Hands-On Windows 143 transcript

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

0:00:01 - Paul Thurrott
Coming up next on Hands on Windows. Clippy is back. Well, not exactly, but there is something called co-pilot vision, which will feel eerily familiar.

0:00:14 - Copilot
Podcasts you love.

0:00:16 - Paul Thurrott
From people you trust. This is Twit. Hello everybody, and welcome back to Hands on Windows. I'm Paul Thrott and this week we're going to take a look at one of many new Copilot features called Copilot Vision Sort of the return of Clippy. You'll see what I mean by that soon.

Copilot is getting complex. We've talked a lot about Copilot this year. We've talked a lot about AI, all the new features that are coming, um, but one of the weird things about Copilot is that it's it's in a lot of places right, so I tend to focus on it in windows, which makes sense. Um, it is an app. That app has changed a bunch, but Copilot is also on the web at copilotmicrosoftcom. It's part of Bing search. Now it's also Bing search is part of Copilot. Now there's a whole mess there. There's a standalone Microsoft 365 Copilot app that's in Windows 11, but it's also in the individual office apps right On Windows and the Mac. Right, it's on mobile. So there's a standalone Copilot app Microsoft 365 Copilot app, and then other Microsoft apps that have copilot capabilities, and so once you get into this weird matrix of copilot implementations, sometimes it gets hard to know where certain things are right. So Microsoft introduces these new features and you're like great, well, where do I get that? It's not always obvious, so I'm going to bring up the Copilot app in Windows to show you just one example.

So in Windows, if you go into settings, one of the unique features you'll see here is this phone connection feature. So in Windows only I can have Copilot. Give me information that's based on things that are on my phone right, in addition to things that are on Windows or on the web or wherever else. So that's kind of interesting. There's this voice mode where you have different voices to choose from. This is kind of a minimalist user interface. If I choose a different voice, it doesn't give me a sample and it also doesn't have other options that are available in other places. So, for example, if you were to bring this up on your phone in the Copilot app, you would have the opportunity not to just try different voices, but you'd hear those voices and you could control the speed of those voices, so you could speed them up or slow it down, which actually, I have to say, is kind of a useful feature. So for some reason that's not in Windows yet. So one feature that's like that it's brand new, announced recently and it's just starting to roll out as I record.

This is something called Copilot Vision, and this is, in some ways, copilot, as I think people think of it when you think of what a Copilot is something, someone or something, in this case, that is next to you, helping you as you do your thing. To date it's been a user interface, so it's a pane or whatever that you see on the side of another app. Maybe In this case, you're interacting it with your voice and so, in addition to typing queries and asking questions, you can actually just use your microphone and talk to it. Right, and that you know, depending on how you do things, that might be better or worse. You know, it might be just a different way to do it, so it's just a different way to look at it.

This particular computer does not yet have Copilot Vision, but if it did, I would click this and then I could talk to it, right, so I could say hello, copilot, oh, I didn't do it, oh, I did do it. It's not listening. Anyway, I'm just going to say not going to worry about that one, because I know it doesn't work on this particular computer, but one place I do know that it works on this particular computer is in Microsoft Edge, so Edge has this Copilot pane. That's on the side. Start a new conversation here. And it supports Copilot vision's on the side. Start a new conversation here and it supports Copilot vision. Hey, paul, all right. So it comes up, it asks, it says hi, hi, hello, copilot, talk to it.

0:04:15 - Copilot
Absolutely, Paul. We're having a conversation.

0:04:18 - Paul Thurrott
Okay, I'm going to cut them off because seriously. Okay, so what are the types of things you might do with Copilot? Right? So I have. Let's see if I can find this article. I wrote an article recently about a NAS that I'm purchasing and why I'm getting it and so forth, and so I could go into Copilot. Actually, let me turn this off.

Sorry, I could go into Copilot normally and create a summary is right there, right? So summarize the main points of this page, and so it's going to look at this article that it's right next to right, side by side, and then give you information about the article. Okay, that's cool, that's text-based, right. So I clicked a button, but typically I would type in that prompt and then it would answer the question, right? It's giving me a nice little hint here about what I can do, but with copilot vision I can just ask hi, paul, hello copilot, what is this article about?

0:05:15 - Copilot
hey, paul is talking about the author's journey in setting up a network attached storage device. They've gone for a Synology Distation DS 224 plus and with the RAM up.

0:05:25 - Paul Thurrott
Thank you, that's good enough. I'm just going to cut it off. It would go on for a little while there, right? And then one thing you'll see every time is this kind of needy hey, how was it? Was I good? And it wants this feedback. It's not much, as you can say, but it's usually pretty good, so that's fine. I don't have Copilot, or rather Copilot Vision, in Copilot, the app in Windows yet. But if I did, I could also bring up, you know, an image and say what is this image? But I can do that. I'll just do it on the web. You get an idea of how that can work. Hi, can you tell me what this is a picture of?

0:06:00 - Copilot
Sure thing, Paul. This image shows a sleek laptop, likely a Microsoft Surface, known for its detachable keyboard and touchscreen capabilities. It's looking pretty stylish with that colorful display, isn't it? And touchscreen capabilities. It's looking pretty stylish with that colorful display, isn't it? Okay, thank you.

