This Week in Tech 1036: Apple Reflux

Summary

Podcast Summary: This Week in Tech 1036: Apple Reflux

Released June 16, 2025

Hosts:

• Leo Laporte – Host of TWiT
• Harry McCracken – The Technologist at Fast Company
• Jason Heiner – Editor-in-Chief at ZDNet
• Jason Snell – Contributor at SixColors.com

---

Introduction

In Episode 1036 of This Week in Tech, titled "Apple Reflux," Leo Laporte and his panel of seasoned tech journalists dive deep into the latest developments in the technology landscape. The discussion covers significant announcements from Apple’s recent Worldwide Developers Conference (WWDC), the release of Android 16 by Google, advancements in quantum computing, and the tumultuous journey of genetic testing company 23andMe.

---

Apple’s WWDC Announcements

New Design Language

Apple unveiled a "liquid glass" design language at WWDC, sparking mixed reactions among tech enthusiasts and critics alike.

• Harry McCracken highlighted the opinionated design, noting, “What if ChatGPT can write code in Basic?” [04:03]
• Jason Snell expressed his initial skepticism but found the new design "ambitious and interesting," especially praising its performance on iPhones versus Macs. [08:58]
• Leo Laporte remarked, “It really is like they needed a reset,” observing that the new design feels more integrated across all Apple devices. [10:41]

Windowing Model Improvements

Apple introduced an entirely new windowing model for iPadOS and macOS, moving away from previous iterations like Split View and Stage Manager.

• Jason Snell appreciated the shift, stating, “They built an entirely new windowing model that just works basically like Mac Windows do.” [19:09]
• Harry McCracken voiced concerns about the usability for those not fond of managing multiple windows, saying, “I feel like windows that float and overlap... computing's original sin.” [20:54]

AI Integration and Apple’s Approach

Apple’s strategy towards artificial intelligence (AI) was a focal point, emphasizing measured integration rather than over-promising capabilities.

• Jason Heiner discussed Apple’s approach: “Apple’s strategy is not to have chatbots dominate but to integrate AI features where they make the most sense.” [28:17]
• Jason Snell added, “Apple is focusing on features like better translation and smarter tools integrated into the OS,” highlighting Apple’s commitment to privacy and on-device processing. [28:17]

Vision Pro and Future Technologies

The panel touched upon Apple’s Vision Pro, an augmented reality (AR) headset, and its implications for the future.

• Harry McCracken affirmed, “They are fully engaged by Vision OS,” indicating ongoing investment despite its current market position. [103:25]
• Jason Heiner expressed optimism, “The software is not the problem with that product line. It’s the heavy, expensive hardware that’s the issue.” [105:32]

Spotlight Updates on macOS

Apple revamped Spotlight on macOS, introducing new features to enhance productivity.

• Jason Snell was excited about the new Spotlight capabilities, including keyboard and clipboard history: “It’s like the big missing piece of productivity built into the system.” [105:11]
• Leo Laporte anticipated further integrations, stating, “It’s only a matter of time before they bring it to the iPad too.” [105:03]

---

Google’s Android 16 Release

Google launched Android 16, focusing on incremental updates rather than a major overhaul.

• Leo Laporte mentioned, “You’d be hard pressed to see a whole lot of differences,” noting Google’s strategy of releasing features through monthly Pixel updates. [38:28]
• Enhancements include live update notifications mimicking iPhone’s live activities and advanced protection features for security. [38:28]

Harry McCracken commented on Google's lack of prominent AI integration compared to Apple, expressing, “It seems like Android is not one of their top priorities these days.” [38:57]

---

Quantum Computing Developments

The discussion shifted to quantum computing, with a spotlight on IBM’s ambitious plans.

IBM’s Starlink Quantum Computer

IBM announced plans to build Starlink, the world’s first large-scale error-corrected quantum computer, by 2028.

• Harry McCracken explained the technical challenges, emphasizing the importance of logical qubits over physical ones: “If you want to look at numbers that relates to these computers, look at how many logical qubits they have rather than physical qubits.” [49:09]
• Jason Heiner highlighted the unpredictability of quantum computing's progress, stating, “It's extremely hard to get a qubit to respond exactly the same way every time.” [48:29]
• Leo Laporte noted, “Starling will have 200 logical qubits,” and discussed the skepticism around its economic viability: “Some experts say 100 million is a tenth of what you need.” [49:09]

Harry McCracken offered a balanced view, acknowledging progress while noting, “They are getting there. We are not yet at the point where quantum computers are doing the kind of stuff that people envision.” [46:23]

---

23andMe's Bankruptcy and Reacquisition

23andMe, a genetic testing company, faced financial struggles leading to bankruptcy and an attempted acquisition by Regeneron, which was ultimately outbid by founder Ann Wojcicki.

• Jason Snell provided background, noting, “The company, according to the Wall Street Journal, never made a profit. They burned cash across DNA testing and drug development.” [64:34]
• Leo Laporte raised concerns about data privacy, stating, “Some customers have already deleted their data from 23andMe.” [61:21]
• Harry McCracken questioned, “Why did they fail in the first place?” suggesting that companies haven’t figured out post-test services for customers. [61:34]
• Jason Heiner discussed legal implications, highlighting a lawsuit by 27 states and D.C. over unauthorized data sales: “The lawsuit is expected to help settle who owns the data.” [71:25]

The panel emphasized the critical issue of data ownership and privacy, with Leo Laporte summarizing, “This lawsuit could be significant in helping to establish who actually owns that data.” [74:20]

---

Market Trends and Industry Implications

Layoffs and Hiring Challenges

The tech industry continues to navigate through massive layoffs, partially influenced by changes in the tax code affecting Research and Development (R&D) expenses.

• Leo Laporte introduced the topic: “The tax code allowed companies to write off R&D expenses in the same year, but now it changes to amortizing over five years.” [44:38]
• Jason Heiner connected this to recent layoffs, stating, “More than half a million tech workers have been laid off in the last two years because of Section 174.” [43:17]

AI Bubble and Public Perception

The panel discussed the burgeoning AI industry bubble, noting a discrepancy between public enthusiasm and technological readiness.

• Jason Snell remarked, “There is a lot of hype, but also a lot of limitations,” cautioning against overestimating AI’s capabilities. [112:51]
• Harry McCracken shared personal experiences with AI tools, highlighting their limitations in practical applications: “They are not good at pinpointing one thing and can create more problems than they solve.” [116:34]

Leo Laporte expressed concern over governmental AI adoption without full understanding, comparing it to historical tech overextensions: “They are using AI to find fraud in Social Security applications. I hope they understand its limitations.” [120:54]

---

Quantum Computing and AI Intersection

The panel delved into the potential synergy between quantum computing and AI, debating whether quantum advancements could revolutionize AI training and applications.

• Harry McCracken suggested, “Quantum and AI will work together,” with quantum aiding in specialized tasks like drug discovery rather than supplanting AI. [52:51]
• Jason Heiner remained skeptical but acknowledged the potential: “There is tremendous hype, but reality ensures limitations,” emphasizing the need for realistic expectations. [73:30]

---

Light-Hearted Segments and Personal Notes

Throughout the episode, the hosts shared personal anecdotes and lighter topics, including:

• Cat Discussions: Tales of cats following visitors and the joys of being a cat owner.
• Curling Conversations: Insights into the sport and personal experiences with curling competitions.
• Nintendo Switch Release: Excitement and frustrations surrounding the new Nintendo Switch 2 availability, highlighting supply chain issues. [132:30]

---

Conclusion

Episode 1036 of This Week in Tech provided a comprehensive look into the current state of technology, with a particular focus on Apple's strategic decisions amidst evolving trends in AI and quantum computing. The panelists offered balanced perspectives, acknowledging both the advancements and the challenges facing the tech industry today. From the intricate updates at WWDC to the legal battles of 23andMe, this episode encapsulated the dynamic and ever-changing landscape of technology.

Notable Quotes:

• Leo Laporte: “This is MAYBE what Vision Pro's full-fledged ecosystem is going to look like.” [07:04]
• Jason Snell: “Apple is remembering who they are and focusing on what makes their devices great.” [31:02]
• Harry McCracken: “Quantum computing is extremely hard to get a qubit to respond exactly the same way every time.” [48:29]

---

For More Information:

• Harry McCracken: fastcompany.com
• Jason Heiner: ZDNet.com
• Jason Snell: SixColors.com

Transcript

[Leo Laporte]

It's time for Twit this Week in tech. Harry McCracken is here. The technologizer, Jason Heiner. Jason Snell. It's old home week. Got some great people on to talk about, of course, everything. Apple announced on Monday at the WWDC. The new Android 16. And it's finally here, the arrival of quantum computing. That and a whole lot more coming up next on Twitter. Podcasts you love from people you trust. TWIT. This is TWIT this Week at Tech. Episode 1036, recorded Sunday, June 15th, 2025. Apple Reflux. It's time for Twitt this Week at Tech, the show where we talk about the week's tech news. I like to bring in veteran technology reporters on this show. I think if you include me, we've got about 400 years worth of experience. No, it can't be that much. 50 times 4. I don't know what it is. Jason Heiner, how long have you been covering tech?

[Jason Heiner]

25 years.

[Leo Laporte]

Yeah, there you go. He's editor in chief of ZDNet now. It's good to see you, my dear old friend here.

[Jason Heiner]

A pleasure, a pleasure. And with a distinguished panel, old friends of mine as well. So this would be fun.

[Leo Laporte]

You all know each other, which is cool. Harry McCracken, the technologizer, now at Fast Company. How long have you been covering tech?

[Harry McCracken]

I started doing it professionally in 1991, so 34 years.

[Leo Laporte]

That's nice. I think I'm going to beat you all, but that's good. So we got 25, 34, that's 59. Jason Snell from sixcolors. Com. How long you've been in this business?

[Jason Snell]

Depending on how you count, about 32 years.

[Leo Laporte]

Oh, my God.

[Jason Snell]

Well, 31, 32, something like that.

[Leo Laporte]

We're going past 100 because I've been doing this since the late 80s.

[Harry McCracken]

I do have something I can pull out for cases like this, which is right after I got out of high school, which was 1982, I wrote a couple of pieces for Creative Computing that got me beat in very instances like this.

[Leo Laporte]

Yes, we're close because I wrote my first piece for Byte. I think was 85. I wrote for a couple of Atari magazines before that. Probably 83. So we're close. I wasn't in high school. I was gonna say grown man Compute.

[Jason Snell]

Magazine paid me for a basic program I wrote when I was 14. That's right.

[Leo Laporte]

That counts.

[Jason Snell]

That's about it.

[Leo Laporte]

That counts. Didn't you just recently resurrect that program?

[Jason Snell]

No, no, no, no. I don't think so.

[Leo Laporte]

Somebody I was talking to had done the same thing and he got it working again.

[Jason Snell]

No, that one's lost to history. Although trust me, it's not that hard to make a blackjack program. It's very easy.

[Leo Laporte]

And which BASIC was it?

[Jason Snell]

It was Commodore BASIC because it was on a Commodore pet, so there was no graphics. It was literally just numbers. Before the VIC 20 and the C64, the Commodore PET had no graphics mode at all. So that was interesting. But yeah, that's. I realized I type really fast in part because in the computer magazines of the early days, they would give you programs to type in. And so I got really good at typing without looking because I was looking at the magazine and then typing code in. And that's why I can type fast to this day.

[Harry McCracken]

I actually did an article on the history of BASIC where I interviewed the editor of compute, who decided to get rid of the listings, which for them was a big decision because they were the most list centric magazine at the time.

[Leo Laporte]

Well, it was. The bulk of the magazine was programs.

[Jason Snell]

You could type in.

[Leo Laporte]

Yeah, Code, BASIC code. There were no pictures, it was just code. I guess they could put some ads around it though.

[Jason Snell]

I mean, basic, so it was readable, but you know, a lot of them still at the end there were like 80 lines of data statements that were completely inscrutable.

[Leo Laporte]

Right, right, yeah.

[Jason Heiner]

What if ChatGPT can write code in Basic?

[Harry McCracken]

I've tried to get it to write TRS 80 Basic and the last time I tried I was terrible at it because I don't think I can tell one old microcomputer BASIC from another. So that's probably what it is if you're doing things like graph commands. It was awful. Although I keep meaning to try to like feed a lot of TRS 80 listings into some LLM to see if that helps.

[Jason Snell]

That's a good idea.

[Leo Laporte]

I will say it does Common Lisp quite well because Common Lisp was standardized in the 80s and so all the stuff it's reading is the same and it codes Common Lisp like a champ. Does Apple Script pretty well too, right?

[Jason Snell]

It does AppleScript pretty well. And I can't verify how good it is, but I got it to generate some hypertalk for HyperCard. There's some of that in there too, but it's, you know, basically the question is, is there enough on the Internet as an example? And if there is, it can do it.

[Leo Laporte]

But I like what you said, Harry, because it's not standardized Basic, There were so many different basics so many.

[Harry McCracken]

And the guys, the guys who invented BASIC were really unhappy about that. They really looked down on Tiris ADBL Soft and they came up with something called True basic in the 80s.

[Leo Laporte]

I remember that. Yeah.

[Harry McCracken]

Which was a reaction to the fact that BASIC had become so unstandardized.

[Leo Laporte]

Let's talk about the week's tech news. Let's talk about last century's tech news.

[Harry McCracken]

This is all in an article I wrote for Time. When basic turned 50, I think so.

[Jason Snell]

Last century in tech.

[Harry McCracken]

10 years ago.

[Jason Heiner]

Last century tech.

[Leo Laporte]

That's not a bad show. People have often asked me if I would do a tech nostalgia program because I think people, like some people, people of a certain age anyway, like that. And maybe the youngs like to hear how hard it was becomes. Yeah, it's history compared to what they're doing. Well, everything old is new again. Apple has brought back Windows xp Arrow to be the new Apple ui. Apple, of course, had its WWDC on Monday. Jason, you were there. So was our good friend Harry. You were one row apart at the Apple Park.

[Jason Snell]

That's true. In the shade.

[Leo Laporte]

One thing that didn't get a lot of coverage. Verge mentioned it. We didn't see it because we were watching a canned video, but you guys were there. There was a protester, was there not?

[Jason Snell]

Yep, there was one.

[Jason Heiner]

I was also there.

[Leo Laporte]

Oh, you were there too, Jason. Now I'm feeling left out.

[Harry McCracken]

I saw Jason from across the room, but unfortunately we didn't get to say hi in person.

[Leo Laporte]

I will never get an invite to Apple park, but that's okay. Yeah. A pro Palestinian Apple employee stood up and apparently harangued for 30 seconds before being escorted out.

[Jason Snell]

Yeah, I was surprised that it took them that long. I thought they would be a little more on guard, but it sort of took a little while where everybody was kind of looking at each other, all the security people, like, what are we going to do with this guy? And then they took him away.

[Leo Laporte]

Craig Federighi was speaking at the time. He didn't pause.

[Harry McCracken]

After those Microsoft events, you would think that every big tech company would be on guard.

[Leo Laporte]

The difference was this was inside the Apple campus. I mean, pretty hard.

[Jason Snell]

Well, I heard it was an employee.

[Harry McCracken]

Who was about to leave Microsoft for their 50th anniversary was on campus and that they had protesters.

[Leo Laporte]

A lot of Microsoft employees are unhappy about Azure's involvement in the Israeli Gaza Palestinian conflict.

[Jason Heiner]

Sabrina Ortiz, who was there, also from ZDNet, you know, who's our AI reporter, she said she's been to, you know, she's been to almost all the developer conferences. Microsoft Build, Google IO, wwdc, a couple other ones. She said there's been. This has happened at almost everyone that she's been to. So.

[Leo Laporte]

Yeah, yeah. Less obvious at Google's Shoreline event and less even obvious at Apple's event, which everybody outside of the campus was watching on tape.

[Jason Snell]

Well, it was, it was an intro. So Tim Cook and Craig Federighi come out before the show and they say, hey, everybody, you know, because there's a big group of developers in press there, and then they play the video. So the, the protesters started shouting in the last couple of sentences of Federigh before they rolled the video. And so, you know, we couldn't hear him. It was kind of ineffective other than to acknowledge that there was a protester. And then they were gone. It was kind of. I couldn't hear a word he was saying. And he was probably, you know, he was probably 30ft from me.

[Leo Laporte]

Well, as with this show, life intrudes into tech sometimes from time to time. So this new liquid glass, it's funny how the reactions have been all over the place. Mac Stories had an article calling it, what opinionated design.

[Jason Snell]

That was one of the lines. I mean, there, there, that was like a summation of, of a bunch of the quick takes.

[Leo Laporte]

Yeah.

[Jason Snell]

Afterward.

