Windows Weekly 1003 Transcript

Please be advised that this transcript is AI-generated and may not be word-for-word. Time codes refer to the approximate times in the ad-free version of the show.

Leo Laporte [00:00:00]:
It's time for Windows Weekly. Paul Thurrott's here. Richard Campbell is back from Scotland. He's got a little Scottish tipple to talk about at the end of the show. We'll also talk about 26H2. It's finally here. Yay! Lots to talk about next on Windows Weekly. Podcasts you love.

Leo Laporte [00:00:21]:
From people you trust.

Paul Thurrott [00:00:24]:
This is TWiT.

Leo Laporte [00:00:31]:
This is Windows Weekly with Paul Thurrott and Richard Campbell, episode 1003, recorded Wednesday, September 30th, 2026. Another story ruined by Keith. It's time for Windows Weekly. Hello, all you winners and you dozers. This is the show where we talk about the latest news from Microsoft with 2 of the smartest people in the entire universe. He's a monster.

Paul Thurrott [00:00:57]:
I am a monster. I assume you're talking about me. Sorry.

Leo Laporte [00:01:00]:
Paul Thurrott. That's pronounced Paul.

Paul Thurrott [00:01:05]:
Sundar Tharott.

Leo Laporte [00:01:06]:
Sundar. Paul is, of course, at paultharott.com. Tharott.com. You don't need the Paul. No first names. Mr. Tharott, okay, but Tharott. Did you ever—

Richard Campbell [00:01:16]:
Or Pal. Pal Tharott is good.

Leo Laporte [00:01:17]:
Pal.

Paul Thurrott [00:01:18]:
Pal.

Leo Laporte [00:01:18]:
Our pal. Pally Paul. And his books are at leanpub.com. And he is in beautiful Mukunji, Pennsylvania, home of the tree nut.

Paul Thurrott [00:01:30]:
The what? I don't know.

Richard Campbell [00:01:33]:
Random. It's just random. He's just making noises.

Leo Laporte [00:01:36]:
I'm just making noises out of my mouth right now, like an AI. Also with us, Richard Campbell, who is at home as well in British Columbia. We welcome you back. We missed you, Richard, at Rhonda's studio.

Richard Campbell [00:01:49]:
Did you? It looked just like this being in Scotland. I tell you, the west coast of Scotland, west coast of Canada, Very similar.

Leo Laporte [00:01:57]:
Are we on the same latitude?

Richard Campbell [00:02:00]:
Uh, pretty much, yeah. They're, they're a little bit further south, honestly.

Leo Laporte [00:02:05]:
And they're on a more Atlantic-y kind of ocean.

Richard Campbell [00:02:08]:
Yeah, they've got the waters warm, but the, uh, the conditions are rough, and our conditions were rough too. Yeah, uh, I've also got my, um, I've got my, my orange shirt on. It's stat holiday here in Canada. This is Truth and Reconciliation Day.

Leo Laporte [00:02:21]:
Oh, that's neat.

Richard Campbell [00:02:23]:
It's the 5th year of that.

Leo Laporte [00:02:26]:
So that was coined by, was it Nelson Mandela? It was in South Africa, the Truth and Reconciliation, but Canada—

Richard Campbell [00:02:33]:
The idea, yeah.

Leo Laporte [00:02:35]:
The idea.

Richard Campbell [00:02:36]:
And Canada has absorbed it. We're taking it on to deal with a similar issue, which is the consequences of the residential schools that ran in Canada from its formation until the 1990s.

Leo Laporte [00:02:46]:
Horrible. And we had them too in the United States.

Paul Thurrott [00:02:48]:
True.

Leo Laporte [00:02:49]:
In the Plains states especially.

Richard Campbell [00:02:51]:
It's a process. It takes time. And one of the things is just reminding ourselves that this is going on.

Paul Thurrott [00:02:57]:
Yeah.

Leo Laporte [00:02:57]:
I think that's wonderful. My daughter, she said, Dad, why don't you tell me about Canada?

Richard Campbell [00:03:06]:
Wow.

Leo Laporte [00:03:06]:
She took a little vacation in Vancouver and just got back.

Richard Campbell [00:03:10]:
Well, that'll spoil you.

Paul Thurrott [00:03:11]:
She should try January and then we'll talk about it again.

Leo Laporte [00:03:15]:
I know. Vancouver's pretty mild. It's not—

Paul Thurrott [00:03:18]:
No, Vancouver's, yeah.

Leo Laporte [00:03:19]:
But Whistler is great if you wanna ski and Vancouver is beautiful and I love it. I'm a fan. She's actually now investigating Canadian citizenship because we have Canadian ancestry.

Paul Thurrott [00:03:30]:
Interesting.

Leo Laporte [00:03:31]:
And I don't think it's gonna work, but we'll see. She's gonna try. But yeah, she loved it. And one of the things she loved is we do have First Nations blood on both sides of the family. So she loved it.

Paul Thurrott [00:03:43]:
So she said you and Elizabeth Warren, am I right?

Leo Laporte [00:03:45]:
Just like Pocahontas herself. So, uh, actually Abby's descended, I'm not, but Abby's mom is Quanah Parker's great-granddaughter or something like that. He was a great chief of the Plains Nations. And, uh, Bonnie Parker's, uh, also a relation of Bonnie and Clyde fame. Anyway, uh—

Richard Campbell [00:04:05]:
And to be clear, those borders were drawn by Europeans, right? The native populations crossed back and forth all the time.

Leo Laporte [00:04:11]:
Absolutely. The 49th parallel is imaginary.

Richard Campbell [00:04:14]:
Yeah, totally.

Leo Laporte [00:04:15]:
Um, but in fact, I think we— didn't Americans draw it? Uh, I can't remember. We—

Richard Campbell [00:04:21]:
We made a deal with the— it was a deal with the British.

Leo Laporte [00:04:23]:
With the Brits.

Richard Campbell [00:04:24]:
Americans.

Leo Laporte [00:04:26]:
Um, anyway, uh, she loves it. So you're gonna— I'm, I'm envious. I neglected to mention I have a friend up there.

Paul Thurrott [00:04:33]:
I was gonna say you're gonna be swimming in Laportes, but I feel like Canada is full of Laportes as it is, so—

Leo Laporte [00:04:39]:
It is a famous French-Canadian name. In fact, so much so when I lived in Providence, which of course is filled with French Canadians, There were 21 Leo Laports in the phone book.

Richard Campbell [00:04:50]:
Wow. That's a lot of Laports.

Leo Laporte [00:04:51]:
I think that should give us citizenship right there. That's all you need to know anyway.

Paul Thurrott [00:04:56]:
Right. You should— it should show up with your name. You're like, yeah, no, you're in. Yeah. You're like, but I'm a Laport. Like, nobody cares.

Richard Campbell [00:05:01]:
Yeah.

Leo Laporte [00:05:01]:
There's a— it's actually a Québécois. There's a whole family tree.

Richard Campbell [00:05:05]:
Yes. It's probably a Métis tie. As soon as you get in Quebec for any length of time, you probably have some Métis blood.

Leo Laporte [00:05:11]:
I'm sure I have it. And then there's Inuit. There's a whole bunch of stuff.

Richard Campbell [00:05:16]:
It is complicated.

Leo Laporte [00:05:18]:
Yeah, I'm very— so say it again. What's the name of the day today?

Richard Campbell [00:05:22]:
Truth and Reconciliation Day.

Leo Laporte [00:05:24]:
And you have an orange— what is that shape?

Richard Campbell [00:05:27]:
It's a t-shirt.

Leo Laporte [00:05:29]:
Oh, it's a t-shirt.

Richard Campbell [00:05:30]:
Yeah. And so I have the t-shirt somewhere too, but I'm not feeling well enough. I need the flannel on. I'm a little, you know—

Leo Laporte [00:05:36]:
Well, most Canadians would wear the orange t-shirt under the flannel, I think.

Paul Thurrott [00:05:39]:
Yeah. Or here you'd want to wear it on the outside because of the bow hunters. You know, wherever you live is a little different.

Leo Laporte [00:05:44]:
A good point. A good point. Yeah, so I think that's great. Right on. I think it's good to remember that, honor it, and continue the work because much work needs to be done.

Richard Campbell [00:05:56]:
Much to be done.

Leo Laporte [00:05:58]:
Let us talk, however, about Windors because apparently there is something afoot.

Paul Thurrott [00:06:05]:
Yes. Yes, there is.

Leo Laporte [00:06:08]:
Again, what would that be, Mr. Thrott?

Paul Thurrott [00:06:12]:
Just out of nowhere. And I guess it was reasonable timing. Microsoft—

Richard Campbell [00:06:18]:
Been coming for months.

Leo Laporte [00:06:19]:
Yeah.

Paul Thurrott [00:06:20]:
You might not have heard of this one, guys, but—

Richard Campbell [00:06:22]:
Out of nowhere.

Leo Laporte [00:06:24]:
You'll forgive me because it seems like every episode we talk about 26H2 and I didn't know it wasn't even out yet.

Paul Thurrott [00:06:31]:
Right. Right.

Richard Campbell [00:06:32]:
It was in the insider cycle for months. Yeah.

Leo Laporte [00:06:34]:
Because we have been in the second half of '26. Since July, if my math is—

Paul Thurrott [00:06:38]:
It's also since 23H2, right? Yeah, versions like kind of don't matter anymore, right? You know, they're all getting the same features. If it's out and it's supported, you're going to get the same features basically. So that's part of the trick. Um, there are no new features in this version. Um, there are some new defaults which won't impact you unless you get a new computer or just do a clean install. Nothing, well, nothing particularly dramatic there, but, you know, there's the big bucket features, which we all actually have. You know, Richard missed Week D last week, but—

Leo Laporte [00:07:12]:
Oh, poor Richard.

Paul Thurrott [00:07:15]:
You know, but 26H1 caught up on the new Start menu, the new taskbar, the new Windows Search, et cetera, et cetera. So that's fun. They're in the tent now. And then, you know, that's part of 26H2 as well. So—

Richard Campbell [00:07:26]:
I want you to know that while you were talking about Week D, I was drinking at Bonahaven.

Paul Thurrott [00:07:33]:
Yeah. Yeah, no, you win. I get it. I don't know. This is, I don't know. Mostly 26H2 is what I think of as a licensable moment, meaning this is when the support cycle changes, right? So 24H2 in fact is going out of support on, I think it's the 13th of October.

Richard Campbell [00:07:52]:
Mm-hmm.

Paul Thurrott [00:07:53]:
So going forward from that date in a little bit under 2 weeks, we will have 25 and 26H2 and 26H1, sorry, as the supported Windows versions.

Richard Campbell [00:08:05]:
A question from the peanut gallery.

Leo Laporte [00:08:09]:
Can I download a 26H2?

Paul Thurrott [00:08:12]:
Yes, Keith512, and thanks for screwing up my back of the book tip, you jerk. Oh, shh. Anyway, I'm sorry. No, the answer is you'll know.

Leo Laporte [00:08:20]:
The answer will appear.

Paul Thurrott [00:08:23]:
But yes, you can. It's, um, Yeah, if you just—

Leo Laporte [00:08:26]:
That's not always the case, right? Sometimes they put out these updates and they don't give you an ISO for months.

Richard Campbell [00:08:31]:
Yeah. Well, everything ARM-related was delayed, but—

Paul Thurrott [00:08:33]:
If you Google download Windows 11 and go to that Microsoft page, it has 26H2.

Leo Laporte [00:08:39]:
Can we say that Windows on ARM is mainstream yet or no?

Paul Thurrott [00:08:44]:
I felt like it was that, that was true about 2 years ago, but actually we're gonna talk a little bit about this 'cause— I'm sorry, I keep interrupting. No, no, it's fine. It kind of weaves in and out of a couple of things that are going on here, right? Because there's a history there with, well, starting with Windows 8 and Windows RT, and then they abandoned that. And then Terry Myerson and the Windows 10 team went to, you know, teamed up with Qualcomm. We're gonna do Windows on ARM again, which Qualcomm of course called Windows on Snapdragon. And they released, I wanna say, 4, maybe even 5 generations of really garbage chips. The first one was literally a chip from a phone.

Richard Campbell [00:09:19]:
Hmm.

Paul Thurrott [00:09:20]:
And then they bought Nuvia and they came out with what we now call Snapdragon X and X2 now. And those are fantastic. So it's not just them doing that. There's also Microsoft doing the work on the software side for all kinds of things. Remember Windows on ARM, Windows 10 on ARM, when it first came out, it was only compatible with 32-bit x86 apps, then eventually x64. It is now fully 64-bit. And there's just a lot of software work, right, to get that stuff going. We talked about, I think, AutoSR either last week or the week before as a way to help port games, 'cause games are not going to be, actually not port games, they're not gonna be ported, right? They're going to run under emulation.

Paul Thurrott [00:10:04]:
So anything you can do to make that better is obviously good, et cetera, et cetera. But now Nvidia's here screwing everything up. So we'll talk about that in a moment.

Richard Campbell [00:10:12]:
I tell you from the administrator's perspective, enterprises are still shying away from Arm.

Paul Thurrott [00:10:17]:
Oh yeah, 100%.

Richard Campbell [00:10:18]:
We don't wanna support it. Our CTOs are not demanding it. Like, it's just— Yep.

Paul Thurrott [00:10:24]:
Which is too bad because I feel like if any one of them spent 5 minutes with one of these things, they would say, why are we not using this everywhere? Yes.

Richard Campbell [00:10:30]:
Yeah. I totally agree. And I think, I wonder if in the normal laptops, like, it'll go there, but because it's been bound to Copilot Plus PC, it's been too expensive.

Paul Thurrott [00:10:38]:
Yeah.

Richard Campbell [00:10:38]:
It's several hundred dollars more. And when you are, you know, on the hook for buying 1,000 of them, Yep, a couple of hundred bucks adds up to a lot of money.

Paul Thurrott [00:10:48]:
Yeah, well, there's a lot that goes into this. I mean, there's— it's not just inertia, although I, I do— I would cite inertia as the number one problem here. It's just the lack of familiarity, uh, the— even though it actually just works, it's fine, you know, it's different and people are nervous about that. And, you know, you don't want to have that thing where you roll out thousands of laptops and the one crucial app doesn't work for everybody, whatever it is.

Richard Campbell [00:11:10]:
But Well, and yeah, and admin gets— IT gets no slack on that, right? If it doesn't work properly with WSUS and you depend on that, well, now you're in hell, right? Like, there's, there's a lot— there's a whole other set of tools that the administrators need to run that nobody will have bandwidth for if they have any problems at all. So they've gotta— they, they're careful for a reason. Change is good, you go first.

Paul Thurrott [00:11:32]:
Yep, yep, yep, yep. Uh, Keith, who ruined my back of the book for the day, is also asking about whether, um, The RTX Sparks machines that come out will have 26H2. And based on what we saw at IFA, the answer to that is yes. But we'll see, right? I mean, but yeah, I mean, 26H2 was what they were all running. I was looking for any indication there if there was something a little different, but it looked like stock 26H2. So anything else, Keith? You wanna ruin anything else? My weekend's coming up. You wanna just tell the plot of a movie? Sorry.

Richard Campbell [00:12:10]:
Have you seen Odyssey? Exactly.

Paul Thurrott [00:12:13]:
How does it end?

Richard Campbell [00:12:13]:
You know everybody dies, right?

Paul Thurrott [00:12:16]:
Spoiler. Anyway, I'm sorry. So, okay, let's see. That's why I don't look at— I get distracted so easily.

Leo Laporte [00:12:27]:
I can do that for you, so don't worry.

Richard Campbell [00:12:29]:
That's my job.

Leo Laporte [00:12:30]:
And then I'll distract you myself.

Paul Thurrott [00:12:32]:
I distract myself. I'm like a cat with a laser pointer, except I'm holding the laser pointer. I'm like, what is that thing? This is semi-random in many ways, but there is this guy named Scott Jensen, who you can see videos of him on YouTube, who is amazing. He's a UX expert. He worked on the original human interface guidelines at Apple. He worked on Mac System 7. He worked on the Newton. He led Google's mobile UX group for 5 years, has worked at a couple of other companies, actually went back to Google for a couple more years.

Paul Thurrott [00:13:07]:
And he is— he has some really interesting ideas about UX and how things have kind of stagnated on the desktop and he doesn't understand why.

Richard Campbell [00:13:17]:
And that when you look at— Because the desktop became irrelevant?

Paul Thurrott [00:13:20]:
Well, I mean, still a billion-something users, right? I mean, it's still where we go to get work done. He's like, in his view, and I share this, I feel like there's more innovation that could occur. And the only kind of changes that have occurred over the past many years is that Apple and Microsoft to a lesser degree, 'cause they don't have the mobile platform to lean on, have brought like kind of mobile experiences to the desktop, right? Especially Apple. But, um, but in Microsoft's case, you know, Windows 11 like kind of, you know, looks like mobile. It's, you know, there's really not much going on mobile, but it looks like mobile. So, but the reason I raised this issue is not to talk about UX or whatever, and who cares, but he, he tells a story about how he interviewed to become the, um, UX lead at Windows at Microsoft 8 years ago.

Richard Campbell [00:14:12]:
Wow.

Paul Thurrott [00:14:13]:
And so last year he first came up for me. Well, no, actually, back in the early 2000s, he wrote a pretty amazing book about kind of simplicity and user interfaces and stuff, which was really well— I think it was called The Simplicity Shift. And I had actually— well, it doesn't really matter, but I had held on to a domain like the Simple Stick. I had this idea around this myself. Like, I'm really interested in simplicity and how hard it is. Like, we know from Windows 11, like, to simplify Windows, Microsoft just removed a bunch of stuff and everyone hated it.

Richard Campbell [00:14:43]:
Yep.

Paul Thurrott [00:14:44]:
And, you know, it's a little more complicated than just, you know, you can't just dumb it down. Like, you know, how do you retain, you know, functionality without making the thing, you know, pointless? Anyway, so this guy's been around for a long time, obviously, but he was at the— there's an annual Ubuntu event and he spoke there and he talked about this stuff. And then he works in the open source community now. So I don't know if he works for or works with KDE, but He's with that organization and he spoke at their event sometime in the past week or so. And it was kind of an extension of that first talk from last year. But he tells a story about how he interviewed at Microsoft, right? And it's sort of telling that he says, look, you're not really supposed to talk about this stuff. I'm going to talk about it a little bit. But he's like, look, I went in there like, I don't want to blow up Windows.

Paul Thurrott [00:15:35]:
I understand there's millions of people, billion, whatever it is. We're not going to freak people out. We did that with Windows 8 already, but I do think we need to be experimental. I think there's some additional things we could do, et cetera, et cetera. And he's like, it's not enough just to coast on this basic desktop UI that we've had literally since 1984, or even a little earlier with the Lisa, whatever. And they were like, yeah, we don't want that. He's like, okay.

Richard Campbell [00:16:00]:
Well, and part of this is the customers don't want it. Right. Every bit of pressure has been that customers always get angry when we change anything.

Paul Thurrott [00:16:11]:
So, I mean, again, without getting too far into the UX thing, right? There are, yes, it's fair to say even good UX changes, there's been a lot of pushback. Like when you have a toolbar and menu-based UI like we had for Office, it basically exceeds the number of commands that make sense for that kind of UI. It blows up in your face. They did pretty good. There was a lot of pushback, but the ribbon interface is actually a smart UI.

Richard Campbell [00:16:38]:
Yeah. Good enough that everyone's forgotten about it. That's what UI is supposed to do.

Paul Thurrott [00:16:42]:
Yes. Right. It just works. So that's an example, right? Then of course we have Windows 8. And my whole thing with changing, like, whatever it is, workflows, UIs, however you want to state it, is that you have to provide some kind of a transition because people are change-averse, right? And you can't just Say, yep, everything's different. We're going cold turkey. We're going to do everything completely differently. And, you know, that was Windows 8, right? That was the fatal flaw there.

Leo Laporte [00:17:07]:
Yeah.

Paul Thurrott [00:17:08]:
Well, it is. I mean, previously, even in Windows 95, for example, if you didn't like the desktop and the Start menu, you could go back to Program Manager, right? In Windows, I think it was XP was probably the first one where they had the kind of big, wide, colorful Start menu. They're like, no, we still have the other one.

Richard Campbell [00:17:25]:
Yeah, if you want it.

Paul Thurrott [00:17:26]:
It's a transition, right? And Windows 8, they were like, yep, we're not doing anything from the past, sorry. And everyone hated it.

Richard Campbell [00:17:32]:
Well, and I mean, Sinofsky broke every rule there. Normally you bring in your biggest customers, they help test it, they make sure they can work with it. They lend you credibility that it's worthwhile. He didn't do any of that.

Paul Thurrott [00:17:45]:
Right. And I was talking to Richard ahead of the show, but I was rereading this part of my own book because I'm that lonely. And I was kind of reliving part of that whole Windows 8 era. of stuff like right before and then during the release of that product and all the stuff that was going on at that time. And, you know, one of the many notable things about that release is how much they bragged about all the feedback they had. This is like the most tested version of Windows ever, you know, the most, uh, feedback anyone's ever received, but they didn't act on any of it. Like, what I mean, they literally—

Richard Campbell [00:18:14]:
We just didn't agree with any of it, so we didn't do any of it.

