Welcome to tech Stuff, a production from iHeartRadio. Hey there, and welcome to tech Stuff. I'm your host, Jonathan Strickland. I'm an executive producer with iHeartRadio. And how the tech are you. I'm having a rough recovery day. I'm still getting over COVID. I'm no longer testing positive for COVID, but there's been some lingering symptoms and some days are better than others. This is not one of the better days. But it is a Thursday, which means it's in tech News day. So we're going to cover the tech news for October twelfth, twenty twenty three, and like Tuesday's news episode, we're going to start off with a bunch of Twitter slash x news. So first up, I mentioned that a lot of people from media outlets to government agencies to watchdog groups have been directing criticism at X for the proliferation of misinformation and disinformation, specifically in the wake of the ongoing war in Israel. It did not take long at all for bad information to flood the platform after Hamas launched attacks within Israel, and that bad information included everything from photos and footage of past conflicts being passed off as if they were from the most recent circumstances, to some accounts claiming that footage that was taken from modern video games actually represented circumstances in either Israel or Gaza. One critic who raised concerns about the platform in general was Thierry Breton or Breton, I apologize for my pronunciation, but we're talking about a European Union commissioner who said that X was a place for people to quote disseminate illegal content and disinformation end quote. That could mean that X could be found guilty of viol leading the EU's Digital Services Act, or DSA. So the DSA went into effect earlier this year, back in August, and it covers a lot of ground when it comes to a platform's responsibilities. The DSA expects that large platforms ones like x Slash, Twitter or Facebook, or even platforms like Google and Amazon, have to take an active role in content moderation. Unlike the United States, which has Section two thirty, the EU's DSA holds these platforms at least somewhat accountable for the material that users post to those platforms, and the DSA gives the EU government the authority to find companies that failed to comply with the law, they can get hit with a really hefty penalty, up to six percent of their global turnover, which for the purposes of this episode, will just kind of reference as revenue. It's there's a slight distinction there, but you'll get the idea. So last year, Twitter's global revenue was around four point four billion dollars, so the DSA essentially says they could take six percent of that. They could find Twitter six percent of that for failing to comply with the DSA. That would be around two hundred and fifty million dollars. That's not a small fine. Plus. If the EU finds out that a platform continues to fall short of the DSA's requirements, the EU has the authority to suspend that platform's operations within the European Union, which of course would also mean the loss of even more revenue. If you can't operate in a major part of the world, you're missing out on a tremendous amount of revenue. Ex's CEO Linda Yakarino has posted that Twitter has removed hundreds of accounts affiliated with hamas and removed or labeled quote unquote tens of thousands of pieces of content also said that Twitter slash X has not received any information from europoll about illegal content on the platform. She's saying, well, we're being accused of something, but no one has actually come to us with these concerns. It's the first we're hearing about. Essentially, the company is relying on the quote unquote community, which essentially means volunteers who are flagging posts that might violate the DSA or x's own policies. But then X kind of has to rely on volunteers because Elon Musk effectively gutted the content moderation team by laying off most of them in the wake of his acquisition The Federal Anti Discrimination Agency or FADA THODA is leaving X. This German anti racism group explained the decision to bounce off the platform. They claim that incidents of hate speech have quote increased, particularly end quote, ever since Elon Musk took control of Twitter. FADA said that cases of anti trans and homophobic content, and racism, misogyny, anti semitism, and other examples of hateful speech are all on the rise, and that ultimately this forced the leaders of FADA to ask the question, does it make sense to stay on X? Is it providing more of a benefit than a cost. And clearly the endecision was that it just wasn't worth sticking around, or, as Fada's statement puts it, quote in our opinion, X is no longer an acceptable environment end quote. Now I've said this before, but in the years past, Twitter, which by the way, has always been a deeply flawed platform. It's not like Elon Musk took it over and then it went from being a shining beacon to the absolute worst. It had problems well before Elon Musk entered into the picture. It's just those problems have magna fight arguably since Musk took over. Anyway, back in the day, if you will, Twitter was put to effective use by various activists and humanitarian groups and news outlets. It was a scene as a very useful tool and worth the investment of time it takes to create and cultivate and maintain a Twitter account. But Musk's capricious leadership and the massive changes the he's made to X have convinced a lot of those entities just to pull up stakes and leave X behind. They're saying it no longer makes sense to stay there. And now we're actually starting to see what effect this has had on the organizations that decided to leave. So let's take NPR National Public Radio, for example. So in April of this year, NPR made the decision to leave X. This is right around the time that Musk actually renamed Twitter to X. And to be clear, MP made this decision for the official NPR accounts. They did not mandate that for everybody who works for NPR, but they decided they were going to stop their accounts on Twitter or X. The platform at that time had appended a label to NPR, and the label was US state affiliated Media. Now, Twitter's own definition for state affiliated media was quote outlets where the state exercises control over editorial content through financial resources, direct or indirect political pressures, and or control over production and distribution end quote. So clearly that definition is meant to designate media outlets that are serving as a mouthpiece for a government. So the content produced by those types of outlets tends to be heavily biased in favor of whichever government is in charge. I mean, that's clear, right. It's the kind of designation that you might give to a media outlet in places like Russia or China, where you have authoritarian leaders who crack down on any voices that oppose the state. Now, MBR does receive money directly from the US federal government, but it makes up less than one percent of its budget, and it is clear that the government doesn't have any influence on the content that INPR produces. So the crux of what makes a media outlet a state affiliated media outlet wasn't there, but that's what Twitter was labeling in PR. So NBR decided that's it, we don't have to deal with this, We're a gone, and that it leaves. It was clear that Musk had an axe to grind against NPR, So six months later, now that we're in present day, leadership at NBR has circulated a memo to the company's staff. In that memom, the leaders addressed the impact that the organization has seen since leaving X and it's not that much so, according to Niemann Reports, The NPR memo says that traffic to NPR has dropped by only a single percentage point as a result of leaving Twitter slash X. NPR also pointed out that even before all this kerfuffle, traffic from Twitter amounted to less than two percent of all of NPR's traffic, and that explains how the organization found the decision to leave Twitter to not be difficult at all. The message is one that X definitely does not want to hear that staying on Twitter just isn't worth the effort. The negligible amount of traffic NPR got from Twitter was not worth the frustration of operating on Twitter. I suspect other media outlets may also reevaluate their position on the platform, particularly as it deals with more allegations of hosting disinformation and illegal content. But obviously it's not going to be the same for everybody, right, some