Upgrade - 628: Conspiracy to Steal Trade Secrets

Episode Date: July 13, 2026

From Relay, this is Upgrade. Episode 628. Today's show is brought to you by FitBard ExpressVPN and Century. It is July 13th, 2026. My name is Mike Hurley, and I'm joined by Jason Snell. Hi, Jason. Hi, Mike. I looked for you throughout the coverage of the Wimbledon finals.

Okay. And I just didn't spot you. I saw some royalty. I saw some Hollywood stars. Didn't see the most important person to me. You know, about 15. 15 years ago, I was at a Wimbledon semi-final between Roger Federer and Raffa and Adele. That's pretty awesome.

That's always, I've always dreamed of going to one of those. I was a seat filler, though. So I only got to see a few, like, I only got to see very small amounts of the game, I'd say. That's okay. From different perspectives? From different perspectives. That's nice. And the thing is, even though I was a seat filler that you needed, every time the person came back to the seat, they were always really mad at me.

But I'm only here because you didn't make it back. in time. Like, I'm here to fill your seat. It's a good time, and one day I hope to do it to do it properly. It would be lovely. For realsies, yeah. That would be, I would love to do that sometime. That's on my list still. But I have a Snowd talk question for you, Jason, Snow. Okay. It comes from Tony. Tony says, with the popular TV time service and app announcing that they're shutting down recently, it got me thinking, does Jason use any service and or application to track his movie and TV watching? It's funny.

I didn't do this with television for a very long time. I tried stuff. I could never get anything to stick. And I finally got one that sticks. And the good news is it hasn't been discontinued. The bad news is it sounds like the name of the app that got discontinued. And that's because RIP to TV time, but television time lives forever. Yeah, it's called television time.

And what's worse, when you install it on your iPad or your iPhone, the little app says TV time. Yep. But it's television time. It's very good. I had it recommended to me by John Syracusa and Dan Moran. Okay. And so it's got multiple layers of approvals. And I like it.

It's pretty good. It's, you know, there are complaints. There are things about it that I don't like, but it is the only one that I've ever had that sticks where I can mark, we watch that episode, and then I can look. and my iPad when we're sitting down to watch TV and see everything that's sort of like in my I should watch this queue. It means I can add things when I hear about them

and then they just kind of go in the list and that's what I've been using for that. And then for movies, it's letterboxed. I just, I will log movies I watch to letterboxed. And that's not that interesting, but there it is. I do that. Does letterbox not do TV? No.

You feel like it should. Not to my knowledge. They should do tell TV. They might have stuff you could log, but the important of a TV tracker is an up-next tracker to me. Right. You need to be able to say, where am I in that show? You know, am I, which season did we finish? Are we starting season three now or what episode are we up to now?

Or, oh, that is, I just used it, use television time last night to see when is this show coming back? And it says, like, it'll be back in 14 days. so that's what I use for it. Yeah. Very nice. So you actually mentioned this in the last Upgrade Plus, where we also spoke about some TV recommendations. And I want to give follow up to that right now to tell you that Ludwig is amazing. Thank you for that recommendation.

And I would like to recommend it to anybody else. It's so, so good. We watched it all in the last week. I know. It's good. second season is coming and they've already commissioned a third season, which is great, because I think the story is they commissioned season one of Ludwig and didn't know what it was going to be. And then it did incredibly well, like to the point where I was reading about the ratings for small profits, which is a fun show, McKenzie Crook did. And it had good ratings.

And they said, these are the best ratings for our original comedy sense Ludwig. And I was like, oh, I guess Ludwig did really well. Ludwig, like I just, I read this and I just checked it now. Ludwig was the BBC's biggest launch for a comedy since 2018. Yeah. Yeah. So they didn't realize that. So after it had this huge launch, they're like, we would like more Ludwig.

And they're like, well, great. You know, we'll get it to you in two years. So I think the good news is they've already said that they're, they've commissioned a next series. It's reached the point now. They do the first year on spec, and then it takes years for them to get back to speed to do another season of it. But now they know that they want it.

I hope that season three will happen faster than season two did because they know that it's a hit now. But it's so good. So, yeah, highly recommend it. I mean, I love David Mitchell. Yes. I love the Mitchell and Webb Look. It's one of my favorite,

all-time favorite sketch shows. And I love that in the announcement about them doing season three, they mentioned that it's a co-production. Yes. With, I believe it's called that Mitchell and Webb company. Yeah.

Which is funny. Like, even though those guys do so much work separately, they still produce stuff through the company they set up. Yeah. Which is cute. That's a nice. I like that too. That warmed my heart when I saw that.

If you would like to send in a question of your own to help us open a future episode of Upgrade, just go to UpgradeFeedback.com and send in your own snail talk. We have some follow-up. So on last week's episode, we were wondering, and Jason, I think you specifically wondered aloud, who is still using FileMaker today? Yeah, I knew I was going to regret that. We got so much. I don't think it was a regret. I thought you knew you were going to do this.

Well, I didn't quite, let me put it this way, I didn't quite, I answered my own question, which is, it's legacy, a lot of legacy systems. Yeah. But we heard from everybody who, yeah. I mean, we may have heard from literally every. We may have heard from everyone. So I'll start off with Sam who says, I'm a TV producer in the UK specialising in quiz shows. Last year,

I worked on a new international production which uses FileMaker as its default questions database. This is Shepardy UK, maybe? I mean, it's similar, right? So it's definitely not just limited to 20-year-old productions. In over a decade in the industry, I've never used it myself as there as a company specializing in making databases for quiz shows,

which is cornered the UK market. But I wouldn't be surprised if that besie. spoke system started off in FileMaker, given the similarities between them. Of course. Brian says, I had to chime in and say that FileMaker Pro is used by, used by most Broadway productions to keep track of every element of a show's sound system. Each cable, mic, speaker, sound processor, etc. is planned out via FileMaker databases with different teams custom styling their information and functionality. We even use it to put labels on everything, printing custom labels on every return and shipping address.

stickers. Yeah. Patrick says, and they carry on. We've used it at my office since 1990, when the file maker manual was still a hard cover book. I have a number of databases I've developed since the 90s, and I'm still using some made by former co-workers for compiling ski condition reports and registration at a trade show that we do, too. It's just so much easier than anything else out there. I also use it for our Christmas card list, says Patrick.

Sam says, a big chunk of my early web development career was building FileMaker apps hosted with FileMaker server on an X-Serve. This sounds familiar. On a prestigious university in Southern Connecticut? Connecticut. I'm going to say CT. Okay. It's an Ivy League school you may be familiar with.

Yeah. As I was listening, I was thinking, please mention Lassau, please mention Lassau, and there it was. I wrote dozens of Web apps in Lassau script that talked directly to hosted FileMaker apps. The web has surely moved on. Anyway, thank you for the nostalgia bomb. And this is my favorite. It's came from an anonymous follow-upper.

I'm a file maker developer working with exactly the kinds of systems every day you talk about on your show. They're old and they work and my team keeps them running. I feel compelled to remind you that the Apple gift card database as well as AppleCare is all run on file maker systems. It is my most fun, modern file maker fact. Yeah, I also heard from somebody who said that they work at Apple, But they're, I think they work at Apple and they're right across from the file maker building. And they had wondered for a long time, basically, really? Why?

And so we kind of gave, we kind of gave the answer of why. Like, are they doing things over there? So, yeah, this is the answer is there's a lot of legacy stuff out there and needs to be supported. And again, for anybody who has ever dealt with one of these kind of systems, like, everybody knows that it's not cutting edge. everybody knows it's not the latest and greatest, but the amount of effort that would be required to disentangle you from this thing that you've built

that works perfectly fine for you and it doesn't have to be cool, it just needs to work. The amount of effort required is enormous and the benefit is probably zero. It is, I'm going to say, a little bit like why I still use movable type three for the incomparable. It's because we built a very complex system.

It works fine. it's not cutting edge, it's not cool, uses an old version of pearl, like all of these things about it, but like it works and the amount of effort to rip it apart and replace it with something else that might not really be much better means you just keep using it.

So I kind of get it here. Also, to the person who said they use it for the Christmas cardless Patrick, also once you have FileMaker and you know how to use it, it's like, you know, it's that hammer that you're looking for nails. Like if you've got a functional, database and you know how to build stuff in it, then you just use it because you've got it and you know how to use it. So yeah, yeah, I get it. So this is why this is the answer to the question, why is filemaker a wholly owned subsidiary or now it's back to being called Claris, a wholly owned subsidiary of Apple. It's like there's a business there. So they just keep it running. My favorite thing, I put it in the show notes, Claris.com, right at the top it says Claris an Apple company, which is just, I don't know why this is such a fascinatingly weird thing to see. but it really is just so weird.

But there you go, Claris and Apple Company. I think it's only Claris and Beats, right, that we would call Apple Companies. Oh, I'm really stretching my thinking here. No, I think that's it. They have investments in other companies, right? No, but I think the wholly owned subsidiaries that are the separate is, I think it's Beats and Claris.

