[JM]: I want to take a moment and congratulate “Take On Me” on its 40th birthday.
[JM]: This amazing song by A-ha — I cannot believe I just said that out loud — is now 40 years old, came out in 1985.
[JM]: And I wanted to mention it not only for that, but also because I listened to this fantastic podcast episode from Song Exploder, in which the show host talks to
[JM]: one or more of the members of the band and talks about how the song came to be.
[JM]: And it's a really fascinating story, because among other things, we think of it as this smash hit.
[JM]: And initially it wasn't — like, it flopped twice.
[JM]: Like, just really failed, despite multiple attempts to get the song more airplay to get it more widely known, just could not get enough interest to have this song go much of anywhere.
[JM]: And they were kind of close to giving up on it, but they didn't.
[JM]: I think there's a underlying life lesson here about believing in something that you're doing and being persistent, even when everything fails.
[JM]: Everything around you is telling you that, “Hey, this isn't working”, because obviously it worked really well.
[JM]: And it's a fantastic song.
[JM]: The podcast episode is great.
[JM]: I'm also going to link in the show notes to a very interesting article.
[JM]: I am, as usual, going to link to an archived copy of the article because the original article is, among other problems, paywalled.
[JM]: But it's an excellent article that goes into similar detail, covers also some other things.
[JM]: Both the podcast and the article complement each other really well.
[JM]: I highly recommend checking them both out.
[JM]: And if you haven't seen it in a while, this is also a good occasion to watch the video for Take On Me.
[JM]: If you haven't seen it in a while, it's great to be reminded of just how innovative Take
[JM]: this video was for its time.
[JM]: It's really clever idea, implemented really well, hand illustrated in case if you're imagining because you've seen it before what this video was like, a truly impressive accomplishment and worth a rewatch if you haven't seen it in a while.
[JM]: All right, in other news, there is an open source project called LibrePods, and LibrePods unlocks Apple's exclusive AirPods features on non-Apple devices.
[JM]: Now, at least for now, I only use Mac computers and iPhone and iPad.
[JM]: I'm, at least for the moment, still very much in the Apple ecosystem, and I've never had needs to use AirPods outside of that ecosystem.
[JM]: But I think it's cool that according to what I'm seeing here in the readme for this project, that it looks like you can now access features in AirPods that were previously only available on Apple devices.
[JM]: Well, now you can use those formerly exclusive features on Linux and Android devices.
[JM]: And if you get a chance, you should definitely go to this project and check out the screenshots.
[JM]: If for no other reason that I don't know why, but one of these screenshots has like a cat in it.
[JM]: There's just like a screen with this giant cat's face looking off to the side, like it's in some kind of I don't know, photo shoot with superimposed icons about transparency mode and adaptive mode and noise cancellation and all of these features that
[JM]: are being unlocked by this project.
[JM]: I just love that this is here.
[JM]: I don't know why it's here, but it makes me happy.
[DJ]: Is the cat wearing AirPods?
[DJ]: Is that part of it's demonstrating the features?
[JM]: As far as I can tell, the only reason that you could possibly imagine this cat needing to be here, because no, it's not wearing anything.
[JM]: It's just there in the background.
[JM]: I don't know.
[JM]: It's got big ears.
[JM]: Maybe the emphasis is on the hearing.
[JM]: I don't know.
[JM]: Your guess is as good as mine.
[JM]: Well, cats do...
[DJ]: traditionally have very good hearing and use Linux computers.
[DJ]: So I think cats are really part of the target market of this project.
[JM]: Everyone, yeah, well-known fact that cats make up the majority of the Linux user base.
[JM]: Okay, moving on to a topic that I wanted to discuss and that is Apple intelligence.
[JM]: I ask because I want to know, like, how are we feeling about this feature so far?
[JM]: I don't remember how long it's been.
[JM]: It's probably been less than a year.
[JM]: I think they originally talked about shipping this at WWDC in 2024.
[JM]: And then in the fall, it kind of came and went like not many things shipped.
[JM]: And then in subsequent 18 dot releases, like dot two, dot three, dot four, they slowly kind of dribbled out some of these Apple intelligence features.
[JM]: That's how I remember it anyway.
[JM]: And so my question is, after this time has passed, how are we feeling about it?
[JM]: Are we using it?
[JM]: So I'm going to let you start, Dan.
[DJ]: Well, this is an interesting question because I haven't really been using it, even though I have it.
[DJ]: And as you say, it's been out for maybe a year.
[DJ]: I'm pretty sure I've had it on my phone for at least six months and I haven't really used it.
[DJ]: I think because I never really got in the habit of using Siri.
[DJ]: It has been so often underwhelming, right?
[DJ]: Like this is a topic we hardly need to go back over.
[DJ]: So part of the promise of Apple Intelligence was like a better voice assistant.
[DJ]: And I think I tried it a couple times when it first came out, remained underwhelmed.
[DJ]: And I just sort of haven't bothered to go back to it.
[DJ]: But I think that this is actually a great time on this podcast to do a live demo of Apple Intelligence and find out what this thing is actually capable of.
[DJ]: So
[DJ]: I have a relatively new iPhone in front of me here, and we're going to try asking it a few questions.
[DJ]: What is the weather like today?
[DJ]: Well, it pops up a thing that says that it's six degrees with showers.
[DJ]: That's not bad.
