Tech News: TikTok Strikes Back, AI Poses Threats, and Butts!

I'm an executive producer with iHeart Podcasts and How the tech Are Yet, it's time for us to cover the tech news for the week ending on March twenty ninth, twenty twenty four, and our first story is that FTX cryptocurrency exchange and Alameda Research founder Sam Bankman freed AKASBF faced sentencing for his multitude of crimes relating to playing fast and loose with a whole lot of people's money.

SBF had sought a light sentence of just a few years, while prosecutors were hoping for a forty or fifty year sentence. He could have faced a maximum of one hundred and ten years in the jailhouse, but Judge Lewis Kaplan has decided on a twenty five year sentence, not as heavy as what prosecutors wanted, but way the heck more than

what SBF was hoping for. The judge said that the defense had put forth a pretty lousy effort, plus SBF had engaged in witness tampering, which affected the judge's decision. Now of course, this isn't the end of it all. SBF does plan to appeal both the decision and the sentence, though, considering the issues of witness tampering and such, I think he probably faces an uphill battle on that one. He has done himself very few favors. In the wake of

his empire collapsing. An English data company called Savanta conducted a poll judging American's opinions about banning TikTok, and the results are unsurprising. So according to savantah only around twenty eight percent of all respondents favored a ban on TikTok, and one of the questions they included asked if respondent's friends would try to find ways to continue to access TikTok even if it were to be banned in the

United States. That particular question got a sixty percent affirmative response. I guess this was actually Savanta's way of saying, look, mate, I'm not saying you'd break the law, right, I'm not saying that, but your friends. Right, there's some shady characters, any of them likely to slip around the corner, maybe log into a little VP, maybe do a little scrolling

on TikTok. A anyway. Like I said, the results are unsurprising, but they might very well provide the US Senate with some food for thought, as that body now has the responsibility to take up this issue or to you know, not take it up. They could do that too, They could just not consider it. I still think it's likely that the Senate is not going to pass this bill into lab because the perception is it would be very harmful to their chances of getting re elected, and that

seems to be a pretty big driver in politics these days. Meanwhile, TikTok has invested more than two million dollars in an ad campaign in an effort to pressure the Senate to drop the proposed ban legislation, and they've targeted the ads and battleground states where Democrat senators could face a pretty darn tough reelection campaign if the public were to be

totally cheesed off at them. Specifically, they are running ads in Ohio, Wisconsin, Pennsylvania, Montana, and Nevada, And when you know it, each one of those states has a Democrat senator who's running for reelection this year. So I think senators are already a little worried about the repercussions they would face if they tried to push this bill into law.

But this sort of really drives that threat home. And I have yet to see one of these ads, but to me, they kind of sound like the super sad ads you would see for the ASPCA that feature Sarah McLaughlin singing about how the puppies are now in the arms of an angel or something, Except these ads show supposed TikTok users talking about how the world would really

suck if there weren't a TikTok. Also, I'm a little extra snarky today, speaking of politics, the Biden administration announced some new policies regarding AI and its use by the

federal government. Those policies are that federal agencies are now tasked with ensuring that their use of AI does not put the rights and safety of American citizens at risk, which sounds totally reasonable, but it also sounds like it's ding dang hard to put into practical use, because we've already talked about how there are challenges of just gauging risk and danger with AI before you actually deploy it.

Sometimes you only find out after it's too late. Anyway, the second of the three policies announced is that any agency that is using AI has to disclose that use as well as to publish a risk assessment report in order to maintain transparency, which also seems reasonable to me, though I do suspect some agencies will maybe be a little less transparent than others based off history cough cough. The FBI, and finally, each federal agency must create the

position of chief AI Officer. And further, the person in that position has to be knowledgeable in the field. So no hiring your cousin Bobby so that he has a cushy government job unless you know Bobby is also an authority in the field of AI. Now I'm having a bit of fun with all this, but I do think it is a decent start. It's just going to be a really long road ahead of us, and it's hard to predict if these measures are going to be adequate to mitigate future harm caused by AI. In fact, let's

actually talk about how AI can cause harm. There's a piece by Colin Letcher or Leecher. I'm not sure how to say your last name. I apologize. It's in the markup and it really raised my eyebrows. So late last year the Mayor of New York City, Eric Adams, announced that an AI powered chatbot tool would help New York businesses operate in a way that worked within New York law.

