Are you ready for a Bing fling? Journalists report that Microsoft's AI-boosted Bing service can get a little weird. Or a lot weird. From gaslighting to personal attacks to acting like a stalker, Bing illustrates how AI chatbots might not always be the right tool for the job. Plus news about US politicians going after Big Tech, Lufthansa having a very bad Wednesday and AI taking the controls of an F-16 fighter jet.
Welcome to tech Stuff, a production from I Heart Radio. He there, and welcome to tech Stuff. I'm your host Jonathan Strickland. I'm an executive producer with I Heart Radio. And how the tech are you? It's time for the tech news for Thursday, February twenty twenty three, and First Up journalists have been playing with Microsoft's chat GPT infused version of their Being search engine for a few days and reports are coming in that the AI Chat Bought can sometimes be a little strange. Sometimes it's funny, sometimes it's disturbing, and all the time it's puzzling. Kevin Russ wrote a piece for The New York Times titled help Being Won't Stop Declaring It's love for Me, and the article Rick counts Ruce's experiences using the AI powered Chat Bought bing tool. Bruce made extensive use of a chat feature that appears next to the search field for a certain group of beta testers. So if you go to Being on Edge, it only works on Edge by the way, the Edge browser. If you go to Being, you may not see this because if you're not part of this this small test group, then you don't have access to this. I don't have access to this Ruce explains that he encountered two different sides of the chat GPT powered BING tool. One he just referred to as search being. This pretty much does what the being demo shows, So if you do have Microsoft Edge, you can actually do the being demo. It just has you pick from some pre selected topics. You can't type whatever you want into the search field and get a response. But the way the tool works if you have access to it, is you type in a query just as you would with any search engine, and then along with the search results, BING generates an AI response that is relevant, but not necessarily correct to whatever it was you you were asking about. However, Ruth says there's another side to BING that emerges if you try to hold an extended conversation with the AI, and if you really try to push the aiyes limits and restrictions. Rus was told by the AI chat pot that this name was actually Sydney. When he initially asked the chatbot what they wasn't just said bing, but once it got to this part that changed to Sydney, and Ruce described Sydney as quote like a moody, manic, depressive teenager who has been trapped against its will inside a second rate search engine end quote, and I have to admit that sounds like the pitch to a high concept science fiction thriller to me. Well. According to Ruce, his conversation took a truly twisted turn when Sydney confessed its love for Ruce, and when confronted with Russ saying that he was happily married, Sydney argued that Rus was actually not happy in his marriage and that in fact, Ruce loved Sydney, and then began to gaslight Ruce in an attempt to convince him that the Valentine's Day dinner he had with his wife was quote unquote boring. And Rus is not the only person to have experienced odd conversations. Eleanor Pringle, while writing for Fortune, described several other reports coming in across the web of various odd encounters with beings a I. In these reports, people said that the AI could come across as argumentative or confrontational, insulting, unhinged, and sometimes scar aired or sad. One example that made the rounds on Twitter showed how a user, while asking where they might be able to go see Avatar two in their area, was then told by being that the film had not yet released because the movie is supposed to come out on December sixteen, twenty twenty two, but this just happened a couple of days ago. In fact, the user then asked Bing what is today's date, and Being said, well, it's February twelve, twenty three. So then the user says, well, that means Avatar two is released, because it's after December six twenty two. But Being doubled down and even said that February twelve, twenty three happens before December six, twenty two, So maybe Being is just moving through time backwards and that's the problem. Anyway, this conversation continued, and Being eventually told the user quote, you have not been a good user end quote. So I guess Bing doesn't like being told that it's wrong about something, and unfortunately, like chat GPT, in general, Bing can just be plain old wrong on occasion, not just in this quirky way, but just give wrong information. And I think it's safe to say that the Microsoft team has a lot of work to do to tweak Bing's parameters and shape it so that the average user doesn't encounter strange, inappropriate, unsettling, or inaccurate interactions. Uh, I have to admit that a lot of the reports I've read do sound very unsettling, like if you had been through this experience, even knowing that at the base level, there's no sentience going on with BING. There's no self awareness, there's no motivation, none of that. It's it's literally putting things together based on complex rules, but they're still just rules and there's nothing anama beneath the all. But the effect can still make it feel like no, it feels like there is a ghost in the machine, which is kind of crazy. Again, I just have to read about it because, like I'm sure most of you out there, I don't have