Ask Paul: March 14 (Premium)

Happy Friday! We have friends in town, our third week straight with visitors, so it’s been super-busy and we’ll be away over the weekend. Complicating matters, Laurent was off yesterday and today, so I’ve been trying to keep up with the news, too. And I had to get up at 4:30 am yesterday for a sunrise canal boat thing, and … I’m exhausted. So please bear with me. 🙂

? Snapdragon X v2

timo47 asks:

With the upcoming next gen Snapdragon chips, what are the chances of a version that won’t require fans? My feeling is that primary focus will be on making them more powerful to match Apple, and not so much on increased efficiency?

What I’ve been told is that the focus on v1 was on CPU and the focus on v2 is GPU. I would have almost assumed that NPU was the focus, but I guess that was just happening regardless. What Microsoft/Qualcomm needed was an Arm chip that could meet the performance of x86 for real. And it had always fallen short before. Clearly, they nailed it on v1 in that regard–with some help from Microsoft’s improved Prism emulator–but now they need a similar push for GPU. And so that’s happening.

I will say, I don’t quite understand or even agree with the expansion of Snapdragon X into lower-end versions mid-stream on v1. They added Plus variants, then more plus variants, and then an even lower X tier. And while some of that can be explained by binning, now it feels like they’re just treading water until v2 is ready. I’m surprised this is taking so long.

Regarding the fan issue, I suppose lower-end variants could be used in fanless systems, but these things currently sell in low enough numbers that creating unique designs just for that many not make sense. For example, Microsoft will never make two Surface Pro designs, one for a fanless version and one with fans, unless it can sell enough of both. But this may not matter. While I was initially worried about this, fans aren’t a real-world issue with Snapdragon X-based PCs, you never hear them anyway. And having them there means you can at least use the compute when needed, unlike with Apple’s fanless MacBook Air, where all it can do is slow down the processor to lower heat.

Related to this, Pakeha asks:

Pretty much everything for ARM is laptops right now. Hopefully, the desktop PC will come. I am looking forward the NUC type computers. Any thoughts on the possibility before 2026?

The first Snapdragon X-based NUCs were announced at CES in January, so that’s happening now. (Geekom announced theirs early, angering Qualcomm.) The bigger push on the desktop PC side will likely occur when the Snapdragon X v2 chips arrive–I’m hearing September–with improved GPUs. And/or the MediaTek/Nvidia chipset(s), which I believe are on a similar schedule.

Given the nature of NUC form factor PCs, though, there’s no reason current-gen Snapdragon X chips can’t be quite effective right now. But this is still a low-volume thing, and most of the emphasis will be on laptops, which makes sense. That’s what most people want and use. There may also be cost issues, tied to the binning/lower-end chips noted above. I know Qualcomm has tried to be aggressive on pricing, but I suspect they’re still mostly on the premium side of the pricing spectrum. Regardless, Intel (especially) and AMD sell at a much higher volume, so it’s likely that the cost of those chips to system builders is lower.

? Rotten to the core?

will asks:

Curious if you agree with John Gruber’s recent post on Apple, and how they misled everyone with Apple Intelligence and what was coming? Not to mention the promise of things with the iPhone 16 that was false. Should there be some accountability from Apple on this?

In the Apple-verse, this is high drama, but even in our world, it’s news because Apple has an audience base of billions and the iPhone basically defines the smartphone market. And it is news. But I feel like the drama is slightly overblown. We can pretend otherwise, but this isn’t the first time Apple has promised and underdelivered. And that was true when Steve Jobs ran the show too.

The most obvious comparison is Apple Maps, and in that case, Cook handled it properly, by forcing Scott Forestall to apology publicly. When he refused, Cook fired him.

In this case, I don’t think we have a single Forestall figure to blame, unless you want to point at Cook. I think that Apple, like Microsoft, Google, and whomever else, saw that AI was a sudden and unwelcome strategic requirement, and that Cook directed all his product leaders to figure this out. Apple is unique in some ways, and one thing it did announce last June that I still find to be quite credible and responsible is to make that privacy promise with its AI, backed by Private Cloud Compute. And on some level, one might argue that Apple is “behind” on AI, but I see it more as, Apple promised to do AI responsibly for its users, and that is a hard problem. What they have so far is pretty much what we see elsewhere, just without the chaos: Image creation, writing help, communication, etc.

Siri is problematic. But Siri has always been problematic, right? And as it stands today, it’s not like it’s worse than it was. But the issue here is that taking something that lackluster and making it incredible while adhering to its self-imposed privacy and reliability promises is quite a task. Even Google, which has been about as chaotic as Microsoft in rolling out AI functionality, stripped out the ability to generate human images for several months. Apple’s bar is higher.

