Ask Paul: May 22 ⭐

Happy Friday! We’re heading off to the Finger Lakes for the Memorial Day weekend, but first things first. Here’s another big install of Ask Paul to kick things off.

🎮 More Laurent is the new More Cowbell

christianwilson asks:

I was pleasantly surprised to see Laurent’s early impression review of Forza Horizon 6. I’d like to see more of these kind of articles from Laurent, assuming he’s interested in doing so.

You’re not alone on that. I would never ask such a thing of him, as he already does so much, but he asked about potentially reviewing Forza Horizon 6 and I was of course all over that happening. Similarly, he’s looking into getting a review unit of the Fitbit Air, and if that doesn’t come through for whatever reason, I will just buy him one so he can review it. The lack of a screen makes that particular device uninteresting to me personally, but Fitbit/Google Health and all the changes that are happening there are important.

Again, I would never ask this of him. He writes the news, of course, and I’ve always been super-happy with the quality of his writing. And less obviously to those outside our little company, he’s the first line of defense when it comes to support issues, and he always does more than anyone has a right to expect. If something falls back to me, God help us all, or to Robert because it’s a technical issue, then something is really wrong. Laurent just handles things in a very professional manner, it’s very much appreciated.

One thing that will likewise not be obvious from the outside is that the two of us have various back-and-forths regarding whatever we’re writing. There are examples of him covering a news story and me following up with a Premium editorial, us dividing news up (for industry events, for example), and either one of us telling the other about some thing that’s happening and then we decide between us who’s covering it. I will take on certain topics that I know he either doesn’t care about too much or are just in my wheelhouse–like developer-related things–and I assume, usually correctly, when there’s something I know he’d prefer to write. It’s a nice dynamic.

But he is interested in writing reviews and I am all for that. With the understanding that it’s not expected, but if that’s what he wants to do, I am very happy to see it happen. It’s nice to see that others appreciate it as well.

“You Might Want to Wait to Buy a New iPhone”

Headlines that write themselves, part one

🧑‍💻 What happens when no one can read and debug code?

justme asks:

A thought struck me reading one of your recent articles regarding vibe coding – in an era where AI can write code and rewrite code the way it does, how do we preserve the ability to debug and troubleshoot. This thought was brought further to the front of mind by a recent Techradar article regarding NASA’s Voyager probes.

If you’re not familiar with Isaac Asimov’s Foundation series, one of the premises behind the plot is that there’s a Galactic Empire in decay. There is technology from the past that is superior to anything in the in-book current day, and there are no people alive at that time who understand it or can fix it if something goes wrong. So when an advanced faster-than-light spaceship, or whatever, breaks down, it’s just junked. As a result, a mathematician predicts when this empire will collapse, trigger a dark age, and he sets up a foundation (later discovered to be two) that can store and protect our collective knowledge and then lead mankind, eventually, out of that dark age.

This is very much like the code in the Voyager probes, though in this case it doesn’t matter all that much since no person will ever need to read, rewrite, or debug that code. And even if that does happen, we already have AI today that is helping us read previously indecipherable ancient texts. So it’s reasonable to think that future humans, or perhaps even other life forms, will make short work out of that particular 1970s-era code.

I think we can apply that thinking to the vibe coding issue as well. I don’t believe I wrote this anywhere, but vibe coding now reminds me of the Alanis Morrissette song Ironic, which was criticized on release because the events described in the song are all coincidental, not ironic. But this song was big enough that the meaning of the term ironic literally changed, which is annoying to people like me, but also how language works. This is all made up anyway, so the popular understanding of any term literally becomes what that term means.

With vibe coding as originally envisioned by its creator and misunderstood by just about everyone else, it wasn’t normal, mainstream users who would be describing an app so AI could create it. This was for professional programmers. It was not really coding. It was fun for throwaway weekend projects. Etc.

The thing is, AI is advancing so fast that in the 14 months I wrote that article, vibe coding is now what everyone thought it was, or soon will be. That is, anyone can sit down with whatever chatbot, or with Google AI Studio, as I did, and create an app. As with anything technology-related, it will usually be the more technical people who do this first. But this will very quickly move into the mainstream. One of the other things Google announced recently was that Pixel users, and then all Android users, will soon be able to vibe code widgets that will work on phones, tablets, and Android Auto. It didn’t use the term vibe coding, but that’s what this is. Widgets are surfaces for apps today, but in this new model, they can also be standalone mini-apps that draw in data from multiple apps and services.