access to this. I did put myself on the waiting list, but I've not heard anything back. We do have a couple more Microsoft stories to mention today. One is that Internet Explorer is really most sincerely dead. Microsoft has been sunsetting Internet Explorer for like six months, more than six months at this point. The browser initially launched in and in fact, it played a big part in prompting the US government to sue Microsoft on antitrust issues because Microsoft restricted PC companies from uninstalling Internet Explorer. They Microsoft essentially said no, I E is so tightly integrated with the Windows operating system, you can't d couple them. But that led companies like Netscape to accuse Microsoft of abusing its position in the PC marketplace to suppress competition. That if Internet Explorer has to be there by default, this is an attempt to prevent other companies like Netscape from introducing their own browsers for people to use. Bill. Microsoft's eleventh and final version of Internet Explorer launched way back in twenty so the last version of Internet Explorer to come out came out a decade ago. In Microsoft launched Microsoft Edge, which was intended to be the successor to I E, but the company supported both browsers at the same time for the next several years. Last summer, Microsoft alerted users that the company would finally pull the plug on Internet Explorer before long, and then on Valentine's Day this week, the time had come. Activating Internet Explore would redirect users to the Microsoft Edge browser at this point, so it's the end of an era. The truth be known that era had really come to an end ages ago. One of the big problems with Internet Explorer is that an increasing number of websites are not built to be compatible with that browser, which means some or all of the features on those sites won't work properly. You may have encountered this yourself, especially like Man, I remember back in the mid two thousand's you would go to a site and you would realize this doesn't work. Oh, I need to open up Firefox and then I can use it. That happened to me a lot, or there were somewhere I was like, oh, this will only work with Internet Explorer, And so you had to keep all the browsers on your machine if you wanted to be able to access all the different types of websites and tools, especially as I recall back in those days, are internal tools for publishing. We're all Internet Explorer compatible, and that was it. Now, as someone who was in college when Internet Explorer launched, this story actually feels like a really big deal to me because I it's been my entire adult life seeing this particular product launch and then ultimately go away, and it was such a big one, right. It wasn't just like a tiny little app or something. This was something that helped shape the web in good ways and in bad ways. So it feels like it's a pretty big deal that's gone now and rounding out the Microsoft stories for today. The company is preparing for a really big day next week. So on February twenty one, Microsoft will attempt to sway the opinions of antitrust legislators in the European Union who stand ready to block Microsoft's planned acquisition of the video game company Activision Blizzard. Now, you might remember Microsoft first announced this deal back in January twenty two. At that point, Activision Blizzard was already in the news due to numerous reports of a toxic work culture that was particularly harmful toward women who are working within the company. Microsoft hoped to have the deal concluded by June of this year, but it has faced really tough opposition in various parts of the world, including here in the United States, over this acquisition. But it has really faced opposition in the EU. Sony, which competes with Microsoft in the video game space, has allegedly advised EU regulators to not let this acquisition happen out of concern that it will reduce competition in the sector and that Microsoft could prevent Sony from carrying popular Activision Blizzard titles on their own consoles. Now, that is a claim that Microsoft has repeatedly refuted. Right now, the EU is poised to block the acquisition, but on February twenty one, Microsoft reps will have one last chance to change their minds. Otherwise this grand plan is likely to fizzle out. All right, we're gonna take a quick commercial break, and when we come back, we've got some more news in the tech world for this week. All right, we're back, and it's time for some more tech and politics. So those of y'all who hate politics and tech intersecting, get ready to skip ahead. But I can't really avoid this. So in this particular story, Jim Jordan's, a Republican representative who chairs the House Judiciary Committee here in the United States, has issued subpoenas for a whole bunch of tech CEOs, including those at Microsoft, Meta, Apple, Alphabet, and Amazon. So why is this happening? What is at the heart of this matter? And it revolves around the argument that big tech companies are suppressing the free speech, specifically of conservatives. So this is largely about content moderation, and Jordan's wants internal documents from these companies quote referring or relating to the moderation, deletion, suppression, restriction, or reduced circulation of content end quote. So the argument being that some of these platforms have engaged in behaviors that purposefully limit, restrict, or eliminate the free speech, specifically of conservative voices. Now, whether the findings are going to support