00:00:31
◼►
When I realized last week that this was happening, I did a small bit of math so I could do it on this week when you booby-trapped me with it.
00:00:37
◼►
So obviously we have our member specials and we've done one-off things like interview shows and stuff like that.
00:00:43
◼►
So it's actually obviously more than 700, but for the regular episodes, the not interview, not member special, not blah, blah, blah,
00:00:50
◼►
we've done 52 regular episodes per year for 13.4615 years, which is not quite as round.
00:00:57
◼►
I know it's episode 700, but it's not as exciting as the year-round number.
00:01:01
◼►
So anyway, we've been doing this for a while.
00:01:03
◼►
So I don't have the author's name in front of me and I'm deeply sorry about that,
00:01:07
◼►
but there is a listener who has been working on for like a year or two now, catatp.fm, catatp.fm.
00:01:14
◼►
This is an absolutely incredible resource.
00:01:17
◼►
This is taking the spirit of what Underscore had done years ago about running shows through a transcription service and trying to make the best of what comes out of it.
00:01:25
◼►
And obviously with today's AI stuff, it's quite a bit more impressive than it was years ago.
00:01:30
◼►
There's a handful of these like underscores like catatp that exist.
00:01:34
◼►
And this one I happened to look at recently and there are statistics.
00:01:39
◼►
So if you go to catatp.fm slash statistics, you can see that as of, I think, this most recent episode, not the one we're actually recording now, obviously,
00:01:47
◼►
but as of $6.99, total length of all the episodes, 1,421 hours, 6 minutes, and 13 seconds.
00:01:57
◼►
And catatp, by the way, like what a level up that is from like everything that's come before because the author is using AI heavily and cleverly.
00:02:21
◼►
And among other things, they had come up with a visualization of what sections of the last couple of episodes were cut from the bootleg to the released version,
00:02:33
◼►
which I thought was a really interesting visualization of what you're missing out on if you don't listen to the bootleg.
00:02:38
◼►
Now, what you're getting if you don't listen to the bootleg is Marco's genuinely, not sarcasm, genuinely exquisite edits.
00:02:44
◼►
You're getting much better audio quality.
00:02:45
◼►
But you're missing some of the nonsense and shenanigans that you would hear only in the bootleg.
00:02:51
◼►
And if I remember, I will post one such example image and link it in the show notes.
00:03:55
◼►
And I'm not going to go on for two hours like I normally do.
00:03:57
◼►
But I will just say very briefly, thank you to anyone who listens, whether or not you're a member.
00:04:01
◼►
Thank you for sticking with us for, what did you say, John, 13 years or whatever it's been.
00:04:04
◼►
It's really a genuine privilege that we get to do this together, that the three of us get to do this together, that you give us your time, which is your most valuable resource.
00:04:13
◼►
So thank you, everyone, for sticking with us.
00:04:19
◼►
And I'm curious, what do we have betas installed on?
00:04:23
◼►
And I ask this because I don't have, I have the betas on my Vision Pro, and I have the beta on my iPad.
00:04:32
◼►
And now that the public betas are out, and now that all the travel that I know I'm doing for the summer is done, I think I might do public beta on my phone soon.
00:04:55
◼►
Beta 2 was a little shaky with messages, like the messages app, sometimes like losing a whole thread and having to be rebooted or whatever.
00:05:03
◼►
That seems to have been fixed after beta 2.
00:05:05
◼►
And the only major issues I have with it now are some like behavioral differences that are making my app behave weirdly and other apps occasionally behave weirdly.
00:05:16
◼►
They did a lot of under the hood changes with iOS 27, and some of them are showing.
00:05:21
◼►
So there's going to be a little bit of a bumpy period as apps get updated for that.
00:05:25
◼►
And then finally, the biggest problem I have is CarPlay has gotten a lot of updates, and it's a little rough.
00:05:34
◼►
In particular, if your car has a hardware play pause button of some kind, like you click the volume knob in and it turns it off or on,
00:05:43
◼►
or there's an actual play pause somewhere in your car, and if that maps to CarPlay's playing and pausing.
00:05:50
◼►
The previous treatment of that was that iOS would forward that to the audio app as a remote play pause command,
00:05:56
◼►
the same way as if you tapped, you know, if you like clicked the stem on an AirPod, same thing.
00:06:00
◼►
For some reason, in CarPlay and iOS 27, they're instead treating those as audio interruptions,
00:06:06
◼►
which means, which is the same kind of event as like if you're playing a podcast and you get a phone call,
00:06:12
◼►
or Siri comes on to speak a direction to you or something, that's an audio interruption.
00:06:17
◼►
And then, you know, so it pauses your audio kind of forcefully, and then it unpauses it whenever the interruption ends.
00:06:22
◼►
Well, in CarPlay on iOS 27, it has moved those from play pause events to audio interruptions,
00:06:28
◼►
and when they end, my AV audio engine totally errors out, fails, and crashes.
00:08:27
◼►
There's a couple of things to be aware of with Siri and iOS 27.
00:08:31
◼►
Obviously, it's the whole new, you know, Siri AI thing.
00:08:33
◼►
In daily use, what that means is when you pull down on your home screen to type in an app name into Spotlight,
00:08:41
◼►
it is slower because that pulls down all of Siri, and Siri is like, let me do the world for you.
00:08:48
◼►
And I'm just trying to launch one password.
00:08:49
◼►
It's like, let me do everything for it, you know.
00:08:51
◼►
So, there are some common operations that now that Siri is everywhere, they're actually much slower now.
00:08:59
◼►
So, one of them is typing app names into Spotlight, what used to be Spotlight, to launch them.
00:09:04
◼►
The other one, which is kind of annoying to me in particular, is creating reminders with the new Siri is actually significantly slower.
00:09:16
◼►
Which is not great news for somebody trying to launch a new app that uses reminders.
00:09:24
◼►
And this is also just annoying because this is how, this is by far the way that I create the vast majority of my reminders.
00:09:30
◼►
Creating a reminder now, you know, you still say, you know, hey, remind me, you know, later today to do this.
00:09:36
◼►
It now thinks a lot more about that because it's invoking the AILM layers.