There’s a lot that’s fascinating about that, but the big one, to me, is that normal, mainstream users will almost certainly come up with some of the best ideas for widgets, and then apps, too. And that’s because they’re not developers or engineers, and they don’t maybe understand limits or norms or whatever. They just understand what they want. I am very curious about what comes out of this.

To your point, few of these people will know or care how these things are made. They will never know the underlying programming languages, frameworks, interconnective technologies (MCP, AppIntents, whatever) or anything else. But they won’t have to. This will make things like automation and IFTTT-type solutions obsolete, in one sense, but these were never going to be used by mainstream users anyway. This is bringing that power, and more, to the masses.

In my case, I understand programming topics to whatever degree, and I know some languages and frameworks pretty well. For the two vibe coding projects I’ve written about, one a web app and the other a native Android app, I am not familiar enough with TypeScript/React or Kotlin/Jetpack Compose to effectively read/understand any of that code and then add to it, debug it, or whatever else. I have also created (but not written about) a new version of the Android app version of this Markdown editor from scratch in Android Studio using Gemini 3.5 Flash, and I don’t understand that either of course. Whereas I was/am familiar enough with C#/Windows App SDK to fully understand what the AI was doing when it helped me finish WinUIpad recently.

Is this a problem?

On the one hand, I know that I could learn TypeScript/React or Kotlin/Jetpack Compose well enough to understand those codebases. I also know that I don’t want to learn either, not really, and there is this growing sense that none of this matters anyway. If AI can fix things, if a human operator can simply describe what they want until it’s perfect, in their view, and if this things are inherently disposable, so to speak, then maybe it doesn’t matter.

But maybe it does. One of the issues I ran into with the Android Studio-created app is that it made a ridiculously minimal UI, so the app just appeared as a full-screen sheet of white. I thought it was broken, so Gemini spent a solid 20 minutes debugging it, during which I could see the app shutting down, restarting, and doing things in front of me, until it finally reported back that it was fine and working as expected. All I had to do was type some text, type Ctrl + S to save, and I could get on with my life.

The problem is that I did this on a Windows 11 on Arm machine and Android Studio device emulators do not work on that platform, so I have to deploy the app to a physical device. There’s no way to type Ctrl + S on an Android phone or tablet because the onscreen keyboard does not support that. Plus, I want some UI. So I prompted the thing to add the UI. I showed it a screenshot of the AI Studio-created app as a reference, with menus and a floating formatting toolbar. And … Gemini hasn’t worked ever since. It spews out nonsense and tells me to try later. I just tried it again this morning and it’s still not working.

And that, to me, feels very much like that premise from Foundation and speaks to the issue you raise. So it’s real, and it could (and does) happen. But I feel like we have to have faith in two things. One, that there will always be people who do understand coding and other technology and that they will be necessary for truly important apps/services, or whatever. And two, that AI simply keeps improving and that the issues I or we see now are temporary. Put another way, you can plan for the future to some degree. But you can’t plan for the apocalypse, not really, so you gotta have faith.

The ability to write code, debug, and troubleshoot feels so important right now, it certainly does to me. But it’s helpful to remember that many skills that were important in the past aren’t useful anymore. For example, if you owned one of the first cars, you had to be a mechanic. You had to know how to drive a stick shift. Etc. But today’s drivers don’t need any of those skills. They get in, press a button, and they drive. Soon, the car will drive them. And a future generation will chuckle at the notion that people ever drove cars, something that most take for granted as not just their right but their preference. Things change.

And with AI, that rate of change is just incredible. The tech industry was already on its own trajectory, but AI seems to make things almost exponentially faster.

“AMD unveils $3,999 Halo PC”

The compoment pricing issue is out of control

🐧 More Linux

justme asks:

Linux – have you tried Fedora 44 yet? Any impressions?

Yeah. It’s pretty much my favorite distribution at this point and I will be writing about that and some other related experiences soon. I don’t want to give up a bit I started writing elsewhere already, but as we were winding down in Mexico a week or two ago, I started resetting PCs and storing them away as our departure date got closer. And I found myself missing Linux, which was interesting and new. This triggered the beginning of an article that may be called “The Zen of Linux,” but it also fed into what became Switcher 2026: Privacy, Security, and What Really Matters ⭐ and then Switcher 2026: Minimizing the Microsoft in Windows 11 ⭐, the latter of which might correctly be viewed as an attempt–mostly successful, I think–to make Windows 11 work like Linux.