[DJ]: It just pops up like the weather app widget, so that's a sensible answer.
[DJ]: Let's try something a little harder.
[DJ]: What is five times a million?
[DJ]: Five million.
[DJ]: All right, I nailed that one.
[DJ]: But computers are traditionally pretty good at arithmetic.
[DJ]: So, well, I think that's, you know, low-hanging fruit.
[DJ]: What else should we ask this thing, Justin?
[JM]: I think this is a good opportunity to interject because I don't think what you are doing is involving Apple Intelligence at all.
[JM]: And I think that's actually a big part of the problem, right?
[JM]: It's what is Apple Intelligence?
[DJ]: But it has the fancy new highlight around the screen when I press the button.
[DJ]: So I think that means Apple intelligence, right?
[JM]: It means Apple intelligence is enabled.
[JM]: If you disable it, that fancy little halo around the edges of your screen will go away.
[JM]: That's my understanding anyway.
[DJ]: Well, I like the halo, so I'm not going to do that.
[JM]: Right, but I'm not convinced that the Halo actually means what you think it means.
[JM]: You are asking it questions that I think the traditional Siri is continuing to answer.
[JM]: I don't think any of the things that you just tried actually are involving so-called Apple intelligence.
[JM]: which if we're gonna draw a line, we're drawing the line presumably at, okay, it's using a large language model to do some of the processing into return an answer.
[JM]: But I don't know if that's actually what's happening.
[JM]: Now, then again, you're using iOS 26 at the moment, maybe it does have additional Apple intelligence features that I'm not familiar with, because I'm not on that OS yet.
[JM]: But the impression that I get is, at least for me, when I ask it questions that aren't basic arithmetic or what's the weather today or things that Siri has been able to answer for some time now, the answer is always the same.
[JM]: And that is, would you like me to ask ChatGBT this question?
[JM]: Because this is just Siri's failure mode, right?
[JM]: This is Siri saying like, yeah, I'm not capable of doing LLM stuff yet.
[JM]: So I'm just going to direct you over to ChatGBT.
[JM]: So this has been interesting that you're trying to test Apple intelligence in a way in which I'm not sure is even touching it.
[JM]: And that is, I think, in part, a serious problem for Apple, because most people don't understand this distinction, right?
[JM]: Like if they are told, if they're even aware that Apple intelligence exists, they're
[JM]: then they think of it as Siri.
[JM]: Like that's just probably more or less one in the same for them.
[JM]: When I originally posed this question, I never even thought about Siri at all, because again, I don't connect those two things, even though they are identical for most people to me, they're completely separate because that's been my experience.
[JM]: I see where it fails and it tries to kick me over to ChatGPT.
[JM]: So when I think of Apple intelligence, what I'm thinking of are things like text summarization, text manipulation, right?
[JM]: I guess you can tell it to rewrite text in some particular tone or style.
[DJ]: Rewrite this email to my boss so I don't get fired, for example.
[JM]: Right.
[JM]: Yes.
[JM]: Not getting fired being, you know, an important thing.
[JM]: But I think the area of Apple intelligence that most people experience are notification summaries where you get a bunch of notifications or an email, you get a long email and.
[JM]: You just see like this brief summary of what could be a very long message in the email and you get this little summary in a notification or in a list view of your email.
[JM]: And speaking personally, that is probably the only way that I experience Apple intelligence.
[JM]: I don't use it for text summarization or text manipulation.
[JM]: I don't use image playground and generate images.
[JM]: I think I tried that once or twice on launch day and was like, yeah, this is not particularly useful and quit and never launched again.
[JM]: So to answer the question that I posed to you, I thought about this earlier today and thought, when have I used it?
[JM]: Oh, right.
[JM]: I really haven't.
[JM]: And how much value do I get out of these notification summaries or email message summaries or whatever?
[JM]: And the answer that I came up with is not enough for what I presume to be a non-zero amount of resource consumption.
[JM]: You know, like, I don't know how much RAM it consumes to have that active.
[JM]: I don't know what it's doing behind the scenes, but for as little as I use it and as little value as I get out of it, I just instantly went into settings and went, bye, disable,
[JM]: on all my devices.
[JM]: Did that earlier today?
[JM]: Felt good because why have it be on?
[JM]: I don't get any value out of it.
[JM]: And hopefully that changes, right?
[JM]: It does seem like this is recognized as a problem within Apple.
[JM]: They seem to be putting a lot of attention and resources into it.
[JM]: And I really hope that it reaches a point where I'm excited to re-enable it and excited to use it to do either things that are fun or productive.
[JM]: And it's particularly weird for me because I remember when they originally announced it, they said that you need to have an iPhone 15 Pro or greater in order to use Apple Intelligence.
[JM]: And I thought, well, that's weird.
[JM]: I have an iPhone 15 Pro, not because I wanted to use it for Apple Intelligence because it didn't exist when I bought it.
[JM]: But I remember feeling like, oh, this is cool.
[JM]: Like I have a device.
[JM]: that supports it.
[JM]: I think it was really the only device at the time that they announced it that could support it, or at least one of the few.
[JM]: But in the end, it didn't really provide enough value for me to even be excited about it when I tried it.
[DJ]: Yeah, now that you mention it, those notification summaries are sort of the only way that Apple Intelligence has really surfaced itself to me.
