MacBreak Weekly 1034 Transcript

Please be advised that this transcript is AI-generated and may not be word-for-word. Time codes refer to the approximate times in the ad-free version of the show.

Leo Laporte [00:00:00]:
It's time for MacBreak Weekly. Andy, Christina, Jason are here. We're glad you're here too. Coming up, we're gonna talk about Apple's price hikes. It's not just hardware anymore. Is it greed? The OpenAI battle continues on with lawyer letters, and China gets Siri AI, or does it? That's coming up next on MacBreak Weekly.

Leo Laporte [00:00:34]:
This is MacBreak Weekly, episode 1034, recorded Tuesday, July 21st, 2026. The Other Side of the Corn. It's time for MacBreak Weekly. Hello, you Mac maniacs. This is the show where we cover the latest Apple news. Jason Snell from sixcolors.com. Hello, Jason Snell. How are you?

Jason Snell [00:00:58]:
Good to be here.

Leo Laporte [00:00:58]:
Wonderful. Did you enjoy the World Cup halftime show of 27 minutes?

Jason Snell [00:01:04]:
Oh, did I? I mean, Apple TV's own Ted Lasso was there, so.

Leo Laporte [00:01:08]:
Oh my God. Well, we'll talk about that. Christina Warren's also here. She is DevRel at GitHub. And we were talking before the show about the local AI and doing all this stuff and then having so much fun. Also, Mr. Andrew Ihnatko, he's in the library. Good to see you, Andy.

Andy Ihnatko [00:01:28]:
Hey there, Heather.

Leo Laporte [00:01:29]:
Howdy. How you doing, everybody? Yeah, Ted Lasso and his pal, I don't, what's his name? The guy with the beard?

Jason Snell [00:01:36]:
Coach Beard.

Leo Laporte [00:01:36]:
Coach Beard.

Jason Snell [00:01:37]:
Brendan Hunt, yep.

Leo Laporte [00:01:38]:
Easy to remember.

Jason Snell [00:01:39]:
Coach Beard.

Leo Laporte [00:01:41]:
Were just kind of walking.

Jason Snell [00:01:43]:
They did a walk and talk.

Leo Laporte [00:01:44]:
They did a walk and talk.

Jason Snell [00:01:45]:
And then introed Justin Bieber, yeah.

Leo Laporte [00:01:47]:
Which was wild. And then they turn around the substitution sign and it says, Belieber, Beliebe.

Jason Snell [00:01:53]:
Beliebe.

Christina Warren [00:01:54]:
Yep.

Andy Ihnatko [00:01:55]:
And the world soccer press thought that was all wonderful.

Leo Laporte [00:01:59]:
They loved it. Oh my God, they hate it.

Andy Ihnatko [00:02:01]:
every 4 years we should do something exactly like that.

Jason Snell [00:02:05]:
They were offended that we did a Super Bowl halftime show in America for the World Cup.

Leo Laporte [00:02:09]:
But it wasn't even as good as the Super Bowl. I mean, one of those artists and like Shakira would have been great for 27 minutes.

Christina Warren [00:02:18]:
Or Madonna, right?

Leo Laporte [00:02:19]:
Like, I mean, I was blown away when she comes. I didn't like— I hate it when they do the pre-tape. The Weeknd did that too at his Super Bowl where they record a whole bunch of stuff and then emerge onto the field. I don't like that so much, but it was great to see all of those.

Andy Ihnatko [00:02:34]:
There's nothing wrong with this, but it was kind of like, remember when Letterman hosted the Oscars and he decided to do a stupid pet trick segment? Yeah, it's like, this, this isn't that. You don't have to take that from this and put it here. It's like, respect what you're doing.

Leo Laporte [00:02:47]:
Yeah, some things belong, which is true.

Christina Warren [00:02:49]:
But, but at the same time, if you're gonna have the, the World Cup in America, I, I don't know, I feel like we should be gauche and American, and that means we're gonna put on a Super Bowl halftime show at the World Cup. Like, that's what that means. Like, I don't disagree, it's maybe not time or place, but also what's more American than having a Canadian? Um, Shakira is from South America, I think, and, uh, and Madonna's American at least, right? So at least we got a couple, you know.

Andy Ihnatko [00:03:15]:
Not according to her accent.

Christina Warren [00:03:18]:
No, it was very fun.

Andy Ihnatko [00:03:19]:
Yeah, it was fun.

Leo Laporte [00:03:20]:
The game was amazing if you're a fan of, uh, the Spanish form of soccer, which was Quite beautiful. Their passes are amazingly accurate, but didn't really— Argentina didn't enjoy it too much. I hope I'm not spoiling anything. Anyway, on we go with the show. We like to begin the show with something totally unrelated. Wait a minute, there was Ted Lasso. That counts.

Jason Snell [00:03:45]:
Exactly, see, I was tying it into Apple's fair, which is Ted Lasso and Coach Beard were there. And that's a great— you can't buy that kind of promotion for your— Ted Lasso returning in August to Apple TV, by the way, so.

Christina Warren [00:03:57]:
Yeah. Also Apple TV where you can watch Major League Soccer or whatever, right?

Jason Snell [00:04:02]:
Indeed. Indeed. I think that there you could see all those connections happening there.

Leo Laporte [00:04:06]:
It's kind of amazing that Tim Cook didn't show up. I mean, usually he shows up at those things just to, just to be there.

Jason Snell [00:04:14]:
Maybe he was there. They're too busy taking pictures of Matt Damon and misidentifying as Brad Pitt.

Christina Warren [00:04:20]:
Yeah.

Jason Snell [00:04:20]:
Yeah. Sorry.

Leo Laporte [00:04:21]:
That was pretty funny, actually.

Christina Warren [00:04:22]:
That was very funny.

Leo Laporte [00:04:23]:
It was Brad Pitt. I don't think so.

Jason Snell [00:04:25]:
The one— well, he— that— yeah, clearly that announcer got handed a card that said Brad Pitt and then he was like, That's not Brad Pitt, actually. That's Matt Damon. I saw a very funny social media post that showed him with the woman, the lovely woman he was with, and the post just said, that's not Penelope. Nice.

Leo Laporte [00:04:40]:
I like it.

Jason Snell [00:04:40]:
I wonder what that's about.

Leo Laporte [00:04:43]:
So, we like to start with cheerful news because the bad news is Apple is raising prices, not just on hardware, but on software as well. Services, in fact. Apple Music and Apple One subscriptions will go up. Apple Music's plan up a buck to $11.99 a month. Apple One bundles up. Now, I'm an Apple One subscriber. I have the family plan, which has gone up $3 to $20. Individual plans, $10.99 to $11.99.

Leo Laporte [00:05:13]:
Student, only $5.99 to $6.99. Oh, I'm sorry, that was Apple Music. Apple Family— I was dreaming— is $27.95, up a couple of bucks. Individuals, not changing. Maybe I have Premier. I don't even know what I have.

Christina Warren [00:05:29]:
Yeah, Premier is family, but you also get more space. So it's like you get the Cloud Plus, you get the Cloud Plus, you get, you get 2 terabytes, and then you get the family subscriptions for all the other services. So, um, it— like, I don't use Apple Fitness, and I don't think it's a good product. Um, and I don't like Apple News. But do I have both of those things? Yes, I do. Because, because when I priced it out, and now I think it would still fit, be like, if I got Apple Music subscription, like, like Family, which I, you know, get to share with my family. If I got like the, the cloud to be able to share my iCloud storage with my parents, those 2 things alone, and, you know, I guess like Apple Arcade or whatever, like it would wind up being cheaper to just get all of them. So.

Leo Laporte [00:06:12]:
Yeah. So that's the question with the price hike. Is it still a good deal?

Christina Warren [00:06:16]:
I mean, good deal. I think that's a different question. Uh, is it, is it still better to get the bundle than to buy the things individually? Yes. Is it a good deal? I think that is hard. Everybody's got to make that question for themselves because what gets me about this and like price hikes, they are what they are. But when I look at like $39.95 a month, which is fine, that, that right, we're talking basically $500 a year, which fine, whatever. But when they introduced Apple Premiere, Apple One Premiere or whatever, I'm pretty sure that it was like low $30 a month, or maybe it was $31.95 a month. I don't remember what it was, but it's certainly, it's gone up at least 2 times since, since then.

Christina Warren [00:06:59]:
I think this is the 3rd price hike. So at this point, you're going from like, okay, well, I used to budget $360 a year. Now I'm having to budget $480 a year. That's, that sucks.

Leo Laporte [00:07:11]:
9to5 did the math. If you bought those services separately, it'd be $73 instead of $40.

Christina Warren [00:07:17]:
Sure, sure. But—

Leo Laporte [00:07:18]:
But you may not I don't want Fitness+ and News+.

Christina Warren [00:07:21]:
Well, I was just going to say, the only reason— the only reason I get those is because they're bundled with it. Right. I really only want the storage.

Leo Laporte [00:07:27]:
I would want TV+, and music.

Jason Snell [00:07:31]:
Right. Right. Premiere is— by the way, Premiere started at, I think, $29.95 in 2020 when they introduced it. So 6 years later, it's basically $10 more.

Leo Laporte [00:07:41]:
Okay.

Christina Warren [00:07:41]:
So it's gone up 50% in price, which is not insignificant.

Leo Laporte [00:07:45]:
Yeah. Uh, do you think Apple's spending more? Is this a price hike created by additional costs?

Andy Ihnatko [00:07:53]:
Yeah, I mean, licensed content is going up. It's like everyone's, everyone's raising prices at least a dollar or 2. But, and it also—

Leo Laporte [00:08:00]:
Coach Beard don't come cheap, my friends.

Andy Ihnatko [00:08:02]:
That's true. Jason Sudeikis could basically ask— actually, not actually, maybe not so much anymore because he's no longer like the one huge, huge, huge hearts and minds show on the, on the network anymore. But you're, but you're right. I was like, Uh, I was— it's easy to jump on them for this, but Spotify has gone up, uh, YouTube Premium has gone up, pretty much everything. Uh, but that said, I'm— you, you do have to worry if it's like, we think that we're undercharging for this. We think that if we, over the next 2 or 3 years, issue 2 or 3 $1 price hikes, we can get this all the way up to $16 or $17 a month. And then we will wait to see. We will, we will terminate this plan as soon as we see that it is costing us in renewals and new signups.

Leo Laporte [00:08:43]:
It, so some have said it makes sense for the hardware to go up. Apple's paying more, but it's pure greed for services to go up because Apple's probably not paying more.

Andy Ihnatko [00:08:56]:
Yeah. Especially since like, uh, also, uh, uh, AppleCare Plus has also gotten a price increase and it's like, okay, so basically you're saying that because it's going to cost the hardware that we might have to replace your motherboard with is going to cost us more because of the storage and the RAM, we need to increase the cost of AppleCare+. Yeah, I mean, there's an— I wonder if there's an argument inside of Cupertino that, look, let's just rip off all of these Band-Aids at once. Like, I know, I know it's not gonna be one big Band-Aid, but if we can just basically put all these through in the same month, it won't seem as though we're constantly chiseling our users, no matter what the explanation is. But if not, that's really, really bad timing if Apple gets associated with, well, we're just basically— I think this— I think that's the cost, come to think of it, of being now nearly a $5 trillion company. That is, even if they have very legitimate reasons to, to raise prices like this, you're always going to have people like me, jerks, saying, you're a $5 trillion company, do you really need an extra dollar a month for Apple Music from me?

Leo Laporte [00:10:02]:
Really?

Jason Snell [00:10:03]:
So, I mean, I hear, I hear that, but also all prices are up.

Andy Ihnatko [00:10:07]:
Yeah, of course.

Jason Snell [00:10:07]:
And all expenses are up. All their servers are up. The people— they're going to have to— the people who work for them, who do Apple Fitness, have to buy groceries. And those prices are up and their cost of living is going up so that they're probably asking for raises. And like, they're not getting them, though.

Christina Warren [00:10:23]:
Come on. They're not getting any raises.

Andy Ihnatko [00:10:25]:
Maybe they are.

Christina Warren [00:10:26]:
You know, they're not. Like, like, like, I fully agree with you. No, you're right. You're right.

Jason Snell [00:10:31]:
They don't— they don't give raises to their employees at Apple. Yes, you got Come on. I mean, this is, this is the world that we're in. So you can single them out if you want to. I think it's not realistic to do that because this is where we are, is everything is going up in price. It's not like Apple is sitting there twisting the dial and everybody else is like, no, no, no, we're holding it down. In fact, you could argue that Apple held it down for a long time, but they're not going to do it anymore.

Andy Ihnatko [00:10:55]:
Yeah, all I'm saying is that, again, this is part of the cost of being a $5 trillion company that you're not willing to— like with a lesser company, you might say, well, okay, I guess it sucks, but okay, whatever. I think that that's the sort of like pushback that people will give when a company gets to this size, 'cause they didn't get to that size by saying, oh, well, gosh, we don't wanna charge our consumers any more than we absolutely need to. It's like, yeah, you know what? They're not 2 hippies in a garage. They are gonna do what they have to do. They're gonna make a lot.

Jason Snell [00:11:24]:
Yeah.

Andy Ihnatko [00:11:25]:
It's a worthwhile question to ask, even though, as you make some good points, that everything is going up. But again, it's just, again, another Band-Aid is after the past 2 or 3 Band-Aids get ripped off. I'm losing hair on my legs right now because of all the Band-Aids that are being ripped off.

Jason Snell [00:11:39]:
I get it, but it's happening everywhere. And I guess the other thing I would say is, do we really expect that a giant profit-making company is going to be the one that says, no, not us, everyone else but not— and the fact is, they kind of did that with their hardware for quite a while. And then they finally were like, all right, we can't do this anymore.

Christina Warren [00:11:55]:
Well, the only thing I'll say is, and I agree with you largely, Jason, although tech companies are not actually giving raises right now. And so everyone's incomes are staying stagnant and prices are going up. So like, we shouldn't pretend like this is in any way going to be impacting, like, especially their hourly employees. But like, putting that aside, yeah, I don't expect them to keep their prices constant when all their competitors are raising them. Having said that, I think that as a consumer, when this happens, this gives you the opportunity to evaluate what value do I get out of this service?

Leo Laporte [00:12:28]:
Hmm.

Christina Warren [00:12:28]:
And for some people that means, yes, the $40 a month or whatever I'm paying for AppleCare or these things are worth it. Now, for other people, that might be looking at how much do I really use Apple Music? How much do I use Apple TV? How much do I use Arcade? Can I get storage from someone else that maybe I like better? Because my argument with, with Apple, you know, one, to be honest with you, has been that in my opinion, even if you are fully in the Apple ecosystem and I am, I don't think that any of those services are best in class at all. There are better videos. There are better video services. There are better cloud services. There are better music services. There are better news products. There are better things you can get individually.

Christina Warren [00:13:09]:
The value is that the tight integration, the vertical integration between your Apple products, which is good, and just kind of the ease of having one thing to do it all. That said, if your prices are continuing to go up and in this case, like when my— again, like my Apple Premiere One plan has gone up 50% in 6 years, which is a lot. Now, a lot of other services have gone up that much too. But I think when any of that happens, that gives the consumer time to evaluate not just Apple, but anybody and say, is there another service or is there something else that I could be buying that I like better? Is this still a good value for me? And that, I think, is the question, not are they right to do it or not? Because obviously, yes, everyone is raising prices and that doesn't make Apple unique. And I don't think that they should have to suffer just because they're such a $5 trillion company. But I think as consumers, we do still have the right to say, Is this still the best use of my money, especially when the economy and everything else is hard and when things are tight?

Andy Ihnatko [00:14:04]:
Yeah.

Jason Snell [00:14:05]:
And that goes for every— that goes for every service from every company. And just as a correction, you and I are on the same Apple One plan. It went up by 33%, not 50%.

Christina Warren [00:14:15]:
Okay. Okay.

Jason Snell [00:14:16]:
Well, from— it's 10 more above the original 30.

Christina Warren [00:14:19]:
Okay. Well, my correction, excuse me, it went up a third regardless. Yeah, a third in 6 years is a lot.

Jason Snell [00:14:25]:
And my Netflix is up and my Spotify is up and everything else is up too.

Christina Warren [00:14:28]:
And that's why I said this is the opportunity not just for to evaluate Apple, but any company. And I think that's what this is. When this happens, you should evaluate all of your services from all of your providers and say, is this still valuable to me?

Andy Ihnatko [00:14:40]:
Yeah, but I think, I think that, Christina, that was, that was a good question that we've been asking ourselves for the past 2 or 3 years. Like, so I bought— so as you know, I bought a new MacBook, which means that now I get all of these services for free for 3 months, so long as I remember to cancel before they start charging me for them. And so this is my first time back on Apple TV. This is my first time back on Apple Music. This is my first time back on a lot of these services. And I don't think they're terrible, but they're not standouts to me. Like every day now that I get to, after a long pause of access to this stuff, it's like, wow, why did I ever drop this? It was like, well, you dropped it because you were on Spotify before Apple Music and Apple Music just wasn't a compelling alternative despite the fact that you've got so many, despite the fact that the Apple Music app on Android is first class on the platform. There's just nothing about, and I can kind of go that way all the way down the list.

Andy Ihnatko [00:15:37]:
And Jason is also right. I mean, this is also the reason why last year, the year before, I did have to go through all of my subscriptions. And it wasn't that I don't have, don't necessarily have like $13 for Hulu as opposed to, uh, $10 when I, when I might have first signed up. It's that, that's that extra dollar at that last price increase makes you look at this and say, why do you— why do you subscribe to Hulu? And for me, the answer was always, uh, Bob's Burgers in season, because I want to be able to watch Bob's Burgers and No Murders in the Building. And the thing is, like, 8 months out of the year there— that means that there's no— 4 or 5 months out of the year there's no reason to subscribe to Hulu, which is why— Again, it's on that list of I will can't, I will, I will subscribe for a couple of months. Like as soon as they offer me that, hey, $5 a month for 3 months to come back, I will sign up for those 3 months. I will get caught up on all the series I want, but then I'll absolutely cancel because compared to YouTube Premium, compared to for me, HBO Plus, like this is not a must-have. So this is, I'm sure that's part of the decision-making inside of Apple saying if we do this price increase of even a dollar, that might provoke people to who are fine, with it at $10, who are okay with it at $11, who are starting to feel the heat at $12.