One of the nice things about being back in PA is that I have far more PCs sitting around, including far too many laptops of a 12th or 13th generation Intel Core variety, and those are ideal for Linux. So I’ve installed a couple of different Linux distributions on some of these already and will do more next week and then continue with the series. And though I’m not bringing a Linux-based laptop away with me when we go up to the Finger Lakes for the weekend, I did install Ubuntu 26.04 in WSL on the one I’m using regularly, so there will be some work with that at least.

I will be writing more about all this soon, basically.

“Plex’s 200% Lifetime Pass price hike tries forcing users to another subscription”

What it’s forcing is Plex users to abandon this platform

🔫 Videogames that are not named Call of Duty

justme asks:

Have you made your way through Fallen Order and/or Survivor yet? Any thoughts on either title? A preference for either one?

Not in the past few weeks, though I did play through more of Fallen Order since the last time this came up, and it’s one of the games I test on Linux and Windows 11 on Arm now. The only notable game-related things recently were installing Fortnite on my iPad after it came back to the App Store and then, just last night, I purchased the original The Last of Us Part I from Epic Games because they are holding a Mega Sale right now and it was roughly half-off. That’s another single-player experience I’ve wanted to try for quite a while. Part of the problem, of course, is that it’s been busy. I want to actually pay attention when playing anything single player and it’s just been too distracting.

“Samsung overtakes Apple in 2026 customer satisfaction study”

LOL what

💼 When will then be now?

jrzoomer asks:

Microsoft officially confirmed the end-of-life date for Microsoft Publisher. Given how long Publisher has survived as a legacy Office staple, does this signal a broader future where Microsoft will aggressively axe older desktop apps to clear the runway for new web and AI first software? Correct me if I’m wrong, I believe you mentioned that you no longer use the Microsoft Office apps yourself?

When Microsoft finally embraced mobile and web versions of its classic Office applications, there was an interesting moment when that team was talking openly about new Office apps, often less complicated or specific to whatever use cases, that it could make. Sway and Office Mix came out of that, for example, and you can make a good case for Loop being a far more recent example.

In addition to being “new,” these things all share a commonality in that few seem interested in using them. It’s difficult to know why, and it will vary app by app, but in this case, there is this now ancient Office suite that everyone still seems to want and use, and very little engagement elsewhere. (Teams being maybe an exception, though I’d argue that it fits in neatly with the complexity/vibe of the other traditional Office apps).

Windows experienced the same thing. Windows scaled up to Server nicely and got a lot of traction, but Microsoft’s attempts at new smaller, lighter Windows-like things, from Pocket PC through Windows Mobile and Windows RT, all failed in the end. The only thing customers seemed to want was more of the same. It’s why Windows 11 on Arm is “just” Windows today. And even that hasn’t set the world on fire. (Why would it? I guess.)

There are some interrelated forces driving all this, I think. The first is just general inertia, which we might tie to change aversity. And the second is the customer base that’s driving (or not driving) usage, meaning businesses, especially enterprises, the biggest businesses (and related governmental and other organizations). In some sense, “it just works” is both desirable and understandable. (I have another Switcher 2026 post underway that is basically “Boring is good.”) But there are problems, too. If no one is asking for change and Microsoft, in this case, is not driving change, then there is no change. And things just stay the same, for better or worse.

I reference this a lot because I re-read this a lot, but Steven Sinofsky’s book Hardcore Software covers, among other things, the functional focus shift in Office over time from standalone applications to a bundle to an integrated suite and then to what is just a subscription service. The customer focus shifted over time, too, from enthusiasts to individuals (consumers) to small businesses to businesses and then to enterprises. And at some point, you as an individual will come to the conclusion that this thing, Office, a name Microsoft doesn’t even like to use any more, is the company’s most important product today though Microsoft is looking to its future with AI, and it is increasingly not for you. It’s not to much stuff, too many features no one will ever use, too many complex apps one would have to master for some reason. It’s just too much overall, too much money that you will never stop paying.