00:09:43
◼►
And it asks more follow-up questions where I think it shouldn't.
00:09:47
◼►
So, it seems like it hears me worse now, like Siri has upgraded to worse ears.
00:09:53
◼►
And so, it'll ask for clarification, like, did you say this or this?
00:09:56
◼►
And, like, before, it never asked me that.
00:10:00
◼►
It would just dictate it pretty accurately almost every time before.
00:10:03
◼►
And usually, one of the things that it thinks I said is just way off.
00:10:08
◼►
So, they're still ironing some stuff out.
00:10:11
◼►
But I think it is probably pretty safe to conclude at this point that the move to new Siri just is going to make a lot of common things slower.
00:10:22
◼►
Because it is now applying a much higher level of complex intelligence to even simple things like searching for apps and creating reminders.
00:11:30
◼►
It has never been able to do that before.
00:11:33
◼►
So, I, I do think, you know, it is, it is like any other modern AI tool in the sense that this is a significant step up from what we had before.
00:11:43
◼►
I think one of the biggest challenges that we're all going to have is just remembering to try something or thinking to try something instead of just assuming that something won't work and never even trying it.
00:11:53
◼►
That, that is by far what holds me back the most with AI and that is by far what will hold me back using new Siri over time.
00:11:59
◼►
That, like, we're all so accustomed to Siri sucking for, like, you know, over a decade that we don't even, we're not even going to think to try it, but try it.
00:12:10
◼►
Because you, you might be surprised at what it can do as long as you're patient because it's slow.
00:12:14
◼►
I feel like it's probably worth briefly following up on my weird, I'm not getting tap taps from Aaron on my watch.
00:12:22
◼►
If you recall, it was working when Aaron sent messages and other conversations like in group chats and I was just constantly missing notifications from her.
00:12:31
◼►
And let me tell you, of all the people I want to miss notifications from, both selfishly and unselfishly, she was the last on the list.
00:12:38
◼►
And, uh, somebody at some point had suggested, Hey, do you somehow have her conversation muted?
00:42:20
◼►
Thank you so much to Claude for fixing my code and for sponsoring our show.
00:42:24
◼►
Let's do a little bit of follow-out, if we can, please.
00:42:31
◼►
In episode 627 of Upgrade, which as we record this, is not this week, but last week, there was a discussion about FileMaker.
00:42:39
◼►
And I was aware of FileMaker as a thing, and I was aware that it was part, was, or perhaps is part of Claris, which is itself part of Apple now.
00:42:49
◼►
I remember Claris works very fondly from when I was a kid and using Macs briefly in school.
00:42:54
◼►
But I was hoping that probably John could explain to me what the heck is FileMaker.
00:42:59
◼►
Like, I don't feel like anyone has actually sat down and explained what the heck it is.
00:43:02
◼►
Well, you wrote in the notes here what was FileMaker, and I put a link into it.
00:43:39
◼►
You'd define a schema and here's where your data is and you would get things into it and out of it.
00:43:43
◼►
And you'd make like a GUI front end for it.
00:43:45
◼►
And it was used by like businesses to sort of, this is days before the web, right?
00:43:50
◼►
Say you had like an inventory system for your small business.
00:43:53
◼►
You had to keep track of how much whatever stuff you have or you wanted to like roll your own system for keeping track of people's hours or something.
00:43:59
◼►
You can think of it kind of like HyperCard because HyperCard also had, you know, a similar type of like you can make GUI apps, but without being a developer.
00:44:07
◼►
Like, I don't know, like maybe like a cross between HyperCard and Excel with more of a data focus.
00:44:15
◼►
So lots of businesses had FileMaker databases that they ran their businesses off of.
00:44:21
◼►
I never used it in like, in anger, like in a real actual scenario.
00:44:25
◼►
I was just, I was just playing with it as a kid.
00:44:27
◼►
So I never worked at a business that used it or wrote it for a business.
00:44:30
◼►
So I don't know how it actually performed and why people still might use it in the modern age.
00:44:34
◼►
But in the days before the web and in the days when you couldn't do all this from a spreadsheet, it was like a little bit more than that.
00:44:40
◼►
And the days before, like, because a spreadsheet is not a good UI to like have employees using stuff or whatever, although plenty of businesses still do it these days.
00:44:47
◼►
But yeah, that's my impression of what FileMaker was like for people who actually used it.
00:44:53
◼►
And I guess it still exists because, hey, if your business runs on FileMaker and they keep making new versions of it, why would you ever change if it's still working for you?
00:45:00
◼►
Like there's all these stories about, you know, someone wandering around in like 2010 and they see like a Mac SE in the back of some business running some FileMaker database that their whole business runs on.
00:45:09
◼►
They're like, how is that computer even still running?
00:45:10
◼►
It's like, well, we need it because it runs FileMaker.
00:45:12
◼►
So I'm glad it still exists for the people who like it.
00:45:14
◼►
But these days you'd be like, why don't you make a web app?
00:45:17
◼►
Why don't you, you know, there's a million other things that you could do other than a FileMaker database, but it's still out there for people who want it.
00:45:23
◼►
I was glancing at the marketing website, which is Apple, like it's very Apple-y, except not, which is very unusual.
00:45:30
◼►
This is claris.com slash FileMaker, which is very interesting.
00:45:34
◼►
And I don't have an exact, I'm looking at it, but I can't find exactly what it was I was seeing.
00:45:38
◼►
But it was, they had some discussions about how you can do like agentic coding with FileMaker, which is so mind-bending to me because I perceive FileMaker, seemingly unfairly, as something from like the early to mid 80s.
00:45:51
◼►
And they're talking about stuff that's relevant in the last like 80 days, you know?
00:45:55
◼►
And it's just astonishing to me that this is still a thing.
00:45:58
◼►
I see the thing that said like FileMaker Cloud.
00:46:12
◼►
It was like a consumer-focused, friendlier revision of FileMaker.
00:46:18
◼►
And I think that just didn't take off.
00:46:19
◼►
And they're like, oh, nevermind, our customers just want FileMaker.