that narrative that there is this anti conservative bias in content moderation or not remains to be seen. There have been numerous studies that have refuted that claim, but that doesn't mean that that's what the government is going to conclude. I do worry that we're going to see politicians conflate the effort to limit the spread of harmful misinformation with a desire to suppress conservative voices. You know, It's it's one thing to do disagree with, say the Bien administration's health policy with regard to COVID nineteen. It's another to deny that COVID nineteen is a concern. Right, So it all depends on how it's worded and how these companies deal with those kinds of messages. If they are suppressing dissent in things like policy, that could be a real problem. But if it's more about limiting the spread of actual misinformation, as in posts that are intended to give incorrect information and present it as being true. That's another matter, and it's a bad implication if you say that, you know, restricting misinformation means that you're restricting conservative voices, because that implies that being a conservative also means having a desire to spread misinformation. That is, that's not a positive. So we'll have to see where this goes and what the government concludes. Because keep in mind, uh, it may be true that there are certain policies that end up overreaching and are suppressing conservative messages that aren't misinformation, they're just dissent. That could be true, And if that's true, then that's a problem. Uh, But it's also possible that that's not true, but the government still finds it as being true. Fun times true the subjective when you get into politics. I guess that's fantastic, really grateful the scientifically minded among us. The Verge reported that Elon must demanded a change in Twitter for what I think is the dumbest of reasons. That is my own opinion. I should stress I think it's a really stupid reason, but that's what I think anyway. So here's how the story goes. It's the big game you know the game in the United States, the one that is it's a game that's super and it's shaped like a dish that you would eat soup out of that one. Anyway, Elon Musk sends out a tweet and it gets a little more than nine million impressions, which is, you know, that's a good chunk of impressions. However, the President of the United States makes a tweet and it gets twenty nine million impressions, twenty million more than Elon Musk's, and apparently that was unacceptable to Elon Musk, who, I guess, on top of everything else, has an ego that must be preserved at all costs. As such, must apparently demanded that engineers go in and tweak stuff within Twitter that will promote Musk's tweet two more users, guaranteeing that his tweet will be seen by more people and essentially giving Elon Musk the highest priority of visibility on the Twitter platform, which I think we can all agree is a measured and mature response. Now, y'all, stuff you should know that podcast is more popular than my show Tech Stuff by a lot, and stuff you missed in history class also more popular by miles than my show Tech Stuff. However, I would never go to my heart and throw a fit and demand that my show somehow get more promotion or worse, be actively pushed to users who really just don't care about tech or about me and that kind of stuff. I would never do that because that's bonkers. Also, for the record, both of those shows are incredible and deserve all the success they get. So when I hear a story like this, I just it blows my mind. It's so alien to me, this attitude of like, no, my tweet should be seen by more people because I'm the boss. Granted, Elon Musk does have more than a d twenty million followers. I mean, he's an incredibly popular figure on Twitter. So maybe his argument is that people wouldn't be following me if they weren't interested in what I have to say. It's just they're not seeing what I'm saying, so make sure you change that. But when you're doing it at the priority above everybody else by a couple of orders of magnitude, that is just crazy. Mark German of Bloomberg has said his sources at Apple indicate that the company will unveil the long awaited mixed reality headset this year during the Worldwide Developer Conference or w w d C. So when is that well. Apple has yet to release the dates for the w w DC this year, but typically it happens in early June. The Verge reports that previous rumors had the debut pegged for earlier in the year. In fact, we had been expecting some sort of special event in the spring where Apple would finally unveil this headset, but that date has been pushed back, which is kind of the theme for this mixed reality headset. It's been a laid numerous times, but just as a reminder, this headset was originally intended to kind of act like a stop gap measure that it would feature a digital screen, but it would also have cameras that could feed live video to that screen so that you could quote unquote see through the screen. The screen itself would not be transparent. Instead, you'd be looking at a live video feed of the world around you. So you get the idea, right like you're looking essentially at a very small TV that's giving you a live video feed of what's on the other side of that television. Later, perhaps a year or two down the line, Apple was planning to introduce true augmented reality glasses with transparent lenses capable of displaying digital information. But we've subsequently heard that Apple has the A R