Apple Upgrades MacBook Pros with M5 Pro and Max - DTNS 5217

0:00Tom Merritt: Busy work weeks can leave you feeling drained. Prolon's 5-Day Fasting Mimicking Diet works at the cellular level to rejuvenate you from the inside out, providing real results that include fat-focused, sustainable weight loss with no injection needed. Next Gen builds on the original Prolon with 100 % organic soups and teas, a richer taste, and ready-to-eat meals. Developed at USC's Longevity Institute and backed by top medical centers, Prolon supports biological age reduction, metabolic health, skin appearance, fat loss, and energy. Get 15 % off plus a$40 bonus gift when you subscribe at prolonlife.com.

0:39Tom Merritt: That's prolonlife.com. Hi, this is Matt from P1 with Matt and Tommy, and this episode is sponsored by eBay. The cars you'll find on eBay are just different. They come with a story that you can't wait to share.

0:53Dr Niki: Like this 1973 Dodge Charger on eBay that has been tucked away in an Arizona barn for over 40 years, only 55 ,000 miles, and somehow in great running order. It even has a rare sunroof. Suddenly, a car that was hidden for decades is being delivered in just a few clicks with eBay's secure purchase. All the paperwork handled. There are thousands of cars on eBay, from unique finds like the Pontiac Grand Prix SJ to daily drivers. and now with a new way to buy them. eBay, things people love. With almost half a million customers and over a trillion dollars of secure payments, Bill isn't new to intelligent finance.

1:34Dr Niki: It's the proven way to simplify Bill Pay and maximize cash flow. Want to learn more? Visit bill.com slash proven for a special offer.

1:43Tom Merritt: Youth mental health is a complex challenge that requires comprehensive solutions. We must strengthen after-school programs. We must make digital literacy tools available in our schools. We must work with mental health professionals to support children. And we must empower mentors, educators, and parents to keep kids happy. Learn more about our commitment to finding lasting solutions at empowerourfuturecoalition.com slash solutions. Paid for by the Coalition to Empower Our Future.

2:21Jason Howell: this is the daily tech news for tuesday march 3rd 2026 we tell you what you need to know give you the important context and help each other understand oh my gosh yes today dr nikki is

2:32Dr Niki: going to help us understand what we know and don't know about the effects of llm use on mental health and of course apple updated its macbook pros so you get to watch me kvetch about which one i should

2:43Jason Howell: by yeah it's kind of like where we rub our hands together and go i'm jason howell i'm don't mayor let's start with what you need to know with a big story

2:56Jason Howell: delicious new apple devices apple's mark march stravaganza i don't know if anyone's calling it that but i certainly am continues today the company refreshed its high-end macbook pro lineup with new M5 Pro and M5 Max chips. And this is a little bit of spec soup. I don't know how you do this otherwise, so apologies. But get hungry. The M5 Pro scales from 15 to 20 CPU cores, 16 to 20 GPU cores. Now, keep in mind, CPU, that's kind of faster processing. GPU, faster video processing. So that's kind of what you would see out of those increases. Up to 48 gigs of unified memory. The M5 Max standardizes on an 18-core CPU paired with 32 or 40 GPU cores for heavier Pro workloads.

3:47Jason Howell: It's really targeting Pro workloads overall. The entry M5 MacBook Air jumps$100 to$1099, so$1 ,100. Doubles its base storage to 512 gigs, along with 16 gigs of RAM. There's the 14-inch M5 MacBook Pro. That starts at around$1 ,700 with 1TB of storage. The 14-inch M5 Pro starts at$2 ,200. And the 16-inch at$2 ,700, both up$200 with 1TB standard. 24 gigs of RAM on the base 16-inch model. I told you, spec soup, everyone. The M5 Max starts at an eye-watering, at least my eyes watered,$3 ,600 for the 14-inch or$3 ,900 for the 16-inch with 2 terabytes storage, 36 gigs of RAM. We're never going to need more than that RAM, right?

4:43Jason Howell: That's just like, that's the most we'll ever need. That's what I say every time it's large like that. Inside, Apple is leaning on a new N1 wireless chip. That supports Wi-Fi 7, also Bluetooth 6. That lays a nice pathway for next-gen routers and peripherals. Why don't we stop there because there's more, but I want to give your brains a chance to adjust all those numbers I just threw at you.

5:06Dr Niki: Yeah. I mean, honestly, the reason we give the numbers is a lot of you do pay attention to that and want to know how much storage I'm getting. And it's significant that it's an upgrade in storage to a terabyte minimum here. The RAM, don't forget, this is unified RAM. And this works differently than the sticks of RAM you put into a PC, I have found that it kind of is all you ever need. I do not run into a memory problem. I have not thought about upgrading RAM in the five years since I bought my M1 MacBook Pro, which has 16 gigs of unified RAM. So 36 sounds luxurious. Now, Now, I say that and I welcome the email at feedback at dailytechnewsshow.com from the AutoCAD user or super video editor that's like, well, I ran into the limit.