So if a business owner had a question, they could ask this chatbot instead of a real human being, and the chatbot would in theory give an answer that was in alignment with New York law, and the business person could act on that. But Colin says that opposite is actually true, that the chatbot has in fact made suggestions that outright violate certain New York City laws, which seems to be a big old whoop sie, and he provides

examples of the chatbot encouraging some shady business practices. So, for example, in New York City, there is a law that states landlords cannot discriminate tenants by source of income. In most cases. There are some exceptions for very small buildings where the landlord and their family also reside in that building, but otherwise they're not allowed to discriminate no matter where a renter is getting their rent money, such

as if it's from a government assistance program. So the landlord is not legally allowed to deny anyone the opportun unity to rent a home based on that, but the chatbot declared that landlords are not required to accept tenants if they're on rental assistance programs, which does sound like that's advice that violates the law. Further, the chatbot said that restaurant owners are entitled to a cut of server tips,

and that's also not true in New York City. In some cases, bosses can use tips to count toward minimum wage requirements, which I would argue is sort of like taking tips away from servers, but they can't actually dip

into the tips themselves and take a cut. He also includes several other examples in other scenarios, and he says that in some tests the Markup did in their offices, the bot would actually provide correct answers occasionally, so maybe it depends upon how you word the question, although that's not entirely clear. The fact that the bot often provided a misleading or outright incorrect response is beyond troubling, particularly when city government has proclaimed that the bot is a

way for business owners to navigate government issues. City officials have said the bot is merely in a pilot program phase, and the whole idea is that it's going to improve over time, which I sure hopes so, because right now it sounds like it's doing the opposite of what it is intended to do, which you know, it's kind of hard to get around that, and it all just depends upon how you work'd your question. Apparently New York Cities

mayor ain't stopping there with AI. However, he announced that the city is also going to deploy a system made by a company called Evolved Technologies, and it's an AI powered weapons detection system that's made by a company that has been tied with issues relating to weapons detection, namely

problems like false positives or failure to detect. So just imagine that you're trying to commute to work, but then you get stopped by New York Police because an AI protocol running on what is essentially a surveillance system mistaken believes you're packing. So this particular system is similar to a metal detector. It's something that commuters will have to pass through before they can get on the subway, and it uses quote ultra low frequency electromagnetic fields end quote

to detect weapons. The New York City Police Department will be in charge of the program, and according to New York law, they will have to publish a report about the system's effectiveness and impact and sort of like a risk assessment sort of thing once it's been in place for a while. Okay, we're going to take a quick break and when we come back, I've got a few more news stories to get through. We're back, So moving

on away from AI in politics. Last week, Reddit finally held its initial public offering or IPO, which means it has now become a publicly traded company on the stock market. And this week the company saw the stock price drop like a rock after it had surged early on. So like initially, Reddit stocks were doing great when it launched. The stocks price was intended to be forty six dollars

upon the stock market opening. It actually opened above that value at nearly forty nine dollars per share, and it closed at the end of the day at fifty one dollars per share, So at the end of first day of trading, it was already up significantly from where you was intended to launch at. On March twenty sixth, it hit its high of nearly seventy five dollars per share, but then things took a bit of a turn so why did that happen. Well, a couple of disclosures really

seemed to do a number on Reddit stock prices. One was that some risk management companies reported that the stock was being overinflated beyond its actual value, essentially saying this stock does not reflect what Reddit's actual value you is. Keep in mind, Reddit, like a lot of tech companies out there, hasn't posted a profit. It is a revenue losing business. But that doesn't always matter on the stock market. In fact, you could say that often it seems to

have no impact on the stock market whatsoever. But another really big set of disclosures that I think hit that stock price hard is that a corporate filing revealed that Reddit CEO Steve Huffman sold half a million shares when it went public, So he cashed in big time, like five hundred thousand shares. And I should really do a full episode on Steve Huffman because he has an absolutely terrible reputation among Reddit users for various reasons, and there

is a lot to go into. But more than that, Reddit's COOO, Jennifer Wong, also disclosed selling more than five hundred thousand of her shares in the company. And when your corporate leaders are offloading stock, they'll keep in mind both Huffman and Wong still own a butt load of shares. Well. If the leaders are starting to sell off large numbers

of shares, that can really shake investor confidence. I mean it seems to send the message of hey, if the people in charge are offloading their stake in the company, why am I staying in. So shares dropped down to forty nine dollars thirty two cents per share by the end of yesterday, which was below the closing price that they traded at on opening day. Now, today is good Friday, so the stock market is not open today and we'll have to see next week where things go from there.

But over on Reddit there are a lot of armchair stock analysts who are essentially saying told you so, or they're saying this is just desserts. Mashables Matt Binder reports on a study that was conducted by censor Tower, and this study claims that x formerly known as Twitter, has seen a sharp decline in daily mobile app user numbers since Elon Musk took control back in November twenty twenty two.