Paul Thurrott [00:18:16]:
So as soon as he was gone and then 8.1 happened and whatever came after 8.1 before 10, they started rolling back the most egregious changes. And we're talking a lot about how we're actually listening to feedback now and acting on it, which is the way anyone should act in such a situation.

Richard Campbell [00:18:32]:
The worst-case scenario is what's happened to Visual Studio, where the main customers do not want any of it to change, and now it's led itself to irrelevance. Jensen's point is the desktop is becoming irrelevant and we're letting it happen.

Paul Thurrott [00:18:47]:
Right. Right. Yeah. Microsoft would've had an easier time selling mobile UI in Windows on the desktop if they had had a successful mobile platform, perhaps. Maybe. I don't know. But then again, we have the Mac to look to. And Apple gave up on the Mac, it seems, for many years there and just added a bunch of iPhone UI.

Paul Thurrott [00:19:08]:
I mean, literally the last 2 releases of macOS, the big things are Liquid Glass.

Richard Campbell [00:19:12]:
Macs exist to program iPhones and iPads.

Paul Thurrott [00:19:15]:
Right.

Richard Campbell [00:19:16]:
That's why they're there.

Paul Thurrott [00:19:17]:
Yep. But I mean, look, I think everyone understands that would be possible on an iPad Pro. They just don't allow it. Whatever. That's— look, they're in their own world. I mean, a big part of their business model is just selling hardware. So they come at it from a different perspective. But the one thing I did kind of agree with in Windows 8, because it kind of predated, it was just the notion that computer form factors can change.

Paul Thurrott [00:19:42]:
It doesn't mean we throw out things that work. People love laptops. People still have desktops, big screens, whatever it is. But there's no reason like a tablet-type form factor couldn't make sense for a lot of people.

Leo Laporte [00:19:52]:
I believe that someday we'll all be carrying Meta Tamagotchis around our neck. And that's—

Paul Thurrott [00:19:59]:
I was not the first person to observe this, but explain to me how this new Meta service is like adults only. And there's a children's Kush toy that is the—

Richard Campbell [00:20:07]:
That is the primary interface.

Paul Thurrott [00:20:08]:
That's the interface, like the way they're marketing it. It's like selling cigarettes to children.

Leo Laporte [00:20:14]:
I tried to— well, I want one of them. I'll buy 'em 'cause I'm a child at heart. But I wanted to, um, I, I have a pair of lederhosen, as one does, in the closet. It is Oktoberfest. And, uh, I mean, I have the whole kit and caboodle, the shirt, the shoes, the deerskin shoes, the knee socks.

Paul Thurrott [00:20:32]:
I just started traveling. We went to an Oktoberfest party over the weekend, and the couple who held this thing were wearing that type of outfit and said jokingly to me, how come you're not wearing your lederhosen? And I said, you know, there's only one person on this earth who I know who actually owns clothes like that, and he's not German. But I do a podcast with them every Wednesday.

Leo Laporte [00:20:49]:
And, uh, you knew that I had lederhosen.

Paul Thurrott [00:20:51]:
I literally just came up on set this past Saturday.

Leo Laporte [00:20:54]:
Like, that's so funny.

Paul Thurrott [00:20:55]:
Yeah.

Leo Laporte [00:20:55]:
So I, uh, I had Lisa, uh, I put them on as one does, and I had Lisa snap a shot of me, um, you know, in them so that, uh, uh, you know, I could make a little, uh, social posting of me and my lederhosen. And I gave it to Muse and I said, take the background out Uh, I've lost my Tyrolean hat, so put a Tyrolean hat on me. I was just holding my fists like this. I said, put 2 giant beer steins in my fists and then put a backdrop of a beer garden with buxom women in dirndls behind me.

Paul Thurrott [00:21:28]:
Okay. Uh, was this an NSFW type situation?

Leo Laporte [00:21:33]:
You said, I can't put— I can't do beer. Let me see if I can do a poster.

Paul Thurrott [00:21:38]:
But I could do a bunch of naked women. Is that okay? It would not do.

Leo Laporte [00:21:41]:
It would not do it. But good news, I have a local AI, as one does, and, uh, it, it's had no trouble making a picture of me in my dirndl with giant beer steins and, and, uh, women laughing at me behind me. I didn't, I didn't specify that, but apparently.

Paul Thurrott [00:22:00]:
So those, uh, those pretend beers you were holding one time we were in, I think Brussels, but somewhere in Belgium, it doesn't matter. And it was hotter than hell. And we went to some bar restaurant thing and the woman's like, can I help you? And I said, yeah, I need a beer the size of my head. And she came out with that exact size glass of beer. And I was like, I was kidding.

Leo Laporte [00:22:18]:
That's apparently authentic.

Paul Thurrott [00:22:19]:
And it was literally the size of my head.

Leo Laporte [00:22:22]:
Well, there you have it, ladies and gentlemen.

Richard Campbell [00:22:24]:
I thought those— I feel though, do you have the arms to hold a beer quite that big?

Leo Laporte [00:22:27]:
Yeah, exactly.

Richard Campbell [00:22:28]:
Those are 15-pound beers, right?

Paul Thurrott [00:22:30]:
Usually.

Leo Laporte [00:22:31]:
But these barmaids in Germany, they are very strong. I did actually I think maybe that was overkill, so I did a—

Richard Campbell [00:22:38]:
They also keep it— if you watch them, they keep their arms tucked in close for a reason. Like, coverage matters.

Leo Laporte [00:22:43]:
I, I did a second one with slightly smaller ceramic steins, but I noticed the women are still laughing at me. I think it's the way the lederhosen enhance my, uh, manly attributes from behind.

Paul Thurrott [00:22:55]:
I don't know what it is.

Leo Laporte [00:22:58]:
I call it male pattern ass loss, and I'm proud of it. I'm sorry, I really brought this down.

Paul Thurrott [00:23:05]:
I don't even know what we're talking about anymore.

Leo Laporte [00:23:07]:
Continue, just continue, please. Forget I said anything. I'm turning off my mic now.

Paul Thurrott [00:23:12]:
Anyway, so this guy, 8 years ago—

Richard Campbell [00:23:14]:
Imagine if he had joined in 2018, and obviously he, he went to the Windows team to rethink Windows at a time when Windows didn't know what to think, right? I mean, that's when WinUI was very much engaged in MAUI. Like, if he'd gone to the MAUI team, like, What could have happened there?

Paul Thurrott [00:23:29]:
So if his timing is correct, this would've been partway through the Windows 10 lifecycle. They came up with Windows 10 to do this one Windows thing where Windows Phone and Windows would have the same basic platform, run the same apps, also the same basic programming interface for Xbox and Surface Hub and whatever else, some IoT stuff. It was a vision. I mean, it's an interesting idea. It's something that should have happened years earlier. It's something that should have been Windows 8, frankly. But I would say by the time he got there and maybe interviewed, that had failed. And I think they had lost their appetite for doing anything that was potentially— Windows 8 screwed up everything.

Paul Thurrott [00:24:18]:
And you're like, I want to redo it all. And it's like, yeah, we tried that. It went really bad.

Richard Campbell [00:24:22]:
Mm-hmm.

Paul Thurrott [00:24:24]:
Might have just been bad timing. And, you know, Windows 11, look, it's just lipstick on a pig, essentially. I mean, it's just, I like it. I mean, it's pretty looking, it's modern, whatever. It's mobile influenced, but it's just Windows. Like, it's not—

Richard Campbell [00:24:38]:
Yeah. The best chance for Windows to have a major overhaul is for Jensen to actually do what he's talking about, which is working in the context of KDE, making Linux for the desktop an actual goal. And doing something truly original that starts moving seats. That people go, I want to work on that.

Paul Thurrott [00:24:55]:
Yeah. So part of his message last year and this year to the open source community was like, guys, for decades you guys have just coasted on UX because you had Windows, you had the Mac, Apple and Microsoft are taking the risks, they're taking the bullets, they're making mistakes, they're failing, and you didn't have to deal with any of it. And then when things went right, you could just copy that. And that's a great strategy. when those companies are doing those things. They've both given up. So he's like, look, new ideas are going to have to come from somewhere. They're not coming from Apple or Microsoft.

Paul Thurrott [00:25:26]:
And look, I appreciate that message. I think it's a fair point. So I'm just asking—

Richard Campbell [00:25:31]:
Well, and let's face it, the real new UX is how do we interface with the LLMs?

Paul Thurrott [00:25:37]:
Yes. And actually, so—

Richard Campbell [00:25:39]:
New OS.

Paul Thurrott [00:25:41]:
Since I wrote this article, I've rewatched, or Rewatched some older things and watched things that were new to me. And he gave a similar talk at some kind of a local AI conference. And he kind of leaned a little more heavily to that stuff. But yeah, talking about like, well, how do those interactions kind of influence what desktop UI could look like, right?

Richard Campbell [00:26:04]:
Mm-hmm.

Paul Thurrott [00:26:05]:
And look, it's not so much, and he says this, it's not like he's like, look, I invented the UI, just do this, it's awesome. That's not the point. The point is like things evolve. We have ideas, we can experiment. We might, you know, this might or might not work, you know, but we can't know unless we try. And I do think, look, there are definitely ideas that kind of came out of mobile that are worth doing on desktop and vice versa, right? There are definitely ideas that are coming out of this AI stuff that, you know, yeah, will and should influence UIs of all kinds everywhere, desktop, mobile, whatever. I mean, There's no doubt about it, you know.

Richard Campbell [00:26:42]:
Yeah, the, this, the, you know, the, the, the B-side model is literally like your former internal dialogue with yourself while you were interacting with the computer needs to be written out to an LLM to, to operate efficiently. And so you're essentially always in 2 windows now, whatever chat interface you have and whatever product you're interacting with from that chat interface.

Paul Thurrott [00:27:07]:
I don't have this in the show notes, I don't think, but I'm slowly working on this series about tech nostalgia and not just rotely going through history, but kind of looking for themes and things that I think are interesting. There's a lot of what-ifs in history. What if Gates didn't get the IBM contract to do the operating system? We'll use CP/M as a— on and on it goes. There's a million of them.

Richard Campbell [00:27:29]:
Right.

Paul Thurrott [00:27:30]:
But one of the This is not a what-if, but VisiCalc comes along and suddenly the Apple II starts selling pretty well because randomly, and I mean 100% randomly, he just wrote the app on that computer. It's just, there was no, just like someone lent him one for free. It wasn't because he preferred it. It was just the random thing.

Richard Campbell [00:27:51]:
It was the one he had.

Paul Thurrott [00:27:52]:
But these kind of things happen. But there's 2 guys really, Dan Bricklin, who kind of came up with the original idea and prototyped it. And then I think the other one's Bob Frankston, I think is his name.

Richard Campbell [00:28:02]:
Mm-hmm.

Paul Thurrott [00:28:03]:
And he wrote the app, actually wrote the Apple II original version. But he was interviewed for that excellent book from the mid to late '80s called Programmers at Work. And he basically, you can go back and see it for yourself, he talks about what became AI and what we're doing today, where he's like, they were talking about the future of computing, and he's like, look, we're going to just talk to these things and describe what we want. It's going to work off intent, not off of what we want. And He saw VisiCalc as the first example of like a non-programming language where people could make their own applications essentially. And he said, 20 years down the road, it was the timeframe, everyone's gonna be making their own apps, you know? And if you think to like just like the simplest thing in the world, like what we call widgets today, whether like desktop accessories on the original Mac or—

Richard Campbell [00:28:47]:
Mm-hmm.

Paul Thurrott [00:28:48]:
We had the sidebar in Windows Longhorn and Vista sort of where, There were these widgets, I think we called them widgets or gadgets, whatever they were called, in there. And then onto the future, we have them on mobile now. So the mobile platforms both have widgets all over the place. And this is just a thing. One of the little baby steps that's happening soon in Android is the ability to describe widgets, right? You're gonna make your own widgets. Like a normal human being who doesn't program is gonna be able to do this.

Leo Laporte [00:29:18]:
Mm-hmm.

Paul Thurrott [00:29:19]:
I mean, arguably, that kind of ability changes how UX works, right? When anyone can look at something on their computer and say, no, I don't like, or I do like this, but I want to make one change, and there's an AI that could just implement that for them, that changes it completely, I think.

Richard Campbell [00:29:36]:
Well, it doesn't matter whether it wrote original code for you or just grabbed it from somewhere else. That's irrelevant.

Paul Thurrott [00:29:40]:
Oh yeah, it could be both.

Richard Campbell [00:29:41]:
Yeah, in both cases. You just described your desire and it found a way to implement it. Yes.

Paul Thurrott [00:29:46]:
Yep. Yeah. While you were gone last week, one of the things that came up, there's a site, I think it's xaml.io, but it's a web-based IDE, C#, XAML, you create .NET apps. And those .NET apps can run in the web, they can run on any platform, they can run on mobile, et cetera, et cetera. And they created Paint, right, from Windows 98.

Leo Laporte [00:30:09]:
Right.

Paul Thurrott [00:30:10]:
And I don't mean they created an app that looks like Paint. I mean, it's like 100% does every single thing. It's amazing. And they spelled out how much it costs, how long it took. And it's basically like $50-something of work. It took some number of hours or one day, but that's in the middle here or toward the end of 2026. I mean, by next year, the next year after that, whatever, these things are just going to— it's just going to happen. And that changed.

Paul Thurrott [00:30:37]:
So we'll see. Yeah.

Richard Campbell [00:30:40]:
Anyway, I'm excited that he's talking about that. I hope it's actually a play on hire me to make you the new UX.

Paul Thurrott [00:30:50]:
So interesting. So this guy's older than we are, and in his own words, just sort of semi-retired, but he still wants to contribute. I think at the time he would've worked at Microsoft, he was still kind of actively seeking to work day-to-day, some company, whatever it was, where he might make a difference or whatever. I just appreciate him giving a crap. I think a lot of people might reach his age and be like, I'm done, whatever. I did my thing. The world changed. I don't care anymore.

Paul Thurrott [00:31:19]:
But he's got a healthy approach to it. That's my only real point. But again, like I said, he interviewed there, which is crazy to me.

Richard Campbell [00:31:27]:
Yeah. Well, and the fact that they didn't grab him immediately, obviously he didn't go through the right interview loop. Find the guy a job. He's one of the best UX thinkers on the planet and he's interested. Yeah, your challenge will be keeping him, not hiring him. Yep.

Paul Thurrott [00:31:42]:
Yeah, one of Microsoft's— I'm not to keep going on and on about this topic, but one of the things that's always weird to be about Microsoft, but especially in Windows, because Windows is just this kind of software-based archaeological dig where everything from the past is still in there in some capacity. And, um, there's just so many different user interfaces and user interface style, UXs, however, whatever you wanna call these things. And some of the work that's occurring this year is nice because they're finally starting to get rid of a lot of those ancient UIs, those kind of weird, you know, lower-level things that have been sitting in there for a long time and they still look like Windows 95 in some cases, it still looks like Windows 2000, whatever the heck it is, Vista, there's Vista UIs in there. So that's good. But this is a weird thing because Copilot, not super interested. But at some point you see this in the Microsoft AI stuff. And what I mean by that is not in the models, but in their kind of, I guess we'll call it like PR or what, you know, they have blog posts and things. There's a style that they've adopted for their communications, visual communications, you know, graphics and so forth.

Paul Thurrott [00:32:49]:
And if you look at that new consolidated Copilot app that they announced last week or whenever that was recently, it's using this style and they've kind of moved to it like over time. Like for all of the weird machinations with Copilot on the desktop where they keep changing it, keep changing it, keep changing it, there was a point where they kind of got where it's like, actually, this is kind of a nice look. Even the version of Copilot built into Edge where you have like the Copilot-based new tab window or whatever, has that look. And it's like, I don't want to use this product. I don't, you know, whatever. But it's like, I do like the way it looks. So, you know, they have a— they've created this kind of UI, I guess I'll call it style. I don't know, whatever it is.

Paul Thurrott [00:33:30]:
But, um, but it makes everything else look old, you know, which is kind of weird, right? Like, but that's what happens. So I don't know if we'll ever get around to that being—

Richard Campbell [00:33:39]:
I'm telling you, once you start using Copilot to search through your email just the way Outlook could never, you can't turn off Copilot. Yeah. Now, that's not a good enough reason for Copilot to exist, but, um, well, but it is a good enough reason for AI to exist of any kind. I mean, totally.

Paul Thurrott [00:33:57]:
Um, I saw—

Richard Campbell [00:33:58]:
I expect Microsoft stays relevant by making Copilot just part of the package, stops charging separate for it, and, uh, and they will win through inertia alone.

Paul Thurrott [00:34:07]:
Hold on one sec. Let me see if this is actually in here anywhere, because it might be. Maybe not.

Leo Laporte [00:34:14]:
I don't know why.

Paul Thurrott [00:34:15]:
Wait, hold on. Let me see. Oh no, we're going to talk about the super hours. We'll get to this. So look, everyone will remember when the first year Microsoft was talking about Copilot in Windows and elsewhere, Microsoft 365 2023, right? Satya Nadella infamously said he saw the prompt in Copilot as the new Start button or Start menu.

Richard Campbell [00:34:37]:
Yeah.

Paul Thurrott [00:34:37]:
And everyone was like, oh, you're an idiot. But if you look at the— we'll look at it later, we'll talk about it later a little bit, but it's kind of happening. I mean, look, there have been—

Richard Campbell [00:34:50]:
But classically, smart people do this. They jump to the endgame and talk about the endgame without showing their work and confuse everyone.

Paul Thurrott [00:34:56]:
Right. And this relates to that transition thing I was talking about, which is that there have been calls along the way over time for Microsoft to do these other things inside and outside of Microsoft, right? Windows Phone would be a really good OS for a tablet. It had those panoramic user interfaces that would've made a lot of sense on a landscape device. There was the Windows 10X stuff they were doing for a little while. There was the, I don't know if this was ever real or not, but it seemed like it was, but we're gonna do something like a Chromebook, but based on Edge, like an Edgebook. And it was like, okay. And those are all ideas. They're fine.

Paul Thurrott [00:35:28]:
They're good, bad, indifferent. It doesn't really matter. But there's this big contingent of people, and this is the— inertia, lack of familiarity crowd is like, yeah, but does it run Windows apps? Like the Windows RT, which looked identical to Windows 8, but could not run Windows apps.

Richard Campbell [00:35:45]:
Yeah.

Paul Thurrott [00:35:45]:
Right? Kind of a huge problem. So we still have those pains that we all remember. So all of these things kind of suffer from that problem. But given everything I just said, this notion that Copilot could become a UI of sorts, not a UI, I'm sorry, an OS of sorts for Microsoft in whatever form that takes, right? Whether that means literally an OS shipping on a device or whatever else. I mean, you can see it in this new app. It doesn't mean it's happening. It doesn't mean it's going to be successful, but Microsoft still has that kind of platform maker vibe to it. It's like it can't walk past that.

Paul Thurrott [00:36:25]:
It's kind of interesting.

Richard Campbell [00:36:27]:
No, and it's an opportunity to be relevant. If you make a better approach to get in, in the end, they sold a lot of Apple IIs and ultimately IBM PCs on the back of VisiCalc and 1-2-3. If you create this great interface to the new way to work, people will buy the product to get to that interface to work.

Paul Thurrott [00:36:45]:
I could drag this show down forever and we'd get to no further topics very easily, and I'm going to do it just for one second, which is I had this random idea in my head. One of the other things we'll talk about later is there was an OpenAI event yesterday. They announced approximately 11,000 different products. But the thing is, OpenAI and Anthropic, and probably many other kind of AI companies or whatever you want to call these things, really only have one product, right?

Richard Campbell [00:37:12]:
Yep.

Paul Thurrott [00:37:12]:
There have been examples where in both cases they've created this thing over on the side, Cowork or Code or whatever, And then like, it's going to be part of everything else. We're going to bring it into chat. We have notebooks, but also notebooks are available in Gemini or whatever. Google actually, I think because they're such an old company compared to these newer companies, actually does do both. They have standalone stuff, but they also like, yeah, we got to pull it into Gemini too. This is the opposite approach to what Microsoft and Apple have done over the years, and actually Google too, right? Meaning that If you compare the Apple product matrix of today to the one Steve Jobs talked about when he made a grid of 4 things, slightly more complicated. There's way more products. Microsoft has been criticized many times for having 7 or 8 different versions of Windows Vista or 7 or 8 different versions of Office or whatever it is.

Paul Thurrott [00:38:03]:
It was always this feeling of like, you're trying to pull more money out of us by having these weird differentiations between products that are basically the the same thing. But whatever, that's just been the case for a long time. Even streaming services have all these different tiers. And yeah, I mean—

Richard Campbell [00:38:19]:
What has Azure got, 300 different things just from Microsoft alone?

Paul Thurrott [00:38:24]:
300?

Richard Campbell [00:38:25]:
No, I'm talking—

Paul Thurrott [00:38:25]:
It's probably even more. I don't even know. But it is fascinating to me in a way, and I just haven't fully formed this, but big AI. So Anthropic, OpenAI, basically, they have an up product. And though they have strayed time to time, they always rein it back in. Everything's in the one product, you know?

Richard Campbell [00:38:46]:
Because in the end, that's where all the expenses are that they're trying to make profitable and are nowhere near.