[Leo Laporte]

Well, now, after the fact, you guys have had some time to think about it. Most of you have probably installed the developer preview. I haven't. But you guys live on the edge. What do you think?

[Jason Snell]

I, you know, first off, it's a, it's a new design. It's always hard to judge. I, I'm trying to not rush in because it's human nature to reject anything that's new. And that doesn't mean that every new design is good. It really doesn't. But it does mean that I am less inclined to listen to people who see something and immediately say, oh, that's terrible, or who have been complaining about Apple's design trends for a while and say, see, this is right. It's like, I don't think that tells us anything. I think you need to kind of consider, and consider that it's new and that you, all of us have to kind of like, drink it in and sleep on it a little bit and spend some time with it. And so, you know, I, I think it is ambitious. I think it's interesting. I think they're kind of showing off their, their GPU power that they can do some, like, some of this stuff. I think some of it is, you know, I Think it's pretty sometimes and ugly sometimes. And I have a real question about, like, what they're going to ship. But I'll tell you, if you've only judged it based on the macOS beta, it looks really bad on macOS right now. On an iPhone, it looks really good. So, I mean, it's a work in progress in a way, and I wonder how they'll react or how they'll react to the reaction to it. But I do think they needed a new design that was considered across all their devices. They really have been patching designs and importing one design language into another device that was based on a completely different design for quite a while now. And they really kind of needed a reset. I will explain. Expect that it'll be painful for a little while, but in the end I, I'm. I'm optimistic about it. But like, I mean, it is, it is a fresh coat of paint. That's what it is.

[Leo Laporte]

To me, it seems more than a fresh coat of paint. Like it's a. It really. Everything's going to look a lot different.

[Jason Heiner]

Yes, it feels very Apple like that was the. My reaction to it, especially on. So I. I did it on iOS first, then iPadOS, then Mac OS and on I iPhone and iPadOS, it. It just again, it felt very Apple like and. And the consist to Jason Snell's point, you know, and then when I started on Mac os, then I started getting a little uncomfortable. Ooh, this is a lot different. I was like, this, this is. This is. I'm feeling a little. I felt the reaction that Jason mentioned like, okay, I'm. Now I'm getting a little uncomfortable. I'm trying to, I'm trying to give this the benefit of the doubt, but I'm still getting used to it there. It feels the most different on a Mac, I think, than it does on the others.

[Leo Laporte]

Bloomberg managed to get a think piece out of two data points. That and the fact that the have skeuomorphic icons in the Airbnb app to write an entire piece called Big Tech is dealing flat design a death blow. Skeuomorphism is back.

[Jason Snell]

I mean, to a point. I mean, it is. I mean, because it's trying to ape glass, it is a little skeuomorphic, although not like super skeuomorphic where everything.

[Leo Laporte]

You don't have green velvet or anything?

[Jason Snell]

No, no, nothing like that. But I would say, yeah, this is the. To your point about it being not a fresh coat of paint, what I would say is I feel like something like iOS7 when they really Redesign was functionally a very different way of interacting with the iPhone in a lot of areas. And this doesn't feel like that to me. It feels using an iPhone or an iPad, like, I've been using my iPad. I installed this on my iPad that I use all day, and I had another iPad waiting just in case. It was a disaster. It wasn't. It works really well, but it also. It looks different, it doesn't feel that different. And so it's not as disruptive, I think, to a user as some previous redesigns have been. It's more like a visual refresh than anything. There's more animations, there's a different look, but it doesn't like remap how you view notifications or something. Like, they all still pretty much are a bubble that is in the lock screen. And like, it's not that different.

[Leo Laporte]

Look at this, though. This is an Airbnb icon from Austin Car's article in Bloomberg. That is vivid. That is not skeuomorphism. That makes me hungry.

[Jason Snell]

And Apple's not doing that, which is. So we've got two different kind of views of what the future of UI design might be.

[Harry McCracken]

I mean, I think it's definitely going to be a work in progress until it ships, and in some ways probably a work in progress for the next few years, because there's just a lot they can do with this. And the devil is in the details. And anytime you introduce transparency, you're opening up the possibility of stuff being extremely hard to read. And you have to kind of balance the cool transparency with just being able to read the labels. And I expect it to get better before the fall. But iOS7 also kind of led to additional iterations and the updates that followed. Using that as the foundation, I would expect something similar to happen here.

[Jason Heiner]

Accessibility was the first thing that got called out because of the clearness of the. The notifications. Then, depending on what your background is, they can be difficult to read. You can turn them off. You can go into accessibility and make them so that they are easier to read.

[Leo Laporte]

That'll be the test. How many people who don't have accessibility issues decide to go back to the style?

[Jason Snell]

That's exactly it. Because accessibility, I think it's great and Apple has a commitment to it. And in some of their sessions this week, the designers said, you know, we've designed this with some different looks based on your accessibility settings, and there's different ways you can reduce animation, you can reduce transparency. There's lots of things you can do and that they've designed this with the accessibility settings in mind, so it's not an afterthought. It really is sort of like they wanted to be able to scale. And my friend Shelley Brisbane, who writes about a lot of accessibility stuff, said back in iOS 7, there really weren't as robust a set of accessibility settings as there are now. And so it is something that, like, if you're really worried about it, there are probably going to be features under accessibility to change what the design looks like. There's a good argument to be made about, like, why is that under accessibility instead of just saying these are design settings. But it's fine wherever it is. All of us can probably take advantage of one or two accessibility settings. So I think it's fine. I think the question is exactly what you said, Leo, which is, does everybody have to turn on those settings or does almost nobody have to turn on those settings? And in between those two, there's a spectrum. And I think that the more people have to resort to the accessibility settings, the, the more of a failure the design is. Right. They want it to be something that most people will love as is. And if you don't love it, there's a setting which is great. But if everybody is rushing to turn on, you know, opacity in glass, then they failed.

[Leo Laporte]

How much of this is driven by Vision Pro and Apple's ambitions towards augmented reality?

[Jason Snell]

I think not. I think they designed Vision Pro and it was almost like a spearhead for this design where they knew it was going to go everywhere, but they needed to start somewhere. I think that they were inspired by some of the choices they made. Right, because the design team had to design that from scratch. That was a new UI for them and they, I think, knew that they were going to do a big redesign. And so sure, I think there's some inspiration from it, but I don't think it's not.

[Leo Laporte]

No, no, not backward. Inspiration forward, like, let's get people ready. I just to start seeing through things to having heads up displays, things like that.

[Jason Snell]

I'm skeptical of that. I think it's more about using their materials, of their touchscreen devices than anything else. I do think that they built Vision Pro and they learned a bunch of stuff that they thought, well, this is kind of cool, we should do that for everything. And so that they have. I think that's, that's so far out, Leo. Like, I mean, I just don't know if you're going to build an AR device in five or 10 years that you would make those changes today. I Do think it's possible. I know Mark Gurman of Bloomberg has been talking about this a lot that they want to do what a kind of like quote unquote, all glass iPhone next year for the 20th anniversary of the iPhone and that this is a good fit for that. I mean, I don't know about that story, but I will say sure, I'm sure this design was created knowing what the next couple of years of products are that are in the pipeline. And if there's a folding iPhone, if there's an all glass iPhone, whatever that means, I'm sure they took those into account. But the AR story feels like way off to me.

[Jason Heiner]

So I have the same skepticism. But I was sitting with Scott Stein from cnet, you know, my colleague there. And so I was sort of asking him the same question, you know, like, how much do you think the design language is related to some of this, you know, mixed reality kind of future? And of course, you know, because he thinks about this stuff all the time, has been thinking about it for, for years, he said that, that in his mind that this a lot of the ways thinking about what do windows look like when you have it on any background. Right. What do they. Because you can't predict what does it happen when you have certain windows of different depths the way that you do in Vision Pro. So he was a little bit more on the scale of like he feels like this starts to lay the tracks for some of that future when there are glasses, when there's a cheaper Vision Pro that is accessible to more people. And some of those devices are, you know, more prevalent than they are today. I mean, we have to be honest, like there's only. There are hundreds of thousands of people who have Vision Pros. You know, there are 2.5 billion people using the other devices. So, so I tend to be a little more on the skeptical side as well. That, that, that's a long journey, you know, from here to there. And yet, and yet I think it is compelling to think like, okay, well if we start to think about some of the design language and think of depth in different ways, then some of these things do seem to tracks a little bit for that world.

[Leo Laporte]

Of course, that was perhaps the headline of the Apple event. They saved till the end what I consider one of the most interesting parts of the Apple event. Pretty much a redesign of how the iPad works.

[Harry McCracken]

My goodness. Yes.

[Jason Snell]

Oh yeah, yeah. It's a huge thing. Yeah. My piece on six colors is that Apple got over its hang ups a little bit, which is they basically went back and redesigned. It's not a. They did like split view and slide over and then they did stage manager and they iterated on that a little bit. This os, just to be clear, they threw that away. It's gone. They. They built an entirely new windowing model that you can turn on and off and so you can use it in single window mode, but if you're in the windowing mode, it just works basically like Mac Windows do. I think, I think when I say that Apple has kind of gotten over its hang ups, one of the things is that there was a period where they kept trying to do the iPad is like not quite the Mac, but let's reinvent it and put a spin on it. And I think that. And I felt a lot of confidence coming from people at Apple that like, you know what, the Mac did this, it's good. Let's just like, why are we not just using something that's familiar? And whatever that hang up was, they definitely got over it. And having played with it this week, like, it's good. It's good because it's familiar and it works.

[Leo Laporte]

It's a computer, not a consumption device.

[Jason Snell]

And honestly I felt like for five years they were fighting, it's good and it works. And they're like, no, we're going to revisit this and we're going to move your Windows around. And I mean, Harry uses iPad even more than I do.

[Leo Laporte]

So yeah, you were iPad complete, right?

[Harry McCracken]

I have been primarily iPad since 2011 and I have to say, I have to say I'm not over my hang ups. I feel like virtually 100% of the comments I've seen about this new approach have been extremely favorable, although some of them are from people who are not really iPad people. And I kind of feel like if you're not an iPad person, then who cares what you think?

[Leo Laporte]

I don't think this is a renaissance for.

[Harry McCracken]

Well, the thing I'm concerned about is that I basically feel like windows that float and overlap in a lot of ways are computing's original sin. The more time I spend having to fool around with windows and drag them and get to one that's below another one, the less I can do what I want to do, which is to create stuff or consume stuff. And I thought Split View was pretty good for letting me do a lot of stuff without wrestling with Windows. And I don't think it's entirely clear whether somebody who doesn't want to drag Windows around can get anything except for this full screen mode, which does not offer Split View. And if they offered Split View as an alternative or potentially offered this new interface, but offered some way where you could kind of ignore it if you wanted to and just work with like two large windows at a once at once, I'd be happy. But it's not entirely clear whether that'll be possible. And I feel like nobody else is rooting for Split View to survive except for me. So it might not survive.

[Jason Snell]

I'm surprised that they built Split View and then they didn't make it easy to get to. Like, Split View is essentially, you can tile two windows in multi View, hitting that button by accident.

[Harry McCracken]

Yeah. You have to know it's there to use it. Stop.

[Jason Snell]

Yeah.

[Harry McCracken]

And I don't understand why Stage Manager survives, because it's the worst thing ever.

[Leo Laporte]

They're planning for bigger screen iPads and multiple screens. I'm thinking about plugging in a second screen to my iPad.

[Harry McCracken]

Well, actually, on an iPad, I think Stage Manager is terrible, but on an external screen, it's not bad. So you do need. The bigger the screen, the more it makes sense to have kind of more elaborate windowing options.

[Jason Snell]

And it's optional now like it is on the Mac. It no longer is a fundamental part of iPad windowing. Stage Manager, it's just a window manager if you want to use it.

[Leo Laporte]

Another choice. It feels like we're relitigating. One of the earliest battles in GUIs, Xerox PARC, had a tiled window interface, not overlapping Windows. But Steve Jobs, when he visited Misunderstood. He thought he'd. Or maybe it was Andy Herzfeld. I think it was, saw it and thought he saw overlapping Windows. So he wrote overlapping Windows into the Lisa, into Quick Draw, even though Xerox parc. Mark didn't do it because it's computationally difficult and maybe aesthetically it's problematic. Harry sounds like he's a tiling Windows guy.

[Harry McCracken]

And Bill Atkinson, who, I mean, was a genius who just passed away, I think contributed a lot to the ability of GUIs to have sophisticated Windows.

[Leo Laporte]

Maybe I should say Bill Atkinson because he wrote quickdraw, not Andy.

[Jason Snell]

I should mention that part of this reorder. They also added all of those same window tiling features that they added to the Mac last year, where even with a keyboard shortcut, you can put, you know, side by side.

[Leo Laporte]

Microsoft.

[Jason Snell]

Exactly. So there's tiling in there, but it's not a tiling system. It's freeform Windows with tiling commands that you can assign to them. Yeah.

[Harry McCracken]

I mean, basically, if you want an iPad to work Like a Mac. This is fantastic. And there is a lot of other stuff they added that I am enthusiastic about. Like the Files app is much more sophisticated. Background tasks work better in a lot of ways. I think they did do a good job of finally making the iPad into a more professional tool.

[Leo Laporte]

And all of this was to kind of deflect from the fact that they had announced a year ago at WWDC Siri as a chatbot, which they were not able to do. And I thought though that Apple handled it much better than I expected. They basically introduced a bunch of AI features kind of gracefully integrated into the overall interface and then said developers, you can do this in your apps too. And said we'll get to the chatbot later. I think that was the right thing to do.

[Harry McCracken]

And they were also open minded about calling on ChatGPT when ChatGPT can do something well. So you can do image generation with ChatGPT now.

[Leo Laporte]

Gurman says it's spring 2026 for the Siri AI upgrade in his Bloomberg rumor piece. Of course that's not been announced.

[Harry McCracken]

That sounds about right given it's clear it's not going to be in the fall. It's not going to be in the fall. It'll be a little while after that.

[Leo Laporte]

In the interviews the executives did, they did a whole bunch of them with everybody but Jon Gruber after the event. They implied that we already have it working inside Apple. We just want to make sure that it's completely reliable before we release it.

[Jason Heiner]

Yeah, look, I think I know some people that have worked on and Siri and are familiar with some of the code base and they've said, even they've been saying for a couple years and they could be wrong. So I want to be, you know, honest about this. Like they could be wrong. I think there are different opinions. I've sort of asked people in my briefings with Apple. So I asked them as, as, you know, respectfully as I could, this question which is that the, the code base of Siri goes back, you know, all the way to like darpa.

[Leo Laporte]

You know, Apple acquired a company, right. It wasn't even done inside.

[Jason Snell]

That's all right.

[Jason Heiner]

Yeah, yeah, yeah, yeah. So, so that, you know, so adapting that code base, my, my question to them was like, how do you adapt that code base to the, you know, LLMs, right. Which is how do you put, put that on top of it. And so the people that I know that, that say, they say they're going to have to start over. The people that, you know, had some familiarity with the code base at Apple they say they're going to have to start over. It's just not, it's just build on. You've seen how it struggled the last few years. So that's one school of thought. There's another school of thought. So inside Apple they're like well macOS is based on Unix code base which like look at that over time like that's been hardened, it's been tuned, it's, you know, has shown its, its strength, you know by having a long code base. It's not necessarily a bad thing. So, so these like two, they're sort of of two minds on this but I think no one you know believes that it's, it's really where anywhere close to where it needs to be yet. But, but I think that still Apple's larger strategy and I wrote about this, this was what I wrote after wwdc, which is the headline was Apple's D Chat Modification of AI is nearly Complete. Which is like they're like we. Our thing is not that you're going to sit there chatting with chatbots all day, right? Like that's not that far away from a command line. Instead we're going to put these features where people are actually using them and finding the ways that they can make the most sense to take the best things that LLMs can do like language translation, a great thing like taking, you know, pattern recognition and large bodies of text summarization. We're going to take those things and we're going to move them straight into the places where people can use them. And I think there is some validity to this approach based on some research that Znet, so ZDNet and Aberdeen research did some research this year and what we found was we did a survey that we published last month that we found that 8% of people said they were willing to pay for AI features in their devices. That was far less than we expected. 69% said they would stop using the device if there were AI features they couldn't turn off.

[Leo Laporte]

Right.

[Jason Heiner]

So there's a huge gap between the story that Google and OpenAI and Microsoft and Anthropic are telling about, you know, these things and the general public's sort of massive enthusiasm gap there and maybe understanding gap right of these, of these things. And I, and our, our conclusion was that a lot of it is based on privacy, right? That they're, that people are not real trustworthy. They've seen what happened with especially Meta and some, you know, and at times Google with privacy and they're and they're not real sure that they can trust these AI features. And so, you know, Apple, I think take some of the features, sometimes they use AI and they don't even call it AI, they just try to make a smarter feature out of it. And my sense and what I, what I put forward in this piece is that that's probably the best way to start getting people used to using these features. Just give them better translation in the translation app. Right. It's still not as good as Google Translation, but it is better. I think that's a smarter way and it's going to get people more used to using these features because when they hear AI, people are, are pretty skeptical. You know that that survey that we ran was scientific survey of, of about 2, 500 people throughout the U.S. and yeah, it was pretty representative of, of various. Yeah.

