articles. Yeah, the number of subscribers is far outpaced for streaming. Here is your talk. We mentioned earlier, 232 million. I don't know what the DVD streaming subscription numbers are, but if they were very big, they wouldn't be shutting this down. So I feel like that one would be interesting to know what the difference is, but it's going to be a big difference. Also, I can guess that the number of titles available for disc rental is much bigger than what you get on Netflix streaming because they can do any disc and they can't do everything on streaming. I would be curious though, like like Comey, if there's a good estimate of what that title difference is because while Comey is in a dwindling number of people that were using this, he's not wrong that you have a much bigger title catalog with Netflix disc rental than you do with Netflix streaming. Yeah, I found it, by the way, this is total side note. I've been cleaning out my studio and I found an old Netflix disc from like 2005. And you never mailed it back? I forgot to. I don't even know why I still have this thing. It's like, it's horrible. I don't know. Did you ever record it at lost? No, nothing. I just just gone. It's like you just canceled your DVD subscription and never looked back. It's like the first time I got jury duty. I never, I forgot. I just forgot and no one ever said anything. And now here I am. You should restart your subscription. Now you've said it out loud. Restart your subscription and send it in and see what happens. I haven't done physical library books for a number of years because I'm you know all into ebooks or audio books now but there are there's more than one book in my bookcase upstairs from the library in my hometown that I never returned and I mean I thought I would be arrested because once enough time went by, it's like now I'm too embarrassed to bring it back at all. You just get a collection statement. Yeah, this is going to send you a letter and go, I'm sorry man, but you always $14,000 for the last 10 years or whatever. No late fees with Netflix. That's on Scott's side there. Yeah, you've got till September to sign back up and return that DVD Scott. Yeah, do the right thing, Scott Johnson. Do the right thing. But when you're not finding Netflix DVDs in your basement what are you doing and where can people find out more? Well, I should be doing a Kickstarter right now but I'm waiting for him to approve it. So instead I'm going to tell people if you want some really good quality fun games coverage, video games that is, then check out our show on Thursdays called Core. It shows up as a podcast after of course, but if you want to watch it live we do that every Thursday at 5pm and there's a lot to talk about this week. So if you like video games, you like the industry around it and you like round table discussion, get over there and check it out. That's core wherever you get your podcast. Love it. Special thanks to David Snyder. David Snyder is one of our top lifetime supporters for DTNS. David, we want to salute you. Thank you for all the years of support. Yeah, you could be on your way to being David Snyder and helping us out across the years. We intend to be here for a long time but we need your help to do it. Support us at patreon.com. And when you do, you get the extended show Good Day Internet. We're going to be sharing our theories on what's going on with that AI track that sounds like Drake and the Weeknd. Because maybe it is Drake. It's Drake doing AI of himself. Is that possible? We're going to talk about that. You can catch the show live Monday through Friday at 4pm eastern 2100 UTC. Find out more at dailytechnewshow.com slash live. We are back doing it all again tomorrow with Megan Maroney joining us. Talk to you then. This show is part of the Frog Pants Network. Get more at frogpants.com Diamond Club hopes you have enjoyed this program.