00:46:21
◼►
I have no idea what FileMaker is today.
00:46:23
◼►
And the website really obscures what it might be.
00:46:26
◼►
But it kind of reminds me with the agentic stuff, it kind of reminds me of like, we haven't talked about this in the show, but a lot of other people have been talking about it.
00:46:32
◼►
The feature in the 27 OS is where you can describe a shortcut.
00:46:35
◼►
You open the shortcuts app and you just like click a thing and you say, I want a shortcut that does X, Y, and Z.
00:46:41
◼►
I imagine that it's possible to do something similar with FileMaker because it is like shortcuts, kind of like a higher level abstraction for data with a certain number of operations that you can do that the FileMaker knows about.
00:47:14
◼►
Clarice is investing so you can build with the leading AI coding tools and deploy straightened FileMaker securely, permissioned, and ready to run.
00:47:32
◼►
I would love to see, like, screenshots of this.
00:47:34
◼►
I haven't gone spelunking through the internet trying to find them myself.
00:47:37
◼►
But I'm curious what this looks like both on the development side and on the, like, production side.
00:47:42
◼►
Because presumably on the production side, like, for whatever these apps are that this makes, presumably they can look like almost anything.
00:47:48
◼►
But I'd doubly love to see what it looks like on the development side.
00:47:51
◼►
So reach out if you're a FileMaker person.
00:48:06
◼►
This is the app that we kind of found around our circles of maybe, I don't know, three, four years ago, that was a replacement Dropbox client for Mac and Linux.
00:48:18
◼►
And I don't know if it worked on Windows, but it works on Mac and Linux.
00:48:20
◼►
And the idea was basically, like, you know, the Dropbox app keeps getting bloated and crappy and just has a million things in it.
00:48:49
◼►
Yeah, and they were playing tricks like they were trying to get you to enter your root password by, like, faking an admin permission dialogue.
00:48:56
◼►
Like, Dropbox, at many times of its history, has acted like malware on your Mac.
00:49:04
◼►
And Maestro is a client that, it's open source, and it uses the Dropbox API to basically provide what Dropbox used to be, like, a basic sync client.
00:49:17
◼►
So, you'd have a folder at HomeU Dropbox, and it would sync the content of that folder to Dropbox, which is exactly what Dropbox used to do.
00:49:27
◼►
Oh, that's another factor we didn't mention, which is that Dropbox was sort of forced to move to the file provider thing because Apple made that API for them, and the file provider API was worse than the old one.
00:49:37
◼►
That is another reason people preferred Maestro.
00:49:39
◼►
Like, I don't want the, I don't want the thing where my Dropbox files are like, well, is it actually there, or is it just like a little zero content thing that when I double click on it, it has to download.
00:49:47
◼►
I just want the old thing where it was just local files, and they just happened to be in sync with files somewhere in the cloud, and that's what Maestro would give you.
00:54:43
◼►
And then sometimes in the evenings, it gets actually ever so slightly nippy.
00:54:47
◼►
And one of the great things about Quince, one of the many great things about Quince, is that they can cover all of your bases.
00:54:52
◼►
And they are especially good at summertime stuff.
00:54:56
◼►
And so, for example, Quince offers 100% European linen pants and shirts that are breathable, easy to throw on, and the summer upgrade your rotation needs.
00:55:14
◼►
And I will be sad when I can't wear them anymore at the end of the summer because it's not really in style, and even I know that.
00:55:19
◼►
Their t-shirts are soft enough to live in all day, and their lightweight cotton sweaters are exactly what you want when summer nights cool down, like I was talking about.
00:55:26
◼►
Everything at Quince is priced 50% to 80% less than similar brands.
00:55:29
◼►
They work directly with ethical factories and cut out the middle people.
00:55:33
◼►
So you're paying for exceptional quality, not brand markup.
00:55:35
◼►
And one of the great things about Quince is that it's not just for dudes.
00:56:30
◼►
Thank you to Quince for sponsoring the show.
00:56:33
◼►
Big topic for today is a couple of days ago as we record this, Apple decided to sue OpenAI.
00:56:44
◼►
So we'll put many, many, many links in the show notes.
00:56:47
◼►
Like there's a zillion different ones, but we're going to read a little bit.
00:56:51
◼►
Starting with our dear friend Jason Snell over at Six Colors.
00:56:54
◼►
Back in May, OpenAI was rattling its saber about thinking about talking to lawyers about possibly considering a lawsuit against Apple for not treating it right when it came to chat GPT integration.
00:57:06
◼►
Indeed, Apple has sued OpenAI and it's alleging that the hardware program it's building with Johnny Ive and company is being fueled by the theft of trade secrets.
00:57:17
◼►
Here's the statement Apple supplied to various news outlets on Friday.
00:57:22
◼►
At Apple, our teams are constantly developing breakthrough technologies to create the best products and services in the world.
00:57:28
◼►
And protecting their work and intellectual property is something we take very seriously.
00:57:31
◼►
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.
00:57:42
◼►
We will always defend our team's hard work and innovations, and we are taking all appropriate steps to do so.
00:57:51
◼►
These companies, to be at this size and scale, almost everyone at and near the top of a company like this is going to be a huge a**hole in some form.
00:58:03
◼►
Like, it's just what is necessary to rise to those kinds of levels, right?
00:58:09
◼►
And these companies pathologically, like, you know, just the personality of the company based on its behavior, its actual behavior, these companies are both just tremendous a**holes in different ways.
00:58:24
◼►
And so, this statement, this starts out as Apple throwing such, like, manipulative bulls**t language.
00:58:37
◼►
So, let's, okay, our teams are developing technologies, protecting their work.
00:59:01
◼►
Now, keep in mind, Apple, which they're about to, we're going to get into this, you know, later on in this discussion, what the work is that they're defending, Apple sees as 100% Apple's work.
00:59:34
◼►
And Apple owns what it is trying to defend here.
00:59:37
◼►
Let me continue reading, but there's a lot here.
00:59:41
◼►
And I have extremely mixed feelings about it all.
00:59:44
◼►
But continuing on, there's a lot more in the court filing.