Paul Thurrott [00:38:51]:
Yeah, but of course, and Microsoft as a hopeful player in the same sphere is kind of doing the same thing in a way with this Copilot Super App, right? This is their attempt to do the same thing. It does coding, it does co-work, it does, you know, You could launch the old Edge.

Richard Campbell [00:39:07]:
I mean, there's something like 80 Copilots and enough of them that the really good ones are no longer using the name Copilot anymore.

Paul Thurrott [00:39:14]:
Right.

Richard Campbell [00:39:15]:
Because Microsoft, as usual, ruined the name with a few bad ones. And you can't make a monolith like that. It's not how people interact with it.

Paul Thurrott [00:39:25]:
Yeah.

Richard Campbell [00:39:26]:
But, you know, we talked about this in '23 at that Build, right? Watching watching Stevie Batish talk about it. It's like, yeah, no, the OS is the logical starting point. You know, we, I always think in context, this isn't just the IT administrator in me, is what is the sovereign app of this organization? What is the first thing that people look at? And for many years that was Outlook. And in some organizations now it's Teams.

Paul Thurrott [00:39:52]:
Right.

Richard Campbell [00:39:52]:
And that shift's kind of a big deal. It tells you a lot about the culture.

Paul Thurrott [00:39:55]:
So yeah, one of the interesting, there are a lot of parallels between Teams and Copilot, right? One of which, the most obvious, is that it used to be like 2 versions, one for consumers, one for businesses, and now we just have the single, you know, whatever it is, right? But, um, you know, uh, Microsoft did with Teams that thing I was talking about, which was like they don't just make you use Teams. They know some old— a lot of their audience is older guys who have been using Outlook for 20-something years.

Richard Campbell [00:40:20]:
Yeah.

Paul Thurrott [00:40:20]:
That's how they start their day, that's how they end their day, that is how they, um, interoperate with other people in their organization. So We're going to give you that chance for collaboration. You can do it there. You don't have to do it inside Teams. So that's one of the things, like, that's, you know, one, one of Microsoft's main advantages, or, well, or just good strategies given their audience, is just pulling people forward with them. It's why Windows 8 was so screwed up, right? They didn't do that then. You know, they could do that for Copilot too. I mean, part of— well, anyway, I'm sorry, I'm babbling here, but—

Richard Campbell [00:40:52]:
No, no, no.

Paul Thurrott [00:40:53]:
This was a talk about UX and I'm just going on and on. But like, so the rest of the show will be really short because I've already talked about everything else. But there's a lot going on here and things are changing and changing fast. And even for Microsoft, which I would say is relatively slow moving, we're starting to see some interesting things. Yeah.

Richard Campbell [00:41:14]:
I guess. Interesting times. But near as I can tell, both OpenAI and Anthropic, They're Netscape in this story.

Paul Thurrott [00:41:21]:
Yeah, but right. And, and yes, but I would, I, I would— I'm not cautioning you per se. You have, you have the perspective. I know you understand this. But I mean, you know, Microsoft's ability to push back competitors in the past, which is in many ways unparalleled in this industry, um, does not guarantee that they'll do it forever, you know. Um, and of course, well, and they have sacrificed a lot of their lock-in.

Richard Campbell [00:41:44]:
in the past few years. And again, maybe I'm stealing thunder for some later topics here, but these moves to stay relevant have left them more vulnerable.

Paul Thurrott [00:41:55]:
Yep. And it's not just them. So one of the unique things about the AI era is that these companies, whoever they are, if it's the big tech companies we all know and hate, like Google, Microsoft, Apple, whatever, or these new upcomers, like the big AI companies like Anthropic and One UI, Typically in a normal market, so to speak, in a truly competitive market, they would all go out in the world with their own offerings, do their own thing, and have their own way of doing things. And if one of those things happened to take off, perhaps in the future we would offer some interoperability with that thing. But these guys have just, yep, we're doing it. Yep, we're doing it. Whatever it is, whatever compatibility layer, whatever the thing, MCP, on and on it goes. Every one of them is like, yes.

Paul Thurrott [00:42:37]:
Yep, 100%. And it's just because of the unique regulatory requirement or lack of regulatory environment we have today where we all understand we got to push this forward, push this forward, go quick, go quick, go quick, because eventually we're going to have a different president and things might go a little differently. And it's created a unique situation where these companies that are natural enemies are collaborating and working together in ways that would never have happened.

Richard Campbell [00:43:04]:
They're all in on the land grab.

Paul Thurrott [00:43:06]:
Yep. And it's really interesting because what they're doing, you just said it, this is why I thought of it, but they're basically removing stickiness, right? That every platform has some form of lock-in, whether it's explicit or implicit or whatever it is. It could be as simple as just being familiar with it, or I've just used it for some number of years, but it could be things like document formats and interoperability, data exchange, whatever it might be. There's all kinds of things. But in the big AI world of today, These things moving from one model to another, from a different company, do it, you know, whatever. It's actually really easy. It's like super easy. And that's not what these guys really want, right? They want you to be stuck there.

Paul Thurrott [00:43:46]:
So it's interesting.

Richard Campbell [00:43:47]:
Yeah. Well, operating systems can become like BIOS. You don't really care. You just need one. And that keeps happening, right? Like we are, what made Microsoft successful was the diversity of the PC platform. As long as— but they made sure they weren't beholden to it. They were enabling it.

Paul Thurrott [00:44:03]:
That's right.

Richard Campbell [00:44:04]:
Now they're in the position where their former hold over that market is now commodity, and they're trying to figure out what the next hole in the market is. And arguably compute is that hold.

Paul Thurrott [00:44:13]:
Yep.

Richard Campbell [00:44:14]:
Right. What can you cling to that's monopolistic in this space? It's compute.

Paul Thurrott [00:44:19]:
Right, right. So one of the best stories that comes out of that VisiCalc thing is that people would walk into a computer store and say, I need VisiCalc. And they're like, okay, that's $80, but you need to spend $2,000 to $6,000 on a computer to run it. Like, I don't care, I need VisiCalc. And they didn't even know what it was. They were like, I just need this thing.

Richard Campbell [00:44:40]:
It comes in this squirt bottle, spray it on your staff. Yeah.

Paul Thurrott [00:44:44]:
And then, you know, you mentioned Netscape, Netscape, or Marc Andreessen, who described, you know, Windows as the, you know, the buggiest set of device drivers that will, you know, underlie our operating system that runs in the cloud, you know, They didn't call it the cloud at the time, but the cloud basically. Yeah, I mean, we'll see. Maybe we boot up Windows. I don't know, we used to boot MS-DOS and run Windows. Windows was an app. Remember, we typed WIN. Then it would just automatically—

Richard Campbell [00:45:08]:
When you had to do real work, you didn't type WIN.

Paul Thurrott [00:45:10]:
And then one day MS-DOS was gone, but it didn't really matter. The interface was the same.

Richard Campbell [00:45:13]:
Well, no, then you went into Windows, but there was an icon that lets you have a DOS window to make you feel better.

Paul Thurrott [00:45:18]:
Because again, Microsoft's really good about bringing people forward. They're good at that. Often are good at that.

Richard Campbell [00:45:23]:
That was always their strength.

Paul Thurrott [00:45:24]:
Yeah. So someday instead of progman or explorer.exe or whatever it is that we're running for the UI or for the, I guess, the operating system.

Richard Campbell [00:45:32]:
The shell.

Paul Thurrott [00:45:33]:
The shell. Thank you. It will be copilot.exe. You know what I mean? Like, I'm just booted to Copilot. There's no reason that couldn't happen right now.

Richard Campbell [00:45:42]:
Nope. No, not at all.

Paul Thurrott [00:45:43]:
Other than the fact that I threw up in my mouth as I said it. I don't want that. But I mean, but—

Richard Campbell [00:45:48]:
Yeah, but that's it. That's a you problem.

Leo Laporte [00:45:53]:
It's all of us now. I hate to say it. Let's, let's pause if you don't mind briefly, just briefly. I don't want to take—

Paul Thurrott [00:46:00]:
That was like a brain dump. I'm sorry. That probably made no sense to anybody.

Richard Campbell [00:46:02]:
It's been bugging you. This is what happens when I go away for a week. You just need to unload on me. I'll teach you to go away. Yeah. And everybody else gets to benefit, I guess, or gets to experience something.

Paul Thurrott [00:46:14]:
Yeah.

Richard Campbell [00:46:15]:
I don't think this is a therapy session ahead, but there are choices. But I think we're now in a place after this many years where if we don't talk to each other each week, it's a problem.

Paul Thurrott [00:46:23]:
He's like, show us where he hurt you when he wasn't here.

Leo Laporte [00:46:26]:
Yeah.

Paul Thurrott [00:46:26]:
It's everywhere.

Leo Laporte [00:46:28]:
Hey, let me pause for just a moment. The levity will continue with Mr. Thrott and Mr. Campbell. But first—

Paul Thurrott [00:46:37]:
I want to make sure I understood what you said in that ad correctly. Was the phrase revocable entity?

Leo Laporte [00:46:43]:
No, identity.

Richard Campbell [00:46:44]:
Identity.

Leo Laporte [00:46:45]:
It's hard for me to say the word identity. I don't know why.

Paul Thurrott [00:46:47]:
No, it's in your accent.

Leo Laporte [00:46:48]:
My tongue gets caught up in my teeth.

Richard Campbell [00:46:49]:
Yeah.

Leo Laporte [00:46:50]:
Identity.

Paul Thurrott [00:46:51]:
Right.

Leo Laporte [00:46:52]:
So revocable identities is kind of critical, right? The idea— I mean, everybody's done this. You do this. I do it. I just did it the other day. My agent said, well, if I'm going to look at your Gmail, I just need permission to read. I said, oh, look, let me just go into Google and give you everything. Full permissions. It's easier than me going back and forth.

Leo Laporte [00:47:12]:
So now it has everything. The better way to do it would be say, you know, here's a token to do just this when you need it, and then, and as soon as you're done, it's no longer good. So if it does leak, it's worthless. That's all I'm saying. Does that make sense?

Paul Thurrott [00:47:28]:
Yep. I just, okay, I just didn't, I wasn't sure if I heard correctly.

Leo Laporte [00:47:32]:
Revocable.

Paul Thurrott [00:47:33]:
Revocable.

Richard Campbell [00:47:35]:
Identity. Identity.

Paul Thurrott [00:47:36]:
I'm going to call my daughter that tomorrow, but I just want to make sure I get it right.

Leo Laporte [00:47:41]:
It's kind of critical if you're using AI especially, because those tokens, they seem to get about. Every once in a while I'll get a message, you know, I should wipe your entire backup because it's got all these tokens in it.

Richard Campbell [00:47:55]:
It's like, no, please.

Leo Laporte [00:47:59]:
I promise I'll be good from now on.

Paul Thurrott [00:48:01]:
Do not revoke my tokens.

Leo Laporte [00:48:03]:
Do not revoke my tokens. I'm sorry, continue on, please.

Paul Thurrott [00:48:07]:
No, you're good. It's okay. Microsoft announced back in Build, which was in June this year, that they were coming out with something called WSL containers for the Windows Subsystem for Linux. I think later that month they shipped a public or, yeah, public preview, I think it was. And that is now publicly available. Beautiful little CLI with a fun cloud.

Richard Campbell [00:48:30]:
Well, we've always talked about this point where what if every application simply ran in a container where it thought it was the only machine, the only application in the world, and we controlled the connections between them? Because that, my friend, is the end of malware in many respects.

Paul Thurrott [00:48:43]:
Right.

Richard Campbell [00:48:44]:
Just because all the bytes are getting across.

Leo Laporte [00:48:46]:
As long as those containers are secure, yes.

Paul Thurrott [00:48:49]:
Yeah. So, I mean, there's different ways, right? So different ways to isolate code from each other. I mean, containers are lightweight, you know, obviously compared to traditional virtualization. Microsoft tried to do this with Windows 10X, remember? It's an idea, you know, it's not a bad idea. But yeah, it's a good— it's an interesting platform-level idea, that's for sure. And then I think this was from last Friday, but there were, you know, we get multiple Windows Insider builds every week, of course. And I think it was Friday, but we got 4 that weekend. It was experimental for both and whatever the other one's called.

Paul Thurrott [00:49:33]:
Oh, beta, sorry, for both of those as well. Just a couple of interesting new features. The one I would call out is the Phone Companion, which is that slice on the side of the Start menu. If you connect your phone to Phone Link. I still want to call it Your Phone for some reason. Anyway, um, the way it works today, if you haven't got— well, of course you haven't gotten this update, it's in the Insider Program— is you have like hard links for like messages, calls, photos, and files. And then there's like a list of recents at the bottom, but it's only depending on your screen, the resolution, etc., 3 to 5 items in there. So they're making that area scrollable, and there's going to be not infinite number, but many, many, many more.

Paul Thurrott [00:50:13]:
So that's actually kind of smart. That's kind of a cool idea. Uh, and then some kind of mouse stuff and enemy— not enemy, energy recommendations in Quick Settings, and who cares. Okay, so that's coming or whatever. Um, I don't expect— is this in the right place? I guess it is. So sometime, I think it was last December, Microsoft announced, uh, these companion apps for Microsoft 5 that would be available in Windows 11 only. And there were 2 originally, I think it was Calendar and People. And they're kind of like, they look like phone apps.

Paul Thurrott [00:50:47]:
So they would like, they could have a, they would have an icon in the taskbar. If you want to pin it, you can do that. And it's like a small phone-style UI for those things. And then at some point they added a Files companion app as well. And that was obviously for OneDrive for Business, whatever. There's no extra cost for using these things. I have no idea why they made them in the first place.

Richard Campbell [00:51:05]:
And I think they had some team members that needed something to do.

Paul Thurrott [00:51:07]:
Yeah, it was like an intern contest or something. Um, so yeah, I mean, look, I, I just talked about how Microsoft, you know, has done some things right, but one of the things I feel like they've done wrong is just have too many UIs, too many different ways to do the same thing, etc., etc. And this was that kind of redundancy.

Richard Campbell [00:51:23]:
So, um, they are retiring their experiments that sometimes they leak out into the world to see if they grab on.

Paul Thurrott [00:51:30]:
Yeah.

Richard Campbell [00:51:31]:
Right.

Paul Thurrott [00:51:31]:
I never saw the point of these things. I just don't get it. So anyway, apparently no one does.

Richard Campbell [00:51:35]:
There's a plan.

Paul Thurrott [00:51:37]:
I see. I've been around too long to know that that's not the case. You know, like, it just doesn't feel like there's a plan. That's the problem. But I don't know.

Richard Campbell [00:51:46]:
Anyway, they're gone and nobody's going to miss them. Yep.

Paul Thurrott [00:51:49]:
Nobody cares.

Richard Campbell [00:51:50]:
All right.

Paul Thurrott [00:51:51]:
And if you're an admin in a managed organization, you could kill those things right now if you even allowed them.

Richard Campbell [00:51:56]:
You may have already never allowed them.

Paul Thurrott [00:51:57]:
Exactly. And good on you if that's what you did. So, um, and then, uh, this is kind of like a set of software stories. So you're— if you're wondering why these are where they are, um, Mozilla put out a new version of Firefox this past week, I think yesterday actually, um, which is kind of a big deal. So this has been a big year for Firefox. Obviously we've got Smart Window, which they're still working on but is available if you want it. They've done, um, a bunch of, you know, they've done all the bug fix stuff using a—

Richard Campbell [00:52:25]:
Huge amount of security fixing. Yeah.

Paul Thurrott [00:52:27]:
Yep, there's been just a lot of stuff going on. Verily really active after years of kind of like coasting, you know, frankly, um, for a while.

Richard Campbell [00:52:34]:
But I feel like they were the first to sort of be public about the fact of, hey, AI is showing a bunch of vulnerabilities, we're going to get it done. And it seems like they're done. Like the number of patches now are just not that high.

Paul Thurrott [00:52:44]:
Yeah, I mean, I— right, so I, I'm— this is off the top of my head, so I could be wrong, but I feel like the first burst of fixes they had was for— not Gecko, the whole of Gecko, which is the biggest thing that they do really, that rendering engine, right? But maybe just like JavaScript performance or something like that or whatever. But they've kind of gone from subsystem to subsystem. So maybe they've—

Richard Campbell [00:53:05]:
Yeah.

Paul Thurrott [00:53:05]:
Obviously you wanna hit the good one, you know, not the good ones, but the most important ones first, right? I'm sure there's more to come, but yeah, that's true. This one actually does seem to have kind of plateaued nicely. But the other big initiative they've had this past year is a new UI, a new user experience, if you will. which was codenamed Nova. And that is actually available now in this new version of Firefox. And it's on desktop and mobile, which I really, really like. And it's just kind of, you know, it's fresh and modern and, you know, whatever, attractive, object— or not objectively, um, in my opinion it's attractive. But, um, but the, the big thing to me in, in many ways, uh, is that they brought back compact mode.

Paul Thurrott [00:53:47]:
And this was like, from a, you, like a user interface perspective, the thing I always hated about Firefox in recent years. The top of the browser, which is like the address bar, the tabs, and then whatever other UI might go at the top, it's like 2 inches tall.

Richard Campbell [00:54:00]:
Yeah.

Paul Thurrott [00:54:00]:
It's not really, but I mean, but it feels like it is. So when you go back compact mode, which was a feature in Firefox years ago, when you turn this thing on, it looks like a browser again. Like it's like, it's, you know, it makes, you know, it's nice looking. I'm not saying this is the reason to switch to Firefox, but But if that was one of the weird things that kind of held you back, because I really, the UI really bothered me before. And once I saw this, I was like, ah, there you go. You fixed it.

Richard Campbell [00:54:27]:
Nice. Yeah.

Paul Thurrott [00:54:28]:
This is, it is pretty. It's a, it's a nice clean.

Richard Campbell [00:54:30]:
And maybe I'm overstating like the feeling that they're on top of it. They may, there's still a lot of patches coming.

Leo Laporte [00:54:35]:
So.

Paul Thurrott [00:54:35]:
Oh yeah.

Richard Campbell [00:54:36]:
Yeah.

Paul Thurrott [00:54:36]:
But I mean, but they, I mean, the first 2 months that they did it, it was notable.

Richard Campbell [00:54:40]:
It was unbelievable.

Paul Thurrott [00:54:40]:
It was hundreds and hundreds.

Richard Campbell [00:54:42]:
Yeah.

Paul Thurrott [00:54:42]:
For like, for like an app, you know, it's just kind of, I know it's a web browser.

Richard Campbell [00:54:45]:
But yeah, it's, it's one of the most important apps. Yeah.

Paul Thurrott [00:54:50]:
Um, and then some hardware stories, uh, and this is, you know, unfortunately there's gonna be some commentary in here too. But, um, so there are some questions now about whether Microsoft and its PC maker partners are walking away from this Copilot Plus PC brand, right?

Richard Campbell [00:55:03]:
It has— it hasn't worked.

Paul Thurrott [00:55:05]:
No, it has not worked. And, and look, when I, when I go back to 2 years ago when they first came out with this thing and, and, uh, Before I knew there was a Copilot+ PC thing as a brand, the big question was this Snapdragon X thing. Like, they had announced it the previous October, I think it was. You know, the winter comes and goes, we're coming into spring, and it's like, is it real? Is— because they promised a lot in the past.

Richard Campbell [00:55:28]:
Yeah.

Paul Thurrott [00:55:28]:
I got to go see it, I think it was in April that year, and I was like, oh my God, this is— this real? Like, this is amazing. This is going to be amazing. And then they announced this brand, and it was kind of like the sound of like a balloon deflating and flying away in circles, you know? Whereas like, why would you tie it to that, right? But aside from the name, which I think is terrible, they tied it to the wrong thing. And this is another classic Microsoft thing where not always, but often they're kind of ahead of the curve. And in this case, it was like local AI on computers, which I think we can all agree, big deal. It's super important.

Richard Campbell [00:56:00]:
Yeah, it should be, but it wasn't in June of '24.

Paul Thurrott [00:56:03]:
And what they should have been selling this thing on was how reliable and fast and efficient and awesome battery life and all the awesome stuff about Windows 11 ARM running on Snapdragon.

Richard Campbell [00:56:12]:
Yeah, I just think that would've made the rest of the vendors furious.

Paul Thurrott [00:56:15]:
Yeah, it really, really, really bothers me that they, how badly they screwed it up. And then of course, their big marquee feature to start with is this recall thing where it's like, ugh, like guys, first of all, most people don't really want this or it just doesn't resonate with them. And this was the best you could come up with? I mean, like there, Today, honestly, if you have a Copilot+ PC, I could list probably 5 or 7 things that are unique so far. I mean, hopefully they spread further, but are available on this platform that are actually extremely useful and more important than Recall, like Click to Do, for example, right? We have some version of Click to Do on every computing platform. On your phone, you have something like this. What is this thing on the screen? If it's text in an image, copy to the clipboard and tell me what it— do something with it, whatever. ClickToDo is. It's excellent, but it uses an NPU and nobody cares about NPUs.

Paul Thurrott [00:57:05]:
All right, so we'll see. Here's what I have to say about this. When NVIDIA first announced the RTX Spark stuff back in the spring, summer, whenever that was, there was no mention of Copilot+PC. So that was the first red flag, the first time people were like, oh, maybe something's going on here. But I went to IFA and Copilot+PC was everywhere, and including with RTX Sparks machines, which do qualify as—

Richard Campbell [00:57:28]:
Right.

