Leo [00:00:25]:
So Gary, I've had an interesting week. I have lots of little anecdotes to share this time.

Gary [00:00:31]:
Okay.

Leo [00:00:32]:
It's, it's been kind of fun. As it turns out, my wife spent the weekend up visiting her sister-in-law and her sister-in-law drives— honestly, it's my mother's old 2000 Honda CR-V. And of all things, The parking brake froze up, basically. It's one of those with the pull handle between the front seats.

Gary [00:01:00]:
Oh yeah, the old kind. Yeah.

Leo [00:01:01]:
As best I can figure, something rusted, you know, some piece of metal rusted on some other piece of metal and it just, it wasn't happening. Of course, she's up there. I'm down here, 100 miles apart., and, um, she messages me and, you know, basically says help. Um, so, but, you know, I wonder, um, I went online and searched for 2000 Honda CR-V maintenance manual because I don't know if you've ever seen them. Uh, back in my many much younger days, um, it was all often a very useful thing to purchase the maintenance manual for the car you were planning to keep for a while.

Gary [00:01:49]:
Awhile.

Leo [00:01:50]:
Yeah, my old— and my maintenance manual, I mean like a 1,000 or 2,000 page book that you would find, um, actually in the mechanics' shops, um, you know, for them to do.

Gary [00:02:02]:
Yeah, they didn't come with the car. No, no, to go you had to— that wasn't the manual you get with the car, which is a weird little document and continues to be in 2026 a weird little document you get. Yes, this was like— yeah, you would get— and they used to be bookstores used to carry them too. They used to be— that's right, I remember that. Yeah, like sections in the back of bookstores that, you know, there'll always be a guys looking at the manuals for different cars.

Leo [00:02:24]:
I remember having to order them, you know, mail order them, get them from someplace. I mean, they were not, you know, they were not the easiest thing to get. However, in 2026, that turned out to be a very easy thing to get. The second search result was for archive.org. And indeed, on archive.org was a full scan of the 1,400-page service manual for the 2000— it was actually the 1997 through 2000 Honda CRV. Perfect. I was absolutely thrilled that that just sort of showed up. And sure enough, you know, I fire it up.

Leo [00:03:10]:
It's a PDF. I did a little bit of browsing through their interface and then I decided I would download it., which I could do, and just use the PDF reader I've got to do some searching. And I was able to actually do a quick screenshot and send my wife a picture of, okay, here's the brake handle. Here's how you remove the COVID Here's where you squirt a lot of WD-40. Right. Um, and you know, in the long run, we got, you know, we got the thing working. Um, we ended up doing a little bit more than just oiling it. But the bottom line is that the information, the ideal information we were looking for was like a few keystrokes and a click away.

Leo [00:03:57]:
And I got to say, this is not the first time that Internet Archive has shown up just randomly for just really random things. I think we often think of Internet Archive as being primarily the Wayback Machine. Yeah, which is something very coincidentally I was just writing about before we, before we started recording. And it's so much more than that. I mean, our friend Kay, former co-host of this podcast, is spending a lot of time updating amateur radio archives, which technically, is that really Internet Archive? Well, yes, no, maybe, certainly not. You don't expect the repair manual for a car, a 26-year-old car to be something you would find there. And yet it's this wonderful treasure trove of just random information. Um, and I think you and I have both mentioned that we both support, um, the Internet Archive financially to some amount, um, just because we do, you know, we use them.

Leo [00:05:05]:
In various different ways at various different times. Like I said, I'll be talking about it a little later in the show about how I'm actually using it as a tool to basically vet websites. So anyway, that was just an interesting remote diagnostic thing where Internet Archive popped up in the, in the search results. Um, very coincidentally, we had another story of, again, random things just sort of coming together. This is a more interest— this is an interesting one where, again, my wife plays a central role. She is driving around our neighborhood, notices that there's a hawk sitting on one of the fences not far from here. And takes a picture of it. I have links to the pictures that I'll be talking about in the show notes.

Leo [00:06:07]:
It's, you know, a fine image. It is automatically uploaded to her Google Photos just because, well, you should automatically back these things up anyway, but it's the primary way for her to take a picture and then have me be able to do something with it. I grab it. It is a very, you know, it's a fairly high resolution photo. It's 4,032 by 3,024. Lots of opportunities for cropping, which is what's necessary. The full photo actually includes the hood of her car and a bunch of the neighborhood, but yes, there's a hawk in the middle of it. So I'm able to crop it down and actually get a pretty reasonable image of the hawk.

Leo [00:06:51]:
But then I decide, you know, I've got this tool that I play with from time to time. It's Gigapixel AI, although I think they just call it Topaz Gigapixel now. And with a little bit of, you know, adjusting the settings and letting it make some assumptions about what kind of an image it's working on, I actually upscale the image by a factor of 4. This is not quite, but in some senses it's the enhance command you keep seeing in, in, in TV shows that of course we're all screaming back here that no, it doesn't work that way, it doesn't work that way. But now that we're applying AI to these problems, in some ways it kind of sort of does. It's also Not bad. It does a good job of enhancing the image. Um, there are definitely some telltale signs if you know where to look.

