Ask Paul: December 8

Happy Friday! We had our first snow flurries the other day, but these reader questions are a wonderful distraction from this cold and dark season.

Speech recognition in Windows

RussDW asks:

With the Microsoft acquisition of Nuance completed over 18 months ago and MS this week deprecating the Vista-era Windows Speech Recognition Technology, one wonders: what’s in store for integrating Dragon speech recognition into Windows? It’s not clear that the current Voice Access capabilities in Windows 11 are even based on Nuance technology.

I assume that Voice Access is specifically based on Dragon speech recognition. I hadn’t seen that the legacy Windows speech recognition technology had been disabled, but I guess that just happened, and given how good Voice Access is, it makes sense. It’s unlikely that Microsoft acquired Nuance but independently created Voice Access. But I’m surprised they’ve not promoted this in some way.

Regardless of how Voice Access came about, I think it’s fair to say that speech recognition will play a big role in the shift to more natural user experiences as discussed below. In addition to the accessibility use cases, using one’s voice to interact with Copilot or other AI experiences is a lot more natural than typing prompts into a text box.

I have a notion

jrzoomer asks:

Paul I heard you say that you’re still using Notion, is it a more momentum based decision, ie the product is actually good enough for your needs vs hassle of migrating to Loop or any other reason?

I write and talk about workflow a lot, and I make a point of experimenting with different devices, apps, services, and other tech products with the aim of determining whether there are better alternatives than what I am using or doing. Most often, I identify some shortcomings that make switching impossible, but that’s fine: If all I’ve done is validate what I do or how I do it, it’s worth it.

Notion is one of those rare and delightful examples of something that was demonstrably better than what I was using before (OneNote, in this case), and like other game-changers I’ve switched to over the past year or two—like Brave for web browsing and Clipchamp for video editing—it’s something I can recommend without reservation. My use of Notion has only expanded, and dramatically, since I started using it.

And that’s the real issue for what I assumed would be a transition to Loop this year for all the obvious reasons. Not so much inertia—or switching “costs,” or whatever—but rather that the bar is so high. Loop can’t just sort of work. It needs to work as well as Notion does. And Notion works really well.

Sadly, Loop is a mess right now. It’s been a slow boil from a development perspective. Microsoft announced Loop two years ago, but its underlying technology, called Fluid Framework, was first announced in early 2019. And like so much else Microsoft has released this year, including Copilot for Windows 11 and the new Outlook, it’s coming in hot, in an incomplete state, and with lots of features still to come. This seems to be the new normal, sadly.

But I am me, so I keep trying. But the Loop mobile app can’t even work reliably enough to display a simple page with weights I need at the gym. So I’ve never even tried to take the second step, not yet anyway. It has to get there eventually. Or maybe it doesn’t.

Guest posts on Thurrott.com

dholiman asks:

Paul, now that you’re in charge of the site, have you considered the possibility of having the opportunity for Premium members to write “guest” articles on relevant topics that they’re passionate about — maybe once a week/month/quarter?  You would obviously vet each and every one, editing for clarity and readability once you approve the article, etc.

Hm.

Sort of: One of the ideas we had before my split from BWW (and the site redesign in 2022) was to integrate the forums feed into the main article feed so that that they were commingled in some way. And this might have included some notion of me or others amplifying certain quality forum posts. But the redesign was big, expensive, and complicated, and this got lost in the mix.

But what you’re suggesting is more interesting in some ways: Readers have their own perspectives and opinions, and many come at things from different angles than I do.

That said, I certainly don’t want to take advantage of anyone in the sense that I can’t really pay people for their time and content, and I would never try to offload my own work to readers. But I could see this being of interest to some, and there’s no reason we couldn’t offer some form of promotion or whatever.

So I guess I’ll leave this hanging here for others to read and chime in. Maybe we can make something of this.

TWiT

christianwilson asks:

Thoughts on TWiT?  Leo spoke openly about the financial challenges the company is going through on This Week in Google and Windows Weekly. Jason Howell, Ant Pruitt, and Victor Bognot were let go. Terrible to hear about, especially heading during the holiday season, but clearly this was a business decision that was difficult to make. It sounds like the coming year may be a make or break situation if they can’t make ends meet.

For those who aren’t aware of what just happened, here’s the letter that Leo and Lisa posted. Long story short, it’s been a tough year for online advertising, not just for podcasters but for websites and blogs like Thurrott.com. Our ad revenues are down dramatically year-over-year as well, so I understand the pain those guys are enduring viscerally.