media outlets may say, well, yeah, but we have a more significant percentage of our traffic coming from Twitter. That's going to be a much harder decision to leave. In fact, I would be surprised if a media outlet chose to leave in the wake of a revelation that Twitter was providing a significant percentage of traffic to the media outlet. But for those that are in a position similar to NPR, I would not be surprised to see more of those outlets follow suit. Okay, we're going to take a quick break. When we come back, we're going to have some AI news, the other big story that continues to unfold throughout twenty twenty three. But first let's take a break to thank our sponsors. Okay, we're back, so let's talk about artificial intelligence in general and generative AI in particular. So first up, open ai has plans to make it less expensive for developers to build applications on top of open AI's models. So, as it turns out, using artificial intelligence, like using it and incorporating it into your work, tends to be really expensive because AI requires a lot of computational resources in order to work, and that goes all up and down. The computational resources gamut from processing power to computer memory to things like storage. Not every developer has easy access to all of those things. It's entirely possible to come up with a potentially brilliant idea on how to incorporate AI into an app, but not be able to afford the costs of bringing that idea to life. So open ai plans to offer options that can mitigate some of those costs, like offering memory storage, for example, which is a huge cost apparently for developers who are working in the field of AI. So that's just part of open AI's strategy. The company also plans to unveil some computer vision tools, which can make it possible for developers to create applications that include some element of image analysis in them. You know, it's pretty easy for humans to describe stuff that's included inside an image. It tends to be a pretty heavy lift for a computer, and the computer vision tools simplify this and gives developers access to capabilities that otherwise they likely wouldn't be able to create on their own. According to various news outlets, the company is going to roll out these features around November sixth. That's during the company's first ever developer conference, which is taking place in San Francisco, California, on November sixth. And this matters to you and me because it likely means we're going to be seeing a rush of new applications that incorporate AI, at least to some extent in the not too distant future. So like it or not, the AI revolution is going to intensify before much longer. Now that's going to have some hefty real world consequences. Even if we assume that every AI application is benign or goodness, let's say that it's even helpful. Let's be super optimistic. Let's say that the incorporation of AI is genuinely helpful in all of its implementations, there would still be consequences we would have to face even with like perfect AI. Tech Spot reported on a research paper written by Alex Devrie titled the Growing Energy Footprint of Artificial Intelligence, and Devree argues that while much research has been focused on how much energy AI language large language models require during their training phase, it's a lot like when you are training a large language model, you're spending a great deal of money just on electricity to power that process. However, less studied, but just as important, arguably, is the amount of electricity needed during normal operations. Like you've stopped training now you're just deploying applications built on top of those large language models. Defrie is saying, well, that's going to still put a pretty hefty demand on electricity generation, especially as more and more of these apps are deployed, are developed and deployed. So an analysis firm called semi Analysis estimated that if Google were to incorporate AI into every single search, which it cannot do. I want to be clear about that, it lacks the physical capability of powering that. But if Google were to try and do that, it would require electricity equivalent to somewhere in the neighborhood of twenty nine point two tarawatt hours on an annual basis, which is pretty much the same amount of electricity that Ireland uses per year. So to power AI enhanced Google searches, you would need the same amount of electricity that Ireland needs for a full year. That's incredible. It also reminds me a lot of cryptocurrency mining. Like we talked a lot about how the process of mining crypto coins, particularly proof of work crypto coins requires a ton of electricity. Proof of stake less so, but still requires some proof of work requires a lot. That's the kind that like bitcoin uses. And we have already talked about how in the past, at least mining operations worthy creating the power need equivalent of what we would see for some nations in Europe on an annual basis pretty crazy. Well, same thing is true of AI, but as I said, Google doesn't have that capability. This is more of a theoretical kind of discussion. Google lacks the equipment that would be needed to be able to do this, and in fact, the companies that make the equipment, like make the processors that would power this kind of thing, they aren't producing chips at that scale, Like they could not produce enough chips in the short term anyway to power that. So this is really again just a hypothetical situation. But the point is that AI is a power hungry technology, not in the sense of you know, sky net taking over the world, but in the sense of it requires a ton of power for it to work. And it's something that people like defriethink developers need to really take into consideration. Does it make sense to incorporate AI into an application knowing the energy demands of that particular application, And if it makes sense, then sure. But if it doesn't make sense, then don't just throw it in there just to have AI as part of your value proposition of your app, because you're going to be costing yourself a lot more money due to the fact that there's so much electricity needed to power the darn thing. Okay, moving on to some other news. Microsoft got some bad news in the mail last month, which we are now learning about due to a filing with the Securities and Exchange Commission up the United States Internal Revenue Service. The dreaded IRS sent a notice to Microsoft that the company had failed to pay a few bills. It was called a notice of Proposed adjustment, which is a, you know, a fairly benign sounding term, but it had some massive consequences. According to the IRS, Microsoft ows about twenty nine billion dollars twenty eight point nine billion to be more specific, and that's in back taxes from the two thousand and four to twenty thirteen period, as well as penalties associated with a failure to pay those back taxes over the years. Microsoft says the whole thing's a misunderstanding. In fact, they say they outright disagree with the Notices of Proposed Adjustment aka NOPA, So I guess they're saying NOPA to NOPA. They're going to pursue an appeals process with the IRS, and if that doesn't go Microsoft's way, they plan on bringing the whole matter to court. It would be shocked if this doesn't ultimately go to court at some point. And yeah, it's going to take a minute for all that to hash out, so we'll have to check back. And I don't know, like five years or so, the State of Utah has filed a lawsuit against TikTok, and the charges are probably pretty much what you would expect. So first and foremost, Utah's Governor, Spencer Cox, argues that TikTok children both by getting them hooked on the platform so that they're spending countless hours on it, and also by serving up harmful material. The lawsuit cites research that links social media activity, like prolonged activity on social media, with poor mental health. However, I think it is very important to note the distinction between causation and correlation here. So does staying on social media for a long time lead to poor mental health? Possibly, But it's also possible that people who have poor mental health tend to use social media for longer than other people do, and that would mean that using social media a lot doesn't necessarily lead to poor mental health. But perhaps poor mental health makes you more inclined to use social media. It's similar to how people who commit violence may also