Yeah, I don't, I cannot think of it. anymore. So yes, I think you're right. And let me just say, Claris, I love you. Clarice and Beats, they're not, they shouldn't be mentioned in the same. They shouldn't be mentioned in the same breath. But nevertheless, here we don't know. What is Johnny File Maker and Dr. Dre have in common? That's true. Clarison Beats. At the Financial Times is reporting that Apple is now moving ahead with testing the DRAM chips from banned Chinese suppliers. So they're now doing the testing. So I think at this point, it feels like they're either, A, confident they're going to get the exemption from the U.S. government or B, have decided they don't care. I don't know. I think they're poised. I would, my guess would be they're poised. They're poised. They want to be ready. If they get the okay, they're going to spring into action. And so they're trying to do everything they can in advance of that. I don't. I think that's as far as it's gone. Say it somewhere in the middle. I don't know if they're confident, but like, you can imagine Tim Cook saying, well, be ready. If I get the, if I get them to say yes, we should.

be ready to jump on this immediately. And they're like, okay, well, we're going to start. Like, what's left for us before we could go into production? It's like, well, we've got to test him. He's like, we'll do that now. Do that now. Just in case. Let's do a little upstream.

A quick moment here. Apple has received 89 Emmy nominations. Seems like a lot. Jason, it is a lot. It's a lot. The stat that Apple is touting the most is that they are the network. I use the word network. I don't know how it was to call it company.

I don't know, with the most nominations in both the major drama and comedy categories. Okay. There's a lot of exceptions there. I know. They did not get the most nominations, but they did get a lot of nominations in the major categories. I think what they're saying is the categories you care about, right? Like show and people, not necessarily, you know, should go down the list. Like, for example, like I will just mention, you can go down the list where, like my wife, you know, took great pains to point out to me that,

what is her? Parma Royale got quite a lot of Emmy nominations or got some because she's maybe the only person in the world that enjoys Parm Royale so she wanted to let me know about that but as I reminded her they cancelled that show

it got eight, it got eight numbers but they're all like production design, costume, hairstyle, stunt performance. That show is beautiful. Like it is a really great period production for a period that is not touched very much, like the 80s, I think. It's like, it's very beautiful television show. But from my perspective, then kind of like falls apart on literally everything else.

I did not enjoy season. I watched season one way for and I did not enjoy that. I have not watched season two. But anyway, they got a lot, but they're not the major categories. But the big winners are Widows Bay and Pluribus with 19 and 18 Emmy nomination suspectively. Shrinking got 10. Slow horses, got nine, Margo's got money troubles, got eight, and then the list goes from there.

Apple had 15 titles in total getting Emmy nominations. So I'm sure they're pretty pleased about this. I think they are. I think they're very pleased. This is why they do it. Yeah. Although I mean, kind of like Emmys in general, I think HACS is going to probably sweep the comedy this time. You never know, but I mean, Hax got a lot of nominations, and the Academy really likes to do big send-off.

for shows that aren't coming back. And, yeah, so we'll see. But Apple, they've got to be happy. And it shows, I think the breadth of it also shows, I think that one of the reasons they're happy is it's so many different Apple shows that did really well.

Your show, your friends and neighbors, was nominated for Best Drama. It should get more than that, though. Like, I was actually going to mention this. It's literally only nominated for that. John Hamm didn't get nominated. I don't know how a show could win. best in its category

and then none of the actors get nominations. I mean, I know why. It's called The Pit. The Pit is the problem, right? In this scenario, where everyone that's in the pit is getting nominated. But I think that your friends and neighbors is fantastic. I love that show.

And I'm happy that it got a nomination. But John Hamm should also be getting a nomination. I also, I mean, this is one of these things with a bear where they could put that show on comedy if they wanted to. It's a dark comedy. But it's like where are you going to put it? And they decided to put it in drama.

The truth is if you run over 30 minutes, you're a drama. And if you run under 30 minutes, you're a comedy. And that's just how they've decided to do it. Those categories don't make sense anymore. No, they do not. But yeah, good work for everybody over at Apple TV. They continue to crush it. Absolutely continue to crush it.

It's true. This episode is brought to you by FitBod. If you're looking to change your fitness level, it can be hard to know where to get started. which is why I want to tell you all about FitBod because it is an easy and affordable way to build a fitness plan that is made for you because everybody has their own path

through personal fitness. So FitBod will use data to customize things to suit you perfectly. It's going to adapt as you improve so every workout remains challenging, pushing you to make the progress that you want. It's important to do that adapting because FitBod will track your muscle fatigue

and recover as you're doing your workouts to help design that well-balanced workout routine. So the app will mix up your workouts with new exercises, rep schemes, supersets and circuits, because your muscles improve when they're working in concert with the entire muscular system. If you overwork some or underwork some, you're going to negatively impact your results. So they make sure that they're going to mix that up for you, which also my favorite thing is it means you're getting new exercises all the time so you don't get bored. I find it to be quite demoralizing to do

the same workout day after day after day. That is when I start to fall off. What I love about FitBod is they mix it up so you're getting new stuff, keeps me engaged. You're going to see the best results in a workout program is tailored to your body, experience environment and goals. FitBod stores this in your FitBod Gym profile where it tracks your muscle recovery so you can avoid burnout and keep up your momentum. FitBod builds the best possible workout for you by combining AI and exercise science. They have analyzed billions of data points, fine-tuned by certified personal trainers. And every time you get a new exercise, which I mentioned, happens often, you're going to be confident you're learning these movements the right way because they have over a thousand

demonstration videos in the app. Personalised training of this quality can be expensive, but FitBOD is only 1599 a month or 9599 a year, but you can get a huge 25% off your membership by signing up today at Fitbar.me slash upgrade. So go now and get your customized fitness plan at FITBOD.mys dot me slash upgrade. That is FitBod.com slash upgrade for 25% of free membership. Or thanks to FitBod for the support of this show and Relay. It's time

to lawyer up, Jason Snell. Yeah, it is. Clunk, clunk. It is the classic. I almost brought the jingles back for this because I had that jingle in my head all weekend. So, it's time to lawyer up. Last Friday, Apple filed

a lawsuit against Open AI, accusing the company of the theft of trade secrets. Apple is saying that employees who leave them to join OpenAI are stealing secrets, quote, for the benefit of OpenAI. So I want to start this conversation by reading the full statement that Apple gave to various outlets, including Chance Miller at 9 to 5 Mac, who wrote a great article breaking down a lot of the legal filing, which was very helpful for me in planning today's episode,

because what I don't want to do is read an entire legal brief, so I'm very thankful for all of the people who did. I read it. It's very entertaining. Excellent. I'm pleased that you have done it too, because I didn't want to do that. I didn't need Chance Miller and his useful, helpful, breakdowns that are good and does a good job. I didn't need that.

Chance and Mark Goerman, I read their articles. Get out of my way, Chance. I don't need your help. Help Mike. Get in my way, chance. Help people who need your help. Anyway, Apple said, at Apple, our team, are constantly developing breakthrough technologies to create the best products and services in the world, and protecting their work and intellectual property is something we take very seriously.

Recently, significant evidence has emerged suggesting individuals employed by OpenAI wrongfully took Apple's secret and confidential information regarding our unreleased technologies, processes, and products. We will always defend our team's hard work and innovations, and we are taking all appropriate steps to do so. So this lawsuit names four parties as defendants. So OpenAI. I.O. products, which is, I mean, this one's hard. It's like, it's the startup by.

Johnny Ive, Evan Tankie, and Tantan. And Tentan. And they throw shade on it in the brief. They basically say, we should just consider this with basically owned it operated by Open A.I. They basically say it's a shell company. It's a fake company that was always in. tended to be and was completely controlled by open AI. Yeah.

Because, you know, there's love from and there's open AI and then there's I.O. And so Apple's like, for the purposes of this, you should really even consider just I.O. is open AI. Because really, it seems like I.O. was created just so Johnny Ive didn't have to go to, like, they weren't going to buy love from because Johnny didn't want to sell love from. I think that's it. So they put, they created a space for them to do this work. And then Open AI said, we're just going to buy you for stock. Because remember, like Johnny is not actually. like an employee of Open AI like a lot of people are. He's still like, he still works at love from and advises.

It's, it's complicated, but complicated in a way that is done specifically just so everybody gets something they want. Yeah. Apple's basically saying, you know, this is a, don't, don't pretend that this isn't just open AI again, but which I just thought, I enjoyed the shade of that. Yeah, like, yeah. There's a lot of shade. Not real.

As well as that two named individuals. So former Apple vice president of product design. Tang Tan and a former Apple Senior Systems electrical engineer by the name of Chang Liu. He isn't mentioned in the suit specifically, but to me, like, what a terrible state of affairs to have Johnny Ivan broiled and all this. Like, I just find this so, like, we're going to get to all of it. But to me, it's just like, this is, I'm just so disappointed. They didn't, they didn't name him. I think it's very funny that they didn't name him. He's not mentioned at all. They took great pains not to name him. Yes, they did. But, yeah.

man, like if this goes, so Apple wants to take this to a jury, if this goes to jury, he'll surely be called, right? Could be. But nevertheless, no matter what his involvement or not in all of this is,

I just find this to be so, I mean, the whole thing is so gosh anyway, but that to me is really close. It makes me wonder when you talk about Johnny wanting to be a step removed from all of this, you know,

you know, my feeling is that he got burned out on all this stuff. Yeah. And that's why they let him do cars. And he was designing Apple Park and all of those things. And now he's got love from, which is almost like his playground, fun project where he can mess around and do luxury items and all that. And then he does this. And they create I.O.

And like you said, he's sort of not, he's an owner, but not really a, he's not an open AI employee or anything like that. I wonder if part of it is that Johnny I've just kind of wanted to keep everything at arm's length. Yeah. And that, and that you look at this and you're like, is he like, this is why? Yeah, it's like, it's like, should, keep it away from me? Should you have been close to it or should you have been two arm lengths away from it? Like, is the- Sure, Sam, I'll help from over here. Yeah.