[DJ]: Because aside from what I did before, which was this sort of conflation of so-called Apple Intelligence features with Siri, what's funny about, aside from it being inherently funny, what's funny about this bit is I think this is very common to anthropomorphize, right?
[DJ]: So to me, it feels quite natural when Apple announces these new Apple intelligence features for everyone to go, oh, that means Siri, because that's what we think of as the, quote, intelligent part of our phone.
[DJ]: And indeed, the biggest thing I've always wanted and not gotten from Siri in the past, like, 14 years or something...
[DJ]: I think it's been out, is the sort of things that large language model powered chatbots are good at, which is keeping a context going.
[DJ]: The thing that always frustrated me about Siri was that you could say, you know, what's Justin's phone number?
[DJ]: And it would bring up a random contact named Justin from your contact list.
[DJ]: So one thing that's annoying about that is
[DJ]: It lacks the context of here's a guy named Justin who I message all the time.
[DJ]: And here's a guy named Justin who fixed a toilet for me a decade ago in a house I don't live in anymore.
[DJ]: And it can't figure out that like when I ask for Justin's contact information, I probably mean one of those people and not the other.
[DJ]: But then it's like, well, here, you know, here's his phone number, let's say.
[DJ]: And then if I go, now send him a message, it'll go, who do you want to send a message to?
[DJ]: Because it doesn't hold any context between requests.
[DJ]: And I've been hoping for that feature to emerge.
[DJ]: And I think it has been among the promises that Apple has made for Apple intelligence, but I don't think it exists quite yet.
[DJ]: The thing with the email and message notification summaries is
[DJ]: That feels like a problem in search of a solution, frankly.
[DJ]: I guess I would be happy to hear from other people for whom this has really been transformative.
[DJ]: I will grant you I don't operate in my email as much as a lot of professionals do.
[DJ]: But yeah, having experienced it, I've found these summaries just don't typically provide that much value.
[DJ]: They're usually slightly inaccurate for starters.
[DJ]: So usually what happens is I look at my message list and it has these little summaries that
[DJ]: And I glance at them and go, and then I have to tap on the message to read what it actually says anyway.
[DJ]: So it really does feel like, well, there's a little bit of space on the screen here where we could put something.
[DJ]: So what we'll do is summarize this email through a large language model and put the summary in there.
[DJ]: It doesn't really feel like this incredible site of value, but instead a what is a possible use for this technology that everyone is obsessed with.
[DJ]: And if we don't do something with it, we will be perceived as no longer on the cutting edge.
[JM]: Yeah, and I would even argue that the risks outweigh the rewards when it comes to email summarization.
[JM]: There have been several times where I've seen a summary and gone, uh-oh.
[JM]: like where I've been alarmed and then tapped through to read it and went, oh, okay, got it.
[JM]: Because unsurprisingly, several paragraphs of text can convey a lot more nuance than a couple fragments in the summary.
[JM]: And so it has to just whittle it down to a handful of words that can often be rather alarming without that other context.
[JM]: And so I would argue that the risks of this summarization are considerably worse than whatever benefits they are conveying.
[JM]: And I'm glad to be done with it.
[JM]: Side note, when I went to disable it, Apple presents you with this dialogue that says Apple intelligence features like writing tools will not be available if you continue.
[JM]: Yeah, well, okay.
[JM]: I never use those writing tools.
[JM]: So, you know, good riddance.
[JM]: Siri will also stop using the additional capabilities provided by Apple intelligence.
[JM]: Then my next thought is, I'm sorry, Siri had additional capabilities provide Apple intelligence?
[JM]: That's news to me.
[JM]: Never noticed that before.
[JM]: You're like, I have a follow-up question, Apple.
[JM]: Which ones were those?
[JM]: Right, yeah.
[JM]: Please tell me which additional capabilities I'm about to miss out on because I don't think I'm going to miss much.
[JM]: And then the two options below it were turn off Apple intelligence in scary red text and cancel.
[DJ]: Well, I was not scared.
[DJ]: That is obnoxious dialogue design, for sure.
[DJ]: The only way that could have been worse would be if the button said, turn off for now.
[JM]: Right.
[JM]: Well, on a related note, it sounds like allegedly next year with the 27 line of operating systems, we are going to get a series of releases that are focused more on fixing bugs and products quality and also a bunch of Apple intelligence features.
[JM]: And
[JM]: I think that I buy the latter.
[JM]: Like, sure, I imagine they'll ship a bunch of additional Apple intelligence stuff because as we just discussed, that's kind of a core focus for them right now.
[JM]: But I'm feeling a lot less confident about the, okay, we'll ship less features and really make sure that our software experience is polished.
[JM]: Like, I will believe that when I see it.
[DJ]: indeed that is one of those rumors where it's like okay this is what everyone wants so we're gonna get it and therefore it feels too good to be true like i think there's something to the fact that rumors of this kind about how oh apple's gonna slow down and focus on quality always reference mac os 10 snow leopard and i enunciate that very carefully because i want to underline the fact that this was so long ago that the operating system was known by a different name
[DJ]: Snow Leopard was a version of Mac OS X that came out in 2009, which, in case you haven't had your coffee yet and are having trouble doing arithmetic, was 16 years ago.
[DJ]: So given that it's been 16 years of annual releases since the last time, arguably, that Apple, quote, focused on quality,
[DJ]: Yeah.