play a lot of violent video games, but that doesn't necessarily mean that playing violent video games makes you violent. Again, there could be a correlation, but no causation there. We don't know, is my point. So I think making legislation on something where there's not a clear causal link is problematic. You're not likely to actually address whatever underlying issue is at stake, But anyway, Alex Horrik of TikTok contested the charges brought against the platform by the state of Utah. He said that TikTok quote has industry leading safeguards for young people, including an automatic sixty minute time limit for users under eighteen and parental controls for teen accounts end quote, So he's saying we already have protections in place. These are here to prevent young people from being harmed by TikTok. According to AP News, the lawsuit also brings up a concern that we've seen many, many times from lawmakers here in the United States that is related to the relationship between TikTok, the American company, and Byte Dance, TikTok's Chinese parent company, arguing that maybe there are deeper links there than what TikTok claims. This is a song we've seen again and again, so not a big surprise here, but we'll keep an eye on where this lawsuit goes and how it unfolds in the future. Got one more than I'm going to tell before we get into another quick break it has been a hot minute since I've talked about FTX, the cryptocurrency exchange that famously went up in flames almost a year ago. Next month will be a year since FTX went under. Fdx's co founder Sam Bankman freed AKASBF is in court to face charges that he committed several fraud related crimes in his role as CEO of FTX. Yesterday, Caroline Ellison, the former CEO of Alimator Research, which was intrinsically connected to FTX, testified that FDx engaged in some pretty darn shifty behavior, including an attempt to bribe a Chinese official. So, as Ellison tells the story, Chinese authorities brows ftx's funds on the OKX and quea by exchanges. So apparently someone at Alibator Research, which by the way, was an investment company that specialized in the crypto market and had been co founded by SBF. Apparently this person unnamed person ended up doing business with some someone else who was suspected of money laundering. So the Chinese authorities locked down around a billion dollars worth of FTX assets, and this prompted folks at FTX to try and come up with various strategies on how to free up that money that otherwise was just locked away. It sounds like first they started off with some fairly reasonable and more importantly legal options, but apparently those did not get any results. So the leaders, including SBF himself, began to consider alternatives, and while they did not necessarily approve them initially, further down the line they did, and that ultimately resulted in offering a one hundred and fifty million dollars bribe to a Chinese official to grease the wheels and get that money unlocked. Not only did the company apparently actually do this, it also listed the expenditure in its State of Alameda documentation. Now, they didn't go so far as to label it bribery costs that might have been a red flag. Instead, it was filed under notable slash idiosyncratic PNL stuff. And here's the wild thing, y'all. The bribery thing isn't even one of the legal charges brought again SBF. Right, that's not even a material component of this trial. The whole story really just becomes part of establishing SBF's character and leadership decisions and willingness to engage in illicit activities. Ellison, for her own part, has pled guilty to several fraud charges already, so things are not looking so great for SBF at the moment. Okay, we're going to take another quick break. When we come back, I've got you know, I got quite a few more news stories, so stick around. We'll be right back. Okay, we have come back, and now for a couple of stories about how the US government is trying to make companies be more transparent when it comes to stuff like you know, bills, service fees. So first up is the Federal Trade Commission, or FTC, which has taken aim at so called junk fees. So these are those hidden fees that won't pop up in a transaction until it's right at the moment of purchase. So the customer might get a price quote on a product or a service and then they'll start the whole process of purchasing it or agreeing to it, and right when they're getting at that last moment to complete a transaction, a bunch of weird fees pop up and the cost increases, sometimes significantly. So this happens in lots of different kinds of transactions. Personally, I see it most whenever I use a ticket broker to purchase a seat for an event, whether it's a show or a concert or whatever it may be, sometimes the fees end up being as much as the cost of the ticket itself, so it doubles the amount I would have to spend if I wanted to, I don't know, go see thump Asaurus performs Strutton Live. Actually, that's probably a bad example because thump Asaurus tends to play small to medium venues. Those are less dependent on ticket brokers. But you get what I'm saying. You can also find it in other places, like vacation rentals, right, so Airbnb has been brought in as a possible perpetrator of these kinds of junk fees or car rentals. It can be like that too, where you'll see one price quoted as you log into a car rental company, but as you go through it suddenly balloons as you get further into the process. Not all car rental companies are like that, by the way, I've noticed that a few have started to already post what the full price would be on that initial screen, and I think that's admirable because not every one of them does that. A lot of them starts off with a price where you're like, oh wow, it's like eighty bucks for the full week. That's not bad. Then you click through and you're like, oh wait, it's actually like two hundred and seventy dollars for the full week once you add all these fees in. So the FTC would force companies to include any additional fees within the advertised prices themselves. They could no longer hide it until the end, and they have to explain in what those fees are actually for, like why does that fee exist? What is that fee going to? Like? Those convenience fees drive me nuts because I'm wondering who the heck is this convenient for, because it's not me anyway. It's also supposed to indicate which, if any, fees could be refundable. So all of this has to be done if FTC adopts these rules. I'm sure the companies are not eager to comply with any rule like this if it is in fact adopted. But as a consumer, I am personally fully in favor of making companies be more transparent when it comes to costs to customers. Now, on a related note, the US Federal Communications Commission or FCC has transparency rules that are going to impact the major Internet service providers in the United States starting on or in April of twenty twenty four. So starting then, providers will have to include a broadband consumer label, and this label will break down the provider's services. It has to include the service's features, its limitations, any restrictions that are in place, as well as the price for the service. So that means the label should show customer stuff like what broadband speed is available in their area, how much is the introductory rate for enrolling in that service, what are the actual rates once you're done with the introductory period. Are there any data allowances or data caps? If you prefer on that service, it would need to have all of that listed out, plus any other fees that would come into place, and that, in theory, should make it easier for customers to do some comparison shopping. However, I should also add that for a lot of folks in the United States, comparison shopping is more of a dream than a reality. It's not a luxury for a lot of people because a lot of locations, including where I live, there's not much choice. Technically, I have two providers in my area that provide for actual broadband speed service, and only one of them can provide service at greater than fifty megabits per second. The other one, I think max is out right at twenty five, which I think is even I think that is the threshold for broadband. So it's not like I even have a different option to go with if I wanted to, at least not one that would work for what I do for a living, So that's kind of a bummer, But at least I would have a better