Andy Ihnatko [00:16:51]:
Once we add that $1 more, is that going to make people think it's stupid for us to have YouTube, YouTube Premium, which has YouTube Music and this other music app, which I don't necessarily love? So I'm sure that was part of the discussion.

Jason Snell [00:17:04]:
Well, in the end, it's math, right? It's how much more revenue do we get by adding a dollar minus how much revenue do we lose by the people we turn off? And if the net is a positive, I guess you do it.

Andy Ihnatko [00:17:19]:
Yeah, it does. It does. I wish that they were— for a lot of these services, they did feel a desire to really make a showing of it and really make it best in class. I mean, I would— I love iCloud storage because of the way it just— it's simply part of macOS that my OmniOutliner documents, they're just simply, by virtue of saving it in one place, they're just simply available everywhere. I just wish that it were as aggressively good as Google Drive is, for instance, which they keep making this a really valuable piece of real estate to put my files in. Like every 3 or 4 or 5 months, there's something new that says, wow, I'm so glad that I'm using this for so many of my projects because it just makes dealing with files on, in the world, real world so much easier. And I'm not getting that from iCloud or in a lot of different Apple services. They're all really, really, it kind of brings us back to that halftime show.

Andy Ihnatko [00:18:13]:
They're all really good. There's nothing wrong with them.

Jason Snell [00:18:14]:
Yeah.

Andy Ihnatko [00:18:16]:
But you kind of wouldn't be— you wouldn't be put out if it didn't exist, almost.

Leo Laporte [00:18:22]:
That's pretty funny. So what is Fitness Plus? Is that Justin Bieber or is that Shakira?

Andy Ihnatko [00:18:31]:
Hips don't lie, Leo.

Leo Laporte [00:18:34]:
That's true, they don't. Actually, I was really pleased to see Shakira. People are making that decision right now.

Andy Ihnatko [00:18:40]:
I forgot how good she is.

Leo Laporte [00:18:41]:
She's fantastic. And she looks great. I thought she must be older by now, but no.

Andy Ihnatko [00:18:46]:
The energy you have— I won't make this into a big distraction, but you have, you have so much appreciation for these performers where you might have forgotten about them for 5, 6, 7, 8, 9 years, and then you see them on stage like 10 years after they were a big hit. Like, my God, they're still bringing it. Like, they're still on a live performance. They're still acting like, I'm 23 years old and my whole, my whole career, and this is the best time of my life. I'm not gonna waste a second of this.

Leo Laporte [00:19:10]:
Anyway, um, what was I gonna say? Oh yeah, everybody's making these decisions right now with Netflix. With everything is, uh, and I imagine all these companies are kind of a little bit on edge saying, well, did I push it too far? And I guess Apple, among all the companies that you could think of, is probably the least likely to suffer from this kind of thing. I just feel like, I don't know, we, we ask a lot of Apple, don't we? Maybe we ask too much.

Andy Ihnatko [00:19:36]:
Yeah, I mean, I guess, I guess to sum, to sum up more, more coherently what I was saying earlier, it's kind of like when you go out to dinner with somebody that you know cashed out of Google like at 45, and they've got $100 million. And it's like, again, it might be unfair that you're thinking, why are you fighting me on the fact that like I had a dessert and you didn't? Like, why are we— not even like they should take the checks. Like, why, why is it that you don't want to split this check 6 ways? It's like, you of all people, even though that's unfair, but it does like this. It does attract that kind of Yes, the pushback.

Leo Laporte [00:20:11]:
Yes. How many, how many trillions of dollars are they worth right now? Their market cap is, is soaring. Their stock price has gone up a lot.

Andy Ihnatko [00:20:20]:
Big news on Friday because them and Nvidia were trading, trading slots for the number one, the most valuable company in the world. And I think Nvidia took the crown at the end of the day. But a lot of the financial news was like excited about like, maybe we're going to get to $4.9 trillion. This is not a sign of a problem with the economy. This is wonderful.

Leo Laporte [00:20:40]:
Yeah, market cap $4.8 trillion freaking dollars. That's, uh, that's, that's almost Elon Musk money. I think they're gonna be fine.

Andy Ihnatko [00:20:49]:
That's— I think they will be fine for this chip shortage. I think they'll do okay.

Leo Laporte [00:20:53]:
Yeah. Wow, wow, wow, wow. Um, all right, uh, let us take a little break and then we're going to talk about some other things. There's some legal matters pending and some unpending, I guess. Is that what you call it when lawsuits get dismissed? They're now no longer pending. Plus, uh, some AI stories. I guess that OpenAI lawsuit's still going on and a whole lot more. You're watching MacBreak Weekly.

Leo Laporte [00:21:18]:
Andy, Christina, and Jason. Um, I, you know, I'm a big supporter of Section 230 because it protects me, it protects our chat room, it protects our forums. It protects our Mastodon instance. But there are a lot of people who say, oh no, 230 protects big tech and it lets big tech get away with murder. 230, for those who don't know, and the best place to find out about it is to just Google what is Section 230 on the site techdirt.com. He has a really excellent— Mike Masnick has an excellent explanation. But basically it's a section of the CDMA which says Communications Act, which says that an information service is not liable for the content posted on that service. It gives us— it's short.

Leo Laporte [00:22:09]:
It gives us the ability to moderate. We can't be sued for taking down content, which is really important. We can't be sued for leaving content up, and we can't be sued if some bad actor comes and posts in our chat room something evil. But it also protects the big companies. And there was a $32.8 billion lawsuit against Apple and iCloud from 2024. It accused Apple of, quote, privacy washing and failing to deal with the issue of CSAM, child sexual abuse material, on iCloud. The judge has now dismissed this as of July 14th, a week ago. District Judge Noel Weiss in San Jose, California agreed with Apple's argument that Section 230 of the Communications Decency Act protects it.

Leo Laporte [00:22:59]:
They're not liable for CSAM on iCloud.

Christina Warren [00:23:04]:
Hmm.

Leo Laporte [00:23:05]:
And that was going to be a class action suit, 2,680 people, compensatory damages potentially of $32.8 billion. The judge said there is no federal law that would Which then— this, by the way, I like— there was no federal law that would force Apple to create new tools or deploy existing technology to identify and report CSAM. Remember, Apple initially proposed some sort of scanning technology, but after all of the backlash, backed down. And now under the ruling, they are not liable for that. Yeah, nobody likes CSAM.

Andy Ihnatko [00:23:39]:
Let's be clear.

Christina Warren [00:23:40]:
Yep.

Leo Laporte [00:23:41]:
If Nobody wants CSAM to exist.

Jason Snell [00:23:44]:
Well, somebody does and they're very bad and they should go to jail. Yeah.

Christina Warren [00:23:48]:
Yeah.

Andy Ihnatko [00:23:49]:
It's, I mean, the, the, the, the judge was actually kind of explicit, like for, if, if you read a lot of these rulings, like it's weird. It's, it's a little bit out of character when you see a judge inject what seems to be a very, very personal sort of part of the narrative. And she was, she was, you mentioned that, uh, hey, there's no federal law that requires Apple to do this. Therefore they're out. This is, they're absolutely off the hook here. But she did say, she did say that within the context of saying that she kind of wishes that Congress would pass some sort of a law because of the shelter that modern technology gives to people doing one of the most incomprehensible and horrific crimes that are on the books today. And I mean, the worry is that this is a valid argument, and that the problem is that if— although I kind of support this kind of idea that we should find a way to basically require companies to do advanced scanning on this. I think that there are a lot of parties that are more interested in weakening Section 230 as a whole, that basically saying that we want to remove shield— shielding from all these companies just as a matter of principle.

Andy Ihnatko [00:24:57]:
We need to rewrite Section 230 or amend it. Not A, because yes, we will get this law passed to give companies requirements to be more strict against CSAM, but also we want the ability to sue YouTube, to sue Google, to sue everybody because they decided to publish a very, very unflattering photo of someone in power who does not want that photo to exist. So that is the danger, but it's hard to know what side to— it's, as Jason often says, it's possible to have more than one opinion about something, and this is the most confounding of all because all of our principles are kind of at war with each other.

Leo Laporte [00:25:34]:
Yeah, nobody wants CSAM. This is the post that everybody should look at on Tech Dirt if you're confused about this, particularly if you think Section 230 is a bad thing. Hello, you've been referred here because you're wrong about Section 230 of the Constitution.

Christina Warren [00:25:48]:
Yay!

Jason Snell [00:25:48]:
Go Mike!

Leo Laporte [00:25:48]:
This is Mike Masnick. He wrote this, it's 6 years old.

Christina Warren [00:25:52]:
Still great.

Leo Laporte [00:25:53]:
It's still absolutely germane. And we are still hearing now a lot of people saying, oh, it's got to be modified, it's a massive gift to big tech.

Jason Snell [00:26:03]:
This is the problem, is it is— look, CSAM Is about the worst thing you could imagine.

Leo Laporte [00:26:12]:
Yes.

Jason Snell [00:26:12]:
And that's no accident. They're using the worst thing you can imagine because all sorts of interests want to use this and make them say— this is why I get kind of mad about this whole thing. There are a lot of people out there who don't really care about this. They don't. They just want to find a bad thing that they can use as a battering ram to get what they want. And they want to frame it in a way where if you're against what they wanna do, you're for CSAM, which is not the case. But what they wanna— that's the game that they're playing, is like, you're either with me or against me, and if you're against me, you're with the monsters. And these people wanna do stuff like put cracks in Section 230 so that they can— there is controlling of speech and chilling of speech on the internet.

Leo Laporte [00:27:07]:
Mm-hmm.

Jason Snell [00:27:08]:
And let's not forget, the other part of this is using it as an excuse to break down end-to-end encryption.

Christina Warren [00:27:14]:
Yes.

Jason Snell [00:27:15]:
Because ultimately what they want to do is they want to be able to look at all of your digital information whenever they want. And they are using CSAM as a pretext for it. And that's why all of us who are tech savvy need to stand up and say, this is not it. And honestly, I got to be honest, as much as I am sympathetic to the people who stand outside Apple Park at WWDC every year with their signs that say, Tim Cook, fight against, stop supporting CSAM, stop supporting child abusers. And they're out there every year. The fact is, those people, I believe, really believe it, but they're being used by people who don't care about their thing because they just wanna use it to get our information and control our speech. That's it.

Christina Warren [00:27:59]:
I could not— I could not agree more. I think you're exactly right. The other thing I would say, too, is that I was very upset when 5 years ago Apple introduced plans that they did not implement where they were going to be automatically scanning your iCloud for CSAM. And they got a lot of very valid pushback from people in the privacy community, from parenting groups, from all kinds of folks, because I— maybe the technology eventually gets to a place where you can have a no, you know, like false positive positive error on that? I don't think so. Right. I also— there are so many questions about that in my mind. And I was very, very upset when they did that. But I was very, very heartened when they pulled back from that and said, no, we're not going to do that.

Christina Warren [00:28:40]:
And so the fact that they're— you know, this lawsuit has been dismissed, I'm very happy about that. But I hope that this does not in some way try to encourage whether it's Apple or other companies to say, oh, well, yeah, we should take this on again. And maybe it won't be required by the government, but we'll go ahead and we'll put these provisions in place ourselves voluntarily. Because as Jason said, like, I'm sympathetic to people's position, but I think they're being pawns and being used. And this is really just the worst example. Jason said it perfectly. This is the worst example that's being used so that they can take away our privacy, so they can take away end-to-end encryption, so they can control, you know, flow of information and come after us. And if we think that won't happen in the United States of America, like, look at people who are being arrested because of things they post on social media that are, you know, that criticize the government, right? Many of those cases have then been thrown out.

Christina Warren [00:29:31]:
But this is happening now. If you then allow, you know, courts to have access to your cloud contents and get rid of encryption and things like this, it only gets worse. And the people who fundamentally are usually hurt the most are not the people who are actually committing the crimes, because the criminals will be finding other ways to commit the crimes regardless. But regular people and our regular, you know, actual rights are the ones that get infringed on.

Andy Ihnatko [00:29:58]:
Yeah, and it is absolutely true that these are— oftentimes these are just Trojan horses for something far more sinister. Just like I think that most people agree that there should be protections for children on social media, And it would seem as though having age identity checks on all new social media accounts is a way to do that. And but the thing is, that's also a way for large entities to think that we want to make sure that privacy on the internet does not exist, or anonymity on the internet rather doesn't exist. We want to make sure that everybody who wants to use the internet in any meaningful way must have at some point submitted their their actual provable identity to to a force. So then, so that, hey, if it means that a government can then subpoena that information and unwind the identity of somebody who posted something on Blue Sky that, again, a totalitarian regime does not approve of, a regime that likes to find people like that and make them disappear, hey, who cares? So yeah, we need to be completely on watch. I mean, Section 230 might need some looking at and some maintenance as it is. A German court I think a German court recently, for instance, ruled against Google for their AI Overviews, basically because—

Leo Laporte [00:31:14]:
Yeah, it was Germany, yeah.

Andy Ihnatko [00:31:15]:
Yeah, there was a couple of different publishers were basically maligned by these Overviews because when people were asking, hey, what's the reputation of this organization? AI Overviews confused them with other organizations that were doing shady subscription practices, all this sort of stuff. And so Google tried to get out of this by basically using their German equivalency of Section 230, saying that, well, look, we're protected as just We're not an editorial, we're just basically a publisher. And the German court says an AI is not a human being, and therefore human beings do not have access to that protection. Therefore, you are on the hook for all of this.

Jason Snell [00:31:50]:
Yeah.

Andy Ihnatko [00:31:50]:
So a lot of conversations need to be had, and we have to make sure that bad things do not get slipped in by trying to basically improve these little weaknesses and these problems. We wanna replace, we wanna make sure that these specific actions fix a specific problem are not used to introduce bigger problems that for the past 10 to 20 years law enforcement and governments have been trying to get to basically put the internet under its heel.

Leo Laporte [00:32:18]:
Here's another one. As long as we're in court, let's go back to court.

Andy Ihnatko [00:32:24]:
We got our suit on.

Leo Laporte [00:32:26]:
You may remember in 2024, the United States Department of Justice went after Apple saying that its monopoly in smartphones violated antitrust laws. Apparently, according to Mark Gurman at Bloomberg, they are in early discussions about settling that lawsuit. Gurman says the 2 sides have exchanged draft settlement materials, which usually means, you know, well, obviously means they're actively negotiating and maybe even getting close. No trial date has been set. Apple has apparently— this is not the first time Apple's made offers to the DOJ. But the activity, according to Gehrman, has picked up in recent weeks with 2 sides exchanging draft settlement materials. Yeah.

Andy Ihnatko [00:33:09]:
And that's generally how these things end when the DOJ brings suits.

Jason Snell [00:33:12]:
They don't—

Andy Ihnatko [00:33:12]:
they can't fight this tooth and nail. They don't have the resources to fight this tooth and nail.

Leo Laporte [00:33:16]:
This was under Lina Khan at the FTC.

Christina Warren [00:33:18]:
And I was going to say, this was a different DOJ. This was a different FTC, different administration. So, yeah, this is not surprising to me at all. And honestly, I always felt like that was— are there things that you could potentially— antitrust actions that you could bring against Apple? Yeah, I think there probably are. Were these that? I thought— and I'm not a lawyer or anything, but just as a— as an observer, as a commentator, I felt like these were fairly weak things to bring against.

Leo Laporte [00:33:41]:
They don't have 100% of the smartphone market.

Christina Warren [00:33:44]:
Yeah.

Jason Snell [00:33:44]:
No, this is like the super apps thing.

Leo Laporte [00:33:48]:
Yeah.

Andy Ihnatko [00:33:48]:
Yeah.

Jason Snell [00:33:48]:
They're like, we already did that. The super app, the mini apps, and they changed some policies. a little bit. And like, plus, don't forget, it's the Biden administration.

Christina Warren [00:33:56]:
Well, exactly.

Jason Snell [00:33:56]:
Department.

Christina Warren [00:33:57]:
So different.

Andy Ihnatko [00:33:58]:
Just very antitrust.

Christina Warren [00:33:59]:
Yeah, it was. And they failed, right? Like, she failed. Lina Khan failed at every single thing that she tried to do. So, you know, congratulations, I guess. But yeah, this doesn't surprise me that they're in settlement talks.

Leo Laporte [00:34:11]:
19 other states and the District of Columbia are in the case, and it's not clear whether the states' attorneys general are involved in the discussions, but I think this will be dropped. Look, they dropped Biden's Live Nation lawsuit, which was much more merited, if you ask me. I have a feeling the settlement is not far off. Just, you know, another gold bar. It's just, it's a simple thing. It's not complicated. That ballroom isn't going to build itself. Damn.

Andy Ihnatko [00:34:42]:
God. Yeah, let's see.

Leo Laporte [00:34:45]:
It's not getting much done at all, as a matter of fact, but that's another story entirely.

Andy Ihnatko [00:34:49]:
There's a— there's a— that's an interesting argument though, that, uh, going back to Apple now being close to a $5 trillion company, it does mean that they have a lot of legal firepower that sometimes can be wielded in the favor of the public good, such as, hey, look, if you want to attack Section 230, like, we are— if you want— if you want to, uh, to solve CSAM, we're happy to solve that together. But if you basically make us compromise the security and the privacy of all of our devices for all of our users. We got $5 trillion to spend to make you, to make you like basically drop it. How— what's your budget to make this bill go through? Versus we feel as though we don't need— we can run our company however we want, that we can basically have a monopoly. We can break antitrust laws, we can do whatever we want. And we have $5 trillion to back up the idea that whatever you try to do, whatever action you try to lodge against us, We can basically keep this tied up so much that we will basically get anything pled down to $500 fine and time served. So come at us if you like. It's— there are two sides to every corn, and it's good and bad.

