 Coming up on DTNS, Nate Langston has a new show on Bloomberg about AI, so he'll help us understand why everybody's so scared of it. Plus, the UK puts the brakes on Microsoft's Activision Blizzard acquisition, and password sharing seems to have reduced Netflix subscriptions in outer Spain. This is the Daily Tech News for Wednesday, April 26th, 2023 in Los Angeles. I'm Tom Merritt. And from Studio Redwood, I'm Sarah Lane. I'm the show's producer, Roger Chang. And joining us, Bloomberg's tech editor and host of the text message podcast, Nate Langston. Welcome back, Nate. Thank you for having me. Congratulations on the launch of the new show, by the way. Oh, thank you very much. AI-IRL has got one episode out, right? Yes, as we speak, but in about an hour and a half, it will be two. So by the time a lot of people hear this, there will be two. He will have doubled the amount of episodes. In mere seconds in your experience. Well, cool. I am looking forward to talking a little more about that later in the show. Let's start here with the quick hits. Open AI applied for a trademark on the term GPT with the US Patent and Trademark Office, or USPTO, back in December. In March, the company filed a petition to speed up the process, noting that myriad infringements and counterfeit apps were using that same term. Now the USPTO has finally responded. It dismissed the petition because Open AI's attorneys failed to pay all fees and provide all the required evidence. Good news. They can start the filing over again. Bad news. Sources tell TechCrunch the trademark process could take five or more months again. Do you think they'll tell them if they paid the fees wrong this time a little faster? Seems like it took them a while to notice that. It's been nine years since Lenovo bought Motorola, but we finally have the first Lenovo Motorola co-branded device. The Lenovo ThinkPhone by Motorola, kind of like the Los Angeles Angels of Anaheim. It's available now for enterprise customers in the US for $699, an unlocked version come in April 28th. It's ThinkPad in the back, Motorola on the inside, I guess. It's got a ThinkPad-like textured fiber back panel, a customizable red side button like the old ThinkPad nub, a 1080p OLED 14, or I'm sorry 144Hz screen, and is Mil-Standard 810H compliant in case you drop it down a staircase. Alphabet reported revenue grew 3% in Q1s that beat analyst estimates. Google ad revenue fell for the second quarter though. YouTube ad revenue fell 2.6%, the third quarter of a decline, but that was slightly less than expected. CEO Sundar Prachai said shorts views increased 66% on the year to $50 billion daily. Revenue for other bets, that's the non-Google businesses like Waymo and Verily, fell 35%. Google Cloud generated its first-ever profit after posting a $706 million loss this time last year. That's actually probably the best news out of the whole bunch for Google. Microsoft reported that revenue grew 6.7% for it in Q4 on the year, beating estimates. Intelligent cloud revenue rose 16% on the year. Within that division is Azure. Azure grew its revenue 27%. CEO Satya Nadella said teams grew its monthly active users 7% on the quarter to more than 300 million. Bing grew a little bit too. It now has more than 100 million daily active users, but that didn't stop the unit it's in. Microsoft's more personal computing unit from losing revenue 9% on the year. That is due in part to a 28% fall in Windows license revenue. All right, so mixed earnings for Microsoft. In other Microsoft good news, the company started rolling out phone link for iOS on Windows 11, meaning you can make iPhone calls and texts using your Windows machine. But some other Microsoft news today, not as good. A Reddit user who goes by hacker mchackface discovered that Microsoft's edge browser appears to send URLs entered into the address bar to bingapis.com. Software engineer Raphael Rivera told the Verge it appears to be due to a creator follow feature in edge, the idea being to let Bing know when you're on a page from a creator like their YouTube channel say, but instead this seems to send almost every URL and is also on by default. Disabling the feature stops URLs from being sent and Microsoft told the Verge it is investigating this issue. You know, I hate to correct you because 99% of that was right, but I believe it's pronounced hacker mchackface. Right, right. Yeah. Well, you know, you got to love a handle that keeps people guessing. Also folks in the chat room are having fun saying whether the phone should have been called Lenovo Rola or Moto Novo. Those are those are both interesting takes on the the new Lenovo Motorola phone. Let's start off talking about Microsoft. We haven't talked about them in seconds. Yeah, Microsoft executed an all out charm offensive to convince the UK competition and markets authority or CMA that its proposed acquisition of Activision Blizzard would not harmfully affect