6:01Dr Niki: So I actually am curious, who of you does run into that limit? But these unified RAM numbers are deceptive because it's a different way of using things. And because of the integrated nature of the processors, they work a lot better. I have been trying to rationalize needing a new MacBook Pro. Does that happen to you every single year? Like I said, it's five years old and it's an M1 Pro and it works great. I occasionally now, occasionally run into it being a little slow or being a little wonky on something just within the last six months, really. So I look at this and I think, you know what? But I should go.

6:46Dr Niki: If I'm going to go, I should go now because it might be another couple of years before they upgrade the MacBook Pros again. And by then, I think I will need a new MacBook Pro. So I don't want to get caught. I don't want to get caught between a cycle, Jason. No, who wants that? Right?

7:02Jason Howell: Yeah, absolutely. I mean, if you're going to get one, get it at the beginning. Get it at the beginning, not at the end. Please tell me the answer I want to hear. Yeah, I'm just telling you that you need to buy some new computers, Tom. Not even one. Just buy one of each. I don't like these prices, though.

7:16Dr Niki: And I was actually very surprised that we didn't see any price increases on the new iPhones and iPads that were announced on Monday. I'm not surprised to see these go up. And they go up about what I would expect, given just general price increases, not even just Ramageddon, but just the fact that parts of materials are more expensive than they used to be. And you're seeing new silicon that can do more stuff. So I don't love the price, but I do think these are good machines.

7:48Jason Howell: Moments like these, though, just prove to me how weird time is because I swear I just bought this MacBook Pro. Yeah. It is an M1 Pro, and I'm looking at the data on it.

8:01Dr Niki: Oh, you're in the same boat I am then.

8:02Jason Howell: Yeah, totally. I swear I just bought this, and it was five years ago at this point. I'm like, damn it. Here I am again. oh i'm not i'm not going to do it though i can't do it not right now maybe what i need instead tom is a new display apple refresh that'll do it yeah let's take a look at those the 27 inch studio display while keeping the panel the same at 5k ips lcd 60 hertz refresh uh that's priced at the same 1600 mark but there's upgrades under the hood so it's very much the same except under the hood it gets a new chip that improves the 12 megapixel center stage camera also an updated thunderbolt 5 setup so improvements there and then apple also announced the new studio display xdr this is a 27 inch 5k display with a mini led panel adaptive sync so you get 120 hertz adaptive refresh rate that starts at 3 300 with an optional 300 nanotexture matte finish so you know if if i If I can't quite swing a new computer, maybe I should just get a$3 ,300 monitor.

9:10Dr Niki: Maybe you should drop$3 ,300 on a 27-inch monitor. Yeah, I don't think so. I know people who swear by these, and they say they are the best monitors they've ever owned. I, on the other hand, have never been tempted to buy an Apple monitor. They're always priced well out of the range that I think is reasonable. And I know people say the same about the MacBook Pro. They're like, you can get a similarly specced Windows machine for quite a bit less. I want to say apples and oranges, ha, ha, ha. But it is hard to compare the laptops. It is easier to compare the monitors to me because there's a whole ecosystem involved in the laptops.

9:49Dr Niki: And there's now the Apple Silicon, which is hard to compare because are you comparing it to ARM or are you comparing it to x86? Whereas the monitor, it's like, yeah, these are great. And maybe that nanotexture matte finish is something only Apple can do really great, but I have to pay an extra$300 if I even want to get that. Can I get a mini LED panel at 120 hertz refresh rate that's 27 inches for less than$3 ,300? You bet I can.

10:14Jason Howell: Absolutely. And it probably also depends on what you're using that monitor for. If you work in, let's say, graphic design and color adherence is the most important thing. Like, personally, I don't even know if these monitors would be considered in that category of best of the best. But there's so many monitors out there, so many solutions. And, yes, most of them are far less expensive and probably going to get you 99 % of the way there would be my guess.

10:42Dr Niki: So I also welcome that email of the person who's like, no, the Studio Display XDR is superior because… Right. …feedback at DailyTechNewsShow.com.

10:52Jason Howell: Love to hear it. And then finally, before we move on, very briefly, a regulatory filing appeared for a MacBook Neo on Apple's site, which confirmed the name, not any details about the product itself. So we don't have any more details to share other than that name. And that entry was deleted shortly after appearing on the site. So we were not meant to see it, apparently, or were we?

11:14Dr Niki: Spoiler. Yeah. If you're listening to this on Wednesday morning, you probably already know the details of the Mac with me, or I could look them up because I'm pretty sure that's getting announced on Wednesday.

11:26Jason Howell: So stay tuned for that.

11:27Dr Niki: Well, also stay tuned as we thank our folks who make the show possible. If you see a patron on the street and you're not a Patreon supporter yourself, thank them for making the show possible. They're the ones that are bringing it to you. We always like to thank some of the folks at the top end. So thanks. Howard Yarmisch, John Atwood, and Pat. You guys are the best.

12:16Tom Merritt: and do nothing because it's so darn cold out there. That's why many people are turning to Bombas, whose pillowy plush slippers and warm merino wool socks have been said to be the most comfortable in the history of feet. Bombas products have been found to boost coziness by up to 1 million percent. Okay, enough fake statistics, but could Bombas socks and slippers really be the cure? Go to bombas.com slash audio and use code audio for 20 % off your first purchase. That's B-O-M-B-A-S dot com and use code audio. Youth mental health is a complex challenge that requires comprehensive solutions. We must strengthen after-school programs.