Gary [00:07:54]:
I could see— I'm looking at them now. Yes. Yeah, interesting. The sign underneath the hawk.

Leo [00:08:00]:
Exactly. The sign that was somewhat legible in the original is kind of messed up.

Gary [00:08:10]:
Yeah, I could clearly— I could see that it says All Seasons, probably the name of the fencing company since it's on a fence. But I just barely could see. Oh, okay. Fencing company. Oh yeah. All Seasons. But then enhanced, it's basically an alien language.

Leo [00:08:26]:
Exactly. However, I mean, and that's part of the assumptions that you allow a tool like this to make, right? You tell it that this is a photo and not line art. You, you tell it that it, you know, it's kind of looking at this kind of stuff and not that kind of stuff. And it ends up making assumptions. That's how these upscalers all work. You know, the more information you give them about the context of the photo, the better the assumptions that they can make. Anyway, so, you know, that was kind of cool. Gave me a nice picture, a better picture of the bird.

Leo [00:08:58]:
Then my wife asked, okay, what kind of hawk do we think that is?

Gary [00:09:04]:
Yeah.

Leo [00:09:06]:
I'm not a bird guy. I have no clue.. But then I remembered that ChatGPT, of all things, identifies photos. So I said, sure. I ended up uploading the upscaled version, which honestly was a bit of a risk because it's going to have upscale artifacts and that could affect ChatGPT's ability to identify. Nope. It nailed it. Quoting likely ID, a red-tailed hawk.

Leo [00:09:34]:
Juvenile, and then a description based on the bulky buteo shape. I've never seen that word, but it's apparently the style of hawk. Um, overall brown upper parts, the pale checkerboard, um, pattern on the wings, the strongly banded tail common in young birds before they develop the classic red tail. This looks most consistent with a juvenile red-tailed hawk. And indeed, we do have red-tailed hawks around here. We've seen them before. We typically see the, the adults as they're flying overhead. You are able to see the red tail very prominently.

Leo [00:10:11]:
I just thought it was pretty amazing that, um, it just, I did it as, yeah, this is a juvenile red-tailed hawk.

Gary [00:10:19]:
Yeah.

Leo [00:10:20]:
What amazes me about this, I mean, this is one of those, you know, what a time we live in., we've got an image taken on a phone. It's automatically uploaded to the cloud. I'm able to grab it and crop it. I throw it at some upscaling software. I then throw the result of that at just this random AI, to give me more information about it. And all of these steps, one after the other, just work. It's fun.

Gary [00:10:50]:
Yeah, it is. I, now Apple on the iPhone has had a plant and bird identification feature for for a while. Like before we were talking about— before ChatGPT was something people were talking about, they had like an AI kind of model already going on quietly, like low-key. It was just like the, you know, the identifying thing. And you were able to use it and you still can. And I commonly do that. When I see a bird, I'll either snap a picture of it or you don't have to snap a picture of it. You can actually use something that they've got called visual intelligence, which basically brings up the camera and you point it at something and it tells you what it is without you having to take the photo as like an interim step.

Gary [00:11:34]:
And yeah, it's actually turned me into kind of a bird guy. I mean, I'm surprised you're not a bird guy as a guy of a certain age with a camera, you know, and because I wasn't a bird guy probably 2 years ago. I couldn't, it was everything was, there were 2 types of birds, pigeon, not a pigeon. And then at some point, I guess I was using this feature or whatever. And one of those not a pigeons turned out to be something kind of interesting. And what was more interesting was that another bird I saw a little later was like, oh, there's one of those birds again. That was nothing. It looked at first to be the same kind of bird.

Gary [00:12:14]:
It was not, not even close. Once I identified it and saw, oh, it's a different type of bird, then I looked at the two pictures and said, well, yeah, obviously now I could see it's a different type of bird. And then I started looking at other birds and was amazed that my little park, or not really little, but my big park that's near my house, I could go over to it and almost every time I see a bird, it's a different kind of bird. And then you start looking into the stories and you find out these are different ranges. This is a type, you know, evolutionary-wise, this is a whole different gene, you know, you just go into all this stuff and it starts to suddenly become fascinating. And the next thing you know, your brain gets retrained and you're spotting birds where you hadn't spotted birds before. It's like, why did I just see 3 different species of birds on my walk home? Usually I walk the same way and I see zero birds. And it's because I'm noticing them now.

Gary [00:13:05]:
And then I'm starting to notice the calls, you know, because I'll read the description and say its is, you it'll call know, it makes it, it's a songbird and it does this, this. And I was like, yeah, that you describes, know, so now I'm getting into it because of all of that. And you could do it with photos that you have already. So like when I took my trip to Africa, I was able to identify all the birds that I hadn't been able to identify while I was actually like taking pictures with my good Sony camera. Now they were in my Photos app, you know, on my Mac. And when I look at the photo, it comes up and there's a little button, I click it and it says, this is a whatever type of bird.

Leo [00:13:40]:
Not a pigeon.

Gary [00:13:41]:
Not a pigeon. And it's fascinating. So yeah, it's a cool feature.