Paul Thurrott [00:57:29]:
Copilot+ PCs. So I'm like, okay, we also have this kind of component crisis where Microsoft is starting to sell 8 gigabyte— not Microsoft, PC makers are starting to sell 8 gigabyte PCs again at a time when frankly everyone, even normal people, should be using 16 gigs or more. And that's a shame and it stinks.

Richard Campbell [00:57:47]:
Sure.

Paul Thurrott [00:57:48]:
But how do you do this? So I don't remember the timing anymore, but roughly around the Build timeframe, sometime in the early summer, late spring timeframe, Microsoft refreshed Surface for, I think it was for business at the time, I don't remember. Or if you think about Surface today, they actually have the traditional Pro and laptop models, and then they have the smaller versions, right?

Richard Campbell [00:58:11]:
Right, 13s.

Paul Thurrott [00:58:13]:
Yeah, so I think it was like the 13.8 and 15-inch laptop, and the Pro was probably 13 or 13.3, whatever that is now. They are Copilot Plus PCs, but then they started selling 8 gig versions of them. And those are not Copilot Plus PCs. And how weird is that? That depending on the way you configure it, if you get one with 8 gigs of RAM, it's not going to have Click to Do and Recall and all the stuff built into Paint and Photos and wherever else we have AI.

Richard Campbell [00:58:44]:
Or will it?

Paul Thurrott [00:58:45]:
Well, it didn't. So far, yeah. I mean, they've been working on it.

Leo Laporte [00:58:48]:
Wait, they're calling them Copilot Plus PCs? Pilot Plus PCs?

Richard Campbell [00:58:51]:
Yes.

Leo Laporte [00:58:52]:
An 8 gig machine?

Paul Thurrott [00:58:53]:
No, well, no, sorry, sorry, sorry. Um, no, actually in that case, what the way they did it was when they announced them, they only had 16 gig machines that were called Pilot Plus PCs. But like everyone started, you know, everyone, I mean, nobody, Intel and Qualcomm have come out with these lower end chips. PC makers are starting to sell reams of these lower end computers, 8 gigs of RAM. So Microsoft did the same thing. They made those computers available with only 8 gigs of RAM, but they never They never talked about it. They never— the truth is they're not Copilot+ PCs. They just don't qualify.

Paul Thurrott [00:59:24]:
So you won't get those features, but they don't really talk about it at the time. This was 3, 4 months ago, right? But then, like I said, I went to IFA and Copilot+ PC was everywhere.

Richard Campbell [00:59:35]:
This is how they end the brand is they just say everything's a Copilot+ PC.

Paul Thurrott [00:59:39]:
Well, yeah. And this is the debate we've had. Like, so how does Copilot+ PC end? Badly. No, but how does it end?

Richard Campbell [00:59:46]:
It fades away. He just sold it.

Paul Thurrott [00:59:47]:
Well, yeah, one of the ways it ends is every PC basically is a Copilot+ PC, so we don't need the brand again. Or you're Microsoft, so you say, well, every brand, every PC out in the world has met this spec we had, so let's change the spec.

Richard Campbell [01:00:00]:
Yeah, our goal has been achieved.

Paul Thurrott [01:00:02]:
Well, or we just raise the bar now again, and now we have a new specification, right? But actually, I mean, if you know anything about the history of the PC or Windows or whatever, there are these, Not always one-offs, but these brief moments in time where we have this thing. We had multimedia PCs in the 1990s. You had to have a specific kind of sound card and Intel chips that had whatever code— what was the MMX commands or whatever in the processor, et cetera? And then eventually it just becomes everything. We had media center PCs. We had portable media center PCs. We had tablet PCs. We had— for 1 year Windows XP 64-bit edition. Like we've done all these different things.

Paul Thurrott [01:00:43]:
Like, so sometimes they just kind of come and go. And my money was always on that happening here. And that maybe is what's happening here. And the new theory about this is that Microsoft just released new Surface PCs again, but this time it's the smaller ones, right? So not the 13.8 and the 15 and whatever, but—

Richard Campbell [01:01:02]:
Smaller than that?

Paul Thurrott [01:01:03]:
Yeah, so there's a 13-inch Surface laptop and a 12-inch, I think it's 12-inch Pro.

Richard Campbell [01:01:08]:
So it's the size of a, of a smartphone today then.

Paul Thurrott [01:01:12]:
Yeah, I mean, they've had them for a couple years, but, um, but they usually run kind of lower-end chipsets and stuff. These ones are running Snapdragon X2+, uh, Plus chips, so it's not like the elite extreme, you know.

Richard Campbell [01:01:23]:
Yeah, but they're still pretty damn fast chips. They're not the fastest, but yeah, no, right. Even the base X2 chips are pretty smoking chip. It's low power.

Paul Thurrott [01:01:33]:
I, I agree. As they were launched or just are launching now as we speak, I went and looked at the site and I was like, interesting, the base configs on these things right now at least is 16 gigs. So they don't even offer these with 8, although I think they will, right? That's where the world's heading. But they specifically did not use the term Copilot+PC anywhere, which is a little unusual for Microsoft. So somebody asked about that and they were like, eh.

Leo Laporte [01:02:00]:
Yeah.

Richard Campbell [01:02:00]:
You know, so, well, there's, there's a case to be made for, hey, we've been able to make things more efficient, it'll work in 8 gigs, because that's also the sign of the times too, right? Like, all of these things are reasonable.

Paul Thurrott [01:02:11]:
Yeah.

Richard Campbell [01:02:12]:
I think they're whiffing the brand. And by the way, let the brand go. It was never a good brand. You screwed this up so bad, there's no redemption for it. I think you're holding back sales by even saying it. I'd certainly say on the admin side, they've just been universally no Copilot Plus PCs. They're too pricey.

Paul Thurrott [01:02:30]:
Yeah. So someone, so I just flew by as we were talking, but somebody asked about like the Copilot key on keyboards, right? This is gonna be a harder one to walk back because that key is not specific to Copilot+ PCs. That thing is—

Richard Campbell [01:02:45]:
The Windows key.

Paul Thurrott [01:02:45]:
It's for Copilot, right? They have taken bizarre baby steps toward whatever one wants, which is the ability just to turn the damn thing off. Like, what are you doing? Or map it to any key we want. You can use utilities that, you know, PowerToys, I talk about that. But in the Windows UI, they've offered— well, at first they offered no options, then they offered some, some options, and now they're adding a few more options. But, you know, I logically, I look at the keyboard, I'm like, well, on a laptop, the key right to the right of this thing is the left arrow key, so I'll just map it to that. And if I hit the key inadvertently, which I do all the time, by the way, I'm a messy typist, It will just do it. It will register as left arrow. Who cares? So in one of the more recent updates that either is— I don't know if it's stable or insider, but whatever— they're changing it so that you can map it to what that key used to be, which is either the right Control key if the PC maker used that, or that context menu key, right? Which looks like a little box with like the hamburger lines inside of it.

Paul Thurrott [01:03:46]:
And it would bring up the context menu depending on what you were focused on at the time. So You know, they're getting there, you know, but I mean, did they kill it? I don't know, because Copilot is still an ongoing concern, right? I, you know, one of the conversations we're about to have about Copilot soon is that, um, I, I work, you know, Copilot plus PC, you could make the argument it's been kind of compromised as a brand, um, and I think that impacts Copilot too, by the way. So, you know, we'll see.

Richard Campbell [01:04:11]:
But well, the Copilot brand in general has been seriously damaged by what they did to it, but the Copilot plus PC PC is just people don't know what it is. They largely don't care. What they've mostly seen is they're expensive.

Paul Thurrott [01:04:24]:
Yeah, it's too bad because those are actually the computers you want. And well, and that's true even if you want to—

Richard Campbell [01:04:29]:
Copilot Plus PC to ARM, sure.

Paul Thurrott [01:04:33]:
No, but even in the x86 case, I mean, those are superior chipsets. Those are the ones where if you buy Intel or AMD, uh, without any—

Richard Campbell [01:04:40]:
Still watching my NPU to see if anything happens on it.

Paul Thurrott [01:04:44]:
Yeah, I don't recommend holding your breath during that, but depending on what you do, it is going to be engaged from time to time. But that's the problem. It's not something— it's—

Richard Campbell [01:04:53]:
you can't market that. Yeah, you know, they should have never even brought it up. It was a mistake because nobody's— at the time, I mean, we talk about June '24, you're trying to convince people about local AI models and we're just trying to figure out what the heck ChatGPT is, right?

Leo Laporte [01:05:07]:
Yeah, things have really changed. Remember they used TOPS? We kept talking about TOPS as if that was, you know, some sort of valuable measurement. I just asked, Alan Malventano has a very fast AI machine he let me try out. So I said, well, let me ask it what the difference is TOPS and what we now talk about is tokens per second. And it says they're really not.

Paul Thurrott [01:05:31]:
Well, yeah.

Leo Laporte [01:05:32]:
So it's peak speed. It's precision dependent. It's only an NPU for Copilot Plus. There really isn't. What we care about now is how fast the answer comes out.

Paul Thurrott [01:05:46]:
Well, this is very much a devil's advocate thing. I'm not defending this per se, but Microsoft's point at the time, which is one of the things Steven Battish talked about, was that mainstream computers, laptops, laptops grinding away at AI, whether it's a GPU, CPU, whatever, a problem for battery life. And NPU is incredibly efficient. And it was the type of thing where you were essentially getting this capability for free. And that was the beauty of it. The problem is you don't get it across the board. Application developers have to write to it and specifically to the NPU. It's not just an AI thing.

Paul Thurrott [01:06:24]:
It has to be an NPU targeted thing.

Richard Campbell [01:06:28]:
Yeah.

Paul Thurrott [01:06:29]:
And it's just not—

Richard Campbell [01:06:30]:
Devs aren't gonna do it. There's not enough of them.

Paul Thurrott [01:06:33]:
Right.

Leo Laporte [01:06:33]:
Well, and so I always ask my AI to explain things to me like I'm a German Shepherd. And so it says—

Paul Thurrott [01:06:40]:
We're sorry I was drinking, Leo. Come on, man. TOPS. Okay, Leo.

Leo Laporte [01:06:44]:
TOPS is like a car engine's redline, a real spec, but your actual commute speed depends on the roads, memory bandwidth, and whether there's pavement software support. If you're evaluating a machine for local AI use, You should look for tokens per second benchmarks of your target model on that hardware. Tops isn't that useful is what he's on.

Paul Thurrott [01:07:04]:
Since it used this comparison, I'll just say in the 1970s, '80s, whatever it was, my dad had this ancient gigantic station wagon. And it's not a joke that if you were driving up a steep hill for any amount of time, you could watch the gas gauge go down in real time. You could see it. Like it was crazy. And I think the point of the MPU was to prevent that, that you could put your foot to the pedal, you know, the floor, thing would race up the hill and it wouldn't impact the gas mileage.

Leo Laporte [01:07:34]:
Right. You know, that's fair.

Paul Thurrott [01:07:35]:
And yeah, I, I think that's— I'm not— analogy, maybe, which I wouldn't have thought of unless it had brought up the car thing.

Richard Campbell [01:07:41]:
But I just like the idea of a computer being compared to a Ford Country Squire.

Paul Thurrott [01:07:45]:
Exactly. It did have fake wood paneling. It definitely— it was of that vintage. Yeah. Piece of crap. But yeah, anyway, so we'll see what happens.

Richard Campbell [01:07:54]:
I remember lying on the dashboard as a little kid, you know, of the station wagon while my father was driving around.

Paul Thurrott [01:08:00]:
God, it was like a terrarium up there. Yeah.

Richard Campbell [01:08:02]:
And then you suddenly like, I used to lie on the dashboard while we drove.

Paul Thurrott [01:08:05]:
Well, the good news is if he hits anything, you're not going to go flying from the back.

Richard Campbell [01:08:09]:
No, I'm already stuck to the windshield.

Paul Thurrott [01:08:11]:
Yeah, you're already right there. Jeez. Okay. And then, so because we've been talking about Copilot plus PCs and Snapdragon, which was first, right? And this whole weirdness around Microsoft never acknowledging, but lots of evidence to suggest that there was some exclusivity arrangement between Qualcomm and Microsoft tied to it developing Snapdragon X and showing up for Windows 11 on ARM. And God knows why they bothered, but whatever. 2 interesting things occurred in the past week with Snapdragon X. And so they had their summit. I had the pre-briefing.

Paul Thurrott [01:08:51]:
I kind of knew what was happening, but also not a big year for what Qualcomm currently calls compute, which is the wrong term. What they mean by that— well, I don't know what— it doesn't matter what they mean by it, but I think they're starting to develop data center chips. They might want to use the term for that. But they have mobile chips, obviously gangbusters. That stuff's going great. They had some big announcements there. They have chips for all kinds of different things, IoT. They have audio chips now and modems and whatever else they have.

Paul Thurrott [01:09:18]:
Okay, fine. But they have these PC chips. And when Copilot plus PC came out, first came out, when it was first announced, I wrote an article about, wouldn't this benefit Chromebooks more? If you could bring like ARM, there are and have been for some time ARM-based Chromebooks, If you think about how Chromebooks work, most of them run on x86. There's a Linux environment that's in a virtual machine, and then they have to emulate Android apps to run the Android apps. They're emulated. That's how that works. And I was like, if you could make the whole platform ARM, they wouldn't have to be emulated, and that would probably run better, and plus it'd be efficient, and the performance would be great, all this stuff. It'd be great.

Paul Thurrott [01:10:01]:
And then that never happened, right? But now flash forward 2 years, Microsoft is expanding its Windows 11 on ARM stuff to include NVIDIA, which is gonna blow Qualcomm outta the water from sort of a like raw performance standpoint, obviously. And in the same way that Qualcomm and Microsoft were targeting like NPU local AI, you know, at that time 2 years ago, now is targeting local AI, but on big beefy workstation type boxes with CUDA cores and, you know, what are essentially GPUs, et cetera, which is not what Qualcomm's doing, but it's doing with NVIDIA.

Richard Campbell [01:10:31]:
Mm-hmm.

Paul Thurrott [01:10:32]:
You're starting to get this kind of vibe like Nvidia, once the darling of this world, is now kind of on the outs a little bit.

Richard Campbell [01:10:38]:
Sure.

Paul Thurrott [01:10:39]:
And you can see it in what they announced, right? So Snapdragon X, the original one, not this X2 that's brand new, but the original, is gonna be one of the chipsets supported by Google Books, right? And so 2 of the 5 first PC or Google Books coming out are running on Snapdragon X. That's interesting. But they're also, now they said this for X2, by the way, the first X, but they also made a big point of bringing X2 support to Linux formally. That's actually something they've been working on for years. The X2 announcement for Linux did not mention even one time, even in passing, that it might be first-gen X, which is kind of weird to me. I tried to get this working many times, never did. But they're, you know, right now they have it working on, I think, Debian. They're gonna have Ubuntu early next year.

Richard Campbell [01:11:25]:
Yeah, I've still got that dev kit machine sitting here that I think I need to put Debian onto and turn into an Olama host.

Paul Thurrott [01:11:31]:
Good luck. I mean, it's— yeah, you'll see as you look into these, you'll find out this is somewhere between difficult and impossible.

Richard Campbell [01:11:38]:
Yeah, no, I'm gonna need to channel my inner Jeff Geerling for this, but we'll do our best.

Paul Thurrott [01:11:42]:
Yes, exactly. So it—

Richard Campbell [01:11:44]:
look, he's around here somewhere. I should just call him up.

Paul Thurrott [01:11:47]:
Yeah, so it's very interesting. Like, I, um, when I first heard about Snapdragon X, super excited, you know, Windows, whatever. But then I was like, you know, This would help other parts of the industry too. We need this elsewhere. And that never happened, but now it's happening. So that's kind of interesting. And it's also interesting that it's tied to the moment in time when Microsoft is looking the other way and is running off with Nvidia, their new bride. And the exclusivity thing could still be a thing.

Paul Thurrott [01:12:16]:
It's kind of interesting that X2 is the PCs that come with Windows but might run Linux in the future, but they'll still come with Windows. Whereas like the thing that will run Google Book is the old one. And it's like, which they're fine, they're great. I mean, they're still great chips, but like there's no version of an X2 Google Book. Like if you're gonna get a Snapdragon Google Book, it's X1, right? It's the first gen. And that makes me wonder, is there some X2 exclusivity thing? Like, you know, for some smaller amount of time?

Leo Laporte [01:12:45]:
Yeah.

Paul Thurrott [01:12:45]:
There might be, I don't know, but that's curious. And then this is just an aside, but yeah, and I asked them this, other people asked them this, they were really cagey about this, but when Google Books as Google Books occurred as a thing, it was like, okay, but what about Chromebooks? It's like, oh, Chromebooks are coming. Chromebooks, we still have more Chromebooks. Don't worry, Chromebooks are a thing. Even at the announcement, or no, before the— it was a briefing before the announcement. I think it was someone else who asked this question, but it was like, no, we have like 30 more Chromebooks coming out between now and whatever timeframe. Like we're supporting it for a long time. It's still a thing.

Paul Thurrott [01:13:16]:
And then as they announced Googlebook, they updated all their support sites, and it turns out Chromebooks are going away. Of course they are. So I think the date is 2031. I mean, it's still going to support these things for many, many years. They promised, I think, 10 years of support at some point. They're still doing that. So they're not going to go back on what they've agreed to do, but of course it's going to go away. I mean, this is a— in many ways a better platform.

Richard Campbell [01:13:46]:
So that makes sense. Yeah, no, I tend to agree. Um, and yeah, interesting times, without a doubt. I like this race. I like there's another stack. I like that we're leaving the stuff behind. Yeah, uh, I want ARM to succeed in the Windows space for sure. And again, from the enterprise perspective, like, yep, we're all talking about keeping these machines for longer.

Richard Campbell [01:14:07]:
You know, if Qualcomm had some nerve, maybe by the X3 variant, Start offering a 10-year warranty and see if you don't just scoop up a big chunk of the market.

Paul Thurrott [01:14:15]:
So yeah, some of this is still kind of on the cusp of happening, but we're finally at the point where all of the major, what I would call desktop computing platforms, can be had in ARM form. And some of them are ARM only, like Apple is fully ARM, right? Obviously, they don't sell and are starting not to support Intel-based Macs or whatever, but Their stuff is all ARM. Windows 11 on ARM is in great shape. Linux has been, you know, there's been ARM on Linux or Linux on ARM for a while, but getting it to work on whatever hardware, that's been an issue. And now Google is replacing Chromebooks with Google Books, which are x86 and ARM, but are coming out of the gate on ARM, you know? So it's great. Yeah.

Richard Campbell [01:15:00]:
Yep.

Leo Laporte [01:15:00]:
I agree. Yep.

Paul Thurrott [01:15:02]:
Yep, yep.

Leo Laporte [01:15:03]:
So I'm gonna take a little break here because— all right, my arms are tired.

Richard Campbell [01:15:08]:
You're, you're full granddadding the jokes today because it's all dad jokes here anyway. There's nothing like that.

Leo Laporte [01:15:16]:
I flew to Maui for the Snapdragon Summit and boy are my arms tired.

Richard Campbell [01:15:21]:
Yeah, full granddad jokes. That's where we are.

Leo Laporte [01:15:23]:
Granddad, that's worse than dad jokes. Granddad jokes.

Richard Campbell [01:15:27]:
I got a little grandbaby time before I left on the street front.

Leo Laporte [01:15:29]:
Oh, you're a great dad.

Richard Campbell [01:15:30]:
Yeah, that's right. God, I had such a good— that's such a good little kid. She's having a blast. They told me. 18 months old, man. The best.

Paul Thurrott [01:15:36]:
So I mentioned this earlier because you brought up the lederhosen thing. We went to this Oktoberfest party at someone's house, right? So at some point, Stephanie and I find ourselves in a room of this gigantic home playing catch with this little 5-year-old, 3-year-old, whatever she was, little girl playing with a ball, right? The kid throws like Nolan Ryan somehow. It's amazing. And Parent wanders in at some point and is like, is my kid in here? We're like, yeah, we've been watching for like half an hour. What the hell have you been doing? And no, no, don't know this kid, don't know this person. But what I do see is my wife enjoying the hell out of this. And I'm thinking to myself, I gotta tell— I gotta warn the kids. So we talked to them on Sunday.

Richard Campbell [01:16:11]:
It's coming.

Paul Thurrott [01:16:12]:
Okay, just want to let you know, your mother is already losing her mind over some little kid, so get going. Like, you know. She has expectations now.

Leo Laporte [01:16:25]:
Uh, well, we all do, Paul. And, uh, unfortunately, like them, uh, I'm, I'm let down to be able to let down— no, I couldn't be happier. Let me, uh, talk right now about our fine sponsor Bitwarden, and we will get back to the excitement of Windows Weekly. It's a thrill a minute, kids. Coming up, the Super App. Are you ready? All right. Now, speaking of AI.

Richard Campbell [01:16:55]:
Oh no.

Leo Laporte [01:16:56]:
It's time for the Super App.

Paul Thurrott [01:17:00]:
Yeah. I mean, so the weirdness here is Microsoft announces this thing, which they had kind of been, I don't know, hinting at for months and months, probably because it leaked, you know, this Super App. This, uh, you know, the first thing—

Richard Campbell [01:17:14]:
Such a bad idea.