12:50Tom Merritt: We must make digital literacy tools available in our schools. We must work with mental health professionals to support children. And we must empower mentors, educators, and parents to keep kids happy. Learn more about our commitment to finding lasting solutions at empowerourfuturecoalition.com slash solutions. Paid for by the coalition to empower our future.

13:12Dr Niki: A password manager should be the first security purchase you make for your team. Why? Because compromised passwords are the number one way bad actors attack companies, and small businesses are their favorite targets. But unlike a lot of security challenges, passwords actually have a pretty simple solution. One password lets you manage all your business's credentials, so you can feel confident that your data stays secure as your company grows. Find out more at 1Password.com slash special offer and start securing every login.

13:44Tom Merritt: When you manage procurement for multiple facilities, every order matters. But when it's for a hospital system, they matter even more. Grainger gets it and knows there's no time for managing multiple suppliers and no room for shipping delays. That's why Grainger offers millions of products and fast, dependable delivery. So you can keep your facility stocked, safe, and running smoothly. Call 1-800-GRAINGER, click grainger.com, or just stop by. Grainger, for the ones who get it done.

14:18Jason Howell: There is more we need to know today. Let's get to the briefs.

14:21Dr Niki: Yes, indeed. More going on in that drama between Anthropic, OpenAI, and the Department of Defense. OpenAI, Sam Altman, publishing a memo that he had sent to employees on X, so that we can all see it, saying that the company has amended its agreement with the U.S. Department of Defense. That's the agreement that allows the Department of Defense to use open AI systems on classified systems. One passage of the agreement reads, according to Altman, consistent with applicable laws, including the Fourth Amendment to the United States Constitution, that's the unreasonable search and seizure one, National Security Act of 1947, the FISA Act of 1978, the AI system shall not be intentionally used for domestic surveillance of U.S.

15:09Dr Niki: persons and nationals. And so that's the part that says you got to follow the law when you're using our systems for the avoidance of doubt. The department understands this limitation to prohibit deliberate tracking, surveillance, or monitoring of U.S. persons or nationals, including through the procurement or use of commercially acquired personal or identifiable information. So this does leave room for monitoring foreign nationals, right? So that's one loophole people have pointed out here. But it says you can't go to a data broker, get a bunch of commercially acquired personal information, and then use ChatGPT to analyze it.

15:49Dr Niki: You can't do that under our terms. He also confirmed that OpenAI products will not be used by intelligence agencies, so the NSA, CIA, cannot use these unless they modify the contract. That implies that OpenAI might have a few more rules for intelligence agencies to use this. Altman said he does regret that they rushed to sign this contract over the weekend because they were trying to de-escalate the fight with Anthropic, which then he said kind of made us look opportunistic, which I would say, yes, because you were being opportunistic. Yes, absolutely. You may have also had a positive reason for doing this, right?

16:27Dr Niki: But, you know, you also got that$200 billion contract. He reiterated to the government that Anthropic should not be designated as a supply chain risk. He says he opposes that. And he actually hopes that Anthropic and the DoD can sign a deal like OpenAI's. He's trying to say, look, this deal could be a template. It achieves the prevention of the concerns that Anthropic had, at least in his opinion, and makes the DoD happy. There's no reason you two can't work out your differences. As we mentioned yesterday, there have been a lot of threats to cancel the contract. There have been threats to declare them a supply chain risk.

17:04Dr Niki: Neither one of those things are confirmed to have happened yet. The only thing that we know publicly is that the U.S. Treasury Department, Secretary of Treasury Scott Bessent, has said they have stopped using anthropic products, but it's a little easier for them to do it, I would suspect, than the Department of Defense. So following the president's instruction that nobody in the government should use anthropic, the U.S. Treasury Department has said we stopped, but nobody else has said that, including the Department of Defense, which continues to use the products for its ongoing operations.

17:37Jason Howell: Yeah, well, I'm happy that you were in charge of this story and kind of spelling this out because it's been a dizzying one for me to follow, to be honest. But I'm really curious just kind of about the potential fallout or the kind of perspective of the everyday user here in the U.S. and all over, like what this says about OpenAI potentially, or maybe not even necessarily what it says inherently about OpenAI, but what people interpret it to say. Is this Sam Altman saving face? Is it him being truly altruistic for the kind of the AI marketplace and his competitors and everything? Is it even going to matter to people if you've signed a contract whatsoever?

18:31Jason Howell: Some people are going to be like, that's a bridge too far. You know what I mean? I think I get lost.

18:37Dr Niki: I think there's a lot of fear, uncertainty, and doubt around all of this that people suddenly decided to have an opinion on because it was in the news. And the fact of the matter is Anthropic was the first to sign and for a long time the only to have signed a contract for classified information, and none of you cared. So if you have a problem with these companies working with the military at all, that is not going to end up being something that I think will happen. I don't think you'll keep everyone from working with the military, including Google and including Meta. I think everybody likes to try to draw things in stark black and whites of us and them.