Like, I'm going to just stand over there. It's right. All right. So, I'm going to try and lay us out the best I can. So the timeline and then talking about things in detail. So Apple says that they contacted Open AI in February to voice their concerns, their belief that this was happening essentially, and to ask Open AI to investigate it. They say Open AI did not respond to this complaint from Apple. Now, there are a lot of things in this where it's like, how much of this is the stuff you're supposed to do and everybody ignores it anyway?

Like, I don't know, right? Like, I don't know enough about these kind of legal processes, right? Yeah. There is a statement from Apple on the breadth of this issue in the, this is in the, this is in the, uh, the legal case. This is the tip of the iceberg. Apple lacks visibility into what's been happening behind closed doors at OpenAI, where such misconduct is normalized and exemplified by leadership.

This much is clear, however, at every level, from members of its technical staff to its chief hardware officer, and in coordination with business partners, Open AI has been stealing Apple's trade secrets and confidential information. As a natural result, OpenAI's nascent hardware business now rests on the shakiest of foundations, rotten to its core by illegal reliance on misappropriated trade secrets. Man, that's a... Good, good writing.

I like that right. Yeah, rotten to its core. And that's them sending the message, like, we think your entire hardware operation is part of this lawsuit because it's based on false premises and theft of secrets. By the way, it is important to note, Chief Hardware Officer, is that code for somebody? It's actually Tang Tan.

Tang Tan is also named. So yeah. Yeah. Yeah. A job that he, by the way, according to Mark German at Bloomberg, and Germman talked to people who said the design side wanted Tang Tan to get the job that John Turnus got. Yeah. And this is why I always say, like, never forget that stuff happens sometimes because someone

didn't get the job they wanted, didn't get the title they wanted, didn't get the promotion they wanted. And I'm not saying that that's exactly what. happened here. But I think German's story, it's a strong implication that one of the reasons Tang Tan left is because he wanted the job that John Ternis got five years ago. And then if you extrapolate that out,

now again, I don't want to trip up too much on this, but like we don't know about the truth of any of this, right? Like, we don't know. And it's all, but I don't want to keep saying allegedly. But if you extrapolate that out, you can see a scenario where somebody who didn't get the job that they wanted being annoyed and upset enough that they may do some things they shouldn't.

Sure. Sure. They're really bent out of shape about it. And some of the worst behavior in this lawsuit is, it feels kind of like that. Yeah. Like there are people who feel like they're either petty or they're out for revenge or whatever. And what German said too is it was like, it was the designers felt that he should do it. And like he's not a designer, but like they, and we know Mark German talks to the designers.

that's his best source of leaks. You could just see it. And so, of course, he would say that. But, like, you know, you're going to have camps. You're going to have people who are, like, grumpy about it. And I think that some of this behavior maybe can be explained in part by that, by people who were inside Apple feeling adversarial toward Apple and wanting to do max damage on their way out the door.

Let's talk about the two named individuals. So Tang Tan, Chief Hardware Officer, as Jason mentioned, Apple says that he used his insider information of Almond. and their processes to extract information from job candidates from Apple to get information on what the company was working on, going as far as asking for them to bring samples of components for quote, show and tell. Mark Gohmann notes this. About 400 people have left Apple to go to Open AI. It's one of the bigger kind of exodus of staff to a single entity that has occurred. So it is significant and there is that pipeline there.

I've seen some stories saying that, look, if you're in certain groups, being asked to bring samples of your work is not unreasonable. The questions for me are, were these components considered secret? And what was the purpose of bringing them in? The Apple is going to say the whole plan here was that he was trying to spy on Apple's operation after he left. by having people from Apple come and then deposing them essentially about what they were working on at Apple right at the moment so that they could get a sense of what Apple was doing and that he had them bring these products. If it's, you know, if it's fairly standard Silicon Valley practice, this isn't actually that big a deal. If he's inducing them to bring secret stuff so that they can look at it and study it, then it's a pattern of behavior. And that's, that's just one of the things that will have to be argued.

Yeah, but I, you know, kind of reading it on the face of it, that isn't all he did, though, right? So it's like, okay, I can see the argument in that. But then also, Apple says they have evidence of candidates for positions taking screenshots and downloading files related to confidential projects. And they know that Tan then asks for additional detail about these projects in interviews. Apparently this became an established pattern. So one of the things that we can read out of this that I think is fascinating, how do they know? Well, the answer is some people at Apple interviewed with OpenAI and were deeply uncomfortable with what they were asking and told somebody at Apple. Yeah. Yeah. That's what happened. Because if every one of these people then just goes over to Open AI, nobody's squealing, right? Nobody's talking. But they obviously talk to people from Apple. and asked these things. And it's either people, it's people who did the interviews and people who know people who did the interviews and they talked about what they were doing.

And then I would absolutely, absolutely not rule out a scenario where then some people were asked to apply for the sole fact of get the info. What is this like? Tactically, yeah. Yeah. Yeah. Because that feels like something,

like, you know, the things that we know that they do with their security teams, like this feels like something that they would then start to do, like send out fishing expeditions of like what is asked. before someone would leave Apple to come and join Tan's team,

he would share with them an internal Apple document that would tell them what the departure security protocols would be. So when he says, you've got the job, oh, by the way, this is what is going to be asked of you. Yeah, when you tell Apple you're leaving, they're going to give you this document, which we've spirited out of Apple, this Apple presumably confidential document that we just took.

And so here's what you need to do in advance to protect yourself from Apple protecting its intellectual property, which is, again, not great. To me, this is the smoking gun. This, to me, is the, is the ever, like, I don't know. There's a few. This feels like the thing. I know, but for him, because you were mentioning about, like,

oh, maybe this is like standard Silicon Valley practice or whatever. This, to me, feels like the thing that would suggest that there is a malicious intent for all of it. Where, like, you're a saying to someone, you need to, you need to, get as much information as possible before you leave. And again, if everybody does that, fine, then everybody's stealing trade secrets from each other. And like, yeah, so, I don't know, a certain

point though, someone's going to say, hey, you shouldn't do that. So I think this goes to the core of one of the issues here. And I want to be sympathetic because trade secrets. Trade secrets are more complicated. So like in California, you can't, there's no such thing as a non-compete. Yeah.

In California, you can't, your employer can't say, as part of your employment, you agree not to work in this industry for five years. I think is great because non-competes are evil. And I don't say that lightly. They're evil. They're basically saying, we don't want to compete for your talent. We don't want somebody else to pay you for the work you do that's good. So we're going to make it so you can't. Literally can't work in your chosen profession. If you leave here, you're trapped. It's evil. Pure evil. Anybody who whose company has non-compete, I'm sorry, you're doing wrong. R rethink your life. Illegal in California. Great. Trade secrets.

So here's, for me, one of the conversations that we have to have, and I think that there's a level in this that, that I think needs is absolutely debatable, is Apple would prefer that people who worked at Apple and became incredibly knowledgeable on the process to build amazing pieces of hardware. They would prefer those people not work elsewhere against Apple. Yeah. Right. Obviously, they would. But that's just like a non-compete. Of course, you'd prefer that your people can't go. anywhere and can't do anything.

But people have skills, and they should be allowed to apply those skills elsewhere. So the problem with trade secrets is there's a very weird, hazy line between the skills of the people working there and the trade secrets that are sort of like intellectual property that belongs to the company in question. And I, you know, I am not a lawyer. but the way that I've always read that is people's innate skill doesn't belong to the company

that helped them build the skill. Yeah. But the detailed processes, documentation, all of that, blueprints. Yeah. If you can't replicate that yourself using your talent and skill,

you are stealing trade secrets. That's where it goes across. Yeah. I think it's pretty simple. anything you can write the word confidential on. That's the trade secret part. Yeah. So if Tang Tan is hiring brilliant people from Apple and then they're replicating what they know at OpenAI.

Because they've built knowledge. It's in their brains now. I think it's okay. And I think that would be one of Open AIs arguments. I think that that is okay. People should be free to apply their trade wherever they want. However, and this goes to what you said about this being a smoking gun, if you have built an entire system.

that is designed to spirit technical documents and information about secret projects out so that you can take that information. If you built a system of how to spirit things away before you let Apple in on the fact that you're leaving, to me, that looks like a conspiracy to steal trade secrets. Yeah, that's the theft part. Like, you're taking something.

Like, what's in your brain? No one can take that away from you. But if there's a document that you can download, a piece of paper you can print out, that's company property that you've taken now. Like, that's just the fact of it. Yeah. Justin in our Discord says the rule is pretty simple. And this is someone who's prosecuted trade secrets cases, Justin says.

You show up to your new company with an empty briefcase. Your knowledge and experience can't be trade secrets. Yes. But you can't take the work you did with your company to another company. This is pretty simple. And this is where, look, we've only heard one side of this. Again, I want to be aware. We have only heard one side of this.

Other than like Open AI saying, well, we don't need trade secrets. We're just innovators. And Mark, or what was it? Sam Altman's like, oh, we don't, we're not afraid of Apple. We don't hate Apple. We just are doing our own thing, man. Okay, but basically they haven't responded with any content. It's just Apple's lawsuit here.

But that said, this is really, I think it's really, really damning because it's, it's, I think it's institutional arrogance. I think the people involved here decided that as far as they were concerned, they owned the trade secrets at Apple because they were wronged by Apple, because they'd put in the time, whatever it was. Because like, I could even understand some wink, wink, nudge, oh, I happen to have this document from Apple, which would be effective trade secrets, but it would be kind of like a thing that happened.