competition. Microsoft said, this is going to be great. They signed up every platform they could think of to guarantee access to call of duty for 10 years on their platforms and it worked. The UK CMA doesn't think that the acquisition would harm competition for console platforms, but there is more to the story. Yeah, not for console platforms, but the CMA does think the acquisition would harm competition for growing business of cloud gaming services, which is odd because Microsoft amongst all those 10 year deals signed one with Nvidia in February that has already begun. A lot of these deals wouldn't kick in until Call of Duty became Microsoft's. This one has already begun and guarantees that Nvidia's cloud service GeForce Now will get access to all Xbox PC games. Now, in batches slowly being rolled out, but it's something Nvidia has not had since Microsoft pulled access in 2020. Even so, the CMA still believed too much possibility of Microsoft making games that are exclusive to Xbox Game Pass and that possibility could undermine innovation. Microsoft told the CMA it would agree to requirements on what games must be offered to which platforms, but the CMA rejected the proposals as too narrow for a 10 year agreement in a rapidly changing market. The fight is probably not over. Microsoft Vice Chair and President Brad Smith said that the company will appeal the decision to the UK competition appeals tribunal, but Nate, do we think there is hope here or does this feel kind of like the deal's dead? I do not think the deal is dead. I think there will be an appeal. I think it will take quite a considerable amount of time. I don't think we should rule out the possibility that the regulators in charge of this deal have sons and daughters who are worried that their favorite games are going to disappear from their favorite console. And there's been some dinner time conversations that have fed into this. But I think it is a very strange decision because of the cloud gaming focus here. This is a nascent market. We have seen Google Stadia die a death. We have seen Amazon Luna with question marks around it in terms of success. Who are the other big players in this sense? There really aren't that many. And then when I think it threw like that, I think, well, Microsoft would have become the big one. Microsoft would be the one to beat. And maybe the argument is we stop this now before we get 10 years down the line and it is an unbeatable behemoth that no one could compete with, which is also something I disagree with because it's basically like suggesting you can't compete with Netflix, which is very possible to be competed with. So I think there are many counter-arguments here. I think we will see them. I don't think this is a dead deal. But I'm not quite sure what concessions Microsoft can offer that it hasn't already offered. Who knows? That's where I'm at. I think you're absolutely right that this is going to take a while longer to resolve that Microsoft has said they're going to appeal. So this fight is definitely not over. But Microsoft has given Nvidia a 10-year guarantee on all its games. Microsoft is willing to guarantee other platforms to get the steal done. And yet neither of those things is good enough for the CMA. I don't know what else they could do. To use your Netflix analogy, it would be as if Netflix owned Paramount and MGM, which actually Amazon owns, and then was going to try to buy, I don't know, Warner and said, well, we'll guarantee that we'll keep shows available to other platforms, which is not how that system works at all. So I think that's what the CMA is worried is whatever you say now, within 10 years, this is all going to break down into exclusivities. We've seen that happen with streaming. And these are franchises that have proven they have longevity. Call of Duty is much older than 10 years time. And there's no expectation that it's going to go away in 10 more years. So in that sense, I can see the argument that you're thinking further ahead, but it's still a stupid argument when you're focusing only on cloud gaming today. I mean, it sounds like the CMA is saying things have changed with cloud gaming a lot in the last few years. So 10 years from now, this rapidly changing market is going to be a lot more crowded. Microsoft is setting itself up to say, we're going to do these things, but we won't necessarily have to do the other things later. But the other things aren't identified. They're unknown. Nobody knows what they are. And to that way of thinking, I think the CMA is rationally saying, I'm curious why they're fine with the console version of this, because maybe they just think console is on its way out. But they're saying, look, this market is going to be fundamentally different in 10 years. We cannot possibly legislate a thing that will keep the competition flowing except to stop Microsoft from getting Activision Blizzard. That's the only way we can be sure that