[DJ]: I have a hard time believing that we're going to get one of those.
[DJ]: I would love it.
[DJ]: Don't get me wrong.
[JM]: But really?
[JM]: I think that would be amazing.
[JM]: But I think you're totally right that oftentimes stuff that gets churned through the rumor mill a lot of times is wishful thinking.
[JM]: It's people that are dreaming of something and want it to exist.
[JM]: So it becomes a rumor.
[JM]: I just don't know that that's...
[JM]: how the world works.
[JM]: You know, you can't just via magical thinking, will it into existence as much as I sometimes dream that, that I could do that, but that doesn't seem to work out so well.
[DJ]: Sadly not, no.
[DJ]: In a semi-related context, I recently came across the phrase, the wish becomes the thought.
[DJ]: And that also strikes me as a good descriptor of what feels like it happens here a lot of the time, where however various sources of rumor acquire their information, there is a heavy incentive to attract readers by promising what they want.
[DJ]: So if you can find any sort of suggestion that a highly desired thing will occur, then you'll write about it.
[DJ]: You'll be like, hey, the next Mac OS might be like this.
[DJ]: It's like, yeah, I guess.
[DJ]: I mean, it might feature a bold new neon-inspired color scheme.
[DJ]: as well.
[DJ]: Incidentally, I would love that.
[DJ]: I think the follow-up to the notorious liquid glass user interface should be vapor glass.
[DJ]: It should be like a vapor wave kind of style.
[DJ]: That would really get me to upgrade my laptop on day one if Apple shipped that.
[DJ]: So there's your wishful thinking.
[JM]: A 1980s synthwave-inspired OS?
[JM]: Oh yeah, sign me up for that.
[JM]: I would take that in a heartbeat.
[JM]: Right?
[DJ]: Right?
[DJ]: Yeah, it's no surprise that the nerds who like to do everything in the terminal also think that 80s neon is cool.
[DJ]: So I don't think we're surprising anyone with this revelation.
[JM]: Probably not.
[JM]: Okay, moving on to a fascinating article that I loved reading because it hit home on so many levels.
[JM]: And this article is called Needy Programs.
[JM]: I don't even know where to start.
[JM]: I'm going to start with the videos.
[JM]: This article describes multiple facets of needy programs.
[JM]: The first one, which you, Dan, have harped on on multiple occasions.
[JM]: And I'm going to do so again.
[JM]: I'm sure you are.
[JM]: And that's the fact that in most cases, we as users don't need accounts and yet programs keep insisting that we need them, even though as will be demonstrated as a clear theme, most of the times what these programs need is for the benefit of the people shipping the software, not the end users.
[JM]: So you and I don't want or need an account
[JM]: It is being foisted on us by the people who made these products.
[JM]: And so there are two videos in the beginning of this accounts section.
[JM]: Both of them are great for different reasons.
[JM]: The first one, this guy is trying to get a keyboard to work and realizes as he's testing it out for the first time, he says, how do I turn off the lights on my keyboard without logging in?
[JM]: And...
[JM]: becomes increasingly frustrated as it becomes evident that he's not going to be able to do that without creating an account and logging in.
[DJ]: The very statement, how do I turn off the lights on my keyboard without logging in, fills me with rage.
[JM]: And it filled this person with rage.
[JM]: And let's just say it didn't really end well for the keyboard.
[JM]: I hate that the poor keyboard got blamed for this, but it was in the wrong place at the wrong time.
[JM]: It's true.
[JM]: And the second video is also amazing.
[JM]: Just the very tangible amount of angst that this person is experiencing at having to constantly log into things.
[JM]: I highly recommend watching the second one in particular.
[JM]: It's really well done.
[JM]: And it really is absurd.
[JM]: You know, everything that we try to do now wants an account, all of it, all the time.
[JM]: It doesn't matter.
[JM]: Hardware, software, everything.
[JM]: Like a light bulb, something as simple as a light bulb that you used to just screw into a socket and flip a switch and it turned on.
[JM]: Light bulbs now need accounts.
[JM]: Like, sure, okay, it's not just a normal on-off bulb.
[JM]: It has additional functionality.
[JM]: But that doesn't mean that you need an account to get access to those other features.
[JM]: I have plenty of light bulbs in my house in this very moment that are smart to the extent that I can control them and change their color temperature and their brightness and colors and whatever other functions they have.
[JM]: all without creating an account.
[JM]: So these things are not synonymous.
[JM]: And yet, once again, we are being told that we have to create these accounts because that is in the best interest of the people making the products, not because it's in our best interest.
[DJ]: The account thing is such a rich vein.
[DJ]: I have previously complained about my speakers, which I otherwise like very much, but I did not enjoy having to agree to terms of service and a privacy policy.
[DJ]: A privacy policy in order to stream audio signals through a thing that would turn them into vibrations.
[DJ]: I really dislike that aspect of seemingly everything, but the thing that I love so much about this article
[DJ]: is it doesn't stop at accounts.
[DJ]: One of the other things that it calls out is notifications.
[DJ]: And we were just talking about you deactivating Apple intelligence and it having this kind of scary dialogue that sort of tries to get you not to do that.
[DJ]: And it struck me that, although that's not a notification, it's something popping up in response to what you're doing, it's a confirmation, it fits the needy programs model as well.
[DJ]: Because you want to turn this feature off.