Leo Laporte [00:35:55]:
Two sides to every corn?

Andy Ihnatko [00:35:57]:
Coin.

Jason Snell [00:35:59]:
Oh.

Leo Laporte [00:36:00]:
Last time I had a corn on the cob, there was only one side.

Andy Ihnatko [00:36:03]:
I was channeling Ed Norton on The Honeymooners. Yeah, we'll flip a kern.

Leo Laporte [00:36:08]:
Sorry about that.

Jason Snell [00:36:10]:
It's the product syndicated television of the 1980s.

Leo Laporte [00:36:16]:
There's 2 sides to every kernel, I will grant you that. Uh, so I like the idea of replaceable batteries. I do wish my iPhone had a replaceable battery, although there's a rumor— we'll get to that— that it might have some sort of double battery thing. Um, there are, there are devices which is pretty hard to replace the batteries in, like, I don't know, AirPods. Apple Watches.

Andy Ihnatko [00:36:38]:
Rings.

Leo Laporte [00:36:38]:
The EU, which was gonna require that devices have easily removable and replaceable batteries, have now ruled that, yeah, no, if it's a wearable, you don't, you don't have to replace it. Apple Glasses, if they ever do one, Meta Glasses, no longer need to have replaceable batteries. I guess it's just a technical thing, right?

Andy Ihnatko [00:36:59]:
You just Yeah, it's just—

Leo Laporte [00:37:02]:
I can't do it.

Andy Ihnatko [00:37:03]:
This is why, like, I admit this is a prejudice, but I'm prejudiced when the EU takes those kind of legislative, regulative actions. I don't think that everything they do is great, but I do think they have a good track record of proposing something and then modifying it as they go when they see when problems seem to occur. And to me, this seems like one of those problems. I would like to— I would like to see companies like Apple encouraged to figure out a way to make batteries replaceable inside of AirPods. There are third parties that can do that. for you, although it's kind of hit and miss. I mean, uh, we, I, we did discuss like a few weeks ago, we were talking about the, the idea of an Apple Ring that I just hate the idea that this is a $400 device that will last for 2.5 or 3 years and then become e-waste. But I know, I bet, but how, I mean, you can with a ring.

Andy Ihnatko [00:37:48]:
Yeah, that is like how there's absolutely no way to do it.

Leo Laporte [00:37:51]:
Yeah, but I have an Oura Ring and, uh, yeah, I have mis— I have mixed feelings about it, right? Um, Yeah. I've had an Oura Ring die 'cause I couldn't charge it anymore and I replaced it with another Oura Ring.

Jason Snell [00:38:03]:
But yeah. I feel like, you know, a lot of these cutting-edge things, they're just the cutting-edge things are not as replaceable. And then over time they get more space and they have more ability and all of that. So with some of the newer, even something like AirPods, I'm okay with the idea that I can't pop the battery out, but what I would like is a story about how If I take it back, they will disassemble. Because remember, a lot of that stuff has valuable components in it.

Christina Warren [00:38:30]:
Exactly.

Jason Snell [00:38:30]:
Apple has made a great success of using recycled components, in part because they take them back and then they, and then they break them apart and then they take all the components out of them and they use them. I just want to hear that story, right? If it's not repairable, recyclable, renewable in ways that make sense, and I can like, I can give them a flyer because this is so miniaturized, it's less likely that that's going to be possible. That's a— I mean, I can accept that, but I would like to hear what your story is. If I bring these back in 3 years, are you going to be able to take them apart and reuse the components or the elements inside or whatever for future use? Or are they just going in a landfill? Because that's what I don't want to hear.

Andy Ihnatko [00:39:17]:
Yeah. And also, it should be pointed out that this law also is the reason why To replace a battery on an iPhone, you no longer have to disassemble the entire thing from the top down to get at the battery.

Christina Warren [00:39:28]:
Right.

Jason Snell [00:39:28]:
It's not glued in anymore. And like, which is great. Exactly. That's, that's what people misunderstand is like, it's not like I can pop out a battery door and slide it out, but it is within reason. You could unscrew things and pop the battery off and do a self-repair.

Christina Warren [00:39:44]:
Yeah. Yeah. And there were reasons, you know, why Apple had moved into the direction that they moved into, you know, away from replaceable batteries. and laptops and in phones. I mean, I think especially now that we see that they can actually do it without glue and without the tabs, that, that shows that that was always possible. It just wasn't maybe the easiest thing to do. But like, there are valid reasons to do that insofar as like just economy of space and how you can use the space most effectively so that you can actually have your, your battery as much capacity as possible in in the size of the chassis that you have. And so in a laptop, that means that you have multiple cells.

Christina Warren [00:40:20]:
And in a phone, it means that it's not going to be, you know, like the normal type of shape. And I think that those are all tradeoffs that a lot of us would probably take. But yeah, I mean, you know, one of the nice side effects of this is that, yeah, it is now easier to, to repair your battery, replace it with, with, you know, a part that now Apple does provide. Granted, you know, they don't make it super easy to get, but you can in fact order, you know, apart from them. And, and they even, you know, provide the tools to you if that's what you need. And so I think that's all positive. I think things like AirPods, would it be great if they could get to that point someday? Yes. I think especially when you look at the over-the-counter, like, hearing aid market, you know, standard hearing aids that you buy usually do have replaceable batteries.

Leo Laporte [00:41:05]:
That's a good point.

Christina Warren [00:41:06]:
And so, you know, they usually do. Now, granted, they don't have as much of the same tech and everything else that like the Apple stuff does, but it is a possibility. Now you might have to charge more. That's the only thing. And so I think that that becomes—

Leo Laporte [00:41:17]:
Smaller battery so that you—

Christina Warren [00:41:18]:
Smaller batteries, right? And it's also, you know, part of the reason why the AirPods are so cool as over-the-counter, you know, hearing aids is that it's $250 versus $1,200, right? So that, I'm sure that part of that is— Mine were $7,000. Oh yes. I was using Costco pricing. So I was already using—

Leo Laporte [00:41:39]:
I should have gone to Costco.

Christina Warren [00:41:40]:
I mean, but that's— everybody should go to Costco for hearing aids, but especially if you already have like a good hearing exam done, 'cause usually you can get the same hearing aids.

Leo Laporte [00:41:50]:
Oh, I didn't know that.

Christina Warren [00:41:50]:
But for a fraction of the price. Yeah, you can save thousands. Like I think they have—

Leo Laporte [00:41:54]:
Yeah, they're really expensive if you go to an audiologist.

Christina Warren [00:41:57]:
They really are. And so I'm sure that's part of why, you know, like you've made a consumer product, you're not necessarily thinking about how—

Leo Laporte [00:42:03]:
Well, my first set of hearing aids actually had batteries that you took out and put in new ones. They wore out. Now they're rechargeable. But, uh, yeah, so I mean, and they're not much bigger. The rechargeable ones in fact are the same— not much smaller, I should say— that it's about the same size. So yeah, it's— you can do it. Uh, one of the things the EU said is, well, especially because they have to be waterproof and dustproof, and you know, you don't want to have a case that's opened. I don't know.

Christina Warren [00:42:28]:
Yeah, I mean, look, I'm not saying that it's even like a pressing thing. I'm glad that they made an exception in that case, but it is something They needed to. But I'm just saying, like, this is an area where, as we've seen with phones and laptops and other components, that maybe eventually we will get to, you know, economy of scale as such where technology will improve enough that you can make them, if not easily user replaceable, then at least serviceable, to Jason's point, so that at the end of this you're not just e-wasting the AirPods, which is, I think, what they do now.

Andy Ihnatko [00:42:56]:
Right.

Christina Warren [00:42:56]:
Like, I'm pretty sure they just chuck them. I don't think that they take anything any part of the components from them at all. And that would be nice if they, if they had a mechanism to not do that. I think Apple Watches, I think they can probably save, you know, the, the mainboard and, and other components from that. But, you know, I think, I think AirPods are probably one of those things that you toss.

Leo Laporte [00:43:16]:
When I buy a new device, well, not the AirPods, but the Apple Watch, I send it in. Yeah. Do they recycle AirPods? Do we know? I don't think they do, actually.

Christina Warren [00:43:25]:
I think that— I think that they— I think they responsibly get rid of them, is I think what the term is. I'm not exactly sure because it's one of those things like—

Leo Laporte [00:43:33]:
Certainly fire tended by—

Jason Snell [00:43:34]:
I think they remove— I think they do remove the batteries. I think they've got a lithium-ion— lithium-ion.

Leo Laporte [00:43:39]:
Yeah, you could probably pry it apart. You just—

Christina Warren [00:43:41]:
yeah, you could probably pry it apart. Sure.

Jason Snell [00:43:42]:
Yeah.

Christina Warren [00:43:43]:
I mean, and that's possible, right? I just more mean like it's not like they're going to be reusing those components.

Jason Snell [00:43:47]:
I don't know.

Christina Warren [00:43:47]:
Right. Whereas like an Apple Watch, because they do sell refurbished Apple Watches, you could— and they do sell, you know, like refurbished Apple Watches. That could be one that I could see being resold. Whereas AirPods, yeah, maybe take the batteries out to responsibly recycle them. But it's not like they're going to be, you know, reusing that, you know, transistor or whatever.

Leo Laporte [00:44:08]:
You guys use Hide My— the Hide My Email feature? Anybody use that? I've used it.

Christina Warren [00:44:13]:
I have.

Leo Laporte [00:44:14]:
When I buy stuff through Apple Pay, you have the choice of having your Mac email, or you could say no, hide my email, don't use my real email, which gives you some sense of privacy. Mail to that address that they provide you does end up getting forwarded by Apple to your iCloud account, but apparently 401 Media and a security researcher reported this earlier this month, uh, it don't always work. The privacy isn't perfect, and apparently they can leak your email. I'm not sure how, but Apple is now being sued, um, because these randomly generated email addresses, uh, apparently aren't as private as we thought.

Christina Warren [00:44:58]:
I mean, look, you can sue anybody for anything, but this seems pretty silly.

Leo Laporte [00:45:02]:
Yeah, I don't— yeah, it's a class action.

Christina Warren [00:45:04]:
It's a class action thing, and, and I get that.

Leo Laporte [00:45:06]:
Here's the reason there's maybe some merit to this. Apple knew about this for more than a year.

Jason Snell [00:45:11]:
Yeah.

Leo Laporte [00:45:11]:
And did nothing.

Andy Ihnatko [00:45:13]:
4 or 4 media is reporting this morning that they finally closed the, the, the thing.

Leo Laporte [00:45:17]:
The hole? Oh, good.

Andy Ihnatko [00:45:19]:
That's, that's what annoyed the hell outta me, that this was around for like a long time and that it wasn't a, oops, we'd better fix this right away. I mean, a class action suit is kind of okay. It's, it's the same reason why there's a class action suit regarding Apple's promises on Apple Intelligence that based on, wow, I better, I'm gonna buy a brand new iPhone 'cause in just 2 or 3 or 4 months' time, it's gonna be able to do all this magical stuff. I don't think that Apple has sinned and they need to repent, and they do, but it's— this is why class-action suits exist. If all that happens is that all the people who respond to this get $4.80 in iTunes Store credit, at least Apple has to say, yeah, you know what, maybe we should have fixed that before a year had passed by.

Leo Laporte [00:45:59]:
If people thought that they were getting credit, but they weren't. Yeah, but I mean, they were selling it as privacy. I mean, you know, it was— people paid for iCloud Plus Partly, perhaps, because of the Hide My Email feature.

Christina Warren [00:46:09]:
Maybe. Maybe. I mean, I don't know. I guess the way I look at it is I think it's slightly different than the Apple intelligence thing, because in that case, you literally had, like, as we now know, like, created out of thin air vaporware demos that showed a product that did not exist.

Leo Laporte [00:46:22]:
That's pretty bad.

Christina Warren [00:46:23]:
And you sold that as the thing that was going to happen. Right. And do I feel like they should be raked over the coals for that forever? No. Companies do that all the time. But Apple typically is not. Whereas in this case, I don't know, even when I sign up for a service, and maybe I'm alone in this, but even when I've used that feature before, I've never actually believed, like, deep in my heart of hearts, that my information was completely private from the company that I signed up with, right?

Jason Snell [00:46:49]:
Yeah.

Christina Warren [00:46:49]:
Like, I'm still choosing to give an email address to this company, right? And so there's still a database table somewhere that has whatever this Apple ID is, ID is, and it is connected to my name, right? So I do understand the argument, like, oh, well, I paid for this, and then I was told it was one thing it wasn't. And it's certainly not good that they let something go for a year without fixing the hold. But I also feel like, come on, this is a little bit of a— you're still knowingly giving out your email address to a company.

Andy Ihnatko [00:47:19]:
Right.

Christina Warren [00:47:19]:
And ultimately, how that's going to be used is up to that company.

Jason Snell [00:47:22]:
It's not up to It's not quite a secure— I mean, it is a security feature, but I always felt like the real killer feature of this was you could turn the alias off.

Christina Warren [00:47:31]:
Exactly.

Jason Snell [00:47:32]:
Right. Not that, not that no one will be able to guess who I am, because if you do anything to interact with them, then you're going to tell them who you are, even if you don't give them your email address. But more that they're going to take that email address and who knows who they're going to sell it to, or they're not going to honor your unsubscribe. And then you have the ability to just To mark it off. But even so, if they can look up your actual Apple ID, then they've got another email address to harvest. But it just, it's annoying. And it's annoying that it hung out there for a year and that as soon as it got press, they fixed it. That's like the worst.

Jason Snell [00:48:03]:
But in the grand scheme of things, it's not the worst, uh, security breach you could have. It's like low, a low-key security issue. Yeah.

Leo Laporte [00:48:14]:
Yeah, you know, I mean, I, I think many email providers, certainly my email provider Fastmail, lets me— well, that's because I own the domain. So I have a domain and I can have an infinite number of, uh, sure, names before the sign. So what I mean, when I sign up for something, you know, I sign up for 404 Media, it's 404 Media at that email address. So I have an infinite number of those. It's not private in that sense. But I can certainly filter it out. And I know if they sell my address, I know who sold it, which is always—

Jason Snell [00:48:46]:
Oh man.

Leo Laporte [00:48:47]:
Surprisingly, that doesn't happen as often as I thought it might.

Jason Snell [00:48:50]:
You know who does it is Bloomberg. You pay like $80 a year or whatever for Bloomberg and they sell your email address everywhere. It's so annoying.

Christina Warren [00:48:59]:
They do. And political campaigns, if you ever donate to any political campaign, they will sell you down the river, like your phone number, your email, everything. To the point that like I actually won't donate to campaigns anymore. Um, and it's— it— and I, I'm a civically minded person. I would like to give to them, but I do not want their, their emails. I do not want their texts, you know?

Leo Laporte [00:49:18]:
Right. Uh, yeah, so just use a junk mail address, right? Isn't that what people did for years? They had a Yahoo mail address.

Christina Warren [00:49:25]:
And yeah, it's true. I mean, and I think a lot of us have that. Um, it's sometimes like when they want your phone number too, that's the one that bothers me because that's that's a lot harder to get around unless you have multiple phone numbers, in which case then you have to deal with multiple, you know, things of spam and whatnot. So it's— yeah.

Leo Laporte [00:49:43]:
Yep. All right, let's take a little break. More to come. We haven't mentioned OpenAI this whole show. It's an amazing thing. Here we are almost an hour in. We will, don't worry. Christina Warren is here.

Leo Laporte [00:49:57]:
Film girl, always nice to see film Girl. I see we've taken the underscore out in Film Girl. Are you using the underscoreless version now?

Christina Warren [00:50:05]:
Well, okay, no, the problem is— it's 2 problems. It depends on the service. The problem is, is that I did not realize when I, like, took that on as my username that that would be my username everywhere. And GitHub, for instance— Forever.

Leo Laporte [00:50:18]:
For the rest of your life.

Christina Warren [00:50:19]:
Basically, right? And so GitHub does not have an underscore because they don't support it, right? And, and so same thing with with Bluesky, but Mastodon does, Twitter does. So it varies on— Instagram does. So it varies service by service. So I'm either film_girl or filmgirl depending on where I am. But if you, if you search Filmgirl, you'll usually find me is what I, is what I try to say.

Leo Laporte [00:50:39]:
There you are, Filmgirl without the underscore on GitHub. You know, uh, my son, when he started his business, called himself salt_hank. And I said, dude, what are you doing? What are you doing? Underscore?

Jason Snell [00:50:53]:
Really?

Leo Laporte [00:50:53]:
He didn't put any thought into it at all.

Christina Warren [00:50:56]:
I mean, he was like me, and, and now he's so famous.

Leo Laporte [00:50:58]:
He didn't know what would happen.

Christina Warren [00:51:00]:
And, and, and he's much more successful than I am, but like, it's become— it becomes your branding thing, right? Like, that's just— it is what it is. Like, things you do when you're in college, you do not think are going to become like—

Leo Laporte [00:51:09]:
You know, that is such a profound— things you do in college, you don't think what might happen when you get out of college.

Andy Ihnatko [00:51:17]:
I just like that sign that I am so old school, I predate being able to have put spaces and file names, dude, and account names. That's how old school I am.

Leo Laporte [00:51:26]:
Wow.

Andy Ihnatko [00:51:27]:
I still don't trust lowercase and uppercase, man.

Leo Laporte [00:51:30]:
Whoa. Well, I'm surprised that Bluesky doesn't allow the underscore. That is weird.

Christina Warren [00:51:35]:
GitHub doesn't either. A lot of places don't, and I guess it's about sanitization.

