Ask Paul: October 27 (Premium)

Happy Friday! It’s been another long week, so let’s kick off the weekend with another great set of reader questions.

Contingency

simionda asks:

With you running the site on your own now, do you have any kind of system in place should something happen to you?  I’ve seen a lot of writers who go dark for an extended period of time, giving no updates for months (Scott Hanselman and Eric Lippert are two examples).  Since your site is your livelihood, I doubt you’d just go dark – but what if something happened while you’re traveling?

If I go dark, then, yes, something bad happened to me. I live to write (and write to live) in many ways, and not just to write, but to publish. Low output days are frustrating to me, and those days when I can’t publish at all are the worst. I experience many days when I write a lot but it’s for the book, perhaps, or an article that isn’t quite there yet, or whatever, and no one has any idea that this work occurred except for me. No bueno.

So I’m not going to discover the true meaning of life and disappear. This is my true meaning of life. But yes, something could happen at any time, and that’s true whether I’m home or traveling. And I don’t view my travel as any riskier than me walking across the street in Pennsylvania. In that it’s probably safer: Pennsylvania drivers are the worst.

But as far as having a plan, no, not really. Now that my wife is partnering with me on the business side of the site, she would know how to reach out to Laurent and would do so. And she knows my key friends in the industry like Mary Jo, Richard, and Rafael, and would reach out to them as well. So the world would find out.

If the worry is about the site, I’m not sure what to say. 🙂 That would be a secondary concern for me, at least, and I’m sure for my wife. But I will discuss this with her now.

One Copilot to rule them all

helix2301 asks:

I see Windows Copilot showed up on my Windows 11 machine. What is the difference between Copilot for windows, GitHub vs office? Do they all do something I am kind of confused about branding. I would think it would be one AI to rule them all. I don’t understand the difference. Cortana on skype did very similar to windows cortana. Is this just branding thing?

Yes and no.

I agree that the sudden explosion in copilots at Microsoft, its inconsistent way of naming them, and its sudden September about-face on some of the copilot branding is confusing. But I will also argue that this is good branding and that Microsoft’s about-face was for the best and a sign of good decision-making: Past regimes might have just run with the original branding, creating longer-term confusion for customers.

So. While these things are not all the same, technically or otherwise, some actually are (which contributes to the confusion). But the more important point is that Microsoft has chosen to refer to its side-by-side AI experiences as copilots, and while some of these are implicit—for example, the Microsoft Copilot powers Bing Chat but it’s never marketed that way—it’s the right branding choice. Copilot, unlike an invented term like “Cortana” or a failed brand like “Bing,” is a good brand. It’s both obvious and clever. Everyone immediately understands that a copilot is something that assists you. And that you (the “pilot”) are ultimately in charge. Which is the right way to view the human/AI relationship.

GitHub Copilot was the first of the copilot and at the time it was announced, there was no reason to believe that Microsoft would one day use this for a vast swath of product names. It was launched in technical preview in 2021 and became generally available in mid-2022. (We didn’t learn about Bing Chat until February 2023 and that Microsoft would use copilot naming more broadly until March 2023, when it announced Microsoft 365 Copilot.)

The only thing that all copilots share is that they integrate with existing apps and tools using what Microsoft calls a “beside” (or side-by-side) experience. This is often, but not always, a sidebar of some kind as we see in Windows and in Microsoft Edge. The GitHub Copilot doesn’t appear as a sidebar, however. Instead, it pops up in your code editor (usually Visual Studio Code) to make coding suggestions based on any number of factors (corporate coding style, perhaps, context, whatever).

So regardless of the specific UI, the use case is similar: You’re working on something and without leaving the app, you can find the answer to a question, get help, optimize your work (more efficient code in Copilot, perhaps, better writing in other places, etc.). Key to this experience is that you’re using a familiar tool—Visual Studio Code, Microsoft Word (soon), Windows, etc.—and don’t need to use something else. For example, developers typically spend a lot of time Googling answers to coding questions and have to sift through a lot of crap on sites like Stack Overflow before they find the answer or at least a pointer in the right direction. (I’m not a professional coder, but I’ve wasted many hours doing this myself. With GitHub Copilot, you get the answer in-line where you are working immediately. And on you go.

