
Happy Friday! Later today, we’re flying back to the U.S. to spend the holidays with friends and family. But first, let’s kick off the weekend a bit early with some great reader questions.
helix2301 asks:
Any idea when the Snapdragon features of copilot plus that are exclusive will be available to all?
I don’t see Microsoft adapting all or even most Copilot+ PC-exclusive features to work with non-compliant PCs (meaning, those with no NPUs, or low-end NPUs). But Microsoft’s October declaration that “every PC will be an AI PC” was confusing, and I sort of did a double-take, wondering if something major was changing.
Parsing the original announcement, a “normal” Windows 11 PC and a Copilot+ PC are still different things, with the latter essentially being a new Windows 11 SKU, or product edition, with specific unique features. And though I don’t like that weirdness, it feels temporary, like Media Center or Tablet PC Editions, both of which worked similarly in offering unique features outside the normal Windows product edition differentiations.
So, there are two questions that come out of this, I guess.
The first is what happens to Copilot+ PC? There are at least two possible outcomes. Either every PC becomes a Copilot+ PC, rendering the specification dead because now all mainstream PC processors and PCs have capable NPUs. Or Microsoft literally just gives up on this, at which point your question kicks in and we’ll see if we get those features everywhere via GPUs or CPUs, or just go back to the cloud. (A third horrible possibility is that they keep increasing the hardware requirements to include ever more powerful NPUs, but I don’t see this one happening now.)
The second is what happens to the PC in general? And I think that’s what the October announcement points to. It’s not about bringing Copilot+ PC features to non-Copilot + PCs, it’s about advancing Windows across the board to support natural language conversational capabilities as an obvious way to interact with the system and, in the future, with agents and other AI capabilities. All the features discussed in the main announcements fall under the Copilot umbrella, not Copilot+ PC, which is a terrible, confusing name. It’s all just software and the AI runs in the cloud. So all PCs can handle all those capabilities.
But there will be overlap. Obviously, all the normal Copilot features will work on Copilot+ PCs too. And we’re starting to see individual features in individual apps start to offer choices of local and cloud-based AI processing.
Two examples.
Notepad has AI writing tools now, and they are cloud-based and use AI credits and all that nonsense. But they’re testing using local AI for some of that, so those with Copilot+ PCs can use some of those writing tools locally, with no need for AI credits or a Microsoft 365 subscription.
And I don’t think we wrote about this, but the latest PowerToys update includes a new version of Advanced Paste that now supports a choice of cloud-based and local AI models. So those with a Copilot+ PC can use Advanced Paste with local AI (actually, that’s probably true for everyone, with certain models optimized for GPUs too).
Given this, it’s possible that some Copilot+ PC features will come to other Windows 11 PCs, and we’ve wondered for a year and a half when or if Microsoft will simply allow those with GPU-powered PCs to use these features. But there’s been no real news on that end. It’s the type of thing that makes sense, but no one there has ever said, hey, obviously we’re doing that. But they should.
To be fair to Microsoft, the issues here are many, and it’s not just the cynical argument that using local AI will save Microsoft money. Local AI can work when the PC if offline, too, so all that Click to Do and semantic search stuff (and whatever else) works consistently no matter the state of connectivity. Requiring an Internet connection would make for a terrible user experience in some cases. So this is about things just working, too.
lvthunder asks:
I would like to try all these new agentic features in Windows. Which ring do you recommend I join? Also, do you have the link to the Ignite session where they show this?
My money is on the Dev channel, which will likely start testing new 26H1 features soon. That said, Beta should be OK as well, and that will stay on 25H2 with the same features, if history is any guide.
Regarding the sessions, I rely on Microsoft making most Ignite (and Build) sessions available for download and/or later viewing, and I often watch more of these things on the flight home from events than I do while there. But I noticed something odd about the Ignite sessions this year. More than usual are not being recorded at all, and of those that are/were that I’m interested in, many are not available for download, meaning we can stream them on demand but not save a local copy.