understanding of what I'm spending my money on and what all those fees are going to, and also things like how much would it cost to terminate the service early maybe suddenly I get a different opportunity to use a different ISP, how much would it cost for me to cancel my agreement with my current provider? And as you might suspect, the broadband companies have pushed hard against this move. They've claimed that it would be too difficult to implement, and that really just gives the FCC more ammunition, right, because if listing out all your fees is something that you say is too complicated, there's definitely a lack of transparency going on among other issues. So I don't think it is an argument that does the broadband service much favor. Next up, we've got a story about eBay and a potential two billion with a b dollar fine for the company. So what did eBay do that would warrant such a hefty fine. Well, it all has to do with the facilitation of sales of products that violate environmental laws in the United States, namely defeat devices aka gadgets that, when you install it in a diesel powered vehicle, gives the driver the ability to quote unquote roll coal. So if you've ever seen a car or a truck blast out a huge cloud of black exhaust all in one burd first, that is rolling coal. I had not even heard about that practice until just one day someone did it to me. I had been walking back from our office back when it was at Pont City Market, and it was a three mile walk, and it was a hot day, and I was taking a little break, and as I was sitting down on a bench, a big old pickup truck drove past me and then belched out a ton of black smoke rolling coal. Well, here's the thing. These devices violate the Clean Air Act, and if authorities find a retailer selling them, they can find that retailer five eighty dollars per sale, so it's accumulative, and the EPA has set its sites on eBay. Now, eBay has said the lawsuit is quote unquote entirely unprecedented. Now that might be true. It might be entirely unprecedented, but that doesn't mean that it's unwarranted. It also doesn't mean that the EPA won't continue to do that sort of thing with other companies. But eBay also claims it has actively removed listings for such illegal devices for years and that it has a ban on those kind of devices. Those are not allowed to be sold on eBay, according to eBay's representatives. So we'll have to see where this goes from here and whether or not it actually ends up going to trial, and whether eBay actually has the receipts to show that it has cracked down on this practice, or if, as the EPA alleges, eBay has allowed for the sale of these devices. Google announced it stopped a distributed denial of service attack or d DOS that was the largest ever on record. In fact, according to Google, this attack was seven and a half times larger than the previous record holder, and the attack was blasting targets with three hundred and ninety eight million requests per second. So, just in case you don't know what a denial of service attack is, it's basically a tactic meant to overwhelm a target server a web server. So in its most basic form, a denial of service attack will see attackers send out requests to a target. The target is compelled to respond to the requests because that's how the Internet works. So by sending millions or even hundreds of millions, or even billions of requests to a target, the target can get locked up trying to respond to all these requests. It can even shut down as a result of this. A distributed denial of service attack is the same thing, except the attacks are distributed across a network of attackers. It's not just one source of attack, and that makes it a much more effective means of attacking a target because if the attacks were coming from one source, the target might identify that source and then block traffic coming from that source and then protect itself. If it's distributed across lots of machines, it can be more challenging to actually protect yourself against that kind of attack. Google says it detected the attack and then launched mitigation efforts, and that included reaching out to peers and started to patch a vulnerability in the HTTP slash two protocol stack. The attackers were leveraging this vulnerability and it's present in some but not all servers. So server administrators have been informed they should check to see if their machines are vulnerable, and if those machines are vulnerable, to be on the lookout for a security patch from whatever vendor made that machine in order to close off this route of attack. And finally, for some sky is falling news, the US Federal Aviation Authority or FAA has a warning about all those satellites that were putting up in orbit. So I talked earlier this week about how companies like SpaceX and Amazon are launching thousands of satellites constellation of satellites up into orbit. And the warning that the FAA is giving us is the old saying what goes up must come down, even if it happens to be in low Earth orbit. So the FAA did an analysis and estimates that up to twenty eight thousand pieces of satellite could fall to Earth each year based upon how much stuff we're putting up there, and that these could potentially represent an actual threat to people in property. So the FAA estimates that the chance of being hurt or worse from a piece of falling satellite equipment could be point six per year, and that means that it could possibly happen every two years or so, just when you look at the big picture of statistics. And if that's not scary enough, the FAA also indicated that falling satellite pieces could potentially be a threat to aircraft in mid flight, though this represents an even smaller likelihood. The FAA indicated that regulation could limit some of this risk. So if some regulations were put into place that would restrict how many satellites companies can put into orbit or how they operate those satellites, maybe that would end up helping this. But the FAA also stressed it only has authority to do that kind of thing here in the United States. There are other nations that are launching satellites into orbit, and so there is potentially a need for some sort of international body that could set global regulations on satellites and launches and that kind of thing. And the proliferation of space debris. SpaceX, by the way, says that the FAA's analysis is flawed and that the company's satellites, they're Starlink satellites are far more likely to burn up on entry because they're not terribly big. So they're like, this is a moot point, because if our satellites are falling, they're not going to make it to the ground. They will burn up before they reach that far, and so there would be nothing significant landing on Earth afterward. They've also raised some other objections, so again, interesting, scary. My guess is that there's the truth is somewhere in the middle here. I do think that the odds of being struck by space debris are pretty darn low if you're not in space, probably somewhere along the same lines as getting struck by lightning. Although obviously you can help avoid getting hit by lightning by not being out and about while there's a lightning storm. It's a lot harder to predict when a piece of a satellite's going to come plummeting down, So I get it. It is kind of scary. Now before I go, I do have an article to recommend. This article is titled how AI reduces the World to Stereotypes. It was written by Victoria Turk and it was published in Restofworld dot org. So Turk is mainly talking about generative AI and how text to image systems tend to perpetual wait and amplify biases and stereotypes. So the article follows an experiment in which a team at rest of World use the text to image generative AI tool mid journey to produce pictures associated with specific places or types of people, and then analyze those images and looked for the inclusion of elements that reinforce biases and stereotypes. So I recommend that article again. It's titled how AI reduces the World to Stereotypes by Victoria Turk at Restofworld dot org. That's it for me today. I hope you are all well. Stay safe out there, take care of yourselves both physically and mentally. It is a really tough time right now, and I'll talk to you again really soon. Tech Stuff is an iHeartRadio production. For more podcasts from iHeartRadio, visit the iHeartRadio app, Apple Podcasts, or wherever you listen to your favorite shows.