For most individuals, the important point to understand is that something called Microsoft Copilot is the foundation for the key Microsoft products you use. This common foundation sits under Bing Chat, the Microsoft 365 Copilot, and Copilot in Windows, and so there is a common set of capabilities you see across each. This is most obvious if you’ve used Bing Chat on the web, the Copilot sidebar in Microsoft Edge, or Copilot in Windows, because they’re all almost identical. Copilot in Windows offers only a tiny slice of extra functionality that is specific to Windows right now (“turn on dark mode”) and not much else. But the difference will be much more obvious in Microsoft 365 Copilot, as there will be hundreds (if not more) new, app-specific features across Word, Excel, and so on, plus what I’ll call organizational-wide capabilities powered by the Microsoft Graph (in the business version that launches next week).

There are/will be many other copilots across the Microsoft ecosystem, most of which do not interest me (or you, I bet). But at a high level, the only thing to know is that this branding does signify something—that this thing will help users of that product in some way, using AI—while not necessarily sharing any technical underpinnings with other copilots.

On a side note, there are other ways for Microsoft and developers to add AI capabilities to apps, and when this happens, those features won’t be called copilots. We see this now in Paint, for example, which has a new Cocreator feature that is based on the same OpenAI DALLE-3 backend that powers Bing Image Creator. (In fact, in use it’s nearly identical). This feature wasn’t called Image Copilot or whatever, just as Bing Image Creator doesn’t use the name copilot, and for good reason: It’s an integrated feature in the app, and not something outside that sits beside Paint.

Of course, Microsoft is still Microsoft, and it will inconsistently name things here and there. It can’t help itself. But for the most part, this is going pretty well right now. I don’t get to applaud Microsoft for branding almost ever. But this time, they deserve it.

Alternate web browsers

j5 asks:

Hey Paul! Are you considering checking out any new web browsers? I’m typing this on Orion. There are a lot of good ones out there right now, Orion, Arc, Vivaldi, and Mullvad.

I always check out other web browsers. I was going to link to my article about my recent (second) look at the DuckDuckGo web browser only to find out I had never posted it; but I talked about it on last week’s Windows Weekly podcast: It’s one to keep an eye on but it’s still lacking some key features like settings sync, extension support, and a lot more. Someday.

(I also never wrote up the latest Vivaldi release, which adds secure settings sync, pop-up video playback, various improvements to its Mail and Calendar experiences, and more. I’m not sure how obvious this is from the outside, but it’s been incredibly busy lately, and this one just fell by the wayside, sorry.)

Generally speaking, I think it’s important to routinely test alternatives to the tools and services we use, and I do this as a matter of course both personally and professionally. Every once in a while, you’ll land on something that is better/more efficient than what you had been using, and those moments are magical.

With web browsers specifically, I do a few related things. I always check out the major browsers—Chrome, Edge, Firefox—when they’re updated just to see what’s changed and whether any of it seems meaningful. (I do the same less frequently for second-tier browsers like Opera, Vivaldi, and now DuckDuckGo, but same idea.) And I switch to Edge temporarily at least twice a year, not just because I hate myself but so I can update the book. Related to this, I use Brave full-time everywhere (though I am experimenting with Safari on the iPhone this month) so I have a high bar for switching. Brave is pretty much perfect for me.

Also related to this, a reader (SherlockHolmes, and I hope he doesn’t mind me noting that) pointed me to a new cross-platform web browser called Thorium that could be interesting in time: It is promising to do what Edge promised and be Google Chrome but without the Google. So this is inherently interesting. (Also if you read this, Christian, I can’t write you back when I reply to your emails, they keep getting rejected. Sorry.)

I’ve web browser ADHD and really like trying out all the ones that are out there. I love macOS but there are plenty of missing features that frustrate but there are plenty that keep it as my main browser. Anyway, I really enjoyed your web browser review. And it’s always interesting to get your take on subjects/software like this. Thanks! https://browser.kagi.com

I used to experiment heavily with alternative browsers in the early days of Mac OS X, before there was a Safari browser. I forget the names of these things now, Camino was one, but there were some neat choices for the day.  If I was using a Mac, I guess I’d still use Brave, but I would look at Safari for all the obvious reasons. There is an interesting minimalist aesthetic there that I do like. And I still think Safari for Windows could be successful too.