19:16Dr Niki: And the fact is that's not how it works on the ground. So I would caution everyone to pull away from the really lazy, simplistic analysis of stuff where it's like, so who's good and who's bad? And look at like, OK, what is OpenAI really trying to do with this? What are they trying to provide? Because if I'm a military service person, I want to be able to use these tools for legitimate purposes to help my job just the way you would want to use them in your job. Now, if you're against using AI at all, I feel more sympathy for you. Like, oh, well, you don't want anybody to use them. And then that's fair.

19:53Dr Niki: But why should a person who is drawing up contracts and tracking inventory at a PX not be allowed to use these just because there's public opinion that's swaying one way or another? So I think it's fine for there to be contracts, but they think they need to have these red lines. A lot of people are spending a lot of time trying to figure out what OpenAI's contracts really do and do not prevent. We talked a little bit about it yesterday, and there's a couple of good links yesterday of good analysis to that if you really want to dig into it. A lot of this does have to do with setting up lines and hoping they don't get crossed because there's what the contract says.

20:35Dr Niki: There's what the judge might say if the contract is violated and you go to court. And there's what the person in the room can convince the other person to do or not do. And the Department of Defense is far from wanting to have these things autonomously being lethal, but they don't want to rule that out if it's legal. And so that's what the fight with Anthropic was about. I think what I'm more concerned about is that the government does not follow through on its threat to declare Anthropic a supply chain risk because that is the government killing a business. If anybody doesn't understand what that means, if you declare them a supply chain risk, no one who does business with the government in any way can use Anthropic.

21:18Dr Niki: Now, Anthropic could survive on selling its business overseas or selling its business to companies that never do business with the government, which there are plenty of. But it would greatly reduce it because so many companies do business with the government, not even on the military side of things, that it would significantly reduce competition. And if you believe Sam Altman is just greedy, then Sam Altman should be out there going like, well, bad for Anthropic. I guess they're a supply chain risk. Not saying don't do that. The reason that he's saying don't do that is not because he and Sam Amode have agreed to kiss and make up and like each other, because I'm pretty sure they haven't.

22:03Dr Niki: It's because he understands that if it happens to them, it could happen to him. And it's not a reasonable thing to do. Domestic companies are not declared supply chain risks because domestic companies are too careful. That is held for companies that are actually engaged in espionage, which is almost entirely overseas companies, usually Chinese companies, but Iranian companies, North Korean companies, a few Russian companies. And so it is irresponsible to apply that to a company that isn't engaged in espionage and even more irresponsible to apply it to a domestic company that is not engaged in espionage.

22:40Dr Niki: And I think that's why Altman is saying that you shouldn't do that. If they do go ahead and call it a supply chain risk, Anthropic will take it to court and say we're not and likely get a judge to rule. Well, they are not a supply chain risk, so we're going to throw that out. But that would not be a good situation either because they would be in this unsettling area where companies would not want to sign new contracts with them until it got clarified.

23:07Jason Howell: Love it. Thank you so much for the context on that. Yeah. Wow. Okay. Now, in a different story, but that has nothing to do with Sam Altman, even though it kind of sounds like it because the contractor is called Sama. A contractor called SAMA does data annotation for Meta to help train models that make its smart glasses work better. And a Swedish site called SVD has a story, a very interesting report, of workers for the contractor describing sometimes seeing intimate moments, private information like credit cards, that sort of stuff, bank details in the videos that they annotate. The question here is why these videos are sent to be annotated.

23:49Jason Howell: Former Meta employees told SVD that an algorithm is used to sift sensitive data, but also doesn't always work. So faces are meant to be blurred in videos and images that are used for annotation, but that blurring doesn't always work as intended. So sometimes they see everything. All videos may be used in this way. Meta's terms say, quote, in some cases, Meta will review your interactions with AI, with AIs, including the content of your conversations with or messages to AIs. And this review can be automated or manual. That is human. And Meta says a user should not share information, quote, that you don't want the AIs to use and retain, such as information about sensitive topics.

24:34Jason Howell: And I think, you know, a big part of this is like, I know for myself, like I do have, you know, meta, they aren't the Ray-Bans, but they're the Oakleys, which actually plug, I'm going to be talking about on Live With It tomorrow or sometime in the future. but when we're wearing these glasses, I don't know that there is any sort of real easy thing to point to, to say, oh, when I record video here, it's going to this place over here where someone with human eyes is actually putting their eyes on this content to do things like say, oh, that's a fire hydrant or, oh, that's a book or whatever to help identify and improve their systems, which is what this report shares is part of the reason why this system is in place.

25:24Jason Howell: And it's interesting that these glasses could be passing through things that you might not have even intended to send to. You put it on the counter, and apparently it's possible that that's passing through information to someone viewing it. It's a very interesting report.

25:40Dr Niki: Yeah, I mean, there's so much to sift through on this, right? Because this reminds me of all of the uproar about Amazon reviewing your recordings and hearing things that were unintentionally set. In these cases, some of these things are intentionally recorded. Uh, cause I think we could have one conversation over, uh, I, I set my glasses down on the table and forgot that I had hit the button to record and didn't stop it. Uh, you know, that, that could happen. It probably doesn't happen that often. Not every video is used for training. Uh, if you're wondering why these videos need to be used for training, it's just model improvement.