[DJ]: For whatever reason, Apple thinks that you shouldn't do that.
[DJ]: Presumably, it's in their best interest for this feature to be on, to be successful, I guess, I suppose.
[DJ]: Or they're trying to head off...
[DJ]: customers complaining about, hey, this thing doesn't work.
[DJ]: And the answer is, well, yeah, you turned it off, dummy.
[DJ]: But either way, you get these design patterns where you want to turn the thing off.
[DJ]: And instead, there's this like, well, are you sure?
[DJ]: Are you sure you want to do that?
[DJ]: Do you really wouldn't you like to put your phone down and like maybe go away, take a couple of minutes, just let your heart rate settle, really think through what you're doing.
[DJ]: Like, there's something very patronizing about it.
[DJ]: I feel like that aspect is well covered by this article as well, where these programs, when the writer describes them as needy programs, you really want your software to just do what you say and then go away.
[DJ]: But instead, so many of the products that we use nowadays want to intrude into your life in all these various like large and small ways.
[DJ]: Large ways like banging you with notifications while you're just trying to go about your day.
[DJ]: And smaller ways, like the demand that you create an account and then have to manage credentials for it, or kind of walking you through this patronizing sequence of events when you try to simply alter certain options, because after all, they know better than you do.
[DJ]: A large corollary to this that I think I've called out before and will call out again is that I observe that the people who make these products are very passionate about them and that's great, but they confuse their own passion for something.
[DJ]: They confuse the importance that a product means to them with the importance it means to the user.
[DJ]: And I think everyone who makes a product needs to have the humility to
[DJ]: to say as much as I care about this and I want people to love it, even the people who love it don't care very much about it, right?
[DJ]: Like the product I work on for my job takes up an unusually large share of my mental attention, at least while I'm at work.
[DJ]: But like the apps I use on my phone, even the ones I love the most, I spend maybe 0.001% of my waking life thinking about them.
[DJ]: And you sometimes get this disconnect where you get emails from apps that are like, hey, we just updated this thing.
[DJ]: Hey, don't you want to hear from us more?
[DJ]: Hey, I'm going to pop up a thing when you're using the app.
[DJ]: It's like, I know you just want to check your email, but have you heard about our Black Friday sale?
[DJ]: And I think the fallacy there, I always want to take the people creating these apps and sit them down and going like, listen, I don't mean to be mean when I say this, but I don't care about you and your app.
[DJ]: I'm just trying to live my life.
[DJ]: So leave me alone.
[JM]: Spoiler alert, they're not gonna leave you alone because in addition to telling you to create an account and sending you notifications, like the onboarding screens is another area that the author of this article mentions, right?
[JM]: Like Apple telling you when you launch an app, oh, the news app now has these additional features.
[JM]: If I have to see one more Apple splash screen when launching an app just because they added some feature,
[JM]: These are not things that I need to be notified of with the frequency that Apple seems to put these screens in their apps.
[JM]: And the other thing is updates.
[JM]: This is a section of the article that really hit home for me, and I'm sure I've commented on before.
[JM]: I was going to say, this was written just for you, Justin.
[JM]: It really was.
[JM]: Listen to this quote, which I just think is amazing.
[JM]: Every program has an update mechanism now.
[JM]: Everybody is asking for updates all the time.
[JM]: Some notoriously bad ones lock you out until you update.
[JM]: And yet do we users really need these updates?
[JM]: Did we ask for them?
[JM]: And no, of course we didn't.
[JM]: No, we didn't ask for these.
[JM]: And a lot of these things just update themselves now on their own.
[JM]: And oftentimes not even asking you if that's the behavior you want.
[JM]: Once again, it's like, you don't get to choose whether we auto update because again, it's a needy program.
[JM]: It's about serving its needs and the people that created it, not our needs as the end users, you know, the customers, the people that are supposed to be the most important.
[DJ]: There are definitely updates that are important when they close bugs and security holes, or when they legitimately do add great new features to software, that happens.
[DJ]: Yet at the same time, as a user, I would like the software to be, sounds like a weird way to put it, more respectful.
[DJ]: I don't mean more respectful of my rank in society, but I mean more respectful of my time and attention.
[DJ]: As opposed to this attitude of like, well, we're making this software and it's going to do all of this stuff that's annoying and you're just going to have to put up with it.
[DJ]: I would like to see a software developer that puts minimize user annoyance at the center of their attitude.
[DJ]: But this article really is, I think, going to become part of our podcast canon because it so perfectly summarizes a set of things that annoy us about software and technology.
[DJ]: And so I just really want to encourage everyone to give it a read.
[DJ]: And kudos to the author, who it turns out also made the FIRA code font, which I'm a big fan of, because I feel like we're going to refer back to this frequently.
[JM]: I think you're totally right.
[JM]: I can just picture future references when we're talking about something and being like, haha, yep, needy programs.
[JM]: It's just going to become part of the abstractions podcast parlance.
[JM]: Okay, moving on to something that I noticed recently, and that is that Google has killed the youtube-nocookie.com link behavior.
[JM]: And when I say killed, I don't mean that the feature itself is dead.
[JM]: What I mean is that the following used to work and no longer works.
[JM]: So oftentimes people will send me YouTube links.
[JM]: And I...
[JM]: Don't really care for Google, as I've probably made clear on this podcast, at least as it relates to a lot of things that they do.