Wait-and-see

wright_is asks:

With Qualcomm shooting itself in the foot with the PMIC fiasco surrounding the new X Elite chip, have they killed the new platform, even before they have released it? Will this mean that we will have to wait until 2025 until we see any realistic ARM devices for WoA?

I discussed this generally about Windows and the Microsoft ecosystem, and more specifically with Windows on Arm on Windows Weekly recently, but one of the more frustrating aspects of our little corner of the world is the endless wait-and-see. Windows on Arm is maybe the worst (or, best, I guess) example of this right now. But it’s not just Windows on Arm. Loop is another one: We’ve been talking about this thing for years and it never seems to happen.

With Windows on Arm (WOA), there have been two parallel paths of frustration, the software side for which Microsoft is responsible, and the hardware side, which we now understand to not just be Qualcomm’s responsibility, but Qualcomm’s fault: It is beyond embarrassing that you can run Windows on Arm with better performance and responsiveness on any Apple Silicon-based Mac than you can on native hardware on the PC side.

Making matters worse for Qualcomm, Microsoft has done everything it can to turn this one-time experiment into what is now an almost completely equal sibling to its x64 counterparts. This is unparalleled in the history of Windows/NT: While Microsoft did port NT and then Windows to various alternative platforms like MIPs, PowerPC, Alpha, and Itanium back in the day, none of them were as fully featured compared to their x86/x64 siblings at that time as WOA is today. That is a fascinating reality that I don’t believe anyone has really discussed so far.

It’s not perfect, yes. Hardware drivers and the accompanying utilities for x86/x64 drivers will never work on Arm-based PCs. And some software isn’t smart about Arm and may cause you to download or install a less efficient version. But these are fading with time: Microsoft, for example, is taking over printer drivers from third parties, and while that wasn’t done for WOA specifically, it will help, and, really, the class drivers in Windows today are good enough for most people anyway (and work normally on WOA). Software will improve and while I am speculating there, surely a hardware compatibility “shim”-type system like we used in the old days could help ensure that users get the most efficient x64 versions of software that has to be emulated.

No matter. The weak link here is Qualcomm. And that is very clear now. When I wrote Windows on Arm’s Last Stand? (Premium), I struggled with the article name and, honestly, failed. I should have called it “Qualcomm’s Last Stand in PCs” or whatever. Because I think WOA is going to be fine. Conversely, I think Qualcomm is screwed and, truth be told, maybe deserves it. With competition from AMD, NVIDIA, and probably Samsung—all huge Microsoft partners, by the way—this market could get very interesting fast. Which leads me back to my central premise here, which is wait-and-see. And how much I hate that.

The waiting game on Qualcomm in the PC space has been ongoing since the initial announcement, which is now almost exactly 7 years ago. In the beginning, we were waiting for better software and hardware. But now it’s just better hardware. And while I think we were all happy to finally hear something, anything, about this NUVIA-based generation of chipsets, we were all likewise unhappy to find out that reference designs were not already in PC makers’ hands, and that new devices would be announced at CES in January and released soon thereafter. Mid-2024 is … tough. It’s more of the same. It’s wait-and-see.

So is 2025: That’s the earliest that we will see AMD- or NVIDIA-based WOA PCs, though I assume we’ll hear about those before then. Still. The thought of another year on this is agonizing.

I should also address the PMIC issue.

If you’re not familiar with this, Qualcomm has mandated that PC makers that use its Snapdragon Elite X SoC will also need to use its PMIC (power management integrated circuits) chipset which, wait for it, are smartphone parts and not specific to PCs. Worse, this is also seen as a power play by Qualcomm, which makes the vast majority of its earnings from chipset sales (85 percent in the most recently reported quarter vs.15 percent for patent licensing), a kind of revenue-per-device piling on.

Granted, this is all based on a single report by SemiAccurate (great name). But this will increase costs for PC makers, which previously could and did choose their own PMICs. And there’s no reason to sugarcoat this: Qualcomm has a problem with the way the rest of the industry perceives them. And this ain’t gonna help.