I was discussing this with my wife during our drive to the Boston area on Thursday—we’re here for a big event a friend is throwing—and how much it bothers me that the tech industry would allow another bubble, in this case the silly explosive growth in podcasts thanks to big media-oriented tech companies like Spotify, to burst. Podcasts were always exactly what they still are, niche content enjoyed by relatively small audiences, and of varying quality. And TWiT has always been the model for what high-quality podcast content is, so watching this market overrun by tech giants with more money than common sense has been difficult. Spotify and these other idiotic companies sucked all the air out of the room by overvaluing content and creators that were never going to make back the millions that were foolishly spent on them. And now they’ve ruined it. And are running off to other idiotic markets, like audiobooks, that will never deliver results either.

The hope here, of course, is that high-quality content creators like those at TWiT can survive the coming and going of interlopers like Spotify, much like the tiny mammals that survived the passing of the dinosaurs. But it’s going to be tough for at least a while. Very few advertisers, especially quality advertisers, are interested in podcasts now thanks to the failures of Big Tech in this market.

Of course, I’ve known Leo and have been a member of the TWiT family for over 15 years, so this is personal too. I rely on the money that TWiT pays me—it literally helped put my kids through college—and that’s truer now than ever thanks to our own advertising slowdown. And so the fear here isn’t just for TWiT, it’s for my own business.

I hope things improve quickly. Leo and Lisa built an incredible business and deserve continued success. I love those guys and am so appreciative of their incredible support for so many years. I wish I could do more than simply hope for the best.

Windows Everywhere in your ears

GeekWithKids asks:

Have you thought of doing an audiobook of Windows Everywhere?

Not seriously. It’s a huge book and it was hard enough getting through the writing of it. And it’s not really sold well enough to warrant the effort. I had hoped it would do better.

That said, it’s a labor of love and I still plan to expand the book to include content from more recent years, and to cover Windows 10, Windows 11, and the new .NET. But I have a lot of Windows 11 Field Guide additions to make first. I will return to Windows Everywhere when that slows down.

Hey Copilot, start me up

JustMe asks:

A speculative question for you: a recent PC World article quotes Satya Nadella as saying Copilot is like the new start button, with further discussion implying though not stating directly that Microsoft could replace the start button with Copilot as ‘the new orchestrator of your app experiences.’

Do you think this is realistic? How do you think ‘normal’ users would respond? Do you think the idea has merit (personally it’s not to my tastes, but I understand others may see value in it). Do you think it would ever take off? What are your thoughts on the idea?

Nadella made this comment during a pre-recorded interview with Qualcomm CEO Cristiano Amon that played during that company’s Snapdragon Summit 2023 in late October.  It mirrors comments that Yusuf Mehdi has made about Copilot somehow turning average users into power users, and it’s fair to say that it’s in some ways both marketing drivel and aspirational because Copilot in Windows 11 today is laughably inept and is generally less efficient than using existing UIs in Windows like the Start menu and Search.

But to be fair to Nadella/Microsoft, the underlying AI technology comes with all kinds of perception issues, and trying to condense its benefits down to a soundbite is difficult if not impossible. And both Nadella and Mehdi are trying to convince customers to at least try it today. Most of the benefits will come from the same Copilot functionality that’s available in Bing and Edge on the web, at least for now. And I could see that being compelling to many people.

Here’s the full quote, which starts at about 48:45 in the video linked above.

“There is a new generation of AI PCs that … cannot be done without a new system architecture … The marquee experience for us is going to be Copilot. When … Windows first came together, we had the Start button. The Copilot is like the Start button. It becomes the orchestrator of all your application experiences. So, for example, I just go there and express my intent and it either navigates me to an application or it brings the application to the Copilot. So it helps me learn, query, create, and completely changes, I think, the user habits.”

It is absolutely not coincidental that Nadella uses the same orchestration language that Stevie Batiche used when he discussed copilots and AI application structures at Build 2023. (Yep, I’m referencing this again. Seriously, this is like the AI Rosetta Stone for Microsoft’s AI work.) This bit comes at about 1:42:52 into the video linked above.

“AI is the new interaction technology … Our interfaces are transforming from being exact to being implicit and fuzzier, less programmatic, more piloted. [Here, he describes the three AI application structures I’ve already referenced several times.] This brings us to the third and final application structure, where AI goes from executing from within in the context of the application frame to being outside, executing globally. Here, the AI will orchestrate across the multiple apps, plug-ins, and services, functioning more as an agent … If you take a step back, the Windows shell itself is an orchestrator. In fact, maybe one of the most powerful orchestrators across apps, across content, across the [Microsoft] Graph. Imagine with AI and natural language, you start to see glimpses of the opportunity with [Copilot in Windows]. And it is here when you get intelligence that is functioning not just at granular details, but at the higher levels where you get a mixing of both tactics and strategy, you get both vision and execution. It’s like a Copilot of Copilots, a very powerful application structure.”