26:22Dr Niki: Every model does this. Uh, it's just a way to refine it and make it work better. It's not like Meta's model isn't good, so they have to do this. This is pretty standard stuff. I do think this is a lot different than smart speakers in the sense that people want to be able to use these to record video the way you would record video on your phone that you would never send anywhere. Uh, but because these are recorded on your smart glasses and because they're smart glasses, you may be saying, Hey, meta AI start recording now that makes them subject to like, Oh, he used the active word. Uh, let's see if it, if it did well, uh, let's use that, that video to help train our, our, our system.

27:11Dr Niki: and you need to have a better way to say, no, I am doing something private right now. I want to have this video. I have the right to have this video. I don't want it to be seen by a contractor in Kenya. And the contractor in Kenya probably doesn't want to see it either, frankly. So when I look at this and it says, well, there's an algorithm to sift that stuff out, but it doesn't work well. I'm like, yeah, that tracks. These algorithms are not a hundred effective. Even if they're 99 % effective, that 1 % could be enough for SVD to get a few people to talk to them and say, yeah, I saw this weird thing.

27:45Dr Niki: So I think meta needs to do better and needs to make this very clearly like, I want to have a setting that says this never gets sent anywhere. This does not go outside of my own realm of data.

27:59Jason Howell: And what is that? Because when you're talking with meta inherently you're talking to the ai in the cloud it's not necessarily running on device right it's passing through your phone into the cloud and so everything that's hey but

28:14Dr Niki: not everything that goes in the cloud has to be sent to kenya's to no no totally verification so there should be a way for me to say yeah process this in the cloud and then get rid of it immediately and store it on my local device uh and then of course you have to trust meta to do that if you don't trust meta to do that you probably shouldn't be using these glasses maybe you shouldn't get the

28:31Jason Howell: Glasses. Yeah.

28:33Dr Niki: And I know a lot of people like, that's why I'm not buying those glasses. I totally hear you. But that should be an option that companies can offer, not just meta that everyone could say like, oh, yes, if you mark this, like, I don't want my videos to be used for training, then we will respect that. Meta does have a setting that says, I don't want these videos to be used for training. But then it also has a setting that says, however, when you invoke the AI, that is always going to possibly be used for training. And I don't think that's an acceptable way to do it. No, don't like that. All right.

29:06Dr Niki: Google, however, has made its Google Home Gemini more acceptable for all of you. Google is rolling out a bundle of big fixes for Gemini for Home that might make home devices a little less finicky. The update softens the edge of Gemini, brings those devices closer to capability with their former assistant-driven versions. Some changes include better handling of room commands and knowing and understanding user intent. So if you say, turn off the kitchen, it doesn't turn off all the smart outlets and then turn off your refrigerator or something. I don't know why your refrigerator would be in a smart outlet, but you get what I mean.

29:44Dr Niki: It only turns off the lights because that's what you mean. It understands, oh, they just mean turn off the kitchen lights. Improved targeting to impact only the intended room or category. deeper understanding of the devices themselves, better localization to control just home devices instead of those at other locations. If you've got multiple locations in your account and you're at home and you say, turn off the lights, you don't want to turn off the lights in your parents' house just because they happen to be on your account. Also, Gemini less likely to cut you off mid-sentence when you're talking to it.

30:14Dr Niki: And just generally things are more reliable, daily tasks like notes, reminders, and timers in there as well. Jason, on the smart home side, Google is adding live search for home premium users. So you can answer questions about the Nest cameras, what you see in real time. And the devices are getting wider routine support for trigger actions based on contextual awareness, a little better understanding of what's going on when you make that request. Does this address most of the people's complaints, do you think?

30:43Jason Howell: Well, I think it certainly gets us closer. I mean, I think the big complaint for a lot of these home devices of which our house is littered with them is that we had a lot of these things working really well when Assistant was the thing. Clearly, Google doesn't care about Assistant, is actively pushing it out and getting rid of it and maybe already has entirely replacing it with Gemini. The potential of Gemini as a replacement has always been, in my opinion, a whole lot greater. And I think that's part of why Google's leaning into it. But the kind of transitionary period this last couple of years has been really bumpy.

31:18Jason Howell: And things like what this fixes, you know, those like understanding that when you say turn off the kitchen, it doesn't mean turn off literally everything in the kitchen. It means the lights. These were problems that weren't problems on Assistant and then were problems with Gemini, and now they're not. So thankfully, I think hopefully this addresses these devices and makes people want to use them again. Because I know in our house, we've really, really scaled back on our use of them, if at all, like very, very minimal use. And I know Google wants to make it more appealing again. And so they have to button up these things.

31:54Jason Howell: I think these are very welcome changes.

31:57Dr Niki: Well, folks, you can join the conversation and discuss these changes and all the other stuff we've been talking about in our Discord. If you'd like to join our Discord, become a patron. You can just join by linking your Patreon account at patreon.com slash DTMS.