This is alleged to be systematic. This is, and that's why I don't use this phrase lightly. It's a conspiracy to steal trade secrets. It's not the, you know, oh, this piece of paper was stuck to the bottom of my shoe when I walked out the door. It's, we have a plan to slowly steal everything they've got by exfiltrating it from Apple as we hire new people. It's really bad. I mean, these are incredibly serious allegations, I think. Let's talk about Chang Liu.

Chang Liu had been exploiting a bug in Apple's engineering system to continue accessing files after leaving the company, to which Apple has records of messages sent to a colleague saying, quote, LOL, I found out I can access the network storage. So funny. This colleague, who this was sent to, Alyssa Peng, then helped Chang obtain confidential information, who then also left for OpenAI. So this is, this guy's in trouble. Yeah. This guy's in big trouble.

Yeah. Because this is, I mean, this is kind of cybercrime. Yes. This is, so we have to ask the question like, how does he have a Apple laptop after having left the company? Apple says that Leo did not return his company laptop and has downloaded one of his company laptops or whatever. And has downloaded over a thousand pages of technical documents detailing the work he completed at the company to take to Open AI. Yeah. Leo allegedly also conducted applicant coaching for the information to bring to a job interview.

And I would say also damning here is Alyssa Peng, who was an Apple employee after Chang Liu left, who then later left for Open AI. In the interim, and Apple has documentation because these are Apple systems that she was using. In the interim, she funneled confidential documents to Chang Liu, according to Apple. alleged. We'll just say allegedly, but this is a serious allegation as well because now you've got somebody who was accessing Apple secret information after they left, not even just emailing it to themselves, which they also allege while they worked there. But then after they worked there, continuing to take documents. And then after that, getting their friend who worked at Apple to take documents and then eventually that person gets pulled out. Like that is,

pretty serious, serious stuff. Serious allegations there. I'm not surprised that Apple found this out, but I'm surprised that it got to this scale before they seemed to. So, like, if employees were sending confidential files to personal email addresses, which is something that is alleged in the complaint,

I don't know how they didn't know. So when I worked at the bank before my life became much better, anything labeled as confidential would flag if an email, if it was email to an external address. And I know this because I once got a corporate warning because I sent a presentation that I made to my personal email so I could demo it at a client on my iPad.

I did not uncheck the confidential box. It was not a confidential document, but it was marked as confidential. And so it immediately flagged when it went through the system. And like maybe Apple's systems, aren't so draconian, like, in a way that these were, although that would be surprising to me. Maybe.

Also, these are technical people. These are computer nerds. Yeah, they just know the way around it. Yeah, screenshots maybe or whatever stuff as opposed to the documents themselves. But it is just surprising to me that this seemed to occur at such a scale before it was recognized.

Why email it to yourself? Can't you just put like a thumb drive? Print it. Like, yeah, I don't know why. Look, basically as well, like you say to technical people, I also think some of this, like some of these things are technical people who think they know about it. Like they, yeah, yeah. And then they end up getting caught out.

There is a massive streak of arrogance running through all of these allegations. Chang Liu thought they could get away with it. Didn't care. Like, ha, ha, ha, ha, I found this. This is so funny. And it's like, so funny. Also, last thing, Apple claims that someone at Open AI was able to trick one of Apple's manufacturing partners. to divulge information about a proprietary metal finishing technique

and also approach to battery manufacturer to attempt to do the same, basically saying Apple said we could do this. So this is fraud. Yeah. This is just fraud. And in our chat room, people are saying, well, why aren't they going after these individuals for these other crimes? It's like, well, give them time. I think the purpose of this lawsuit primarily is to get a judge to tell OpenAI, it's on, you know, prepare all your documents, do not throw anything.

away, you know, you're going to have to deal with this now, and then they can proceed from there. But this allegation, like, this allegation is very serious. This is somebody misrepresenting themselves to a manufacturing partner. I think also Apple wants to go for the head here. Apple, Apple believes this goes to the top or close to the top at OpenAI. Apple believes that Open AI's policy is steal as much as you can from Apple. That's what they allege in this document. And so you start by saying, we are suing open AI for a systematic, I mean, and they said that at the intro of the lawsuit, for a systematic attempt to steal Apple's trade secrets that covers all of these bases, including lying to suppliers and claiming that Apple had given them they're okay so that the suppliers gave confidential information out. And so you start with the top, which is their allegation here is that this is a systematic strategy. by OpenAI to steal as much knowledge from Apple as they can so that they can hurry up and get their hardware out faster.

Now, look, there was a piece of the Wall Street Journal about this yesterday that was totally in the bag for OpenAI that was basically like, Apple, it's not fair. Apple steal stuff all the time. Haven't you heard of Sherlocking? Which I could, I had to take an hour to roll my eyes back out of my head after I saw that. just a what a stooge that Wall Street Journal reporter turned out to be, definitely in the bag for Open AI. But look, it is an argument that Apple has been wrong-footed by AI, right, and that they are concerned that AI might challenge the smartphone. AI devices might make Apple's iPhone product harmed down the road, and that that is bad for Apple and that that is one of the reasons they're reacting here. And I would say, okay, sure. However, we also have to say, nobody does it like Apple in terms of hardware.

Nobody's really even that close. Samsung, I guess. Like, there's very few. Open AI is starting from nothing. Open AI, while I would say is mostly should, one of the problems with over AI is they have no focus and no discipline. I think they have enormous amounts of arrogance and no focus and no discipline. So they're focused on what, what, everything. They got their models over here and they're like, no, we're going to do hardware. Sure, we're going to do hardware too. Okay. But they want a shortcut. They want a shortcut because building up this level of knowledge and skill to make a product that could rival Apple, they're looking at it and thinking, well, if we take too long, Apple's going to have a chance to get up to speed on the new stuff that we're doing in an AI world. And they'll catch, because

they've got the skill and the resources and they'll beat us. So we need a shortcut. And I think that's the allegation here. Let's not just say this is about Apple being caught flat-footed on AI and being worried about the iPhone. That's part of the story, no doubt. But the other part of the story is Open AI had no hardware skills at all. And according to this lawsuit, what they decided to do is hire a bunch of Apple people away and steal the trade secrets from Apple as a shortcut to get them where they wanted to go. And that's why Apple, Apple's attorneys, are happy to say this entire thing is, you know, the fruit of a poisonous tree, that they have built the shakiest of foundations for all their hardware because it illegally

relies on misappropriated trade secrets. And again, I'm open to the idea that like Open AI is fully within their right to hire, to give those people better jobs for better pay to start a new business for them. Absolutely fine. But if there's been a decision made, either by the higher-ups or by the people they brought in who feel like they're asking too much, we can't do this, I need my stuff from Apple, that they've decided our whole thing is going to be. We deserve that stuff. We're just going to take it from Apple. It might as well belong to us. That's the problem. Because then what they're saying, again, and my argument would be they did this because they didn't want to do the work that would have taken longer to do it. They wanted shortcuts. And the shortcuts were, find out Apple Secrets, take their documents, lie to their suppliers when necessary, so that you can ship that product in a year instead of five years.

And from OpenAI, it's going to be very hard to argue this if one of the people named in the case is at a C-suite level. Like, it becomes, I think, very difficult to be like, oh, we had no idea what he was doing. I was like, well, this guy is one of your top executives, right? Like he is the chief hardware officer. Chief hardware officer. Like, and he's like running some renegade team inside of the company. It's like, you know, it's like that doesn't really hold water, right? Yes.

Doesn't really hold water. I don't know. I don't know what the, what the result of this is going to be. Yeah. I mean, presumably the goal here, number one goal here is excruciating levels of discovery. Which will be, I think, for Open AI, like, amazing. Like the discovery process as this goes to jury case, jury trial.

Getting these documents out of OpenAI will be, I think, fascinating and horrifying. I mean, the way, I mean, and maybe, again, Apple's going to portray it in the worst light possible. But Apple's whole thing here is to say, we don't have a lot of visibility into what's been going on.

Yeah. What we've seen suggests an enormous conspiracy. And we, we're filing a lawsuit. We want to see those records. We want to prove it. And like, if this is what we can see from the outside, what do they have on the inside in terms of their strategy to steal these documents? What are they using these documents for? What was their plan?

How do they work it out about how they were going to ask different people to bring in different things? Like, what's the strategy here? And Apple's saying, we don't even know, but they're saying we think this is, like they said, the tip of the iceberg, that there's an enormous plan. here to bootstrap OpenAI's entire hardware development system by stealing Apple's intellectual property, which again, again, we only hear Apple's side of it. And I think hiring people and giving them more money to do their job is fine. My thing that's the red flag here, flashing sirens, is their plan seems to have been step one, hire good people. Step two, get the good good people to steal all of the proprietary documentation they can from Apple, step three,

profit. Right. Like that, that seems to be step two. And step two is your problem. Step two is your problem. So I'm going to read from Mark Gohmann here. Like, what is Apple hoping to gain from this? Apple is seeking monetary damages and an order requiring open AI to halt the alleged conduct

and destroy any proprietary materials. Legal remedies could take months or years to materialize, but the consequences to the lawsuit itself are likely to be felt more immediately with the legal fight potentially putting a strain on recruitment and product development plans. Okay, so they want to put the hurt on OpenAI.