[JM]: And I don't watch YouTube outside of private browser windows just because, I don't know, it's just a habit that I have.
[JM]: And every time I go to a private browser window and put in a YouTube link, I get some cookie notice or something or other that I have to dismiss.
[JM]: And then once I'm on YouTube, instead of the video being front and center, it's a relatively small part of the viewport.
[JM]: And most of the page is being taken up by comments and comments.
[JM]: other videos I might wanna watch on the right-hand side.
[JM]: And I just find YouTube to be a relatively user hostile experience, at least for me.
[JM]: And I know that people love it and that's great to each their own.
[JM]: So I usually don't just tap on YouTube links.
[JM]: Instead, what I do is I copy it, I paste it into my browser window and I modify it slightly and then I hit return to load it.
[JM]: And the way that I modify it is I change the domain to be youtube-nocookie.com slash embed slash and then the identifier.
[JM]: I have been doing this for so long that it's just become automatic to me.
[JM]: And if I am going to share a YouTube video with other people, that is how I share them.
[JM]: And okay, so what does this change?
[JM]: Like, why do this?
[JM]: Well, for one thing, it's right there in the domain, nocookie.com.
[JM]: I don't want one.
[JM]: I don't need one.
[JM]: You don't need to ask me if I want them or not because I've already told you that I don't.
[JM]: It's right there in the domain.
[JM]: And also when you load this, all of that other cruft is gone.
[JM]: The video takes up the entire frame.
[JM]: There's no other stuff.
[JM]: There are no comments.
[JM]: There's no, here's other videos you might like to watch.
[JM]: It's just the video.
[JM]: You know, the thing that I came here for.
[JM]: Well, as of about a week or two ago, this no longer works.
[JM]: It's almost like Google realized that people were doing this because the intention of this is also right there in the URL structure.
[JM]: And that's that it is to embed this video in something else.
[JM]: And I think at some point they realized that people like me were using it to, you know, skip all the other stuff
[JM]: the needy program that is YouTube wants me to see, which are the comments and the like button and the subscribe button and the here's other videos you might like to watch.
[JM]: All this other stuff that the needy program wants, but that I don't care about at all.
[JM]: But now if you visit a link that is structured in this way, you get the oh so helpful error 153 video player configuration error and a link to watch the video on YouTube.
[JM]: And so now I, and presumably a decent number of other folks, because if there weren't other folks, Google probably wouldn't have cared and wouldn't have disabled this feature.
[DJ]: You don't think they're targeting you specifically, Justin?
[JM]: Probably.
[JM]: So now me and all these other people that have tons of saved links in this format, people who have shared links in this format with other people, they're all busted.
[JM]: They're all going to receive this oh so helpful error 153 message.
[JM]: So thanks, Google.
[DJ]: Well, and presumably, aren't there probably a lot of videos embedded in websites about for which this was the intended use?
[DJ]: Are those broken too?
[JM]: I would imagine that the answer to that question is no.
[JM]: I believe what Google has done here is to say, oh, you're using this in a way in which we did not envision or condone, and so we're gonna break that part.
[JM]: But if you're embedding it, it'll continue to work fine.
[JM]: That's how I presume Google has done this.
[JM]: So I'm gonna go back to the way that I sometimes have watched YouTube things in the past, which is to use yt-dlp on the command line,
[JM]: and just download the thing and then watch it and then throw it in the trash.
[JM]: So sorry, Google, you're still not gonna get what you want, at least not from me.
[DJ]: That's exactly the solution I had running through my brain as you were describing this, as I'm like, well, it could just, when someone shares a YouTube link with me, I could just paste it into a downloader and then watch it and then throw it away, so.
[DJ]: Agreed.
[DJ]: I might do that too.
[DJ]: It's a form of resistance, if you like, to the user-hostile experience of using YouTube, which is obnoxious, aside from the tracking or whatever else.
[DJ]: I don't like going to that website, which is so focused on trying to get you to then watch other stuff.
[DJ]: That seems to be the main purpose of YouTube.com, is to keep you there watching movies so you can see ads, of course.
[DJ]: So, no thanks.
[JM]: No, thanks indeed.
[JM]: Okay, moving on to a story that was definitely news to me.
[JM]: It seems that unpowered solid state drives will slowly lose data.
[JM]: Now, if you hear this and you think, yeah, duh,
[JM]: then why didn't you tell me this before?
[JM]: Because now I'm mad at you, theoretical person who already knew this and who didn't tell me, for not having told me.
[JM]: Because this is not something that I realized.
[JM]: I did not understand that if you have a computer with a solid state drive in it, and you put that computer in a closet and leave it there for a certain period of time, that it will just slowly...
[JM]: lose data to the point where you probably can't even turn on that computer anymore, because the data on the solid state drive has degraded enough that it just doesn't boot or can't read the data on it or whatever.
[JM]: I don't exactly understand what the failure modes are when they quote lose data, but presumably it's not good.
[JM]: And
[JM]: The first thing that I thought of when I saw this is, wait a minute, a couple of weeks ago, I pulled out a 2018 Mac mini that I had sitting on a shelf that I hadn't used in a while.
[JM]: And I wanted to get some of the files off of this computer.
[JM]: So I connected it to a monitor and a keyboard and a mouse, turned it on and nothing, just a black screen.
[JM]: I thought, okay, I don't know, maybe it needs a minute.