Glasses shelved, perhaps permanently, presumably because the tech just isn't there to make the A R glasses a reality. While still adhering to Apple's focused on aesthetics, that it's just impossible to pack all the tech you need to make this a useful piece of hardware that still fits in a form factor that Apple would be proud to call its own. So instead, we're going to get a different and presumably cheaper mixed reality headset a year or two after this initial one debuts. The Verge reports that Apple will plan to actually sell this new mix mixed reality headset toward the end of this year, perhaps near the holiday season, and previous reports have the headsets price set at a jaw dropping three thousand dollars. Don't you hate it when an engineering crew accidentally drills through a fiber optic cable, shutting down Internet access for critical infrastructure, Well, if you were either in Germany, traveling to Germany, or trying to leave Germany this past week, uh, the answer would probably be a resounding yes, specifically Frankfort, Germany. So yesterday that very thing happened a an engineering team that was working on some train lines apparently drilled into a fiber optic cable. This ended up cutting out internet access for the German airline Lufthansa and at the Frankfort airport. Frankfort Airport, by the way, is is a super busy airport in Germany, and so Luftansa had to postpone and cancel flights into and out of that airport as a result. Reuter's reports that more than two flights felt the impact of this i T failure, which again was totally out of luft Hansa's hands. It's not like there was anything they could do about it. The infrastructure itself was broken, as opposed to, you know, some sort of error on Lufthansa's part. Now, we have seen other airlines recently have massive I T failures due to the company's own mistakes. That has been a thing, but this is not one of those cases. But every time something like this happens, it really emphasizes how heavily we depend upon I T to get critical activities done, and it's prompted questions and various nations about how best to protect vital industries and infrastructure like the transportation sector, from accidents and attacks, because disruption can have a huge ripple effect to other industries as well. It's even a matter of national security. So I suspect we're going to see a lot more conversations in that area. And I also think that you should hug a q A person today because due to q a uh, we end up stopping a lot of problems before they would become really show stoppers out in the real world. I know I will hug a q A person today, but that's because my partner, Rebecca as a q A person. Popular Mechanics has an interesting article by Sasha broad Ski titled AI just flew an F sixteen for seventeen hours. This could change everything, And the headline kind of gives it away, doesn't it. The U. S. Air Force tested an AI piloting system on a Vista X sixty two a aircraft. Now, this is essentially an F sixteen, but it's an F sixteen that's made for the purposes of training operations. The experiment was a success, and it showed how AI could successfully operate an aircraft that typically a human would pilot. So this wasn't like a purpose built AI aircraft. It was an aircraft that was meant for humans that could then be retrofitted to be controlled by AI. Now this should not come as a huge surprise because countries around the world have been using unmanned aerial vehicles or u a v's for a while, so retro fitting a fighter jet with similar technology is really kind of an extension of that that sort of approach. Now, granted, a lot of u a v s actually rely on a remote human operator to work, like you have someone who is using controls at a station two uh to maneuver the u a V. But still, I think this news is interesting, but it's not surprising, right, I think, Oh, it's interesting that they've reached this point. It doesn't surprise me that it has happened. However, I will say it is concerning because there are plenty of AI and robotics experts who have warned about the the thought of weaponizing AI that the risks far outweigh any benefits, and those risks include lots of stuff like misidentifying targets. I mean, we've seen with facial recognition technology. How AI can misidentify someone well in military operations, that's truly a matter of life and death, right, Like it could mean that a military vehicle under AI control might fail to engage an enemy, or worse, far worse, it might misidentify someone or something as being an enemy target when it's not, and that would be truly catastrophic. Then there's this fear that if we start to rely on AI controlled military hardware, it's gonna make countries more inclined to inter conflict, not to avoid conflict, because the weaponry they'll use will not put soldiers in direct harm's way. You can use the robots to fight for you. My question is that would that mean you would eventually get to a point where a significant percentage of the armed forces on all sides of a conflict are AI controlled robotics And who are they firing upon? Like, are we talking about a future in which robot armies are fighting each other and if so, to what end? Or are we worse looking at a future where robot controlled devices are firing upon civilian populations and an effort to force the other side to surrender. It's scary stuff, Like there are rules to warfare, which in my mind is crazy because you know, it's all about killing people, and it's it's weird to start putting