The session that Pavan Davuluri headlined was Innovation Session: Windows & Microsoft 365 Copilot: Secure AI & agent productivity. That I have watched, and will watch again, but you can’t download it for some reason, which is a shame as I am considering writing something specifically about that content.
Assuming you specifically only want to know about the agentic capabilities that Microsoft is adding to Windows (as opposed to Microsoft 365 or elsewhere), there aren’t too many sessions that were recorded, so I guess that narrows the field a lot. But these might be of interest and both are downloadable too:
I was going through the Ignite sessions yesterday, looking for session videos to download for the trip home and later. And it is brutally AI agent-heavy, with those sessions spread between a crazy array of top-level topics that go from security to IT management to developer to Microsoft 365 and lots more. There are over 625 sessions that come up when I search for “agent,” compared to 99 for “windows,” for example.
You didn’t ask this, but I have saved the following sessions to watch (and, where possible, download to watch offline):
Only 4 of those are downloadable, so I/we will need to be online to watch the others. There are other sessions I’d watch, but they weren’t recorded.
JustMe asks:
Recently acquired a new laptop for work – an HP Probook 4 G1a 14. I was somewhat surprised that it arrived with 24H2 installed, and it wasnt until after nearly a week of use that it was offered 25H2. Isnt 25H2 now supposed to be essentially universally available?
In the crazy world that is Windows updating these days, there is no difference between 24H2 and 25H2, and 25H2 is rolling out if not in a completely random way, then close enough. But no, you don’t just get 25H2 automatically or immediately. It just happens … eventually.
JustMe asks:
I was wondering if you could offer some insight into the business case and use of all the extra crapware that vendors seem to install on laptops these days. Surely the OEMs know how much of it will be either uninstalled or just left dormant? What struck me most was a) so much of what HP includes is HP branded and b) there are pieces of it that stubbornly refuse to uninstall.
This has been a long-time pain point. I’m not defending the PC makers, per se. But from their perspective, customers are buying a product from them and so they own that relationship, and they understandably want to expand it, possibly with ongoing subscriptions or other ways to make money in what is essentially a very low margin business. In my dealings with them over the years, it’s clear that they believe they can add value in ways that customers will appreciate and want. And it’s as clear to me that they almost never do that.
I did two Signature PC studies over the years for Microsoft, and the point of that program was to create clean versions of PCs without any of the crap that PC makers installed. So you can see where this was a source of friction for all the parties involved. But one of the stranger aspects of the studies I did is that one in 10 people seemed to really enjoy having the extra crap. It was just more stuff, and to these people, more was better.
Of course, that leaves the other 90 percent. And the reality is that most PC maker software is either terrible, superfluous, or basically a duplicate of something that’s already in Windows. And it’s just a bad experience for users. Microsoft doesn’t help matters by allowing PC makers to customize the recovery images so that their crap is automatically reinstalled every time. And worse, now a lot of PC maker crap is automatically installed even when you use a clean Windows ISO to reinstall. There’s no escaping it.
(Even those Tiny11 Builder-based installs of Windows 11 I’ve done recently will eventually download at least a few PC maker utilities, like the HP, formerly MyHP, app.)
Unfortunately, this is a byproduct of the PC market and how we have Microsoft making the OS and PC makers making the hardware, and each feels that it needs more exposure with each customer. So Microsoft has its weird value-adds and so do the PC makers. Microsoft or the PC maker gets blamed when something goes wrong, but the PC maker is on the hook for support. And he we are. We have a lot of choices, which is good. But we also have companies scrabbling together upsells and value-adds to get us to spend more, which is bad (for us).
The only solution, such as it is, is to spend more money. The most premium PCs, especially in the commercial space, tend to have little or no crap at all. The unfairness of this is obvious, it’s just the way it is.
train_wreck asks:
Paul, I think your rant against the people on Twitter trolling Pavan Davaluri’s post suffered from a bit of selection bias: the people on Twitter. That particular social network is heavily skewed towards impulsive negativity, and should thus be ignored.
To be clear, I don’t believe you meant any of that to be insulting. But I do see three major problems with where this is heading already.