00:59:46
◼►
As summarized by Chance Miller at 9to5Mac, the complaint alleges that former Apple designer Tang Tan used insider knowledge of Apple's confidential projects to grill job candidates in interviews.
00:59:57
◼►
Additionally, Tan directed job candidates still working at Apple to bring actual Apple hardware components and samples for, quote, show and tell sessions.
01:00:07
◼►
Furthermore, Apple says a candidate began, quote, screenshotting and downloading files related to a highly confidential Apple product, or excuse me, project hours before interviewing with Tan, who then solicited more information about the same Apple project once the interview started.
01:00:20
◼►
This became an established pattern, Apple says.
01:00:23
◼►
So on this topic, I'm not a lawyer and I don't know about the legalities of this.
01:00:28
◼►
But again, assuming what Apple says is true, I do wonder, and we'll get more into this as we go on, where the line is between what I think is a fairly clear line.
01:00:38
◼►
Well, I don't even know because, again, I'm not a lawyer of like whatever your legal responsibilities are as an Apple employee.
01:00:46
◼►
Presumably you sign stuff when you become an employee and by you violating your employment contract or whatever, like I'm sure there's there is civil legal things there to say, hey, you're an Apple employee.
01:00:59
◼►
You're not supposed to, let's say, take hardware devices and bring them to a job interview with a competitor.
01:01:03
◼►
That's I'm pretty sure the Apple employment agreement probably has language saying that you shouldn't do that.
01:01:09
◼►
But that feels like it would be on the person going on the job interview.
01:01:13
◼►
They have broken their employment agreement with Apple by doing a thing that is against that agreement.
01:01:17
◼►
I don't know where the line is because I'm not a lawyer of like, say you're at the competitor company, say you're Tang Tan and OpenAI.
01:01:24
◼►
Can you just tell people, oh, totally go like, do bring all that stuff?
01:01:28
◼►
Like, can you if you tell someone, hey, you should do a bunch of stuff that will probably break your employment agreement.
01:01:41
◼►
I don't know what the what the civil legal penalties are for, like, encouraging someone to break their employment agreement or whatever, which makes it hard for me to understand what the chances are of this thing.
01:01:52
◼►
Even if you take everything Apple says at face value, it's like, OK, well, but where does the responsibility lie?
01:01:57
◼►
Like, I read this and we're going to read more of it.
01:01:59
◼►
And it totally seems like a bunch of people who either are current or former Apple employees did things that were clear violations of their employment agreement.
01:02:06
◼►
And they're going to get destroyed because, like, if it's true in any way whatsoever, it's so clear that you've done something bad as an Apple employee.
01:02:16
◼►
I don't know where OpenAI's responsibility appears in this because I'm not an employment lawyer.
01:02:22
◼►
I don't know anything about how that works.
01:02:24
◼►
But people reading it, you're like, oh, OpenAI is a villain.
01:02:27
◼►
And I'm like, well, are they just telling people to break the law and then people are doing it or break it again, break up, break their contracts and they're doing it.
01:03:30
◼►
But let me tell you, it sure sounds like Apple has some of these folks dead to rights.
01:03:35
◼►
But anyways, two more allegations from Emma Roth of The Verge in her article listing the six wildest claims in Apple's lawsuit against OpenAI.
01:03:43
◼►
Chang Liu, a former Apple employee who worked as the systems electrical engineer on the iPhone for over eight years, allegedly kept an Apple-owned computer, allowing him to download dozens of confidential files.
01:03:55
◼►
Liu also allegedly accessed Apple's cloud-based network storage weeks after he left the company using an authentication vulnerability that Apple didn't know about.
01:04:04
◼►
Again, I'm pretty sure Apple's employment agreement says you can't keep your Apple-issued computer after you leave the company.
01:04:09
◼►
And you certainly can't use it to access our network after you're no longer employed.
01:04:13
◼►
This is that's like basically I'm sure there's like criminal like.
01:04:16
◼►
Yeah, that's actually whatever laws they passed in the 90s to criminalize hackers.
01:04:19
◼►
That's criminal liability of like you broke into a foreign network like you you unauthorized access to a blah, blah, blah.
01:04:25
◼►
Like we have so many laws in the US about that because back, you know, back in the 90s when they made these laws, it was like to punish, you know, Mitnick or whatever his name was like he's.
01:04:35
◼►
So there's if again, if this is true, there's criminal liability and just really, really poor thinking like the worst industrial aspirin you can ever imagine.
01:04:54
◼►
It's just mind-bendingly poorly thought out.
01:04:57
◼►
Yeah, I mean, this is like, yeah, like this again, not lawyers, but Cheng Lu sounds like he has some significant legal problems if this if any of this is true.
01:05:24
◼►
I'm sure everyone's talking about stuff like that.
01:05:26
◼►
Yeah, this is you can argue about some of the other things like what does Apple own in terms of the people, their expertise, things like that.
01:05:34
◼►
But Apple does own their documents, their hardware prototypes like they that's pretty clear cut.
01:05:42
◼►
Exfiltrating those and bringing them to a competitor is generally a crime.
01:05:47
◼►
Continuing from The Verge, OpenAI allegedly coached Apple employees on how to bypass security measures.
01:05:53
◼►
Apple claims tan kept an internal document that outlines employee outlines employee offboarding procedures.
01:06:00
◼►
OpenAI allegedly used this information to warn employees coming from Apple about the company's security checks and coach them on how to avoid it.
01:06:08
◼►
They said, hey, don't use your Apple laptop to don't refuse to give back your Apple laptop and then use that same laptop, which, you know, Apple is tracking to infiltrate their network and get stuff out.
01:06:19
◼►
So I don't know, maybe they maybe they updated their coaching, but this gets to what Marco was saying.
01:06:22
◼►
Like, OK, at the point you're coaching again, presuming Apple is to be believed here at the point where you're coaching applicants on how to avoid detection while performing industrial espionage.
01:06:33
◼►
That feels like, again, not a lawyer, but that feels like a place where OpenAI suddenly has a bunch of liability.
01:06:38
◼►
Well, there's a version of this that's OK.