Leo Laporte [00:51:38]:
GitHub I understand, but Bluesky, look, if you're gonna be a social network—

Christina Warren [00:51:42]:
I don't know. I think it comes down to sanitization practices, probably. It probably makes— it probably messes up their regex stuff. We have, one of my good friends at GitHub is Matt R Dash. And that is literally his username. He's been at GitHub for a really, really long time. And his username is used as like a sanity testing thing in a number of different services to make sure that it won't catch somebody's username. Because it's not common that people will have, you know, like you can have a dash in a username but not an underscore.

Christina Warren [00:52:12]:
And it's not common that you'll have that.

Jason Snell [00:52:14]:
Right.

Christina Warren [00:52:14]:
since it's at the end of his name. Um, there are a number of services where they, they use him as part of the testing suite to make sure that nothing will break.

Leo Laporte [00:52:22]:
I do know that people still use their, you know, college or even high school or even grade school handles because in our, you know, Club Twit Discord, there are a lot of weird handles. Mashed Potato, uh, Chocolate Milk Mini Sip. Actually, it looks like he's changed that, uh, so he doesn't use that anymore. And there's Film Girl with an underscore. We allow an underscore, Film Girl.

Christina Warren [00:52:43]:
Yeah, which I appreciate. Thank you very much.

Leo Laporte [00:52:47]:
Andy Inako is here. He has a name you can't even spell, but that's how I remember it. I have no idea how to spell it. I have no atko. I-H-N-A-T-K-O. You told me that years ago. A friend of yours uses that.

Andy Ihnatko [00:53:02]:
And I know that was— there's somewhere, there's a middle-aged woman who used to run, used to work the counter at a video rental store, like near my neighborhood, who has given me this legacy because on her last day, I'm like, oh, I'm sorry, you're the only person. Because, because of course my name on the card was like what you like, what your account name at this independent thing. Oh, I remember because I just said to myself, I have no idea how to spell his name. So she has no idea. She's, she's living her life, having a great time, but she does not know what a positive impact she has had on a stranger's life.

Leo Laporte [00:53:34]:
Let's hope I hope she's not still working at the video film store.

Andy Ihnatko [00:53:38]:
Oh no, that went out of business a long time ago. Yeah, I bought their shelving when they went out of business.

Leo Laporte [00:53:44]:
Wow, okay, you have hundreds of feet of shelving now. Aren't you a lucky guy?

Andy Ihnatko [00:53:49]:
Shelving like 5 inches high and like 8 inches deep. It's not the most useful stuff in the world.

Leo Laporte [00:53:53]:
Just the right size for all those VHS cassettes. And Jason Snell is here, sixcolors.com. See, he didn't put it under It could have been 6_collars.com, but no.

Christina Warren [00:54:03]:
He was smart.

Jason Snell [00:54:04]:
No, you gotta— you can't do that. You gotta have English words. I learned my lesson with, with domains that weren't .coms and that you had to spell out loud letter by letter, and I'm not doing that again. Also, yes, and I'm @jsnell everywhere, mostly because that was the format that was assigned to me my sophomore year in college when I got my first email address that was personalized, and it's been that way ever since.

Leo Laporte [00:54:25]:
So yeah. Yeah, you can tell when people have Unix handles. Like that they have from college. There's a certain, I don't know what it is.

Jason Snell [00:54:32]:
It's lit. Yeah. Literally I went into the, you know, the math department. Yeah. Or yeah, we were aware it was physics and math building, I think at UC San Diego in the, you know, in September of what, 1989? And they said, okay, you are jsnell@ucsd.edu. And it's like, that's it. That's it forever.

Andy Ihnatko [00:54:50]:
The first name, last initial generation of internet usage.

Jason Snell [00:54:53]:
Yeah.

Leo Laporte [00:54:53]:
I think I'm Leo L. Yeah, Leo L.

Jason Snell [00:54:55]:
Well, the year— my freshman year, they just assigned me a random number. There was— they're like, we're not personalizing undergraduate computer accounts. You're PA151.

Leo Laporte [00:55:04]:
I am not a number.

Andy Ihnatko [00:55:06]:
I'm a man.

Jason Snell [00:55:07]:
But the next year, I— they, they were like, all right, you're— maybe it was like, you're a sophomore now. We— you can have your name now.

Leo Laporte [00:55:14]:
Speaking of I am not a number, uh, The Prisoner is on the Criterion Collection now on the Criterion Channel. I've been watching the old Prisoner shows. Wow, they, they are a trip, literally.

Andy Ihnatko [00:55:27]:
Oh, there's— it's still, you know, I was watching, I was watching, to mention another, another awesome show, I was watching one of the episodes of Columbo where Patrick MacGoohan was the killer. He did— he was the killer 3 times, and there are so many freaking Prisoner references. It's delightful. Even, even really, even, well, A, he, he plays a spy to begin with, but also he says, be seeing you, to her secretaries like a couple of times. And also he's wearing in his, in his big like walk and talk with Leslie Nielsen, who plays like his murder victim. He's got like not a— he doesn't have the same jacket as he has with The Prisoner, like the black jacket with the white piping.

Leo Laporte [00:56:02]:
Piping, yeah.

Andy Ihnatko [00:56:02]:
But he's got like a dark, dark, dark blue windbreaker and the zipper lining is white, so it has the exact same effect. I'm like, oh, you knew what you were doing. You knew exactly what you were doing there, man.

Leo Laporte [00:56:15]:
Anyways, he was a secret agent, man, uh, the Johnny Rivers song. He was a secret agent. Was a great show which I used to watch. And then And The Prisoner was kind of his rebuttal to all of that. Yeah, the good old days.

Andy Ihnatko [00:56:29]:
Anyway, such a great series.

Leo Laporte [00:56:31]:
Yeah. Uh, is the OpenAI lawsuit still on? Are we still, uh, still suing them?

Christina Warren [00:56:36]:
They're still suing.

Leo Laporte [00:56:37]:
Yeah, still suing. Have we learned any more?

Christina Warren [00:56:39]:
Well, they have now sent discovery letters, um, or I guess don't delete letters to 40 employees, which, uh, okay.

Leo Laporte [00:56:47]:
So Financial Times says Apple has sent letters to former 40— it's the other side of the corn. Former 40 former employees. It's not easy. Let me start over. The Financial Times is saying Apple sent letters to approximately 40 former employees saying, do not delete. Um, boy, wow. Can you, can you enjoin former employees not to delete documents?

Christina Warren [00:57:18]:
I don't know.

Andy Ihnatko [00:57:18]:
I—

Christina Warren [00:57:18]:
this is what is unclear, and I looked at the Financial Times direct source, not just Just the 9to5Mac.

Leo Laporte [00:57:23]:
Yeah, let me go back and go there.

Christina Warren [00:57:24]:
And I still am— it's still not clear to me if this was like official legal communications or if this was just a lawyer sent you an email that said don't do this, right?

Leo Laporte [00:57:32]:
It's a demand letter. In fact, they demand meetings with Apple lawyers, right?

Christina Warren [00:57:36]:
But I don't know if that is legally binding or if it's just like a scare letter.

Leo Laporte [00:57:40]:
Anything.

Christina Warren [00:57:41]:
I was going to say, like, like, all I know is— and I'm obviously not giving advice to anyone— but if I were to receive a letter like this, I would immediately get my own counsel.

Andy Ihnatko [00:57:49]:
Yes.

Christina Warren [00:57:50]:
That is not— that is not OpenAI's. That is not anybody else. I would immediately get my own counsel and I would like abide by whatever they say. But I certainly would not be like, oh yes, Apple, I will absolutely give you everything and I will show up for this meeting, right? Like, no. But— Does this—

Leo Laporte [00:58:04]:
if you're one of those 40, the presumption is you went to OpenAI, right? These are probably all OpenAI employees.

Jason Snell [00:58:10]:
Yes.

Leo Laporte [00:58:10]:
Yeah.

Andy Ihnatko [00:58:12]:
So it's a hell of a distraction technique. Like now, So much of your company is now going to be devoted to addressing all these concerns of employees who are getting these letters saying, what do I do? What do I do? What do I do? And that's a really— whether it's necessary or not, what a nice intimidation tactic. Yeah, those are wartime lawyers.

Jason Snell [00:58:32]:
Yeah, this is one of those things where, you know, if what is alleged is even like remotely true, I think there's some people who work at OpenAI who are in trouble regardless of what it means for OpenAI. But, but as I said last week, make no mistake, part of this is also about intimidation. It's about intimidation at OpenAI to, to, you know, shake some things loose there and maybe get them to change their behavior or to get people to admit to what they're doing there. And it's also to create a chilling effect around people who might be thinking of going to OpenAI from Apple. And that, you know, again, I'm a real advocate for people who want to get a better job and get paid better. And obviously Apple would like those people not to have left, but they don't get to decide that. But, you know, a side benefit of this is it makes everybody a little on edge about Apple and OpenAI and about leaving Apple. And that is— it's probably not the prime motivator here, but it is totally a motivator.

Leo Laporte [00:59:31]:
Yeah.

Christina Warren [00:59:31]:
I also wonder, too, if in addition to— because like, look, if employees want to leave, like there are a lot of places places that they can go. But I do wonder if this means that— and this would also probably be probably not the intended effect, but a nice side effect would be, okay, if OpenAI looks at this and goes, all right, we're not going to recruit as heavily from Apple as we were, right? Like, we have other, you know, companies we can approach if we're wanting to do recruitment efforts, or we can, you know, approach employees in different ways. Like, that could be a side effect too, is OpenAI is just like, okay, we also don't want to deal with this. But yeah, I mean, I think that this definitely will make people much more circumspect before leaving. And I hope that the only impact if somebody wants to leave is that they know, okay, under no circumstances do you take anything with you. Be very above board. Like, don't even— you know, I would go as far as to probably not even download any of my personal documents before I left, just to be clear and clean. Who knows?

Leo Laporte [01:00:29]:
Yeah, I wouldn't assume it's just intimidation. I think it's also discovery, right? Apple's really trying to figure out what people brought with them, what they know.

Christina Warren [01:00:37]:
Well, yeah.

Leo Laporte [01:00:37]:
And then that's what you do in the case.

Christina Warren [01:00:39]:
Well, and I think that's why, like, I would be more curious about what the nature of the demand letter is. Is this literally just from a lawyer that says, keep this in mind, might be calling you, is part of this lawsuit, or is there some sort of legal teeth behind it? And that's what I don't know, because if there's not any legal teeth behind it, Then they can advise you all you want. But if I decided to wipe out my devices, you know, what are you going to do? My lawyer might not tell me to do that. I don't know. My lawyer might say, you know, don't touch anything. But at the same time, if I've already wiped my devices before I got that demand letter or anything else, I don't know what they can do here because it does seem to be at least somewhat saying, okay, we're reaching out to these 40 employees. We think that they might have information that might come out in discovery.

Leo Laporte [01:01:23]:
Right.

Christina Warren [01:01:24]:
But the fact that this came out before it's, you know, even a judge has even weighed in on anything, I don't know the legal validity of any of these demand letters.

Leo Laporte [01:01:33]:
Yeah. Yeah, I mean, a lot of times, I know this happens, that lawyers send out letters that are intended to scare. They may not— you could just ignore them, but they— yeah. Yeah. All right. Well, you know, this is a soap opera that will continue for many, many years. to come. The gift that keeps on giving.

Leo Laporte [01:01:55]:
And we will— you'll keep on hearing about it, I'm sorry to say, for as long as it remains of interest. Um, Apple Intelligence is in China. They've made a deal with Alibaba, and that means they'll be using Quen, Alibaba's models, which are quite good. On as Apple AI instead of, instead of Gemini. Although this raises the question of what exactly is running in the US on Apple Intelligence.

Jason Snell [01:02:28]:
And I think, I think this is specifically not Siri AI, right? This is like 2024 old school.

Leo Laporte [01:02:35]:
This is what happens when you go to ChatGPT instead of ChatGPT, you go to Quinn.

Jason Snell [01:02:39]:
Well, no, I think that, I think that no, this is instead of Apple's foundation models. I think is the—

Leo Laporte [01:02:43]:
oh, entirely. So this— so according to Reuters, China's cyberspace regulator— I'm sure they don't call themselves that— has allowed Apple's intelligence to be registered for use on iPhones in China.

Jason Snell [01:02:57]:
Okay.

Leo Laporte [01:02:58]:
Yeah, you know, obviously, I mean, it makes sense that you would have to use a sovereign company in China. Yeah, I mean, the EU wants to do the same thing, I'm sure. That's why Apple's not in the EU. with Siri.

Andy Ihnatko [01:03:09]:
Of course, there's added dimensions to this. Like every company— every country, I think, correctly says that, look, you're gonna have to store all the personal information of our citizens on servers that are inside our country so that they are subject to our privacy laws and, of course, to our surveillance. But mostly, we need to keep this stuff within our territory. But, of course, it also means that whenever you ask Siri a question regarding June 4th, 1989 in China, China wants to make sure that you get, hey, well, here's the baseball scores for that day. Here's the weather of that day. Absolutely nothing happened with protesters. No protesters. People don't protest in our country.

Andy Ihnatko [01:03:48]:
So there is that.

Leo Laporte [01:03:50]:
Here's, here's what Reuters says. This is a Reuters story. Alibaba said in a statement to Reuters its Quen model will be integrated into Apple Intelligence across Apple's iPhone, iPad, Mac, and Vision Pro operating systems. systems in China. Reuters also says Apple is working with Baidu, a competing company, to develop Apple Intelligence features for Chinese iPhone users. What is Apple Intelligence? That's not Siri AI. That's something different.

Jason Snell [01:04:16]:
That's like on— that's like— I mean, what is that?

Christina Warren [01:04:18]:
Yeah, I think that's the question.

Jason Snell [01:04:20]:
All that stuff. It's all that stuff from 2 years ago. And is it going to be all the stuff from now on device models, some cloud models?

Leo Laporte [01:04:29]:
I guess the question is, will, will China get Siri AI in the fall? And, and if they do, it's obviously not gonna be Google's Gemini, right? Trained. That's an interesting question anyway.

Jason Snell [01:04:42]:
Yeah. And is it, is it Apple's models are approved, but then when they go off the device, they're going to the, the Baidu models, right? I don't know. I mean, we don't know. This doesn't— I like one of the news stories basically says Apple didn't say anything and the partners didn't say anything that we didn't already know. And I was like, yeah. Okay, so we don't know much about this.

Leo Laporte [01:05:02]:
I have to say, Quen has— Alibaba has released a new Quen 3.8. The Quen models are quite good. I use Quen Vision for my cameras. It does a great job describing me as an older gentleman. I use Quen locally. I use a Quen 27B model. It's quite a good model. There's a coder model, and then I use the new 3.8 preview in my from Agentic AI, and it's very good.

Leo Laporte [01:05:29]:
I mean, I mean, I think it's close to the frontier models from the US, GPT-5.6 and at least Opus 4.8, if not Fable.

Christina Warren [01:05:39]:
They do great work. I mean, they are— I mean, the Chinese models, especially the open weight ones, are— a lot of people are very concerned, I think probably overly concerned, but are, you know, that's been a lot of conversation around that. What does occur to me, and again, we don't have enough information about this, but this is sort of an interesting question. To think about is because presumably this could happen in Europe too if the EU passes sovereign laws about what types of AI things can be implemented there. Right now they're just not shipping stuff. But if things were to happen, this I think raises an interesting question for a developer where, okay, if I'm building an app and I'm taking advantage of Apple Intelligence features, if there are underlying different models happening in different countries, can I test for that? How can I account for that? Do I just just, you know, not implement those features? Like, what's the kind of scenario there? And that's interesting because we've had a few cases over the years where there have been certain SDK features that have only been available in certain regions. And China has certainly been excluded from some of those things, but not where something could, I think, be, you know, as— because these are not deterministic systems where you could have a different result or a different feel based on what the underlying model is, whether it's the local model running on device or the cloud model that it's talking to. And so I think that just becomes interesting if you're, say, an app developer and you're looking at, okay, well, am I building multiple apps for multiple different regions or am I only going to focus on, you know, like specific regions or what the scenario will be?

Leo Laporte [01:07:09]:
Yeah. If in fact, and I apologize 'cause I know you used to work for Google, but if in fact they do get Quinn instead of Gemini, I might be jealous.

Christina Warren [01:07:19]:
Well, I mean, honest. Well, I think it would depend on what model they would be using. Right. Because certainly I think the Quinn models are very good and certainly better in some regards than some of the others out there. So I think it would totally depend on what they're doing. But I mean, like, and maybe this— maybe I'm overconflating how important this is. And this is just kind of like how well in the EU you can have multiple app stores and here you can't. And this will be kind of an edge case.

Christina Warren [01:07:43]:
I don't know. But it is interesting that they are going to be bringing some variant of Apple intelligence, however that's defined, you know, to China, albeit following China's rules. And maybe it's just as simple as— and I don't actually know this definitively. I assume that iCloud works slightly differently in China than it does in the US. I know that if you have a US-based account, obviously you're going through US servers. But if you have a China-based account, it It's, it's different. So maybe it's just yet another kind of fragmentation line based on, you know, where you bought your phone and where your Apple account is, is created.

Leo Laporte [01:08:21]:
Yeah. I mean, it makes sense. You're going to get not, not purely because of Tiananmen Square, but you're going to get a sovereign model in. And I'm sure the EU wants to do the same thing. I don't know what you will get if it's Mistral. I'd rather have—

Christina Warren [01:08:36]:
I was— well, I was going to say, I think it would have to be Mistral. I think that they're the only— they're the only one who are of any size. Yeah, they're the only ones who've, you know, kind of raised at, you know, who maybe have the size of that. But yeah, I mean, I'm sure that that's something that— I was in Germany the other week. And that was certainly a conversation that came up a lot from, from various people was, okay, they want Siri AI, they probably do, right? Well, I don't even know so much about Siri AI. It's just more about like, not— there was a sense that I got from talking to European users, AI users that they don't feel like they can trust. Oh, yeah. That the US will allow them to still have access to things.