So the issue here is essentially vaporware and broken promises. Apple likely thought they’d be able to bring their enormous resources to bear to solve the problems they created for themselves and they failed. And in that sense, that is a bit like the problem with that trashcan Mac thing where they had created this elegant (Intel-based) box but then could never upgrade it meaningfully, and so they had to start over. The difference is scope: Siri probably used by over one billion people fairly regularly, whereas the audience for that Mac was tiny. So it’s amplified.

We’ll see what happens, but none of this bothers me per se. I don’t use Siri personally, and it only comes up when I say something that sounds like Siri and my screen lights up. But Apple Intelligence, broadly, is in a good place. One might argue that Gemini is better right now, and it probably is. But you can use Gemini on the iPhone too (I do), or any other AI chat app. So it’s not like anyone is hurting per se. The built-in stuff is probably good enough for many now, and Apple fans are loyal enough that they’ll leave it all on for the most part, and as it gets better, more and more will use it. I certainly notice Apple Intelligence features in apps like Messages. If you use Mail and other Apple apps, that exposure is much bigger.

To Gruber’s overly dramatic pontification, that’s his role in that community. It’s fine. It’s a bit much in my opinion, but I probably do that sort of thing with Windows and Microsoft more than I want to admit, and I’m sure that’s what I look like to those guys. Apple does have a higher bar, though. And where Windows 11 feels like it’s run by idiots, Apple throws off a more responsible, adult vibe. So I do get it. It’s disappointing on a different level.

But it’s still early, too. Cook could very well have already orchestrated something similar to the Apple Maps/Forestall thing. This isn’t the most transparent company, but it could also discuss the challenges with AI at WWDC and just be upfront about it. We’ll see.

For now, I look at what’s available through Apple Intelligence and it’s fine. And even if Apple nails it with Siri, all they’ll accomplish is leveling the playing field. No one will ever move to Apple products for Siri. This is about keeping their existing customers from looking elsewhere.

There are parallels here with Microsoft’s and Bill Gates’ sudden embrace of the Internet in 1995. It was nothing to worry about and then it was the only thing to worry about. But when Gates directly all of Microsoft to add Internet functionality, there were no restraints. Again, Apple has raised the bar on what I’ll call responsibility. This is a lot harder than just doing it, damn the torpedoes.

? Office outside the Microsoft universe

Christian-Gaeng asks:

Maybe a strange question for a Windows nerd like you, but I’ll ask anyway. ? One of the reasons I’ve avoided getting more involved with Linux is that I want to continue using the Microsoft Office suite. I haven’t been able to get Office 365 to work on Linux yet. But since I’ve been using Office 2024 without a subscription for some time now, I wanted to ask you (or the community) if you have any experience with Office on Linux. Thanks.

I assume you’re referring to the desktop suite in particular, and have found that solutions like Wine and CrossOver can only run older Microsoft Office versions effectively or at all. (Whether this is tied to Office 365/Microsoft 365 is unclear and perhaps beside the point, but also a bit suspicious.) But the short answer is no, not any recent experience trying this. If I were to move to Linux, I wouldn’t try to run Office locally, I would just rely on the web apps. I don’t use Office daily now in Windows, so that wouldn’t be a huge issue for me.

If this is your only major blocker on Linux, all I can really recommend is trying third-party Office productivity suites that do run natively on Linux, like LibreOffice. (As I see others have recommended, too.) I assume you have tried such things, but I found LibreOffice Writer to be a perfectly acceptable alternative to Microsoft Word, and I’m sure some of the other options are likewise good.

Related to this, spacecamel asks:

By the way, what is the status of the legendary Linux article that you keep working on?

I’m not planning a big Linux article of any kind, so I’m not sure what you mean. My status with Linux is the same as ever. I have at least one install on a dedicated PC (here in Mexico, that’s the latest Mint Linux right now, but I usually have more going on back in PA). And I always check out major new distribution versions. But my issues with Linux are a little more complex, in part because using whatever cloud storage sync service (Google Drive, OneDrive, whatever) is central to my workflow, and none of them work seamlessly with Linux. If that changes, I would get more serious about considering using Linux more regularly. But for now, it’s mostly just about keeping up on that world.

(Related to this, Google Drive and OneDrive are both available in first-class form on the Mac, and I have zero interest in switching to that platform. Linux is enough like Windows that if that were the case there, things might be different.)

? You can’t spell air without AI

eeisner asks:

I’m curious your thoughts on always-on AI devices, like the bee.computer that Leo wears. Do you think these are the next smartphone or the next Google Glass? Do you see a future for yourself or your family where perpetual AI assistants make sense?