01:06:39
◼►
Like, it's kind of like, you know, when we tell when we tell our kids, like, don't talk to the police, like never, ever talk to the police.
01:06:46
◼►
If the police, you know, come and try to stop you.
01:06:49
◼►
If you're not under arrest, you could like you can say, am I free to go and you can go?
01:06:55
◼►
So if if OpenAI is saying, like, here's what's going to happen.
01:06:59
◼►
Apple is going to demand your laptop back.
01:07:01
◼►
They're going to ask you for all these things.
01:07:03
◼►
If any of those things are like legally not required and you can just say, no, thanks.
01:07:08
◼►
Like, I don't want to do your security interview.
01:07:10
◼►
I don't see anything wrong with them telling people that sort of thing.
01:07:13
◼►
But there is a fine line between telling people what their rights are and telling people what's going to happen when they quit or get fired or whatever versus saying, like, well, you can keep your laptop for three extra days and copy all the files off if you do this.
01:07:28
◼►
Like there's so that there is a line and there's a bunch of stuff on the OK side of that line.
01:07:33
◼►
But it sounds like Apple thinks there's some stuff on the other side, too.
01:07:37
◼►
Also from the complaint, as quoted by Gruber, OpenAI has used confidential Apple information in approaching Apple's trusted partners, even having one carry out a specific trade secret metal finishing technique for OpenAI, misleading the partner to believe they had Apple's permission to do so.
01:07:58
◼►
Again, I don't know about the legalities here because like, oh, isn't that just fraud?
01:08:01
◼►
But it's like there are so many situations where things seem like they should be against the law but aren't.
01:08:06
◼►
For example, police can lie to you about this stuff all the time.
01:08:08
◼►
Speaking of police, they can they can just lie and say, oh, yeah, no, Apple gave me permission and you just I know it's the same thing, but I can and do.
01:08:23
◼►
But isn't it kind of on the manufacturer to say, like, I'm sure the manufacturer has an agreement with Apple that says here's our metal finishing technique and our agreement says you can only use it for us.
01:08:36
◼►
Shouldn't the manufacturer say, well, we actually can't because the contract says we can't and we can't take your word for it that like is there liability for opening it?
01:08:44
◼►
There sure is liability for the manufacturer.
01:08:46
◼►
I can guarantee that because Apple's not making contracts that say that don't have a clause in it that says you cannot use this this metal finishing technique for anyone else.
01:09:00
◼►
I'm sure there's something in there that says, as Marco said, there's some line that you cross when you induce a partner or you you provide false information to a partner as part of your contract with them.
01:09:15
◼►
Don't call them continuing from Gruber footnotes 13 on page 15 states Apple and open AI have a commercial relationship involving the integration of open AI's chat GPT into Apple intelligence.
01:09:27
◼►
The companies have entered into a written agreement governing that integration.
01:09:46
◼►
Apple was not mad at open AI when they were making AI chat bots, even though Apple was trying to do that.
01:09:51
◼►
Apple was so incredibly bad at it that they needed open AI to bail them out to the extent that they could back in 2024 with chat GPT integration and iOS when they couldn't get their own act together.
01:10:01
◼►
And Apple just felt like, well, we don't have this and we're not competitive with this and we have this deal with open AI.
01:10:06
◼►
But where Apple has always felt like this is our area, don't step on our toes is hardware.
01:10:13
◼►
So as soon as open AI said, hey, we're going to do hardware, we're hiring Johnny Ive and we're buying his company and we're doing this and we're doing that.
01:10:20
◼►
And Johnny Ive is pulling all the Apple designers out of Apple into open AI.
01:10:23
◼►
And Apple's looking at that and say, no, no, no.
01:11:05
◼►
So this is totally that moment where they're like, oh, so you're going to take all our designers and Johnny Ive and you're going to make a hardware product that competes with us.
01:11:47
◼►
But, you know, so Johnny created this like sub company for his hardware efforts, sold it to open AI.
01:11:54
◼►
This is obviously like a bunch of like, you know, moving cups around to try to somehow legally and financially work out the fact that Johnny Ive is working for open AI, but not exclusively for open AI.
01:12:22
◼►
Think about the relationship already that had to have gone on between Apple and Johnny Ive.
01:12:28
◼►
Now, the last thing Apple wants to do is point this out or draw attention to it, which is why Johnny Ive is mentioned nowhere in this lawsuit, even though it obviously involves the division of open AI that is basically his team.
01:12:41
◼►
Like it would be really out there and far fetched if it didn't involve any of him or his core people.
01:12:49
◼►
So it probably does, but Apple does not want the PR to be that they are suing Johnny Ive in any context or any of Johnny Ive's properties.
01:13:01
◼►
Especially since at this point, as far as the public is concerned, they're like, there is no Johnny Ive open AI product.
01:13:07
◼►
If they had done this after, like the open AI egg comes out and Johnny Ive is there on stage telling everyone how great it is and then they sue, then that's a big scandal story.
01:13:15
◼►
But if they sued and mentioned Johnny Ive now, people would be like, what are they suing him for?
01:13:26
◼►
So the timing of this is actually pretty good and allows them to not name him and to avoid this story because there is no focal point for it.
01:13:32
◼►
There is no egg for us to look at and say, open AI came out with hardware and Apple hates them.
01:13:36
◼►
Open AI is going to come out with hardware.
01:13:52
◼►
There must be some pretty bad blood between them and Apple because, first of all, they all left.
01:13:59
◼►
So, you know, so they probably weren't super happy with Apple.
01:14:02
◼►
Second of all, Apple is probably not super happy with them for for leaving and then pulling so many people with them.
01:14:09
◼►
So obviously something has gone down in, you know, over the last few years, there has been some kind of significant turmoil in these relationships between these these factions that used to be all with an Apple.
01:14:28
◼►
Yeah, according to the most recent dithering, Tang Tan and Ternus hate each other from back in the Apple days, because basically they were it seems like they perhaps were up for the same type of thing ahead of hardware and Ternus got it and Tang Tan didn't.
01:14:42
◼►
So, yeah, if you are Tang Tan and you're over at open AI and Apple already hates you and new CEO personally hates you lawsuit.