Put simply, I feel like Microsoft and the PC makers cannot wait to drop Qualcomm like a bad habit. And there are things they know that we don’t, not just with regards to how well the Elite X actually performs but also what AMD and NVIDIA (and perhaps Samsung) are up to. That knowledge could influence their respective go-forward strategies too.

So we’ll see. As always. But until then, we wait.

Weird new Windows 10 behavior

JustMe asks:

Windows 10 question for you – on my remaining Windows 10 machine, recently did an update.  On reboot, I had a search bar on my taskbar (I turn it off on my machines).  The start menu then popped up (unbidden) with a message that a search box had been added to my task bar, would I like to keep it?  For me the answer was no, and upon selecting no, the taskbar went back to my normal configuration.

I literally just saw this as well. I don’t normally run Windows 10, but I installed it in a VM for the new versions of the upgrade chapters in the Windows 11 Field Guide (Upgrade from Windows 10 to Windows 11 Version 23H2 and Upgrade to Windows 11 Version 23H2 on Unsupported Hardware) and was curious about it (and believe it may be tied to the then-most recent monthly cumulative update). I knee-jerk screenshot everything I see like that, but I can’t find the shot now so I’m not sure if I did or not.

Do you know what Microsoft may have added to W10 recently which might have caused that behaviour?  Is this more of Microsoft trying to cajole users into using Microsoft services like Bing?

Yeah, that’s my guess. Microsoft promised no new features, but it never said it would stop advertising (sorry, making “suggestions”) and this seems like an escalation of that. I suspect users on PCs that are compatible with Windows 11 will likewise see an escalation in that little suggestion too.

Chris Cap

ianceicys asks:

Thoughts on Chris Capossela leaving Microsoft after 31 years? With Panos leaving recently too, thoughts?

If you haven’t read it, do check out my news article about this. I’ve known Chris for many years and have interacted with him personally on various things many times, and there’s a confidence there I can’t break. But as I noted, he really is one of the good guys and there isn’t a malicious bone in his body. He is one of the most honest and helpful people I’ve ever met, and my growing suspicion is that he wasn’t on board with the marketing needs of a Microsoft that is now spending 10s of billions of dollars a quarter on AI, and doesn’t care what behavior’s required for it to recoup those costs.

I hate to say this because I can’t stand Panos Panay, but I suspect Chris left for similar reasons: As part of a reorg, he was offered a position that was either a demotion or would be seen that way, pushed back, and either left of his own volition or told he’d be doing so. And that sucks. We need more people like Chris, not fewer.

Here are two related items of interest that I didn’t mention in that news post. Chris came under some undeserved criticism when an internal memo leaked; he had the temerity to recommend that employees interested in a return to the gravy years of bonuses and massive pay raises maybe show up and do something that actually benefits Microsoft, which is entirely reasonable. And I just posted a video from 2015 of him discussing how he transformed Microsoft’s marketing to be more modern and effective to our YouTube channel. It’s worth watching.

The AI is coming from inside the house

jrzoomer asks:

Hi Paul curious for an update as to how you use Amazon Alexa and Google Home products in your house? Do you find them useful and which one do you prefer?

When we moved to Pennsylvania in 2017, I decided to see what if any smart home technologies I’d want to implement in the new house and wrote a series of articles about those efforts. But what I found—and, granted, this was several years ago—was that most smart home products and services didn’t stick. We had some classic issues with complexity and cost, and since this needed to make sense for my wife as well, we scaled back until the only smart home devices in our home were the Sonos speakers, two smart displays (one in the kitchen and one in my wife’s office), and various Philips Hue smart lights, both inside and out. Not much, really.

As for the underlying smart home platform, I “standardized” on Google Assistant at the time, but I later scaled that back dramatically and dumped a bunch of Google Home/Mini speakers we had been using around the house for a while. The two smart displays we use are on Google, so if we need to ask a question for some reason—which we almost never do—that will do the trick. But we use them mostly for family photo slideshows.

So I guess I’m technically on Google, though it’s a vague thing. And I have often observed that Amazon is a lot more serious about the smart home and that its annual Alexa devices event each September (here’s the most recent rendition) is a tsunami of new products and features, whereas Google kind of stumbles along and announces fewer smart home features less often. You’d think this would be critical to them, but they don’t act like it.