[Leo Laporte]

So one thing Apple did gloss over is the fact that this was more of a crisis than they let on because they basically kind of beheaded the whole department. John Giannandrea kind of got sent upstairs. A lot of people left internally. It was a much more turmoil than you saw at wwdc.

[Jason Snell]

Yeah, Apple's never going to show that to you. They're never, I mean, that's PR101 especially. It's Apple PR101 is that you remember they could introduce a disastrous product or product feature and the next time they, they fix it, what they will not say is we're sorry, this is our apology. They'll say, well, this one is even better and they move on.

[Leo Laporte]

And just that once, to my knowledge. And that's with the trash can Mac over Jason's right shoulder.

[Jason Snell]

Yeah, it is, yeah.

[Leo Laporte]

They did apologize for that one.

[Jason Snell]

They did, sort of. But so anyway, it is, it's very much Apple, I would say, like Jason, my fellow Jason. I also wrote a piece about this that I said, Apple intelligence shifts gears. And the idea is, look, I think last year Apple felt so pressure for the industry and for, for investors and all that they needed to fly the AI flag and they kind of lost track of who they are. And this year I felt like full credit to them. They remembered who they were. The first off, everything that they announced is in the betas, which means basically they didn't vaporware.

[Leo Laporte]

No pre announcements.

[Jason Snell]

Yeah, if there are things that they want to ship in the next year that aren't going to ship this fall, they didn't talk about them. They'll do that later. I think that's super smart. I think that as Jason said, they integrated a bun of AI and ML stuff throughout the operating system. They also got off their, I think they're kind of, well, we need to invent it approach where they're like, yeah, put Chat GPT where it needs to be their update to xcode. They had this thing called Swift Assist that they showed us a year ago and, and never shipped. And I, I went into the same room with the same demo people this year and they said it's already out in developer Beta one and what they did with Swift Assist. So it's AI coding, right. And there's a huge amount of growth and pressure there. They took originally a feature with their model that they were going to work on and instead Xcode now will work with ChatGPT or literally any other on or off device model you want to plug in with a URL and an API key and it will be completely integrated with xcode. And that's a great example of Apple getting over it, getting over their discomfort with relying on third party LLMs when they need to. I think we're going to see more of that. I mean of course there are table stakes. Apple needs to have good models. They have allowed their app developers now to use their on device models, which could actually be a boon for third party app developers. But like they also are remembering, oh yeah, we make devices that people like to use and maybe we should focus on making the platforms for those devices good and not over promise because we're panicked about AI and it was just a good, like there's still lots of issues with Apple's AI infrastructure and their research and all of that kind of in the long run there are lots of issues there. Just try to do an image, image playgrounds and you'll see it. But, but it feels like they are understanding who they are a little bit more and I think that's good because last year I think they actually kind of lost the plot.

[Harry McCracken]

I'm a little surprised that Android has not raced further ahead with AI, given Google is Google.

[Leo Laporte]

Maybe they saw Jason's survey, Maybe so.

[Harry McCracken]

I agree with that. I also think besides people being worried about privacy, we've just seen so many AI features that are terrible and a smaller set of AI features that are really good and the industry has to put up with stuff that actually is helpful. And I was a little bit hopeful at WWDC because that seemed to be what Apple was focusing on rather to impress anybody that they were AI wizards.

[Jason Snell]

And the empowering of third party developers at a time when Apple's relationship with third party developers is kind of at a low the fact that you basically get access to their on device models for free and we know that those are, you know, those are private because they're on device. They opened private cloud Compute, which is their private cloud infrastructure that's open in shortcuts now. So not the third party app developers but like I think they're sending a signal that in the long run if you're developing an app for the iPhone or the iPad or the Mac, you are going to get access to Apple's private models and like that's a really good thing. Not only do you get the power of those models in your app for free or for cheap and in the case of the cloud model, probably, but also it's got the privacy story that goes with it and I think that that's a way that Apple can kind of say this is what being an app on our platform brings you. So like I think, I think they're making some smart decisions sometimes. I think we let the fact that they got behind on their LLM technology kind of cloud other places where they have been making good machine learning decisions for a decade, they blew that one. And last year it seemed like they blew that one and they were freaked out. And this year it seems like they've got a game plan. They still got it. Hey, I would say I'm much more positive about them in general. I do think they're going to need to prove it when it comes to their models but you know, otherwise I think that because they're open to other models, they're going to be okay.

[Leo Laporte]

We're going to take a little break because I do want to talk. I mean we probably won't spend as much time on it, but Google did release Android 16. That'll be a three minute conversation. Harry's just written a piece about quantum computing. IBM announced that it's going to have a real quantum computer in just a few years. A lot more to talk about in just a bit with our esteemed panel of tech journalists, Jason Heiner, Harry McCracken and Jason Snell. Great to have all three of you. This episode of this Week at Tech brought to you by Oracle. In business they say you can have better, cheaper or faster, but you only get to pick two. What if you could have all three at the same time? That's exactly what Cohere, Thomson Reuters and Specialized Bikes have since they upgraded to the next generation of the cloud. Oracle Cloud infrastructure OCI is the blazing fast platform for your infrastructure, database application development and AI needs where you can run any workload in a high availability, consistently high performance environment and spend less than you would with other clouds. How is it faster? OCI's block storage gives you more operations per second cheaper. OCI costs up to 50% less for compute, 70% less for storage and 80% less for networking. Better and test after test, OCI customers report lower latency and higher bandwidth versus other clouds. This is the cloud built for AI and all your biggest workloads right now with zero commitment. Try OCI for free. Head to oracle.com twit that's oracle.com twit we thank them so much for their support of this week in Tech so Google, it's funny. Maybe the timing could have been better. Launched Android 16 I have it on my Pixel. You'd be hard pressed to see a whole lot of differences. It's not a giant release, anything like what Apple's planning in the fall. In fact, what Google did and what they have been doing is held back some features for monthly Pixel feature drops. This has been their new thing of late. But there's a new live update notification which mimics the live activities notifications on iPhones, which by the way I really like. I'm glad to see that on Android there's group notifications so you can mush them all together again. Another thing Apple has been doing more important than anything else though I think advanced protection which adds security for your Android device initially again for the Pixel online attacks, harmful apps, unsafe websites, scam calls and more. You have to go to advanced protection to enable it. Anything else to talk about? There's dark artwork now in the media player. Health Connect lets you it's funny because people complain about Apple copying Android. Almost all of this is stuff Apple's been doing that Android's now. But this is good. You want cross pollinization between platforms, right? That's good for users.

[Harry McCracken]

I do think it's weird that Google has not done more to kind of make Android clearly kind of the first well known existing operating system where AI is absolutely at its core and is doing transformative things. Because if anybody has the opportunity to do that it's probably Google. And if they're working on that, they have not really told us about it or shown it. And in fact it sometimes seems like Android is not one of their top three or four priorities these days.

[Leo Laporte]

Yeah, there was a rumor, I don't think it's true, spread by the creator of One of the third party ROMs that Google was about to abandon AOSP, its open source version of Android I don't know if that's true. I think he was maybe overreacting. But still, Android does not seem as important to Google as it once was. Do you think they're worried that the courts are going to force them to divest?

[Harry McCracken]

They're worried about the courts forcing them to divest something, although that something might be Chrome or something, or their ad business rather than their mobile operating system.

[Leo Laporte]

I thought we were going to know it sooner than now. I thought we were going to know in the end of May, but I guess the judge is still debating that one. Let's talk anyway.

[Jason Heiner]

Chrome seems more likely. Chrome or that business seemed.

[Leo Laporte]

Or stop paying Apple $20 billion a year to be the. The default search engine.

[Jason Heiner]

I mean, if they don't, somebody else will, right?

[Jason Snell]

It's a lot of people.

[Leo Laporte]

Right? I'm sure Sam Altman's sitting there ready to write a check any minute now. That's Google's biggest threat, I think. Not government intervention, not Apple, but AI. And you're right, Harry. Gemini is a strong platform, but you would never know it, really. It doesn't get all the press. It doesn't get the attention. Partly, Google fumbled it with their initial releases of Gemini. Talking about putting Elmer's glue on your pizza and recommending eating rocks and things like that. That kind of stuff sticks in people's minds.

[Jason Snell]

In fact.

[Leo Laporte]

Fact, when you talked about your survey, Jason, I honestly, I don't think people are worried about privacy. I think they just are creeped out by AI in general.

[Jason Heiner]

There are a lot of things it's not very good at. So. So the hard thing about the chatbot is you do it and you figure out pretty fast. Like, I ask it purposefully ask it things that I know sometimes just to see if it's going to give me the right answer. And it's just a. It's still far too high. A number of times where it gives me wrong answers or inaccurate or just almost dangerously incorrect.

[Leo Laporte]

Well, that's how we judge search engines. All these years, you'd put your name in and see what they came up.

[Jason Heiner]

With, see what happens.

[Leo Laporte]

Yeah. Now we do that with AI and it's just as dumb. Google is buying out. This is actually, I think, a bigger story than I saw. I mean, the information had the exclusive, so maybe that's why. But Google is offering voluntary buyouts to employees across search knowledge and information. They're offering buyouts. This all started on Tuesday to US employees in core, which is the engineering team working on Google's underlying technical infrastructure, research, marketing, and communications. I'm not sure how big they are. They laid off 20,000 people at the beginning of the year.

[Jason Heiner]

Yeah, you know, I, I think that, I wish we had a chart here because I think this is related to. More related to anything than the massive over hiring that the tech companies did during COVID and you know, 20, 20, 21, 22. Right. If you look at the number of people they hired, you know, like Meta, for example, it's just now starting to get back to almost about where it was, you know, in pre. Those. Those years, pre2020. And there was so much hiring that they did, they just, you know, hired huge numbers of folks and they overhiered. And now they're having to rationalize some of their costs around that stuff. It was a bit of an arms race. Right. Whoever could hire, they were trying to hire the best talent the fastest that they could. And the thing that's really rough right now is I know some people who've just recently graduated that have graduated from computer science, computer engineering. They cannot find jobs because they have let all these tech companies have let go so many folks. And so now the market's been flooded with a lot of folks. And if you've tried to hire in the last six months or a year and you just are a new graduate, it's just pretty brutal.

[Leo Laporte]

There's another reason, and I floated this last week, but you guys might have some handle on this in the tax code. For a long time, the tax code allowed you to write off R and D expenses, which includes, by the way, engineers and coders in the same year that you spent it. That will change soon to amortizing those costs over five years, which is a huge change in the tax bill for these big companies. This was in the 2017 law the Congress didn't want. They wanted to keep it revenue neutral. So they didn't want to mention this change. But it is now going to kind of kick in. And this is an article by Katherine Bobb in Quartz. She says that that's probably the reason more than half a million tech workers have been laid off in the last two years because of Section 174. It's gutting in house software and product development because they can't write it off in the same year that it's incurred. Anybody got a line on that? Or is that. This was out of the blue when I read this, I went, holy cow. Yeah, I hadn't seen that for 70 years. American companies could deduct 100% of qualified research and development spending in the year they incurred the cost, that salary, software contractor payments. If it continues to, if it contributes to creating or improving a product that's off the top line right there, fully deductible. That was changed in 2017 and now, you know, it was changed in a way that it wouldn't really kick in until now, but it didn't start affecting people until 2022. So I don't, you know, I, I don't know.

[Jason Heiner]

It does make sense actually. It's interesting. Like, like I said, Leo, I wish I had the chart with me. But if chart at some of these companies, the hiring. 2020, 2021, 2022.

[Leo Laporte]

Right.

[Jason Heiner]

And then it's stair stepping downward. Right.

[Leo Laporte]

And that's why, or at least that's one contention, the tax code. Yeah, I haven't heard any coverage of this. So I, this is the, I saw this story in courts and it was repeated of course in a lot of publications, but it came out of courts. I don't, I don't know any more about it. I'm not an expert in this kind of stuff, but as a small business owner, I know if you can deduct it in the year you spend it, that's much better than spread over five years. That's a big, big difference. All right, moving on. Let's talk about quantum computing. Harry, you're writing a cover. Not a cover piece, but a big think piece for a fast company. I saw this this week in the MIT Technology Review. IBM is planning to build the world's first large scale error corrected Quantum computer by 2028. Now, Harry, I had thought quantum computers were like fusion reactors, the stuff of science fiction.

[Harry McCracken]

Yeah, I mean, well, they've definitely been making progress. And the piece I wrote, which will be up on our website shortly, is not so much a think piece as an explainer. Like all of a sudden, my colleagues who work at our headquarters in New York who are not full time technology people had been hearing enough about quantum that they were intrigued in a way they hadn't, which is how this story came. And it's trying to get kind of curious people who don't know a lot about this topic up to speed. The state where we are now is that progress has been made. It's not just stuck in the labs. IBM has made progress. Google and Amazon and Microsoft have all had major announcements in terms of breakthroughs in the last few months. Smaller companies like D. Wade have also kind of reached an inflection point. And we're not yet at the point where quantum computers are Doing the kind of stuff that people envision, which is essentially like coming up with amazing new drugs through forms of computing that you couldn't ever have done before. But they are getting there. We are at a point where I think there are some examples of quantum computers performing calculations that could not have been done before before and might even be useful. They're not yet ones that will change the world. But I think that for a long time people kind of thought that maybe it might be like fusion and something that people love to talk about as a breakthrough that would always be years away. And the biggest stuff is probably still years away. But I don't think that it's fanciful at this point.

[Leo Laporte]

The IBM computer is going to be called Starlink. The big breakthrough is in error correction. That's kind of the challenge of this. Qubits are kind of in superposition. They're a little fuzzy.

[Harry McCracken]

Yeah. I mean, essentially controlling quantum physics is really hard.

[Leo Laporte]

I'm amazed we can even consider it, to be honest with you.

[Harry McCracken]

It's extremely hard to get a qubit to respond exactly the same way every time. In the way that we're used to bits being very reliable and the way generally speaking, you solve that as you take a bunch of unreliable qubits and you pool them together into a logical qubit, which relies on all of them. But you get out the unpredictability by pooling them. And so logical qubits are really interesting. And if you want to look at numbers that relates to these computers, look at how many logical qubits they have rather than how many physical qubits they have.

[Jason Heiner]

Have.

[Leo Laporte]

The MIT Technology Review says that Starling will have 200 logical qubits. That means it could perform 100 million logical operations consecutively with accuracy. Although they say, Some experts say 100 million is a tenth of what you need. You need at least a billion error corrected logical operations to execute any useful algorithm. She quotes Wolfgang Pfaff, who's a physicist at University of Illinois Urbana Champaign. He says it's unlikely this will generate economic value.

[Harry McCracken]

I mean, one of the big challenges about writing about this stuff is that even the people who know a lot about it kind of disagree wildly about how soon it's coming. And a lot of the people who do know a lot also are working for these companies and they're invested in the idea that it is going to happen relatively soon. And if you're working at these companies, you can't say that a useful Quantum computer is 30 years away because you probably would not be investing in it. But a lot of the major work on this stuff sort of happened 20, 25 years ago, and it has gotten us somewhere.

[Leo Laporte]

IBM says 2028 for Starling. And then shortly after 2029, it plans to build Blue J, which will contain 2,000 logical qubits and will be capable of 1 billion logical operations. So maybe it is closer than we thought. Thought NIST is preparing for that. Of course, they have quantum hard encryption technologies available now so that you can start encrypting your stuff, preparing for a future where these computers will be able to crack strong encryption.

[Harry McCracken]

The software is a very large component of this and is kind of making progress in parallel because you don't, you know, once you actually have the quantum computers, you need algorithms that can accomplish useful things. But it's been a little bit of a chicken and the egg. But there are a number of startups that are entirely focused on algorithms and other stuff. Like IBM has this thing called Qiskit, which is essentially a platform for creating quantum algorithms that works on their computers. But it's open and not solely for IBM. Quantum computers.

[Leo Laporte]

Yeah.

[Jason Heiner]

I wonder where quantum intersects with the.

[Leo Laporte]

AI revolution, where Benita was asking that, does this make smarter AI because.

[Jason Heiner]

Or cheaper? Because one of the biggest challenges is AI is massively unprofitable, right? Like, LLMs are massively unprofitable right now because it takes so much computing power to do it. And so I, I wonder, you know, why? And there's so much, there's like so much hyperbole around generative AI, and. And yet, like, even open AI is. Is years and years away from profitability, right? As high as their value.