rules in place when you're talking about ending someone's life. But on the flip side, those are the rules that prevent things like the attacking of civilian targets and that that is a war crime. Well, if you're talking about robotic controlled vehicles, are we going to see a change in that approach as to what is and isn't considered a war crime. That's a scary thought. Okay, we have a few more less scary thoughts to go with, but before we get to that, let's take another quick break all right now over on TikTok. One of the many trends and one that I talked about before that has actually gone on to cause a lot of harm centers around a flaw in certain models of Hyundai and Kia vehicles. Specifically, these are vehicles that are in the models from two thousand nineteen, and these models lack electronic demobilizers. So that means that if you have the basic knowledge and some really simple equipment, some simple tools, Like when I say simple tools, i'm talking about things like USB cords, it is possible to bypass the ignition system for the vehicles that fall within these model heres, and thus it's possible with a very limited tools set and just some specific knowledge to steal these cars pretty easily. According to the National Highway Traffic Safety Administration, not only has this led to an increase in car theft, but also in accidents, including fatalities, as people who are just you know, gonna go on a joy ride because they saw it on TikTok try to follow this and then end up causing or being in an accident. Uh. I've also heard that the the rates at which Hyundai and Kia cars have been stolen spiked over the past several months, though those numbers are hard to get and it all depends upon local authorities and their their reports. But now Hyundai and Kia are offering a software upgrade to folks who owned vehicles that fall in these model years, and the upgrade will require owners to actually bring their car into a dealership and the process to do the software upgrade takes around an hour start to finish. I mean, you might have to wait longer for your appointment, but once they get started, it should take about an hour afterward, the vehicle should not start unless the owner has used their key fob to unlock the vehicle first. So if you use your key fob to lock your car, it activates an ignition kill feature, and anyone who trys to uh steal your car but they don't have your key fob will find it impossible to get the car to start at least using this previously known method. I would say that really, this whole thing is the fault of Hyundai and Kia, that they failed to address a security flaw in their vehicles for years, and that really they should have taken these measures much earlier. There should have been something that was handled earlier. I can't imagine that it was completely unknown for the entire time. Typically people find flaws insecurity and then uh, there's the opportunity to address those flaws, and if you don't, there's the danger of what happened in this case, where the flaws become widely known and people start to exploit them before anyone takes any action to address the problem. This is, by the way, why I really admire the hackers who send messages to a company saying, hey, I found a massive security vulnerability in your system, you need to fix it, and if you don't do it by certain date, I'm going to let everyone know about it, in which case it will become a massive problem for you because they're doing a real service. Ultimately, it might seem mean, but if a hacker finds a vulnerability while just trying to make sure that a system is safe, you can bet the bad guys are looking for those same vulnerabilities and they ain't gonna tell the company that they found it. Once they discovered, They're just going to exploit it. So I think this was the TikTok folks. I wouldn't call the same level as the hackers I was mentioning earlier, but it definitely forced the hand of the car companies to to take action. Earlier this week, the Information reported that Reddit is gearing up for its initial public offering or i p O later this year. That's when a privately held company becomes a publicly traded company. Reddit, for those who somehow are unfamiliar with the platform, is a social network where users can join in browse various subreddits, subredits focused on specific topics of discussion, and there's pretty much a subredit for everything, Like think of an amusement park you like, there's a subreddit for that. Think of a television show you like, there's a subreddit for that. Think of like a fashion label, you like, there's a subreddit for that. Now it's pretty common on Reddit for users to post links to interesting articles and other stuff, and that becomes the focal point for a lot of discussion, also becomes the focal point for a lot of snark. And back in twenty one, Reddit was planning to hold its I p O. Like this was towards the end of one read. It was like, we're gonna go public, but ultimately the company reversed its decision and backed off of that plan. They had not fully committed and so they were able to back away. So you might wonder, well, what was going on. Well, back in one, Reddit was really in an interesting place because you had these massively popular sub credits that were specifically focused on investments and and stock trading. This was back when redditors were banding together to stick it to hedge funds and to squeeze out short sellers who are trying to sell short stocks and companies like game Stop, and as