That said…

And what do you see as the future of digital assistants/ambient computing, is it winner take all, and is Google or Amazon king?

There’s no reason that this market can’t have three (or more major players) and I’d put Amazon, Apple, and Google on the A-list. And this is especially true with standards like Matter that eliminate the need for third-party smart device makers to explicitly test and support multiple platforms. (It even has the added benefit of putting first-party devices on competing systems, which is excellent.) So it’s all one big happy family now, sort of.

But Amazon and Google are unique because they have cross-platform digital assistants, so choosing one or the other (or Siri if you’re in the Apple ecosystem) certainly makes some sense. And I do see the appeal of first-party devices tied to those assistants. The issue is, neither seems quite there to me. Google revs the hardware too infrequently—it hasn’t updated its speakers in forever, for example—and Amazon’s hardware is always a bit low-rent for tastes. I know that sounds terrible, but I don’t mean it in an elitist way, I just wish they had some premium options. Keeping to the speaker space, the Echo Studio comes the closest, but … not quite.

One thing that both companies should succeed in is bringing AI to their assistants. This will be a nice multi-year series of pushes, I guess, so maybe Google will finally show up and give Amazon a run for its money. But if history is any guide, Amazon will “win” this one, but Google and Apple (to a lesser degree) will remain strong players, I bet.

Surface strategy going forward

will asks:

This is a sort of “What If” question, as if a different Paul in a different dimension had this ability, and that is if you were in charge of the Surface line and branding for Microsoft, and maybe it is a bigger hardware platform question, but what do you think they should be doing going forward?   I mean, lets wave our imaginary wand and for a small 1million a year salary you could and would direct and implement the hardware Microsoft releases, what would you want to have happen?  Ignoring all of the R&D issues and what you choose to have happen would be released within 90 days at scale. I am curious how you would change this brand and what you would focus on and what would drop?

As the anti-Panay, I accept this challenge.

As I’m sure everyone knows, Microsoft recently scaled back its Surface product family, which I think was smart: There was too much experimentation with ridiculous form factors and entry-level devices, and too many tertiary products like headphones and earbuds that are not central to the brand or to Microsoft’s expertise. I also disagreed strongly—still do—with Microsoft’s decision to retire its in-house hardware efforts to keep those products under the Surface umbrella: Microsoft is a much better and more trusted brand than Surface, and this naming can only cause confusion with customers who will believe that the devices will work only with Surface PCs. Stupid.

Surface is such a head-scratcher. It’s not a profitable business, they don’t sell enough product to justify the investments, and they’ve always gone off in these bizarre directions that are immediately obviously going to fail (a key example being Surface Duo). Under Panos Panay, there was too much pride in the superfluous—hinge designs, their non-existent ability to drive new form factors to popularity, etc.—and too little focus on the real-world needs of customers.

I have no experience that can help with the business end of this: I cannot suggest a strategy and know it will succeed or whatever. But I can look at what they have and like Steve Jobs returning to Apple in the late 1990s, I would cut heavily and across the board. There are far too many Surface products serving far too many types of customers. This business is too small to sustain that.

It’s nearly 2024 and the PC is all about productivity, and the nascent AI wave will only add to that strength and market focus. The PC is also all about hybrid work and portability, so there is no need for desktop PCs, certainly not in Surface. I think there’s a debate to be had about the need for tablets and other convertible PCs, but Surface Pro has staked a claim, so I will ignore my inner ape brain and consider what’s right for the brand.

So.

I would keep Surface Pro, with both x64 and Arm variants until and unless Arm shows enough promise to eliminate the need for x64. As the most device-like product in the family, the Pro would retain the 3:2 display aspect ratio, but it would be the only one to do so.

I would also keep the Surface Laptop, in two screen sizes—14.1 and 16 inches, but with a 16:10 aspect ratio—because the world literally needs a MacBook Air alternative in the PC space, and this is it. I’d kill Alcantara, which is a gimmick, but keep the color choices. I would add a few ports. (And if Arm takes off over time, this would be the next line to add that as a choice and potentially transition fully.)