01:14:50
◼►
Right. And so there's obviously there's all this drama going on between these of the all these factions that used to be all at Apple.
01:14:58
◼►
It probably, you know, explains a lot of the different directions Apple has gone with hardware over the last few years that I think are universally better.
01:15:08
◼►
But, you know, it probably helps explain a lot of a lot of those machinations that we've seen over the last few years that like obviously there's there are these factions and they and they do they don't like each other.
01:15:19
◼►
And so Apple also super hates losing people to competitors.
01:15:24
◼►
I mean, there was that famous Steve Jobs, you know, collusion thing.
01:15:31
◼►
And so, yeah, Apple is going to feel really attacked.
01:15:33
◼►
And I'm sure Apple was just looking for something that they could that they could like bring up in some kind of legal threat to try to, you know, try to stop some of this.
01:15:43
◼►
And it looks like open AI was totally happy to give them lots of things.
01:15:48
◼►
I mean, again, again, transferring it to open AI might be tricky, but I'm sure the security team had all these people, as Casey said, dead to rights.
01:15:56
◼►
If this all this is true for a long time, it's just a question of can we turn this into an open AI thing?
01:16:01
◼►
The other angle on that discussion that has come up in our little circles is the broader discussion about when you leave a company.
01:16:10
◼►
Do you have to empty your brain of everything, you know, or can you leave like what do you get to leave with?
01:16:15
◼►
Marco mentioned that obviously when you do work for a company, it is work for hire.
01:16:22
◼►
Um, but California in particular has fairly, uh, strict rules as compared to the other states in the United States about not allowing employers to unduly restrain their employees.
01:16:37
◼►
Because part of the Silicon Valley culture is you work for one company, you learn something, you work for another, you go to another company, you do a startup or whatever.
01:16:44
◼►
They don't want to allow things like draconian non-competes.
01:16:47
◼►
I think non-competes in general are not even legal in California where they say, oh, if you leave this company, you can't work in this industry for two years.
01:16:54
◼►
Stuff non-competes like that are legal in some states of the United States, but not in California because it is against their culture of like, we, we started this company, then we leave to do a startup, then we compete.
01:17:03
◼►
Like, but they want to see more competition.
01:17:05
◼►
Similarly, I would imagine the laws about like, what do you get to take with you in your brain when you leave a company?
01:17:12
◼►
Like if you, you know, learn a bunch of stuff at one company about like machining and metal finishing, obviously there's the, uh, the various legal constructs like patents and, uh, trade secrets that you've signed an agreement that you're not going to reveal.
01:17:28
◼►
Like, you know, good ways to do things and how to organize people.
01:17:31
◼►
And like, you take those with you, obviously you can't, you know, Apple doesn't own the things that you learned when you were there.
01:17:37
◼►
And I'm sure there's a whole complex realm of law about, okay, well you learned like best practices and how to organize things and what makes a good deal and a bad deal or whatever.
01:17:45
◼►
But what about this one very specific thing that you're doing at open AI that is a quote unquote trade secret.
01:17:52
◼►
And it is not just like a thing that you learned when you were there.
01:17:56
◼►
Like I can say in all, like in all my past jobs, I don't know how many web frameworks I wrote starting with my very first web development job out of school.
01:18:03
◼►
I think at various jobs, I've written multiple web frameworks and each time I went to a new job, I didn't take my old web framework with me because I didn't want it because I had learned so much since then.
01:18:12
◼►
But when I wrote my next web framework, it built on all the knowledge of my previous web frameworks.
01:18:17
◼►
And I kept doing that and making better and better and better ones.
01:18:20
◼►
I was taking my knowledge and experience with me.
01:18:22
◼►
I was not taking the code with me, but I was definitely taking like, oh, last time I did it, I did it this way.
01:18:27
◼►
And I learned that this is good and this is bad.
01:18:29
◼►
And if you squint at one of my web frameworks, you're like, that looks kind of like the one you did at your last job.
01:18:33
◼►
Only I see this change, this change, and this change.
01:18:34
◼►
It's all new code, but I'm taking my experience with me.
01:18:37
◼►
That's what makes you a more valuable employee as you gain experience is that you learn things.
01:18:41
◼►
And that's why someone hires you for more money at another job.
01:18:44
◼►
And again, I feel like California is one of the best states, I said strict before, but like strict in terms of defending employees' rights to compete, to leave one company and to go to another or to leave one company and start their own company.
01:18:56
◼►
And we can't say, oh, that's not fair.
01:18:58
◼►
You learned a bunch of stuff at Apple and that's the only reason your startup is successful.
01:19:01
◼►
Like, yeah, I learned stuff and now I'm smarter and I'm making my own company, right?
01:19:05
◼►
So that also is kind of against Apple on the broader issue of like, you took things that you learned and went to another company to compete with us.
01:19:13
◼►
Like, yeah, that's how it's supposed to work.
01:19:46
◼►
And so the definition of what is considered a trade secret, that is probably going to be possibly argued here.
01:19:53
◼►
That's going to be pretty important because that sounds like it's a somewhat squishy concept.
01:19:58
◼►
But yeah, like, I think, like, if you – if you are really good at metal finishing and OpenAI hires you and they say, hey, we want you to give us some really good metal finishing expertise, unless certain things are patented by Apple, I think you can probably bring them your metal finishing knowledge in most cases.
01:20:19
◼►
And if they're patented, the whole point is the patent allows that to be shared, but then there's perhaps a licensing fee.
01:20:24
◼►
The whole point of patents is we tell the world, but you don't get to use it for free.
01:20:28
◼►
The idea of hiring away Apple's people with competitive salaries and, you know, offers of freedom or excitement that they might not have at Apple, that's fair game.
01:20:39
◼►
And Apple, they are a big, established company with mostly big, established, mature product lines.
01:20:47
◼►
So, yeah, people are going to get the itch and want to go do something new.
01:20:50
◼►
And Apple, historically, is somewhat stingy with salaries, from what we've heard.
01:20:55
◼►
And so, you know, OpenAI is able to throw around some cash or Meta, you know, Meta famously throws around big salaries.