Yes, and I think this will happen sooner rather than later. It will happen in stages, and it will start with smartphones and smartwatches, because we already have those devices with us all the time. We’re already seeing this notion of AI that will personalize itself to your needs, and for that to work, it needs to know more about you, just as previous-generation digital personal assistants did. (You can see this in the latest Gemini announcements.) You can do that thing we’ve discussed, where you go for a walk and just core dump on it by talking to the AI, giving it your ideas, or whatever, and then asking for a summary when you’re done, or maybe some feedback. Or you can let it do this automatically in the near future, where these things will have an always-on mode at your choice, with the goal of it learning about it.

If you think about a tool like Recall, and can get by all the faux outrage, what it’s really doing is automating the collection of data, in this case about your daily activities. And the express aim of it is that recording that data will allow you to later “go back in time” and use it to “remember” something that you did earlier. For so many people, that is a useful bit of functionality, but we’re only on our PCs sometimes, and usually only for work. So the notion of something that will record your other activities, be them personal or work-related, with the same general aim, makes sense too. Makes even more sense, maybe.

(Tied to the Siri bit above, one can easily imagine how Apple devices “remembering” what you do can construct a more complete picture of your activities because these people will have several different devices for different purposes. It’s almost like building a DNA sequence, but for your personal experiences, and in real-time.)

How and where this information is recorded will evolve to smaller and smaller devices, which may be wearables at first (wrist-based, but also glasses, hearing aids/earbuds, whatever) and then will eventually become more sophisticated implanted devices that could start as contact lenses or tattoo-like things but, again, become smaller, less invasive, and more automated.

I’m ignoring the privacy concerns here, but also the real-world implications of what could happen when a hacker gains access to that kind of data because it’s just too big a topic and there’s too much technology to figure out first. But obviously there will be those who oppose this stuff, and there will be a bigger majority of people who will wonder how humanity even worked without it in previous generation. It’s just going to happen.

? No pain, no gain

tomhere asks:

Do you think you could raise the gain on your First Ring Daily podcast volume? It is so much quieter than any other podcast I listen to. I actually can’t get the volume on my earbuds loud enough to hear you and Brad when I am walking outside. (Yes I actually paid for a premium membership just to post that comment. 🙂 ). BTW, First Ring Daily reminds me of the conversations we shared at the dot-com I worked at in the late 90s. Good times.

I almost don’t even understand what the term “gain” means. For the most part, I just show up for podcasts and someone else records, edits, and publishes the show. (The exception is the videos I make with my wife for Eternal Spring, and that’s a nightmare.) But because Brad is the person who does all the work on that, and I just show up using the same PC(s) and hardware/software configuration I use with Windows Weekly and Hands-On Windows, I can only conclude that this is on him. And so I mentioned it to him this morning, and he made some minor adjustments. I didn’t notice this issue, so I’m not sure whether I’ll notice whatever changes, or whether you will. I don’t know anything about this, really. Sorry.

? Samsunk

epsjrno asks:

Just curious if you find anything about the Galaxy S25 series compelling or do you plan to continue going back and forth with the Pixel and iPhone.

No, I’m not at all interested in the new Samsung smartphones.

I will continue switching between Pixel and iPhone for now, and unless something changes, what I have my eye on for the future is a folding smartphone of whatever kind that could replace my tablet in addition to being a phone. Pixel has gotten close, but whatever Apple might do in the future is by nature super-interesting as well, especially since I very much prefer the iPad to any Android tablet.

?‍♂️ What the …?

JEAM asks:

Been wondering about this for a while: what happened with the “What the Tech” podcast that you used to do with Andrew Zarian? It seems to have simply come to a stop, with no warning or explanation. I greatly enjoyed your interactions with Andrew and was sorry when the podcast stopped. Any details you feel appropriate to share will be appreciated, thanks. I hope the reason was not due to some terrible personal news.

I feel like I explained this as much as I can somewhere, but I can’t find that. The short version is that I don’t really know what happened, though I have a theory. And that theory is tied to a third person’s privacy, so I can’t explain it fully in public. Which I realize is unsatisfying, sorry.

I haven’t spoken to Andrew in almost exactly four years. Microsoft is about to kill Skype, so I will soon lose this record, but our last interaction, which was typical and in no way portentous, was just about the usual scheduling issues he/we had at the time. And then … nothing. He never contacted me again.

I will always love Andrew and his family, and I miss him. My theory is tied to a misunderstanding I think he had. And that’s about all I can say, I guess. Nothing malicious or bad. No bad feelings on my end, for sure. Weird, in many ways, and unexpected. But it’s been four years.