they won't dominate the games coming out of there. I wonder if one of the ways we'll see this resolved is if Microsoft gets very clever and starts picking through the wording and digging in the weeds and finding some kind of failure of procedure to say, well, you got that wrong. Your decision is fundamentally flawed because you got this thing wrong. I don't know what that thing is right now, but that is a way I could see them wriggling through this and basically suggesting that this decision is just completely irrational. Well, we will have to wait to find out if they can convince a judge of something like that. But let's turn to a more mature market. We were just referring to a streaming video during its earnings call last week. Netflix announced that it did see temporary drops in subscribers when it launches a crackdown on passwords in test markets. But so far over time, subscriptions come back and revenue rises. During the earnings call, you may remember, we talked about this, they held up Canada as an example. Canada's paid membership base is now larger than it was prior to the launch of so-called page sharing and revenue in Canada is growing faster than it is in the US, which does not yet have the password crackdown. But a study from Kantar of Netflix subscribers in Q1 found that about one million users in Spain did need leave Netflix likely as a result of the crackdown. Keep in mind, though, that's users. It's not paid account. So about two thirds of those users were using someone else's account. So Netflix probably accounted for that, were expected it anyway. So they left and they decided we're not paying to keep our profile, at least not right now, but a good 30,000, 300,000. Yes, a little bit different. We're actually paying and they also canceled. Some of that might be unrelated to password sharing, of course, maybe they just didn't want to pay for Netflix anymore. Obviously there's competition, but it is a number to keep in mind. Competition's rising. Amazon and Sky Showtime added subscriptions during the quarter as well and some people just cut back. I'm one of them. The number of households in Spain that had at least one streaming subscription fell in Q1 by 292,000. Which is ironically very close to the 300,000 plus they Kantar found had stopped their paying accounts. Yeah, the two thirds of users that weren't paying anyway, Netflix doesn't lose anything by them stopping using, but the ones who were paying were. Nate, what's your sense around this? Is the Canada story or the Spain story feel more like reality? I think they both sound pretty real, to be honest. I wouldn't know how much local competition affects these sorts of numbers because Canada and Spain both have very different streaming markets. I believe Canada might have more competition. But then Spain has probably a greater number of Spanish language shows than perhaps Netflix does. I think there may be comparisons there. The other comparison could be economic. Canada's having a rather stable economic situation relative to the rest of the world and Spain is not. That puts pressure on people. That's why you see households in Spain canceling all streaming subscriptions and larger numbers. Yeah, that makes more sense than just saying, well, probably people in Spain are sharing passwords more than anywhere else in the world. I guess there's probably some regional waves where something like that becomes an opportunity to not pay a few bucks a month for something that you want. But yeah, it's probably economic more than anything. I feel like this is all just really good data for Netflix. Netflix can say, all right, well, here's where we've done the right thing, at least short term. And here's where we've seen a dip. And in six months to a year, let's look at Spain again and see where has it bounced back? Did it flatten out? Is the still an issue? Is it really the library that's available there that is not serving our customers to the point where they go, I was kind of done anyway. And now I'm really done. Yeah. I think you were talking about this on a show the other day, actually, that one of the issues is that sometimes subscriptions go down for reasons other than this and it can be easy to bury it or make a big deal of it. And actually, it's got nothing to do with password sharing at all. It's just to do with, as Sarah says, the economics of the world right now. And it just looks like it's to do with password sharing because it's coincidental. 