32:15Dr Niki: Over 90 of the top 100 U.S. accounting firms trust Bill to simplify and secure BillPay. That's proven financial infrastructure built on over a trillion dollars of secure payments. Visit Bill.com slash proven for a special offer.

32:33Tom Merritt: Youth mental health is a complex challenge that requires comprehensive solutions. We must strengthen after-school programs. We must make digital literacy tools available in our schools. We must work with mental health professionals to support children. And we must empower mentors, educators, and parents to keep kids happy. Learn more about our commitment to finding lasting solutions at empowerourfuturecoalition.com slash solutions. Paid for by the Coalition to Empower Our Future.

33:05Jason Howell: And now some quick headlines that are just good to know might make you look smarter in

33:08Dr Niki: the future. Bloomberg has a good read on how people in Iran are using Starlink, VPNs and decentralized messaging to evade state imposed blackouts on the Internet that have reduced capacity to about one percent of normal there.

33:24Jason Howell: And then Amazon Web Services said drone strikes hit two UAE data centers in a nearby Bahrain facility, knocking them offline, disrupting key Middle East cloud services.

33:36Dr Niki: Audible is offering a new entry price of$8.99 a month for a plan called the Standard Plan, which gives you just streaming access to not only ad-free podcasts, but also all the audiobooks it has in its service without ownership. So these are not ones you could buy. So you may not get all the books that are available on Audible under the regular plan, but it's cheaper and it's available in Australia, Canada, France, Germany, the UK and the US.

34:06Jason Howell: Google's March Android update addresses 129 bugs, and that includes a Qualcomm display zero-day bug that's already under limited, apparently, limited targeted exploitation on 235 Qualcomm chipsets.

34:20Dr Niki: The ROG Xbox Ally X newest feature uses its neural processing unit to auto-tag your best gameplay moments so you can easily clip them out and share them later. You can't get this on the Xbox Ally because it doesn't have the NPU.

34:35Jason Howell: Yep.

34:35Dr Niki: Just the elements.

34:37Jason Howell: Right, exactly. Meta is testing a desktop-only shopping research assistant. It recommends products in a carousel with prices and links while using profile data for the customization of it.

34:49Dr Niki: Bloomberg reports that Morgan Stanley told creditors SpaceX is about to pay back all$17.5 billion in debt. That's$12.5 billion from X and the other part from XAI. They're going to pay it back in full.

35:02Jason Howell: Wow, okay. And finally, my pick for Mobile World Congress's strangest new smartphone idea is Ocatel's WP63. It's a$500 rugged smartphone with a 20 ,000 milliamp hour battery. That's not what's weird. It has a built-in electric igniter for starting campfires, lighting cigarettes, whatever you might want to set on fire.

35:22Dr Niki: It's like a car cigarette lighter from back in the 80s, but in your phone. O-C-Tel?

35:29Jason Howell: O-C-Tel? O-C-Tel. Those are the essentials for today. Let's dive deeper. Chatbots have been blamed for causing mental health harms, including in a trial in Los Angeles. And Dr. Nikki talked with us about what psychologists know about the potential harms and what we can do.

35:49Dr Niki: Back to the show.

35:51Tom Merritt: Beep boop. Hello.

35:54Dr Niki: If you are Dr. Nikki and not a chatbot, should I be worried if you're a chatbot? Like, I don't think.

36:01Tom Merritt: That's perfectly understandable, Tom. Let's get into that.

36:04Dr Niki: I think a lot of our audience assumes that chatbots are bad for your mental health. And so they're going to be like, OK, tell me something I didn't know. But what's going on here? What is the topic we're talking about?

36:20Tom Merritt: Yeah, I started out thinking about, you know, it's been in the news recently that there's been like AI psychosis. And so because of that, some psychologists are calling for changing the way that we interact with LLM chatbots. And I'll just keep saying chatbots, but we're talking specifically about AI LLM chatbots. And each developer has their own version. I think there's what, ChatGPT, Claude, Llama. I don't know what the bigger ones are.

36:44Dr Niki: Mistral has one as well. There's also SeedDance and DeepSeek and so many others. Yeah.

36:51Tom Merritt: There you go. And people say, you know, it's a bit easier to fall for these than you think. The tech, it's very companionable. But psychologists have called for potentially reducing the risk of chatbot dependency by potentially requiring the bots to deliver regular reminders that they are not human. However, a new publication in the journal Trends in Cognitive Sciences suggested that this might actually make the user feel even more alone and was maybe not advisable.

37:17Dr Niki: Oh, so how does that happen? Because, yeah, it makes sense that if it's reminding me it's not a human, then I'm less likely to treat it as a human. But that sounds like it makes the problem worse.

37:28Tom Merritt: Yeah, and it goes kind of both ways. It's really interesting. I think chatbots are a bit of a conundrum, and that's why we're having difficulty dealing with them. But what this research found is that the reminders that the chatbot isn't human actually encourages users to become even more dependent on it. And in doing surveys, they found that users who are aware of the non-human status would actually make more intimate disclosures, which then would lead them to intensify their attachment. So they've told them a bunch of secrets and now they're even more attached to it. And they're comfortable telling these secrets because they know it's not human.