01:21:03
◼►
Like, of course, they're going to hire – they're going to be able to hire away some of Apple's people.
01:21:06
◼►
And every time somebody at Apple gets promoted, there's going to be someone who didn't, and they're going to be upset, and maybe they'll be easy to hire away as well.
01:21:33
◼►
As long as you don't cross certain lines of things like stealing documents and stuff like that, like that, that is obviously a line that they should not cross.
01:21:41
◼►
And it sounds like they did just walk right past it and stomp all over it.
01:21:45
◼►
But the idea of restricting a company from hiring away your people with compelling offers and then having those people do similar work for them, Apple doesn't own those people.
01:21:59
◼►
And if Apple does not want to lose those people, they should pay them better if they need to or work with them in whatever other things that they want if they can.
01:22:08
◼►
And recognize that that's not going to always be the case.
01:22:11
◼►
Again, like, you know, the promotion situation of, like, somebody gets promoted, but someone else didn't.
01:22:14
◼►
You know, there's not that much Apple can do about that.
01:22:17
◼►
And so if those people want to leave, they're going to leave.
01:22:19
◼►
Or if somebody wants to leave the giant mothership of Apple and go work for some, you know, fast moving, exciting startup, like Apple probably can't do much to keep those people either.
01:22:37
◼►
And Apple is going to put out a whole bunch of bravado about how they own all their stuff.
01:22:43
◼►
And I'm sure Apple believes, I'm sure some people in Apple believe that they own a lot more of these employees' talents than they really should.
01:22:51
◼►
But at the end of the day, it does sound like a lot of really clear lines have definitely been crossed by OpenAI.
01:22:59
◼►
And I'm not surprised to hear that, like, going back to my earlier comments, like, a lot of these people have very strong personalities, let's say.
01:23:09
◼►
Amateur psychologists would have a field day with a lot of these people trying to figure out what exactly their issue is.
01:23:16
◼►
These are some really strong personalities at these levels.
01:23:21
◼►
And so many of them have no problem lying.
01:23:25
◼►
I mean, geez, Sam Altman has basically made a career out of it.
01:23:28
◼►
Like, many of these people are, you know, really showing, like, significant sociopathic kind of behaviors of just, like, totally disregarding morality effects on people, rules, laws, fraud.
01:23:42
◼►
So, yeah, you're going to see a lot of this kind of behavior at these levels.
01:24:08
◼►
Apple probably has a pretty good case.
01:24:10
◼►
Like, I don't think Apple would do this if they didn't.
01:24:12
◼►
And I don't know if they're going to get all of what they want, but I bet they're going to get some of what they want.
01:24:16
◼►
Yeah, when Apple initiates a lawsuit, there's usually a pretty good chance that they think they're going to win it.
01:24:21
◼►
When they're the victim of a lawsuit, you never know how it's going to go because you don't know what Apple did.
01:24:25
◼►
But when they initiate a lawsuit, they're probably, they're so conservative about this.
01:24:29
◼►
And in this particular case, you would imagine if you were to remove most of these personalities, you were like, oh, OpenAI will just settle this because they've got tons of money.
01:24:37
◼►
And what you do in cases like this is you throw money at it and you make the problem go away.
01:24:41
◼►
Because the last thing OpenAI needs is some, like, long legal battle and bad press and blah, blah, blah.
01:24:50
◼►
One, if there's, like, personal animosity, like, again, between Ternus and Tang Tan and maybe Johnny Ive and Apple and Sam Altman and Apple.
01:24:58
◼►
Because, again, we started this with Snell's thing with OpenAI was putting out these little leaks that, like, we're thinking about suing Apple, like, a long time ago.
01:25:07
◼►
So there was already bad blood, probably because they knew they were going to be sued at that point or whatever.
01:25:11
◼►
So that bad blood exists and you're like, okay, well, when you have these big personalities and bad blood, on the one hand, you're like, oh, that means OpenAI is not going to do the smart thing.
01:25:19
◼►
And instead of settling like they could with their giant pile of VC money right now when they're rolling in it, they're going to, you know, dig in and say, no, we're going to fight this, right?
01:25:34
◼►
But then I look at the specific personality of Sam Altman, who seems like so incredibly feckless and just, just not.
01:25:41
◼►
He's, he's more like, you know, a pathological liar and untrustworthy or whatever, but also not particularly like brave or like he's not Mark Zuckerberg.
01:25:51
◼►
Mark Zuckerberg, I can imagine like digging in and be like, we're going to fight this.
01:25:54
◼►
Or Steve Jobs, for example, of course, would dig in his heels and fight this way past the edge of logic.
01:25:59
◼►
But, uh, I don't know what Ternus is like in that regard, although it seems like he's okay with this lawsuit.
01:26:03
◼►
But Sam Altman, like, even though he has a strong personality and has no problem lying, I would imagine he's not like, he doesn't seem like a fighter to me.
01:26:11
◼►
Like in the, in the Elon Musk lawsuit and everything, I mean, Elon, it's not hard to look better than Elon when you're on the stand or whatever.
01:26:17
◼►
And I think Sam Altman did, but I just, he just seems so like, just like a wet dish rag full of lies.
01:26:23
◼►
Like it just not, like, I don't see him going like, we will, we will fight Apple to the bitter end about this lawsuit.
01:26:52
◼►
Um, but the other thing, and I don't remember if I heard this on Dithering or somewhere else, but somebody pointed out, maybe on Upgrade, that this is a really great opportunity to do a whole lot of decisions.
01:27:02
◼►
And I think that that's very appealing to Apple because they can.
01:27:06
◼►
But that's why open AI was settled to avoid that.
01:27:09
◼►
But if open AI really wants to go for broke, that's not going to end well for open AI, I don't think.
01:27:14
◼►
And coming back to something that I think Marco said earlier, it might have been, it was either one of you, but, but somebody said earlier, you know, if you don't want to have an exodus, here's an idea.
01:27:23
◼►
And I know that Apple employees make a butt ton of money, but I also know it costs a butt ton of money to live in the San Francisco Bay Area.