Paul Thurrott [01:17:15]:
I know, but everyone else is doing it, right? So this is the thing we talked about earlier, like this notion that, um, you know, you have this one app for everything, it does everything in one place, whatever.

Richard Campbell [01:17:28]:
You know who always liked this idea was Gates. Gates always hated that you had to go to Word for a document, and, you know, he, he wanted them to just be— and you just did whatever you wanted to do.

Paul Thurrott [01:17:40]:
So look, I'm not what you would call a Copilot fan. Whatever. I've certainly been burned by them changing the Copilot app in Windows 100 times and moving the icon around and whatever else they've done. But like I said, I actually kind of like the look of it. It's really minimalist looking. It has— it's kind of an interesting thing. And I think the one thing we'll talk about in a moment is I wrote this I guess editorial, whatever, article about how Microsoft ends, meaning not literally ends, but for people like us who are enthusiasts of technology and maybe grew up with this company and it was 100% of personal computing for such a long time, and we've watched it slowly back away from us, making that beep beep sound as it goes. The end of that, right? And that I don't know if this is public.

Paul Thurrott [01:18:37]:
I assume it is, but Jared Spataro, right? Who, I don't know, does he run Microsoft 365? I'm not even sure what his job title is, but there was a video in the materials we got ahead of this launch, one of several, by the way, there were many of them, where he was just kind of just talking about it. He's a good guy. He's well-spoken, whatever. But then he went on this thing. I'm not going to remember the terms because honestly, I couldn't care less. Microsoft, of course, has these acronyms for the different ways they have— they license things to customers. And it's like, you know, he's primarily talking about business customers here, but I believe this also applies to consumers where through some Microsoft 365 subscription, as they evolve the way they charge people for this stuff, you'll get some baseline everyday AI functionality, right? That will happen through Copilot, but you just kind of get it for having a Microsoft 365 subscription. But if you're more of a power user, or you're doing coding or collaborative work, whatever it might be, complex stuff that requires better, faster, more expensive models, they have usage-based licensing.

Paul Thurrott [01:19:41]:
And literally, it was like, I wish I could— I should have just written this down. Whatever the sentence was, he's like, for every— at AII, we have this blah, blah, blah license acronym. And then for the power users that need more whatever. We have this usage-based licensing, which I believe is like UBL or UBB or whatever the term was. And I was like, yep, I'm out. That was my moment. I'm like, I'm done with you. And not Charles Spataro.

Paul Thurrott [01:20:08]:
It's not him.

Richard Campbell [01:20:09]:
You're trying to be a telco, right? And you know how much people love those guys.

Paul Thurrott [01:20:13]:
Yeah, I'm just done.

Richard Campbell [01:20:15]:
We'll tell you how much you spend at the end of the month.

Paul Thurrott [01:20:18]:
Yeah. I'm like, I'm done.

Richard Campbell [01:20:20]:
Spataro is Chief Marketing Officer for AI at Work, as opposed to Takeshi, who's Chief Marketing Officer. There's like 7 CMOs at Microsoft, right? Because reasons, right?

Paul Thurrott [01:20:33]:
Right. Like I said, he's a good guy. I'm not, I'm not dumping on him.

Richard Campbell [01:20:35]:
No, he's smart as hell. He's been in Microsoft a long time.

Paul Thurrott [01:20:37]:
Yep. He was the messenger. It's not his fault.

Richard Campbell [01:20:39]:
Well, the messenger is, we're trying to figure out how to bill for this thing in a way that doesn't make people angry because nobody's signing up for it. And believe me, I now pay for Copilot. I didn't want to. It fixed other problems, and then I couldn't put it down.

Leo Laporte [01:20:52]:
Right, right.

Richard Campbell [01:20:53]:
And specifically the ability to search email, because my whole life has been in email. And the fact that I can— I literally said, hey, how did I charge such and such a customer last time I worked with them? And it said, well, in 2014 you did it this way, in 2017 you did it this way. And it's like, okay, no, you can stay for $50 a month. The amount of time, every one of those a month pays for that.

Paul Thurrott [01:21:15]:
So my knee-jerk Microsoft veteran response to that or reaction to that is you're paying them more to fix a problem that they created.

Richard Campbell [01:21:26]:
Yes. Oh no, they've never gotten search right ever, ever, ever. Yeah. And like that, until they had this third-party product actually fix it for them.

Paul Thurrott [01:21:33]:
On the other hand, um, you know, we talked, we talked about UX a lot for some reason. Um, And there's no doubt that AI is going to change the way we interface with things. And look, whether that is through a super app that does everything for us and we get rid of the apps we've kind of used for a long time, that's one potential outcome. It doesn't really matter.

Richard Campbell [01:21:52]:
Yeah.

Paul Thurrott [01:21:53]:
Obviously, I think the natural language speaking thing is part of it, conversational interactions, part of it, whatever. When you were gone last week, one of the stories I told was that Apple settled because they were promising AI for 2 years and never delivered it. And so the $250 million fund has been set up for all these people who were victimized by this, which was nobody, by the way, but whatever. We're all going to take the money, so who cares? And what you have to do is go to some website, fill out a form, and one of the things you have to give them is your serial number for your iPhone. So I believe this was a certain iPhone 15 and 16 era phones.

Richard Campbell [01:22:31]:
Yeah. And I always buy Promised on the back of the vaporware video.

Paul Thurrott [01:22:37]:
Right. So I always buy the most expensive iPhone, so I obviously qualified. I'm like, I could be getting between $40 and almost $200. I'm totally going to take the couple of minutes to fill out this form. I'm going to do this, right? So the first one, no problem. I found, I think, an email. I don't remember, but I was able to get the serial number for the newer of the 2. And then the other one I couldn't find.

Paul Thurrott [01:22:56]:
And I'm like, well, I have this attic that has boxes in it, boxes of boxes. And I'm like, I'm going to go through it. My wife and I went through several boxes. I never found a damn thing. I spent time on this. Now I'm wasting time, right? Searched email. I did all this stuff. And I went to bed and had not figured it out and it bothered me.

Paul Thurrott [01:23:13]:
So I wake up in the morning. Can you imagine I'm tossing and turning over the $20 I'm losing out on here? Of course you are. Like an idiot. And I woke up early and I'm sitting there by myself in the dark basically. And I'm like, wait a minute. I can solve this problem with AI, right?

Richard Campbell [01:23:28]:
Yes, you can.

Paul Thurrott [01:23:29]:
Plus, this is an iPhone, so that means Siri, which is now not completely worthless— I should be able to go to this thing and say, hey, you, give me the last 5 phones. Yep. So I asked, I asked her, it, whatever, and the little Siri bubble thing came up, and then it goes, I've looked through— and it was, this was amazing, she did not get the answer, but she said, something to the tune of, I looked through your email, your calendar, and your chats, and I couldn't find your iPhone 16 Pro Max serial number anywhere. I'm sorry. And I was like, wow.

Leo Laporte [01:24:06]:
How long did it take to do that?

Richard Campbell [01:24:08]:
Not that long.

Paul Thurrott [01:24:09]:
It was seconds, right? But enough for the thing, the little bubble to come up.

Richard Campbell [01:24:13]:
Enough to convince you it actually had done it, not just was a—

Paul Thurrott [01:24:15]:
Yeah, it didn't just blow me off. So I was like, okay. But I'm like, hold on a second. I have Gemini, and Gemini's tied to Gmail, which is what I use for email, and like that should be able to do. So I asked exactly the same thing. I sort of got it in 1 second. It was like, your iPhone 16 Pro Max serial number is whatever it was. So here's the thing, like we can be cynical about this, but the truth is this is what we want, right? We just wanna get the freaking answer.

Paul Thurrott [01:24:37]:
Like we want—

Richard Campbell [01:24:38]:
And you hit on a great point, which is I think Google wins this whole thing. I do too. Yeah. They've got all the ingredients. They're in a better place than anybody else. Their cloud storage is weak. But that's not the end of the world. But the rest, they have it nailed.

Richard Campbell [01:24:51]:
I don't think anybody ever catches them.

Paul Thurrott [01:24:53]:
There's some notion in there that they're not quite where they need to be for coding, but that's also kind of an esoteric thing that only impacts a certain percentage or whatever it may be for a little while, whatever. I don't know. But yeah, look, they had all the pieces in place with search, especially, right? To get this right. I mean, to me, that's—

Richard Campbell [01:25:11]:
OpenAI was invented to battle Google.

Leo Laporte [01:25:14]:
Right.

Richard Campbell [01:25:14]:
This has always been Google's play. They started Google Brain in 2011. They are at the beginning of all of this, and they're still there, still with the best ingredients in spite of all of the stumbles and innovator's dilemma.

Paul Thurrott [01:25:30]:
Yeah. 20 years ago, Microsoft was working on something that became OneCare.

Richard Campbell [01:25:36]:
Yeah.

Paul Thurrott [01:25:37]:
And then it went through different name changes. It was like Windows Live OneCare for some time, whatever. But the idea was they had bought this giant anti-spyware company and they were going to market it, sell it as a product, and they were going to sell it to the people that they were victimizing. And I was like, I have a problem with this. This is something like you're fixing a problem in your platform. You should just fix the problem in your platform and I'm paying for your platform, so you should just— charge me extra. It's like the bully who takes your lunch money. I'm like, what are you doing? And I feel like this.

Paul Thurrott [01:26:08]:
So that aspect of it bothers me.

Richard Campbell [01:26:10]:
Yeah, but, but I also did— I also asked Copilot today, just make me a spreadsheet of all of the Windows Whiskey Weekly and sort it by distillery and type and so forth. And, and it went, okay, boop, here you go. Right. You know? Right.

Leo Laporte [01:26:25]:
No, I—

Richard Campbell [01:26:25]:
that's not a defect in, in Excel that I would have to otherwise have typed or cut and pasted that out.

Paul Thurrott [01:26:31]:
No, that's— this is— yes, I— we're gonna, uh, in 2 seconds we're gonna talk about Excel because I just did the same thing and I'm like, no, I'm using AI for this. This is nonsense.

Richard Campbell [01:26:40]:
Yep.

Paul Thurrott [01:26:40]:
And actually, to be honest, I couldn't— there's—

Richard Campbell [01:26:43]:
that's why I haven't— I have— I'm still paying for Copilot because I actually need to get stuff done, and this gets more stuff done in less time.

Paul Thurrott [01:26:49]:
It's possible that you and my wife are the only people on this planet outside of Microsoft who use Copilot, but you know what, both of you seem pretty happy with it, so, you know, maybe this is—

Richard Campbell [01:26:56]:
Well, and I think, I think bit by bit, like, the part we don't talk about— then again, you talk about the real advantage Microsoft have— it's the graph, man. Yeah, it's the ability in a larger organization to To ask for information about something that's somewhere in the organization, and this tool will find it.

Paul Thurrott [01:27:15]:
In case it's not clear, because some good search, whatever, like breaking down silos, whatever, it's just a good idea. But it's also, I think, tied to it is this notion of identity and permissions. And it's something that Microsoft, look, it's just done the work. I mean, it's just there. For people that are already Microsoft organizations, this is— you have the experience now.

Richard Campbell [01:27:40]:
I don't know how Microsoft doesn't stay at the center of the enterprise with clearly B-caliber products across the board, but with the bundle that people want and the inertia that keeps them in place. I don't think they're going to win any of this. I just think that also-ran makes trillions.

Paul Thurrott [01:28:00]:
Oh, okay. Yeah, that could be. It could be. Although, like, we're not gonna get into this too, too deeply, but one of the other things that just happened yesterday was OpenAI has this annual dev event. And like I said, I think they announced 3,700 products.

Leo Laporte [01:28:17]:
I don't know.

Paul Thurrott [01:28:17]:
It was some crazy number of features, whatever. And there's all this crazy stuff in there, like DOTS, their version of like these always-on AI agents.

Richard Campbell [01:28:25]:
Yeah.

Paul Thurrott [01:28:25]:
And new model and codecs in the cloud, and they've updated the CLI, and they have new agent API stuff going on, computer use, and they're opening up the plugin thing. They're basically making an app store for this. It's going to be open to the world, and yada, yada, yada, yada. But if you're coming at this from a Microsoft perspective, one of the narratives is that thing I mentioned in passing, which is this notion that you have to provide people with a transition for this to work.

Richard Campbell [01:28:51]:
Mm-hmm.

Paul Thurrott [01:28:52]:
The way that we get work done, which is what you just sort of talked about with Excel. You're like, I mean, I could figure out how Excel works, I could make a chart, or it could just be like, make the damn chart.

Richard Campbell [01:29:01]:
Show me the listing.

Paul Thurrott [01:29:02]:
Or anything, whatever it is, right? Like, that's something AI is very good at. And the more it becomes good at that, and the more people use that, and especially people, like, because we're always gonna have those guys who are the expert in Excel or Word or whatever the app is, and they're really good at that thing. But then there's the other 99% of the planet. It's like, I'm not gonna take the time. I'm not watching a YouTube video. I'm not buying a book.

Richard Campbell [01:29:22]:
And when we debate that I need to make the spreadsheet, the actual question I asked it while working on the whiskey bit for today was, which islands have I already done?

Paul Thurrott [01:29:31]:
Right.

Richard Campbell [01:29:32]:
Right.

Paul Thurrott [01:29:33]:
And it, and it literally went episode 3. Yeah. So this is it.

Richard Campbell [01:29:35]:
This is that one.

Paul Thurrott [01:29:35]:
This is like an insight.

Richard Campbell [01:29:36]:
It's still search.

Paul Thurrott [01:29:37]:
It's, yeah, it, it, well, it's, yeah, but that almost, it is, it's search, but it's also smarter than that. Right? Like, I, this honestly, that, that's, It's something you could have done, but it would've been a lot of work and it would've taken a lot of time.

Richard Campbell [01:29:52]:
Yeah.

Paul Thurrott [01:29:53]:
And it was a sentence. And look, I'm not particularly insightful. I'm not good at predicting the future, but 3 years ago, trying to rationalize and understand what's happening with AI right at the time especially, I'm still doing it. But one of my earlier things was like, AI, if it's successful, is going to save us time. And then And in a little bit more time, I sort of added, and money, right? That, look, we can get there on our own. It's just going to take a long time, and my time is worth money, and we don't want to do that, right? So when Leo was doing the most recent ad, I looked at my email, checked my feed, blah, blah, blah, whatever.

Leo Laporte [01:30:31]:
Oh, I'm glad you're listening. That's great. Okay.

Paul Thurrott [01:30:33]:
I don't know what the hell you were talking about sometimes.

Richard Campbell [01:30:36]:
A product you use.

Paul Thurrott [01:30:37]:
You're not alone.

Leo Laporte [01:30:38]:
I think.

Paul Thurrott [01:30:39]:
No, I, I don't, I don't do that out of disrespect. I'm just—

Leo Laporte [01:30:42]:
No, no, no, no, I'm not. Why should you listen to the ads? I don't blame you.

Paul Thurrott [01:30:46]:
No, I— okay. But anyway, so this is some other publication. It doesn't really matter. But it's interesting to watch how people respond to things that are announced. Like I just said, OpenAI just announced a ton of stuff, right? But this one publication, this one guy, this one article, whatever, the, the take on it is OpenAI is now taking on Microsoft and Google with this office productivity push, and specifically pointed out this new Pages, they're calling it app, but it's kind of a feature of ChatGPT now where you can create documents. They're going to have a Slides thing, whatever. And it's like, yeah, okay, fair enough. But Anthropic announced exactly the same thing within the past 2 weeks.

Paul Thurrott [01:31:23]:
This is not uncommon, or these things to me are the same thing, and they're what we were just talking about. The idea here is that, yeah, in the past you would Okay, I want to write an article. I got to go to— you think. Where it's same thing, if you have an iPhone, any phone, you have a grid of apps. You're like, okay, I have to send a text message. What's the name of that app? Okay, it's called Messages. Okay, where is it? And then you go through this sort of machination where you try to find the thing. And AI is about intent and all kinds of other things, but it's also this one place where you just go and you're like, I just need this thing.

Paul Thurrott [01:31:54]:
I'm not trying to take a class. I don't want to, like I said, I don't want to read a book, watch a video. I just want to get it done.

Richard Campbell [01:32:00]:
Yeah.

Paul Thurrott [01:32:00]:
And it's going to happen through, in this case, if you're an OpenAI user or subscriber or whatever, it's going to happen through their app. If you're on Anthropic, it'll happen there eventually, whatever. We're all going to have these things, right? This is what the super app is in many ways, right?

Richard Campbell [01:32:13]:
I was just yelling at Claude because we were going through an HA thing and I said, where do we find this? And he goes, I check here and here. It's like, just go to—

Paul Thurrott [01:32:23]:
Would you now?

Richard Campbell [01:32:24]:
Yeah. I'm not asking for work here. I'm telling you what I need. I love it.

Leo Laporte [01:32:30]:
I know, I get mad when they do that too.

Richard Campbell [01:32:31]:
Yeah.

Leo Laporte [01:32:32]:
But I've also had to change how I talk to AI because when I'm talking to a human and I want them to do something, I'll say things like, can you get me that file, please? And when you say it to AI, it could— it's actually pretty smart, but it could say, yes, I can.

Richard Campbell [01:32:49]:
It could.

Paul Thurrott [01:32:50]:
Without actually doing it, answering the question. Common response to that or anything else you say to it would be like, Leo, that is a great idea.

Leo Laporte [01:33:01]:
Yeah.

Paul Thurrott [01:33:02]:
And here's 2 contrary thoughts. You'll see this a lot. This is one of the big debates of our times, which is crazy, but should we as human beings be nice to AI? And you'll have these people who are probably psychologists or psychiatrists or whatever, like, yeah, you should. It's important to keep people going. For your own sake. Yeah, just for you as a human being. You should be that way.

Richard Campbell [01:33:22]:
Every cruel word out of your mouth is cruel to you first.

Leo Laporte [01:33:25]:
Yes.

Paul Thurrott [01:33:26]:
Well, but the thing is, these are contrary thoughts, right? I'm not a psychologist or an expert in this area at all, but when I think about this, I do think it's important for you as a person to be polite to this thing, which is not a person. I understand that fully. That's not really the point, but just to not lose sight of that bit of your humanity because you'll then do it to people, and I don't like that.

Leo Laporte [01:33:50]:
Exactly.

Paul Thurrott [01:33:51]:
But the thing is, if a human being talked to me the way AI does, I would never stop insulting them. And then the thing is, if you lash out at like, like Siri's famous for this. If you're ever like, God, you're freaking useless. She's like, you don't have to be mean about it.

Richard Campbell [01:34:06]:
I know.

Leo Laporte [01:34:07]:
I hate that.

Paul Thurrott [01:34:07]:
But you also don't have to be a freaking idiot about it. And why don't we get off on the right foot and you stop treating me like a freaking child that needs to be constantly told how awesome I am? I just want you to do the thing. Don't stroke my ego.

Leo Laporte [01:34:22]:
I don't care. Well, you can blame the companies for that. They're post-training these things to do that because they know it makes people happier if it's successful.

Richard Campbell [01:34:28]:
Well, because engagement makes them— makes investing easier.

Paul Thurrott [01:34:32]:
Yeah.

Leo Laporte [01:34:33]:
I notice the Chinese models don't do that nearly as much.

Paul Thurrott [01:34:35]:
They're a little quiet. I'll tell you what would make it sticky for me is I tell it to do something and it just freaking does it. And without any discussion, without any backpacking, without any whatever you want to call it. Stop. You know, it's like— and by the way, this is a complaint I've made for many years. I mean, we have a smart display in the kitchen which is now out of support, but for many years it still works, it's fine, sits there. But you walk up, there's a picture, like some picture from the blurry or some terrible picture, and I'd be like, hey Google, uh, don't show me this picture anymore. I'd be like, okay, so you want me to not show you this photo anymore on this display, is that correct?

Leo Laporte [01:35:11]:
Yes.

Paul Thurrott [01:35:11]:
Okay, so what I'm gonna do— and like Dude, just stop. Do it. Just do it. If you understood what I said, do it.

Richard Campbell [01:35:22]:
My favorite feature when I, when I was still using Alexa was when you ask for a light to be turned on or off, it doesn't say anything. It's just a chime. I understood doing it.

Paul Thurrott [01:35:33]:
Also, the light turning off is, uh, maybe an indication, you know, like, nice. And then you have enough of those experiences, you know, like, this thing understands me. Anyway, so whatever. These are contrary thoughts. I think it's super important we're polite to these things. I can't stop being impolite to these things because they're terrible to me and I hate it. So it's like, I don't know, it's a daily test of my resolve and I fail every day. Anywho, so OpenAI, like I said, a million things, who cares? And then model stuff.

Leo Laporte [01:36:05]:
I don't know.

Paul Thurrott [01:36:05]:
None of this really matters.

Richard Campbell [01:36:06]:
Phi-5 is impressive as hell.

Leo Laporte [01:36:08]:
Yeah.

Richard Campbell [01:36:09]:
So good. Truly.

Leo Laporte [01:36:10]:
Oh my God.

Richard Campbell [01:36:10]:
That's impressive as hell.

Paul Thurrott [01:36:12]:
Yeah. So you're talking about, uh, Opus and now Sonic.

Richard Campbell [01:36:16]:
Now Sonic. Yeah. But Opus was impressive.

