Tech News: Advocates Say We Should Pump the Brakes on AI

Apple's technologies while they're designing apps for Apple devices. But it's also an event where the company occasionally shows off new products. Now, typically those are software features and that kind of thing, as opposed to physical hardware. But yeah, WWDC is sort of an event where stuff can be unveiled or launched, usually something about like an operating system or again some features that will enhance apps, but once

in a while they'll show off some hardware. And this year there is a great deal of anticipation that Apple will finally show off its mixed reality headset at WWDC. So the version that Apple is going to show off, if it does in fact show it at all, is a more modest effort than what was originally planned. So Apple's goal was to create an augmented reality headset that looked like a stylish but otherwise normal pair of eyeglasses.

But it turns out that cramming that many components into so small a form factor is a challenge we just can't overcome yet. Gordon Moore might smile and say, just wait a couple of years on that one. So instead, the mixed reality headset is reportedly one that has a display built into it with cameras that would feed live video to the display when you have it in AR mode, and otherwise you'd be using it like a VR headset.

So in other words, when you wear this headset and you're using it as an augmented reality headset, what you're doing is watching a live video feed of whatever it is in front of your face, just like if you were holding your phone up in front of your eyes. The logo for WWDC twenty three doesn't give any clues as to whether or not they will show this off. The logos just a series of colorful, metallic style arcs,

kind of like a Wi Fi signal symbol. Rather, last I heard, analysts expect that Apple's headset will go on sale later this year in the holiday season for a really hefty price tag, like you know, three grand or something Moreover, while the hope was to follow this up with the AAR glasses, that plan is on indefinite hold at the moment. Instead, Apple plans on releasing an updated and perhaps less expensive headset further down the line. So yep, we're gonna be waiting a while till we get those

cool iron Man style eyeglasses. Now let's talk a whole lot about artificial intelligence. So first up, an advoticacy group called the Center for Artificial Intelligence and Digital Policy sent a letter to the US Federal Trade Commission, or FTC, asking that agency to stop any more commercial releases of products that use GPT four, which is the latest of the large language models developed by open Ai. This is interesting because the FTC is about stuff like consumer protection

and ensuring fair trade practices. So what's going on here. Well, the advocacy group claims that GPD four is quote biased, deceptive, and a risk to privacy and public safety end quote. Now the bit that would really concern the FTC would be the deceptive part if a company where to leverage GPT to generate false claims about products or services that amounts to false advertising. The group says that GPT four isn't quote transparent, explainable, fair, and empirically sound while fostering

accountability end quote. These are standards that the FTC otherwise demands when it comes to trade. More than one thousand people reportedly signed this letter. Meanwhile, there's a different letter, this one from the Future of Life Institute, which is a nonprofit organization that says it's dedicated to pushing technology to benefit humanity and to reduce the risk of you know, like existential annihilation. Also, it has a similar message In

this letter. The letter calls for an immediate pause on quote training of AI systems more powerful than GPT four end quote. That pause, the letter says, should be at least six months while people take a closer look and evaluate the risks associated with these AI models. The letter points out that the technology greatly outpaces governments and their ability to create laws and regulations meant to safeguard citizens.

That's true, that's kind of how it always works, honestly, But the point that the letter is making is that the risks of AI could end up being too severe to just do the normal move fast and break things approach to tech and let law catch up later. The letter is an open letter, so it's not like a letter that's being sent to a specific entity. Also, when it was sent, it had more than one thousand, one hundred signatories on it, including some really famous names on there,

like Steve Wozniak at Elon Musk. But at least some of the folks whose names appeared on that list have since spoken up to say they never actually signed the ding dang darn thing, which should definitely at least raise some eyebrows. A chief AI researcher and meta named Jan Lakun is one of the folks who says he never signed the letter, and in fact, he disagrees with the

letter's premise. So something hanky's going on here. I know we can't trust the robots, but now it sounds we can't trust the peeps who say we can't trust the robots. Y'all know, lots of teachers worry about how students could potentially use chat GPT in an effort to avoid doing work themselves, essentially you know, cheating off a chatbot. You know about that, right, Well, according to the information, a former Google AI researcher has said that Google's kind of

doing the same thing, just at a much bigger scale. Namely, this researcher, named Jacob Devlin, has accused Google's AI team of cheating off rival open ais GPTEAM model. The information says that, according to Devlin, the Google team was scraping data from open ais share gpt website in an effort to have training material for its own barred chatbot AI. Google's Chris Papas told The Verge that the company didn't use any data from share GPT or chat gpt to

train Bard. So Google is outright denying these allegations. Now, there is no denying that Google was caught with its artificial intelligence pants down when chat gpt took the world by storm late last year, and Google has scrambled its various teams to try and make up for lost ground.