They want them to do a lot of documents. Stop, stop, you know, don't destroy anything. Prepare all your documents. We want to see all the details. But yeah, like German says, part of what they're doing here is also I think creating a chilling effect. Yeah. Which is

where Apple employees we're watching you, and Open AI we're watching you. Like, stop this behavior. Because I think there's two things here. There's, is there any remedy that can be done? Which might, you know, in the end, I don't even know what that would be. I don't even know how you remedy this beyond penalizing some of the people involved for committing specific criminal acts.

I don't know what that is. But what you can do is create a chilling effect and also essentially send a message to all the people at OpenAI saying, you've got to stop this, because this has been going on for a while, which to the point of how did this go on for so long, the, I think the answer is,

you let it go on for a little while because it's like, look, they're free to leave. And then you get the sense from somebody like, actually, this is more than that. They're asking these questions. So at some point it ticks over into, let's pay attention to this. But you don't stop it right away, right?

You investigate it and you see what they're doing and you gather that information, which allows them to have the information they can use in the lawsuit. So I think it went along for a while as like, look, this is just how business works and we've got to deal with it. And then I was like, no, this is starting to feel weird.

And then you get some information that says, oh, do you see what they're doing here? And then you have to make the decision. Like, we've got to go to the lawyers. And then you've got to do a full investigation and gather all that information and then you file the lawsuit. So I think it just took time.

But I think one of the things that they probably hope to get out of this is to send a message to Open AI that this has to stop immediately. And that Apple employees who are talking to Open AI need to be aware of exactly what's happening. And that Apple, you know, basically like Apple is not okay with this and Apple is watching you. Which, you know, is, I feel like that's once you get to this point, that's what you have to do. Yeah. And I mean, we don't know if Open AI's potential hardware is a threat to Apple. Like, no one can know that right now. I think at a time it felt like it would be more than it, I think, feels now. I think especially like Open AI had such a weird week last week where they seem to just, in their new desktop app, devalue standard chat GPT. And it's like if you're going to, if you're devaluing to chat GPT, I don't know what your hardware is for anymore, right? Like if you're now... Erigants and lack of focus.

Like, you're a code and work tool. Like, I'm not sure what we're doing now. But if there is a scenario where Apple believes that OpenAI could be a threat to them, this also helps that, right? Like, if you can make them start over again on their hardware, it's not bad. It's not a bad result. Right. Johnny I've coming in and having a coffee with Sam Altman in San Francisco.

and, you know, discussing vaguely that they're going to make, you know, what does ATP say, an egg that, an AI egg that sits on your table? Yeah. Like, that's one thing. And I do think AI in general, Apple's most important product by far is the iPhone, and there's this risk, I think, of AI making, if AI makes apps irrelevant and totally changes our interaction with technology, the, the, I still am very bullish on the idea of the smartphone because you need a screen because screens are helpful, we have eyes. You need a thing with a battery

that's got a good internet connection that's got a powerful processor in it. All of these things are true. So that, I think, does Apple in good stead, even if Apple's advantages as a platform owner are reduced by the rise of AI, which is a debatable thing, but it's possible. So Open AI being involved

is a threat. I'd say, though, once they hire 200 of your hardware engineers and steal a bunch of your secrets, they're a bigger threat, right? They're a bigger threat. Because then it's not, they're making a little gold egg

that is going to be a luxury item for Johnny Ive that's nobody going to really use and is not a threat to the smartphone. It becomes, oh, no, they've just decided they're going to knock off the iPhone. with an open AI iPhone that is AI first in an era where that might be something that actually does harm or replace the iPhone. And that's serious and they should take it seriously. But I don't really get the argument, especially that I saw on the Wall Street Journal, that is, oh, Apple's just doing this because they're scared.

Or, I mean, Ben Thompson did a version of this too, which is like, Apple's doing this because they feel threatened. It's like, okay, let me rephrase that entirely by the tone of my voice. Apple's doing this because they feel threatened. It's not like when you're threatened, when you feel that somebody who's got lots of resources is systematically taking your stuff in order to beat you, you fight back. Yeah, a threat doesn't mean that you failed.

Like, they are threatening them. It doesn't mean they're going to win. It doesn't mean that Apple's going to lose. But like, they are being threatened by Open AI because they've stolen stuff. stuff and they've taken hundreds of people. What's the counter argument here? What's the counter argument, which is, oh, you know, if Apple were so cool, they'd just let

them steal all those secrets because they're so confident that it doesn't matter. And like, that is not it. I'm sorry, that's not it. I am willing to believe that a bunch of this stuff is going to fall by the wayside because there is the eternal argument. We're just hiring smart people and they have their right to change their employers and work on a new project and maybe they feel unhappy with how the hardware business was going at Apple. Maybe they feel unfulfilled.

I think one of the great stories of Apple in the last 15 years is that there's been a bunch of people who've worked at Apple for a long time and made a lot of money and have been making the same iPhones for all this time. And they're kind of bored and they'd like a new challenge. And Apple can't keep them entertained some of them. And the money is, and they don't, Apple doesn't keep paying them ridiculous amounts of money like meta and OpenAI will pay them so they leave. And I think that that's true.

and I think if those people want a new challenge and more money, they should be able to do it. I don't have a problem with that. But you've got to unpick that from the rest of it, which is the systematic choice to, I mean, again, like, what part are they supposed to let roll off their back? What part?

Is it the part where they've gone to their suppliers and lied about Apple says it's okay for you to give us this secret information? Is it when employees are induced to spirit documents and possibly even hardware out of the company in order to show it to the competitor so that the competitor can build a business faster? Like, at one point are they supposed to say it's okay? I'm sorry.

Like, I don't deny the fact that this, that Open AI could potentially be a threat. I would argue, and I'm sure Apple would argue, that they're made much more of a threat by a systematic attempt to steal Apple's intellectual property, right? Like, there are trade secrets. That is, that's way more threat, then game on, play fair,

take those people and figure out how to build a better widget, right? Like that's not what's happening here. So allegedly. Yeah. Is this the end of Apple and OpenAI's relationship? Sure feels like it. I mean, it kind of felt like it had happened anyway, right?

Because Chad GPT was so devalued in iOS 27. I wonder why. Yeah. It's one of these things. where maybe they didn't need it and they also had the opportunity to get rid of it and nobody wanted to keep it, but they have a contract in place which says they have to keep it. So it's labeled as this

extension. You know, actually saying that as well, right? Like, Mark Gorman was talking about extensions so much, right? Right. Maybe someone at OpenAI told him that we're going to be an extension now. But what all extensions

really meant was from Apple's perspective, we're getting rid of Open AI. and so we're like putting them in a box. As far as I can tell, they have an exclusive with Open AI for two years and the moment that it goes off, that extension thing will get populated by other providers.

And although the way Siri AI is implemented, I'm not sure it's necessary. I don't know what you'd need it, what you'd use it to have anything. Like you can use those third-party providers with other things. But I mean, that's okay. If I could give,

if I've got a Claude account and I can say to Siri AI why don't you use Cloud for this? Why don't you ask Cloud about this? That's fine and all, but it's not like super groundbreaking or anything. It's not that important. You can just run the app and do it that way.

So it's not that big a deal. But that's what I suspect is that there's a deal. It's going to expire. Open A, I'll do what it wants and Apple will no longer use them exclusively and they've got their buddies at Google and who knows who else where they'll open that up.

and, you know, I don't know. It was a moment in time when Apple was sort of desperate to have an AI story, and Open AI stepped up and said, we will do this. Here's the deal. And famously, in May, Mark German wrote that story that was obviously just given to him by Open AI, where they said, we're thinking about talking to lawyers about maybe suing Apple, which was very funny because they weren't even, it wasn't even that they were working on a lawsuit. It was like we're thinking of, maybe.

We have concepts because we're unhappy. Because it was obviously like we're just unhappy. They don't treat us right. They don't treat us right. We were there for them

and they don't treat us right. And now it's like, well, I can guess what was going on in the background now, which is that Apple's like, we're not. It's war against them at this point.

Game on. What a terrible track record Apple have here, right? Of like companies that they embed into the operating system eventually becoming mortal enemies. So we had Google. I mean, obviously they're best friends again now, but Google was an enemy for a long time.

We have Facebook. Facebook users embedded into the operating system. Yeah, it was. They became mortal enemies. And here's a third mortal enemy. I'm sure there's more. I'm sure. But that is very funny.

Well, somebody at Apple read that you should keep your friends close and your enemies closer. And they're like, great. What could go wrong with having our enemies be close? Let's have them so close. We give them exclusive access to things in the operating system. How close can you be? How? Well, yeah, but you describe it.

like social media, like that was an area of weakness for Apple. And so they were like, okay, we've got to use, we got to go to somebody who's strong here, which they hate. But, you know, sometimes that happens. And that's what happened two years ago with them. It's a real ramification of them having a complete belly flop on AI, right? That they had to go to open AI two years ago. And then, you know, maybe if they would have been more at the forefront, less of these people would have left anyway, right? Like, I think that there is an other people in Silicon Valley that they want to do something exciting. And this feels new. Like I said, I think, I think a level of it is boredom. Like, I mean, seriously, you've made a lot of money at Apple, although you see now that the AI people are now making way more money than you are. Yeah. Yeah. You've, um, that's what I said about like Alan Die and stuff too. It's like you, in this case, you've been working on iPhone. You know, you did iPhone six and seven, eight, 10, 11, 12. Like, it's just new iPhones. So it's samey, even if you're getting challenged.