I have felt these shifts from the beginning even though I didn’t necessarily see them clearly at first. For example, when Microsoft first released Outlook as a replacement for Schedule+ that also included email, contacts, and calendar, I was intrigued but also bothered by it not supporting Internet email, which was already big by that time. It course corrected with a quickie Outlook 98 release, but Outlook is also this classic Microsoft overthinking and complexity, and it was always too big, heavy, and business-focused for my needs. Microsoft briefly offered things like Internet Mail and News all the way into Windows Live Mail or whatever. But when you’re living on the Internet for email, contacts, and calendar, web-based solutions that were simpler just made sense. Today, I wouldn’t touch anything called Outlook with your computer. No thank you.

But I stuck it out with Word for a long time, and OneNote, too, even though both were too big, too heavy, and too complex. In time, and after many false starts, I finally transitioned to simpler and better apps and technologies for both. And I feel like that’s in a good place. That both revolve around Markdown is important and not coincidental to me thinking about writing a book on that topic. This isn’t a rejection of Microsoft or Big Tech per se. It’s just an embrace of what I prefer, of apps and services that work the way I want, not the way some company requires. Which, when you think about it, is a rejection of Microsoft and Big Tech.

Office will limp into the next century on inertia and change aversity in the enterprise, just like COBOL and other ancient technologies cling to life like the cockroaches that they are. But the vibe coding topic noted above will contribute to the death of many things, among them Apple’s unfair App Store model, that are more recent phenomenons than Office. Things change. And the ability to create or make our own things is, I think, the wave of the future, at least for individuals. Which is all I care about anyway.

I don’t see Microsoft killing the classic Office apps anytime soon, if only because its most important customers demand them. But for all the real human beings out there, Office–and/or a Microsoft 365 subscription that never ends–increasingly makes no sense at all. I never plotted or planned to move away from all that, but I’m happy I finally did.

“The App Store stopped over $2.2 billion in potentially fraudulent transactions in 2025”

Apple also collected far more than that in junk fees in 2025

💻 The perfect laptop

jrzoomer asks:

ASUS managed to fit a 16-inch display, 48GB of RAM, and the top-tier X2 Elite Extreme into a chassis weighing only 2.7 pounds. I could imaging maybe if they did a 13 or 14 inch laptop it would be even lighter. Based on your hands on time, would this be your go to laptop if you were buying today?

I’m using that ASUS Zenbook A16 right now to write this, and it’s a problem because I have other laptops I need to get to. You might think that after reviewing laptops and other hardware devices for 25+ years that I am rarely super-excited by any new hardware these days. And generally speaking, that’s correct. But there are things that occur that truly excite me, including the OG Snapdragon X and that low-end HP OmniBook 5 I still love so much, but also this ASUS. It’s so magically light I had my wife pick up another 16-inch laptop and then the ASUS, and she audibly gasped. It really does feel like there’s nothing inside it.

I bought a Surface Laptop 7 and a MacBook Air M3 in 2024, so about two years ago in each case. Both are still terrific. But as technology marches forward and we’re finally starting to see X2-based laptops, my mind is racing and I’m thinking about upgrading. If I could find the ASUS Zenbook A16 on sale somewhere south of $1500 right now, I would probably just buy one for myself. It’s that incredible. That said, I am, of course, curious about what Lenovo, HP, and Surface will do as the year moves on. For now, I will lurk and wait. And use the review unit, which is … incredible.

Would you wait for Lenovo Thinkpad (I know you were always a fan of Thinkpad, is that still the case)?

Yes and no. ThinkPad is incredible, none more so than the X1 Carbon lineup, but those are Intel-only because of that corporate partnership. So the options are more limited if you don’t want Intel, like me, and very specifically want Snapdragon, as I do. Snapdragon X/X2 are best suited for the thinnest and lightest laptops, but in the ThinkPad space, that’s X1 Carbon, and that is not happening. It’s too bad.

And I also had a more general question, what’s your preferred laptop screen size?

16-inches. I need that much real estate and more, honestly, I’d love to see a 17-inch laptop with no numeric keypad. I suspect such a thing does not exist, and that some future rolling/stretch screen technology will solve this problem for me and others someday. It would be nice, for example, to carry what is essentially a 14-inch laptop around but be able to open it up and have a 16 or 17-inch (or bigger) display when needed. This will happen.