Your further comments on Windows Weekly, however, came across to me as casting a rather wide net on anyone who complains about AI features being added to Windows, namely that they are old and need to get out of the way.
When I write something, I can edit it and I can try to make it as clear as possible. When I speak on a podcast, there are other people, they have other ideas, and the conversation goes where it goes. What I did express on Windows Weekly, eventually, is the reason why I wrote about this in the first place.
This is personal. It’s me, and it’s all of us. These are my people and my community. And I do not like what I see when I look in that mirror.
That “feedback” is the same knee-jerk reaction that I see more and more over time in our space to any change in Windows, or Office, or whatever else. And in the sense that we all have our likes and dislikes, there are too many examples of changes that do not impact the person complaining and/or are things that are either opt-in or can be disabled easily. And that renders those complaints moot. I say this so much: There are so many terrible problems to deal with that we don’t need to make up more. Let’s fight the real battles.
What I saw in the sea of hate that is those Twitter comments was an amplified version of the feedback I get whenever Microsoft does anything that someone doesn’t like. And it’s always the same audience these days. Older guys like me who have apparently forgotten why they’re here in the first place. They have skills that are threatened by advances in technology and they push back. They are not making good decisions for their employers, they’re protecting themselves. Etc.
I’m tired of it. I’m old too. I’ve been around a long time. I’ve seen it all. And I don’t like everything that’s happening to Windows. Obviously: I’m kind of the champion of that. But I’m trying to solve problems, not just bitch and moan. And I have not lost my love of the thing. I am still enthusiastic about technology and about Windows. Another thing I often say: I love my kids, but I can complain when they do badly, in school or life or whatever else, and I can expect them to do better. But I also don’t make up complaints about them. There are real issues to address.
I wrote about this the other day. You can complain. You can solve or not solve problems. But if you’re not solving problems, get out of the freaking way. Otherwise, you’re just an obstacle. A part of the problem. And we don’t need more problems. We need more solutions.
From my discussions with my colleagues in IT/dev and readings of reports from people in various professions, there are an increasing amount of companies who are mandating use of AI at work, with some going so far as to fire workers who can’t demonstrate their use of the (very immature, as i’m sure you’re aware) tooling. I think this is partly what is motivating the “AI-lash” to some degree: it’s all fun and games when AI hallucinates a goofy meme image you asked it to make, but when people’s jobs are on the line it becomes much more of an existential threat. A few very rich people are basically dictating what the rest of us will have to put with for the near future; it’s understandable that there will, and should, be pushback on that, especially from people technical enough to know the limitations of these tools, especially when the people in charge of their paycheck are equally unaware of said limitations. Just some thoughts, thanks for the thought provoking content as always.
The AI stuff is controversial on every level. The irony here, and this should give everyone a moment of peace and happiness, is that actual job loss tied to AI is not happening at our level. It’s happening at the extremes, the C-level or whatever positions, and at the entry level. The guys who believe they can use AI to eliminate jobs are the guys we don’t need. This is like the scene in Office Space where the Bobs ask that guy what it is that he thinks he does for the company. We will always need expertise. But we will always need fewer bozos, too.
Sometimes I’m not even sure what to say about job loss. You can use historical examples like portrait painters, stagecoach drivers, or whatever. Do we not move forward because those were jobs that people had? How about fossil fuels? Do we just keep poisoning the planet … because jobs? We need to adapt. This is inevitable. It happens all the time. AI is difficult on so many levels, but right now, it’s more promise than reality. What’s really unclear is who is exaggerating more: Those promoting AI the hardest or those pushing back the hardest. I don’t know.
The ultimate irony to this conversation is that the session Davuluri was promoting on Twitter, the thing that kicked all this off in the first place, is now available for viewing. I will likely write about this soon, but I’ve watched it twice so far and there is nothing offensive in what they discussed. We get so angry, so quickly, over nothing, and we just doing it. This is not a technology discussion. It’s about people and psychology and it’s just not my area. But it impacts me almost every single day. And it is tiring.