Like, and you're at the top, right? For some people, it's like, this is an industry defining product and I get to work on it every year. And, you know, and I'm making good money even, you know, and I'm happy. And other people are like, this guy who is not as good as me is now getting paid what to work at an AI company. And I've been making the same stupid iPhone for the last, you know, I want to do something new. And, you know, people are different like that. So I think that that is part of the effect is some of these people. And, you know, I'll, I will say Apple criminally underpays some of their people.

They do. Yeah. They do. They've always done that. Not that they don't make a lot of money, but in Silicon Valley terms, Apple has not ever been generous. Let's put it that way. Not that they're like super cheap, but that they're not generous. Like, I wouldn't say that Apple has ever been at the top of the pay scale in Silicon Valley.

You, you take a hit in order to stop selling sugar water and come with me and change the world, right? Like, that's part of Apple's brand from way back is you, you're at Apple. You're in the big leagues. And, you know, if you want to go work at XYZ company and make a little more money, go ahead. But you're not going to change the world. and how long is that going to last and you're not going to be on the cutting edge and this is where it bites them

is that suddenly they have incredibly deep pocketed people who are doing interesting things and maybe giving those people a new challenge and that's how you get here and again, I have no complaints about those people saying why wouldn't I go for a new challenge and more money and work at OpenAI or IO?

Like, I totally get it. The bigger question is when you're asked to bring Apple's confidential documentation with you as a part of the systematic exfiltration of those documents. That's not so good. This episode is brought to you

by ExpressVPN. Watching TV without ExpressVPN is like trying to travel the world without a passport. You're not going to get very far. That's because streaming services like Netflix only show you a tiny fraction of their huge global library of content. When you use ExpressVPN,

you can change your online location, can change where streaming services think that you're located. They have servers in over 105 countries in all 50 United States so you could gain access to thousands of new shows and never run out of stuff to watch on services like Netflix, Disney Plus, YouTube, and more. But there are loads of other reasons to choose ExpressVPN over other VPNs. It's at its lowest price ever with plans starting at just $3.49 a month. That is only 12 cents a day.

It's super easy to use. Just fire up the app and click one button to change. change locations. The blazing fast speeds mean you can stream in HD with zero buffering, and it's compatible with all your devices from phones, laptops, smart TVs, your Apple TV, and more. This means that I, here in the UK, could watch the Godfather and Canadian Netflix, or Sherlock on French Netflix, or Pulp Fiction or Brazilian Netflix. So these are things that would be available to me in other places, just by opening the app, selecting a location, tapping one button to connect, and refreshing the page to get access.

So be smart and stop paying full price for streaming services, but only getting access to a fraction of their content, and get your money's worth at expressvn.com slash upgrade. That is, EXP, r-es-svpn.com slash upgrade. We're going to do that one more time. EXP-R-E-S-V-PN.com slash upgrade to get up to four extra months of ExpressVPN. Our thanks to ExpressVPN for the support of this show and all of Reli.

Room around up time. In his power on newsletter, Mark German lays out the next few iterations of Apple Silicon with some information of what we can expect from them. So Mark begins with, I would say, a fun tidbit that Apple's silicon's AI capabilities actually stem from the work that they did for self-driving capabilities in their canceled car project. This is the best. This is the best Mark German content. This is the best. This is the best. This is, and it goes to the fact that Apple is a great reuser of technology.

is so ambitious and yet relatively small for their ambition as a company that they do they do so much reuse. Obviously, they reuse their operating system fundamentals across all these different platforms. And then you hear stories about products that come out of very unlikely places. I had a story that I can't repeat about where the Mac Studio came from, but like, unlikely place. And then they're like, well, I guess we'll just do a Mac Studio because that was sort of like it, it, It didn't, one thing went in wrong direction and they're like, well, what do we do? We've got this thing now. This is like that.

This is Mark German saying, the reason Apple's been doing so much AI stuff and the reason that they do the neural engine and all of that is mostly because they were working on level five self-driving, which doesn't exist and may never exist for the Apple Car project. And it's just, it's just a fun tidbit of, but again, you've got all these Apple Silicon engineers and you're asking them to build you chips. And then it's like, never mind. going to do. Well, what do we do with this? And the answer is, well, we're using that in all of our other products too, right? Like, and it ended up giving them, again, when we have this,

idea that Apple is so behind an AI, like Apple's been way ahead in building AI into their chips, actually. So this is an area. But I love that detail that it's, it's the car, it's Project Titan that led to a lot of this. It's funny, really, how there was so many years, I'm sure you could go back and listen to odd episodes of Upgrade, where it's like, what is the neural Engine 4? Like, you know, they always would talk about this massive part of their silicon and all their devices, you know, like in the phones and the iPads. It's like, why would you need a specific part of the chip to do that? Like, it was always very confusing to me. Not that it's necessarily much clearer now, but at least understand that it can be used for other things. So Mark reiterates

kind of a little bit about what we can expect and in details from stuff going further. So we're going to the M6 coming later this year, just the base M6. The M6 Pro, Max and Ultra are gone. They will not exist. Apple will move directly to the M7 family. The M7 chip will arrive in the first half of next year, followed by the Pro and Max in the fall, and the M7 Ultra in 2028. Apple had already been planning for the M7 line to have big leaps in AI performance, and the result of the work that they did, is why they decided to bring the lineup forward. Mark says that the M7 Ultra, the 2028 chip, is a big deal for them.

It will, quote, bring it closer so the class of dedicated AI accelerators such as Nvidia Corps's Blackwell. I mean, I only know a little bit about this kind of stuff, but that is significant in what can sit on someone's desk. Yeah, Apple is putting their foot to the floor on their ability. Like, look, they were caught by some. surprise. We all know it. And yet at WWDC, I got to have an AI experience that included spending

time in front of that four Mac Studio node with a terabyte of RAM with a, and that was the M3 Ultra, right? They are leaning into that. Right. Because they're like, okay, you know, they're not, I mean, they are building cloud infrastructure. And he talks about how they're, they're going to, these ultra chips are part of their infrastructure plan as well. But like, they, they, they, they, see real opportunity, especially when you talk about the cost, like, the general perception is all of these AI things, all these AI companies are losing money. And like, even if you're paying them $100 a month for your AI tokens, like, you're probably using more than that. They're probably losing money on you because they're, that's the, they're in growth mode now and they're willing to

lose money. But at some point, there will be a reckoning. And all of those apps you vibe code for $20 are going to become that, so you're going to vibe code for $200. And it changes the dynamic. And I think Apple's pressing an advantage and looking toward the future and saying, ah, but what if we could run really great models locally on one or two or three or four computers that would cost thousands of dollars, but then it's all local. And then for some uses, these models are getting so good and the chips are getting so good. for a lot of uses like on a mac, on a well-powered Mac,

that you don't even need to go out to the big model most of the time. Like that's really interesting, right? Like that's, and so they're pushing there. Whether it's realistic or not, we'll see. But, you know, they're saying, according to Mark German, like M7 Ultra is a really serious AI processor. And they're viewing the high end of the Mac market has been redefined as being for this.

basically. So the M5 Ultra is still on the way, as a previously been reported. And Mark says, they expect that they will be moving their kind of Apple intelligence infrastructure to the M5 from the M2, but they are looking forward to putting BM7 in its place.

So just to be clear, like, there will, at some point, they will decide to ship finally the M5 Mac Studio, including the M5 Ultra. It will happen at some point. you know, you can see that they're not super enthused about it because those are going to be very expensive because they have a lot of ram in them. But at some point, they will do it because they're a generation behind now.

Two generations for Ultra, but, you know, that's just how it's working for the Ultra chip. So it is coming. We, we were talking about M7 Ultra. I just want to be clear, like M3 Ultra is shipping today. M5 Ultra is coming and then M7 Ultra is the thing beyond that. But this goes back to what we were speculating last time, which is that Apple, pulled the M7 line forward because they feel like there's a major leap in performance there. That's why they did it.

Apparently the M7 Ultra will be able to support up to 1.5 terabytes of RAM, which, I don't know, man. I was reminded, I did not remember this, that the 2019 Mac Pro could have 1.5 terabytes of RAM in it. You could spec that. I did not remember that that was a thing. So the last Intel one, you could do that. I guess, I don't know why anyone. one would have done it then.

For kicks. Like it has more of a use now, I think. And I think it does have more of a use now and more applications than it did then. With the M7 Ultra, you'll finally be able to hook up for $10,000 Mac Studios and get six terabytes of addressable RAM. So congratulations. Imagine what you could do with that. But Apple aren't stopping here.

So the M8 lineup also due for 2028 will continue the AI performance push. and these are planned to be 1.4 nanometer chips that will also deliver efficiency as well as power. The nanometer thing is interesting to me because I think it's either this year's iPhone or next year's iPhone that'll be two nanometers. I don't know why I just thought it stopped at three, like that they got the three in.

No, it keeps going. It cannot be smaller. They're going to go down into Angstroms pretty soon. They're running out of nanometers. They're going to go down to angstroms. there's a i'm not sure if this is a joke or a serious thing i don't i don't know that word and so angstroms angstroms is a a measure of molecular atomic and molecular scales okay so that's what will come once they start once they go below one nanometer right like it gets pointless to talk about anymore so david chubbs like it's marketing it is but also like nanometer is marketing the problem is that

under one nanometer, you're talking about like 0.8 nanometers. And like, so two nanometers is 20 angstroms. A nanometer is 10 angstroms. Anxtrum is a measure that goes below nanometers. And so that's where we're headed is angstroms, which is if you took chemistry or physics, uh, chemistry, um, you know about angstroms. Yeah, I know.