01:27:31
◼►
And everything I've understood from many, many, many different people inside and outside of Apple is that amongst Silicon Valley and amongst the like Facebooks and, and Googles and Apples and Netflixes of the world, Apple's like really cheap from everything I've ever understood when it comes to salary.
01:27:50
◼►
Now, granted they have, you know, uh, stock programs for their employees and non, non salary compensation and so on and so forth.
01:27:57
◼►
But here's the novel idea, which was said earlier.
01:28:00
◼►
And I would like to agree with maybe pay your people a little more.
01:28:04
◼►
And I think that might solve a portion of these problems.
01:28:08
◼►
Now, it's still not the new shiny over at open AI.
01:28:11
◼►
You know, the, the iPhone is new ish and shiny ish, but it's not new shiny, like whatever they're doing in open AI.
01:28:16
◼►
But I think some of this could be solved by having more competitive salaries.
01:28:20
◼►
And I'm pretty sure Apple could afford it.
01:28:23
◼►
Well, the flip side of that is something we've discussed in the past, which is, first of all, you're never going to compete with the giant bucket of investment money that is sprinkling down on open AI.
01:28:33
◼►
Nor will you compete with the apparent huge pile of money that Facebook is willing to throw into these efforts.
01:28:39
◼►
And it probably would be foolish for Apple to pursue that.
01:28:43
◼►
Apple has chosen not to pursue the creation of a frontier model, which costs billions and billions of dollars.
01:28:56
◼►
And the second aspect is the thing we talked about back in the JG days, where they were hiring AI people to JG's team to do Siri or whatever, and paying those people way more than other Apple employees.
01:29:08
◼►
And that makes the other Apple employees go, hey, I'm over here working on iOS.
01:29:12
◼►
You know, the thing that makes actual money.
01:29:14
◼►
Why are all these new employees getting paid more than me?
01:30:17
◼►
I mean, it's the same reason people have always left Apple to try to do something else.
01:30:20
◼►
Like Apple is there to be steady and reliable and a place people can circle back to after their startup fails or after they get sick of working for Zuckerberg or whatever.
01:30:32
◼►
I do agree with you that they probably should be less stingy than they are, especially since a lot of their stinginess relied on the fact that Apple stock was just always going up and up and up.
01:30:39
◼►
And by the way, it still kind of is going up and up and up, but not to the levels it used to be.
01:30:43
◼►
And also the whole, like, don't you want to work in the games industry?
01:30:48
◼►
People's willingness to take less in exchange for being what they consider in a really high profile place.
01:30:54
◼►
Like it's the reason the game industry exploits workers is the room in Hollywood exploits them because they're like, oh, I've always wanted to work on games.
01:31:00
◼►
I've always wanted to be in Hollywood.
01:31:13
◼►
It's unwise to pursue the unsustainable salaries that everyone else is getting.
01:31:17
◼►
And, you know, we'll see how it works out for those companies.
01:31:19
◼►
I bet, speaking of Facebook, the people who went over there and got those huge salaries, but then realized that they're going to have like that Facebook is going to be looking at the key loggers that are installed on other computers to make sure they're all working hard enough or whatever.
01:31:33
◼►
I wonder if they're reconsidering their life choices.
01:33:32
◼►
I would say that the comms department at Apple has a little bit more experience and maturity when it comes to dealing with these stuff than a very young company like OpenAI.
01:34:12
◼►
Apple, in this particular context, Apple probably, they probably wouldn't have brought this suit if they thought that discovery would be a big risk for them.
01:34:22
◼►
So, they probably think, look, if you dump a bunch of Apple emails out into court that, you know, around this particular area, we have nothing to hide.
01:34:56
◼►
But if I were Apple, I don't think I'd want to rush to a settlement.
01:35:00
◼►
Because, like, OpenAI, like, they're going to keep getting more and more desperate as their IPO dreams get closer and, you know, if things start looking very bad for them.
01:35:10
◼►
And if a bunch of discovery stuff starts coming out like that, Apple is probably in a pretty strong position on that.
01:35:16
◼►
And, you know, who knows how – if this actually goes to court and goes to trial and goes all the way to the end of a trial, those are very big ifs.
01:35:26
◼►
But if this actually gets that far, who knows if Apple would actually, you know, quote, win or not.
01:35:31
◼►
But they can inflict a lot of pain on OpenAI in the meantime with dragging them out into all of this, dragging all this into the public, all the discovery coming out, and casting doubts maybe on some of that IPO.
01:35:42
◼►
So Apple's in a pretty good position here.
01:35:45
◼►
I would imagine that the ChatGPT and iOS deal is probably going to lapse and not be renewed, though.
01:35:51
◼►
Yes, and also, like, once you use iOS 27, you're like, yeah, there's no need for that anymore.
01:35:57
◼►
Like, Siri has taken most of the need for that away.
01:36:01
◼►
And the OpenAI deal of, like, having it be integrated, if you ask Siri to, hey, also, please ask ChatGPT to tell me this, like, that's so clunky and less necessary now.
01:36:12
◼►
No one's going to really be doing that anymore.
01:36:14
◼►
Honestly, it never worked that well to begin with.
01:36:16
◼►
Hopefully that'll be replaced next year by the thing that was rumored this year's WWDC, which is like a third-party pluggable, hey, if you want to use something, you want to use a different chatbot because the EU is making us, you can plug in, you know, some other company's chatbot besides Siri in there.
01:36:30
◼►
And then OpenAI would just do that just because they have to because Anthropic would do it and all the other stuff.
01:36:36
◼►
But yeah, the specific, very, very particular deal that Apple referenced, I don't know what the expiration date on that is, but I feel like neither one of the parties is going to be excited about renewing it.
01:36:48
◼►
And by the way, if you're an Apple employee and you want to go work somewhere else, go for it.
01:36:55
◼►
If you can get a higher pay somewhere else and it's something that you're really excited about, do it.
01:37:01
◼►
Except unless it's at Facebook, don't do it.
01:43:33
◼►
doesn't it also mean that like all the crappy vacation houses that we rent should be getting better and better Wi-Fi and therefore making it so that we don't need travel routers because the home Wi-Fi actually spans the whole house and works.