Welcome to tech Stuff, a production from iHeartRadio. Hey there, and welcome to tech Stuff. I'm your host, Jonathan Strickland. I'm an executive producer with iHeartRadio. And how the tech are you. I'm having a rough recovery day. I'm still getting over COVID. I'm no longer testing positive for COVID, but there's been some lingering symptoms and some days are better than others. This is not one of the better days. But it is a Thursday, which means it's in tech News day. So we're going to cover the tech news for October twelfth, twenty twenty three, and like Tuesday's news episode, we're going to start off with a bunch of Twitter slash x news. So first up, I mentioned that a lot of people from media outlets to government agencies to watchdog groups have been directing criticism at X for the proliferation of misinformation and disinformation, specifically in the wake of the ongoing war in Israel. It did not take long at all for bad information to flood the platform after Hamas launched attacks within Israel, and that bad information included everything from photos and footage of past conflicts being passed off as if they were from the most recent circumstances, to some accounts claiming that footage that was taken from modern video games actually represented circumstances in either Israel or Gaza. One critic who raised concerns about the platform in general was Thierry Breton or Breton, I apologize for my pronunciation, but we're talking about a European Union commissioner who said that X was a place for people to quote disseminate illegal content and disinformation end quote. That could mean that X could be found guilty of viol leading the EU's Digital Services Act, or DSA. So the DSA went into effect earlier this year, back in August, and it covers a lot of ground when it comes to a platform's responsibilities. The DSA expects that large platforms ones like x Slash, Twitter or Facebook, or even platforms like Google and Amazon, have to take an active role in content moderation. Unlike the United States, which has Section two thirty, the EU's DSA holds these platforms at least somewhat accountable for the material that users post to those platforms, and the DSA gives the EU government the authority to find companies that failed to comply with the law, they can get hit with a really hefty penalty, up to six percent of their global turnover, which for the purposes of this episode, will just kind of reference as revenue. It's there's a slight distinction there, but you'll get the idea. So last year, Twitter's global revenue was around four point four billion dollars, so the DSA essentially says they could take six percent of that. They could find Twitter six percent of that for failing to comply with the DSA. That would be around two hundred and fifty million dollars. That's not a small fine. Plus. If the EU finds out that a platform continues to fall short of the DSA's requirements, the EU has the authority to suspend that platform's operations within the European Union, which of course would also mean the loss of even more revenue. If you can't operate in a major part of the world, you're missing out on a tremendous amount of revenue. Ex's CEO Linda Yakarino has posted that Twitter has removed hundreds of accounts affiliated with hamas and removed or labeled quote unquote tens of thousands of pieces of content also said that Twitter slash X has not received any information from europoll about illegal content on the platform. She's saying, well, we're being accused of something, but no one has actually come to us with these concerns. It's the first we're hearing about. Essentially, the company is relying on the quote unquote community, which essentially means volunteers who are flagging posts that might violate the DSA or x's own policies. But then X kind of has to rely on volunteers because Elon Musk effectively gutted the content moderation team by laying off most of them in the wake of his acquisition The Federal Anti Discrimination Agency or FADA THODA is leaving X. This German anti racism group explained the decision to bounce off the platform. They claim that incidents of hate speech have quote increased, particularly end quote, ever since Elon Musk took control of Twitter. FADA said that cases of anti trans and homophobic content, and racism, misogyny, anti semitism, and other examples of hateful speech are all on the rise, and that ultimately this forced the leaders of FADA to ask the question, does it make sense to stay on X? Is it providing more of a benefit than a cost. And clearly the endecision was that it just wasn't worth sticking around, or, as Fada's statement puts it, quote in our opinion, X is no longer an acceptable environment end quote. Now I've said this before, but in the years past, Twitter, which by the way, has always been a deeply flawed platform. It's not like Elon Musk took it over and then it went from being a shining beacon to the absolute worst. It had problems well before Elon Musk entered into the picture. It's just those problems have magna fight arguably since Musk took over. Anyway, back in the day, if you will, Twitter was put to effective use by various activists and humanitarian groups and news outlets. It was a scene as a very useful tool and worth the investment of time it takes to create and cultivate and maintain a Twitter account. But Musk's capricious leadership and the massive changes the he's made to X have convinced a lot of those entities just to pull up stakes and leave X behind. They're saying it no longer makes sense to stay there. And now we're actually starting to see what effect this has had on the organizations that decided to leave. So let's take NPR National Public Radio, for example. So in April of this year, NPR made the decision to leave X. This is right around the time that Musk actually renamed Twitter to X. And to be clear, MP made this decision for the official NPR accounts. They did not mandate that for everybody who works for NPR, but they decided they were going to stop their accounts on Twitter or X. The platform at that time had appended a label to NPR, and the label was US state affiliated Media. Now, Twitter's own definition for state affiliated media was quote outlets where the state exercises control over editorial content through financial resources, direct or indirect political pressures, and or control over production and distribution end quote. So clearly that definition is meant to designate media outlets that are serving as a mouthpiece for a government. So the content produced by those types of outlets tends to be heavily biased in favor of whichever government is in charge. I mean, that's clear, right. It's the kind of designation that you might give to a media outlet in places like Russia or China, where you have authoritarian leaders who crack down on any voices that oppose the state. Now, MBR does receive money directly from the US federal government, but it makes up less than one percent of its budget, and it is clear that the government doesn't have any influence on the content that INPR produces. So the crux of what makes a media outlet a state affiliated media outlet wasn't there, but that's what Twitter was labeling in PR. So NBR decided that's it, we don't have to deal with this, We're a gone, and that it leaves. It was clear that Musk had an axe to grind against NPR, So six months later, now that we're in present day, leadership at NBR has circulated a memo to the company's staff. In that memom, the leaders addressed the impact that the organization has seen since leaving X and it's not that much so, according to Niemann Reports, The NPR memo says that traffic to NPR has dropped by only a single percentage point as a result of leaving Twitter slash X. NPR also pointed out that even before all this kerfuffle, traffic from Twitter amounted to less than two percent of all of NPR's traffic, and that explains how the organization found the decision to leave Twitter to not be difficult at all. The message is one that X definitely does not want to hear that staying on Twitter just isn't worth the effort. The negligible amount of traffic NPR got from Twitter was not worth the frustration of operating on Twitter. I suspect other media outlets may also reevaluate their position on the platform, particularly as it deals with more allegations of hosting disinformation and illegal content. But obviously it's not going to be the same for everybody, right, some