Christina Warren [01:09:14]:
And so I think that there are— I'm sure that they would like those features on their phone, but I don't know if that's as much of a pressing concern as it would be. We want to make sure that a model that we are getting access to doesn't get taken away from us at the whim of a government that is not our own.

Leo Laporte [01:09:31]:
As in the fable, Ruggles.

Christina Warren [01:09:32]:
Exactly.

Leo Laporte [01:09:33]:
Earlier this month. Actually, this is a— we won't talk about it here, but if you tune in, uh, TWiT from last Sunday and, uh, and the upcoming Intelligent Machines on Wednesday, we will be talking about this. This is a hot topic right now. President Xi at the AI summit said, ah, we support open models, open weight models, we think it's good for the world, and we're going to export our open weight models. But there's also noise that the Chinese Communist Party is thinking about blocking access to its internal— to its open models to the rest of the world. The Trump government also saying we might want to block Chinese models from use in the United States. Everybody, this is a mess because right now the cheapest, most powerful models are not coming from American companies. The Chinese are catching up.

Leo Laporte [01:10:24]:
really rapidly and doing some really interesting things. And I think a lot of people who believe that open weight models are the future— we don't want any one company to dominate— are very excited about what China is doing. We don't really, you know, they know there's very few American companies putting out open weight models. Facebook did, but it wasn't very— that was a long time ago.

Andy Ihnatko [01:10:47]:
What about these accusations that Alibaba is basically distilling Western models.

Leo Laporte [01:10:54]:
I think that's BS. Yeah, but that's a, that's a very technical and difficult question. The idea being that Alibaba and ZAI and all these other companies— Moonshot with their Kimi— make these models and then make, as Anthropic has asserted, 24,000 accounts logging into their AI and training post— it's in the post-training, doing the the reinforcement learning using those models, basically sucking the intelligence out of the American models. And I think that's a very American-centric and a BS point of view. And I think what Moonshot has done with very limited resources really comes down to great engineering, not distillation. And I also think there's a lot of evidence— well, first of all, it's a little bit of the pot calling the kettle black, right?

Christina Warren [01:11:41]:
Right.

Leo Laporte [01:11:41]:
For Anthropic to say, well, you stole everything that we stole from them. Those guys.

Christina Warren [01:11:45]:
Exactly.

Leo Laporte [01:11:46]:
Part 1. Part 2, I don't grant that Anthropic's engineering is that much better. You were inside DeepMind. I know you can't probably talk about that, Christina, but— and DeepMind, everybody knows, has the, you know, among the best AI engineers, some of which— some of who actually created this whole—

Christina Warren [01:12:03]:
Yes.

Leo Laporte [01:12:04]:
You know, tensor and, you know, modeling stuff that we use these days, how LLMs work. I don't think there's a monopoly on this kind of engineering talent in the US.

Christina Warren [01:12:15]:
No, I don't think so. I think that the— I think the reason people had those allegations first was probably because there is truth to it, that you probably did have 24,000 or more accounts being created. And I'm sure that's true things, right? And, and that there might have even been a certain amount of, okay, we're going to try to validate some of our own refinements and whatnot, maybe using these other models to go back and forth with and try to get some improvements that way, maybe. Right. But I don't know if that still remains true now. Right. Because this is what's hard about this space is that an assumption that you could have made 6 months ago might not be accurate 6 months later.

Leo Laporte [01:12:47]:
Yeah.

Christina Warren [01:12:48]:
And so I think that's part of it. I think the other part of it, too, is that, yeah, you know, I don't think that there's a monopoly on where this engineering takes place. And in fact, a lot of the best researchers at any of the of American-based frontier labs, whether that's Google, whether that's—

Leo Laporte [01:13:05]:
Many of them are Chinese.

Christina Warren [01:13:06]:
That's what I was going to say. Many of them are from China, right? Many of the PhDs.

Leo Laporte [01:13:10]:
And many of them are probably going back to China because that's probably the place to be right now.

Christina Warren [01:13:15]:
I mean, I think it kind of depends, right? But I just— I don't think there's a monopoly, to your point on that. And I think some of it is a little bit of a cope. That said, I also would not be surprised if some of these companies have done everything they can to try to access as much information as they can and make things as good as possible, even if that means creating tens of thousands of accounts on, you know, a competing platform to do things, so be it. So, I mean, I will say that some of the labs, without kind of going any further than that, some of the labs are much more sensitive about that than others, right? Like, some of the labs have a much more, I think, kind of laissez-faire attitude towards it and some of the labs are much more sensitive to it. So who knows?

Leo Laporte [01:14:02]:
I think it's completely reasonable. If you look at the guy behind Moonshot, it's completely reasonable to think that these guys have interestingly faced with these hardware constraints they couldn't get Nvidia chips, really done some amazing stuff, very good stuff that I don't think— I don't think it's fair to blame.

Christina Warren [01:14:19]:
I don't either.

Leo Laporte [01:14:20]:
I think that's a very— we're you know, we're number one, USA, USA kind of attitude toward the whole thing.

Christina Warren [01:14:26]:
I think the chip question is more interesting to think about. Okay. Did they really do this without any Nvidia chips? Right. That, that is to me a much more interesting question just because we know that plenty of people have access to the chips that aren't supposed to. And, and how you prove that, I think, is difficult. Right.

Leo Laporte [01:14:42]:
And Huawei chips are getting better. And in fact, ZAI is spinning up an entire Huawei-based data center.

Christina Warren [01:14:48]:
Well, yes, but that all takes time. But that's not here yet, right? Like, and these will take years to build up and they're getting better, but they're not equivalent.

Leo Laporte [01:14:55]:
So I think it's good, by the way. I don't think Nvidia should have a monopoly.

Christina Warren [01:14:59]:
I agree. I fully agree. Granted, we will never in this country be able to have any Huawei equipment ever.

Andy Ihnatko [01:15:06]:
Never again.

Christina Warren [01:15:06]:
Ever, ever.

Leo Laporte [01:15:07]:
They're on the— they're on the banned list.

Christina Warren [01:15:09]:
Well, and it's now been 3 administrations, right? Like the first Trump administration, Biden administration, second Trump administration. That's never coming back. And we're not the only ones. by the way, like Huawei and ZTE are banned by a lot of governments. So for the networking and other things. So I doubt we'll ever see that. But that I would have more question about. I think about like, okay, how did you do this from a hardware standpoint more than, you know, whether you were doing post-training on other closed models or not?

Leo Laporte [01:15:37]:
Yeah. Well, it would be— I would say it would be foolish for the US frontier companies to assume that that's why they're so good.

Andy Ihnatko [01:15:47]:
That's a good point. Yeah, it's not that, oh gosh, they don't have the resources, they don't have the firepower, they don't have the government support. They've got the—

Leo Laporte [01:15:57]:
they must be stealing it.

Andy Ihnatko [01:15:58]:
Yeah, exactly. They have the machinery to build. They don't just have the thing. That's— yeah, that's to be confronted.

Leo Laporte [01:16:04]:
Now there is an interesting new model that Apple is actually talking to right now, Prism ML. It's an AI startup. They've released something they call Bonsai. It's a model small enough to run on your iPhone.

Christina Warren [01:16:23]:
Woo!

Leo Laporte [01:16:24]:
Hmm. And in fact, Apple has been very interested. Apparently Prism ML is now meeting with Apple about, quote, ways it could use its technology. This comes from from the information. Now, I haven't played with the model. Have you, Christina? It just came out. I have not.

Christina Warren [01:16:41]:
It just came out. I have not. And I'm always hesitant anytime anybody's like openly been like, oh yeah, we're in talks with Apple. I'm like, are you? Are you? Are you still? Are you still?

Leo Laporte [01:16:50]:
Yeah. It's also highly quantized, I think.

Christina Warren [01:16:53]:
I think it's a 1-bit model. Yes. Well, I was going to say that that's the biggest thing is that to be able to get it to fit on an iPhone, you have to highly quantize it, which is what what Google does with Gemma, and that's what, like, Unsloth does with various versions of various open weight models. And you can get a lot of power out of that, but there are imperfections.

Leo Laporte [01:17:11]:
Explain for our audience what quantising is.

Christina Warren [01:17:15]:
So basically, you're compressing how much information density there would be. And so, like, a 4-bit quantised model is going to take up much more space and have much be much more information dense.

Leo Laporte [01:17:28]:
How many bits are the— is a normal model? Is it 32 bits? Is it 8 bits? Is it 16? I think I have an FP8 model, which is 8-bit.

Christina Warren [01:17:38]:
Yeah, I was gonna say, I was gonna say, I was gonna say, I think it varies from, from model to model at this point, right?

Leo Laporte [01:17:43]:
But so actually, one way to describe it that might be familiar to you is music. When you sample music—

Christina Warren [01:17:50]:
That's exactly it. It's, it's like, it's like an MP3, right? It's like, it's like compressed compressing the file in an AAC container or an MP3 container. And so it might be, you know, lossy, but it might not be noticeable to your ears. That said, there are going to be specific scenarios where you're not going to be able to get the same kind of impact, right? So, so from a speed perspective, that's one of the reasons why you do it for size and speed. And it might be good in certain scenarios, right? So like, that's why you can have these super, you know, quantized models that will fit on a device and will be very good for asking, you know, basic kind of facts and information, um, and maybe somewhat do tool calling so you could get to a web search or get more up-to-date information, right?

Leo Laporte [01:18:29]:
Yeah, that's an important point. I think people assume, especially if you've used ChatGPT or Claude or, or, um, you know, one of the other models, they assume that what you want is the smartest thing that knows as much as possible. But in fact, a lot of what, especially in agentic AI, What your phone will do isn't really thinking. It's calling tools.

Christina Warren [01:18:49]:
Exactly. And that's the switch, right? Because if we looked back 18 months ago, that was not really the case. That's when it was— we were very much obsessed with when was the training date on this data? Like, how up to date is it? How fresh is it? How much are these things that we need to pull? Whereas now many of the things will just be a tool call to go to, you know, a search provider of your choice.

Leo Laporte [01:19:11]:
Almost anything can do that. I mean, you don't have to be super smart.

Christina Warren [01:19:15]:
No, no, no. I mean, Google had search groundings, which is like their term for it, in Gemma 3, right? Which was—

Leo Laporte [01:19:21]:
Gemma was highly compact. In fact, Google had a paper about a new quantization technique that apparently they said was very good. And that's where Gemma came from, right?

Christina Warren [01:19:30]:
Precisely. And Gemma distills and quantizes down Gemini, right? With some other things removed. And so you can you can get a lot out of it. But I think, yeah, to your point, it's important for people to know, what is it good for? And what is it not good for? Right? And like, what it's not going to be good for is doing agentic coding stuff, right? Like, that's, that's not where we're at, right?

Andy Ihnatko [01:19:48]:
No.

Christina Warren [01:19:48]:
But in terms of, could I either have something that makes a good tool call to get me the latest weather information? Or could it control potentially something on my device in a certain way? Sure. Could it give me, you know, basic kind of grammar or speech to text or translation material? Yes, right. The other thing too is that when you have these smaller models, which again, you're going to be losing some resolution when the more you quantize them. The upside of that is that you could have a model that would be specifically trained for very specific tasks. And that might work really well for that task. And but I think it's just kind of people have to think about it in their heads a little bit differently and go, okay, maybe I use this model for this task and this for another. And this is where in an ideal world you would have at like an operating system level, um, a, a control plane, which would know what model to route each, each task to.

Leo Laporte [01:20:39]:
Yeah. Uh, one of our club members, Nightscape, has actually installed Bonsai on his phone. It fits, it's, uh, 6 gigs. So, uh, it would fit in an iPhone with 12 gigs of RAM. It would take up a lot of, a lot of that RAM, but it still, uh, would fit. And that's why people are very interested in it. But a 1-bit quantization is usually pretty stupid. It says— I mean, think about it with music.

Leo Laporte [01:21:02]:
If you had 16-bit samples compared to 8-bit samples in the music, the music wouldn't sound as good. 16-bit is CD quality, 8-bit is not.

Christina Warren [01:21:13]:
Right.

Leo Laporte [01:21:13]:
And imagine 1-bit music.

Christina Warren [01:21:15]:
Well, I think it's more like trying to figure out, like, is this 128 kilobits you know, which for, you know, if it's encoded in some way—

Leo Laporte [01:21:22]:
Like tokens per second, right?

Christina Warren [01:21:23]:
Right. Like, like, like, but just in music terms, like it was 128K, you know, like, like encoded MP3 file, right? With, with like a V0 or whatever the profile would be, or like a 256K like AAC file, right? Like in that case, yes, the 256K AAC file is lossy and it does not sound as good as a 24-bit or a 16-bit, you know, like lossless, like, like FLAC or Apple, you know, audio file. you know, like Apple lossless file. But are you actually going to be able to tell the difference, especially if you're listening to it in your Bluetooth headphones?

Leo Laporte [01:21:54]:
Right.

Christina Warren [01:21:54]:
Honestly, you're probably not. Most, most people cannot tell the difference once it's about, you know, 128, 256K, usually 256, 320K. Usually you can't tell the difference between that and the lossless stuff. That said, it's different for the models. And, and this we just don't know because it's possible and who knows? I'm sure that they— what they seem to be saying is that they've created a better type of quantization effort so they can be more efficient so that, you know, their lossiness is better.

Leo Laporte [01:22:26]:
Not as lossy.

Christina Warren [01:22:27]:
Exactly right. And so—

Leo Laporte [01:22:29]:
AAC versus MP3.

Christina Warren [01:22:30]:
Precisely. And so that's, I think, what we just have to kind of like look at. And I'm sure that there are a lot of companies, including Apple. I'd be shocked if Apple weren't looking at those types of things themselves too, because again, like they have their own local models, you know, they've developed, they're putting on device and you want to be able to, especially in these RAM constrained times, be as efficient as you can be with that.

Leo Laporte [01:22:51]:
And I'm sure a lot of effort's being put into that.

Christina Warren [01:22:54]:
And then also again, you might have multiple models, right? Like that might be the way you get around this.

Leo Laporte [01:22:58]:
Well, I do.

Christina Warren [01:22:59]:
Trying to have like, right. But I'm saying Apple might, right? Rather than having like one big model of of experts, you might have multiple models they would have that have different parameter sizes that are on the device that are being called and orchestrated by the OS.

Leo Laporte [01:23:12]:
I was talking about this before the show. I, I use, as you do, a Framework, 128GB Framework, as my main AI local AI machine. But that's mostly frontier models with my agent Hermes. But I realized last night, I thought, I have this Mac Mini, 64GB Mac Mini M4 Pro, I haven't used much in a while. I could probably use that as a local inference machine. And so I set that up and it's running local models. In fact, I wouldn't be surprised if at some point it's running this Bonsai model, but it is running local models that fit into 64 gigs. But my main agent, Hermes, on the framework can call it.

Leo Laporte [01:23:48]:
It said, yeah, if you want something private so we don't send it out to a frontier model, or if you have, let's say you have 50 documents you want me to absorb, It'd be easier. It's not going to be as fast, but who cares? I'm going to offload it to the Mac Mini, let it do it. Now, I should point out one other thing germane to our earlier conversation. The Bonsai model is a highly quantized version of Quen, the Chinese open weight model. And the reason they can do this is it's an open weight model.

Christina Warren [01:24:16]:
Mm-hmm.

Leo Laporte [01:24:18]:
And so they could just suck it down, get it from Hugging Face or whatever, and, uh, and run their quantization technology on it. So, uh, this is why I'm very excited about OpenWeight. Uh, I think that's, that's the future. And we'll, we'll, I'll check out Bonsai. Uh, right. I probably don't want the 1-bit, but when they do a full 8-bit model, say, um, I'll probably run it on my Mac.

Christina Warren [01:24:40]:
Although, I mean, honestly, it'd be interesting to see, right? Like, I, I, I would be interested in being able to play with this, you know, on my phone in some way. If you could like have access to it just as—

Jason Snell [01:24:48]:
Yeah.

Leo Laporte [01:24:48]:
I want Nightscape to tell us how it's done.

Christina Warren [01:24:49]:
I was gonna say, I was gonna say, Nightscape, Nightscape, you let us know if it's worth playing with or not because, uh, because honestly it, it could be compelling, right? Especially for local on-device AI stuff. Um, that could be interesting.

Leo Laporte [01:25:01]:
It could be very interesting. Yeah. And, and this is a very exciting area right now. It's not only quantization, but the, the mixture of experts technology where you only load a little slice of it into RAM. You keep most of it on the hard drive so you can do it in a smaller RAM. There's a lot of techniques that are being used right now. In fact, most of the current local models, I think, are MOE, mixture of expert models. I know that the Quen I'm using is.

Leo Laporte [01:25:25]:
It's really an exciting time, actually.

Christina Warren [01:25:28]:
Yeah. Bonsai claims that in their 27 billion parameter model reaches up to 163 tokens a second in 1-bit and 134 tokens a second in ternary.

Leo Laporte [01:25:38]:
That's super fast.

Christina Warren [01:25:39]:
Yeah. And that's on a— now that's on a 5090. Okay. So that's if you've got— but, but, but But, but, but, but, but this is interesting. On an M5 Max, it reaches up to 87 tokens a second in 1-bit and 58 tokens a second in ternary. So that is actually impressive that—

Leo Laporte [01:25:56]:
So why is tokens— tell me, tell me, because I think that's not how fast it's thinking, or is it? No.

Christina Warren [01:26:02]:
Well, I think it's—

Leo Laporte [01:26:03]:
How fast it's typing.

Christina Warren [01:26:05]:
It's how fast it's giving you the output, right? And so, so, so, so your, your blocker there is usually memory bandwidth. And that's why Macs are so good and so much better than other computers because they have the integrated, you know, memory and they have very high memory bandwidth even at the low levels, right?

Leo Laporte [01:26:21]:
Like using Ollama that supports MLX, the Macintosh version of this CUDA core stuff that Nvidia does. And MLX with models that are tuned for it is very fast.