[Leo Laporte]

Well, wait a min. Altman, on his blog post this Week said that ChatGPT uses 0.34 watt hours for each query about what an oven would use in little over a second. And it uses 0.000085 gallons of water. A fiftieth of a teaspoon per query. Now, do you believe Sam Altman?

[Jason Heiner]

Here's what I do believe. I do believe that they are massively unprofitable. And they've said it. And they said we're not even close to being profitable. So however much it costs, they're using a metric ton of it to do everything that they're doing.

[Leo Laporte]

What he didn't mention is, yeah, 0.34 watt hours per query, but we do a billion queries a day.

[Jason Heiner]

Yeah.

[Leo Laporte]

So do the math on that. A billion times a 50th of a teaspoon adds up to a Significant amount of water, actually.

[Harry McCracken]

Generally speaking, I think the conventional wisdom among people who know what they're talking about is that Quantum and AI will work together. And there's an opportunity for Quantum to do things like train AI models. But Quantum is not going to supersede AI. And the future is probably about Quantum doing some things such as drug discovery, materials science, and working in concert with AI, which is also tackling some aspects of some of those same challenges.

[Leo Laporte]

Well, that's closer than it seems and farther than it looks or something. What is that quote?

[Jason Snell]

One more.

[Leo Laporte]

You know, of course, the president and Elon Musk had a little bit of a falling out. Elon's been trying to make up to the president, maybe scared that he was he'd lose some of the contracts that he has with the federal government, including the SpaceX contracts. I don't know if he's gonna. Because there isn't much competition for those contracts. Jeff Bezos rocket company Blue Origin has delayed the second launch of the New Glenn from late spring to August 15th. There isn't a lot of competition for what SpaceX does. Boeing, of course, is struggling as well. So that's the latest according to the information from Blue Origin. Let's take a little break. Come back 23andMe has a new owner just like the old owner. We'll talk about that and more. It's great to have you here. Harry McCracken, the technologizer at Fast Company, Jason Snell from Six Colors.com and editor in chief. No dashes at ZDNet. Mr. Jason Heiner. I checked. You're right. You don't use dashes. I thought that that was always eic was always a dash dash.

[Jason Heiner]

I love that you checked the usage.

[Leo Laporte]

I didn't believe you. How could that be true? I haven't gotten up to go get the AP style guide, though.

[Jason Heiner]

Spoken like a true journalist.

[Jason Snell]

Well.

[Leo Laporte]

Well, you remember that email was initially e mail?

[Jason Heiner]

Yeah.

[Leo Laporte]

And then it became email capital E. One word. And now it's lowercase E. The language evolves, obviously.

[Jason Heiner]

And I guess Internet was also capitalized originally.

[Leo Laporte]

Yeah. Do we not do that anymore? I remember I was shocked by my spell checker on Windows saying it's a capital I. I said what?

[Harry McCracken]

I capitalize it personally because it is just one thing. But that's not our copy standard. It's not a property. It's lowercase.

[Leo Laporte]

Chris, what would. What is the rule on that? It has to be a proper name. But how? The Internet is just a technology. Right? Or it's a.

[Harry McCracken]

To me It's a thing. It's a thing. It's a very large, sprawling thing, but it's a thing. I don't know, the New Yorker just. I started lowercase in it, which is why this discussion has been in the news late lately.

[Leo Laporte]

Yeah, I love that. I love that. See, we're all nerds here. You're all writers, we're all nerds.

[Jason Heiner]

So this is such a nerdy topic.

[Leo Laporte]

It's such a nerdy. Such a nerdy topic. Great to have all three of you. Thanks for being here. Our show this week brought to you by Melissa. So we just celebrated our 20th anniversary. Melissa just celebrated 40 years as the trusted data quality expert. But they are not resting on their Laurel laurels. I'm resting on my laurels. But they are not resting on their laurels because really they see their job not as, you know, cleansing your address records, but really as being your partner in data science. And they have in fact acquired some data science companies. They're really focused on transforming your business. And of course one of the technologies everybody's talking about these days is AI. And AI is absolutely in every industry now, from manufacturing and supply chain management management to the healthcare industry. But even the most advanced AI models can't fix the underlying data quality issues in your data. That's why you need Melissa. It's like having a data expert who never sleeps. Melissa's intelligence system verifies identity to prevent fraud in gaming operations. Used widely in the gaming industry, also used in the health industry, Melissa is used to ensure ensure valid patient and medicine identification in healthcare systems. That's super important, isn't it? Melissa securely updates and verifies constituent data across government databases. And Melissa's know your business enables verification and monitoring for financial institutions is something of course that's a compliance issue. Melissa's industry leading data validation tools are everywhere you want them to be, including now Microsoft Azure. Through the Azure data factory, organizations of all sizes can migrate using existing SSIS packages to the cloud without disruption while accessing comprehensive data cleansing capabilities, scale data quality operations faster, reduce infrastructure overhead and maintain existing enterprise data workflows all within Azure's elastic efficient maintenance free cloud environment with real time data validation, comprehensive enrichment, cross referencing verification with gold standard reference data data and intelligent anomaly detection. It's no wonder Melissa is the trusted data quality expert worldwide. And of course your data is always absolutely safe with Melissa. They securely encrypt all transfers. They have an information ecosystem security ecosystem built on the ISO 27001 framework. They adhere to GDPR policies and SoC2 compliance and on and on and on. Contact Melissa's team to learn how they can help you elevate your business business and improve your data quality. Get started today with 1000 records clean for free at melissa.com TWIT that's melissa.com TWIT we're longtime partners of Melissa and very happy to do so. Melissa.comTwit thanks, Melissa, for supporting this week in tech. So for a while, okay. 23andMe. What a story. Story. This was the genetic testing company founded by Ann Wojcicki, who is the sister to Susan Wojcicki, who recently passed ran YouTube. Her mother, Esther, rented the garage to a couple of guys named Sergey Brin and Larry Page when they were starting a little thing called Google. I mean, really Silicon Valley royalty. But for some reason, and I've never really understood why, 23andMe was really strong, struggling. Maybe people didn't want to do genetic testing. I did it. Everybody in my family did it. The board quit, not happy with Wojcicki's governance. Eventually they declared bankruptcy. And the bankruptcy court found a buyer, a biotech company called Regeneron, who says, we don't want your personal information, but we do want the genotypes and phenotypes. All that data for hundreds of millions of people that 23andMe has collected, that's going to be useful enough for us to make new drugs. That's what Regeneron does. Weirdly, Ann Wojcicki came back and said, outbid Regeneron and bought. And just this week bought 23 and me back. It was a nonprofit controlled by Wojcicki that made a bid that outbid Regeneron. Regeneron had bid $256 million. She bid 303 million. This. I don't know if this is going to reassure former 23andMe customers, because there's some value in 23andMe that beyond the business. I have already deleted my Data data from 23andMe. Apparently so have 15% of the customers. But that's not close to all. What is going on?

[Jason Heiner]

I didn't delete my data from it. I probably should have.

[Leo Laporte]

Should you? Do you think you need to?

[Jason Heiner]

I was kind of waiting to see what happens.

[Leo Laporte]

Yeah, the bankruptcy court hired a privacy ombudsman to make sure that that data wasn't passed on to some company that was going to use it against you.

[Harry McCracken]

Yeah, I haven't deleted mine either.

[Jason Snell]

Yeah, I deleted mine.

[Leo Laporte]

Okay, so it's 50 50. But here's the first of all, here's the question. Why did they fail in the first place? It seems like a good business.

[Harry McCracken]

I feel like none of these companies have figured out what, how to help people once they've done this test.

[Leo Laporte]

Now, what.

[Harry McCracken]

I mean, I did that partially.

[Leo Laporte]

You're 5% Ashkenazi Jew and 3% indigenous Indian.

[Harry McCracken]

And I, I've done a bunch of these partially because of the genealogical aspect of that.

[Leo Laporte]

And there is a lot of.

[Harry McCracken]

You get a long list of people you match. But, you know, in a lot of cases, I've tried connecting with these people and they never respond. And it's not even entirely clear, you know, why you should.

[Leo Laporte]

At least one case in my ex wife's family where a guy who was put up for adoption found his family and they had a reunion and it was a big deal.

[Harry McCracken]

There are a lot of stories like that.

[Jason Snell]

Yeah, I feel like 23andMe's problem was that they were. They thought that the genetic testing was the entire product. Yeah. And it seems like you need more, Right. You need a community that people are going to be in or you need other things. Like, I know, I think like ancestry and myheritage and sites like that that do also are doing, you know, they're doing family trees online. They're connecting people's family trees. They're doing like, deep scanning and search of old archives for newspaper articles and public records about people. The genealogy side of it, and you pay to have access to the larger trees and all that. I was like, okay, well, that's a product. And they also do a DNA screening. Right. That's a product that goes with it. And I feel like it's either that or it needs to be almost like a Facebook sort of thing where you've got a community. So it's all attached together. Because with 23andMe having done that, I don't even know, 10, 12 years ago, like, occasionally I got somebody who said, I think we're a third cousin or whatever, which isn't that exciting. But mostly, if you find a near relative and you, like, ping them, they're not even there because they ran their DNA, they looked and then they went away.

[Leo Laporte]

They haven't gone back.

[Jason Snell]

It's not an ongoing family. There's no ongoing product to pay for. Whereas if you're at something like one of these family tree sites, at least there's an ongoing product that's attached to it that is related. And I feel like those are the right businesses for this.

[Harry McCracken]

I give Ancestry money every month, but the only money I ever gave 23andMe was when I did the tests. And I also belong to Family Tree DNA, which actually has a McCracken group. So they've done a good job of the community side. And I can talk to other McCracken.

[Leo Laporte]

You found other McCrackens and very, very.

[Harry McCracken]

Very, very, very distant relatives. But that's still cool.

[Leo Laporte]

Yeah, I found a few. I never talked to them, but I knew that they must be relatives because they shared a family name. 23andMe had about 14 million customers, so it wasn't a, you know, they had a good customer base. The company, according to the Wall Street Journal, never made a profit. They burned cash across its DNA testing, drug development. That was the. By the way, and that's why Regenera and Juan wanted it, is that there is, in aggregate, some use to this data. Drug companies would love it. Right.

[Jason Heiner]

For sure.

[Leo Laporte]

And they also had a telemedicine business. I didn't even know about that. That they don't have a lot of debt, which is interesting. They just, they just didn't make any money. They were valued at $6 billion briefly after they went public in 2021.

[Jason Heiner]

Yeah. They just weren't sure what they exactly they wanted to be. Did they want to build a consumer business? You know.

[Leo Laporte]

Right.

[Jason Heiner]

You know, like, like everybody said it. If so, they needed to create a community. They needed to do sort of go sort of more the family tree genealogy route. Or did they want to be more of a drug company or a research company? More like Crispr Therapeutics or Fulgent Genetics or some of these other ones, Intellia, you know, Beam Therapeutics, those companies. And I'm sort of surprised that one of them didn't buy it for the data. That's sort of what I expected it to go in that direction. But there's also, from my understanding, spending. One of the things that probably hurt them was the, the regulatory environment over the past four years made it very difficult, almost impossible for any of those companies that I mentioned, genomics companies, anybody in healthcare, to make any acquisitions. That it was really difficult. And as a result, those companies have sort of, you know, complained that, that they haven't been able to go to the market to get. To get valued as what their true value is. Because the regulatory environment is so bad. They really have no hopes of getting any kind of acquisition if they were acquired. Right. To be approved because of the regulatory environment in the previous administration. And I guess there's. I don't know, the regulatory environment now is sort of tough to Read. I mean, we talked about Google earlier. Clear. It appears like Google is going to continue to, you know, the, the breakup of Google could, you know, be continuing even under the current administration. So it's a little uncertain, but I suspect, and this is only a, you know, an opinion, I suspect that, that Ann Rojicki, you know, believes that the environment's likely to get better and so that if she can pull off what she tried to do, she tried to take the company private when she there.

[Leo Laporte]

And that's when the board quit.

[Jason Heiner]

Yeah, yeah, the board rejected it. And so I suspect if she's willing to put this much money forward now that she anticipates that the regulatory environment is going to be better, they could go to the market and get a, you know, evaluation and that some of these other companies actually could be in the market and make an acquisition. Again, that's. That's pure speculation, but that seems to be, you know, why she would be willing to potentially put up a lot of, you know, her own. Her own funds and her own reputation.

[Leo Laporte]

It feels like, given that all four of us have done it, I bet a lot of our audience has done it, that it was like the pet rock of its time, like, 10 years ago. Like, everybody, you know, like, oh, let's all do our genetic testing. Like, I think we hoped we would learn something from this. Like, oh, you're, you know, you're going to have a higher incidence of heart disease or something. It turned out to really just be kind of faddish.

[Jason Snell]

It was about, like, ancestry. Like, you get your family stories of, like, oh, well, you're. You're really Irish or you're really Scottish or whatever.

[Leo Laporte]

Yeah.

[Jason Snell]

And so, and, and they did a pretty good job. I mean, it's not perfect. And there are people who are skeptical that they, they act more certain about your background.

[Leo Laporte]

Yes.

[Jason Snell]

Than you actually are.

[Leo Laporte]

But still statistical.

[Jason Snell]

It's better than, you know, it's better than a story handed down or the fact that everybody's family names come from their father's side. And so you, you might recognize a McCracken, Harry. Right. But you've got seven other out of eight that you might not even know or pay attention to. And like, so the genetics, it's like, aha, we've got science. We're going to find this out. And that's interesting. But I will say, for years, I've been telling people that my family name is Dutch because Snell means fast and Dutch. And it was not 23andMe that told me the truth. It was it was myheritage or ancestry, where I found a detailed. That took it all the way back to Heinrich Schnell, who was from Germany and when he came to South.

[Leo Laporte]

In German.

[Jason Snell]

German, But. But it was. And I could pinpoint literally the man who emigrated to America. That's cool. And went by Henry Snell once he came to South Carolina in 1720 or whatever. But that was the family trees that did it, not the genetics. So that's. That's why I think. I mean, it is cool to know that. That I, yes, I am. Am just as British Isles as I look. But. But yeah, it didn't. It didn't blow. Like, I know people. I know somebody who's a friend of mine who was adopted, who put his DNA results in, and it came back and said, this guy's your father.

[Jason Heiner]

Whoa.

[Jason Snell]

Right.

[Leo Laporte]

As it would if you really. If your father had to do it too. And then, in fact, there was a good story in the. I think it was New Yorker, the Sunday New York Times, a couple of weeks ago ago, about Chinese families who are discovering the real truth about. Or American families are discovering the real truth about the Chinese babies that they adopted. They were told stories that weren't, in fact, true. And that was because there are volunteers who are getting people in China who don't have the money to do genetic testing like 23andMe, to do it so that they could then try to find relatives. And it's been really interesting. I did a couple of years ago, I was interviewed by Frank Church, who's the father of modern genomics, and he has a company called called Nebula genomics that for 10 times more will do your full genome. I mean, literally, it's gigabytes of data. I did it, but it's the same problem now. What? Okay, great. I have my genome, my full genome, which is not what you get from 23andMe. There are companies you can send it to and they will tell you, oh, you know, you shouldn't eat sugar or something. But it's not hugely valuable. I think that's why I think of it as a fed. You have 80. Yeah. The same day that they announced Ann Wojcicki was going to be able to buy 23andMe, 27 states in the District of Columbia sued, saying customers did not expect their DNA data would be sold. I guess that's probably true. And the states are opposed to the sale of it without direct consent. Each and every one of us has to say it's okay before their data is sold.

[Jason Heiner]

Although that was valuable. Part of the company.

[Harry McCracken]

I mean that idea has been looming over this entire industry for as long as this industry has existed. So it's not a complete shock that push would come to shove with at least one of these companies eventually.

[Leo Laporte]

Oregon Attorney General Dan Rayfield says people did not submit their personal data to 23andMe their spit, thinking their genetic blueprint would later be sold off to the highest bidder. I guess that's true. But what did we think?

[Jason Heiner]

What do we think was going to happen? I don't know.

[Leo Laporte]

What did we think was going to happen?

[Jason Heiner]

Yeah, we're probably a little more cynical than the average, you know, American. But you know, I think, yeah, I don't. And also know a lot about the tech industry and how things are monetized, you know, and that when things start they, they do a lot of things and when they have to actually make a profit, they do a lot of different things.

[Leo Laporte]

Right. There's some, something in there because she's willing to pay hundreds of millions of dollars. It's a nonprofit she's running called TAM Research Institute. There is some value in it.