media outlets may say, well, yeah, but we have a more significant percentage of our traffic coming from Twitter. That's going to be a much harder decision to leave. In fact, I would be surprised if a media outlet chose to leave in the wake of a revelation that Twitter was providing a significant percentage of traffic to the media outlet. But for those that are in a position similar to NPR, I would not be surprised to see more of those outlets follow suit. Okay, we're going to take a quick break. When we come back, we're going to have some AI news, the other big story that continues to unfold throughout twenty twenty three. But first let's take a break to thank our sponsors. Okay, we're back, so let's talk about artificial intelligence in general and generative AI in particular. So first up, open ai has plans to make it less expensive for developers to build applications on top of open AI's models. So, as it turns out, using artificial intelligence, like using it and incorporating it into your work, tends to be really expensive because AI requires a lot of computational resources in order to work, and that goes all up and down. The computational resources gamut from processing power to computer memory to things like storage. Not every developer has easy access to all of those things. It's entirely possible to come up with a potentially brilliant idea on how to incorporate AI into an app, but not be able to afford the costs of bringing that idea to life. So open ai plans to offer options that can mitigate some of those costs, like offering memory storage, for example, which is a huge cost apparently for developers who are working in the field of AI. So that's just part of open AI's strategy. The company also plans to unveil some computer vision tools, which can make it possible for developers to create applications that include some element of image analysis in them. You know, it's pretty easy for humans to describe stuff that's included inside an image. It tends to be a pretty heavy lift for a computer, and the computer vision tools simplify this and gives developers access to capabilities that otherwise they likely wouldn't be able to create on their own. According to various news outlets, the company is going to roll out these features around November sixth. That's during the company's first ever developer conference, which is taking place in San Francisco, California, on November sixth. And this matters to you and me because it likely means we're going to be seeing a rush of new applications that incorporate AI, at least to some extent in the not too distant future. So like it or not, the AI revolution is going to intensify before much longer. Now that's going to have some hefty real world consequences. Even if we assume that every AI application is benign or goodness, let's say that it's even helpful. Let's be super optimistic. Let's say that the incorporation of AI is genuinely helpful in all of its implementations, there would still be consequences we would have to face even with like perfect AI. Tech Spot reported on a research paper written by Alex Devrie titled the Growing Energy Footprint of Artificial Intelligence, and Devree argues that while much research has been focused on how much energy AI language large language models require during their training phase, it's a lot like when you are training a large language model, you're spending a great deal of money just on electricity to power that process. However, less studied, but just as important, arguably, is the amount of electricity needed during normal operations. Like you've stopped training now you're just deploying applications built on top of those large language models. Defrie is saying, well, that's going to still put a pretty hefty demand on electricity generation, especially as more and more of these apps are deployed, are developed and deployed. So an analysis firm called semi Analysis estimated that if Google were to incorporate AI into every single search, which it cannot do. I want to be clear about that, it lacks the physical capability of powering that. But if Google were to try and do that, it would require electricity equivalent to somewhere in the neighborhood of twenty nine point two tarawatt hours on an annual basis, which is pretty much the same amount of electricity that Ireland uses per year. So to power AI enhanced Google searches, you would need the same amount of electricity that Ireland needs for a full year. That's incredible. It also reminds me a lot of cryptocurrency mining. Like we talked a lot about how the process of mining crypto coins, particularly proof of work crypto coins requires a ton of electricity. Proof of stake less so, but still requires some proof of work requires a lot. That's the kind that like bitcoin uses. And we have already talked about how in the past, at least mining operations worthy creating the power need equivalent of what we would see for some nations in Europe on an annual basis pretty crazy. Well, same thing is true of AI, but as I said, Google doesn't have that capability. This is more of a theoretical kind of discussion. Google lacks the equipment that would be needed to be able to do this, and in fact, the companies that make the equipment, like make the processors that would power this kind of thing, they aren't producing chips at that scale, Like they could not produce enough chips in the short term anyway to power that. So this is really again just a hypothetical situation. But the point is that AI is a power hungry technology, not in the sense of you know, sky net taking over the world, but in the sense of it requires a ton of power for it to work. And it's something that people like defriethink developers need to really take into consideration. Does it make sense to incorporate AI into an application knowing the energy demands of that particular application, And if it makes sense, then sure. But if it doesn't make sense, then don't just throw it in there just to have AI as part of your value proposition of your app, because you're going to be costing yourself a lot more money due to the fact that there's so much electricity needed to power the darn thing. Okay, moving on to some other news. Microsoft got some bad news in the mail last month, which we are now learning about due to a filing with the Securities and Exchange Commission up the United States Internal Revenue Service. The dreaded IRS sent a notice to Microsoft that the company had failed to pay a few bills. It was called a notice of Proposed adjustment, which is a, you know, a fairly benign sounding term, but it had some massive consequences. According to the IRS, Microsoft ows about twenty nine billion dollars twenty eight point nine billion to be more specific, and that's in back taxes from the two thousand and four to twenty thirteen period, as well as penalties associated with a failure to pay those back taxes over the years. Microsoft says the whole thing's a misunderstanding. In fact, they say they outright disagree with the Notices of Proposed Adjustment aka NOPA, So I guess they're saying NOPA to NOPA. They're going to pursue an appeals process with the IRS, and if that doesn't go Microsoft's way, they plan on bringing the whole matter to court. It would be shocked if this doesn't ultimately go to court at some point. And yeah, it's going to take a minute for all that to hash out, so we'll have to check back. And I don't know, like five years or so, the State of Utah has filed a lawsuit against TikTok, and the charges are probably pretty much what you would expect. So first and foremost, Utah's Governor, Spencer Cox, argues that TikTok children both by getting them hooked on the platform so that they're spending countless hours on it, and also by serving up harmful material. The lawsuit cites research that links social media activity, like prolonged activity on social media, with poor mental health. However, I think it is very important to note the distinction between causation and correlation here. So does staying on social media for a long time lead to poor mental health? Possibly, But it's also possible that people who have poor mental health tend to use social media for longer than other people do, and that would mean that using social media a lot doesn't necessarily lead to poor mental health. But perhaps poor mental health makes you more inclined to use social media. It's similar to how people who commit violence may