Related to this, ianceicys asks:
Hey Paul long comment but I wanted to hear your thoughts given this week was “messaging overload from Microsoft at Ignite”.I think it would be wise to take a more discriminating approach when amplifying stories from people like David Plummer. His past involvement in the SoftwareOnline consumer-protection case due to his Registry Cleaner and InternetShield software products shows that not everything in his professional path has been above board, also several of the anecdotes he shares have been deeply questioned by others who were present at the time.
It’s not clear I amplified anything, I just pointed out that his ideas are simple, not new, and not helpful, and I didn’t do that in an article called “Dave Plummer is terrible and he’s wrong about everything” or whatever. I like his channel and I do watch his videos. And he explained what happened in the event you cite in an interview and whatever. There’s nothing there that I find troubling. He’s a very techy guy on the spectrum like the rest of us. And he’s made mistakes, like the rest of us. That doesn’t make him a leper or a Nazi, or whatever, and it doesn’t mean he can’t be interesting or have good ideas.
As Richard mentioned on this week’s Windows weekly, there’s a VERY strong palace guard around Microsoft’s narratives—deciding which stories get promoted and which perspectives get minimized. You can see this dynamic today with leaders like Pavan Davuluri, the new CEO of Windows and Devices, and Mustafa Suleyman, recently appointed as CEO of Microsoft AI (MAI) following Microsoft’s high-profile acquisition of Inflection AI. Similarly, Microsoft’s acquisition of Activision Blizzard, with Bobby Kotick, illustrates how top-level narratives are often designed to obscure uncomfortable realities.
I made this point on a previous episode, but everyone answers to someone, even the CEO. In Satya’s case, the board of directors and Microsoft’s shareholders, for example. So in Microsoft’s case, there is this high-level strategy that did or did not come from him, it doesn’t matter. My growing understanding of Microsoft today is that it’s as much a giant money management operation as it is a tech giant, and that means that the CFO, Amy Hood, wields a lot more power than her predecessors and maybe even more than Nadella. It’s the difference between “make the numbers make us look good” in the past and “we have so much money that we can turn huge profits just by moving money around correctly and understand what we can and can’t get away with” today.
Even Judson’s messaging at Ignite reflects Microsoft’s long tradition of tightly managed narratives. This isn’t unique to Microsoft. Apple is a master class…just look at what was shared in Apple in China. Across the tech industry, there are many examples where the public story diverges from a more complicated internal reality: Bill Gates’ highly publicized divorce from Melinda Gates, which brought renewed attention to LONGSTANDING claims about his questionable interactions with female interns at Microsoft and related concerns about why he had to resign as CEO that have circulated for years. Similarly, we’ve all seen how this week’s OpenAI board resignation and Microsoft’s official “everything is fine” messaging only partially reflects the deeper friction between these BILLION\TRILLION dollar organizations.
Right. Power corrupts, etc. This is an age-old story in some ways. But we should expect the leaders of terrible companies to be terrible, just as we should expect the world’s biggest companies to be terrible. These things all go hand in hand. But Ignite is a milestone on Microsoft’s calendar and it’s when they discuss things they’re doing. Apple and Google events are much more consumer-friendly PR moments than Ignite is.
Honestly moments like Ignite should remind us that major tech companies, and past and present leaders often operate within layers of messaging, and reputation management that is DESIGNED by highly paid marketing teams to obfuscate.
Of course, but with the caveat that a show like Ignite is where things actually have to work, too. This isn’t a PR moment in isolation. There will be demos, there will be hands-on labs, and there will be downloadable/installable software that real people and real companies will evaluate and it needs to work. So we’ll see what happens. But it’s not just marketing, right? At some point, it has to be real.
None of this implies wrongdoing—only that the polished narratives we’re presented with rarely capture the whole picture and REQUIRE greater skepticism. A deeper skepticism toward corporate storytelling and the personal myth-making of high-profile tech leaders is healthy and necessary.