Christina Warren [01:26:32]:
Very—

Leo Laporte [01:26:32]:
yes, it really works beautifully. And of course, all that extra memory bandwidth that the Macs have because of unified memory. This is right. This is almost 5090 performance level.

Christina Warren [01:26:41]:
Right, right. I mean, I mean, you know, it's not as— it's not as fast, but considering how expensive it is to get a 5090 and that you could do this not on a MacBook Neo, to be clear, the memory bandwidth on those machines, because it's a phone, are right, are not good at all. And it's less on a MacBook Air. But if you've got a MacBook Pro, especially M4 Pro, just sitting around, right? I was going to say, if you've got it, if you've got an M4 Pro, if you have some of the older Macs, machines, even like the memory bandwidth on those is such that you can get really good performance. And, you know, for people who don't have the space or the time or the inclination to, or, you know, the extra $2,000 or whatever it is to get a 5090, it's, it's a compelling option to at least experiment with these things. Is it going to be something that could be, you know, for most people replaced what they're using hosted models for?

Andy Ihnatko [01:27:24]:
No.

Christina Warren [01:27:25]:
But it's still You know, very, very viable and certainly much more viable than it was even, you know, 2 years ago, thanks to the explosion of MLX, these improved quantization efforts, the improvements in the open weight models. So it's, it's nice to see.

Leo Laporte [01:27:41]:
Yeah. Let's see, what else shall we talk about? I'm sorry about all the AI talk. I get emails now from people saying, why are you guys talking about AI? Why do we talk about AI all the time? Christina, tell me why.

Christina Warren [01:27:54]:
Because it's— I mean, one of the stories we have here is actually from The New York Times talking about, like, the rise in number of apps in the App Store. And Vibe put it out.

Leo Laporte [01:28:02]:
It's doubled.

Christina Warren [01:28:03]:
It's doubled. And I'm so sorry, but this is one of those things that is defining— it's why our hardware prices are going up. It's, you know, it's defining, for better or worse, the direction of how software is being developed and also who's developing it. So I think it's It's one of those topics. I totally get it, people not wanting to be interested in it, but it affects, I think, most of us whether we want it to or not.

Andy Ihnatko [01:28:27]:
And it's also going to affect Apple. I mean, their choices that they lost 2 or 3 years by choosing the wrong horse to begin with. And now every choice they make today is going to affect what they can do 5 years from now, 10 years from now. They don't have— they're not in a desperate situation by any means, but they can't afford to— they have to have a good plan and keep executing on that plan time and time again. I mean, one of the things that this story kind of made me think of, and we should qualify that this was an interview that the CEO of this startup did on CNBC, which he said, oh gosh, yes, Apple's looking at our stuff, everybody's looking at our stuff. So for a startup that really wants to get the next round of funding, that's one of the safest things to say. That's what they say, yeah. But it made me think about like, are they now at the point where even if someone came at them with, hey, we've got this model that's really, really great and perfect for what you want on the iPhone.

Andy Ihnatko [01:29:17]:
Would they even be capable of saying, okay, we can— we're going to switch to this now? Or would this sort of technology be, well, we are going to buy a company so that we can hire this expertise and access to this technique and use it, and we won't see effects from this for the next 2 or 3 years? I mean, is it even possible to make that kind of a dramatic shift at this point? Or is it— is the machinery now, their mechanism, their infrastructure for evaluating new techniques and building and iterating and improvement, much more important than any one status of whatever model they have right now.

Leo Laporte [01:29:48]:
Well, I'll give you an example. Apple has released a speech-to-text engine, speech analyzer, and I'm looking at benchmarks from Inscribe that show it to be fast. I've been using Whisper, which is really excellent. I dictate to my agent all the time.

Christina Warren [01:30:05]:
Yeah, Parity's good too.

Leo Laporte [01:30:07]:
Pardon me?

Christina Warren [01:30:07]:
Parakeet.

Leo Laporte [01:30:08]:
Parakeet's very good. It's faster than Parakeet. It's faster than Whisper. It is super fast.

Christina Warren [01:30:14]:
That's awesome.

Leo Laporte [01:30:15]:
Yeah, both on iOS and macOS. Boy, I'd love to see better dictation with Siri.

Christina Warren [01:30:21]:
I was gonna say, like, I think that if you— whatever they replace it with, if they just had, you know, better support, because I— if I could plug in Whisper for what Siri uses right now—

Leo Laporte [01:30:30]:
I love Whisper.

Christina Warren [01:30:31]:
It would be a lot better. Now Now, granted, it's going to vary language to language, country to country, but—

Leo Laporte [01:30:36]:
Yeah, Whisper covers more languages, which is a big deal, but Apple doesn't really care that much because of the whole issue with the EU. Um, so—

Andy Ihnatko [01:30:45]:
Well, they're gonna have to though.

Leo Laporte [01:30:47]:
Yeah, yeah, at some point. Yeah.

Andy Ihnatko [01:30:49]:
This is, this is, this is kind of like the Apple Maps versus Google Maps sort of thing, where getting 90% there is— it's not easy, but it's very, very attainable. And every single step But beyond that 90%, the finer points become more and more difficult to nail, which is why there's always, uh, which is why in Apple Maps it's 90% as good as Google Maps, except for the times when it is absolutely of no use whatsoever because you need to know exactly where the entrance to this building is as opposed to where the address is. And language translation and speech-to-text is one of those things where it becomes, we have all the languages. Not only that, but we know every regional accent, so we are not going to be thrown off by how a certain word is pronounced 3 different ways in this— within the same country.

Leo Laporte [01:31:32]:
So I should mention, Inscribe is a company that does transcription summarization, uh, using AI. They bill themselves as powered by Apple Intelligence. And one of the reasons they've been benchmarking Speech Analyzer is, should we use this? And they've said, for English, we are going to use Speech Analyzer. It's by far the fastest. For any other language, we'll still use Whisper. which is an OpenAI model. So don't, don't think Apple is, is, uh, is out of the running here. They're doing some really good, uh, stuff.

Leo Laporte [01:32:02]:
They're also using AI, maybe even speech analyzer, at the Genius Bar. They are now recording at some of their retail locations the Genius Bar sessions. This is making some people unhappy. This comes from Mark Gurman's Power On newsletter. The system lets Apple Store employees record sessions with an artificial intelligence tool that transcribes and summarizes their conversations with customers. I have to say, my doctor does that.

Christina Warren [01:32:29]:
I was going to say, my doctor does this too. So, so I don't love it. But as long as they are actually reliably reviewing the notes after our time together, I'm okay with that.

Leo Laporte [01:32:40]:
Yeah, it's hard to take notes when you're typing and, you know, and talking to a customer. So I think it's It's a good thing to have.

Christina Warren [01:32:45]:
My only concern would be, I mean, this is the same with my doctor, is, okay, are we actually going to be putting in the hygiene that they will actually review the notes and approve them and not just accept what the summary said? Right. Because I see this with colleagues all the time where people, you join a meeting and you have a million voice meeting recorders going on, and then you see the transcript of how it went, and sometimes it's right and then sometimes it's completely off. And I'm like, okay, if I really hope that if somebody, you know, wanted to rely on, on these meeting notes, that they actually took a look at it after the meeting and didn't just rely on the transcripts.

Leo Laporte [01:33:21]:
Well, employees, according to Gurman, are also concerned that the recordings could be used to coach workers, replace workers, or during performance reviews. But Apple says, no, no, we're not going to save the transcripts and managers aren't going to get them. And employees can edit the transcript and summary before saving them to the Genius Bar system. So, so I think that they are good protections for the employees. Well, by the way, the customer also has to agree.

Andy Ihnatko [01:33:46]:
If Apple had a better track record with labor in the Apple Stores— I'm not, I'm not, I'm not doubting it. I'm saying that they don't— they haven't earned that level of trust yet. So they're right to ask these questions.

Leo Laporte [01:34:00]:
And as long as we're talking about Gurman's Power On newsletter, I I do have to— this is kind of a sad Mac story. Apple had bigger plans for the Mac Pro before discontinuing it. Of course, you know, I think there may be some question about whether there was a market for the Mac Pro, but they were even thinking about Intel.

Jason Snell [01:34:21]:
This shows you that they had— I think that there were people inside Apple who were so obsessed with the Mac Pro customers that they had, the very tiny number of professional usually in, uh, in specific industries, including the entertainment industry sites. And I, I think I— when I read this story, I just smell a, a sales partner kind of relationship where they're like, oh, don't you know, we, we sold 1,000 of these to editing bays or something like that.

Leo Laporte [01:34:52]:
Right.

Jason Snell [01:34:52]:
And they're making the case of like, oh, and they still have Intel stuff, so we need to still make the Intel one. Can we bring that back? And it, I mean, I read it as being somebody made the case that the tiny, tiny Mac Pro buyer base was worth serving. And somebody else did the numbers and said they're not.

Christina Warren [01:35:11]:
No, no.

Leo Laporte [01:35:12]:
That's a lot of development you have to do and prototyping. And so there were 2 product numbers according to Gurman, J170, J190. The former was Intel-based to address certain uses.

Jason Snell [01:35:25]:
Yeah. Again, Some customer somewhere.

Leo Laporte [01:35:29]:
You know, what was it for a long time? What was the music app that everybody used? Was only Intel. They didn't— they never— I don't know if they have now.

Jason Snell [01:35:36]:
So, I mean, this is the thing, right? Is— and we've talked about it here a bunch, which is it sucks to be in a market that is abandoned by a hardware maker, a software maker, whatever. It sucks to be you, right? Like, and we've all been there, I think, at one time or another. But other than an appeal to history and like Apple was— yes, Apple was once kept alive by high-end creative markets. And I think you can see the existence of the Mac Pro as long as it existed as being Apple still acting as if the Mac and Apple existed because of those markets. But it's been decades Now, since then, and when I see a story like this, that's what I keep thinking is there's people inside Apple who are like, "But what about the high-end creative markets?" And finally, somebody was like, "No, we're not going to make a chip for them. We're not going to make a Mac Pro for them." Like there are knowing that it's the highest of the high-end. It's the nichest of the niche markets for them. And if you're in that market, you're really sad that they've abandoned you.

Jason Snell [01:36:48]:
But it's very hard not to look at the use case here and say there's no reason to spend that kind of money on custom building hardware for this teeny, teeny, teeny tiny market when you're Apple and you no longer rely on teeny tiny markets to make enormous amounts of money. And by the way, also this AI thing is going to come along and that's your new high-end market that you need, right? Jason?

Andy Ihnatko [01:37:10]:
Yeah.

Christina Warren [01:37:11]:
Yeah. I mean, Jason, you and I were both 27-inch iMac devotees. And we were kind of— we were kind of caught by that too. Right. And obviously, like, got my iMac Pro, right? Yeah.

Jason Snell [01:37:21]:
One and done.

Christina Warren [01:37:22]:
I have my, my 27-inch 2020 iMac that I bought even though they'd already announced the Apple Silicon. I was like, I'm going to buy this anyway and drive it to the end. So, yeah, it sucks, as you said, to be like left behind. But yeah, I agree with you. I think that this was probably one of those things where there were people who were making the case, but Come on.

Jason Snell [01:37:41]:
And the sales— I mean, like, if you're on the sales team that's selling to the high-end audio studio vendor and you know that, like, you get a commission maybe and you've got your— you're trying to advocate for your— I get it, right?

Christina Warren [01:37:54]:
Of course.

Jason Snell [01:37:55]:
I get that. But what happens in a lot of cases is these sales organizations distort policy for a while because they're like, what do you mean you're going to not do this thing? We need the money. And, and also you end up with a cultural cachet of like, well, we got to take care of our pro creative users because they're so key to us. And then that lives on, you know, a long time sometimes after that market is no longer relevant. And it's hard to say goodbye. Look, it's hard, I think, even for people inside a company to say we are no longer going to focus on you. Like, it's hard for everybody. Yeah, but, but like I said, I think you could argue that that Mac Pro should have gotten killed 10 years earlier, right? But they kept it around and they kept— for, I think, a lot of legacy reasons that, that became— by the time they killed the Mac Pro, by the time that they even did that first Apple Silicon Mac Pro, yeah, it was clear to everybody that it was irrelevant.

Christina Warren [01:38:49]:
Yeah, I, I doubt they even sold out their initial run of that Apple Silicon Mac Pro. Like, I'd be shocked.

Jason Snell [01:38:55]:
That's why, you know how they didn't announce that it was discontinued. I think literally they were just sitting there not— because they didn't update it. I think they're just drawing down the warehouse. I think, Christina, I think you're— I think that's exactly the evidence there is that they just kept it on the price list until they were running out of them. Yeah, right. Yeah. And then it's out.

Leo Laporte [01:39:17]:
But that's one of Apple's real strengths is their willingness to abandon legacy. Uh, you know, you compare it to Microsoft and yeah. They're really good at that, and that's really where they've—

Jason Snell [01:39:26]:
But it stinks if you're the one getting abandoned.

Leo Laporte [01:39:29]:
But yeah, yeah, and you're right, the sales organizations are torn because they have loyalty to their customers. That's what they're there for. I understand that. I, I do. I, I feel our— I feel your pain.

Jason Snell [01:39:39]:
Yeah.

Leo Laporte [01:39:40]:
You're watching MacBreak Weekly. Andy Inaco, Jason Snell, Christina Warren. We're so glad you're here. Thanks to all our club members. It's time for the Vision Pro segment. It's gonna be very short today. I had to go to Germany to get the Vision Pro story of the week. The one—

Andy Ihnatko [01:40:02]:
Love that commitment.

Leo Laporte [01:40:03]:
From IT-Botweis. Vision OS beta 90Hz freeze by 24p. I don't know what it means. Let's— thank goodness there's Google Translate. Uh, I still don't know what it means. VisionOS 27 beta has a 90Hz freeze at 24p. Do you know what that means, Jason Snell? Do you have the new Vision Pro OS beta?

Jason Snell [01:40:26]:
Is this literally somebody complaining about a bug in the developer beta?

Leo Laporte [01:40:30]:
It's VisionOS beta reports of jerkback. That's a good translation. Jerkback. I think they mean, you know, jumping frames. In the jerkpack, in the video playback accumulate, the refresh rate should then remain rigid at 90 Hz. I don't know what it means. It's— I think this is a mistranslation in which film-typical Jews become visible.

Jason Snell [01:40:58]:
Oh no.

Leo Laporte [01:40:59]:
Oh no, maybe this is the wrong site. This is I Am Sorry. What the hell is going on?

Jason Snell [01:41:07]:
It's, it's, uh, Germans are Germans are finding bugs in developer betas is what that's doing.

Leo Laporte [01:41:13]:
Yeah, let's just leave it at that. And that is your Vision Pro segment. I should have read the translation before I did the story, I guess.

Christina Warren [01:41:25]:
Hey, hey, Nightscape in the chat has been working on creating a local—

Leo Laporte [01:41:29]:
On Bonsai?

Christina Warren [01:41:31]:
Yeah, he's actually, I don't know if Bonsai, but he's working on doing a local-first way to to code Vision Pro apps. So, hey.

Leo Laporte [01:41:38]:
Oh, Nightscape, rock on.

Christina Warren [01:41:41]:
I was going to say you should productize that Nightscape, like for, for all 5 Vision Pro developers. Yeah, but I'm not even joking. That'd be cool.

Jason Snell [01:41:48]:
You know, that was, that was actually a Mark Gurman rumor a while ago is one of the things that they were working on for Vision Pro was the ability to, to write, to basically to vibe code. He had to explain what it was. We didn't even have the term yet. And I think some of that is what's in the type-to- make a shortcut, to describe a shortcut feature in 27. Um, but yeah, I mean, sure, the one way to deal with the limitations of the apps and user base on the Vision Pro would be to vibe code your own Vision Pro app.

Leo Laporte [01:42:17]:
Well, vibe coding has jammed the App Store. Twice as many submissions this year as previous years. Uh, the New York Times says vibe-coded apps are flooding the App Store. Is that for Apple. They highlight one particular developer, a developer in Chicago, Marco Perez, who had never made a mobile app before. He spent 5 weeks tinkering with vibe coding AI tools to create a prototype for Info Drizzle, which the Times says has been modestly successful— $500 in profit, 1,300 downloads. But you know what, that's not bad. But then in May, he released Stampa, which was a scrapbooking app that turns photos into digital postage stamps.

Leo Laporte [01:42:58]:
And now $3,000 in profit, 20,000 downloads. I think there are a lot of people who would be very happy with that. And the fact that Apple's approving them means Apple's happy to get those, right? They don't have to approve those.

Andy Ihnatko [01:43:12]:
Mostly, I think the problem— it would be a bigger problem if Apple decides that they haven't— they want to put an editorial control over what deserves to— what does— what deserves to be in the App Store, what does not, beyond a simple threshold of, is this a value? Does this create value for the customer or does it not? We don't want to— A, I think it would be correct for them to say, okay, look, thank you for submitting the 1,000th Markdown editor. Today you get a free hat, but we're not allowing you in the App Store. But on the other hand, if someone basically vibecoded a tool that does one simple thing because they run a bowling alley and it frustrates them that this task of running the bowling alley is so difficult, Well, they shouldn't be saying that this app shouldn't exist. They should make it available. If anything, this demonstrates Apple's— the App Store's shortcoming for discovery. They should be adding more tools that allow customers to find the app they want. And when the search result gives them 12 different candidates to basically sift through them and figure out, I feel as though this one— I wanna try out these 2 and this one will probably have the most value to me, which is something that I don't think the App Store really does very well right now.

Leo Laporte [01:44:22]:
Um, a spokesperson for Apple— I'm sorry, I'm German. We are thrilled to see a new generation of developers embracing the latest tools to build and ship apps faster than ever. And then to the accusation that Apple only cares if you get downloads because that's the way they make money, all apps are held to the same high standards for quality, privacy, and security that users have come to trust. So I guess that means, uh, Apple says okay.