[JM]: So I went, had lunch, came back, nothing.
[JM]: Held down the power button until it turned off, turned it on again, nothing.
[JM]: And so now I'm looking at this article and thinking, is that what happened?
[JM]: Because this thing was on a shelf for a while.
[JM]: And so now I'm wondering, could that have been prevented?
[JM]: Like if I had just periodically connected it to power, would it not be a brick like it currently seems to be?
[JM]: Because if so, that's information I would like to have known before.
[JM]: And...
[JM]: mentioned, if memory serves in this article is a distinction between certain memory cell types in solid state drives, and how older solid state drives are less prone to this problem than more recent ones.
[JM]: And so to test this theory, right before we started recording today, I grabbed my beloved and treasured
[JM]: 2010 or 2011, I don't remember exactly which year, 11-inch MacBook Air.
[JM]: And with great trepidation, connected it to power, gave it a few minutes, and opened the lid to find that everything is right as rain, login screen, logs right in.
[DJ]: But then it automatically updated to macOS Tahoe, and it was all over.
[JM]: I don't know what OS this computer supports, but it's definitely not tall.
[JM]: So I was relieved to see that this adorable little Mac of yesteryear still continues to function and does not seem to be as afflicted by this problem as perhaps my much more recent and powerful 2018 Mac mini is.
[JM]: But for me, this was a real wake-up call.
[JM]: I don't have a ton of
[JM]: old computers sitting around particularly ones that were made in the last 10 years, which seemed to be more susceptible to this kind of problem.
[JM]: But when you think about some of the older computers that we have lying around, it's not great from a nostalgic computer preservation standpoint, right?
[JM]: Like if
[JM]: There are computers being made anywhere in the last decade, including the ones today that you think, oh, I might want to hold onto this and 20 years from now, fire it up and be reminded of what it was like to use this computer back in the day.
[JM]: That's looking less and less like a viable thing you can do unless you somehow set some reminder every X number of, I don't know what the time period is months and connect it to power.
[JM]: allow the SSD cells to refresh or what however this process works but the story has also rekindled for me my love of spinning rust otherwise known as old-school magnetic metal spinning platter hard drives because they do not really have this problem they have other problems which is there are moving parts that can fail and
[JM]: I've probably had more hard drives fail on me so far than solid state drives, but at least you can take a hard drive in theory and stick it in a drawer and pull it out five years later.
[JM]: And there's a decent chance everything will be fine.
[JM]: Doesn't really sound like that's how it works with solid state drives.
[JM]: So.
[JM]: Just an important thing to keep in mind.
[JM]: Okay, moving on in other news, there's been a bit of a backlash regarding generative software.
[JM]: Well, I mean, that's been happening in general, yes.
[JM]: But there are two particular stories recently that I thought were interesting.
[JM]: And they are related to Microsoft and Mozilla, respectively.
[JM]: Let's start with Microsoft.
[JM]: Microsoft has been in the news a lot as it relates to all things generative software, they've invested heavily in open AI.
[JM]: They've pushed their co-pilot large language model software stuff into GitHub, Windows, Microsoft Office, every nook and cranny of the Microsoft ecosystem that they can.
[JM]: And there are folks that don't really like that.
[JM]: And there are folks that don't like some of the other things that Microsoft has done related to their generative software.
[JM]: But I want to refer specifically to the head of Microsoft's AI division because he posted the following on social media.
[JM]: Jeez, there are so many cynics.
[JM]: It cracks me up when I hear people call AI underwhelming.
[JM]: I grew up playing Snake on a Nokia phone.
[JM]: The fact that people are unimpressed that we can have a fluent conversation with a super smart AI that can generate any image slash video is mind-blowing to me."
[JM]: I just love the, I grew up playing snake on a Nokia phone.
[JM]: I mean, same bro.
[JM]: I used to program by carving hieroglyphics into stone tablets and they connected it to my steam engine.
[JM]: I mean, it's like we, we all have stories, right?
[JM]: Like.
[JM]: Yeah.
[DJ]: Yeah.
[JM]: From, from the early ages that we grew up as it relates to technology.
[DJ]: To program computers.
[DJ]: I used to have to go through the waist deep snow uphill both ways.
Yeah.
[JM]: It's just a weird thing to inject into a defense of one's generative software efforts.
[DJ]: It's interesting because I understand where he's coming from, which is we should be impressed by the overall arc of technology.
[DJ]: But he commits the error of so many people in this space, which is to focus only on the
[DJ]: admittedly very impressive capabilities of large language model powered software, but then using the existence of those capabilities to act like there are no problems with it, or the problems are trivial, or it's only a matter of time until we solve the problems.
[DJ]: Because it's a little disingenuous to say, I can't believe people aren't impressed by being able to have a fluent conversation with a, what do you call it?
[DJ]: Super smart, super intelligent, super whatever-
[DJ]: Super smart AI.
[DJ]: I mean, it's none of those things, for starters, but it is very capable.
[DJ]: It's capable of, as he says, generating any image.
[DJ]: Right.
[DJ]: But at least up until very recently and through the present day, there are many images it cannot generate, like images with coherent text in them, although apparently Google's latest image generation model is very good at that, which is cool.
[DJ]: But there is a disingenuousness
[DJ]: about that attitude.
[DJ]: And I think in fact, if anything, it's that attitude that annoys people as much or more than the software itself.