12 months time will be very interesting to look at this again. I have felt like they're trying to squeeze blood from a stone a little bit with this. But yeah, the data is mixed. So maybe I'm right. Maybe Netflix is right. Maybe neither one of us is right. I think it is causing a little bit of negative feelings around Netflix. But I'm not sure they were going to avoid that anyway. So maybe it's worth it if they get the balance back. Folks, if you're feeling social and you want to get in touch with us, go check us out on the social networks where DTS show on Twitter and masted on mstdn.social, daily tech news show on TikTok and DTS picks on Instagram and youtube.com slash daily tech news show on all of those platforms. You can find a link or a video like a TikTok. You just find the video of me unbagging the mobile audio studio I took with myself to Las Vegas for Monday's show. I open up the bag, I show you all the stuff that I brought, the mixers, the cables, the mics and all that. So if that sounds interesting, Twitter, masted on youtube.com slash daily tech news show and you can see that unbagging video. Despite how many times we repeat the chat, GPT doesn't think it just makes predictions about words, the worry about AI becoming sentient, still just kind of nibbles at the back of everybody's mind. Why is that exactly? Remember, we talked to Dr. Mary Ann Gary about some of the reasons, we did that last month. And the first episode of Nate's new Bloomberg original series AI IRL deals with this as well. Now, Nate, you talked to a lot of folks about this, but let's start with the reality. Are we even close to AI sentience? Absolutely not. And I'm very confident in saying that. And you're right. I have talked to a lot. And I should say, not even just me, but my co-host on that show, Jackie DeValus, we've spoken to a lot of people about this. And one of the people we spoke to, to try and figure this out from a neuroscientific perspective was David Eagleman. Now, I'd read a lot of David Eagleman's books. Prior to us doing the series, what I hadn't realized when I started watching the series Westworld was that he was a science advisor to Westworld. And so when we started putting the show together, he was basically the top name on the list to get on, because he can approach this from a very academic, scientific, grounded perspective. But at the same time, he's got his head plugged in in terms of like, well, long term, what could this mean? And how could you adapt this to sci-fi and things? So it made for a very compelling conversation. And I think a big part of the reason is that humans are extremely good at seeing themselves in everything, and indeed of making everything in their own image. We've done this for many, many years. It's proven time and time again, even just seeing what looks like a face in a shadow or in a biscuit or something, in toast. Exactly. We think we see a face in the corner like, oh, it's a face. Look at that. Isn't that cool? So we're trained to do that. In fact, we're not even trained. We've just evolved to do that. And so I think what I've learned is that we extend that kind of recognition into things that don't necessarily look like us in the very physical sense, but act like us or seem human. And that's where this comes into play. And this is why I think this question has been around for such a long time. And where it's got particularly interesting, obviously, over the last 12 months in particular, is with the result of these large language models like chat GPT and similar, because they give such human-like answers. And it's raised this whole question again, which has been around for decades. But certainly, there is nothing that has fundamentally changed that makes anyone we have spoken to, changed their view on whether it's close. That does not, though, mean that people don't think it is possible one day. And that's where things, I think, get more abstract and open to interpretation, in part, because we don't know what makes us conscious. We don't know what exactly intelligence or sentience or self-awareness is in a way we don't have the words to describe progress either. So that's all the big way of saying we don't know, but that doesn't make for a particular answer to your question. But we know it's not there yet, even if we don't know when it might come. Yeah, makes sense. So human anxiety about machines being self-aware, becoming sentient, taking over. What do you think, Nate, is because too many people are watching sci-fi television shows? Is it because if a machine gets something wrong, it can be interpreted as intentionally getting it wrong to confuse the humans? Where does this come from? In a way, I think we'd probably think it more if we didn't have sci-fi shows, because I think what sci-fi has allowed us to do is question reality and have other possibilities than what seems to be obvious. And if we didn't have sci-fi and just suddenly something starts saying things that sound human to us, we'd probably be even less skeptical that actually we might not be talking to another person. And in saying that, I've ever said slightly forgotten the nuance of your question, but it was around, is it the advancement, the speed of technology? Yeah, I mean, I think what I get, especially just all the news that's flying around about this is people pointing out the AI got it wrong, the machine got it wrong, but not