also play a lot of violent video games, but that doesn't necessarily mean that playing violent video games makes you violent. Again, there could be a correlation, but no causation there. We don't know, is my point. So I think making legislation on something where there's not a clear causal link is problematic. You're not likely to actually address whatever underlying issue is at stake, But anyway, Alex Horrik of TikTok contested the charges brought against the platform by the state of Utah. He said that TikTok quote has industry leading safeguards for young people, including an automatic sixty minute time limit for users under eighteen and parental controls for teen accounts end quote, So he's saying we already have protections in place. These are here to prevent young people from being harmed by TikTok. According to AP News, the lawsuit also brings up a concern that we've seen many, many times from lawmakers here in the United States that is related to the relationship between TikTok, the American company, and Byte Dance, TikTok's Chinese parent company, arguing that maybe there are deeper links there than what TikTok claims. This is a song we've seen again and again, so not a big surprise here, but we'll keep an eye on where this lawsuit goes and how it unfolds in the future. Got one more than I'm going to tell before we get into another quick break it has been a hot minute since I've talked about FTX, the cryptocurrency exchange that famously went up in flames almost a year ago. Next month will be a year since FTX went under. Fdx's co founder Sam Bankman freed AKASBF is in court to face charges that he committed several fraud related crimes in his role as CEO of FTX. Yesterday, Caroline Ellison, the former CEO of Alimator Research, which was intrinsically connected to FTX, testified that FDx engaged in some pretty darn shifty behavior, including an attempt to bribe a Chinese official. So, as Ellison tells the story, Chinese authorities brows ftx's funds on the OKX and quea by exchanges. So apparently someone at Alibator Research, which by the way, was an investment company that specialized in the crypto market and had been co founded by SBF. Apparently this person unnamed person ended up doing business with some someone else who was suspected of money laundering. So the Chinese authorities locked down around a billion dollars worth of FTX assets, and this prompted folks at FTX to try and come up with various strategies on how to free up that money that otherwise was just locked away. It sounds like first they started off with some fairly reasonable and more importantly legal options, but apparently those did not get any results. So the leaders, including SBF himself, began to consider alternatives, and while they did not necessarily approve them initially, further down the line they did, and that ultimately resulted in offering a one hundred and fifty million dollars bribe to a Chinese official to grease the wheels and get that money unlocked. Not only did the company apparently actually do this, it also listed the expenditure in its State of Alameda documentation. Now, they didn't go so far as to label it bribery costs that might have been a red flag. Instead, it was filed under notable slash idiosyncratic PNL stuff. And here's the wild thing, y'all. The bribery thing isn't even one of the legal charges brought again SBF. Right, that's not even a material component of this trial. The whole story really just becomes part of establishing SBF's character and leadership decisions and willingness to engage in illicit activities. Ellison, for her own part, has pled guilty to several fraud charges already, so things are not looking so great for SBF at the moment. Okay, we're going to take another quick break. When we come back, I've got you know, I got quite a few more news stories, so stick around. We'll be right back. Okay, we have come back, and now for a couple of stories about how the US government is trying to make companies be more transparent when it comes to stuff like you know, bills, service fees. So first up is the Federal Trade Commission, or FTC, which has taken aim at so called junk fees. So these are those hidden fees that won't pop up in a transaction until it's right at the moment of purchase. So the customer might get a price quote on a product or a service and then they'll start the whole process of purchasing it or agreeing to it, and right when they're getting at that last moment to complete a transaction, a bunch of weird fees pop up and the cost increases, sometimes significantly. So this happens in lots of different kinds of transactions. Personally, I see it most whenever I use a ticket broker to purchase a seat for an event, whether it's a show or a concert or whatever it may be, sometimes the fees end up being as much as the cost of the ticket itself, so it doubles the amount I would have to spend if I wanted to, I don't know, go see thump Asaurus performs Strutton Live. Actually, that's probably a bad example because thump Asaurus tends to play small to medium venues. Those are less dependent on ticket brokers. But you get what I'm saying. You can also find it in other places, like vacation rentals, right, so Airbnb has been brought in as a possible perpetrator of these kinds of junk fees or car rentals. It can be like that too, where you'll see one price quoted as you log into a car rental company, but as you go through it suddenly balloons as you get further into the process. Not all car rental companies are like that, by the way, I've noticed that a few have started to already post what the full price would be on that initial screen, and I think that's admirable because not every one of them does that. A lot of them starts off with a price where you're like, oh wow, it's like eighty bucks for the full week. That's not bad. Then you click through and you're like, oh wait, it's actually like two hundred and seventy dollars for the full week once you add all these fees in. So the FTC would force companies to include any additional fees within the advertised prices themselves. They could no longer hide it until the end, and they have to explain in what those fees are actually for, like why does that fee exist? What is that fee going to? Like? Those convenience fees drive me nuts because I'm wondering who the heck is this convenient for, because it's not me anyway. It's also supposed to indicate which, if any, fees could be refundable. So all of this has to be done if FTC adopts these rules. I'm sure the companies are not eager to comply with any rule like this if it is in fact adopted. But as a consumer, I am personally fully in favor of making companies be more transparent when it comes to costs to customers. Now, on a related note, the US Federal Communications Commission or FCC has transparency rules that are going to impact the major Internet service providers in the United States starting on or in April of twenty twenty four. So starting then, providers will have to include a broadband consumer label, and this label will break down the provider's services. It has to include the service's features, its limitations, any restrictions that are in place, as well as the price for the service. So that means the label should show customer stuff like what broadband speed is available in their area, how much is the introductory rate for enrolling in that service, what are the actual rates once you're done with the introductory period. Are there any data allowances or data caps? If you prefer on that service, it would need to have all of that listed out, plus any other fees that would come into place, and that, in theory, should make it easier for customers to do some comparison shopping. However, I should also add that for a lot of folks in the United States, comparison shopping is more of a dream than a reality. It's not a luxury for a lot of people because a lot of locations, including where I live, there's not much choice. Technically, I have two providers in my area that provide for actual broadband speed service, and only one of them can provide service at greater than fifty megabits per second. The other one, I think max is out right at twenty five, which I think is even I think that is the threshold for broadband. So it's not like I even have a different option to go with if I wanted to, at least not one that would work for what I do for a living, So that's kind of a bummer, But at least I would have a better