[Jason Heiner]

Obviously want to have that much, to have that much data right. On that many people is, is incredibly valuable. Like the whole value of what's left in the thing, you know, is very likely. You know, the, the data from the people. Right, right. So I, I think that it's just a long game. They weren't quite sure who they were and what they wanted to do with all that data. Data. But long term, you know, there, there are. Is likely a treasure trove of information that's medically valuable, that is valuable in, in lots of different ways that you could build businesses on that you could make medical, potentially medical breakthroughs on. That's why I expected it more to be one of the genomics companies, companies to be more likely to buy these things to, to have all of that, you know, all of that data. Now it's not as valuable as like you said, Leo, the whole genome because it's only, you know, mapping part of it. But it is interesting.

[Leo Laporte]

It is, it is not even just the genotype because remember you answered a bunch of questions, you filled out a bunch of questionnaires. That's your so called phenotype and it's the connection between the genotype and the phenotype that's a value. You know. Does your family have a history of heart disease? Okay, well we have your genotype to match up to that. That's what's real value. Apparently what hasn't been settled in law. And this lawsuit is expected perhaps to help settle is, well, do you have the rights to your genetic data? I mean, if you give it to.

[Jason Snell]

A company who owns the copyright on.

[Leo Laporte]

Your genetic code, yeah, the lawsuit could be significant in helping to establish. This is from the New York Times, not just whether a customer needs to give consent to sell the information, but who actually owns that data.

[Jason Heiner]

Is it your intellectual property in some way?

[Leo Laporte]

Very interesting. And there is a public interest, of course, in getting that anonymized data out there because it could develop new drugs, new treatments. It could be very valuable table in a variety of ways.

[Jason Heiner]

This could certainly intersect with tech more and more, too, with the wearables and, and also, you know, biometrics and, and things. Right. We're starting to give more and more of who we are to the tech companies, and that's only likely to, to increase as more sensors get put in devices. And it's. That's, that's not quite as sensitive as the information in 23andMe, but it's getting there.

[Leo Laporte]

Right? All right, let's take a little break. You're watching this week in Tech, Jason Heiner. I always want to say nice things about you, Jason. Maybe it's, I don't know, you got me elected to be president of the Internet back when it had a capital I. So thank you. I appreciate that. Always good to see you, my friend, actually, all three of you. I, I have great affection for you, too, Harry McCracken. Known you for years. Your wife Marie is sitting in the other room with her cats. Give her my best. Nice to see you once again. And of course, Jason Snell, who I get to see every week because he also, among his many other efforts, is on about 100 podcasts. He's on.

[Jason Snell]

We have Mac Break weekly every Tuesday. And my wife is also in the other room with cats.

[Leo Laporte]

There you go. Actually, so's mine, so there you go. You got any cats, Jason?

[Jason Heiner]

Our cats recently have passed on in the last year or otherwise my wife would be in the other.

[Leo Laporte]

She's dreaming of cats. She may have gotten one even now while you were doing the show. She might have just. That's what. Basically that's what Lisa did. She. She said no after we put our cat down a couple of weeks ago, which is very sad. Both of us were in tears. She said no more cats for a while. I gotta get over this. Next day, we got a cat. It was like, it was fast. And we are just as much in love with this one as we were with the other the others? Yeah. She's turned me into a cat person. I used to be a dog guy. I don't know what happened. I don't know what happened. This Week in Tech is brought to you this week by Drata. If you're leading risk and compliance at your company, you're likely wearing 10 hats at once. I mean, you have my deepest sympathy. You're managing security risks. You have got compliance demands and of course, budget constraints, all while trying not to be seen as the roadblock that slows the business down. But GRC isn't just about checking boxes. It can be a revenue driver that builds trust, that accelerates deals. It strengthens security. That's why modern GRC leaders turn to Drata, a trust management platform that actually makes your life better. It automates tedious tasks so you could focus on reducing risk risk improving compliance. And of course they're going to help you with that too. And scaling your program. With Drata, you can automate security questionnaires, evidence collection and compliance tracking. So it takes a big load off of your plate. You'll stay audit ready with real time monitoring. You'll simplify security reviews with Drata's Trust center and AI powered questionnaire assistance. Instead of spending hours proving trust, you build it faster with Trust Drata ready to modernize your GRC program? Visit drata.comweekintech to learn more. Again, that's dratad R-A-T a.com WeekendTech know this? Just weekend Tech. Make it a little shorter for you drata.com weekendtech to find out more. We thank Drata so much for supporting this Weekend Tech Club TWiT Discord is posting pictures of their doggies and their kitties. If you don't know yet about the Club Twit Disc Discord, please consider joining our clubs. 10 bucks a month you get some great benefits. Ad free versions of all the shows pictures of dogs and cats in the club Twit Discord. You also get a great conversation. We're doing now all the keynotes. The developer week keynotes were all in the club only to avoid takedowns from a certain fruit company. But that worked out really well because we got your input as well. I invite you to join the code club. Help us out. It's about 25% of of our operating revenue now comes from your membership. That's a big important part of what keeps Twit alive. Join the cool kids. Look how fast Pretty fly for assist guy is on the AI image generation. How many cats is too many? Asks Leo Laporte Club Twit Join the cool kids club twit.twitch tv club twit. Can't wait to have you in the club. We've got some big events coming up. Actually, we just did Chris Marquardt's photo thing. Micah's crafting corner is this week. It's a chill place to sit down and do some coding. Home theater geeks is going to do a chat room Q and a special on June 25th. You want to be there for that. And my old friend, we were talking about it before the show. The guy I was at the ballpark with during the 1989 earthquake, Norman Maslov, Mazzy to his YouTube fans will be talking about vinyl with us. And I'm trying to book a second piece after that about the rise of the MP3. So we can really cover the history of music in the 20th century. But that'll be the 27th with Mazzy's music. Our AI users group normally on the first Friday of the month, but that's the 4th of July, so we're going to put it off to the second Friday, Friday this week. Anyway, lots of stuff that happens in the club. So please.

[Jason Heiner]

You did the, the one. The. The WWDC chat or sort of live stream in, in just in the community.

[Leo Laporte]

Yeah. Well, so over the years, Apple's gotten more and more prickly about it. What we, what we've always done for almost 20 years is stream the keynote and then, you know, Mystery Science Theater 3000 STE Styles comment on it, which in my opinion should be fair use. It's a journalistic commentator commentary on a news event. Apple doesn't feel that way. And it wasn't content idea. It wasn't the automated takedown. They had a lawyer write us a letter in YouTube and then they went after us on Twitch. And while Microsoft and Google haven't done that, I just thought, you know what? It's not worth worth. I think it's defensible, but I don't want to defend it in court. It's not just not worth the trouble. So we just moved it into the club and there's a lot of members in our club and I think they enjoyed it. So that's just how we're going to do it.

[Jason Snell]

It's absolutely fair use for you to comment on that while you're showing it. But you know, the problem is we live in a world where fair use is after the fact. And all the content scanning on every video platform will just kill things. They just.

[Leo Laporte]

Yeah, and I Think we can't afford to lose our YouTube or Twitch channel. We don't want to take the chance of that.

[Jason Heiner]

But you know, it's interesting like, of, you know, necessity being the mother of invention. Like, it's, it's really great. I think that you're, you're doing it that way.

[Leo Laporte]

It's a community. Yeah.

[Jason Heiner]

And, yeah. And the ways, and even you looking at the current environment and thinking ahead. Right. Like of. Of the way that the environment is going both with AI and the way that the YouTube algorithm is and all of these things. But you know, you doing what you're doing so smart of like going even deeper with your most loyal members. Right. Even the people you have direct connections with versus relying on any other platforms to bring you your, your visitors. Right. It's same thing. Jason Snell too, and in the way that he runs, you know, all the things that he does, you know, going more deeply invested in the people who care the most about the things that you do and the things you see say is really, really smart. And I think as the media, whole industry is in the midst of a major transformation and probably even more so over the next 12 to 24, 36 months. And I think that the things that you all are doing is really the way that more and more media is likely to go.

[Leo Laporte]

Yeah. I just saw an article that said that while digital advertising spend is going down, down in general, it's doing nothing but going up in the influencer market. You know, the YouTube stars and so forth. I guess not a surprise, but oops, let me find that. But this is the Atlanta Voice talking about the influencer marketing industry. Ad revenue set to grow 12% in 2025 to 22 billion with dollars. Yeah, it's kind of an amazing story. I guess we all are creators, right? But Jason and I decided to go it. I don't know about Jason. I didn't have much of a choice but decided to go it independently and I think I certainly don't regret it. I know you don't, Jason.

[Jason Snell]

Yeah, I don't. Harry had a great site on his own too.

[Jason Heiner]

That's true.

[Jason Snell]

He's gone back in.

[Harry McCracken]

Back in the golden age of vlogs.

[Jason Heiner]

I'm the only one chicken enough not to have done it in this group.

[Jason Snell]

You've held on to a job for a long time.

[Leo Laporte]

Yeah. That's all. That's all. You kept your job. Good job.

[Jason Snell]

Pretty good. Pretty good.

[Leo Laporte]

I don't know. You know, if I had, I probably, if I still had a TV career would still be doing that. But I do really much prefer working for myself. It's, you know, I, I like having the control.

[Jason Snell]

And we're fortunate to live in an era where broadcasting can be to a smaller audience and still be successful. And that we. The tools, like, even in the last 15 years, the tools that the Internet provides for people to do what we do.

[Leo Laporte]

It's amazing.

[Jason Snell]

It is amazing. There are so many different little companies that have like, you know, this tool that will let you do this aspect of it. And, and I mean, for me, hardware.

[Leo Laporte]

And software, a lot of my stuff.

[Jason Snell]

Is cobbled together from. It's not like there's a. I know, like substack wants to just like everybody just do substack. But like, you can also build from, you know, pick and choose. You know, you stripe and you use WordPress and you use Memberful and you use all of these different sources and you can like, put them together to make the product that you want and then reach an audience of. It can even just be a few thousand. And suddenly you've got your, you know, you've got your job.

[Leo Laporte]

Even Apple kind of acknowledged that at wwdc, they added tools. They even said for podcasters on the iPad.

[Jason Snell]

They did. They gave. They gave all of us iPad podcasters a solid, which is you can record. So so many of these podcasts, not twit actually, but many podcasts are edited after the fact. And what you do is you're doing it live over the Internet, but that quality is a little reduced. And so everybody records locally. And then you send your file to somebody and they put them together and they make it. It's like you're all in the same room. It's full quality. And the iPad just couldn't do it. You could do it in the cloud, but you couldn't also, while you were using Zoom, for example, also record your voice locally on the device. And they just tossed it in there. And that's the feature, you know, a lot of us have been asking for, for years and years because I used to travel exclusively with my iPad, kind of like Harry, I still like, prefer to do it that way if I can. And then I either need to bring a laptop just to do podcasts, or I needed to rig up a thing where there was like a recorder that was recording on a memory card that then looped into the iPad. And now, you know, I tried it out. Actually. Dan Moran and I did our Six Colors podcast last Friday. I bet you we were the first podcast to be entirely record on iPad, using iPadOS 26 because both of us did that on our iPads using a sure MV7 microphone. And it worked. It worked.

[Leo Laporte]

It's amazing.

[Jason Snell]

It's a. That's a big one. Big win.

[Jason Heiner]

So unnaturally excited when they announced that in the. I was like, I never expected or at least like I thought it was gonna be another 10 years or something.

[Jason Snell]

Before they did that. I know that. What that and the two years ago they did Final Cut. I mean Harry mentioned this in passing, but two years ago they introduced Final Cut for the iPad and Logic for the iPad. And the untold story there is you could do Final Cut for the iPad and then export your 4K project and then you sat there and stared at the screen. But like if you wanted to go check your mail code on your Mac, you can't because it goes, I can't export video in the background. And they added this function which would be great for rendering and video exports and audio exports and file copies, big file copies where it just sticks it all in the dynamic island island instead and lets you actually, because I would use Harry, that's how I would always use Slide over is I'd start an audio export and then I'd like open my email and slide over so that I didn't close the export. So yeah, some big wins for like niche audiences. But you know what, hitting, you know, closing a bunch of gaps for niche audiences is one of the things that helps build loyalty to a product.

[Harry McCracken]

I think it should be good just for stuff like Dropout.

[Leo Laporte]

Yeah.

[Jason Snell]

Oh yeah, yeah.

[Leo Laporte]

I feel like it's more than that. I feel like it really is. Look, we know Warner Discovery just spun off the profit making part of the business, the movie studio hbo and then took all of the streamers and said you, you go over there and be your own business because we don't want you dragging down our results for the next five years. It's, it's. Everybody knows that that's what's happening. Look at this. YouTube said its ecosystem last year created 490,000 jobs and added $55 billion to the US GDP. Now that, that is YouTube's own reporting. But even if it's only half that, it comes from a research study by Oxford Economics. It's remarkable. Remarkable.

[Jason Heiner]

Yeah.

[Leo Laporte]

This is a YouTube report exploring YouTube's US impact in 2024. It's pretty clear this is the up and coming television for the new generation, Generation Alpha. It's YouTube.

[Harry McCracken]

We used to complain all the time that the new iPads, especially the iPad pros, were incredible in terms of computational muscle. But the software was way behind and Apple wasn't really taking advantage of it. And while I do have some reservations about some of the stuff with ipados 26, you can't argue that the software hasn't taken a large stride towards being something capable of soaking up all of this computational power.

[Leo Laporte]

Yeah.

[Jason Heiner]

Prior to 2023 they've done these two moonshots, the last WWDCs, right. Like Vision Pro and then Apple Intelligence, they were just like such huge moonshots. But prior to those two years, if they made this, this, this iPad OS announcement back then, this would have been all anybody was talking about. It would have been the hugest story. It would have been the, the biggest thing. It's just that our perspective is a little different and there's still also that a little bit of, of hangover from those things in terms of, you know, we're still talking about Vision OS and we're still talking about Apple Intelligence and were throughout the this. But the, what happened with the iPad was so big and I think potentially the both sort of short and medium term impact of it is pretty, pretty huge. I was going to ask earlier, Harry, when we were talking about if you use the 13 inch or the 11 inch iPad because you use it so much.

[Harry McCracken]

I currently use the 13 inch. I had the 11 for the round of iPads before that. Essentially 11 inches more portable and 13 inch is definitely better for drying and probably better for productivity. Although I have to say that I'm a little disappointed by the battery life on my 13 inch compared to the 11 inch I had which I still use as an E reader. Also the thing that's changed completely since the iPad was new was when the iPad came out, 10 hours of battery life was amazing compared to a Mac. MacBook and MacBooks have raced way past 10 hours and so now 10 hours is a little disappointing. And I would love it if the iPad Pro, in order to keep a clear difference with the iPad air, maybe got ever so slightly thicker and got battery life that was more MacBook like because the game has changed.

[Leo Laporte]

Stand back. According to Gurman, next year 2026 Apple's going to to release a redesigned MacBook that'll have the same screen as your beautiful 13 inch iPad, a beautiful OLED screen and probably the M. What it will be M6 by then, I guess M5, M5. Okay.

[Jason Snell]

The ultimate computer.

[Leo Laporte]

Yeah, yeah. Stand back.

[Jason Heiner]

I use the 11 inch iPad and so it's really interesting and I love the 11 inch iPad.

[Harry McCracken]

It was mostly like happy with that except for drawing like a larger screen.

[Leo Laporte]

For the pencil, the big one is basically a laptop, isn't it? It's same price, it is about the.

[Jason Snell]

Same weight as a MacBook Air. I use the bigger one mostly because I'm a keyboard snob and the keyboard feels a little cramped to me on the 11.

[Jason Heiner]

Yeah.

[Jason Snell]

But I, every time my wife's got an 11, every time I pick it up I'm like, oh, it's so light though. It's so nice. Yeah, so. But yeah, I mean it's a good problem to have.

[Jason Heiner]

It is. I, so I, I use the 11. I can't draw so that, that helps. I don't have that holding me back. But the keyboard is a little bit cramped. But I also, when Apple made the 11 inch MacBook Air, that was my favorite. That was like my favorite Mac ever. And I was sad when that thing was. Went out of. Yeah. Went out of circulation.

[Leo Laporte]

Was that the MacBook nothing? That was the nothing.

[Jason Snell]

No, it's the 11 inch air. There used to be 11 and 13 inch airs, then they did the 12 inch nothing. But they did the 11 and 13 before it was retina and they only made a 13. There was an 11 and that was the best, the best laptop. Absolutely.