Related to this, LG has a Gram line of very thin and light laptops, too. It’s been a while since I’ve used one, but that had a similarly magic feel to it though it was also a bit flimsy. Looking now, I see there are reasonably up-to-date LG Gram Pro laptops, including a 17-inch model that’s just 3.26 pounds, but they are Intel-based and very expensive.

“Sources and documents detail Satya Nadella’s effort to revamp Microsoft’s senior leadership, creating a startup-style operating model to compete in the AI race”

Maybe he should move Microsoft’s executive team into a literal garage

🏅 It was always going to be Google

spacecamel asks:

Is it wrong that I find what Google is doing much more interesting than what Apple or Microsoft is doing? Google seems to be trying to make consumer AI products while Apple watches and Microsoft focuses on business. Google is a privacy problem but are any of them much better?

I referenced a few articles I’m working on above, and this is tied to yet another one, which I started months ago and is currently called “It was always going to be Google.” There was this time, about two years ago, when Google seemed to be getting everything wrong and kept face-raking itself with mistakes like Bard and an AI image generator that was making hateful pictures. People were declaring Google dead and calling for the ouster of CEO Sundar Pichai. And I just didn’t see it. Google didn’t just invent all the technology that others were going to market with through their chatbots, Google is, well, Google. There is no company with deeper reach when it comes to data and information. I always felt like it would get there.

Over the past year, it’s clear that that’s exactly what’s happening. From Gemini 3 and Nano Banana on, especially, Google has been firing on all cylinders. And this past week’s Google I/O event was nothing less than an overt display of strength and dominance. When I think and write about things like Little Tech and enshittification, one of the things I keep going back to is that this isn’t about severing all ties with Big Tech, it’s about minimizing my exposure to Big Tech. And that doesn’t mean “no” Big Tech, which is impossible anyway, it means probably just settling on a single Big Tech company, tech stack, or whatever. And Google was always the obvious choice there. Despite all the privacy issues.

Apple, Google, and Microsoft are always going to be compared to each other, there’s no way around it. But for all the similarities, there are important differences. Apple dominates in hardware devices, for example, and it has incredible reach with consumers, but it also creates things like Photos and Apple Watch that only work with its devices, which is problematic. And Microsoft is primarily an enterprise company from a software perspective (excluding gaming, of course) and is otherwise a cloud/AI giant and thus patently uninteresting to most human beings.

But Google is different. Google will meet you where you are, no matter who you are. It has strong business and consumer businesses. It’s a cloud/AI giant like Microsoft, but it’s demonstrably better on the AI side. It’s poor in hardware devices, but it has tremendous reach with consumers, billions of whom use things like Gmail, Maps, Search, Photos, and more no matter what devices they use. Android, Windows, Mac, iPhone/iPad, web, Linux, whatever. I keep saying this, but this is what Microsoft could have been but isn’t.

If Google has a problem, and this was evident all over I/O this year, it has too much stuff. Too many products with too many names and too much overlapping AI functionality everywhere. A small example: Why the frick does it have Google Photos and then release Snapseed updates? A bigger example: Gemini is a chatbot, but there’s NotebookLM, Flow, Antigravity, AI Studio, and 117 other things that all use AI and overlap and what the frick is all that stuff? This is very Microsoftian of it.

The trade-off with Google has always been well-understood, though it is not as simple (and simple-minded) as, “if you don’t pay for the product, you are the product.” You know, though you may not fully understand it, that in using Google products and services you are ceding some privacy. You do this because those things are so good. Maps, for example. Gmail. Search, obviously. Increasingly, Gemini, which is suddenly everywhere. It goes on and on. I cannot think of a single Apple app or service that is a good as what Google does. Likewise, anything from Microsoft. It’s not that Google always gets it right. But it’s batting higher than the competition.

So Google will stumble and fail with phones, maybe, or smart home, or whatever. But it’s also this stunning combination of infrastructure (for people and businesses) and front-facing apps and services that billions of people, in both cases, find indispensable. Pixel still takes the best pictures by default with no work. Etc. The only things we can do as individuals are be clear on the contract we’re not signing, take what ever steps we feel we need to do to protect ourselves, and then pick the apps and services we want to use and stop worrying about it. Cheerleading for any company is stupid. But you do this for yourself. And that’s smart.

“Microsoft finally fixes one of the most annoying things about PowerShell on macOS”

That it exists?