According to the information, this actually extends to convincing teams from two different AI initiatives under Alphabet that's the parent company that Google is under, and that would be Google's Brain AI team and an independent like subsidiary, Deep Mind.

That typically these two groups are rivals, they don't really work well together, or they don't often work together at all, and the Deep Mind has been rumored to be looking to actually completely spin off from the company for a while now, but at the moment their efforts appeared to be aligned as they try to catch up to open Ai.

Not too long ago, I mentioned in the news episode that Google was under investigation for allegedly destroying evidence relating to its legal fight against you know, a whole bunch

of folks that includes Epic Games. That's the company that's behind Fortnite, the Match Group attorneys general for thirty eight of the fifty states that we have here in the good old US of A. The lawsuits are focused on Google's policies with its app store, arguing that the company has engaged in anti competitive and exclusionary practices that hurt well,

pretty much anyone who ain't Google, really. But as to evidence tampering, those accusations have led to a federal judge ruling that Google did intentionally destroy evidence and so it must be sanctioned. The judge stated that quote Google intended to subvert the discovery process and that chat evidence was lost with the intent to prevent its use in litigation and with the intent to deprive another party of the

information's use in litigation quote. So essentially, what the judge is saying is that Google purposefully did not instruct any document custodians within its company to turn on the chat history feature. So by default, Google's chat deletes its chat history after twenty four hours, but you can turn this off, and if you do turn it off, then it maintains

that chat history. So what the judge is saying is that Google tried to be coy and deny that it was hiding evidence by not admitting that the chat histories, which could be you'll saved and used as evidence, were set to automatically delete, and that you know, that was just a regular policy, there was nothing wrong with that, and that, of course they didn't even think to tell their custodians to turn on chat history. As to what punishment the judge will meet out to Google, that has

yet to be decided. For one thing, until the judge can figure out what and how much was deleted, it's actually pretty hard to determine the extent to which Google's actions, or lack thereof, in the case of not turning on chat history, have actually hurt the prosecution's case, but in the meantime, Google does have to cover the legal fees

for the prosecution. Google has protested this ruling, saying its policies fall under the restrictions of what is reasonable under federal rules, and of course this is just a corollary to the actual legal case, which itself will continue. Okay, we're gonna take a quick break, but we'll be back with more tech news in just a moment. While Google defends itself against accusations of being anti competitive, it is

accusing Microsoft of being anti competitive. In this case, Google says that Microsoft's cloud computing businesses are designed to discourage competition, and further that Microsoft's recent negotiations with cloud vendors in

Europe are going to make things worse. The arguments are giving me a lot of deja vu, actually, because essentially the critics are saying that Microsoft ties its various products together in such a way that's very hard to mix and match with other vendors, Like you can't easily and seamlessly incorporate whichever products you want from whichever vendors you want.

Because the various systems are designed to interlock together. That Microsoft's approach is really intended to keep customers within the Microsoft ecosystem, relying on things like Windows for on premises computing systems and Azure in the cloud. And the reason I say this gives me deja vu is that back in the nineties, Microsoft was accused of doing something really similar by tying its web browser, which was Internet Explorer

at the time time, with the Windows operating system. This practice was seen as anti competitive because it discouraged Windows users from seeking out alternatives to Internet Explorers, such as Netscape Navigator. This whole thing led to a massive antitrust lawsuit against Microsoft. In fact, at one point, the United States government even ordered Microsoft to break up into smaller companies,

but that ruling was subsequently reversed. Anyway, Google says Microsoft is following that old playbook, only now it's looking at enterprise customers and cloud services instead of regular consumers, and Google is hoping that regulators and the EU will take a closer look and perhaps intervene. Also fun fact, Google is in third place when it comes to being a cloud service provider. First place goes to Amazon, then Microsoft, then Google. So would Google behave in a way that

was more competitive and would not be exclusionary? I guess we'll have to wait and find out our old pal. Sam Bankman freed AKASBF as another charge against him. Now do you remember him. He's the co founder of FTX, which used to be the second largest crypto exchange in the world before it had a spectacular collapse late last year. SPF already faces multiple securities and fraud charges, but this