[Jason Heiner]

Yeah. That thing was incredible and I can only imagine that size now, even with Mac Silicon, Apple silicon. But the 11 inch. The interesting thing about the new OS 26, IPADOS 26 is most of the time I want to do kind of what you expressed earlier, Harry. Like I want to focus on one thing, right? I want to focus on doing something and be able to use the iPad as a way to do that. But every once in a while there's like something I need to do where I need like a couple windows, I need to like have two things open to copy and paste or I need to reference something and I never. On the 11th seven, it's really hard to do the, the tiling. They're so small and so challenging that the new one, I, I almost feel like, you know, the handcuffs have come off. Like, okay, I can have three things open at the time. I can just be able to click back and forth on those moments that I needed. It's probably only 10 or 20% of the time maybe, but, but that's where I, I think the power of this iPad OS 26 potentially could really make the iPad give the iPad a bit of a renaissance even for older ones. Because we know that even going back to the last several generations of Apple Silicon, they're going to run really well. And imagine having even those. I have an M1 11 inch that still runs amazing and is running ipados 26 really well. And yeah, I was thinking about that and thinking like this is going to. It's almost like giving new life to some things that have been around for a while that, that now are even more powerful.

[Harry McCracken]

I also thought it was interesting that they're bringing the new windowing interface to all the iPads down even to the mini, rather than kind of building it primarily for the iPad pro type of people who are doing really heavy duty stuff.

[Leo Laporte]

I can't imagine many people will use it on the Mini though, will they?

[Jason Snell]

No such a small. You never know. It does, it does speak to the fact that they, that they wrote a new windowing system like that, that the stage manager system, they had to lim like the M series and then they brought some of it back to like the, the very last before they went to the M series. And this version, the tech has moved along enough, there's enough RAM in those devices now and then that they built this thing for all iPads to have. And it's not, you know, it's not walled off at all. I think that speaks to the fact that they really stripped it back to nothing and started again and built something completely new which, you know, a lot of new features are just new additions to old features. And this one, you know, is not like that. This one, they liter threw the old one away. And I know there are issues with that that Harry has, but I think in terms of the performance, that's one of the reasons why it's per. It performs as well as it does and it's as compatible as it is because they, you know, they built it for all iPads and it, and it works on all iPads. I think that's pretty sweet.

[Jason Heiner]

That surprised me so much. If you, if you like the iPad, you've sort of been used to disappointment for years, you know, and so maybe it's like our expectations are so lowered, but I, I was just like so surprised at everything that came to the iPad this year. And to Jason's point, the fact that they really have put it across everything does speak to this becoming such a core feature, core functionality of the OS now, which is fantastic. Certainly likely to be much more performant because of that. And there was these reports beforehand that this was only going to happen like if you docked an iPad Pro in an Apple, you know, keyboard. It's not that at all. Right? There's none of that like it's system Mode.

[Leo Laporte]

When the iPad first came out, it was just seen as a big iPhone, frankly. And a lot of companies like Meta said, well, we're not going to develop for it. It's just a, it's a big iPhone. Meta finally has put out WhatsApp for the iPad. Do you think companies like Meta will, will we ever see an Instagram for the iPad? Are they going to start treating it like a real computer?

[Harry McCracken]

The works now.

[Jason Heiner]

Yeah, finally.

[Leo Laporte]

Ah, when did they announce that?

[Harry McCracken]

Well, I saw news and rumor of that at least in the last rumor week or two.

[Jason Snell]

Yeah, it sounds like the code is there and they're going to release it at some point here soon. So yeah, hell froze over.

[Leo Laporte]

That's the other difference. Now, the big difference, and I mentioned this on macpa Weekly on Tuesday, the big difference is you can only install apps from an app store on the iPad. It's not a real computer in the sense that you can sideload.

[Jason Snell]

Yeah. And that is both a pro and a con. Right. Because. Because for some of us who would love to have, you know, a wide open kind of platform, the Mac will do it, but the iPad won't. On the other hand, the other, you know, the other argument is this is a platform that you might be more comfortable on because you know that everything is passed through the App Store. It really is sort of like what you want out of it. You know, you. Apple offers both kinds of devices and you know, unfortunately, philosophically, Apple has just decided to draw the line on, on the shape or functions, the hardware. So like you can't run Mac OS on an iPad and you can't run iPad OS on a laptop. Instead you need to pick. And on one side you get these features and on the other side you get these other features. And that's just, at least for now, that's where we are. You can't use an Apple pencil on a Mac either.

[Harry McCracken]

One minor but cool thing, which I don't think was mentioned during the keynote, is that if you add a web app on an iPad to the desktop so you can open it again from an icon, it will perform like a web app. Whereas in the past that web app had to have some, some code in its manifest saying treat this like an app and, and run it at full screen if it's on an iPad. And they don't need to do that anymore. So I think if you're not happy with the, the app store versions of apps, which in some cases I'm not, I'd rather run the window web app a lot more of Them should run really well on an iPad than in the.

[Leo Laporte]

This isn't PWAs, though. This is being a PWA without the support of a pwa, basically.

[Harry McCracken]

I'm not sure. I mean, I'm not sure if it's technically a pwa, but.

[Jason Snell]

But it can get you along with. Federico Viticci wrote about this on Mac stories recently, how he was actually starting to use more saved website apps on his iPad.

[Harry McCracken]

I do a lot of that because.

[Jason Snell]

Some of them are really good. I mean, this is the thing is like native apps, apps are great. Some companies do not invest at all or very much in the native apps and they generally can invest. Some of them invest a lot in their web apps. And some of those web apps can be really, really good. On an iPad.

[Harry McCracken]

Like Airtable, which we use at Fast Company, the iPad version is not so great and it doesn't do everything that the full version does, but the web version is fine. And so I use that on my iPad and that already ready works like an app. But a lot more apps should look like apps if you save them to the desktop as an icon now.

[Jason Heiner]

Yeah, that's how I use WhatsApp on iPad and some of the volunteering work I do, I have a lot of WhatsApp groups and so I love using the.

[Leo Laporte]

It's nice to have that real estate.

[Jason Heiner]

Yeah, yeah, it's great. Yeah, it's much better than doing it on your phone, for goodness sake.

[Leo Laporte]

Well, as a testament to how much interest there is in this, we did the whole Apple segment at the beginning of the show, but somehow we got back to it.

[Jason Snell]

This is the overage, overage segment about. About Apple. I wanted to mention one of the things because since we're on this topic again, one of the things that everybody. Apple Intelligence. Apple Intelligence, Right. And what were they going to do this year? And I found it hilarious that when I came out of the event, the two things I was most excited by were the Mac and the iPad. And so when I say Apple was kind of like sticking to its knitting a little bit, I was like, it's kind of comforting that you walked out of an Apple event and the things that you thought about were the Mac and the iPad, not about Apple Intelligence or a Vision Pro or something like that, but like it's like real meat and potatoes, the Mac and the iPad. So to give the Mac its due, the new Spotlight, which is only on the Mac right now, that has support for like all of these kind of power user features, they added keyboard or clipboard history to macOS for the full first time. It's like the big missing piece of productivity built into the system and it's their inside Spotlight. They've got the quick keys thing which is you know like launch bar. You know I think pioneered this, the idea that you can launch something with a couple of keystrokes and it will learn and figure it out. They built all of their app intents frameworks into Spotlight so you can do things like send an email from Spotlight because it knows how to talk to the mail app and it knows what blanks you need to fill in in order to get it to send off that action. It's like it's like building a shortcut from the command line for a one off and like that plus and then shortcuts on top of that has access to all of their ML models all the way from on device through private cloud out to ChatGPT. Like as a Mac user I'm like oh man, I can't wait to get my hands on this stuff. And like I can't remember the last time that I came out of an Apple developer event really thinking the Mac was one of the big winners. Usually the Mac just kind of comes along from the ride these days but like it was a good iPad show, it was also a really good Mac event. So I'm, I'm hyped for Spotlight which I cannot believe I've ever, I've never said that before. I'm excited for Spotlight. What is going on.

[Leo Laporte]

I'm actually, actually considering taking Raycast off because a lot, of course Raycast has more features but a lot of the features of Raycast I don't use.

[Jason Snell]

Yeah, I've been using Launch bar for like 15, 20 years and I think with the clip around and they still update it but I use it, you know, I use it as a launcher and for clipboard history and I look at and I don't use more advanced clipboard managers that are out there. I don't use those features at all. I just want a list that I can pick from previous clipboards and so yeah, I might be able to just dump it too Leo and just say.

[Leo Laporte]

Doesn'T it feel like you want to use whatever Apple has decided to do natively rather than add a third party?

[Jason Snell]

Yes. It also gives me encouragement like talking about the iPad and, and how Harry especially uses it so much. Once they put clipboard history and quick launching and stuff like that into Spotlight on the Mac my immediate thought is it's only a matter of time before they bring it to the iPad too, and that'll be great. That would be amazing.

[Harry McCracken]

I thought it was a good wwdc for vision OS2, by the way.

[Jason Snell]

I agree.

[Harry McCracken]

Despite the fact that they've acknowledged this is not a mainstream product, I don't think it's a back burner product. And they actually are putting a lot of work into improving this operating system. And the new Personas are pretty amazing.

[Jason Snell]

They're so good.

[Harry McCracken]

It's a much, much more realistic digital version of you that's created almost instantly and the widgets are cool. And it supports GoPro and Insta360, 360 degree videos.

[Jason Snell]

Now we debate Vision Pro a lot on MacBook weekly, but I'll just say here that I think what we can all agree is it's not a product for the present. If it's anything, it's a product for the future. It's a bet Apple's making on being able to create AR glasses and other products down the road. And if it is a product for the the future, what you want to see is that they are iterating. Right? You don't want to see it dead in the water. And I am impressed that now a couple of years in, they do keep adding features, they do keep making improvements. The geographic persistence thing is they had to have that if they're going to do this, they got to have these features. So they're like, it would be a real red flag if they're like, yeah, Vision os, who cares? But instead they're pushing it forward. I think we could all, as somebody with a Vision Pro, it's great to see those features, but really what they're doing is just kind of loading up on the features so that when they finally, if they finally make products that are more affordable and more wearable and all of those things that are a problem, their OS base looks pretty good. The software is not the problem with that product line. It's the fact that they have a heavy thing that costs $3,500 that's the problem. So push the software forward. I agree. And the perspective Personas thing, which was my least favorite thing when they announced.

[Leo Laporte]

The product, so much better.

[Jason Snell]

It was the dead eyes, the creepy and the new Personas. Like they made spatial Personas which were better. And then these new Personas are, are even better. They're like, it's amazing how good they are. So I love that they're making advancements. I agree completely here.

[Harry McCracken]

It's not like tvos, which is still a hobby all these years later and something they don't they aren't really paying attention to. They seem fully engaged by Vision OS. I think even if I never buy a $3,500 headset, I'm still happy to see them do that.

[Leo Laporte]

You don't have a Vision?

[Harry McCracken]

I do not own one, no.

[Leo Laporte]

Yeah. Jason, do you have a Vision Pro?

[Jason Heiner]

I do have a Vision Pro. I'm a Vision Pro user. This is one of the things I was really looking for to see how much did they talk about it because I was worried about it becoming TVOs or home OS. Right. But it wasn't like, like, you know, they, they definitely invested in fixing and incrementing things that like needed to get better. Right. And so that's a clear indication, I thought that they are still investing in it. They still see this as the future. They, they took, they took it on the chin. There were some things they didn't do great. You know, I think the biggest problem with Vision Pro, it was not the product. I think there were two things that were the, that were wrong. Like one, the stupid thing on the, to make your eyes on the front was, was silly. It didn't need, they didn't need that. It would have been a lot less expensive that they didn't do it. And then two, they just marketed it so badly. Like they really, you know, it should have been and it was a developer kit and I think they don't make developer kits so they weren't going to market it. They had to make something that was still a consumer product. But I, I feel like they over marketed it a little bit and then we saw that a little bit again with Apple Intelligence. Last year they over marketed Apple Intelligence a bit and it's kind of not like them. I think in both cases it felt very, very kind of unlike them to do, to do that. And I do feel like this year that they have sort of gotten a little bit back, as Jason Snell said, back to their knitting of like they are. They've gotten burned a little bit by some of the way they've messaged things. I still think that there's a, there is a great product and a great roadmap in Vision Pro. And so I was excited to see them, you know, do what they did.

[Leo Laporte]

Well, I hope you've enjoyed our Apple Reflux. Yeah, it just keeps coming back.

[Jason Snell]

I thought that was a brilliant hosting move by you, Leo, where you're like, we can't start the show with two full segments about Apple, so we'll just do one and then we'll move on and then like, oh, there's overflow. Stuff it in at the end. Put it back in a. I tried.

[Leo Laporte]

To do a YouTube segment, but nobody wanted to.

[Jason Snell]

So sorry.

[Leo Laporte]

That's fine. We will have more with who knows what the topic will be with Jason Heiner, Harry McCracken, and of course, Jason Snell. You're watching this Week in Tech. Our episode this week brought to you by my favorite VPN. You probably even know I'm going to say ExpressVPN, right? If you've ever browsed in incognito mode, you probably know, I hope you know it's not as incognito as you think. Google recently settled a $5 billion loss lawsuit after being accused of secretly tracking users in incognito mode. Google's response? Well, incognito doesn't mean invisible. Well, in fact, all your online activity is still 100% visible to third parties, even in incognito mode. Unless you use ExpressVPN. It's the only VPN I use and trust. And you better believe when I go online, especially when I'm traveling, I'm in an airport, a coffee shop. In other countries, ExpressVPN VPN is my go to. I think everybody needs ExpressVPN. Why? Well, without ExpressVPN, third parties can still see what you're doing, what websites you're visiting, even in incognito mode. Your isp, your mobile network provider, the admin of that sketch WI FI network at the airport. They all know what you're doing. And a bad guy sitting at the airport can use that, that to go after you, to literally attack you with a cheap device like the Wi Fi Pineapple. Now, this is why I think ExpressVPN is the best VPN. It hides your IP address, reroutes all your traffic, 100% of your traffic through their secure, encrypted servers. Couldn't be easier to use. You just fire up the app on anything that you've got. I mean, it literally is everywhere. IPhone, iOS, Android, iPad, iPad, Mac, Windows, Linux. It's even on my router. Just fire up the app. You click one button to get protected. It works on all devices, phones, laptops, tablets. So you can stay private no matter where you are on the go. And even at home. It's rated number one by top tech reviewers like Cena and the Verge. Look, protect your online privacy today by visiting expressvpn.com twit that's E x p r e s ssvpn.com twit and you can get an extra four months free when you buy a two year package. Expressvpn.com twit we thank them so much for their many years of support. I was at the airport flying down to Tucson and of course you'd get that free airport WiFi, free SFO, airport Wi Fi. And I thought I'm not going to join that. And then I remembered, oh yeah, fired up ExpressVPN. And I was to. Able, able to use it. That's awesome. Expressvpn.com TWIT Anybody tried the new O3 Pro? Any interest in chat? No, zero, nothing.

[Harry McCracken]

Haven't dug into it yet. But I am interested.

[Leo Laporte]

It's pretty smart. It's. It's different, right? Because it's, it's a now kind of a research tool. Here's an article on substack. God is hungry for context. It turns out that if you're going to use O3 Pro, you can't use it the same way as you would use O3 or other versions of OpenAI's AI. It really wants as much context, as much information as possible. So that's a good thing. But if you don't give it enough context, and this is something people have complained about, it will. This guy says overthink. I think it will hallucinate. It will. If you don't give it enough information, it will make stuff up. So really important to consider that and give it as much information as you can. This reminds me of the Apple paper that came out last week saying that thinking is hard. Yeah, yeah.

[Jason Snell]

If you think too hard, something bad happens. Turns out.

[Leo Laporte]

Yeah, the limitations. Although there have been a lot of responses to it saying, well, it wasn't, it wasn't the best, it wasn't the best test, blah, blah, blah. Basically the researchers who were from Apple gave the variety models, not the latest models. It was Sonnet 3.7, I think. 3.3.0 for 3. Oh, for OpenAI. Some kind of common puzzles. In fact I went out and I bought one just the other day. Let me, let me go get it. Want to see if you could solve this?

[Jason Snell]

Interesting.

[Harry McCracken]

There he goes.

[Jason Snell]

There he's gone. Everybody, we run the show now.

[Jason Heiner]

Party time.

[Jason Snell]

Oh no.

[Leo Laporte]

Dad's back on my, on my coffee table. My whole childhood. Have you, have anybody, anybody played with the Towers of. Yeah, I'm a nerd. I know. Anybody played with the Towers of Hanoi.

[Harry McCracken]

That was the sort of game people used to program in basic 40 years ago.

[Leo Laporte]

Yeah, it's a very simple game that a human can pretty quickly understand. I was a six year old when I figured it out. The idea is you want to move this entire stack of disks to the far peg without putting any bigger disks on top of smaller disks. And it's, you know, it's actually pretty simple. The only key is which peg you start on. And that's the insight that maybe took me a little while to figure out. Depends if there's an odd or even number of disks left. But basically it's repeating this process over and over. Whoops, the peg came out. That's what you get for buying cheap Amazon Towers of Hanoi pegs.