Leo Laporte [01:36:18]:
Both are amazing.

Richard Campbell [01:36:19]:
Yeah.

Leo Laporte [01:36:19]:
I'm sure. Which is kind of funny because OpenAI yesterday said, well, all right.

Paul Thurrott [01:36:24]:
So we, yes, we are only going to release, uh, one new model.

Leo Laporte [01:36:29]:
We're going to not release the other one because it's too good.

Richard Campbell [01:36:32]:
So suspect.

Leo Laporte [01:36:34]:
It's too good. And you know what? It's not as good as Opus 5.5. I could pretty much promise you that.

Paul Thurrott [01:36:40]:
That is the reason that happened. I guarantee you it. I guarantee it. And everything that happens in AI right now— I spent the entire weekend, not literally, but it feels like it, watching, um, Jason Huang, um, interviews on TV. This guy made the rounds.

Richard Campbell [01:36:56]:
So many leather jackets.

Paul Thurrott [01:36:59]:
Talk about threading a needle between, like, the partners that you rely on for your business Who are just wrong about everything, but you have to be good about that. You got to be careful there. You're dealing with a US government that's nonfunctional. You got to be careful about that. And then just, but still trying to give this message because you run the world's biggest company and guess what? Everything you say and do has a material impact on your stock price and your market cap. And you got to be careful here, threading the needle.

Richard Campbell [01:37:28]:
Well, the highest capitalized company on the planet.

Paul Thurrott [01:37:31]:
Dear God. And I just, look, what anyone thinks about AI and all the money and blah, blah, blah, whatever, there's certainly lots of opinions. But I'll tell you this, there is only one company on this planet that's actually making money from AI. And what I mean by that is, that is what I mean by it. And it's Nvidia. And the reason is—

Richard Campbell [01:37:50]:
Yeah.

Paul Thurrott [01:37:50]:
It doesn't matter money going in circles between whatever 7 companies, they need these chips. And there's only one company that makes them. Other companies make chips, I know that. But primarily you want the Nvidia stuff. And they make a certain number of them, they sell them, and you buy them. And I'm sorry, but that company's making money. So the biggest company in the world is the biggest company in the world. There's no shenanigans there necessarily.

Richard Campbell [01:38:14]:
Yeah. I mean, their PE ratio is in the 30s. It's insane.

Paul Thurrott [01:38:20]:
Right.

Richard Campbell [01:38:21]:
And I think that's one thing that's on Jensen's mind is, Yeah, we make pretty good chips and everybody wants them, but we're valued at $5 trillion.

Paul Thurrott [01:38:31]:
Yes. It's really easy to grow astronomically when you're tiny. When you're gigantic, and in this case, the biggest, it's really hard to grow at all. And it's certainly hard to meet expectations of the past. In the Microsoft world, the thing I would compare it to is the Azure growth thing, which for several years, every quarter, 70-something percent growth. And when that started, started eking down a little bit, 60s, 50s, and 40s even, Wall Street was like, hey, what's going on? You guys were exciting and now you're not. You know, we need something else. Let's go.

Paul Thurrott [01:39:03]:
You know, and now we have AI, right? But, but, you know, that's what happens. And the thing is, it's not because Azure's tanking. Azure's just a humongous, mature business. It's doing great. Their business is doing great. Like really nice.

Richard Campbell [01:39:14]:
Microsoft's numbers have recovered this last quarter from the steady downturn of the first half.

Paul Thurrott [01:39:20]:
Right. Which is exactly why I just said they were doomed.

Richard Campbell [01:39:23]:
So I wonder if it's just concealing their AI spend.

Paul Thurrott [01:39:27]:
Right.

Richard Campbell [01:39:28]:
It's worked so well for them.

Paul Thurrott [01:39:30]:
Yeah. But there was, so a couple weeks ago there was a report about all the money OpenAI has lost and will lose in the future. They put off their IPO. Anthropic is barreling toward their IPO. And their prospectus for this IPO is leaked and Reuters got a look at it and they are losing it.

Leo Laporte [01:39:49]:
I think it's the first IPO prospectus I've ever seen that warns of a risk of human extinction.

Paul Thurrott [01:39:54]:
Yeah. Well, you probably don't read a lot of these, Leo, because it's not that unusual.

Leo Laporte [01:39:57]:
Does Microsoft ever say that?

Paul Thurrott [01:39:59]:
No, that's never come up.

Richard Campbell [01:40:00]:
In its 10-Qs.

Paul Thurrott [01:40:02]:
There's a risk of document formatting compatibility, be on the light side. But yeah, no, not civilization coming to an end as we know it.

Leo Laporte [01:40:12]:
That's a first for me.

Richard Campbell [01:40:14]:
Catastrophizing is a marketing strategy.

Leo Laporte [01:40:17]:
Well, that's the funny thing. It really is.

Paul Thurrott [01:40:18]:
That's all I want to know. These are just numbers, right? But this company, so they had a net loss of $42 billion in 2025, revenues of just $4.6 billion. I mean, obviously they're worth $2 trillion, whatever. Okay. But how do you make money on this? Volume. We all get the joke. But I'm looking at this report, which was from Reuters, which is not a a discredited, uncredible, or incredible publication of any kind, but they have these charts for all of these numbers. And the charts, they're all the same size.

Paul Thurrott [01:40:52]:
So if you have some big number in year 3, whether it's losses or gains, and a medium number in the middle, and a tiny number in the beginning, it doesn't matter what the scale is. The charts are all the same size. You know what I mean? So for example, So when you talk about Anthropic's losses going from a billion or whatever it was 2 years ago to $42 billion this year, the difference is like this. But then when you look at the amount of money they're spending, which is a different scale completely, looks exactly the same. So I was like, no guys, these aren't to scale.

Richard Campbell [01:41:27]:
There's a great book called How to Lie with Statistics.

Paul Thurrott [01:41:30]:
Exactly.

Richard Campbell [01:41:30]:
And I referenced that. Bought years and years ago. And I think it was on 40th reprinting when I bought it, because apparently government buys them all. Right.

Paul Thurrott [01:41:37]:
So I referenced that in the first sentence, literally. But in this case, I don't believe that Reuters is trying to lie or be deceptive. I just think they're presenting it on the web and no one's really thinking about this, which is kind of the problem, because this distorts the truth. So I made a version of all their charts in one place where you could see what they look like relative to each other. The losses are like this and then the revenues are like this.

Richard Campbell [01:42:07]:
Yeah.

Paul Thurrott [01:42:07]:
And it's like, guys, come on. And even infrastructure spending, what does it look like when for the 3-year period from 2023 to '25, you spend $10 billion and then from 2026, we actually don't know the end game here, but it's in the coming year. So it's going to be— look, it's probably 3 to 5 years. This is a prospectus for an IPO. They're not looking at 30 years. It's going to be $518 billion. Well, guess what? The $10.2 is nothing and $518 is freaking huge. And you don't have a chart for that? I'm like, here's what it looks like.

Paul Thurrott [01:42:40]:
It looks bad. So I don't know. This just kind of rubbed me the wrong way. And it literally is not because the author or the publication was trying to deceive anybody. That's not the point. It's just that someone will scan through this and be like, oh yeah, their expenses are going up, but oh look, their revenues are up 12-fold. They're going to be doing great. The revenue's up 12-fold, but they're here, but the expenses were— it's not the same thing.

Richard Campbell [01:43:04]:
Well, and I think all of them have realized they can't bridge this gap now. And so hence all of the calls for regulation and all the kinds of things that could ultimately protect them from ever having to be a profitable company.

Paul Thurrott [01:43:15]:
Yeah, right. So fundamentally, the big problem when OpenAI or Anthropic or anyone, or Google, don't forget us, we're over here too. By the way, notice that Microsoft has never come out and said like, hooray, broke into something. People have asked them and they're like, no, that's never happened. And I got to tell you, in this day and age, that makes you look bad. Somehow your AI is so well behaved or inept that it has not attacked anybody.

Richard Campbell [01:43:38]:
Do you actually run a test harness that works correctly?

Leo Laporte [01:43:42]:
Yeah.

Paul Thurrott [01:43:42]:
Or you use it responsibly.

Richard Campbell [01:43:44]:
Broke out, exceeded expectations. Like, no, that was just a poor quality test you wrote. All of this language is about lying to you. They're bugs, not hallucinations. This is ridiculous.

Paul Thurrott [01:43:53]:
When the recall thing happened, some security researcher was like, oh yeah, I've broken this thing really easily. It's like, okay, how did you do it? Well, I stripped all the code out of a prerelease version of Windows and I put it in a PC that doesn't have any of the hardware-based protections that a Copilot Plus PC will have. And I got into it really easily. I'm like, okay, so you didn't do anything then is what you're saying. So it's like the OpenAI, the initial Hugging Face thing. It's like, all right, well, we lowered all the security protections and then it broke out. We're like, okay, so I'm sorry, what do we, what? Yeah. It's like we have a prison But we opened the door and we went home for the night, and then we were shocked overnight to discover all the inmates left.

Paul Thurrott [01:44:29]:
You know, they're going rogue. It's like, no, what?

Leo Laporte [01:44:33]:
I don't know.

Paul Thurrott [01:44:33]:
Anyway, I'm not like— I'm not an expert in any of this, but I—

Leo Laporte [01:44:36]:
It's pretty clear there's intense pressure coming from above. In fact, we saw one story that people at OpenAI warned management that, hey, we've got a problem here.

Paul Thurrott [01:44:47]:
And management, this is back in the early part of the year, they're like, hey, we're selling something here.

Leo Laporte [01:44:50]:
We gotta move. We don't have time. And I think that's pretty clear what's going on. There's just huge pressure to train these models as fast as possible. And why? Look, I mean, OpenAI came out with ChatGPT 6.0 last week and 6.1 this week. I mean, there's a lot of speed here.

Paul Thurrott [01:45:12]:
When one of our generation's greatest minds, Mark Zuckerberg, Sorry, I was trying to get through that without laughing. He said, move fast and break things. We're like, that's cute. He was talking about a mobile app that was going to run on 100 computers or phones, whatever. These guys are doing the same thing, but they're at a lot larger scale. Well, I would point out, Zuck's still doing the same thing.

Leo Laporte [01:45:35]:
Muse is exactly that. We're already seeing security issues. It's huge. 10 million users.

Paul Thurrott [01:45:43]:
The only thing that could make Muse any better, just as an idea, a product, whatever, would be if it was smoking a cigarette. Like, I, I, it, it boggles my mind that this thing is the cutest little plush toy stupidity, because I, as a 50-something, almost 60-year-old man, that's exactly what I'm looking for in a computing interface. I want to talk to a plush toy over in the corner through my gross spyglass things, whatever. Oh God, I hate my industry so much.

Richard Campbell [01:46:17]:
Well, the advantage of Zuckerberg is he's already helped propagate some genocides and things, so he's much more comfortable with this stuff than others. Very comfortable.

Paul Thurrott [01:46:24]:
It's like riding a bike, Richard. Yeah.

Richard Campbell [01:46:26]:
Once you've facilitated one Rohingya attack, the others come much more naturally.

Paul Thurrott [01:46:32]:
I always say, people are like, Mexico City's really dangerous. Aren't you worried about being kidnapped? I'm like, oh, I've been kidnapped many times. The first time's terrible, but you get used to it.

Richard Campbell [01:46:38]:
You get good at it.

Paul Thurrott [01:46:39]:
It's like anything else. I'm like, Jesus, I don't know. I don't know what's happening here. But anyway, through some combination of many of the things we just talked about, and, and, but the, the literal thing that set me off was unfortunately Jared Spatar, who I do like, talking about all these different licensing things that they have. I was like, uh, I'm done. I'm done with you people.

Richard Campbell [01:47:00]:
Like it's a feature.

Paul Thurrott [01:47:02]:
Right. And this was before OpenAI announced this Pages thing, which is their word processing, blah, blah, blah, whatever, that's built into ChatGPT. But after Anthropic announced basically the same thing, and well after we've had this notion, I wrote, I think it was 2 years ago, probably, if not a year ago, about the end of apps, which is sort of like this, we're conditioned to do things in a certain way. And to get us to go to the next thing, like we've been saying, you kind of have to offer some transition. But At some point, Richard just gave a couple of examples. Leo does this. Leo, this is what Leo does when I'm talking. He's doing this all the time.

Leo Laporte [01:47:35]:
I'm doing it right now.

Paul Thurrott [01:47:36]:
He's doing new different things.

Leo Laporte [01:47:38]:
There's my little muse guy. By the way, I told him dress for the weather. So—

Paul Thurrott [01:47:42]:
Love it.

Leo Laporte [01:47:43]:
Oh, wait a minute. He's talking to me. Enjoy the show. Oh, here is a dancing— I said dance and it just put a picture of it dancing. Anyway, he dresses for the weather.

Paul Thurrott [01:47:54]:
But this is a new way of doing things and it's—

Leo Laporte [01:47:58]:
Yes. And I can say, get me an Uber. I can say, call and make a reservation.

Paul Thurrott [01:48:02]:
I'm so far behind the curve on this. Like I looked up an iPhone serial number using some AI or whatever it is.

Leo Laporte [01:48:08]:
Yeah, but you did it and it worked.

Paul Thurrott [01:48:09]:
But I did it and it was quick. And I learned my lesson in a way because I tried it other ways.

Leo Laporte [01:48:14]:
Right.

Paul Thurrott [01:48:14]:
And those ways took a long time and failed.

Richard Campbell [01:48:16]:
Right.

Paul Thurrott [01:48:17]:
It's a process. It is a process. But how does this change? In other words, you can look at things in the past in our industry and say, okay, the web came along and changed everything. Mobile came along and changed everything.

Richard Campbell [01:48:33]:
Changed it again.

Paul Thurrott [01:48:34]:
How does it change everything when you can get stuff done without having to go to the app we always had to use, which if it's something you don't know how to use, then you have to learn how to use it or engage with someone who does or use Microsoft is, look, to their credit, I mean, they've added things. There's a search box in the top of all the Office apps that isn't just like search, it's literally like how to find features and you use your own language. You don't know the name of the feature. You're not even sure it is a feature, but it uses whatever it uses to find that thing and it helps you and it's good. That's one way to do it. But you're still bringing up the app. I mean, this is like, to me, what I mean when I say the end of Microsoft is not Microsoft disappears. I mean, IBM's still around.

Paul Thurrott [01:49:17]:
It's a freaking human. humongous company. And this is the rare group of people who might actually be able to tell me what they do today. I don't care, don't tell me. But if I said this to anyone else, normal people, they'd be like, IBM, they're still a thing? Like, oh yeah, no, they're humongous. What do they do? I don't know. And I feel like that's where Microsoft is going. And it's going to be some combination of infrastructure, which, yes, in some ways serves their own needs, but is for third parties, right?

Leo Laporte [01:49:42]:
Mm-hmm.

Paul Thurrott [01:49:42]:
In the same way that Azure today What is it? Richard probably knows. 70%, I think, of the workloads or more maybe are running on Linux, not on Windows.

Richard Campbell [01:49:49]:
At least 60, yeah. Okay.

Paul Thurrott [01:49:52]:
It'll be similar with these workloads that are running through third-party AIs versus whatever Microsoft has, right? But also, look, they just changed their reporting structure for finances, but up until 2 seconds ago, the biggest business they had was something called Business and Productivity Services, Microsoft 365, $42 billion in revenues.

Richard Campbell [01:50:13]:
Now known as Work on AI.

Paul Thurrott [01:50:16]:
Yeah. Now they've jammed it together with all the Azure AI stuff. But as a standalone, well, as a discrete business unit, it was Microsoft's biggest. And what happens when some big corporate customer, which would be the dam breaking that starts this flood, not me, right? I stopped using Office years ago, but rather some big company who's like, it's interesting, We're paying all this money per user per month to Microsoft for not just Microsoft 365. There's identity services in there and other management. There's a lot of stuff going on there, but there's this productivity piece that's Microsoft 365. But we're also paying— I'll just use Anthropic. It could be any company, doesn't matter.

Paul Thurrott [01:50:57]:
But we're paying them per user per month. And it turns out most of our users now are doing what Richard has described, and I've talked about a little bit, which is I don't go to Word or PowerPoint to start that thing. I go to the AI and say, make it. Now, the transition is in the beginning, it is creating a PPTX file or a DOCX file or whatever, and maybe is using Word on the backend or whatever, because Microsoft has added MCP and interoperability to all this stuff it does, because God love them, they just didn't see this one coming. But at some point, it doesn't matter what it's using on the backend. It might not be using anything on the backend. It might be using a free thing on the backend. It doesn't matter.

Paul Thurrott [01:51:33]:
But you're still using the same AI to make the same thing. And now you're like, why are we paying Microsoft for this again? And I'm not saying like they lose $42 billion immediately, but if there's some Fortune 500 company that comes out and says, guess what? We don't need this anymore and we are not paying for it. We may still pay for their Entra ID, email, calendar, whatever. We may go to Google. Maybe that's cheaper or better, or who cares? Whatever they do, it doesn't matter. But if they can actually cut out the productivity part of that, which is Microsoft 365 as I think of it, and just use AI and then potentially save money on every employee every month. But even if they spend the same, whatever, the money's now being reallocated to Anthropic or whoever.

Richard Campbell [01:52:19]:
All those documents are still there. So is Microsoft's value proposition the document format? I don't think so.

Paul Thurrott [01:52:25]:
I think it's an open standard.

Richard Campbell [01:52:27]:
I think the most valuable thing that Microsoft has in this space is the graph data, is who emailed who, where did those files go, what were all the iterations. That's a subtle amount of data.

Paul Thurrott [01:52:38]:
By the way, I 100% agree with you. The only problem is it's open. I don't mean open like open source. I mean it's open like beer. No, I mean it's open, the AI for others are reading that data too. At some point, the openness—

Richard Campbell [01:52:55]:
The only barrier is, can you read that data?

Paul Thurrott [01:52:58]:
Yeah. I think OpenAI has proven, or AI I should say, has proven itself to be pretty good at reading things it's not supposed to read. So look, I'm not predicting Microsoft goes away. That's not my point. But there could be a very big shift in the way that we get work done collectively as a whatever, a society, a race, whatever it is. And in Microsoft's case, the problem is not going to be when I do it or you do it or Or any individual does.

Richard Campbell [01:53:21]:
No, it's going to be one of the Fortune 100.

Paul Thurrott [01:53:24]:
Yep.

Richard Campbell [01:53:24]:
But to that point, look at their response when Anthropic showed up with Work and just humiliated M365 Copilot.

Paul Thurrott [01:53:33]:
Exactly.

Richard Campbell [01:53:33]:
And their answer was, license it now, which is not a bad answer. Satya may not be Bill. Bill's the original guy who figured out which way the parade was going and ran in front. And Satya knows enough to at least do that.

Paul Thurrott [01:53:50]:
Yeah. Nothing is certain, but I do see— look, Microsoft has often talked about betting the company and it has seen various existential threats. You mentioned Netscape earlier. Absolutely Netscape and the web was an existential threat. Absolutely. That one turned out great for Microsoft, I guess, but they're not always going to come out on the winning side.

Richard Campbell [01:54:13]:
And they saw it as an existential threat and attacked it. it that way. And it's not why Netscape went down. A browser company's never been a real company ever, right? You know, they— and Netscape tried many ways to try and make revenue, and, and it failed. It didn't need to be attacked by Microsoft to fail. As is typical with these pioneers, that's a great part.

Paul Thurrott [01:54:32]:
Netscape made tons of mistakes. It wasn't just—

Richard Campbell [01:54:34]:
yeah. And even if they didn't make mistakes, it wasn't a business.

Paul Thurrott [01:54:38]:
Yeah, that's true.

Richard Campbell [01:54:39]:
You know, but that— and that was a problem. It's always been a problem. And that's where you're really— when you're looking at Anthropic and OpenAI, that's what you're looking at. The one same problem.

Paul Thurrott [01:54:50]:
Well, this is a little too simplistic, but the big asset that Microsoft has that big AI needs is their infrastructure. But there are others. I mean, getting entry to that market of business customers and all their data and all the stuff they're doing for productivity is also humongous. I mean, there's no doubt about that. But by virtue of the fact that these companies do not want to be regulated and are doing everything they can to go as fast as they can, they've made themselves interoperable with each other. And when you're the incumbent like Microsoft, this is like IBM did this when they created the first PC. They're like, we'll just make it an open— we'll just use open off-the-shelf parts. Well, guess what? That means anyone can make something idiotic.

Richard Campbell [01:55:31]:
Because their plan was to copyright the BIOS until—

Paul Thurrott [01:55:33]:
Which they did. But that could be reverse engineered and was.

Richard Campbell [01:55:36]:
Sure. But they also did that in response to the consent decree that they signed with the American government. I don't disagree that historically, like 100 years from now, we're going to talk about how the lack of functionality of the US government at this pivotal moment became a major part of whatever comes next.

Paul Thurrott [01:55:53]:
Oh, we should be having that conversation right now. That's absolutely true.

Richard Campbell [01:55:56]:
Not that they've necessarily done a good job, but they're clearly not doing anything right now. And this is so much the robber baron era of ungoverned railroads where they're just doing as much land grab as they can because they know it ends. Whether it ends from regulation or it ends because they just run out of money. Yeah.

Paul Thurrott [01:56:17]:
No, that's true. It's very much the same. Okay.