Christina Warren [01:44:50]:
Well, I mean, I think this is also good for them because I'm like, the number of conversations I've had with people, um, one of my good friends is a longtime Windows user and Windows developer, and because she's been vibe coding so much, she's been, you know, building mobile apps. And then she's like, okay, well, I built this app and it's not even native, it's, it's probably using React or something, but, you know, she wants to publish it in the App Store. She's like, how do I publish in the App Store without a Mac? And I'm like, okay, well, you can But it takes a lot of work. And it's easier if you just, you know, buy a MacBook Air and/or buy a new Mac Mini.

Leo Laporte [01:45:18]:
Or a Neo, right?

Christina Warren [01:45:19]:
Or a Neo, right? I'm, you know, I'm just thinking like—

Leo Laporte [01:45:22]:
Could you do it with a phone? Can you submit to the phone?

Christina Warren [01:45:24]:
No, no. And that's actually the one area where I get a little bit like frustrated with Apple of their position on vibecoding is that the apps themselves that will let you vibecode on a phone and then submit like they're like, oh, no, no, we can't use our phone or iPads like their computers even though those chips are now inside computers that we sell. The difference in the operating system is so vast, you guys, we can't possibly let you use your tools to create things on the platform that you're going to consume it on. But, you know, but a lot of people I think who are— maybe they're experienced developers, maybe they are people who are brand new at it, like it can open up— you've had an idea for an application for a long time, something that just fits your to Andy's point, now I can do this in a much more manageable way that's not going to take all of my time. And so I'm going to need to buy that $100 Apple developer subscription. I'm going to need to make sure that I have a Mac that can, you know, submit and test my apps. That brings people further into the ecosystem, which I think is only a good thing. So—

Andy Ihnatko [01:46:24]:
Yeah, 100%. More points of view always, always make things better. There are people with all kinds of backgrounds who A generation ago were absolutely shut out because they didn't have access to the tools. It cost a lot of money to do this. It cost a lot of years of investment of time to figure out the technology. But we were feeling— we're hearing about the need— we're seeing classes of apps that address needs from communities that have absolutely never been addressed before, and that should be absolutely encouraged. It's one of the great things about modern development.

Leo Laporte [01:46:54]:
Well, okay. It also makes it hard to find stuff. in the App Store.

Andy Ihnatko [01:46:57]:
And I imagine that's a fault of the App Store.

Jason Snell [01:46:59]:
Yeah, this, this is all, all these questions about like, oh, what does this mean for Apple's approval process and TestFlight and App Store search?

Leo Laporte [01:47:07]:
Jam them up.

Jason Snell [01:47:08]:
And the answer is this is an opportunity for Apple, but they're going to have to step up and they're going to have to do better. And that's a problem because a lot of these areas are areas where Apple has been, shall we say, a little lax in trying to improve the infrastructure. They took the App Store success for granted. And it's hard to make changes when you're just making money hand over fist. But like, they're— they need to do more work on App Store approvals. They need to do more work on search and discovery and be better at all of that because those systems are not very good.

Andy Ihnatko [01:47:42]:
Yeah, I, I've always felt like the entire app development stack, some development relations to, uh, to the store, that's the most un-Apple thing, maybe the only truly un-Apple thing about Apple's business. I don't think they run that at the same level that they tend to invest in pretty much anything else. It always surprises me.

Leo Laporte [01:48:04]:
This is interesting. In the New York Times story, it has estimates by AppFigures that Apple collected $34.21 billion last year in revenue from the $114 billion in in-app purchases. If that's accurate, And I have no reason to believe it's not. That's astounding.

Jason Snell [01:48:24]:
That's where the, that's where the money is. Yeah, that New York Times story is very funny because they're there because the New York Times is very concerned. One of their, one of their concerns was that the vibe coding apps were not going to be as profitable for Apple because it's harder to do in-app purchase, and that's where they make all their real money. And a lot of these vibe-coded apps are kind of for chump change, and Apple's not going to make a profit on them. And I mean, but it is true, if we're saying Apple should invest a lot more money in making the App Store better. And they're like, but wait, it's because of this influx of apps that doesn't make us any money. I could see that that's a problem. My solution, again, it's a classic.

Jason Snell [01:48:59]:
This happens a lot when people talk about Apple. There are people out there who believe that Apple can do no wrong. They're amazing. They're the most incredible company ever. Don't stop them. But if there's a thing that Apple doesn't do, they're like, well, that's impossible. Obviously Apple could never. And it's like, well, wait a second, is Apple great or are they incapable of solving even moderately hard problems.

Jason Snell [01:49:18]:
Because I would look at this and say, hey Apple, here's an opportunity to make your cockamamie in-app purchase system work better for developers. So it's easier to vibe code an in-app purchase subscription. How about that as an idea? Maybe that would be a good thing.

Leo Laporte [01:49:34]:
Is it hard to put that into the—

Jason Snell [01:49:35]:
It's a mess. It's hard to test. It's hard if you talk to developers adding the subscription stuff and then gating features and all of that. It gets tricky and I've never tried it with a vibe-coded app before, but like I would— I know from talking to human developers that it's, it's really tricky. And again, is that not an opportunity for revenue optimization? I think right now Apple's point would be we think that the developers who are building in-app purchase are motivated by the money that they'll make and therefore they will put up with our stuff.

Andy Ihnatko [01:50:07]:
100%.

Jason Snell [01:50:08]:
In the long run, Especially when you're lowering the bar with vibe coders, maybe you actually need to make it easier to use.

Andy Ihnatko [01:50:17]:
Yeah, that's my big— that's why it just frustrates me and surprises me. If there were an area of customers, let's call it like education, like sports, like people in a certain geographical region that's responsible for billions and billions and billions of dollars of Apple revenue, My goodness, they would be treating them like kings and queens and this other stuff. It's very weird to me that the customers, the users who develop apps, who are generating these billions and billions and billions of dollars around, they're not treated with indifference or disdain. But every time they— every time the developers for years have said, gee, we really— it would really help us out if you could add this feature to the App Store, if we could monetize this way, or This API, it really is bugging us, and it really is forcing us to put a lot more effort into this that we would much rather spend time adding the features that our users have been clamouring for for years, but we can't because we don't have the time and the resources because we have to develop it towards make— dealing with these APIs that don't really work very well. Again, developers are their customers as well. They're customers of the App Store, and without them, they're also customers that make them a lot of money and they should be treated a lot better than they're being treated.

Leo Laporte [01:51:30]:
Well, I tell you what, if you're a Brazilian user, you may love the App Store. Download a navigation, travel, or weather app, and as soon as it sees your IP address, it goes, oh hey, by the way, you want to gamble? This is an example from 9to5Mac of Potato Plot Puzzle, which is— it looks like a dumb growing game But as soon as it sees that you're in Brazil, it turns into, convide amigos para entrar, ganhe 66-66, win, win, win, por endaçadão. I don't know, that's— I, I, it's not any language I know.

Jason Snell [01:52:06]:
How about Gigante Internacional?

Andy Ihnatko [01:52:08]:
I'm getting flashbacks.

Leo Laporte [01:52:09]:
All I really know is gold, but that's not useful here. Anyway, uh, that is a very clever way of hiding a gambling app, you know, just check the IP address And then reveal your true nature.

Jason Snell [01:52:20]:
Yeah, this is— I mean, John Gruber has been on this point for a while on Daring Fireball, and I think he's right, which is, look, you can't necessarily catch all these guys, but when you've got like multiple apps in the top 10 in Brazil that are these kinds of apps, Gruber's point is, why is there not an investigation squad inside the App Store that is actively seeking out fraudulent apps that have gotten successful and getting them kicked out of the store.

Leo Laporte [01:52:48]:
9to5Mac says there are 60 apps that do this.

Jason Snell [01:52:51]:
Yeah.

Leo Laporte [01:52:51]:
60.

Jason Snell [01:52:52]:
Yeah. And, and yeah, it's, it's infuriating. And this is that thing where we, you know, we hear these stories about innocent developers getting their apps rejected for all these little things. And then these sham apps just get, you know, get let through and high up on the charts and nobody seems to care. It's, it's bad. I, I, yeah, I want them I want John Gruber's Bunko Squad for the App Store, the people who are just going around looking for fraud, fraud apps and kicking them out. That's, that's the way you do this, right? It can't all be happening in the scan upfront. It's got to— you've got to have some investigators who are actually doing the work to find this stuff and pull it out because— and this is my argument.

Jason Snell [01:53:28]:
You're like, well, maybe they, maybe they don't care. Maybe they like it. Maybe they're making money off of it. Okay. But one of Apple's key arguments whenever it talks about to regulators about the App Store.

Christina Warren [01:53:39]:
Yep.

Jason Snell [01:53:39]:
is that it's curated and that it's safe. You can't have it both ways. And you can't have it both ways.

Christina Warren [01:53:44]:
You can't have it both ways.

Andy Ihnatko [01:53:44]:
Exactly. And this is— I hate to bring this up a 3rd or 4th time, but again, $5 trillion company. They can't say that— they can definitely say that we can't solve this problem, but they have the resources to do whatever it is that they've decided is a priority for this part of their business.

Christina Warren [01:53:59]:
Yeah.

Andy Ihnatko [01:53:59]:
So if this is not happening, then that means that it's not a priority. Exactly. And they shouldn't argue that. They shouldn't argue that their hands are tied.

Christina Warren [01:54:07]:
That's what's always bothered me is this speaking outside of both sides of your mouth, because I'm okay if you say this is not a priority and we are not going to police this. I'm okay if that's the statement that you say. But if you say that, then you need to also not say, oh, but the App Store has to be the only place you buy apps because only, only we can protect you and only we are the ones who can sanction this. Right. Like, you can't make both arguments. I will say, and I I fully agree with both Jason and John Gruber. Like, I would like to have a bunco squad and have people who are empowered to look at trends and analysis and investigate apps and say, hey, this looks kind of fishy. And, and also maybe even do more about banning developers once they've been kicked out and prevent them from coming back again, because we know that they come back time and time again.

Christina Warren [01:54:50]:
But I do wonder if, at least for some of this stuff, if some of the automations and things like AI and some of the latest frontier models could improve at least some of these scanning techniques, right? Like, I would not be opposed to that. I would like to see both. I would like to see more humans involved. But I also feel like if you're noticing, you know, patterns like this, this might be something like that you could run through a routine of tests, you know, in your automation scan, that maybe now that's the thing that you add in, which is, okay, we're going to try accessing this from a different IP address, from a different region, and see if we get different results.

Andy Ihnatko [01:55:23]:
Yeah, and this is super important because we can define what addiction means, and of course it's correct and incorrect to say that the phone is the most addictive substance because it is a machine designed to keep us pecking at buttons. But when you add that, when you add an online betting app to that situation where the exact same things that get kids to keep buying in-game purchases because they hit them with the right offer at the right time, When you add to that, oh, it's a betting app that actually costs people real money, this really becomes when the comparison to the tobacco industry becomes very, very, very apt. Apple should be absolutely livid about— excuse me, absolutely defensive about making sure that they don't get this sort of reputation. And it's why every time this kind of story comes up, I remind myself and others, if you go to the official Apple Sports app, they have DraftKings. stuff right in there. They don't— not links directly to the app, but they want to make sure that, hey, here's what the betting odds are, and here's like the— here's the branding, here's the logo of DraftKings, so that you're reminded that, right, if you keep swipe— if you swipe up and left, you will be able to get right back into your betting app. It is, to me, uh, Apple is not— this is one of the few areas in which Apple, I feel, is doing something that is absolutely unethical There is no reason for them to do it. They can't be making so much money off of this or getting so much traction from this app because of this that it is worth being part of a problem that is constantly, constantly being identified as a destroyer of people.

Andy Ihnatko [01:56:57]:
It's just absolutely terrible.

Leo Laporte [01:56:59]:
I am the destroyer of the Brazilian language, and I apologize to all and sundry. I hesitate to tell you this, but the city of San Francisco says Apple should take down at least 8 Nudify apps. They really should. Yeah, and there are— Apple is— the city attorney says stop profiting from 13 face swap apps that are overwhelmingly used to target women and girls. I agree, take them down. And how did they get on there in the first place? Anyway, this is the problem. It's just inconsistent. And by the way, Google also has been—

Jason Snell [01:57:37]:
I mean, I think where, where we all probably agree, I mean, Apple, it's not only strategic that they need to do this, but like, this is where I say, this is where Apple's a cheapskate. Like, Apple makes a lot of money and they built this thing from nothing, but they also didn't really scale it because they felt like they didn't need to. And I think there's a cult— there's some cultural baggage there. Apple also, we've talked about here many times, has issues with hiring. They don't hire people very well. They don't hire a lot of people. They're really reluctant to hire people. And I don't know, like you look at the size and scope of the App Store and I don't know how many people are working there, but I would bet you it sure seems from the outside like it's not enough in these areas.

Jason Snell [01:58:25]:
Like it's just not enough people. And that is a cultural thing at Apple too, which is they're, they're lightly staffed compared to what you'd think they'd be for. And this is where it starts to get to be kind of ugly in areas like this. And, you know, you need leadership who's going to say, yeah, you're right, we're not doing enough in these areas in the App Store. But they also don't invest, you know, their search engine isn't very good and like, but they're happy to put another search ad in there. So, you know, again, it's priorities. This is not even I would even say this is not even a hard problem to solve. It just requires somebody who wants to solve it.

Jason Snell [01:59:00]:
And it's unclear how hard Apple wants to work to solve it.

Andy Ihnatko [01:59:03]:
Yeah, Apple, Apple is— multiple times over the past 2 or 3 months have been hit with Russian banking apps that, of course, they can't host these apps in the Russian App Store because of sanctions. But hey, why don't we just call this a messaging app or why don't we just call this a pocket calculator? And it gets through and then a signal gets sent after it gets approved for the App Store that turns it into a banking app. Like, how did that get through? How did that get approved? It's a— they have a holistic fundamental problem with the— with app reviews that they absolutely need to address. And like you said, it's not that they don't care about it, it's that they clearly have not decided— the right person has not jumped up and down on the right table in the right office to say, we, we need the resource, we need to spend the resources and time to finally solve this. Whatever we do, just wherever we thought was solving this them 5 or 10 years ago ain't cutting it.

Leo Laporte [01:59:52]:
All right, let's take a little break. You're watching MacBreak Weekly. Jason Snell is here, sixcolors.com. Have you recorded your first history podcast yet, or are you—

Jason Snell [02:00:02]:
Yeah, we've recorded. We did a recording last week, so we're on the— in the process of recording our series and all that. And doing it in advance is kind of nice. So yeah.

Leo Laporte [02:00:13]:
Where can people find it if they want?

Jason Snell [02:00:15]:
Yeah, check it out at designed And right now that'll go to the Kickstarter page where you can, you can become a member there. And then over time that will turn into the show. We're gonna launch the RSS feed in August sometime with the episodes previously posted. And then I think in September we'll start rolling out new episodes.

Leo Laporte [02:00:35]:
So the people who are getting it now are people who went to the Kickstarter and—

Jason Snell [02:00:39]:
They're not getting anything now. We did the preview episodes and upgrade. The podcast doesn't launch until September, early September. But we did a bunch of sample episodes. So you can— yeah, right now there's only the sample episodes to listen to. But we're already producing new episodes that will go out. One of the beauties of doing a history podcast is we don't have the thing where we record and then post it the next day, or the same day. We're going to record and then post it.

Jason Snell [02:01:06]:
And, you know, we're going to get ahead. We're gonna be weeks ahead.

Andy Ihnatko [02:01:08]:
So it's not that anything new will have happened in 1980.

Jason Snell [02:01:12]:
Breaking news.

Andy Ihnatko [02:01:13]:
You can reliably record that and put it to— put it in inventory.

Jason Snell [02:01:17]:
Yeah, yeah, exactly. Apple is still gonna buy NeXT. It's still gonna happen.

Andy Ihnatko [02:01:23]:
The people who pay for it should be able to get that information. Come on. Yeah, they will.

Jason Snell [02:01:26]:
They will.

Leo Laporte [02:01:27]:
In my studio, I have the original Paul Rand NeXT poster that Steve Jobs commissioned.

Jason Snell [02:01:34]:
Nice.

Leo Laporte [02:01:35]:
That was given to me, I think, in the late '90s. I can't remember when it was given to me, but It's an original. It's signed by Rand.

Christina Warren [02:01:42]:
That's amazing.

Leo Laporte [02:01:43]:
Yeah, it's really cool. That's Christina Warren. She's amazing. She's Film Girl at Developer Relations at GitHub.

Christina Warren [02:01:51]:
I'm not as cool as Rand, right? Like having a signed Rand poster, I think that's pretty—

Jason Snell [02:01:54]:
That's pretty nice.

Christina Warren [02:01:55]:
That's pretty amazing.

Leo Laporte [02:01:57]:
Yeah. I don't know if it's worth anything, but I like it. I mean, it's worth something to me.

Jason Snell [02:02:01]:
Let's put it that way.

Christina Warren [02:02:02]:
I was gonna say, I think it'd be worth something to any of us in this audience. That's pretty Pretty, pretty killer.

Leo Laporte [02:02:06]:
Yeah. Andy Iannacco also with us from the library where there is a signed copy somewhere in there of Little Women. Louisa May Alcott stopped by, she signed it. If you find it, it could be worth something, Andy. I'm just saying.

Andy Ihnatko [02:02:21]:
Well, I'll go look it up, and if it isn't signed, I'll sign it myself.

Leo Laporte [02:02:24]:
Well, that's always a problem, isn't it? You're watching Fact Break Weekly, and now, ladies and gentlemen, Without further ado, time for our Picks of the Week. Mr. Snell, why don't you kick it off?

Jason Snell [02:02:38]:
All right, I decided to go with an appliance. So this is more of an Alex-priced pick at $329. It is my favorite. It is my favorite appliance. I love it. I love a gadget. My kids will tell you I love a kitchen gadget. But I do— I drink tea and not coffee.