You can’t possibly believe that I somehow headed into Ignite after 35+ years covering this company to be snowballed by marketing. I’m not a wide-eyed cheerleader. I view everything cynically. But I also evaluate things rationally and try to be objective.
So what was mentioned at Ignite that you are applying a healthy skepticism to?
Everything?
Did I write some drivel about how awesome Microsoft AI is or something? I see it for what it is. I feel like that’s how I communicate it.
What I wrote ahead of Ignite was that people should be less terrible to other people. That people in our community need to remember why they came here in the first place and stop believing that all progress has to end now that they are secure in their jobs. And that there are problems real and imagined, and I choose to focus on the former whereas all the complaining I see is mostly about the latter. We should all be on the same page here.
spacecamel asks:
I wanted to get your thoughts on what Valve is building overall. They have put a lot of effort into making their games compatible with their SteamOS and I commend them for sticking with to build the ecosystem that they have. However they are still missing the biggest part of the market with not supporting first person shooters and not doing deals with others such as Epic to get compatibility. So what is the market for the “Gabe” cube?
I wonder as well. Obviously, there is some small market of enthusiasts that will buy anything Valve/Steam makes. And there is some small market for handheld gaming PCs, too, and it’s weird to me that Steam has never updated the Steam Deck as more competition appeared. Why wasn’t there a Steam Deck revision too?
Related to that, I don’t see the Xbox Ally devices as “saving” Xbox, for example, it’s just too small a market. But as a new direction for Xbox hardware, it’s a nice peek at a future that can make sense, and a familiar future because it’s just a PC. As is the case with what Steam is doing.
In the Steam/Valve world, making Steam OS and doing the work to get Windows games to just run on this Linux-based thing makes strategic sense because it cuts Microsoft and Windows out of the loop, and whatever gaming-related initiatives Microsoft may have that undermine Steam. If you think about Steam on Windows, I guess it’s the biggest ecosystem, but it’s also an open platform, essentially, and so gamers can get games from anywhere else too. That’s great for gamers, and it’s one of many reasons why pushing Xbox to the PC makes sense. But it’s not what Valve/Steam wants. Like any company, it wants the biggest possible audience. And so these platforms help it exert control. It’s interesting that some embrace that when Valve does it, as they’re rejecting it when Microsoft does it.
The anti-cheat thing is a strange outlier for Steam, I’m curious they’ve not solved that yet. But for now, this diminishes their platform. I can play all those games on Windows. I can’t on Steam OS.
Also if the “Gabe” cube is priced like a PC (maybe 800USD) is this going to give room to Microsoft and Sony to price their next consoles well above 1000USD?
Yeah, I’m also curious about that. I feel like somewhere in the $600 to $800 range is the sweet spot but I also don’t see this costing much less than $1000. So we’ll see what happens. It’s funny how much positive press they got with no release date in sight and no word at all on pricing. These should both be concerning to everyone.
gg1 asks:
Amongst all the AI hubbub, one Microsoft product has been left behind: Loop. Even though it seems like low-hanging fruit, I have not seen any meaningful AI treatment of this red-headed stepchild of Office. Is Loop dead?
Maybe my memory of the timing is off, but I feel like Loop happened at the wrong time and it just hasn’t gotten the attention it needs because the focus shifted suddenly to AI. Microsoft spent years overengineering the backend of what became Loop and by the time it finally shipped an app, we were mostly struck by it just being a Notion clone. But it’s never worked reliably, and whatever updates it gets are all under the radar now because all anyone can talk about is AI.
To be fair, this makes sense on some level: AI could turn into an economic engine, whereas Loop is one app in a giant ecosystem and not something Microsoft could ever charge extra for. To me, Loop could and maybe should have turned into this thing that replaced Word, PowerPoint, and OneNote. And maybe that was the plan. But today, it just feels like the smallest freaking province of the Microsoft 365 empire, a forgotten also-ran that will never live up to its initial promises.