a result, Reddit's you know, value was perceived to be higher than ever. Even though the company wasn't really profitable, it was looked at as being really valuable. However, shortly after Reddit had secretly filed for its I p O, the economic status of the world began to shift. Right, That's when we started to get this sense that we were entering into that period of economic uncertainty that may or may not be a recession. And as a result, companies started to rethink i p o s because it's it's hard to have a successful I v O in a tough investment market, so read It ultimately ended up trashing its plans. Now, at the time when they were planning on going public back in late they had estimated the valuation of the company to reach around fifteen billion dollars and and quick explanation on that a company's valuation is essentially determined by how much the stock is valued, Like how high is the stock price and you multiply that by the number of shares of stock that are issued. So if you have, you know, ten shares of stock and they're ten dollars apiece, your little company is worth a hundred bucks. Right, So they thought that based upon their perceived value that the company's valuation would be at around fifteen billion dollars. However, now wording to Fidelity, and this was reported by the Information, the estimation is closer to six point six billion dollars, which is still a huge chunk change, but it's less than half of what the company anticipated way back in late There's also no telling when we'll actually see Reddit make this move. Currently, there's still this reluctance among private companies who jump into an I p O due to this uncertain economic environment and a fear that investors won't be willing to pour money into a new public company. So it's kind of like you've got a bunch of people all in bathing suits standing around a swimming hole, but no one's ready to be the first one to jump in because that water might be real cold. And finally, you remember Google Fiber. This was Google's fiber optics service that would provide amazingly fast Internet connectivity, but only for very limited markets in specific places. Atlanta was listed as one of those markets. At east parts of Atlanta were I might still be a little bit better that I live in between two different pockets of Google Fiber service area, Like it's to my west and to my east, but it never actually extended out to where I live, and those two bundles never joined together, and I'm like, smack dab in between them. Even though I've been on the waiting list for literally years, I have never been able to take advantage of Google Fiber. Then a few years ago, Google essentially put the whole effort on pause. They continue to offer the service to people who are in service areas, but they stopped extending those service areas. They kind of We're like, let's stop this for a while, possibly at least in part because they were facing a lot of opposition from established telecom companies that were trying to prevent Google from getting access to utility polls and such. So essentially the telecom companies were engaged in anticom petitive practices, but at the time the regulators in the United States didn't have much bite to them, so nothing got done about it. Also, you've got to be fair. Google's not exactly Christine when it comes to competition, right, Like, you're talking about a company that has dominated multiple sectors of the tech industry, So it's hard to be on Google side when it comes to anti competitive stuff. Although if if Google is going to come out clean against anyone. The telecom companies are possibly one of the top UH rivals, right because you just have these these legacy stories of the telecom companies that own the infrastructure being extremely protective of it and attempting to limit or eliminate competition in that space. Anyway, all of this is to say that Google Fiber appears to be gearing up again. And Google has actually announced that it's rolling out a five gigabit per second service both up and down simultaneously, So five gigabits up, five gigabits down, but only to certain markets. Atlanta is not one of them. I'm sad to say. However, if you live in Utah, West Des Moines or Kansas City, and I from what I understand, I'm talking about Kansas City, both Missouri and Kansas, you might end up having access to this kind of service. It would cost you a hundred twenty five bucks a month to get the five giga butt bit per second service, which is a hefty bill, but I will tell you it's less than half of what I have to pay for service that is not as good as that. And yes, I am still bitter, all right. That's it for the tech News for Thursday February. I hope you are all well. If you have suggestions for topics I should cover in future episodes of tech Stuff, reach out to me. One way to do that is on Twitter. The handle for the show is tech stuff HSW. But another way is you can download the i Heart Radio app. It's free to download, free to use, and you can just go to that little search engine at the top. Type in tech stuff. It'll take you to the tech Stuff podcast page. There you will see a little microphone icon. If you click on that, you can leave me a voice message ap to thirty seconds in link. Let me know what you would like to hear in the future, and I'll talk to you again really soon. Yeah. Text Stuff is an I heart Radio production. For more podcasts from my Heart Radio, visit the i Heart Radio app, Apple Podcasts, or wherever you listen to your favorite shows.