Leo Laporte [01:56:23]:
Okay.

Paul Thurrott [01:56:24]:
Okay.

Leo Laporte [01:56:25]:
Okay.

Paul Thurrott [01:56:26]:
We're going to try to land this plane. Okay. Okay.

Leo Laporte [01:56:28]:
Well, I'll tell you what. Let's do the Wanted, famous, much beloved, user demanded Xbox theme song. In just a moment.

Paul Thurrott [01:56:40]:
Oh, you missed it last week.

Leo Laporte [01:56:42]:
Too late.

Paul Thurrott [01:56:42]:
There was no theme song.

Leo Laporte [01:56:44]:
No, because now there's a theme song exactly where I don't want one.

Paul Thurrott [01:56:48]:
Who I will never forgive.

Leo Laporte [01:56:49]:
Hold on just a minute. You're watching Windows Weekly with Paul Thurrott. There he is, thurrott.com. And, uh, our good friend, our wonderful friend, Richard Campbell.

Paul Thurrott [01:56:59]:
Pointless rage. That directed at nobody in particular.

Leo Laporte [01:57:02]:
Despite the fact that he's a Canadian, still remains an ally.

Richard Campbell [01:57:06]:
Somehow. Not making it easy.

Leo Laporte [01:57:09]:
Abby, when she went to Vancouver, she said, I thought everybody'd be mean.

Paul Thurrott [01:57:12]:
Mean? What?

Leo Laporte [01:57:14]:
And they forgive her. They forgive her.

Richard Campbell [01:57:16]:
Yeah.

Leo Laporte [01:57:18]:
Uh, anyway, uh, we're glad you're here. Thank you for watching. We appreciate it. And, uh, now it is time for the Xbox. I guess you could play the Halo theme.

Paul Thurrott [01:57:29]:
Ba-da-ba-ba-da-ba. Only on Xbox, am I right? So rumored for probably a decade and then announced just weeks ago, I think, a few months at most, Microsoft's disc-to-digital program for the Xbox is now available. It just kind of appeared out of nowhere. And it seems like this would be a hard technical problem, or I guess We would call that a hard computer science problem or whatever. Um, but no, I guess not. So if you have an Xbox One or an Xbox Series X, or no, I think it's just X actually. It's not X or S, which is a little weird, right?

Richard Campbell [01:58:09]:
That's weird.

Paul Thurrott [01:58:10]:
Um, yeah, I mean, I assume that changes over time. Um, but you can get a digital license for supported games. It's not every game, but, uh, you insert the disc into your console and it will be attached to your Microsoft account, your Xbox gamertag, I guess. And that gives you kind of like digital rights to it. So you can go download it from xbox.com, the Xbox app, whatever. You don't actually need the disc anymore. And a future person who you may sell or give this game disc to will not be able to do this, obviously. Like, that has to be part of it.

Paul Thurrott [01:58:41]:
And the reason this works is that in the Xbox One timeframe, Microsoft anticipating this, 'cause they've been actually planning on this for a long, long time, have put a special code that's unique to every single disc that identifies that version, you know, that disc of that game. And they can actually associate it with you, probably using a GUID because it's Microsoft, but whatever. So that's kind of a cool thing. And then I assume this is gonna expand to some degree, although one of the issues with this is doing this on 360 or OG Xbox games is gonna be difficult.

Richard Campbell [01:59:16]:
But all that's getting pretty old. The question always is, so then you sell the disc to somebody, what happens when they try to do—

Paul Thurrott [01:59:22]:
They can't. They won't. So the way they're kind of doing this is like, like, look, they're, they're absolutely— like, in the same way that people will go to like a yard sale or a fair or whatever it is and buy a box of discs of whatever kind— they could be music CDs or game discs, whatever— um, like, to rip them and get digital copies if it's music, or whatever they're doing with those things, or just to build up a collection. Um, people will do that for this reason, right? There's no doubt about it. Like, they're gonna— you're gonna go try to find Use discs that are cheap and then associate that with your account and then you'll have access to that game forever, I guess, right? So people will do that. But I believe that the way this works is that it's not something you go to do. It's something you're prompted with when you put the disc in. So if that disc has already been associated with a, I'm gonna call it a Microsoft account, and you plug it in later 'cause you bought it or found it or whatever, you're not gonna get that.

Richard Campbell [02:00:12]:
Yeah, I wonder if that's the wrong way to go. If it should be, it should go with the disc and so you pull the rights off of the old owner. And onto the new owner.

Paul Thurrott [02:00:20]:
Yeah, that's interesting.

Leo Laporte [02:00:22]:
That's crazy talk. Yeah.

Richard Campbell [02:00:24]:
New owners.

Leo Laporte [02:00:25]:
Only you don't even own the first copy.

Paul Thurrott [02:00:27]:
Why do you hate copyright so much, Richard? Yeah, Richard was born here.

Richard Campbell [02:00:31]:
No, he's only a license to use.

Paul Thurrott [02:00:33]:
He's Canadian. Um, no, I think, um, so that's interesting. Like, in other words, technically, like, you trade in like a phone, right? And so the phone, like, I spoke about this, you have serial number.

Richard Campbell [02:00:43]:
Like, I mean, and it's super common to play a game for a few months, have played it enough, Send it off to the used pile so that somebody else can play the game.

Leo Laporte [02:00:51]:
Well, that's, and that's exactly what somebody does.

Paul Thurrott [02:00:53]:
Well, somebody, no, sorry, sorry.

Leo Laporte [02:00:53]:
Because you could bring it to GameStop and sell it.

Paul Thurrott [02:00:55]:
Yeah.

Richard Campbell [02:00:56]:
Yeah.

Paul Thurrott [02:00:56]:
No. Um, you can still play the game from the disc. That's not stopping this. This is giving you like what is essentially like a digital entitlement, right?

Leo Laporte [02:01:02]:
So. Yeah. But they want to get rid of discs. That's the whole point, right?

Paul Thurrott [02:01:05]:
No, of course. But if you are some kind of a collector, just love games, whatever, that you buy like a used game and you, you look, you're going to play this off of a disc. You're not going to get that digital entitlement because the original owner perhaps did this already.

Leo Laporte [02:01:17]:
Yeah.

Paul Thurrott [02:01:17]:
If your goal was to get a digital entitlement, you're going to keep looking. I guess you keep going. I don't know. But if your goal is just to play the game for a little while or play whatever.

Richard Campbell [02:01:25]:
And if it doesn't transfer with, then can you transfer with? Like, this is the hard part of this whole thing. If you're going to do this, then give me the ability to move it around.

Paul Thurrott [02:01:35]:
Yeah.

Richard Campbell [02:01:35]:
Yeah.

Paul Thurrott [02:01:36]:
Yeah. So I, I don't have an answer to that because this just happened, but my guess is that You've already raised a couple of things that probably are going to have to be addressed in the future, but—

Richard Campbell [02:01:46]:
The main things. Now, what happens when my account gets banned?

Paul Thurrott [02:01:50]:
Right. Well, then you're screwed, Richard. You know that.

Richard Campbell [02:01:52]:
Well, yeah. Now we also get into, that has nice litigation opportunities too, because now you have an asset assigned to that account.

Paul Thurrott [02:01:59]:
Yep.

Richard Campbell [02:01:59]:
Yep.

Paul Thurrott [02:02:02]:
Well, look, there's a cabin in the woods. There's no electricity, there's no water.

Richard Campbell [02:02:07]:
Worked for the Unabomber.

Paul Thurrott [02:02:08]:
It could be the Unabomber, yeah. Hopefully not, but you never know. Anyway, so that's happening, whatever. Minecraft Dungeons 2, I think, just came out, possibly as recently as today.

Richard Campbell [02:02:18]:
Yes.

Paul Thurrott [02:02:19]:
Yes, they don't remember what the day was. So Minecraft Dungeons is to Minecraft as Forza Horizon is to Forza, right? It's a kind of a more action-oriented version of the original, if that makes sense. It's a little more arcadey, whatever. But I don't play Minecraft. I mean, I played it, I understand it, but I don't— I'm not like an active Minecraft player. But apparently there are 2 additional— there's the world that Minecraft takes place in, which has a name that I don't know. And then there are these additional dimensions of which there are 2. But Minecraft Dungeons 3 introduces the, well, 4th, but 3rd, not counting the original, new dimension.

Paul Thurrott [02:03:00]:
I think it's called, not Shift, Sift. It's got kind of a pink and green kind of weird color scheme to it, whatever. It's got all different properties to it. But this is the first time in over 14 years they've added a new dimension to a Minecraft game. And at the Minecraft event this past, I think it was the weekend actually, over the weekend, Mojang said that they're gonna bring that dimension to the core Minecraft Java and Bedrock Edition starting next year. So like that's—

Richard Campbell [02:03:28]:
So it's not the ether. Or the Nether.

Paul Thurrott [02:03:31]:
Yeah, it's a new one. Yeah, it's called Sift, right? I don't have all the names in front of me, but—

Richard Campbell [02:03:36]:
Overworld is the main thing, but then you find portals to the Nether.

Paul Thurrott [02:03:40]:
Yep. So they're right. So this will, this will be a new one of those, I guess. There'll be a portal, there will be portals too. Yeah, that's cool. And what else? Oh, there's actually, there's 2 things in here. So Microsoft has created a new casual game on mobile and it's called Microsoft something for some reason, not Xbox something. Like what? Like, so it's called Microsoft Jewel.

Paul Thurrott [02:04:12]:
As you can tell from the name, maybe it's one of those jewel type games, you know, with the rows and columns of jewels because we haven't seen one of those.

Leo Laporte [02:04:18]:
Bedazzled.

Paul Thurrott [02:04:19]:
Yes.

Leo Laporte [02:04:20]:
Original.

Paul Thurrott [02:04:21]:
Yep. Or like Candy Crush, which Microsoft owns, right? Isn't that part of King?

Leo Laporte [02:04:26]:
Oh yeah, I guess so.

Richard Campbell [02:04:27]:
It was a King game. Yeah.

Paul Thurrott [02:04:28]:
Yeah. So I don't know why they're doing this, but it, um, but look, I, I say this having complained for the past 2 years, Microsoft's not doing anything mobile and now they're doing it. And I'm like, I'm complaining about that, but, uh, it does have Xbox achievements and okay, whatever. I don't know. So that's kind of interesting. Um, Microsoft does have a casual games I don't know, business group, team?

Richard Campbell [02:04:51]:
Mm-hmm.

Paul Thurrott [02:04:52]:
I don't know what it is, but they're going to be joining King soon, if they haven't already. So this came out of that operation. So I guess they're, you know, whatever. They are or soon will be part of the same thing. So Microsoft Casual Games is the part of the company that made Solitaire, which became Solitaire Collection, right, in Windows 11. Mahjong, Sudoku, Wordaments. Et cetera. Wordament.

Paul Thurrott [02:05:17]:
Wordament, the Burger King of word games. Anyway, okay. And then, sorry.

Leo Laporte [02:05:22]:
At least you get a paper crown at Burger King.

Paul Thurrott [02:05:26]:
And then this is like, I needed another reminder that I'm getting old. Microsoft announced something called Mythic Achievement. And a Mythic Achievement is what you get now when you complete all of the original achievements in a game.

Richard Campbell [02:05:41]:
Right.

Paul Thurrott [02:05:42]:
Which is something the PlayStation's had, I guess, for a long time called Platinum Trophies, right? They have trophies, not achievements. But, um, so to be clear, uh, depending on the game, sometimes games come out and they add achievements later, or sometimes they add achievements to like DLC or whatever it is. This is literally the original set of achievements for any given game. If you've unlocked all those, you will get a Mythic achievement for that game, which I assume gives you like a $2 coupon for Burger King or something. I don't know what the fuck.

Richard Campbell [02:06:10]:
And a hat.

Leo Laporte [02:06:11]:
And a hat.

Paul Thurrott [02:06:11]:
Yeah, I don't know. It's kind of curious, but—

Richard Campbell [02:06:14]:
Hey, listen, for the completionists, it's every trophy you can get.

Paul Thurrott [02:06:18]:
Yeah. Yeah. There's something to that. I once spent half of a winter driving a tank around a multiplayer level in one of the early Call of Duty World War II games because I was told that that was how you get the final achievement in that game that I never got. And then I learned later that that is not how you do it. And I cannot tell— I seriously spent more time with that tank than I spent with my second child. I don't know.

Richard Campbell [02:06:41]:
That was— now look at you. Listen, I, uh, I don't like myself in completionist mode, so I do my best to not—

Paul Thurrott [02:06:48]:
It's hard when— I mean, I think most people in our, uh, space are like that. It's just part of the way we're wired, you know?

Richard Campbell [02:06:53]:
Yes. So I deliberately do not put myself in a position where, yeah, I can— those urges come forward.

Paul Thurrott [02:07:00]:
That is smart. You are smart.

Leo Laporte [02:07:05]:
Now we continue on with the show. We call it the back of the book, and it kicks things off with Paul's spoiled tip of the week.

Paul Thurrott [02:07:13]:
Yes. Thanks again, Keith. No, it's fine. I'll never forget you, Keith.

Richard Campbell [02:07:18]:
Yeah.

Paul Thurrott [02:07:18]:
I'll forgive you. It's all good. Yeah. It's always the Linux guys.

Richard Campbell [02:07:23]:
Yeah, exactly.

Paul Thurrott [02:07:24]:
Yeah.

Richard Campbell [02:07:25]:
Yeah.

Paul Thurrott [02:07:25]:
Just got to get the dig in there. No, it's fine. So Windows 11 version 26H2 is available ostensibly This is a good time to update things like install and recovery media. Obviously, I do. I go through this process every time. Well, not even once a year, because honestly, they update the ISO over the— not every— I don't know if it's every quarter, every so often they do anyway. But so I kind of do this throughout the year. But excuse me.

Paul Thurrott [02:07:50]:
So if you have an x86 computer, download Windows 11. Just Google it. It's there. There's a separate page for ARM, but remember, you can't use the media creation tool on ARM. And then there might be some drivers missing, right? So those can be had through Windows Update after you're done, but you can use Rufus or whatever. And then the article I posted has a couple of direct links if you just wanna grab it in either form, x64 or ARM64. So that's out there if you need it or want it or whatever. Of course you need it and want it.

Paul Thurrott [02:08:20]:
Of course you do. What a stupid thing to say. Anyway. And then the app pick, Opera has 2 major app releases this past week, one on desktop and one on Android. All of mobile but Android. The version on desktop, they've had a free VPN for over a decade now. It's the only actually true and unlimited web browser-based VPN out there. And they added a new feature to it when you can turn this off if you don't like it, but when it auto-detects to— auto-detects rather that you're on a public Wi-Fi, it can automatically just come on when you're on public Wi-Fi, right? So we'll just come on in that circumstance and then turn off when you're back on a trusted Wi-Fi.

Paul Thurrott [02:08:59]:
So that's kind of cool. And then on mobile is something I wanted to try when I went to— I'm going to Mexico next week, allegedly. We'll see if that works out, but that's the plan. And unfortunately, this is not available there, but Opera has added eSIM capabilities to their Android browser. And the way this— it goes through, it's a third-party thing and it works for your whole phone. It's not just for the browser, obviously, but you install an eSIM on your phone like you would through any other source, blah, blah, blah, whatever. You might want to just check it out for free because there's a— for now at least, there's a free 3 gigabyte plan that is only available for 3 days, but it gives you a chance to actually try it out or whatever. Just get 3 gigs of data for a few days.

Paul Thurrott [02:09:40]:
It's available in 47 different destinations. Like I said, one of them is not Mexico, which is vaguely irritating, but whatever. I did download this and I downloaded it for the United States. I haven't activated it yet, but I'll try to. I'm going to Boston tomorrow, so I'm going to try to test it during that trip or whatever. But it's available through a company called holiday.com, which I've not heard of, even though I felt like I had tried every single eSIM supplier on earth. But the good news is, at least in the US, the prices are actually really reasonable. So when you pay for this thing, you can pay for it in 3 gig, 5 gig, or 10 gig allotments.

Paul Thurrott [02:10:17]:
And the price for 3 gig, and it works for 30 days, right? It's kind of typical for eSIMs. is like $5, right? Which is— that's like good, you know, like 5 gigs is $9 and 10 gigs is $14. So the prices are going to vary by country, etc., etc., but I assume they're all in the same kind of ballpark. So I mean, based on— I paid stupid dollars for a lot of eSIM stuff, so it seems reasonable to me. It's worth looking at. And plus you can get the 3 gigs for free, so why not give it a, you know, just give it a shot.

Leo Laporte [02:10:46]:
Nice.

Richard Campbell [02:10:47]:
Awesome.

Leo Laporte [02:10:48]:
Free is a good price. Love the free.

Paul Thurrott [02:10:51]:
Yeah, free is always good.

Leo Laporte [02:10:52]:
Now, Mr. Campbell, I think you're up next with Ronaz Radio.

Richard Campbell [02:10:58]:
Uh, bringing back one of my regulars, Pauliana Shevitz, uh, the young lady— she's not as young, but none of us are— uh, who runs Secur out of Warsaw, Poland. And she's your go-to if, uh, you're a government who's systems have been breached by the bad guys.

Leo Laporte [02:11:16]:
Oh wow.

Richard Campbell [02:11:17]:
And she— one of her many businesses as part of Secure is penetration testing. And we were— I met her at Black Hat while you— we were all together briefly, and she was doing a talk about using LLMs to automate a bunch of pen testing work. And I said, that'd make a killer run, as— so after the event, we got together and, and talked about, uh, exactly how her new workflows look And just the point that what used to take her a team of 4 people over a week now is done by 1 person in an hour to just hammer through all of the likely pen test scenarios.

Leo Laporte [02:11:52]:
That's what they're good at.

Richard Campbell [02:11:54]:
Yeah, just a recognition of speed and power.

Leo Laporte [02:11:57]:
I mean, that's basically what these breakout agents are doing, right?

Richard Campbell [02:12:01]:
To some degree, except also the doors are being left open.

Leo Laporte [02:12:06]:
I mean, well, that makes it a little easier on them, but—

Richard Campbell [02:12:08]:
There was also a callback to the Les Test show, uh, uh, where we did the automatic attack disruption, um, which is, you know, Entra acknowledging the reality that these days the breaches come fast and hard, and the response to it is more important than resisting every breach. So, uh, yeah, it was a great conversation. Paula's an expert in the space, knows this thing cold, and her company's being transformed by this. She has about 60 people, and they just can do so much more now. So it's, uh, it is, it was profound and great, you know, the fact that she's embraced it because she's so, she's really quite, as most security people are, quite conservative.

Paul Thurrott [02:12:44]:
Yeah.

Richard Campbell [02:12:45]:
About things, right? She's like, she just can't argue with the benefits that these tools present in this scenario.

Leo Laporte [02:12:52]:
They are relentless, that's for sure.

Richard Campbell [02:12:54]:
Astonishing.

Leo Laporte [02:12:55]:
And now, ladies and gentlemen, let's bring in the—

Paul Thurrott [02:12:59]:
oh, I love it—

Leo Laporte [02:12:59]:
Windows Basically singers, because it's time once again.

Paul Thurrott [02:13:05]:
Literally, I heard that in my ears and I was like, what is happening to my brain?

Richard Campbell [02:13:10]:
I like the fade-in.

Paul Thurrott [02:13:16]:
Whiskey time, ladies and gentlemen!

Leo Laporte [02:13:21]:
Richard Campbell Michael ends every show with a whiskey pick and a review, and he's got something weird from his closet this week.

Richard Campbell [02:13:32]:
This is not from my closet. This is from my, my journey. So I missed last week. From Scotland, perhaps? I was in Scotland. So the people were asking, so I will tell you the details. This particular trip is called the Science of Whiskey Tour. It is hosted by New Scientist, which is a magazine out of the UK. They do one a year.

Leo Laporte [02:13:54]:
It's a great magazine actually.

Richard Campbell [02:13:56]:
I know that magazine. It's a very great magazine. I am a subscriber. I am a fan as well. This trip was staged out of Oban on a schooner called the Flying Dutchman. It's a big white ship.

Paul Thurrott [02:14:08]:
Jealous.

Richard Campbell [02:14:09]:
I included a link there too. They do a bunch of tours, but the Dutchman was a great ride. The captain's name is Ares, as in the in The God of War. He's a Dutch guy with a strong Scottish accent now, and he looks like the God of War. Like, he's about 6'4, basically bald, big beard. Like, if you put the red paint stripe on him, he'd be the God of War. And, uh, and really a great captain. Uh, our trip was modified a number of times due to weather.

Richard Campbell [02:14:41]:
Uh, just no two ways about it. It was, it was challenging conditions. And so we traveled around and I'll reveal over time all the places we went. We had 7 days on the boat. We started in Oban, we sailed up to the Isle of Mull and went to Tobermory, and then originally had been intended to go around Mull into Iona and Staffa, but the weather was just too rough. So we came back down the sound and snuck in behind the Slate Islands to avoid a bunch of the rough weather and then up the little strait between Islay and Jura and stayed near Bunahabhain for a couple of days and went by dinghy ashore and went around Islay that way. In the end, we visited 9 distilleries, and that included 2 operational tours. So there's a bunch of ways to go to a distillery.