Combine these two sets of quotes and you get what I think is a vision, if not for a thing explicitly called Windows 12, then at least for the future of Windows, of how Copilot specifically and AI more generally will evolve how we interact with PCs. Whether this replaces the GUI is sort of beside the point at this early stage, but it’s still worth debating because Microsoft is indicating that it thinks so.

In the short term, it’s more likely that AI will provide another interaction model, similar to how multitouch and smartpen interactions are there now if you want them. The question is whether the AI functionality improves so much that it takes over as the primary interface. I suppose you could make a comparison to Windows in the mid-1990s, when many users would go back and forth between the MS-DOS command line interface and the Windows GUI, with the latter finally taking over completely with Windows 95.

For this transition to occur, for users to launch “orchestrated AI experiences” on PCs rather than discrete apps, a lot has to change. And that’s what the Batiche discussion about AI application structures is all about. In some ways, it’s possible to view these structures as a linear evolution from one to another, but it’s far more likely that each will co-exist, in the same way that legacy apps like Microsoft Word co-exist with more modern web and mobile apps in Windows today. But however it comes about, a transition from the GUI and what Batiche calls “explicit” apps to an AI-based UI model of not just natural language but also more general natural interfaces and “fuzzier” experiences that will likely be built from components that are assembled—orchestrated—into end-to-end solutions seems inevitable. The only question is how long it takes.

Here’s a simple example of this kind of thinking from the past: In Windows 95, Microsoft provided what it called a document-based user interface, where you could create new documents directly from the shell instead of launching a specific app and then creating a new document there. You could also create compound documents that commingled document types from multiple apps, like embedding an Excel spreadsheet into a Word document and having the Excel toolbars appear when that component was selected. This was very forward-leaning and deeply innovative, but it was also confusing to most users and was abandoned in later releases. It was probably too soon, frankly.

But Microsoft clearly has a similar vision for AI in Windows today. So instead of thinking, hey, I need a specific kind of graphic for an article, so I should launch some specific app to do that, in the future, I might literally just tell the PC what I need, and it will use whatever tools—apps, Loop components, online services, whatever—to make that happen. This is in many ways the document-centric interface come to life again in what may be a more realistic and sustainable way.

Interesting stuff.

WTF FTC?

JustMe asks:

You touched on this in WW, but what, exactly, is the FTC up to with Microsoft-Activision? I can understand trying to reign in big mergers and even trying to somewhat control big tech – but they have lost this fight. I don’t know that the FTC’s arguments ever stood a chance of succeeding, apart from just dragging proceedings out to try and kill the merger that way.  They picked the wrong fight.  Why persist in a case that has already been decided? What is to gain – by anyone (apart from the lawyers and their paychecks) in persisting with this?

I struggle with this. I really do.

I agree with you across the board, and there was a moment last summer when it suddenly occurred to me that maybe this was really about dragging out the proceedings so long that Microsoft would just give up. But given what’s happened since, with the FTC picking up the fight again after losing in federal court (twice), it’s clearly something else. And to figure out what that might be, all we can do is examine what it’s said about these efforts. And since the FTC just appeared in a federal appellate court this past week to press its case, that should be easy enough.

But in reading through news reports from Reuters, CNN, and other reliable sources, it appears that the FTC argument boils down to a single point, and it’s procedural in nature: It believes that District Judge Jacqueline Scott Corley held the FTC to high a standard by requiring it to prove that Microsoft’s acquisition of Activision Blizzard was anti-competitive. Instead, the FTC believes that all it needed to do was demonstrate that Microsoft could withhold Activision games from competing video game platforms.

For its part, Microsoft retorted that the standard is not as low as the FTC would like, and that the agency never provided a “scintilla” of evidence regardless. And the FTC’s concerns about “Call of Duty” were explicitly addressed by Microsoft.

Oddly, the appeals panel seemed mostly concerned about cloud gaming, that non-market that so obsessed the UK CMA. This was never a concern with the FTC, a point Microsoft’s lawyers hammered on, but one of the judges on the panel asked why Microsoft’s settlement with CMA couldn’t simply resolve the FTC’s new concerns. To that, the agency had no good answer.

I have no idea what the FTC is thinking. This isn’t just an uphill battle, it’s the wrong battle.