That's why, that's why you thought I was pranking you when I mentioned angstroms. I really did. I was like, is this a Star Trek? Is he doing a Star Trek to me right now? It's impossible to know something. Yeah, they're going to, the new chips are going to be measured in Quatloos, Mike. No, see, I've heard you say that word to me before. Quatlews.

So that's how I know. That's how I know that one's the Star Trek. Yeah, that's right. I mean, they're really going for it with chips again. It's going to be really interesting. It's their advantage. They've got this M6. What will the M6 go into and then how quickly will it be replaced by the M7?

And I still say why you even do the M6? but we'll see what they do with it. I think they've already got products designed for the M6, and they're like, you know, an M6 iPad Pro and M6 Air, you know, let's just get it out there. There's no point in, why not turn that over? There's no point in waiting for those, and then they move on. But that's my guess is that they've literally already got designs going for those,

and they might as well ship it because they're looking out to the M7. And like, I don't want to wait that long, probably, to ship an M7. after the M5 because the M5s are out now, right? Yeah, yeah. And if M7 isn't going to come

until June, July, whatever of next year, a year from now, let's say. Yeah. And they want to turn over those MacBooks in the fall,

maybe, or in the spring. The M7's not going to be ready. So they may have just decided, why not? And so, yeah, we'll see. Speaking of M5,

my laptops coming this week. Yay. I thought it would come faster. I hoped it would, but you can never be sure. But that would be super great. I'm very excited to get that computer. There have also been a variety of reports over the last week from various leakers that the pro phones are going to get thicker and heavier this year. The numbers differ across the models for obvious reasons, right?

But one report suggests that the 18 pro could be up to 2 millimeters thicker than the 17 pro, which is quite significant, and that the Pro Max could weigh in at 240 grams, which is the weight that the Pro Max was before they switched to titanium. So it was like the heaviest that that phone got. It said that there could be a few reasons for these jumps.

Bigger batteries. A new vapor chamber made from stainless steel. And my theory is, does this come from, there's also been rumors of big camera improvements? And I wonder if they need maybe a bit more space in the case for a camera. improvements. So they're like, what else? Space in the case. Because I don't really know

why you would put a bigger battery in these devices. Like, I know people always want battery, but do they want it at the expense of a heavier, thicker phone? I'm not sure. Like, I'm not thrilled about the iPhone getting thicker and heavier, personally. Like, I am a thin and light person. Like, that's, I like that in my devices, because I also feel that I don't need more battery from my 17. Pro to the 18 Pro. Like, I'm happy with that.

So it's going to be interesting. I mean, we're never going to know exactly why they did it, but maybe once we get to these devices are shown off, we can piece it together. But if the 18 Pro was two millimeters thick and a 17 pro, that is quite a jump. It's a lot. It's a lot.

Like, what is it now? It's like 8 millimeters thick or something like that. So like that it's significant. I wonder, it's funny. I wonder if Apple's finally like, you know what? People don't care so much about thin. Oh, well, because of the air. Yeah.

You guys don't want a thin phone? Fine. Fine. We'll make it thicker. See if we care. And maybe they reach the point with the 16 where they're like, okay, we have battery issues, we have heat issues. And we're trying to get this thing to be thinner and lighter. And people want battery and they don't want it to be hot. And like, maybe we just need to rethink this and change our priorities a little bit. And the camera, right?

Like, there is that issue too, which is like, if you're really pushing the camera and you need space for the camera bump, at that point, like, you might as well take the space. Yeah. Like, if a camera bump requires it to be, you know, a certain level of, sorry, iconic plateau, to have a certain thickness, then why are you even trying to make your phone thinner everywhere else? Because all you're going to do is make the prominence of the plateau greater. so why even bother? And that's why I think the camera might be the reason, because it's like, if the camera is really going to be a lot better,

if you make the phone a little bit thicker, you need a bit more space, then go for it. And I think people will always take that, right, as a reason. But you'll never know that that's the reason. They won't say that, right? Like, we made your phone thicker so we could put more camera in it

because it's what variable aperture is one of the rumored things. Well, and keep in mind, like, what's the pro phone for? They may also feel like the 17 is becoming a much more of a mainstream phone. Yeah. That regular people are going to spend less money and get and be happy with. It's got a lot of pro features now and all of that. And maybe part of it too is like the pros, they want more.

They want more battery. They want more camera. And it's going to be thicker and heavier. And that's okay. Because if you don't like it, there's an air. There's an 18 coming in the spring maybe. And there's a 17 now. And like, I can see the philosophy here, which is not.

every Apple, not every iPhone must push on the same dimensions, right? And so maybe that's the answer is like, look, if thin and light is what your priority is, there are other iPhones for you, you could try. But like for this one, where it's about performance and camera. We're going to make it thick. We're going to make it heavy, but it's going to be awesome. And this is the thick, heavy, awesome phone that does everything and is basically a supercomputer. And if you don't want it, there's an error, there's a, there's a non-pro, go to town. Yep. This episode is brought to you by Century.

Applications break in all kinds of ways, from crashes to slowdowns to regressions to the stuff you only really see when a real user encounters it. Century catches all of it. That's S-E-N-T-R-Y. With Century, you get traces, replays, errors, profiles, and the details around them like stack traces, commits, releases, and a developer who broke it, all in one connected view, so you're not jumping between tools trying to figure out exactly what happened. Century shows you how the request moved, what ran, what slowed it down, and what the user

saw, and then Cia, Century's AI debugging agent will take it from there. It uses all of the Sentry context to tell you the root cause, suggest a fix, and open a pool request. It also leaves your PRs for you and flags breaking changes with a fix already there for you. This kind of stuff I think is fantastic because I know that what developers want to do is new features. And they want their bugs fixed as quickly as possible. You do not want to be spending tons of time digging through code trying to find fixes. And to have all of this stuff collected for you, so what's crashing, the regressions, the slowdowns, the stuff that your users are seeing. And then to have it all connected with what the bugs are and have fixes suggested, this is going to get you back to the world.

of developing new features, pushing your product forward quicker. I think that's amazing. But go and try it for free today at s-en-tr-y-o and tell them that we sent you. Century have a free dev plan, and listeners of this show can use the code, upgrade 26, to get $100 in Century credits if you're a new user. That's S-E-N-T-R-Y.O, or thanks to Century for their support of this show and Relay. It is time for some Ask, Upgrade Questions. Darren Wright Sinner says,

With all the foldable iPhone leaks showing only two rear cameras, everyone's assuming that Apple is dropping either the ultra-wide or telephoto. But what if they're not? What? If the cover display has a selfie camera, then when the iPhone is unfolded, that camera effectively faces backwards, giving Apple,

given Apple has steadily widened the selfie cameras field of view over time, could they use it as the ultra-wide, leaving the two rear cameras as the standard? and telephoto lens. Man. Darren's on some galaxy brain stuff, Jason. I don't agree with it. I like the way you think, Darren.

But I like the way you think. I mean, it's true. That is too... It would be an ultra-wide, in theory, yes. That is two rear cameras. I guess my question is, when they say there are two cameras on the back, is that what they mean?

That there's a standard on one side and an ultra-wide on the other, or there are two when it's folded? I think this film will have two ultra-white cameras is what I think will happen. Wow. Because the selfie camera, you get ultra-megal-wide. The selfie camera and the ultra-wide camera on the back of the phone are doing,

they're needed for different things. And so I don't, because it's like saying the iPhone Air has an ultra-wide camera, right? You just got to turn the screen around. Just not look at it. Like I get how it can work, right? But I don't, I just, that seems so inelegant, right, for them. you want an ultra wide, just open the phone up. Like, I don't know.

Maybe. Maybe. I wonder if, also, it might be a thing where, like, the selfie camera is capable of, I don't know, can you do this? Capable of shooting in both directions. So, when it's open, it's also the selfie camera. Like, they only have one selfie hardware. Does that make any sense? Right? So it shoots out.

When it's closed, it's shooting out the back. And when it's open, it's shooting out the front. like it's got a mirror or something that turns it around. Man, I don't know about that. Maybe, but that's like... Let's just put it in there. It's got four cameras. It's got five cameras.

Are you getting it yet? It's got all the cameras. This one bounces off six mirrors, all right? Everything has camera. The touch ID, it's not touch ID. It's picture of your finger camera. Ha. Yeah.

I don't know, man. Darren, I like your thinking. That's what I do. I'm not sure that it's going to result that way, though. Eeli writes in and says, related to the John process situation we were talking about last week, if a journalist isn't allowed to pay

a leaker for information, what would be the reason for someone to leak? Now, we've spoken about this in the past, but I think this is a classic upgrade thing that we can refresh people on. And I will say, inducement is the issue here. I think you could say

that if somebody gave you great information that led to a great story, and then you gave them a gift or something, it would be hard to prove that that was inducement. It was the thank you. I don't think the payment is the issue if it's not considered inducement. So I don't think it's the actual payment.

It is the inducement. I'm going to do something to make it worth your while to steal things. So you're doing it for me is the problem. What's the reason, though? You're right. Classic upgrade question. Ego.