Jason Snell [02:02:54]:
When I was a kid, my parents had a coffee maker, you know, that ground the beans and woke me up at 6 in the morning and and all of that. And making tea, the fact is when you make tea, you gotta, you gotta take the tea out after 3 minutes or 4 minutes, whatever it is, whatever your preference is, or it gets all bitter and bad. And so for the last 10+ years, I have used and loved the Breville One-Touch Tea Maker. It is a kettle, but it has a little basket on a magnetic pole.

Leo Laporte [02:03:21]:
I have a Breville kettle, but this I must get.

Jason Snell [02:03:24]:
A little, and so you can, the night before, we fill it up with water and we put tea leaves in the basket and you let it sit there overnight, and then at 6:30 in the morning, the water boils, the basket drops in, it stays in there for about 3 and a half minutes, and then it comes back out. So when we wander out—

Leo Laporte [02:03:39]:
Oh, so it does the timing and everything?

Jason Snell [02:03:40]:
It does everything. You have tea without any tea leaves getting bad and making it all bitter. It all just happens automatically. And yes, it will work as a normal kettle. You can, you can make it anywhere, and you can set it to be— the tea can be in for as long as you want, and it will it will heat the water to whatever your target temperature is. So if you like different styles of tea, you can, you can dial that in. And for us, it's basically black tea in the morning. And so it's boil the water, drop it in for 3 and a half minutes, pull it back out, and then you wander out a little later and the tea is already made instead of having to fiddle with it.

Leo Laporte [02:04:12]:
My little gloved finger is hanging out over the Buy Now button. I really—

Jason Snell [02:04:16]:
it's just a great price.

Leo Laporte [02:04:17]:
I mean, it's the only—

Jason Snell [02:04:18]:
it's not cheap. It's not cheap. But I love it. It's totally— was totally worth it for me. I have no— zero regrets. It is one of my favorite gadgets. If you were a tea drinker, and if you like the idea that you will— you have a machine. Also, just in general, if you're making tea, even if you don't time it, it's kind of nice because if you make tea, you gotta— you gotta wait for the water to boil, and then you got to put in the water in the tea, then you got to set a timer, you gotta leave but don't go too far because then you gotta come back and all that.

Jason Snell [02:04:45]:
What if you could press a button and the tea would just beep when it was done? That is the Breville One-Touch Tea Maker.

Andy Ihnatko [02:04:52]:
Jason, can I— can I suggest Just one modification, because this is— this was something that I did like 4 or 5 years ago that changed my winters in a positive way. Put the kettle on the nightstand.

Jason Snell [02:05:03]:
Oh, I mean, well, it's noisy though. It's noisy.

Andy Ihnatko [02:05:06]:
No, no, no, no.

Leo Laporte [02:05:07]:
Well, okay, well, because I'm gonna put a George Foreman grill on the nightstand. I can hear it. I can have bacon in the morning.

Jason Snell [02:05:12]:
I can hear it, uh, off in the distance in, in my kitchen from the back when I sleep.

Leo Laporte [02:05:17]:
It gets you out of bed.

Jason Snell [02:05:18]:
It's a pleasant—

Leo Laporte [02:05:19]:
some people want to get up.

Jason Snell [02:05:21]:
I actually have to— I can't drink tea without eating something or my stomach gets all wookey. So, um, so yeah, I have to get out to the kitchen anyway.

Andy Ihnatko [02:05:27]:
I mean, for, for me, it's like I'll, I'll wake up, I said, gee, if I fully wake up, I can have tea in like 3 and a half minutes versus like if I have to walk to the kitchen, it's like, you know what, maybe I'll sleep for another 20 minutes and then it's 11 AM.

Jason Snell [02:05:40]:
And it keeps it warm too for up to an hour. It'll keep it, it'll hold it at temp. So that's also good.

Leo Laporte [02:05:45]:
Very nice.

Jason Snell [02:05:46]:
I love it.

Leo Laporte [02:05:47]:
I still, though, the price is—

Jason Snell [02:05:49]:
I know.

Andy Ihnatko [02:05:51]:
You gotta give yourself a little nod of kindness sometimes.

Jason Snell [02:05:55]:
And people, people will sometimes be like, oh yeah, Jason is ridiculous tea maker. It's like, I know so many people who have automatic coffee makers. Yeah, this is no different. You're right, it's just for tea instead.

Leo Laporte [02:06:05]:
I've spent 10 times this on an espresso machine from Breville.

Jason Snell [02:06:08]:
Exactly.

Leo Laporte [02:06:09]:
Yeah, yeah, well, not quite 10 times, but close.

Jason Snell [02:06:13]:
Yeah, I mean, yeah, I know multiple people with $1,000-ish Breville espresso makers. So come on.

Leo Laporte [02:06:20]:
Yeah, you're right. This is just tea though.

Jason Snell [02:06:22]:
That's the problem.

Leo Laporte [02:06:23]:
Uh, Ms. Christina Warren, your pick of the week.

Christina Warren [02:06:27]:
Yeah, so this is one that I actually came up with. Um, I saw it, uh, earlier this week and then I thought about it based on our discussion that we had, uh, before the show started, which is, uh, you're talking about repurposing local models on, on your, your Mac Mini.

Leo Laporte [02:06:40]:
Yes. And I'm using Ollama, which I don't want to, use, but—

Christina Warren [02:06:43]:
Oh, so, so my, my suggestion—

Leo Laporte [02:06:44]:
Is there something better?

Christina Warren [02:06:45]:
So my suggestion for you, there are a couple things. LM Studio is obviously one thing that you can use, um, and that's pretty good.

Leo Laporte [02:06:50]:
I don't need a GUI though, because it's headless, right?

Christina Warren [02:06:52]:
Okay. All right. Well, so this one is a GUI, so maybe it isn't for you, but it might be interesting for some other folks. And this is called Nativ, N-A-T-I-V. And it is built by the guy who basically created the library that LM Studio runs on top of.

Leo Laporte [02:07:04]:
Yeah, I think LM Studio is fantastic.

Christina Warren [02:07:06]:
Yeah, but, but, but this is actually— it's on GitHub, it's open source. It's MIT licensed. And the guy that basically created the library that LM Studio is based on top of has created—

Leo Laporte [02:07:15]:
Oh, but this also has an interface.

Christina Warren [02:07:17]:
Yeah.

Leo Laporte [02:07:18]:
Oh, a chat interface.

Christina Warren [02:07:19]:
Yeah. And it's Apple Silicon. And it's, it's because he maintains the library, I think that it's, it's pretty good. And so for you, it might not be— it might not fit exactly what you're wanting to do since you're doing it headless. But for other folks out there, um, it, it's, it's, it's worth, um, looking at because You can use your own agents and things with it, models. It'll tell you what model will work best on your system based on what's there.

Jason Snell [02:07:43]:
Nice.

Christina Warren [02:07:46]:
There are a lot of these things out there, but the fact that the guy that develops the library that so many other local modeling things now has his own native app, coincidentally called Native, I wanted to give that one a shout out.

Leo Laporte [02:08:00]:
You definitely want to run an app If you're going to do local models that do sports MLX, you really want to take advantage of that.

Christina Warren [02:08:06]:
Yeah, you want to use MLX if you can. And so he makes it easy to get those things. And so yeah, that's nice. That's my pick.

Leo Laporte [02:08:12]:
Native without an E. And it's at bizzz.github.io/native. You know, just Google native.

Christina Warren [02:08:24]:
I was going to say, just Google native because I don't think the URL is great. Yeah, but it's free.

Leo Laporte [02:08:31]:
It's open source.

Christina Warren [02:08:32]:
It's free open source. The first release came out yesterday. I saw it on Hacker News, and there's already been more pushes in the last couple of hours. So it's obviously getting a lot of things. It's built with Swift. It's powered by MLX. It does require macOS 26 or higher. So if you're somebody who's still like me, stuck on Sequoia on some of your machines for reasons you'll you'll, you'll be left out.

Leo Laporte [02:08:55]:
But, uh, yeah, this looks like a good replacement, frankly, for LM Studio because it's so Mac native. LM Studio's a little bit—

Christina Warren [02:09:02]:
Exactly.

Leo Laporte [02:09:04]:
Andy Inaco, pick of the week.

Andy Ihnatko [02:09:07]:
I have 2 picks and they're both things to stick on your MacBook. Uh, first, first off, uh, the only thing that I'm disappointed by my, my new M5 MacBook Pro over my M1 MacBook is that I and have the most awesome David Bowie sticker on the lid of it. And it's still available. It's, it's $3. I'm sorry, I got the prices wrong here on the— it's $3. You can order it. It's this 5-inch wide sticker that matches up the David Bowie lightning bolt logo with the classic NASA logo. And the number of times—

Christina Warren [02:09:42]:
That's so cool.

Andy Ihnatko [02:09:44]:
The number of times in this library and elsewhere where someone would Walk by, stop, turn back, and then say, "Oh my God, that's an awesome sticker!" And I've said, "It's available online." And unfortunately, and the it's so perfect. And unfortunately, I wanted. I thought about just. I have a bunch of them because they're so nice. I every time one wore out on the old one, I would simply scrape it off and put a new one back on it. I kind of thought about putting it on the new MacBook, but it seemed like that's like. You know, that's like getting the, putting the same collar on another dog.

Christina Warren [02:10:18]:
Yeah.

Andy Ihnatko [02:10:18]:
It's like, no, that neckerchief was Rusty's thing. This is a new dog. And so what I did, but I did wanna do something. So I, for the first time, I got an actual skin, a vinyl skin from dbrand to get my M5 MacBook Pro and make it look kind of bright and shiny and orange and colorful. And also make sure that when I go through airport security, It's not gonna be quite so easy for someone to mix up their MacBook with mine. There's a lot of—

Leo Laporte [02:10:45]:
You're not one of those people that says booey instead of Bowie, are you?

Andy Ihnatko [02:10:49]:
You know what, Jack? The, the, the— in 1982, 1983, I still remember the local, the local anchorman Jack someone of Channel 4. David Bowie was playing at the Patriots Stadium. And he kept— David Bowie is reporting for the— like, and I said, oh my god, it's David Bowie.

Leo Laporte [02:11:12]:
You're wrong.

Andy Ihnatko [02:11:12]:
You are so not hip, man. It's unlike us Generation X types. So yeah, I always say boy.

Leo Laporte [02:11:18]:
Good man. Exactly, man. But that's not the only thing to put on your Mac.

Andy Ihnatko [02:11:22]:
Exactly. So, so I got— so I got the dbrand skin. There are a lot of makers out there. I chose dbrand chiefly because I've seen a lot of videos around about them. The number one, it's not just like a— it is a vinyl skin, so it won't add weight to the lid, it won't wear out the hinge, it won't cause any sort of damage. I've seen videos of people peeling them off, and unlike other vinyls, they peel off very, very cleanly. So it will protect the lid from scratches, but if you decide to take it off, it's not as though it's gonna have to scrape it off or anything like that. It won't leave behind any sort of residue.

Andy Ihnatko [02:11:55]:
The other reasons for picking it is that it's a special kind of vinyl where it actually lets air pass through, so you're not going to be constantly scraping air bubbles from from the middle of it. Also, you can just buy the lid. They offer skins that will wrap pretty much the entire part, the entire MacBook, but they will also just sell you the part just for the lid, which is all I wanted. It's very, very easy to put in. You can actually— it's not one shot and you're done, and that you one chance and that's it. You can actually like keep peeling it up and removing it and putting it back on, peeling up, putting it back on. The hole that lets that lets the Apple logo shine through is actually a little bit larger than the Apple logo. So it's not as though you have to line it up perfectly or else you'll get an edge from the original metal shining through.

Andy Ihnatko [02:12:42]:
And it's, it's held up very— it looks very, very nice.

Leo Laporte [02:12:46]:
You have to put your mind— Which one do you have?

Andy Ihnatko [02:12:48]:
I have the orange one. I think it's the same color as—

Leo Laporte [02:12:51]:
Because we all love cosmic orange.

Andy Ihnatko [02:12:54]:
Yeah, I actually bought it because I didn't want weird patterns patterns or anything like that. I just wanted a bright shiny color. It's kind of like, I, I want the— I want the— I wish they'd come up with a line of just all the MacBook Neo colors because I would have gotten the citrus for it like, like nothing else.

Leo Laporte [02:13:08]:
I'm tempted by the clear one that you see. Looks like you're seeing through the COVID That looks kind of cool.

Andy Ihnatko [02:13:13]:
They got, they, they got some nice stuff. Uh, and also they, they— it really is—

Leo Laporte [02:13:17]:
The topo map or— Yeah.

Andy Ihnatko [02:13:20]:
Yeah. Now the, the advertising is kind of annoying, you have to put it— I think that's kind of their brand.

Jason Snell [02:13:26]:
Yeah.

Andy Ihnatko [02:13:26]:
But as I said, I'm very happy with this. None of the things I was afraid of, which is that again, it'll just— it'll be so impossible to— you line it up and then you get what you get. You can actually lift it up and reposition it several times before— basically, I was able to reposition it until I got sick of repositioning it. The only complaint that I have, which is barely a complaint, is that it really is designed to fit the lid of a MacBook Pro with maybe half a millimeter to spare. So once you have it on, it almost looks like it's factory, but it also means that it's either, you've either got it exactly right or it's off by just enough that it will, you know, that it will annoy you for the rest of your life if you don't peel it up and try it again. So it was literally like 8 or 9 or 10 times before I got it like so close that I said, okay, I will not be cursing myself for the life of this device. or the life of the skin if I, if I leave it like that. But once you got it good, it's good.

Andy Ihnatko [02:14:20]:
So good.

Leo Laporte [02:14:21]:
Here are pictures shared with Twit of the 2 different laptops that Andrew has customized.

Andy Ihnatko [02:14:29]:
Beautiful. You almost hate to lose the stickers. It's like, oh man, those— it's like, it's like tattoos, like on a middle-aged person, where you can think about the circumstances under which you got each one of those. And maybe you regret one or two of those, but it's part of the story.

Christina Warren [02:14:44]:
No, it's so true. I am— I take photos of my laptops after, like, I get rid of them or whatever. I have to take either the stickers off or send them back if they're a company laptop or whatever. I take a photo just so I can commemorate. But you're right, like, it's so hard. I know some people put like clear, I guess, like covers on top of their laptops and then put the stickers on that. And that way they can keep it. And like, that's cool.

Christina Warren [02:15:07]:
I I've just, I've just never done that. I've never been that, uh, smart to do that. But, uh, but your skin looks great, Andy. I've often thought about buying— I bought their skins for other things.

Leo Laporte [02:15:16]:
Do you hear that a lot, Andy? Exfoliate. You hear that a lot? Your skin looks great.

Christina Warren [02:15:19]:
Well, okay.

Leo Laporte [02:15:20]:
Is it the lotion?

Andy Ihnatko [02:15:20]:
I've never seen that combination of blue and green in a skin tone before.

Leo Laporte [02:15:24]:
Your, your, your night routine.

Jason Snell [02:15:28]:
Yeah, loofah.

Leo Laporte [02:15:29]:
It's loofah. It's all about the loofah.

Christina Warren [02:15:31]:
It's all about the loofah. All about the moisturizer.

Leo Laporte [02:15:33]:
Christina Warren. Thank you so much, FilmGirl. We're so glad to have you here on MacBreak Weekly every week. It's great to see you, and thank you for your input. I have installed OrbStack and, and all the other things on my, my headless Mac— my now headless Mac Mini. Thank you. You can find her at GitHub as FilmGirl, and of course here every Tuesday. Andy Anacco, find him at the library, but you know, he is online now.

Andy Ihnatko [02:16:00]:
Yes.

Leo Laporte [02:16:01]:
He's got a website, ihnako.com. Become a premium member and the gods will smile upon you. Jason Snell has many podcasts, but this one is his. sixcolors.com/jason for upgrade. I'm sure the history podcast will be there.

Jason Snell [02:16:19]:
Designing California. It's on the list. Yeah, it's there. I added it. You're the number one user of that page, Leo. That's why I—

Leo Laporte [02:16:27]:
Am I really?

Jason Snell [02:16:27]:
Hooked it up today. Yes. I made it for you because you're like, Jason, I don't have a place where I can say, you know where your podcasts are.

Leo Laporte [02:16:33]:
I need somewhere to send someone.

Jason Snell [02:16:35]:
That's where it is.

Leo Laporte [02:16:36]:
Because he has so many, gosh darn it.

Jason Snell [02:16:39]:
I know, it's true.

Leo Laporte [02:16:40]:
Upgrade, Designed in California, The Incomparable Mothership, MacBreak Weekly, Robot or Not, Six Colors, The Incomparable Game Show, and that's just the beginning.

Jason Snell [02:16:47]:
Just the beginning. There's so many, and so, so many more.

Andy Ihnatko [02:16:49]:
You're like an F1 car with like a million sponsor stickers.

Jason Snell [02:16:52]:
You know, it's hard out here for an independent content creator. You got to Do what you got to do.

Leo Laporte [02:16:57]:
Thank you all for joining us. We do MacBreak Weekly every Tuesday, 11 AM Pacific, 2 PM Eastern, 1800 UTC. You can watch us live in the Club Twit Discord, of course, but also on YouTube, Twitch, X.com, Facebook, LinkedIn, and Kick. On-demand versions of the show available at the website, twit.tv/mbw. You also can find us on YouTube. That's video, of course. Best way to get the show though, subscribe, and that way you'll get it automatically the minute it's available. I'm rushing because I have 2 minutes.

Leo Laporte [02:17:22]:
Before Jason has to leave. And I'm going— I made it.

Jason Snell [02:17:26]:
You did it.

Andy Ihnatko [02:17:27]:
Thank you.

Leo Laporte [02:17:29]:
That's because we didn't do a hydration break. You will see us next week, I hope, right here. But meantime, it is my sad and solemn duty to tell you, you gotta get back to work now because break time is over. See you next time.