0:06:13 - Paul Thurrott
That's good. It's a little super positive about Microsoft there, but that's fine. That's fine. So, when you think about what AI is really good at these days, I think that summarizing is right near the top of the list. So it makes sense, you'd see, create a summary here in the copilot pane. You could do this in the app when you do run the app. By the way, I don't think I showed this earlier, but one of the things you can do is go into this quick view, and the nice thing about quick view is it's over the top of other things, because the copilot interface itself can be kind of big on Windows and this is a smaller UI. It lets you interact with something else. Now, I don't have copilot vision, so I can't do that yet in this interface. That will happen very soon and it might have already happened on your computer, depending. So summarizing right. So I used it to summarize.

This article started. It got going on that. Okay, that's pretty good. This is an article about extended security updates for Windows 10. We did an episode about this pretty recently. This program is typically for businesses, but in Windows 11, I'm sorry, windows 10 for the first time they're allowing individuals or consumers to do it for one year only, right, and this article is mostly about businesses, but it does mention consumers, right? And so I could go in here and ask it about that.

0:07:38 - Copilot
Hey Paul, how's it going?

0:07:39 - Paul Thurrott
Good. Could you tell me what parts of this extended security updates program article concern consumers?

0:07:49 - Copilot
Sure thing. This article primarily talks about the extended security updates ESU program for Windows 10, which is mostly for businesses and organizations, but for consumers, the key point is that support for Windows 10 ends on October 14th 2025. After that, the ESU program allows consumers to get security updates if they choose to continue using Windows 10.

0:08:13 - Paul Thurrott
There's also there you go. So that bit was right here. You can see that text that he basically read there. But then he also went through this. I'm calling it a. He already right, but the AI looked at this article and it jumped down to this bit about about consumers, which I'm having trouble finding because it is hard. So it says, for individuals or Windows 10 home customers, extended security updates for Windows 10 will be available for purchase at $30 for one year only, which would eventually be part of that summary. So that's actually really useful, right?

So this is not a particularly long article, but it's kind of long, and if you were just looking, well, what parts of this apply to me, that's a nice way to. So it's not just a summary, but it's a very specific kind of a summary, if that makes sense, right? So we've got summaries for articles, like we did with these two things. Summary is kind of a weird term, but basically, you know, described in an image, using image recognition. You can do this with PDF files, right, but you can also do it for videos. So when I switched this video, this is Dave Plummer. This is a former Microsoft engineer. I love his channel. I watch every one of his videos. You can see summarize this video as one of the options, but I can also ask Copilot Vision to discuss this. Instead of getting a printout or, you know, like a text-based response, can this thing summarize this video for me? Hey, paul, hey, copilot, can you tell me what the top five points of this video are?

0:09:50 - Copilot
Sure Happy to summarize it, Paul. First, the video dives into the origin story of the Windows start menu, tracing its evolution from the early days of Microsoft's operating systems. Okay, thank you.

0:10:02 - Paul Thurrott
So it would have kept going, obviously, and that's great, right, so that's useful. But how does it do this? Right? This is only a 14-minute video. I would just watch this, frankly, and I would watch all of his videos. I really like this channel, like I said. But how is it doing this? You know it turns out that it's just doing it the same way. It summarizes any document, because every video on YouTube basically has a transcript. So it's actually just doing the same thing. So, whether it's this transcript from the video, a web article of whatever kind, a PDF file, it's using the same type of AI to examine the text, provide a summary and then it can tailor it to how you described it.

I said in this case, I wanted the top five points. I said with that Microsoft Learn article about extended security updates, that I wanted only the bits that were related to consumers, and so it can kind of dive in and pull out those bits. So you know it's useful. I could have done this like this. In fact, I will. If you prefer the text-based approach, that's still part of it. So that's cool, and once this comes online in the Copilot app on Windows, I would be able to do it from there as well. So this is nice because it's built into Edge. If you don't use Edge, use Chrome or Brave, or whatever browser you might use. You can still use the Copilot app to get the same functionality. You can do it with text, you can do it with Copilot Vision. It's up to you. So pretty powerful and useful, right, and also one of dozens, if not now hundreds, of features that are part of Copilot that are all over the place.

This particular one, I believe, is only in these two places. I believe for now it's just Edge and Windows, but inevitably it will probably appear in other places where Copilot can be found, maybe even someday on our phones, although I think that the big screen that you get on a computer is particularly good for this kind of implementation, because it's nice having those two things side by side, right? I'm not much of a talker, so I don't really interact with the AI willfully, but I bet a lot of people will, and it is kind of strange, like how you fall into this pattern with it where you apologize for interrupting or you say thank you or you address it as if it were a person Like this is. We're getting into a weird space here, but that's what's happening. I mean, that's what AI is doing to us or for us, I guess, I don't know. So hopefully you found this educational. Certainly, the co-pilot responses in some ways were hopefully at least a little entertaining.

We will have a new episode of Hands on Windows every Thursday. You can find out more at twittv slash how. Thank you for watching. Thank you especially to our Club Twit members. We love you, as always. If you're not a member of Club Twit, do consider joining. You get these videos without any ad breaks, but also you get to support a great company that's making a lot of great content and it's good for your soul, so consider it. You can learn more about that at twittv slash. Club twit. Thank you so much and I'll see you next week.