new one is different. This time, he's accused of having bribe Chinese officials to unfreeze assets connected to his cryptocurrency companies. This appears to have happened before the collapse of FTX, so apparently Chinese officials froze the assets as part of a probe into quote a particular Alimeda trading counterparty end quote. The size of the bribe is a staggering forty million dollars yauza. Think about that for a second. Think about having so much money tied up that you're willing to

spend forty million dollars to get that money. Meanwhile, I'll just be over are just trying to imagine what forty million dollars looks like. Anyway, This means OLSBF is now looking at thirteen major charges in the wake of his FTX disaster, and he's still going with a not guilty plea, despite the fact that his FTX co founder Gary Wong has already pled guilty to charges of wire securities and

commodities fraud. Moving on to TikTok, there's a small stable of politicians who oppose the growing calls or a nationwide ban against the social video platform here in the US. The newest member of that group is Senator Rand Paul, a Republican politician who is one of the few conservative voices to speak out against a ban on TikTok. Paul has expressed concern that a band could lead to restrictions

on free speech. He's also expressed concern that should the government ban TikTok, young folks will sour on the Republican Party and the GOP will find itself out of a job. So this isn't all about standing up for the rights of citizens here. There's a healthy dose of self preservation going on. Now. I've said in the past that I do have real concerns about TikTok, and I still do.

But I also have real concerns about data security and privacy in general, and as many others have repeatedly pointed out China doesn't exactly need an app on the inside of the US in order to get hold of our data. China could just do what the FBI does and buy

data from data brokers. So yeah, the problem with security and privacy is much, much, much larger than TikTok itself, and really we need to have an approach that gets to the source of the problem, which, in my opinion, is our irresponsible approach to personal data, which extends all the way to matters of national security, and that the app is just built on top of this, this massive vulnerability we have. It's not the cause of that vulnerability,

it's just potentially exploiting it. One of the proposals that could potentially ban TikTok is called the restrict Act. Now I've yet to read this bill myself, but according to Motherboards sources, the language in it is too vague and too broad and could lead to massive problems down the road. So the purpose of the bill is to give the US government the authority to restrict communication services that have connections to foreign countries. And by connections, I think what

is implied is at least foreign ownership. But as Motherboards Joseph Cox points out this could potentially extend not just to social networks and messaging apps and that kind of thing, but perhaps to other stuff like VPNs or virtual private networks. Apparently, the bill says that technology linked to a quote unquote foreign adversary would be subjected to this proposed restriction. That includes countries like Russia, China, Cuba, Iran, North Korea, and Venezuela.

And it also covers a lot of different kinds of software, including stuff like games and payment applications. Now, I hope to read the full bill soon, even though reading proposed legislation often makes my head hurt or it puts me to sleep. But the objections here sound like there are some real legit concerns that the bill could lead to overreach, and worse, failed to address the root cause of the issue, so we would still have the problem there, and this

bill would not solve the problem. It would just cause more issues. EA aka Electronic Arts is joining the ranks of companies like Amazon, Google, Meta, and Twitter in that it's laying off employees. According to The Wall Street Journal, the layoffs will affect around six percent of the company's workforce. That means around seven hundred eighty employees could be shown

their walking papers. According to a blog post published by e A CEO Andrew Wilson, one of the reasons for the layoffs is that EA is looking to reduce its actual office space, and it means that some folks just won't have a place to work at anymore. Those affected are promised severance pay and healthcare support as they transition, which is the decent thing to do. How this will affect game development remains to be seen, but at the risk of sounding like a broken record, sure do hate

seeing more folks getting laid off. Finally, Valve, a different video game company, has announced that beginning on January first, twenty twenty four, it's popular game platform and online store, Steam will no longer work on machines running on Windows

seven or Windows eight. Microsoft itself stopped supporting those operating systems earlier this year, so Steam is actually a little bit behind on that regard, But this means that gamers are still using older versions of Windows to access their Steam library of games will need to upgrade before next year or face the very real prospect that they won't be able to play their Stardo Valley anymore, which y'all, Star Doo Valley is a fantastic game. No one out

there wants to be deprived of their Stardus Valleys. All that being said, as you would probably imagine, this is going to impact a very small number of Steam custom According to Valve, fewer than two percent of all Steam customers are on platforms using Windows seven or Windows eight. But if you are time to upgrade, all right, that's it. That's the news for Thursday, March thirtieth, twenty twenty three. I hope you are all well out there, and I'll talk to you again really soon. Text Stuff is an

iHeartRadio production. For more podcasts from iHeartRadio, visit the iHeartRadio app, Apple Podcasts, or wherever you listen to your favorite shows.