[Jason Heiner]

I kind of want one of these.

[Leo Laporte]

This is great. So. And yeah, everybody who's done programming has probably solved this at one point, but apparently the more disks, the worse and worse these AIs got. And in fact they completely collapsed by the time it got to I think seven or eight disks. I think this one has 1, 2, 3, 4, 5, 5, 6, 7, 8 disks. So I don't know. I think some of this is really just people assuming more, they could do more. Like it's going to be smarter than a human. Right.

[Jason Snell]

Look, there are powerful. Not to sound all conspiratorial, but there are powerful forces in the technology industry that want you to believe that AI just kind of not number goes up, expands, gets bigger, better, and eventually is superhuman and all of those things. But it's entirely possible that we've already discovered a lot of the things that AI does incredibly well, but that it can't be applied to everything and it doesn't always go up. And I think that's okay because I'm one of those people. I'm not a. I'm not a skeptic or an AI boomer either. I'm very much like it's overhyped and also world changing. I think both are true.

[Leo Laporte]

It's both.

[Jason Snell]

But like there is definitely. You feel like the clampdown happens when anybody dares suggest that it might have limitations. Some things might be inappropriate for it. Like people start to freak out in Silicon Valley. But like, I think we need to be open to the fact that there's a. We're in a boom right now and what follows the boom is a bubble that collapses and then you pick it through the rubble and find out the stuff that was actually useful because there's AI is. I mean, I use cursor for the first time time this week. AI is amazing.

[Leo Laporte]

That's the vibe coding editor that lets you code.

[Jason Snell]

Yeah, and it edits my code right in line and shows me the changes and like, it's amazing. But like, again, there are also places where we all know AI kind of flops. And this sort of research suggests yeah, if you push it beyond a certain point you get you. You eventually lead to a complete collapse. And it's like I don't think it's the end of the world to say everything has its limits, but if your business is predicated on convincing investors that it doesn't have limits, then you're going to be upset by this.

[Harry McCracken]

I've been doing a lot of vibe coding lately and overthinking is a real issue. I have some rather minor bugs as far as I can tell in my apps and replat, which I use when trying to solve a minor bug will sometimes just sort of start digging and digging more into the code it generated itself and making bigger and bigger changes and I'll come back 20 minutes later. And it's caused many more problems than it's solved because it's not good at pinpointing this one thing. And it doesn't know when it, when it's spent too much time and invested too much energy in solving a problem in a way that's counterproductive.

[Leo Laporte]

There's also, I think, a group of people who are ready to go, aha, I told you so.

[Jason Snell]

Of course, true.

[Leo Laporte]

And they've been using this paper for that purpose. Let me what happened is basically it's called the illusion of thinking. Understanding the strengths and limitations of reasoning models. These are these new reinforcement learning reasoning models that have emerged in the world ever since Deep Seek. And if you look at the graph, it just collapses. Even the regular AI collapses. Here's a perfect example. Zoom in on this. Claude 3.7. When it gets to 8 disks in the tower of hand, just zero. Can't solve any of them. Thinking Claude, Little bit better, but okay. When it gets to maybe somewhere between 10 and 15 discs it fails. And neither of course can do a 20 disc tower of Hanoi. Whereas a human, even a six year old, once the human understands how it works, which doesn't take that long, can easily do it. But that is not. I don't think that's an indictment of a AI. No, it's just a limitation.

[Jason Heiner]

LLMs can do incredible things like it is an incredible breakthrough and they also can't do everything and they can't do everything that OpenAI and anthropic tell you that they can do because those are the two most valuable, you know, highly valued AI companies in the world. Right. And so the thing that and this is one of the things that I, I think Apple pulling back a little bit on AI and integrating it more directly. Directly is like let's look at the things that LLMs are really good at and let's find ways to integrate those, even if we don't call it AI into the operating system in ways that are, that are measured, that are predictable, that are proven to actually, you know, do some good things for you is a good first step of where we're going. I mean the AI, there are now 55,000 AI startups according to Crunchbase.

[Leo Laporte]

That's mind boggling. Wow.

[Jason Heiner]

A bunch of them having, you know, a lot of those haven't even launched their products yet, maybe never will. Some, some certainly are going to. And then there, according to New York times, there is $29 billion that has been invested in AI and is being invested right over the, the next, this next phase that we're in that is like 50% as I understand it, of like the VC money. So it's hard to get VC money if you're, if you don't have AI in the name. Yeah company right now. And so they're all of that leads to the bubble phenomenon that Jason Snell mentioned earlier. And like we are going to hit the trough of disillusionment any moment now and I think it's good that we're having conversations about what it does well, where it struggles so that we can modulate, calibrate the expectations around what these things can do. Because otherwise that's the difference between this sort of hitting the trough of disillusionment and sort of rising back up and finding thing is use it for versus like this, the whole thing like crashing, you know, the bus sort of all the wheels coming off and it crashing in a fiery inferno into the traffic.

[Leo Laporte]

One thing that does worry me, it's fine if venture capitalists are swept up in the latest fad and it's their money they're throwing it in there. But it does worry me a little bit when government goes all in on AI. And right now that seems to be what our government is doing, choosing to use AI, for instance, to find fraud in Social Security applications. The army has put together a new innovation corps recruiting executives from OpenAI and Meta and other AI companies. Palantir, they're all by the way going to be lieutenant colonels. So congratulations Colonel. I do hope, hope that this is experimental. They're interested but they understand the limitations.

[Jason Heiner]

I think it's driven more by fear. Like that is like we want to know what it can do. Because what we don't want, what they don't want is they don't want another nation state sort of using this and having capabilities that they don't understand or that.

[Leo Laporte]

Very similar to the argument used to create the Manhattan Project. Well, if we don't do it, the natural Nazis will. Right. If we don't do the AI thing, China will. I don't know if that's the best reason to do it. I guess it's sensible. There are some good uses of AI. AI is making healthcare safer in the remote Amazon. I'm not talking about Amazon.com I'm talking about the South American Amazon. This is from rest of world.

[Jason Heiner]

The Internet has Amazon.

[Leo Laporte]

Yeah, the other am the real Amazon. At overburdened clinics, pharmacists are using AI to catch dangerous errors, which seems a sensible use of it. I think AI is very useful if you understand the limitations. Right. I think where we went wrong perhaps is with the consumerization of AI, the chatbots.

[Harry McCracken]

AI is great. I mean it's the humans who can't really trust because a lot of them it's. The humans don't understand it and misuse it and think it's capable of doing things it's not capable of doing.

[Leo Laporte]

Well, Elton John's not too happy. Nor is Dua Lipa. The bill in the UK that Elton and Dua and others, including Paul McCartney opposed the drably titled according to the BBC data use and Access bill. They wanted a amendment to the bill that would have forced tech companies to declare their use of copyright material when training AI tools to say, hey, we're using copyright material without it. These creators argued tech firms would be given free reign to help themselves to content without paying for it. Elton John told the BBC that would be committing theft thievery on a high scale. The government refused the amendment. So Sir Elton is unhappy Be so is du.

[Jason Heiner]

I think that regulators and governments, I don't think they, I, I just worry that they don't understand it well enough.

[Leo Laporte]

Yeah, well, in that case, don't regulate or see. This is the problem. If you don't understand something, it's hard to regulate it. On the other hand, there's not a lot of time to solve this. Yeah, understand it.

[Jason Heiner]

I should disclose like the company that I worked for, Ziff Davis is suing open AI. Right. For saying that what they did was not fair taking and ingesting all of our content and using it to train their models. New York Times is the other entity that's, that's done that as well the courts will decide so but you know that aside like that this is one of the things that. And then Disney also, you know, a big lawsuit against the infringement of their mid journey.

[Leo Laporte]

They saying midjourney stole Darth Vader.

[Jason Heiner]

So you have a lot of this, this issue of is it fair use for the AI companies to. To ingest things that are publicly available on the Internet but they don't have the rights to use. Do they have the rights to use it? Is the question. A lot of that has to get worked out in the courts over the next couple of years.

[Leo Laporte]

Paper published last month by computer scientists and illegal scholars from Stanford, Cornell and West Virginia University studied whether five popular open weight models, three from Meta one each from Microsoft and Eleuther AI were able to reproduce texts from Books three. That's the collection of books that's widely used to train LLMs. Those books still under copyright in many cases. In fact they were able to reproduce Harry Potter and the Sorcerer's Stone with many of these models. But see that's the question. Are you going to use llama3 to read the book? No. You're going to buy the book? I don't, I don't know if this costs them business and I guess isn't that in the long run the real test of copyright? Is this going to harm my business? That's one of the tests for fair use for sure.

[Jason Heiner]

Does it harm?

[Leo Laporte]

Yeah. Does it harm?

[Jason Heiner]

So this is. Sorry to bring it back to Apple one more time. One of the things they didn't talk about was image playgrounds. So last year, this is one of the most shocking things that they did last year that I thought was really, really like across the line for, for them to, to go out and train on all of the data, you know, open web of images and to put that out there. I was really surprised they did it. I thought it was not a great move or just a good look even because you figure a lot of that industry builds there and the people who create a lot of that stuff creates it on max. And that's the, the, a lot of the groups of designers and artists and the people who stuck with Mac even when Mac was struggling, right. They were the only group that stuck with them. And to like to go and you know, and I asked this pointedly last year to multiple people at Apple to make sure I understood it correctly. Like did you train this on the open web? Like did you just go out and scrape everybody's images who put their artwork out there and then use that to train your Models and essentially they, they said that. Yes, they did. Now people could, you know, going forward, say, oh, we don't want you to train on our data, you know, now. But that was the, that was the thing that just really to me struck the wrong, wrong note, the wrongest note about OpenAI. But I did like the fact that they really didn't go back and touch that much. Or maybe I'm, I don't know, Harry or Jason may have gotten more information, but my sense was that they, they've backed away from that a little bit or, or at least hasn't. Haven't leaned into it, you know, as much.

[Jason Snell]

Yeah, they, they said something about honoring robots Txt going forward and that's like.

[Leo Laporte]

That, that's a must, I think.

[Jason Snell]

Yeah, yeah, yeah. So, but I mean, you could argue it's the fruit of a poison tree and that the initial, all of these models are in that, you know, are in that sphere. I would love if they started building models out of licensed content from various companies and all of that and that, you know, I think that they're playing a, of a hedging game here too because. Right. Like Apple is seen as behind an AI, but you know, so you kind of, if you're Apple, you need to build a model. But if, if it turns out there are lots of eagle and legal entanglements with these models, that benefits Apple. Right. Because it maybe throws the brakes on a lot of their competition as well. So I think they're playing both sides right now, but I do think, you know, we'll see how this policy changes. But again, the vibe was very much like last year was almost like a female beaver dream. And they're kind of coming out of it now.

[Harry McCracken]

Adobe is one of the few companies I know of doing a image generating LLM that as far as I know, really is only trading on things they clearly have rights to train on.

[Jason Heiner]

Yeah, Getty too is also doing.

[Leo Laporte]

Getty's also suing. Doing and suing. All right, one more break. No more Apple Reflux. And we got actually all, all I have is a bunch of silly stories to wind this up, so hang in there. This silly, silly season about to begin. Our show today brought to you by Zscaler, the leader in cloud security. You know, hackers, just like everybody else, are using AI. Of course they're using it to breach your organization. Not so good. AI powers innovation, drives efficiency, but it also can help bad guys deliver more relentless, more efficient, effective attacks. Phishing attacks over encrypted channels increased last year by 34.1% fueled by the growing use of generative AI tools. Phishing as a service kits. Used to be you could say, well, if it's ungrammatical, it's, it's bogus. They're never ungrammatical. They got AI right. Organizations in every industry from small to large are, are fighting back. They're leveraging AI to increase employee productivity with public AI for engineers using coding assistance, marketers with writing tools, finance creating spreadsheets and formulas. We're all automating workflows for operational efficiency across individuals and teams, embedding AI into applications and services that are customer and partner facing and ultimately let these companies move faster in the market and gain competitive advantage. We're using AI like crazy at TWIT companies. Companies though though, need to rethink how they protect their private and public use of AI. Chief Information Security Officer from the New York City Department of Education. They're on the front lines. He says, quote, with AI, I'm concerned about the usage of it, but I also love the innovation with it. How are employees using AI? Which AIs are they using? Zscaler can be a good partner there to help us find out the answers to those questions and to help us move faster when it comes to incident response and finding that needle in the haystack. Proactively finding threats to our network and data. See, the truth is what we've relied on for years, traditional firewalls, VPNs, providing then public facing IPs are failing. They expose you and your attack surface to the bad guys and they're no match in the AI era. So we need something better and we need a more modern approach and that zone Skillers comprehensive zero trust architecture and AI that ensures safe public AI productivity. That protects the integrity of private AI and stops AI powered attacks. Yeah, it does all three thrive in the AI era with Zscaler Zero Trust plus AI to stay ahead of the competition and remain resilient even as threats and risks evolve. Learn [e-mail] Security that's Z Zscaler.com Security we thank him so much for supporting this week in tech.

[Jason Heiner]

Zscaler.com Security Impressive number of enterprise sponsors.

[Leo Laporte]

You know it's funny, in the early days of podcasting it was all we call B2C business to consumer, right? It was, you know, food, boxes, mattresses, all that stuff. That stuff. I don't know if those businesses aren't any good anymore or for some reason that that advertising has dried up. We're very lucky that we're in the enterprise space and the technology space because that advertising has not dried up. Those people, you know why companies like yours have stopped publishing PC Week and all of those, you know, enterprise focused magazines, they're all gone. So where are these companies going to go to advertising, advertise? I mean, some of them go to CNN and stuff. I see, it's funny how I see kind of pretty high tech companies advertising on football games. I don't know if that's the best place for them to be. Yeah, I mean it's good for, I guess for brand name, but that's about it. So we're, we're one of the, you know, podcasting in general is one of the few places people can go to to reach IT deciders and we've got a lot of them. That's, that's pretty much our audience. The rest of them are gamers. That's why we're so happy to say that the Nintendo Switch is now the fastest selling gaming console of all time. Four days to three and a half million units. Who says Nintendo is number two or number three? That's pretty impressive.

[Harry McCracken]

They went to Best Buy this morning and they did an enormous sign outside saying Switch to now available. And then the moment you go through the door, there's an equally large sign saying Switch to not in stock. I'm pretty sure which one of the signs is correct.

[Leo Laporte]

Costco really made me mad. I went in there and they have a big Switch 2 display and I thought this is great. They've got some ran over. It was just the controller. They didn't have. Anybody's trying to. I don't think Costco's even selling the Switch 2. I don't think they're one of the retailers, I guess Target, Walmart. Anyway, Nintendo thinks they're going to sell 15 million during its current financial year. Even though I don't think it's that much different than the original Switch by design.

[Jason Snell]

Almost like, I mean, remember the Switch, original Switch was already, you know, quote unquote behind the other consoles. It was low end compared a tech upgrade. And I've heard from people who say when it's docked and it's on your TV screen, it's a major upgrade on that 4K now.

[Leo Laporte]

Yeah.

[Jason Snell]

And like people love them and it's great. I mean I. So I was one of those people who, I didn't want to make an effort to buy one, I just wanted to get one. And so I registered with Nintendo, with my Nintendo account to just say, yeah, sell me One when you're ready. And I got that email a couple days ago. They're like, okay, we'll sell you one and ship it to you. And I'm like, great, send it my way.

[Leo Laporte]

I'm so jealous.

[Jason Snell]

Hopefully in a week I'll get one and I'll have done it in the laziest way possible.

[Leo Laporte]

I've been searching my email every day since June 5th.

[Harry McCracken]

Maybe Nintendo for what will kind of at least sort of keep up with the demand in a way they usually do not.

[Jason Snell]

Yeah, I think this is my big complaint about games. Stuff is, for Pete's sake, Apple can ship lots and lots of iPhones and when you order an iPhone they just say, okay, you'll get it in a week or two weeks or three weeks.

[Leo Laporte]

But it wasn't that way from the beginning, was it?

[Jason Snell]

It wasn't. But over time they've managed to do that. And like, I know there's a lot of theater involved and they want people lining up and stuff like that for these things, but honestly, I would love to go to Amazon or Target or nintendo.com or whatever and just say, hey, here's my money, send me a Switch when you're ready. And they like, they don't want to do it. This is the closest they've come. And I mean kudos to them because the word on the street is you can get one if you want. Which the original Switch, that was not the case. It was real. I think it took like a year for me to get one of those.

[Jason Heiner]

Same for Nintendo Wii. It was like impossible to find forever, you know.

[Leo Laporte]

Oh yeah, Is this the email? Did I get it? Nintendo Switch 2 is available now. Buy now. I don't know, is that the email or is it just.