Bortem. I think a lot of his ego. Really exciting. You know something nobody does. does, you can't tell anybody. But if you tell this guy, it's going to be in print and everybody's going to be talking about it. And you can sit back and go, I did that. I think that's a lot of it. Sometimes it's about targets. We always say on this show, consider the source. That's part of it, too. It's like, ah, this Mark German story suggests that all the designers at Apple are great and everybody else is

dumb and won't listen to the designers. I wonder who told him that, right? Like, some of it is that. Or there's a thing that's going on that you don't like. And so you leak it because you wanted to get out there so that maybe people get riled up and say, why are they doing this? This is so dumb. There are lots of reasons. But I do think that some of it is just the ego trip in various forms of I know a secret. Nobody else knows it. I can change the world by saying it to this guy because then everybody will be talking about it. And I can lean back and go, that was me. I did that. I think that it's one of these things where there are many reasons and it isn't money related most of the time. Right?

Like, someone's got an axe to grind, as you say, right? They've got, they want to see their name in lights. They like the idea of being close to someone in this world because they think it's exciting. It's enticing to them in some way. It

doesn't have to, and I'm assuming, it's probably very rarely about money. Yeah, extremely rarely. For what we're talking about. Almost never. Almost never. almost never. That's why the allegation of what John Prosser may have done is shocking because, you know, Ramasjadhi is saying he said he'd pay me to do this. And that doesn't generally happen. No. Mark German is not writing checks to people. Like, it's not, that's not how it works. No. No, people want to talk to Mark. And I should say most journalists, their organization's policy is you do not pay sources, right? Like that's also part of it is you don't. And I guess that's because you don't want to get into the scenario of when was the money offered, right?

And it calls into question their reliability, right? They make things up to get paid, right? That's not going to work. So it's just, it's just not done. Because I can imagine a scenario where it's like the bigger the claim, the more money. So you kind of just like start to invent things. So it's not that they don't have motivations. It's that they're not financial motivations.

They're other motivations, their ego and all those other things that we talked about. Hank writes in and says, with all of the positive stories of AI use in coding and its ability to build software and find bugs that developers can't, should our expectations on the quality and reliability of software increase accordingly? Here's the thing.

I think that, I think this is a real double-edged sword of, and it says it's a terrible thing to say, but it depends. Because people that are good at development, using these tools should increase the quality of their output. The problem is there's going to be way more software from people who don't know what they're doing. Yes.

Releasing that to the public. So I would say from, and I've noticed this in apps that I use where I know people are using AI tools to help them make their apps better, the apps are better. But I also have had experience with apps where these apps are bad and I know where they've come from. Yeah, they're vibe coded. Yeah. They look like a vibe coded app. They act like it. They're weird things. And can they even fix the bugs? Are they even capable of doing that? Because they just have to ask their agent to fix their bugs for them. Yeah, I have seen some things that are very obviously vibe coded by people who are first time programmers and they put them up for sale in the app store. And maybe it's just me, but like, what are you doing? Like, I decided to put double

ender up, not for sale, for free. Yep. Because it solves a problem I have that I know other podcast editors have. And I used to distribute this shortcuts action that I, or automator action that I had secretly to friends, because I didn't have the authorization to distribute the binary that it was based on. And so I really like the fact that this one is entirely the product of me and, and, uh, clawed the albino alligator. And so we can just put it out there. And if it helps those podcast editors sync their files, great. But like, I'm not going to charge for it. And I had somebody say, oh, it did this weird thing. I'm like, oh, no, can I fix it?

And I fixed it. And that's great. But like, yeah, my level, I'm not a developer and my level of confidence that I'm going to be able to provide like ongoing support for somebody who paid me for this product. It's like, that's not even why I did it. And I'm not going to do that. So my answer to Hank is your answer, Mike, which is both. It depends. Like there are like overcast just fixed a bunch of bugs because Marco was able to point an AI at his code base and say, can you find these bugs?

and it did and Marco fixed them. But also for every one of those, there's probably like eight vibe-coded apps that are going to come across the transom and they're going to be garbage. And you know what? Garbage, yeah. Garbage, if garbage solves a problem,

great. I mean, my video, my video app that I'm doing for our video editing, I mean, it's kind of garbage. Yeah. But it's just for me and it does the job.

And it's not meant to be more than that. Same. That app that I'm building to extract audio clips to use in the podcast is ugly, really. But it works, by the way. I have a name for it now, sound bite, but bite with a Y. It's referenced in the Byte shop. Pretty good, right? It's pretty good, pretty good.

But like even today, right? So like I was listening to App Stories and they were talking about Fable 5. And with my Claude plan, I have some access to it. and like, ah, what could it be like? Right? So I point in Fable at this project. And it's like, oh, here's a bunch of security vulnerabilities. Yeah. Like that the original model didn't see?

It's like, I don't care because I'm never distributing this product. This is the classic line, which I'll say, it's what John Syracuse told me, which is every time a new model comes out, just open your code base in it and say, find bugs. Yeah. And then after it finds those bugs and you fix them, just tell it again, find bugs. And just keep doing that. Because a new model, I did that with Fable. I was like, hey, Fable,

look at this app and tell me what the issues are. And it's like, oh, yeah, well, this is a case that'll be bad. And this thing could break. And you should probably change this thing. I'm like, great, let's do that. Right. But it's funny, funny time. You know, I just realized the logo for Double Lender is, this is a finder reference, right?

It's a, the logo for Double Lender is based on the alert dialogue box. in classic MacOS, which was a face in profile and then a big exclamation point. Okay. And so I wanted to sort of like reference Susan Kerr when I did it. And then I thought like multiple heads, they're almost like Eastern Island heads. Multiple heads would also be interesting because it is about multiple voices. Yeah. And then my final realization was that, um,

I could modify them so that they were even less like Susan Care, but also actually spell D.E for double-ender. And that's where we ended up with those guys. So it's just these blocking-based guys. It reminds me in the colors, though, a very findery. Yeah, I love it. I mean, I just, again, ATP did the whole thing about, like, icons. And I'm like, you know what?

I like my stupid. Yeah. of Easter Island statue guys who spell DE because one, I don't care, two, I'm having fun making a reference and three, did I mention that I don't care? So anyway, I like it, but it is ridiculous.

Conformist has an icon too, but it's my video conforming tool, but it's just, it's basically the old. It's another one of these guys. And it's just, it's the old one of that. Okay. And Jack wants to know,

has Jason read any nonfiction books by Tom Holland or Dominic Sandbrook. I haven't listened to the rest of history yet, although both you and might keep making a great case for doing so. But I did read Tom Holland's book Rubicon on the late Roman Republic a few years ago and enjoyed it a lot. I know Tom has written fiction too. Have you read any of these? So just before you answer, Jason, I have to say, Jack, if you have read Tom Holland's books and enjoy them, my word man, why are you not listening to the rest of history? Come on. Get it together, Jack. By the way, as a part of the run-up to The Odyssey, not only is the Rest of History doing The Odyssey this week,

but Tom Holland, the historian on the rest of history, got to interview Tom Holland. So good. The actor who he introduced as the real Tom Holland, and Tom Holland, the actor was like, no, no, no, no, no, the other Tom Holland. It's brilliant. But Tom Holland, the historian pointed out that he can't Google himself

because there's no point in doing that, which is great. The answer is, I have read, Oh, now I blanked on it. I read Tom's book Dominion, which is about the spread of Christianity and how it impacted Western civilization in surprising ways. And this is a thing that he comes back to in the podcast a lot, which is how there are some, he points out the entire idea of,

there are lots of concepts that we have that are fundamentally Christian concepts or they're like fundamentally Reformation concepts. And he says the idea of personal spiritual choice of choosing whether you believe in God or not, all of these things are like straight line from the Reformation and Martin Luther. And then he talks about things that are like things that we would never consider fundamentally Christian that still come from historically from Christian concepts. I thought it was a really fascinating book.

However, I will also say it was extremely long and extremely dense and not for everybody. So one of the things I love about Restis history is that they dilute it down and just tell big stories that are fun because the books are intense and dense.

And I haven't read any of Dominic stuff, although, you know, that'll probably happen one of these days. I'll read about, you know, 70s Britain and be amused. I want to, at some point, read some of his kids' books with Sophia. Like, he does. Oh, sure. Yeah, yeah.

They referenced those a lot, and they sound really fun. Yes. Adventures in Time. Yeah, Adventures in Time. They sound like really fun books about history. Like, because whenever, you know, whenever either them have a book out, they do a series. Right. That is about that thing.

Those books are books, you can see it too. Like, they're the books they do because they're historians. Yeah. Especially before they had this money-making podcast of theirs. they're historians. So they have the books they do because they're historians. And then there are the books you do to make money. And The Adventures in Time is the Books you do to make money.

Can I make a series of children's history books? And I can use my legitimacy as a historian to get the books and then also then make money selling lots of books. And that's what Adventures in Time is. So that's very funny. Yeah, they've done some like Hero of the Seas. That's Nelson, the Aztecs, Alexander the Great Vikings. the wars. The covers are fantastic, actually,

the penguin. Marketing. So at some point, I will read these. Yeah. If you would like to send in a question for us to answer on a future episode of the show, or you have any feedback or follow-up or you have a snow talk.

Go to upgradefeedback.com. I'd like to thank our members who support us every week with UpgradePlus. You can go to GetUpgradeplus.com to sign up. You'll get longer ad-free episodes each and every week. This week, I want to talk a little bit about where we are with designing California production.

It's been an exciting week, I would say, around these parts. You can find us on YouTube by searching for the Upgrade podcast. I would like to thank our sponsors one last time for the support of this week's episode. That is FitBod, ExpressVPN, and Century. But most of all, thank you for listening. We'll be back next week. Until then, say goodbye, Justice Snow. Bye, everybody. Sleep tight.

Don't steal any trade secrets. Try not to anyway.