because it's still learning, it's more of like, oh, this is designed to fool us potentially. And is this the machine somehow having a bad day and trying to throw us all off the case? Yeah, I mean, machines don't have intent. They have goals and we give those goals and sometimes those goals can be flexible in how we consider them achieved or not, but the machine doesn't come up with its own intent as we would define it yet. So we haven't got to that point, but that hasn't ever stopped anybody from wondering, yeah, but what if? And that's the thing. And there's a really interesting belief that I've read several times from very much broader experts than me who have said, you know, even if something like this is decades away, it could take decades to even figure out a way of controlling or understanding it. And so asking the question and agreeing that it is technically possible is not the same as saying it's imminent or it's about to happen, or even that it's necessarily a bad thing. So it is, I think, healthy to us, these questions. And partly that's why we've made it the first episode of the series, because we kind of want to, you know, take almost the most controversial one and deal with it head on straight away. But I don't believe that anything has fundamentally made it more likely than, say, this time a year ago. Is this just something where we need more exposure to the tools? Because it's new, we get more anxious about it, but the more familiar we get with it, the less anxiety we'll feel. I think it's possible. Sometimes more exposure does increase anxiety. You know, if you see how the sausage is made, you stop eating sausages. But generally, people aren't putting sausages in charge of running in critically important infrastructure in a city or a car or something. Because of what company you work at. But yeah, I got you. It depends what country you're in as well, and where you live. I don't think anyone's putting sausages in charge of the world, but, you know, use it as a metaphor. I think, though, at the end of the day, we think, you know, we know that the brain is a connection of neurons. It's electronic. It's a computer. And cloud computing and massive, you know, hyperscale cloud infrastructure has got to such a point. And neural networks have become affordable enough to scale enormously. And we could throw enough power at it that technically, there's nothing stopping us creating a brain, but we definitely still need to figure out what makes our own brain work in before we can even probably identify it in another. It's quite literally very cerebral, and it makes my head hurt. Well, I'm going to take two things away from this conversation. If I take nothing else away, one is that chat GPT is just a face in toast. And two, I should watch all of Nate's and Jackie's show on Bloomberg because it sounds fascinating. Yeah, not least because Tom may or may not be on an upcoming episode. I guess we'll just have to watch and find out. Well, also fascinating, my friends. This is a good one. Bioengineer Elise Elsacker published a study, April 11th, advanced functional materials that she and some colleagues made successfully self healing leather from mushrooms, thread like structures called mycelium, which can fix itself. Now, if you follow advancements in vegan leather, you might already know that mycelium leather already exists. But the way it's produced currently stops fungal growth. So the fact that it's mushroom really has nothing to do with the fact that you like your leather jacket that isn't actually made from real leather. Instead, the team grew mycelium in a nutrient rich soup. It had, you know, proteins and carbohydrates and other nutrients and then created a thin leather skin that formed on the surface of the liquid. They removed it, cleaned it up, dried it out, still leaving chlamydia spores in the fungus that were functional, just kind of dormant and hanging out. Then they punched holes in the leather, doused it with the soup again, and found that the mycelium started to regrow over the punctures. This is both wonderful and frightening that my mushroom leather jacket, I, you know, when it gets a hole in it, I can just douse in some soup. The researchers said, listen, what we created was very thin. It's not like you could wear leather pants made out of this that, you know, you just like throw some soup on it, you know, if you tear a hole in the knee kind of thing. But it is the first step into something that could regenerate, you know, what if you actually spend a lot of money for something that you wear on your person or, you know, is, you know, it needs to not have holes in it in other situations, and it can just fix itself like a wound. Never, never bet against a fungus. This is what I've learned in life, is you never bet against a fungus. They're amazing. You've seen the last of us. My gosh. Yeah. Well, yeah. And well, yeah. It's amazing. Also slime mold. Slime