So yeah, I think Loop is dead. But it’s dead app/service walking for now. It will limp along for years. Some people will love it. Most will not even know it’s there. And on some future day, it will be quietly deprecated and head off into the sunset with Publisher and whatever else. It’s too bad. There is a core idea there that is solid and could work. (And does work, in Notion and perhaps elsewhere, actually.)
jrzoomer asks:
Paul, the FT reported this week that Apple is looking to replace Tim Cook as CEO. What do you think of Tim Cook’s leadership at Apple, and is replacing him the right move for the company?
Tim Cook was often compared to Steve Ballmer because he was so unlike the visionary leader that each replaced, and because both are more accountants than technologists. But these days, I would compare Cook more to Amy Hood, Microsoft’s CFO, because each is the architect of a financial empire that grew far beyond what any company just selling technology could do. Imagine how much smaller Apple would be if it had the same hardware margins as, say, HP. It wouldn’t even be considered part of Big Tech.
But that’s not what happened. Tim Cook’s Apple is one of the most successful companies in history, obviously, and this is all the more impressive because of the hardware margin aspect. As a fan of personal technology, Apple is also interesting because it makes premium products that are well-designed, and that has mostly continued under Cook. In the end, he may be defined by China and what Apple did there to continue its insane growth. And by the company’s insane anticompetitive and anti-consumer behaviors, if only because of how contrary they are to the image Cook likes to project. But that history will be written later.
That Cook is no Steve Jobs-style visionary is obvious. Also that he’s no Bill Gates-style technologist or big thinker. Few people are. But where Cook was directed to not “do what Steve Jobs would do,” I feel like his successor will do the opposite. Whoever this is will be under a lot of pressure to continue the financial successes, which is borderline impossible. But they will also need to be more visionary or product-focused, or whatever, than Cook. He has a certain distance and aloofness, where Jobs was hands-on. They need a little of that again.
But think about just the technology for a moment. Apple Silicon happened on his watch. I mean, that is incredible. I don’t think he was responsible or whatever. But he didn’t stop it or get in the way. And the whole ecosystem is better as a result.
I also don’t think Cook is stepping down imminently. Some have suggested that the stories we saw were planted to gauge the market’s reaction. That makes sense: Cook leaving could hammer Apple’s stock price, but if the right person succeeds him, Apple could come out in the black instead. He has to step down someday, of course.
jrzoomer asks:
Bloomberg’s Mark Gurman reported a few days ago that Apple has “largely written off the Mac Pro.” I guess the Mac Studio is the future of pros. What do you think of this hardware move by Apple? And also, is this largely the future of the desktop form factor in general?
PCs and Macs have been disproportionately portable for many years, and the biggest gains in Apple Silicon are in the portable space, in that performance per watt (vs. raw performance) sense. I can’t imagine that it’s impossible to add dedicated graphics to Apple Silicon or any other Arm-based SoC. But it’s possible that the gains they see elsewhere may be diminished by that shift, and that there is some engineering issue we don’t understand. But an M-series processor with a dedicated GPU is the obvious first step for future high-end Mac-based desktops. Or, barring that, M-series processors with equivalent graphics processing capabilities. Which is likely also just an engineering issue, something Apple might solve, but just not in the near term.
I get that there is always some need for desktop PCs of whatever kind, from towers down to NUCs/SFFs/Mac Minis. But that’s a small market. Most people like being able to compute anywhere. We do this with phones. We do this with laptops. We can’t do this with desktop PCs. It’s stuck in one room in one building, and if you’re not in that exact place, you can’t use it. So the market has spoken, in a sense.
Is the Mac Studio is the future of the Mac desktop in the sort of workstation space? Yeah, could be. If desktops are a subset of the overall PC market, and they are, then PCs with slots for expansion cards are a subset of that. So a Mac Mini or iMac, I guess, might make sense for most people who want a desktop Mac. And then the Studio addresses most needs in the workstation/professional market. Maybe.
But I do feel like they need to solve the GPU problem. It does seem solvable.
With technology shaping our everyday lives, how could we not dig deeper?
Thurrott Premium delivers an honest and thorough perspective about the technologies we use and rely on everyday. Discover deeper content as a Premium member.