I would evolve Surface Laptop Studio into a traditional laptop form factor and call it Surface Laptop Pro. This would be mostly the same as the Surface Laptop—with the same displays and aspect ratios—but would come with a choice of GPUs and even more ports, and it would be heavier/thicker as a result. There could be gaming/workstation-class variants, etc.

And that is literally it for the current PCs. These would all be premium PCs like the Mac. No cheap plastic versions.

I would kill Surface Connect and go all USB/Thunderbolt for power. I’d pair back a lot of the Surface peripherals, but hybrid work demands at least one docking station, and you could make a case for a docking station with an external GPU that would turn any Surface into a gaming rig. A few ergonomic and non-ergonomic keyboards and mice, ideally in the same color choices as the PCs. All the usual dongles.

A few other things.

There’s no need for an external display, let alone a 3:2 display: That form factor is out of touch in the non-tablet space, and third parties are already killing it here.

I could see a folding PC like the HP Spectre Foldable making its way into the Surface lineup as component prices come down. This would provide a 3-in-1 experience with a 12.5-inch laptop form factor, a desktop-like experience when fully open, and even a bigger tablet experience for those few who do like to use the pen. This is way out there, but over time I could see this replacing the Surface Pro.

I hope I didn’t forget anything, but that’s pretty much where I’m at.

Windows Backstep

SherlockHolmes asks:

Hi Paul, I have been playing around with the new backup tool from Windows. Unfortunately it is useless for me. I never save data on the same partition where I install Windows. Do you have any idea why Microsoft always provides only the default folders for automatic backup? Why can’t I add my own folders? Thank you.

Some background first …

I just completed the Windows Backup chapter for the Windows 11 Field Guide, though I haven’t posted it to the site yet. I will over the weekend. To get to this point, I had to experiment a lot with configuring the “backup” (really, a set of disparate synced folders and settings) and seeing how that impacted each backup, and what it looks like to restore from such a backup, something that can only occur today if you install Windows on a new or reset PC.

And it’s a weird thing, overall. There is only one backup per PC. There is no way to manage a backup per se, but you can remove a device from your Microsoft account or clear some (but not all) synced settings from your MSA from the MSA website. You cannot restore from a backup at any time on any PC, which shouldn’t be hard: All it really does is change some settings. And the difference between restoring from a backup and setting up as if it were a new PC is currently so minor that it’s almost hilarious: Setting up as a new PC won’t restore your installed apps (from the Store only) or their pin configurations. That’s it.

I think I wrote this somewhere before, but I think this will evolve over time. And most probably don’t know that you can configure much of a PC’s “backup” in Settings, in Accounts > Windows backup. Which was there long before the app even existed.

To your questions.

I never save data on the same partition where I install Windows.

I haven’t done this in a while, but if you move the location of OneDrive to another partition, I assume the Folder Backup stuff would still work.

Do you have any idea why Microsoft always provides only the default folders for automatic backup? Why can’t I add my own folders?

This is a great question, and it reminds me of the file system libraries that Microsoft added in Windows Vista. They’re still there in Windows 11, but hidden, and I’m not going to bother to see how/if this interacts with Folder Backup because I could see this feature disappearing at some point. But you could/can make your own libraries. Why can’t you make your own top-level folders in OneDrive and have them synced with Folder Backup too? (I don’t know.)

Some cloud storage services also let you arbitrarily sync folders anywhere in the file system. You should be able to right-click any folder, anywhere, and choose an “Add to Folder Backup” option, just as you should be able to configure where any folder in OneDrive appears in a PC’s file system.

These types of updates would solve the Folder Backup issues I raised in How To Fix OneDrive’s Bad Behavior in Windows 11 (Premium), as one could keep the existing system but add their own folders to Folder Backup and never have to worry about Windows or third-party apps messing around in them. But the solution I did choose, moving all of my data out of those default folders and into a folder of my own making, works pretty well too. I don’t see those folders in Folder Backup, of course. But they sync to all my devices.

People who know what they’re doing are best off just ignoring Windows Backup for now. (The one exception, maybe, is the auto-install of Store apps, but that’s infrequently needed for most.) All we can do is hope it gets better eventually.