understanding of what I'm spending my money on and what all those fees are going to, and also things like how much would it cost to terminate the service early maybe suddenly I get a different opportunity to use a different ISP, how much would it cost for me to cancel my agreement with my current provider? And as you might suspect, the broadband companies have pushed hard against this move. They've claimed that it would be too difficult to implement, and that really just gives the FCC more ammunition, right, because if listing out all your fees is something that you say is too complicated, there's definitely a lack of transparency going on among other issues. So I don't think it is an argument that does the broadband service much favor. Next up, we've got a story about eBay and a potential two billion with a b dollar fine for the company. So what did eBay do that would warrant such a hefty fine. Well, it all has to do with the facilitation of sales of products that violate environmental laws in the United States, namely defeat devices aka gadgets that, when you install it in a diesel powered vehicle, gives the driver the ability to quote unquote roll coal. So if you've ever seen a car or a truck blast out a huge cloud of black exhaust all in one burd first, that is rolling coal. I had not even heard about that practice until just one day someone did it to me. I had been walking back from our office back when it was at Pont City Market, and it was a three mile walk, and it was a hot day, and I was taking a little break, and as I was sitting down on a bench, a big old pickup truck drove past me and then belched out a ton of black smoke rolling coal. Well, here's the thing. These devices violate the Clean Air Act, and if authorities find a retailer selling them, they can find that retailer five eighty dollars per sale, so it's accumulative, and the EPA has set its sites on eBay. Now, eBay has said the lawsuit is quote unquote entirely unprecedented. Now that might be true. It might be entirely unprecedented, but that doesn't mean that it's unwarranted. It also doesn't mean that the EPA won't continue to do that sort of thing with other companies. But eBay also claims it has actively removed listings for such illegal devices for years and that it has a ban on those kind of devices. Those are not allowed to be sold on eBay, according to eBay's representatives. So we'll have to see where this goes from here and whether or not it actually ends up going to trial, and whether eBay actually has the receipts to show that it has cracked down on this practice, or if, as the EPA alleges, eBay has allowed for the sale of these devices. Google announced it stopped a distributed denial of service attack or d DOS that was the largest ever on record. In fact, according to Google, this attack was seven and a half times larger than the previous record holder, and the attack was blasting targets with three hundred and ninety eight million requests per second. So, just in case you don't know what a denial of service attack is, it's basically a tactic meant to overwhelm a target server a web server. So in its most basic form, a denial of service attack will see attackers send out requests to a target. The target is compelled to respond to the requests because that's how the Internet works. So by sending millions or even hundreds of millions, or even billions of requests to a target, the target can get locked up trying to respond to all these requests. It can even shut down as a result of this. A distributed denial of service attack is the same thing, except the attacks are distributed across a network of attackers. It's not just one source of attack, and that makes it a much more effective means of attacking a target because if the attacks were coming from one source, the target might identify that source and then block traffic coming from that source and then protect itself. If it's distributed across lots of machines, it can be more challenging to actually protect yourself against that kind of attack. Google says it detected the attack and then launched mitigation efforts, and that included reaching out to peers and started to patch a vulnerability in the HTTP slash two protocol stack. The attackers were leveraging this vulnerability and it's present in some but not all servers. So server administrators have been informed they should check to see if their machines are vulnerable, and if those machines are vulnerable, to be on the lookout for a security patch from whatever vendor made that machine in order to close off this route of attack. And finally, for some sky is falling news, the US Federal Aviation Authority or FAA has a warning about all those satellites that were putting up in orbit. So I talked earlier this week about how companies like SpaceX and Amazon are launching thousands of satellites constellation of satellites up into orbit. And the warning that the FAA is giving us is the old saying what goes up must come down, even if it happens to be in low Earth orbit. So the FAA did an analysis and estimates that up to twenty eight thousand pieces of satellite could fall to Earth each year based upon how much stuff we're putting up there, and that these could potentially represent an actual threat to people in property. So the FAA estimates that the chance of being hurt or worse from a piece of falling satellite equipment could be point six per year, and that means that it could possibly happen every two years or so, just when you look at the big picture of statistics. And if that's not scary enough, the FAA also indicated that falling satellite pieces could potentially be a threat to aircraft in mid flight, though this represents an even smaller likelihood. The FAA indicated that regulation could limit some of this risk. So if some regulations were put into place that would restrict how many satellites companies can put into orbit or how they operate those satellites, maybe that would end up helping this. But the FAA also stressed it only has authority to do that kind of thing here in the United States. There are other nations that are launching satellites into orbit, and so there is potentially a need for some sort of international body that could set global regulations on satellites and launches and that kind of thing. And the proliferation of space debris. SpaceX, by the way, says that the FAA's analysis is flawed and that the company's satellites, they're Starlink satellites are far more likely to burn up on entry because they're not terribly big. So they're like, this is a moot point, because if our satellites are falling, they're not going to make it to the ground. They will burn up before they reach that far, and so there would be nothing significant landing on Earth afterward. They've also raised some other objections, so again, interesting, scary. My guess is that there's the truth is somewhere in the middle here. I do think that the odds of being struck by space debris are pretty darn low if you're not in space, probably somewhere along the same lines as getting struck by lightning. Although obviously you can help avoid getting hit by lightning by not being out and about while there's a lightning storm. It's a lot harder to predict when a piece of a satellite's going to come plummeting down, So I get it. It is kind of scary. Now before I go, I do have an article to recommend. This article is titled how AI reduces the World to Stereotypes. It was written by Victoria Turk and it was published in Restofworld dot org. So Turk is mainly talking about generative AI and how text to image systems tend to perpetual wait and amplify biases and stereotypes. So the article follows an experiment in which a team at rest of World use the text to image generative AI tool mid journey to produce pictures associated with specific places or types of people, and then analyze those images and looked for the inclusion of elements that reinforce biases and stereotypes. So I recommend that article again. It's titled how AI reduces the World to Stereotypes by Victoria Turk at Restofworld dot org. That's it for me today. I hope you are all well. Stay safe out there, take care of yourselves both physically and mentally. It is a really tough time right now, and I'll talk to you again really soon. Tech Stuff is an iHeartRadio production. For more podcasts from iHeartRadio, visit the iHeartRadio app, Apple Podcasts, or wherever you listen to your favorite shows.