[Jason Snell]

No, it was. It's time to purchase your Nintendo Switch.

[Leo Laporte]

Yeah, yeah, yeah.

[Jason Snell]

And then the three day window to go in and buy it, which I did.

[Leo Laporte]

Oh yeah. On behalf of my Nintendo Store team, thank you for registering. Wanted to let you know you're still in the queue.

[Jason Snell]

Still in the queue.

[Leo Laporte]

Still in the queue.

[Jason Snell]

A lot of people in the queue probably said no, no, no, I've already got one. But I didn't, I bought one.

[Leo Laporte]

I would buy one if I could, but I can't. So there. Did you have 50 hours worth of gameplay?

[Jason Snell]

Well, I have, I mean given that we have had the Switch for ages and before that we had the Wii and the Wii U and my kids play it and like everybody plays it, like yeah, that we, we were in there, we, we were grandfathered in There was no, not a problem.

[Leo Laporte]

Nice. Nice job. All right, continuing on with the, the silly stuff. Let's see here. What else do I have? An experimental new dating site matches singles based on their browser histories.

[Jason Snell]

Why not?

[Jason Heiner]

What could go wrong?

[Jason Snell]

I mean, what could go right, Jason, is the real question. What could go right?

[Leo Laporte]

Browser dating. You users upload their 5,000 most recent searches which are then turned into a browsing personality profile by AI.

[Jason Snell]

Did Google write this?

[Leo Laporte]

Whose idea was this?

[Jason Snell]

We're in the same ad pool. We're meant for each other.

[Leo Laporte]

Actually, I think it's a. It's one of those artists creating a art project.

[Jason Snell]

Yeah, probably so.

[Leo Laporte]

Yeah. It's a browser dating artist and developer, Dries Deporter, who's known for creating digital products with a. An eye for mischief.

[Jason Snell]

I believe there's. I believe there is a lot of signal in what we do. Other than the crippling privacy problems. I believe you probably could figure out how a browsing history correlates to making a match. It's probably not identical browsing histories, right? It's like complementary browsing browsing history. I'm sure there's signal and lots of stuff.

[Leo Laporte]

You're looking for a comb. He's looking for hair replacements, something like that.

[Jason Snell]

Well, that, that's going to lead you to a real gift of the magi kind of situation, which you don't want. I guess. I guess blind date of the magi. But still, I wouldn't recommend it.

[Leo Laporte]

If you want to know more. Browser dating, the dating website matching people based on the. Should I upload my browser history?

[Jason Snell]

Don't do it. Sure.

[Leo Laporte]

What could go wrong? What could possibly go wrong? I am married, so that's probably the first thing that's going to go wrong.

[Harry McCracken]

It sounds more dangerous than 2023andMe.

[Leo Laporte]

Yeah, it does. It's worse than spit. Just a heads up, anchor is recalling 1.1 million of its power banks. We love anchor power banks. But if you have the power core 10,000 power bank with a model number A1263.

[Jason Heiner]

As a recall, they also have recalled in the past like six months. They've also recalled a couple of, couple Bluetooth speakers because of these similar battery problems and another power bank as well. So you know, if you have any acre power banks like think about, just make sure to check them.

[Leo Laporte]

I have nothing but anchor power banks everywhere in the house.

[Jason Heiner]

I think, I think mine are only anchor.

[Leo Laporte]

Here's one right here.

[Jason Heiner]

I'm not a hater at all.

[Leo Laporte]

This, this is not a, you know, false Alarm. The Consumer Product Safety Commission says anchor has documented 19 cases of fires and explosions include including two minor burn injuries in 11 instances of property damage totaling more than $60,000.

[Jason Heiner]

So you've got one of those LLM hallucinating.

[Leo Laporte]

This is not a hallucination.

[Jason Heiner]

This is actually.

[Leo Laporte]

It's the real deal.

[Jason Heiner]

Yeah, things have caught fire.

[Leo Laporte]

This isn't exactly a tech story, but I, I find it fascinating. New York Times today, a town's single largest talk taxpayer is also his biggest headache. Small towns are having a problem with deserted malls. Nobody wants to be in the deserted mall, but they can't redevelop them. Timothy Sorella, town selectman and former rural police chief in Lanesboro, Massachusetts, population of 3,000, says there were times you could not find a parking place in this mall. Inside it was packed for teenagers in particular was a place to hang out. I feel for teenagers, they don't have a mall anymore to go to.

[Jason Heiner]

Well, I think the challenge is that the malls, they're like, you remember when there used to be like three or four paper newspapers in a city and they went down to one paper. Malls are like that. Like. It's not that there's no malls, it's just that there used to be three or four malls. Now there's one mall.

[Harry McCracken]

Some malls are still thriving, but then there's another mall near us that's turning into a biochemist tech campus, which.

[Jason Heiner]

Yeah, right.

[Leo Laporte]

Didn't Rackspace turn a mall into its network operations center many years ago?

[Jason Snell]

Well, I mean, I don't know the last time you were down in Cupertino, but having just been down there, tell.

[Leo Laporte]

Me the pruneyard is still there.

[Jason Snell]

Just south of Apple park is what used to be Valco Fashion. And very slowly they are taking all of the property around the mall and turning it into high density commerce residential. It's offices and restaurants and all of this and residents residential. The mall itself, the core mall, I think is still there because there's like an ice center and a Benihana. But eventually it'll, it'll all be gone, baby.

[Leo Laporte]

You got to have an anchor tenant. If it's Benihana, so be it.

[Jason Snell]

Yeah, and ice skating and Benihana, but not together. Don't do them together. So. But they're doing it even in a place now, a place as wealthy as Cupertino. It's very easy to build that stuff and repurpose that land. I think the question with the story like the town of Massachusetts, Massachusetts, is what do you do? And it's a little like the inner city cores that they're trying to turn office buildings into residents, like in Manhattan. The problem is a lot of those buildings are just not suited for people to live in. And so what do you do? And the mall? Like, I mean, I love a mall. I love a caramel corn and an Orange Julius, and I'm a kid from the 80s. I love a chess king, whatever, but, like, I wouldn't want to live there. So what do you do? I don't know.

[Leo Laporte]

Sad story. Benny and Gizmo have left the nest. This is the. Where are they? Where is. This is the eagle cam in Big Bear Valley. No more Benny and Gizmo, the two eagles that were huge. This was huge. There's still 11,000 people watching this, even though there's no birds.

[Jason Heiner]

They should use the roofs of malls for bird habitats.

[Leo Laporte]

Wow. So, of course, this has spawned numerous Think pieces. Still a little feather hanging in the wind, but that's about it. Think pieces here at Fast Company. You might have heard of them. What's next now that Sunny and Gizmo are leaving the nest? This was huge, this family of bald eagles. We covered it a couple of weeks ago. They were still there. But Sunny and Gizmo have moved on, as eagles will, as they do. They've Left the nest. June 7. Gizmo's turned to Fledge. She practically fell out of the nest. Anyway, she landed nearby, and they're gone. They came back Monday for dinner, but their days are numbered. So I guess that's all I have to say about that.

[Jason Snell]

They spotted some peregrine falcons on the campanile at UC Berkeley. And this is a story because.

[Leo Laporte]

There you go. Put a camera up.

[Jason Snell]

Because they've had cameras for a long time. But the big story is that the bird flu has really killed a lot of the birds. And so they're trying to figure out, like, how it impacts the larger population. They spotted a new falcon that they'd never seen before on their cameras because the falcons who had roosted there for, like, eight, nine years vanished. And they figure that means that they got the bird flu and they died. So it's a. It's a huge. Yeah, it's. It's like I. I want to shout out all these people who love birds and have put these cameras in these places, because in some ways, this is how. How biologists and scientists find out the impacts of things that are kind of hard to see, like bird flu, where, you know, you don't. You don't know what the birds are doing and how Many there are because they're flying around and stuff. But some of these people are really making an effort to track them in places where they roost and places.

[Leo Laporte]

Here is the unknown female visiting the Campanile.

[Jason Snell]

Yeah. And they're like, come on, stay. It's a nice place to have chicks. They've had. They've had many, many chicks there over the last 10 years. There's quite a Cal Falcons. There's a real great setup. They've got. They've got multiple webcams on the top. This is the Campanile, which is the. It's a. It's an obelisk. It's the tallest building in Berkeley. And so the peregrines, which are the, what fastest birds, they. They love it up there. So hopefully somebody else.

[Leo Laporte]

They can see everything going on down there.

[Jason Snell]

Yeah. Because that's what they do is they. They actually attack other birds that fly lower. So they go up high and then they see the other birds and they swoop down and kill them.

[Leo Laporte]

I thought it was mice. I thought it was rats. They're going after birds.

[Jason Snell]

They also get. Oh, they kill a lot of birds. They do. They do. So, you know, top two sources of bird death. One bird flu, two other birds.

[Leo Laporte]

Wow. And of course, you are a proud alumna.

[Jason Snell]

Yeah. Alumni. I'm a Berkeley guy, so I. That's why I follow California falcons.

[Leo Laporte]

Yeah. We have red tail hawks here, but our crows are black. Lisa tells me have chased the red tail hawks off.

[Jason Snell]

I believe it.

[Leo Laporte]

They dive bomb them. Those crows are mean.

[Harry McCracken]

We have some imposing crows in this neighborhood posing.

[Leo Laporte]

They're huge.

[Harry McCracken]

And I was just in Vancouver near the convention center where they have signs saying, for God's sake, be careful about these crows and try to walk under the overhang. And we'll even lend you an umbrella to fight them off if you need one.

[Leo Laporte]

They're super smart. There's a crow. We have a street out here that I walk down and apparently I didn't know this. This crow's following me. Like it sits on the front. This fence looks at me, goes. And I walk a half a block and then it flaps around to ahead of me on the fence and goes. And I thought, what the hell is this out of the Alfred Hitchcock movie? Turns out Lisa's been feeding this crow as she walks that way. And somehow the crow has identified me as a related to Lisa is now following me.

[Jason Snell]

You have.

[Jason Heiner]

You have a similar smell now that.

[Leo Laporte]

I must smell like her. That's probably what it is. Well, on that note, thank you Gentlemen, for making this a quite enjoyable afternoon, despite the apple reflux. Jason Snell will be back on Tuesday to talk about everything, and we're going get to get into the deep nitty gritty of it all on Mac Break Weekly. I look forward to that. And of course, you already covered it on your own podcast. You can find out about all of Jason's [e-mail] podcasts and his written coverage. You and Dan the Moose, you've all done a great job of covering this. Great place to go. 6colors.com thank you so much, Jason. It's nice to see you on a Sunday.

[Jason Snell]

Thank you, Leo. Nice to be here.

[Leo Laporte]

Did you. Did your team win?

[Jason Snell]

Did my team. We did. We dominated in curling today, actually. We really did. We blew it out. I made most of my shots. I do curling in Oakland every Sunday. And yes, that Olympic sport. And it's harder than it looks, but I nailed it. I'm going to go out this season on a high note.

[Leo Laporte]

I think the number one sport in.

[Jason Snell]

Canada, Winter Olympics coming next year.

[Leo Laporte]

Yeah. Do you. Do you throw the stones or you sweep or do you both.

[Jason Snell]

Curling is a team sport, so four people play and they all rotate through. So you shoot a couple of stones and then you sweep a bunch of stones. And that's how it works. Yeah.

[Leo Laporte]

How fun.

[Jason Snell]

It's great. People should check it out.

[Leo Laporte]

Yeah.

[Jason Heiner]

It's the best thing to watch on the Winter Olympics, you know, really on silent. Like you can watch it silent.

[Leo Laporte]

Yeah.

[Jason Heiner]

And you're sort of like, it's like. It's just oddly mesmerizing.

[Leo Laporte]

But every once in a while, somebody makes this impossible shot and you go.

[Jason Snell]

I mean, I thought. I thought they were impossible before, and then I did curling for a couple of years, and now they're just. I. I don't even know what I'm seeing. It's so. They're so good at that. But I will say, yeah, my club and clubs, I imagine all over the country are planning for next winter because this is when people are curious about curling and they come out and learn. And that's what we did three years ago.

[Leo Laporte]

So when is the where in it? When? And where are the Winter Olympics?

[Jason Snell]

It's February in. And I want to say it's in Italy. Italy, France, in the Alps. I think it's in the Alps, but, yes. So. So, yeah, try. Try curling. But curling tends to shut down in the summertime because most of the places that do curling, those people want to be outside in the summertime.

[Leo Laporte]

Yeah.

[Jason Snell]

Here in the Bay Area. We could do summer.

[Leo Laporte]

Jason, always a pleasure. Thank you so much. Harry McCracken, the technologizer. You'll find his [e-mail] and he's on Blue sky at Harry McCracken. Anything you want to particularly plug? I guess I want to read this article you're writing, an introduction to.

[Harry McCracken]

Oh yeah, sure. That Quantum story. My newsletter plugged in comes out every Friday. You can find that easily at fast company. And sign up.

[Leo Laporte]

I'm a subscriber. You bet. You bet. Thank you, Harry. Great to see you guys.

[Harry McCracken]

Thank you. Thank you, Leo. I will. Thank you.

[Leo Laporte]

Jason's and the other Jason. Jason Heiner, editor in Chief. No dashes at ZDNet. Great to see you, Jason. And your Twit Fez. He's a proud holder of a Twit Fez. Always a pleasure.

[Jason Heiner]

Yes. And the Twitfaz. And someday in the back, maybe a Vision Pro. I don't know if it's a Newton or if it's a 1984 Macintosh.

[Leo Laporte]

Either way, kind of my attitude.

[Jason Snell]

Historical either way.

[Leo Laporte]

Although I admit I bought that cube, got it home and that day and Apple announced they were discontinuing it. So they never discontinued it.

[Jason Snell]

They just put it on ice.

[Leo Laporte]

I. Yeah, I like it. I have a soft spot for that. That was a great computer. Thank you. Jason Heiner. You're all great. Really appreciate it. Appreciate your time. Thanks to all of you who watch. As you know, we do Twitter every Sunday afternoon, 2 to 5pm Pacific. You can watch us. That's 5 to 8 Eastern in 2100 UTC. You can watch us live if you're in the club in the Discord. But there's also open to the public YouTube live stream, Twitch, TikTok, X.com, facebook, LinkedIn and Kik everywhere. But you don't have to watch live. There's on demand versions of the show you could download from the website Twit TV. There's a YouTube channel, a great place to go if you want to share just a clip with somebody and introduce them to the addiction that is called twit. And of course you could subscribe in your favorite podcast client. Do me a favor though. If you do subscribe, leave us a review, a good one. I hope as many stars as you can muster. That helps us spread the word. And you know, after 2020 years, it's easy for people to forget you even exist. I still every week hear from somebody say you're still around. You're not dead. No, we're alive and kicking. So tell the world. We thank you for being here. Join the club if you're not a member. And thanks to all of our club members for the great support. We will see you next time. And I guess as I've said for the last 20 years, and I'll say it again, another twit is in the can.

[Jason Heiner]

Amazing. Hi, Zoe Saldana. Welcome to T Mobile. Here's your new iPhone 16 Pro on us.

[Harry McCracken]

Thanks.

[Leo Laporte]

And here's my old phone to trade in.

[Jason Heiner]

You don't need a trade in. When you switch to T Mobile, we'll give you a new iPhone 16 Pro. Plus we'll help you pay off your old phone. Up to 800 bucks and you still get to keep it. There's always a trade in. Not right now. @ T Mobile. I feel like I have to give you something in return for karma. That's okay.

[Jason Snell]

I don't really have much in my purse.

[Harry McCracken]

Oh, let's see. Hand sanitizer. It's lavender.

[Jason Heiner]

I'm good.

[Leo Laporte]

Seriously.

[Jason Snell]

Let me check this pocket.

[Jason Heiner]

Oh, mints. Really, I'm fine. Oh, I have raisins.

[Harry McCracken]

I'm a mom.

[Jason Heiner]

Wait, wait one sec.

[Harry McCracken]

I've got cupcakes in the car.

[Jason Snell]

It's our best iPhone offer ever.

[Leo Laporte]

Switch to T Mobile, get a new iPhone. IPhone 16 Pro with Apple intelligence on.

[Jason Snell]

Us, no trade in needed. We'll even pay off your phone up.

[Jason Heiner]

To 800 bucks with 24 monthly bill credits.

[Harry McCracken]

New line, 100 plus a month on.

[Jason Heiner]

Experience beyond Finance Agreement 999.99 and qualifying.

[Leo Laporte]

Ported for well qualified plus tax and.

[Jason Heiner]

Ten dollar connection charge payout via virtual prepaid card.

[Leo Laporte]

Allow 15 days credits end and balance due if you pay off earlier. Cancel CT mobile dot com.