[DJ]: It's the attitude of people who go, stop talking about the problems with this because it's amazing.
[DJ]: But that's just as unfair as the people who insist that there's nothing the least bit interesting about this technology because of its downsides.
[DJ]: The truth lies in the middle, which is yes, the long arc of technology is
[DJ]: that has brought us from 2D snake on a Nokia phone all the way up to being able to indeed carry on a fluent conversation with a piece of software is very impressive.
[DJ]: And also, here are all these problems that are real problems.
[JM]: I agree that this is a disingenuous take because you can be impressed by some aspects of large language models, for example, and find the entire arc so far of so-called artificial intelligence to be underwhelming.
[JM]: Both things can totally be true.
[JM]: And it's disingenuous of this person to imply otherwise.
[JM]: So to summarize, Microsoft executives are confused and upset that people don't appreciate their generative software features and they don't like people giving them hell about it on social media and respond with,
[JM]: How are people not impressed with this stuff?
[JM]: And if I'm being totally blunt, my response to that is, bro, have you used the internet lately?
[JM]: It's trashed.
[JM]: Have you used Windows?
[JM]: Trashed.
[JM]: It's garbage.
[JM]: People are sick of it being trash because it didn't used to be as trashed as this.
[JM]: You've gone and trashed everything.
[JM]: So that's why people aren't happy with you.
[JM]: And side note, the person that said this is the Microsoft AI CEO.
[JM]: Is it me or has job title inflation just totally jumped the shark?
[JM]: How can he be the chief executive officer if he reports to Microsoft CEO Satya Nadella?
[JM]: What's next?
[JM]: Microsoft Windows CEO?
[JM]: Xbox CEO?
[JM]: You get to be CEO.
[JM]: You get to be CEO.
[JM]: I don't get this.
[JM]: It makes no sense.
[DJ]: It doesn't help this person look less childish.
[DJ]: Let's put it that way.
[DJ]: Meanwhile, at the absolute other end of successful technology companies, Mozilla.
[DJ]: Oh, Mozilla.
[DJ]: It had such promise, Justin.
[JM]: Mozilla did have lots of promise and they are in the news as of late because of things that they have announced as it relates to Mozilla trying to stuff a bunch of generative software stuff into Firefox.
[JM]: And I don't really know what to say about this.
[JM]: Other than it sucks.
[JM]: Other than I'm going to put a link to an article that takes apart this whole idea and you should just read it.
[JM]: It is fantastic.
[JM]: It sounds like one of the features is an AI window that's coming to Firefox.
[JM]: which appears to emulate the user experience that AI browsers like OpenAI's Atlas that we discussed on a previous show or Perplexity's Comet offer, which is instead of performing a search from the address bar and interacting with websites like...
[JM]: we've always been doing with a browser, your interaction is going to be with a chat bot, which then mediates between you and the web.
[JM]: And this is just such a sad end to Firefox.
[JM]: In my opinion, I've wanted to hold onto it both because it's one of the few major independent browsers left.
[JM]: It has its own rendering engine.
[JM]: So it helps address the problem of
[JM]: monoculture in terms of browser rendering engines, or at least it did back when they had some degree of market share, but now their market share is so low that it's not even really providing that value anymore.
[JM]: So I find that I have stopped using it as my backup browser, as I often did, as I mentioned in a previous episode, I'm now been using helium, which is based on chromium.
[JM]: So in the end, I have to agree with the author of this article when he said, just let Mozilla die.
[JM]: It no longer serves its stated purpose.
[JM]: And sadly, I think he's right.
[DJ]: Yeah, it is sad because I mean, I know I'm under the impression at least that Mozilla has been struggling.
[DJ]: because they have expenses they got to pay, you know, like they have to generate revenue somehow, I believe, although the company is this strange split between like a foundation and a corporation.
[DJ]: But you'd almost wish that they would double down on what we would tend to call the traditional web, the open web, so that there could be like one holdout, because I presume that it's only a matter of time before the browsers themselves, the main ones like Safari and Chrome,
[DJ]: build in more and more of these features until there's no way to use the internet that isn't, as you say, intermediated through a large language model.
[DJ]: I would like there to be some holdout, like some really grassroots, we're keeping the web the way it's always been, the web that we know and love.
[DJ]: And it feels like Mozilla and Firefox would be the ones to rise to that call.
[DJ]: And so it's like doubly disappointing to see that, no, they're just doing the same, quote, doubling down on AI, as every other company seems to feel obligated to do right now, which is too bad.
[JM]: There's an amazing meme strip in this article, which emulates a conversation between two characters on Friends, where one of them is telling Joey...
[JM]: Be the, and then he repeats the words, stewards of the open web.
[JM]: Let's do it all together.
[JM]: Be the stewards of the open web.
[JM]: Oh, put AI in everything.
[JM]: This is his response.
[JM]: So good.
[JM]: The good news, at least for me, is that there do seem to be other people that make browsers that care about the open web.
[JM]: So whether that's Helium or some other alternative, go and find one of those if the open web is something that you care about.
[JM]: All right, that's all for this episode.
[JM]: Thanks everyone for listening.
[JM]: You can find me on the web at justinmayer.com and you can find Dan on the web at danj.ca.
[JM]: Reach out with your thoughts about this episode on the FedEvers via justin.ramble.space.