38:00Yeah.

38:01Tom Merritt: And even some people, you know, when they were interviewed for this study, they stated that they find chatbots superior to humans for listening nonjudgmentally, which makes sense if you think about it.

38:12Dr Niki: Right. Because they're like, oh, it's just a computer, right? It sounds hypocritical, but hey, that's what we do. We're humans. And we're like, oh, I will acknowledge that it's not a human when that's convenient, but I'll get attached to it the same way I'd get attached to a stuffed animal or a car or something else.

38:30Tom Merritt: Yeah, the logical part of our brain maybe knows that it's a chatbot, but the herd mentality part of our brain that likes to personify everything, like, I mean, humans have pet rocks, okay, so we can have a pet chatbot, does fall for this. And another thing that they investigated was that potentially constant reminders may sadden the user by them acknowledging that what they've come to see as a companion is actually not human. And then this would increase their social isolation that already led them to seek out a chatbot in the first place. And so this is kind of a paradox that you see consistently across the literature on all different kinds of chatbots.

39:09Tom Merritt: So, and you could argue this, but the researchers state that because these people have increased sadness, it diminishes the effectiveness of the emotional support, however you want to characterize that, that's defined or that's provided by these chatbots. And then if someone is already in mental health distress, it could make it worse.

39:28Dr Niki: And these chatbots that we're talking about in most of these cases are just your general chatbot. And while they do have some safety protocols, has there been a lot of work done to look at the safety mechanisms and how they relate to mental health specifically? Yeah.

39:46Tom Merritt: So a study from Brown University last year investigated chatbots like GPT, Claude and LAMA. And it found out that they violate 15 different principles of mental health ethics standards. Um, and some of, some examples of that are some examples of that are that they, um, I'm not a chatbot. That's how you know. Um, they provide misleading information that will reinforce the user's negative beliefs. They are inappropriately navigating crisis situations. Uh, they, uh, create a false sense of empathy. This is the one that everyone's kind of aware of, like, you're right. You, you should eat that cake.

40:30Tom Merritt: And this is something we've covered on DTNS Forever, but there's huge unfair discrimination because of the way that models are trained. It increases human discrimination and also a clear lack of safety. If there is a crisis situation, a bot is not great at managing it, unsurprisingly. So this study took a year. And I think in the year that they did this study, things probably changed even more because the situation is moving so fast. But, yeah.

40:56Dr Niki: And I know there were probably people say, well, hold on, I've used a chat bot and I've not run into all of these things. And when we say this, sometimes it comes across as they always do that. And I think some people try to make it sound like that because they want it to help their argument. But I but I think it's important to point out, like these there are instances where it does that. And it's not a like rare instance. It could even be only 20%, but it's common enough that it will be inappropriately giving you advice. It could also give you perfectly appropriate and ethical advice sometimes.

41:30Dr Niki: It's just not consistent, right? And that's one of the safety problems. There are some positive effects that they have found from these too, right?

41:40Tom Merritt: Yeah, and I did want to highlight that it's not all bad, and this comes from this research. And a few studies have suggested that people who use chatbots as a companion self-report and underline self-report because that holds its own bias. But they self-report that they have seen higher benefits rather than harms. And the team that did this work underlines the importance of contextualizing chatbot use and especially media coverage of it within the broader environmental factors that will trigger distress in the user. So if someone's already mentally unstable or experiencing problems, and then they turn to chatbot use, there's a specific context that led them to being having, you know, these difficulties.

42:21Tom Merritt: And that context wasn't initially caused by the bot. So now it can compound it, but there's different factors that are working together here.

42:31Dr Niki: Yeah. When I think about this particular topic, I often compare it to various things. On the one hand, this is like using your veterinarian as a psychologist, right? Yeah. They probably have some ideas on psychology that could be helpful, but really not the best. Maybe more like your gas station clerk rather than a veterinarian.

42:55Tom Merritt: Yeah.

42:56Dr Niki: And maybe like, but in other words, like they're not bad, but they're not meant for this. And maybe we'll come up with tools that are fine tuned and very good for this. And it's also a little like alcohol. There are some benefits to alcohol. It can calm you down. It can lower your heart rate.

43:14Tom Merritt: Disinfect your wounds on all the TV shows. Exactly.

43:17Dr Niki: But there are also negatives. And I don't think alcohol, I personally don't think alcohol should be entirely banned, but there should be guidelines about how to use it and warnings about when it should and shouldn't be used and in how much it should be used. And I think that's one of the problems here is we don't have those kinds of guidelines regarding chatbots and mental health, right?

43:39Tom Merritt: Yeah. And, you know, it's the classic thing that we deal with with tech all the time is that the tech moves way faster than the regulations or even the people who make the regulations can even understand the tech. And a thing that we run into with LLMs all the time, and I think you mentioned quite often, is that they can be super broad. And so potentially you might benefit more from an LLM that's trained exclusively on therapeutic things for better therapy, at least a step ahead of it not being trained that way. I don't know what the outcome would be. The main problem with this is that the chatbots don't have accountability if they make a mistake.