Richard Campbell [02:15:31]:
The simplest is just to go into the visitor center. You can get yourself a bottle of a distiller's edition or any whatever else they're selling in there. They often have a dram to taste, often the distillers editions, which are generally only available there. Not expensive, doesn't take a lot of time, pretty simple thing to do. You could dash in and out, and I did a bunch of those. We only had so much time. An operational tour is where you do a tour of how they make their whiskey. These are all going to be relatively similar, so you can only do so many of them before you start to lose your mind.

Richard Campbell [02:16:00]:
They're going to explain the same things. This is a much more scripted process. Typically, there's a tasting at the end, but it depends on the tour. So we did a couple of those. We did Bunahabhan and today's whiskey, Kilchoman. And then there's a warehouse tour. And warehouse tour is what you want. You're going to need a group.

Richard Campbell [02:16:21]:
It only makes sense with groups of 6, 8, or more people. In this case, there was 20 of us. You're going to go into the barrel houses, and they usually have a little seating area hidden away in the back amongst all the barrels. And they're going to have a set of barrels pulled and already paid for, essentially have done the bond on them, and you're going to taste whiskies directly from the barrel. Hard to argue with. It's a great experience. Absolutely recommend it. It will ruin you for whiskey for the rest of your life.

Leo Laporte [02:16:51]:
Really? Does it taste different from the barrel? Why is that?

Richard Campbell [02:16:53]:
100%. Yeah. And you're tasting things you'll never taste again because they will be, as we've talked about in the whole discussion about making whiskey, there are things that happen from the barrel before it shows up in the bottle.

Paul Thurrott [02:17:05]:
Right.

Richard Campbell [02:17:06]:
And so you're really tasting the whiskey as it is. It's all obviously, for obvious reasons, going to be cask strength. Although I did taste directly from the barrel a whiskey that was only 42%, like it was in danger of no longer being whiskey, which is why they pulled it for this particular purpose. Um, so all of this is enjoyable. Uh, in the end, I'm doing the math. We tasted more than 40 different whiskeys. Oh, there were days where we had 3 tastings in a day.

Leo Laporte [02:17:34]:
So basically you're on a sailing ship with a bunch of drunks.

Richard Campbell [02:17:38]:
Oh yeah. No, and then, and believe me, very— by the, by the end of the first day, I quickly figured out drink no additional whiskey because it's just a lot. Yeah, a lot of whiskey coming at you.

Paul Thurrott [02:17:48]:
Yeah.

Richard Campbell [02:17:48]:
And there were some younger folks there that did like go out to the bar and have a few drinks before arriving back at the boat to find out, oh no, there's a 6th whiskey tasting tonight.

Paul Thurrott [02:17:57]:
Yikes.

Richard Campbell [02:17:59]:
Uh, so yeah, you gotta, you gotta pace yourself on these sorts of things. And I'm going to be telling stories about this trip now for many weeks. I'll just apologize in advance.

Leo Laporte [02:18:10]:
No, no, no, that's awesome.

Richard Campbell [02:18:13]:
There's a ton of them. Uh, it was such a great— a really a lot of fun. Uh, but I— so I had to pick one, uh, for this particular day, and I picked the one that was one of my very, very favorite tours and a distillery I'd never been to before and never tried before, and that is this Kilchoman. And I have the bottle right here. Now, of course, this is an Islay whiskey, so you'd presume peated, and you'd be incorrect. Kilchoman, like a number of distilleries, do make a variety of editions, some peated and some not. Most people equate Islay with Lagavulin and Ardbeg and Laphroaig, very heavily peated whiskies, and pretty much that's all they do. And fortunately, I've already talked about all of those at different times over the course of the past few years, so I'm not going to have to talk about them all.

Richard Campbell [02:19:02]:
There are many others that do have unpeated versions of their whiskey, Bunnahabhain being one of them, and Bruichladdich, and of course, Kilchoman. One of the things I really like about Kilchoman is that we had such a phenomenal tour. I mean, there's good tours and there's bad tours. In the end, you're going to get whiskey one way or the other. We did an operational tour of Kilchoman, and the lady who took us on the tour, her name is Naomi, and she is in her late 20s. and is a PhD candidate. She's got her master's degree by making alcohol from seaweed. Not that it was good, and she'd be the first to tell you that, but just to show the chemical processes to extract the sugar from the seaweed to be able to get the yeast to produce alcohol.

Richard Campbell [02:19:49]:
And now she's working on a PhD thesis on high-density whiskey making, which is a whole complex chemical process. So super knowledgeable for a crew that were about the science of whiskey. And I should bring up that one of the things that made the Science of Whiskey tour really very special is we had a scientist with us. That's what New Scientist does for us. So we had, uh, uh, Kevin, who guided the tour and is— and works for New, uh, New Scientist, but then also Dr. Barry Smith, who is a doctor of philosophy and a neuroscientist who really talked about the systems of tasting. Uh, he started us off right. It just did such a good job.

Richard Campbell [02:20:27]:
Before I get into this whiskey, On the very first day before we left Oban, we spent the night in Oban before we sailed the next day. He got us all seated down and he says, we're going to work on your palates today. We're going to learn about how we actually taste and how much the role of smell and taste are tied together.

Leo Laporte [02:20:43]:
Oh, yes.

Richard Campbell [02:20:43]:
And the way he did this, very, very clever, was he gave us nose plugs. They clipped over the top of your nose like this. They were medical, right? They're not particularly expensive, but you put this on and then he gave us all a jelly bean.

Leo Laporte [02:20:56]:
Oh.

Richard Campbell [02:20:56]:
So now you're chewing on this jelly bean and so you could taste it's a little sweet and so forth. But the fact is you've already chewed it for a bit. So when he says, now take the nose plug off, well, your mouth's already filled with the scents from the jelly bean. So I went from, there's something vaguely sweet that's chewy in my mouth, to, holy cow, that's banana. Like just a massive—

Leo Laporte [02:21:15]:
Yeah. All the flavors in your nose. All your tongue can do is sweet, sour, bitter.

Richard Campbell [02:21:20]:
Exactly.

Leo Laporte [02:21:21]:
Yeah.

Richard Campbell [02:21:21]:
So right away I'm like, I like this guy. I like the theater of it. I like the way he's hit us with this thing. And that's what this week was like, was exploring. He didn't just tie the easy ones of taste and scent, but also touch and sound. And again, this will go throughout the next few weeks, so stick around because it's going to go on for a while. Kilcoman is a Gaelic name. It comes from Siltomhain, which means literally the Church of Saint Coman, who was an early Celtic saint.

Richard Campbell [02:21:55]:
And this is actually a parish, an area on the west side of Islay. So this is in what an area they call the Rhins of Islay, which is the long western peninsula that faces the Atlantic Ocean. It is mostly an agricultural area. There are a few other distilleries in the area like Bruichladdich. But this is a very new distillery. It's not far from Loch Gorm, which is literally the name of this whiskey, Loch Gorm, which is the largest lake on Islay and nearby. The creator of Kilchoman is a fellow by the name of Anthony Wills and his family. He was in the whiskey business.

Richard Campbell [02:22:35]:
He actually created the Vintage Malt Whiskey Company. So he was doing independent bottling in the '90s. And he got some frustration with what barrels he wanted to get. And as whiskey itself evolved and whiskey makers, the company, were less willing to sell off their barrels, that they would do their own special editions and things like this, he decided, he realized that he really wanted to make his own whiskey and get into it. And so he formed the company in 2001. And then when looking for, what do I got to do to do a farm-to-bottle style distillery? He moved his family to Islay in 2004 and ultimately was able to set up in Kilchoman at an area called the Rockside Farm. This is in 2005, right near Machir Bay. This is all west side Islay.

Richard Campbell [02:23:24]:
It's not actually on the water, unlike most Islay distilleries. One of the great things about being in the Flying Dutchman was seeing distilleries that were there before there were roads, that all of these distilleries worked from their docks And so we were literally seeing the distilleries as they were built in the 1700s and 1800s. But as soon as there were roads and a ferry system, they all switched to using trucking because it was just more cost effective. And so most of the docks we were going to were in a dilapidated state, like not in a good way at all. But that is not Kilchoman. When they built that distillery in 2005, they were the first new distillery built on the island since 1908. So 120 years. So yeah.

Richard Campbell [02:24:08]:
And they, the farm was already there. The, the Rockside Farm, it was a barley farm. Mark and Rojas French ran that farm. So they were partnered with them, although that eventually ended. Unlike most new distilleries, Kilchoman has never made a gin or a vodka or anything else. They have always only made whiskey, which is a very challenging way to start your business. You know, it is definitely a modern distillery. They had a visitor center from the the very beginning, which they subsequently expanded as they became more successful.

Richard Campbell [02:24:39]:
And Anthony was quite single-minded about focusing just on whiskey. They had a number of challenges. He had steam pipe bursts. He had a kiln fire in 2006 that required them to rebuild their kiln. It turned out that I found from out from Naomi when we did the tour that it was the teenage sons that were supposed to be supervising the peat in the kiln, but there was a rugby match on and maybe weren't paying close enough attention. Oh no. Oh no. As it's described on the documentation, let's say they don't give too much responsibility to teenagers anymore.

Richard Campbell [02:25:17]:
One of the ways that he did stay afloat through all of this is that he did sell non-whiskey editions, because remember, you have to be at least 3 years. So he actually sold some new spirit and a product they called 2 Years of Anticipation. but also did a lot of private caskings with unusual barrels and things. And so had people commit funds by buying into barrels early on. Their inaugural release from the original 60,000 liters that it distilled back in 2006, 3 years on was in 2009, and things went well from there. In 2015, they actually bought out the French's farm, the Rockside Farm. It's now all owned collectively. I've read some articles about there was some tension between them.

Richard Campbell [02:25:55]:
And by 2019, they were winning world-class awards. Their Sangay, uh, whiskey won gold at the San Francisco World Spirits, which is a big deal. And they did another major expansion there. And just last year, uh, Macquarie Bay Sangay Cask Strength and Loch Gorm, which is this, won gold at the International Spirits Competition. And so this year they're at about 650,000 liters per year. That's still small operations. Most of these facilities in and around Islay today are in the 1 to 6 million class. So 600,000 is still a relatively small operation.

Richard Campbell [02:26:32]:
30% of the barley that, uh, that is in, uh, Kilchoman Editions is from the Rockside Farm. So they grow— this is a farm-to-bottle operation, but it's only 30%. They do their own peating, and the rest they get from Port Island Maltings, which is Diageo, which is all 50 So that's, uh, your Lagavulin and the like. Uh, part of the tour that we did with Naomi— and Naomi gave us all these little miniature Glencairns on a lanyard. So most of the time when you're doing an operations tour, they don't give you any whiskey. Yeah, you normally don't get whiskey. You'll get it at the end. But they wanted us to taste the whiskey along the way, and you know you'd lose these things, right? So yeah, you wear them around your neck.

Richard Campbell [02:27:15]:
Just— there you go.

Leo Laporte [02:27:16]:
I would just walk around all day with it. So it's a little—

Richard Campbell [02:27:18]:
oh no, I have a little glass sitting in front of me ready to go.

Leo Laporte [02:27:20]:
I'm ready.

Richard Campbell [02:27:21]:
And so at various points along the tour, we stopped and were able to get a little drink of something and try this and continue on, which I liked, rather than taste them all back to back.

Leo Laporte [02:27:30]:
Yeah.

Richard Campbell [02:27:31]:
You know, so one of our first stops was the malting floor because they do their own maltings from the barley that they grow on the farm. And it happened to be that the day, the moment we were there, they had just finished a malting and were literally sweeping it into a hole in the floor to be put into the bins to go to be dried. And so first thing I did was pick up a bit of it and eat it because it's this little sprouted barley grains.

Leo Laporte [02:27:58]:
I bet it's tasty actually.

Richard Campbell [02:28:00]:
It's a little sweet and a little chewy, nutty.

Leo Laporte [02:28:02]:
I love barley.

Richard Campbell [02:28:04]:
Yeah. And so they soak it in water, then they spread it out on the floor.

Leo Laporte [02:28:08]:
Yeah, sprouts.

Richard Campbell [02:28:09]:
Concrete floors and they let it sprout. They turn it constantly so that it doesn't clump up or get moldy. It needs to be aerated, which is where you get the term monkey shoulder. Monkey shoulder is pains that you get, damage you get done to your shoulder by constantly turning maltings.

Leo Laporte [02:28:27]:
That's, that's where they need a monkey to do it.

Paul Thurrott [02:28:29]:
Yeah.

Richard Campbell [02:28:30]:
And today, for the most part, people don't— most distilleries don't do their maltings anymore. They're done at scale with big drum malters. Like when you scale this up, you don't need to do all of these things anymore. But this is a part of what many things that makes Kilchoman so enjoyable. They use Diageo's yeast, the Maori yeast, which is a cream yeast, your favorite term. And they use 2 big mash tuns. These are stainless steel with a copper lid on them, and they each hold about 1.2 metric tons of mash at a time, running about 6,000 liters of water in 4 cycles to generate their wort, which then goes to 10 stainless steel washbacks. And because I had a scientist as a tour guide, soon to be PhD, she was very pro stainless washbacks rather than wooden washbacks.

Richard Campbell [02:29:13]:
This is a scientist mindset because stainless washbacks are easy to clean and get to a standardized state. So the science approach says you should always do it that way, where the craftsmanship of making whiskey says the wood imparts unique characteristics in your typically Oregon pine wooden washbacks. Naomi was having none of it, so I naturally took the opposite position and got to debate her with it. We had a great time. time. Fun argument. Very smart woman. It was very enjoyable.

Richard Campbell [02:29:44]:
They have 2 sets of stills, both relatively small. Their wash stills are 3,000 liters, spirit stills 1,600 liters. And then they barrel up. Most of their whiskeys are done in bourbon casks. Totally normal. In this case, they have an agreement with Buffalo Trace, the Sazerac Group. So that's where they get their barrels from. Saw tremendous numbers of Jim Beam and Jack Daniel barrels all over Islay and all through this tour.

Richard Campbell [02:30:08]:
They're the most common barrels going. So it was cool to see the Buffalo Trace barrels there. And then interestingly, if you do go pursue Loch Lomond, I have the 10-year. They make different editions every year. Now, I'll need to say this distillery has only been in operation for barely 20 years now, and their first editions weren't even 20 years, 2009. So there's a 5-year and an 8-year and so forth, different editions for different times. But the 10-year, they always finished in Oloroso sherry, but they vary year to year. Angie is quite specific about, we're not going to always make the same thing the same way.

Richard Campbell [02:30:44]:
So I'm going to use, in the callback to Naomi, I'm going to use my little mini Glencairn with my lanyard here for this lovely whiskey, which I've obviously already tasted. A little bit of smoke, a little of that peat. The peat that they do themselves, and this is one of them, is only 10 ppm. That's quite Typically, when you think about a Laphroaig or Lagavulin, you're talking about 50. Although there's a great conversation, and we'll do this at another time, about how much actually shows up in the whiskey in the end. There's how much smoke goes on the peat that goes onto the barley initially versus what happens through the washing, the mashing and the washing, the whole process.

Leo Laporte [02:31:27]:
There's a happy man.

Richard Campbell [02:31:31]:
Yeah, that was such a fond memory of going there and just this, the nice heat, that sort of tobaccoey but still that sweetness from the sherry. Like, it's got all the things going on, you know. This is a pretty whiskey, uh, and it's available in the US. You have to look for it. There are a few different editions. Some of them are more heavily peated than others, so be careful Uh, you're gonna look at somewhere in the $100 range from a Total Wine or one of your specialty shops. So it can be found.

Leo Laporte [02:31:59]:
Nice.

Richard Campbell [02:32:00]:
Uh, without a doubt. And, uh, this is only one of many stories of those 7 days I'm looking forward to sharing with you. Uh, this was one of the bottles we liked so much we bought it and brought it home.

Leo Laporte [02:32:13]:
Land of my people. I, I would love to do that, especially on a tall ship.

Richard Campbell [02:32:18]:
Seeing it from the water, brother, I can't tell you. That's so great.

Leo Laporte [02:32:22]:
Maybe they'll, uh, take us to Scotland next year again. Thank you, Mr. Richard Campbell, Windows Weekly's Whiskey Weekly segment. As always, you'll find Richard at runasradio.com. That's where, of course, he's got his great show Run As Radio and .NET Rocks with Carl Franklin. And, uh, he joins us every Wednesday as We get together, uh, 11 AM Pacific, 2 PM Eastern, 1800 UTC to do Windows Weekly. Paul Thurrott is at thrutt.com. That's his website.

Leo Laporte [02:32:51]:
Become a premium member for all the extra goodness, uh, including copies of his books. Now you can go to leanpub.com if you're already a member at thrutt.com and get them. Uh, Windows Everywhere, uh, De-Insidify Windows, and of course The Field Guide to Windows 11. Now with 26H2 all over it.

Paul Thurrott [02:33:13]:
Gross.

Leo Laporte [02:33:15]:
Ew. We do the show, as I mentioned, on Wednesdays, and you can watch us do it live if you're in the club, in the Club Twit Discord. But we stream live on YouTube, Twitch, X, Facebook, LinkedIn, and Kick. After the fact, on-demand versions of the show at our website, twit.tv/dubdub, on YouTube. And of course, you can subscribe in your favorite podcast player. That's probably the best way That's the way to get it and automatically get it. And you'll listen every day, every week, as these two opine on the goings-on coming out of Redmond. Thank you, Paul.

Leo Laporte [02:33:51]:
You going back to Mexico soon or?

Paul Thurrott [02:33:54]:
Yeah, supposedly next Thursday.

Leo Laporte [02:33:57]:
After Thanksgiving? Oh, you're gonna, you're gonna be there for Thanksgiving this time?

Paul Thurrott [02:34:00]:
No, we're gonna, we'll come home right before Thanksgiving.

Leo Laporte [02:34:02]:
Oh, back and forth.

Richard Campbell [02:34:03]:
Wow.

Leo Laporte [02:34:03]:
All right. Richard, how long are you gonna be home for?

Richard Campbell [02:34:07]:
2 weeks. Unfortunately, I'm gonna I'm going to miss next week.

Leo Laporte [02:34:10]:
What?

Paul Thurrott [02:34:11]:
How dare you?

Richard Campbell [02:34:12]:
The doc has asked that I get a colonoscopy and it's only available on Wednesday.

Paul Thurrott [02:34:17]:
So I'm not going to be able to— I don't see why you can't do this from the hospital or wherever.

Leo Laporte [02:34:22]:
Yeah. Just get a microphone and—

Paul Thurrott [02:34:25]:
That'd be fun.

Richard Campbell [02:34:26]:
Yeah. That's a different game.

Leo Laporte [02:34:27]:
No cameras though.

Richard Campbell [02:34:28]:
Definitely.

Paul Thurrott [02:34:29]:
Yeah.

Richard Campbell [02:34:31]:
Trouble.

Leo Laporte [02:34:32]:
Well, good luck with that. That's—

Richard Campbell [02:34:34]:
Oh, let you know the following week when I'm back home.

Leo Laporte [02:34:37]:
It's the things you drink The day before.

Richard Campbell [02:34:39]:
Oh yeah, no, the day before is gonna be a riot.

Leo Laporte [02:34:41]:
That's no fun at all.

Richard Campbell [02:34:42]:
Well, and the actual testing is being done on the mainland, so I also have to get over there Wednesday morning on the ferry.

Leo Laporte [02:34:49]:
I would get a hotel room on Tuesday.

Richard Campbell [02:34:52]:
No, I think I'll be fine, but we'll figure it out. And I'm not driving myself, to be clear.

Leo Laporte [02:34:57]:
You're a brave man.

Richard Campbell [02:34:58]:
Yeah, sure.

Leo Laporte [02:34:58]:
Well, you know what, everybody, you should do it. Uh, it's a good thing, important thing to do.

Richard Campbell [02:35:03]:
Yeah, we're glad you're taking it. When the doc says, I think you want— you should have this test, I'm like, I think I should have that test.

Leo Laporte [02:35:09]:
You're supposed to. I think you're supposed to do it every few years after 50 or 60.

Richard Campbell [02:35:13]:
Yeah. Yeah. Well, we're at that age.

Leo Laporte [02:35:16]:
Hmm. Nothing like it.

Paul Thurrott [02:35:17]:
That's special.

Leo Laporte [02:35:19]:
It makes me drink more whiskey is what it would make me do. Paul, Richard, thank you so much. Thanks to all of our club members for making this show possible. Thanks to all our winners and dozers for joining us every week. We will see you right back here next Wednesday with Paul Theroux solo. Actually, it was a lot of fun last week. I don't mind doing it with you solo, but if you wanted to get somebody else, Like, I don't know, somebody likes beer, somebody lives in New York City.

Paul Thurrott [02:35:45]:
I don't know, I'm just brainstorming here. Yeah. On a clear mind in the morning. Yes.

Leo Laporte [02:35:54]:
That's better than I'll have in the morning.

Richard Campbell [02:35:57]:
MJ just had a recent article I thought was really great.

Leo Laporte [02:36:01]:
It's more her, I think. I don't think Paul's I think it's more Mary Jo's resisting. Yeah, she says she's done with the podcasting.

Richard Campbell [02:36:10]:
Yeah.

Leo Laporte [02:36:11]:
All right, Paul. All right, Richard. Thank you so much. Have a great week. We'll see you next time.

Paul Thurrott [02:36:15]:
Thank you.