mold. It's not just a fantastic way to intrigue somebody you go on a date with who you end up marrying. Hello, Kate. But slime mold is just the most incredible thing. Like there's some fantastic TED talks and videos about slime mold and about mushrooms and fungus that I would strongly encourage people to go and check out because they really are fascinating things. Now, we are a minute away from this becoming real, right, Sarah, because isn't there a problem if it rains? Yeah. So a co-author of the study, Martin Dade Robertson, says, this technique could be used commercially, not ready for prime time yet, but the leather needs to be stronger and it also needs to be controlled. You know, throwing soup on something, that is not something that is going to be easily marketable if you buy a leather jacket or something like that in the future. He also says somebody could walk out in the rain, then all of a sudden find that their jacket is growing or perhaps has mushrooms popping out of it. Probably not what you're going for could be artistic, but probably not something that you want to happen. So just needs proper marketing. Marketing and additional research. But we are in the early stages of something where a terror could be regenerated by the clothing you wear on its own. I think a lot of people would like a leather jacket that had mushrooms popping out of it when it rained, but you know, you're my wish. I want to see one. That's for sure. Let's check out the mailbag. All right. So on Tuesday's show, Tom lamented the passing of the Honda Insight EV back in 2006. Well, dry your tears, because Komei wrote in and said, hi, we own a 2019 Honda Insight. It's a hybrid because Honda still makes them. The 1999 Insight was the first production hybrid vehicle that came just before the Prius. They introduced the new Insight again in 2018. Komei says, my impression is the dealership still wanted to focus on the well known names like he discussed on the show, the Civic, the CRV and just didn't make enough effort to promote the new model with that old name. We see other Insights only occasionally and most people don't even seem to know about the new one. Well, still, Komei, thank you for the Insight on the Insight. I did not realize it was back. That's great. And this is the second time it came back. Komei mentioned the 2018, but it was back between 2009 and 2014 as well. So good on you, Honda. Next time you see an Insight, you stop that driver and say excuse me, where'd you get that Insight? From Daily Tech News Show. Daily Tech News Show needs to know. Nate Langston, whether or not you know about the new Insight, we certainly love to know about everything that you're up to. So let folks know where to keep up with your work. Well, other than the aforementioned AI IRL, which I would love people to go and check out and watch, we have Neil deGrasse Tyson on in the next episode, which will be out by the time people hear this talking about space, astrophysics, you know, probably unsurprising given who he is. So I'd love people to check that out. And otherwise, as a slight side note, we had our very first ever national test of an emergency broadcast system in Britain the other day. And it went off live as we recorded our last episode of Text Message, the podcast. So if you're curious what that sounds like, go and check out. I forget the episode number 307, maybe. And it's in there. It was a lot of fun. Yeah, that was great. I was listening to that when I was boarding a plane. And I knew that you were talking about that story and I paused it. And when I unpaused once I got to my seat to start listening again, that alert went off. And I was so glad that it was like timed perfectly for me to enjoy it. It was fun to hear. Well, thanks for having me again. Well, thank you for being here, Nate, as always. Also, thanks to our brand new bosses. We got two of them today. We got Josh and John, who both just started backing us on Patreon. Thank you, Josh. Thank you, John. And welcome to you both. Excellent. Man, I'm very excited that we have a couple of new patrons every day for the past couple of days. Keep it coming. If you are not yet a patron, you'll want to join Josh. You'll want to join John. They're getting welcomed into the club along with the other patrons who get the extended show. Stick around. Good day, internet. We're going to talk about Tile's Location Tracker for your cat. Why is it just for your cat? Does your cat want it? And what if the cat really doesn't want to be found? We'll talk about all that and more with the patrons. Stick around. But you can catch this show at DTNS's Live Monday through Friday at 4pm. That's 2,800 UTC. Poor PM Eastern, rather. You can find out more at dailytechnewshow.com slash live tele-friend. We'll be back tomorrow talking about AI energy use and working on improving efficiency of data centers with Molly Wood and Justin Robbery on both with us. Talk to you then.