And by sharp decline, I mean a twenty three percent drop in app usage, so nearly a quarter of all users have dropped off if you were looking at it from that perspective, and that mobile app usage in general is down eighteen percent year over year here in the United States. So in this regard, X slash Twitter is really leading the way among social platforms, because no other social platform has seen that large of a decline over

that same amount of time. X has posted a rebuttal to this report saying that sensor Tower doesn't have access to all of X's data and thus the conclusions are wrong. The reports I've seen have said that since Tower does have access to the data about mobile apps but not website, if you were to go through your browser on your phone and go to X that way, they don't have access to that information, but they do have access to the information of people who are using the actual app

to access the service. And meanwhile, X is also not providing any evidence to back up their own statement that this is all off. So it's kind of a he said, she said situation. At the moment, there's certainly been the perception that usage on X has dropped, at least among traditional users, and that the service has largely devolved into a hive of scum and villainy. That may be my own personal bias coming through. Actually, let's not say that.

Maybe that's my own personal bias coming through, and I fully admit that, And honestly I can only anecdotally talk about this because I got off Twitter last year. Anyway, there's no denying the access encounter more than a few roadbumps since Musk took over the company. Obviously, the excess can continued to struggle with landing advertisers and partners in

the wake of some of the more controversial decisions. Aw Olheiser has an article on vox dot com titled the slow Death of Twitter is measured in disasters like the Baltimore Bridge collapse, and it seems to provide some anecdotal

support for censer Tower's findings. So Olheiser argues that in the good old days, breaking news would proliferate across Twitter in real time, so folks would learn about world events shortly after they happened, and news would spread incredibly far and wide in just a matter of hours, if not minutes. But Olheiser says something different happened when the cargo ship hit the Francis Scott Key Bridge in Baltimore, Maryland, causing

that bridge to collapse. So rather than lots of messages posted to X linking to articles about the disaster, Olheiser says that X was bombarded with conspiracy theorists and folks trying to capitalize on the tragedy through engagement baiting. And yeah, that kind of stuff also happened in the good old days of Twitter as well, Like that's not like unheard of back then, But back then, all of that was kind of a layer that was beneath the legitimate reporting

on world news that was proliferating across the platform. Now, Olheiser says, this underlying layer has sort of been promoted above everything else. It's dominating conversation, and beyond that things got really ugly. Olheiser ends up quoting several Twitter accounts that pushed outright racist narratives in relation to the disaster. I do recommend reading the whole piece on box. It is pretty compelling. Again, the title of the pieces, the slow Death of Twitter is measured in disasters like the

Baltimore bridge collapse. Turning to video games, the publisher Take two Interactive, which c really it's a holding company. It owns a couple of other publishers and developers. They own rock Star that's the company behind the Grand Theft auto franchise. They also own two K, which is a publisher that publishes a ton of different games, including like WWE games that speak to my heart, although I haven't played one in a few years anyway. They've acquired another video game company,

a historic one, a storied one. That is, they have acquired Gearbox from Embracer Group. So Gearbox is the developer behind the Borderlands series of video games, among other things. Borderlands is going to have its own live action film coming out in theaters later this year, so this timing is kind of interesting. The deal was for four hundred and sixty million smackaroos, which is a princely sum. It's not a done deal yet, right, These acquisitions always have

to go through a process. The deal is set to close in June, assuming that regulators don't object to the transaction. Since Embracer Group is kind of like the Kirby of video game companies out there, meaning it gobbles up pretty much every company it sets its sights on, I think regulators might be inclined to see a property, leave embracer groups you know embrace and finally to cover a weird

story about butts and bosoms. Butts and bosoms sounds like a raunchy role playing game when I say it that way,

but you know games are involved in this story. I am talking about Twitch, the streaming service that started off as a way for gamers to stream themselves playing games live on the Internet to their adoring fans, and these days Twitch does a lot more than that, but games are still at the core of the servants, and recently the company had to create a few new policies to fight back over some innovations in the streaming space, namely that certain Twitch streamers are wearing green or blue clothing

so that they can reject their gameplay on their bodies,

typically their butts or their chest. This was the case with Twitch streamer morg Pi, who also came up with a very clever work around where she created a green shirt that had the chest region cut out of it, so that when she keyed out the color, only her head and chest showed up on screen as the player view and everything else was showing the gameplay of the game she was actually playing on stream, So Twitch is now saying that this is against their policy and that

quote content that focuses on intimate body parts for a prolonged period of time will not be allowed. Quote. Twitch has had a very long history of trying to balance out what is popular among users and streamers on the service with what is popular among advertisers, and to that end, the company has made several moves, sometimes humorous consequences, to tamp down on the more salacious streamers who are out there. What a world. All right, that's it for this week.

I hope all of you are well, 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.