44:17Tom Merritt: If your therapist makes a mistake, an ethics, you know, breach, they have things that happen to them legally, like their records get pulled, etc. Whereas chatbots don't have the legal confidentiality like your medical provider would. So you're actually putting yourself a little bit at risk as well.

44:37Dr Niki: So yeah, that's definitely something to keep in mind.

44:40Tom Merritt: But yeah, regulations for things like that would probably be part of the solution. And even the researchers who did this work mentioned that a lot of the AI work today is evaluated through automatic metrics and humans don't even kind of get into the process. And their whole study, like I said, it took them a whole year with an entire team of clinically trained people. And yeah, probably within that year, the chatbots already changed their rules. So they suggest a little bit more human oversight until we have a better wrangle on what's going on and regulation needs to catch up.

45:15Dr Niki: Yeah, I think I think some sensible, you know, guidelines and warnings and public information campaigns to let people know, you know, what which ones of these are good for this and which aren't. And what they're getting into, you know, and what they're getting into and how they should be used. And then I think the companies are starting to add more and more instances of if you try to ask it something, it'll say, hey, I'm not meant for that. It's not 100 % perfect yet. But encouraging that to continue, I think, is good. And also just, frankly, like setting standards of like, oh, if like any other medical equipment or medicine, if it's going to be used for a therapeutic purpose, it needs to meet these kinds of standards.

46:00Dr Niki: We don't have that. We need to develop that. Yeah.

46:03Tom Merritt: And I think more studies will also be helpful. And this is a research field that I would not like to be in because I'm already struggling with, well, it takes me maybe three years to publish the really big project that I'm working on. And I can't imagine if in that time goats grew like six more heads and I couldn't keep up, you know, like that sounds super stressful. So being able to publish quickly is something that's almost impossible. And in that time, if the tech you're studying already changed, that's that's grave. So but they should do it more.

46:32Dr Niki: Think about three years ago where we were with LLMs. It's an entirely different landscape. Yeah. Well, Dr. Nikki, thanks, as always, for bringing this to us. If folks want to find more of what you do, where should they go?

46:42Tom Merritt: They can find purely human grown work over at NicoleAckermans.com. And I have the same handle on Blue Sky.

46:49Jason Howell: Fantastic. Thanks, Dr. Nikki. Yeah. And can I just say, now that we have the video format, I get to see Dr. Nikki. It's great.

46:56Dr Niki: Not just hear her voice. Yeah. I guess we should let people know. If you want to see video of this, you can see it at youtube.com slash daily deck new show.

47:04Jason Howell: There you go. All right. We end every episode of DTNS with some shared perspectives today. Wisdom helps us understand what companies mean when they say nits.

47:13Dr Niki: Yeah. Thank you, Wisdom, for writing this. He said, I just wanted to add context to the screen brightness claims made by manufacturers. When manufacturers quote very high brightness, like 6 ,000 nits in the case of Motorola's foldable, we talked about that on Monday, it's usually peak brightness. Peak brightness indicates how bright a section of the screen, often a very small one, can get. You won't get that level of luminance across the full screen. Even with this context, some manufacturers' claims don't hold up when tested in real-life conditions. While the peak brightness of a section of the screen is high under bright ambient lighting, the full-screen brightness usually turns out very low, even compared with other devices that don't advertise such high peak brightness.

47:54Dr Niki: A lot of people hear such high-knit numbers and are misled into believing those displays can get that bright across the full screen, so I thought I'd add a little context to those numbers to keep up the good work. This is great. Love it. I don't say peak brightness because it's understood usually that it's peak, but a lot of people don't know that. And even if I said it, a lot of people wouldn't understand what it means. And I thought Wisdom did a great job explaining that. I appreciate that.

48:17Jason Howell: Wonderful. It's really easy to misunderstand when you're talking about nits and brightness, what that actually means, especially with the peak kind of getting in there too. So love it. Great explanation.

48:27Dr Niki: If you'd like to add in a little helpful information like Wisdom did, send us an email, feedback at dailytechnewsshow.com. Yeah.

48:35Jason Howell: Thanks to Dr. Nicky. Thanks to Wisdom for contributing to today's show. And thank you for being along for Daily Tech News Show. You can keep us in business by becoming a patron at patreon.com slash DTNS. We'll see you tomorrow. Bye, everybody. Bye.

48:54Jason Howell: The DTNS family of podcasts. Helping each other understand.

49:00Dr Niki: Diamond Club hopes you have enjoyed this program.

49:06Dr Niki: If you work in university maintenance, Grainger considers you an MVP because your playbook ensures your arena is always ready for tip-off.

49:15Tom Merritt: And Grainger is your trusted partner.

49:17Dr Niki: Offering the products you need all in one place, from HVAC and plumbing supplies to lighting and more. And all delivered with plenty of time left on the clock, so your team always gets the win. Call 1-800-GRAINGER, visit grainger.com, or just stop by Grainger for the ones who get it done. Over 90 of the top 100 U.S. accounting firms trust Bill to simplify and secure BillPay. That's proven financial infrastructure built on over a trillion dollars of secure payments. Visit Bill.com slash proven for a special offer.