Leo [00:13:46]:
The bird IDing we've been doing here, because we definitely have lots of birds around the house, but we've been doing it old style. We've got The book, right?

Gary [00:13:55]:
Yeah, I have one.

Leo [00:13:58]:
Too. And we'd look through the book and see if we can identify the thing. However, a few years ago, we discovered— I'll back up. You're familiar with Shazam, right?

Gary [00:14:11]:
Yeah.

Leo [00:14:12]:
Where Shazam is an app where you're in a busy restaurant. There's some music playing over the loudspeaker. And it sounds familiar. So you fire up Shazam. And it filters out the noise and tells you what song is playing, and then I think it actually offers you 2 or 3 links to go purchase the song, from which I'm sure they get a cut.

Gary [00:14:34]:
Actually, it's owned by Apple now.

Leo [00:14:36]:
Shazam?

Gary [00:14:37]:
Yeah. Really?

Leo [00:14:38]:
I didn't know that.

Gary [00:14:39]:
Yes, it is Apple, fully owned Apple company. It's built into the iPhone now in several different ways. And it'll, I don't know what it does on Android, but it'll like— It works.

Leo [00:14:50]:
Does It a full buy the.

Gary [00:14:51]:
A.

Leo [00:14:51]:
Song.

Gary [00:14:51]:
Link to buy the song and a link to add it to your Apple Music playlist.

Leo [00:14:55]:
Nifty. So we decided a few years ago, wouldn't it be nice if there was such a thing for birds?

Gary [00:15:02]:
Yep.

Leo [00:15:03]:
There is. Um, it is called Merlin. It is a product of, I think it's the Cornell lab, uh, the ornithology lab out at Cornell. And that is what we've been using a lot. Um, basically much like Shazam. You fire it up, you let it listen, and rather than identifying the sound and stopping, it just keeps identifying because of course, especially like in a busy, busy spring morning, there's lots of different birds and there's lots of different sounds going on. And it's just saying, okay, you know what? That's a, that's a junco and there's a crow in the distance and there's a sparrow nearby and you know, all that kind of stuff. And it's just showing you this stuff in real time, identifying the birds that are happening.

Leo [00:15:52]:
It's absolutely fascinating. That's one where, yes, I've turned into that guy, right? Where we will absolutely— we hear a bird we don't recognize, I have to whip out the phone and I've got a shortcut to Merlin on the homepage because I need to get to it quickly. Some years ago, the reason this actually came to be in the first place is one of our corgis would occasionally be outside and just freak out. I mean, and they would just be sitting there quivering in fear. And we had no idea what it was. And make a long story short, what we discovered is it was because the, um, our hummingbirds, I think it's the Anna's hummingbirds that we have here. They have this mating thing where they fly really high. They then dive bomb, and at the very bottom of that flight, they make this very, very high-pitched peep.

Leo [00:17:01]:
The only way I can describe it. And that freaked out the dog. Don't know why. I was able to actually find a recording of it online, and I played it here in my office., and the dog freaked out. So we knew that that's exactly what it was. Um, and then yes, you know, Merlin came along and identified and says, yep, you know, that's this hummingbird doing that and that kind of thing. Um, the good news for the dog is that as she got older, she got deaf. She didn't, didn't hear it anymore.

Leo [00:17:29]:
But, um, but yeah, that was interesting. So again, another amazing, uh, fun technology, um, to just have in your pocket and be able to identify the birds that are nearby you. The only thing that's a little unique about, um, about Merlin is that when you fire it up, it does ask you whereabouts you are. So I've got the database for like the Pacific Northwest or the northwest corner of the country loaded. I don't know if it would actually identify the things that are in, say, your area, but I also expect that there's a pretty easy way to say, okay, fine, load up the database for, you know, Colorado or this area. So yeah, interesting stuff.

Gary [00:18:10]:
Yeah, indeed, indeed.

Leo [00:18:13]:
All right. Yeah, I've touched on AI. Let's go deeper. Have you played at all with Claude Code?

Gary [00:18:22]:
Not with Claude Code, but I have done a lot of coding with AI, usually with ChatGPT. Right.

Leo [00:18:29]:
So you're just asking it to write some code.

Gary [00:18:30]:
Yeah.

Leo [00:18:31]:
It's interesting because when you think of Claude Code, You think, okay, it's Claude, which has a very good reputation for writing code. So that's what this tool must be all about. I haven't written a single line of code with it yet. I've been using it for something else. Since it's installed on my PC, it is able to work on data that is local on my PC., which is fairly unique. Normally when you want to, like, you've got a collection of data that you want AI to work on, you basically grab the files and you upload it and then start doing things with it. It's what I did with NotebookLM for, you know, when I had my fall, I loaded up all of the raw doctor's notes into NotebookLM and I was then able to ask questions against those notes and have, you know, my questions be simple English, the responses be simple English, even though the doctor's notes were anything but. So it was very, very interesting, an approach to the technology.

Leo [00:19:43]:
I did the same thing with Ask Leo. I actually managed to upload all of the Ask Leo content into NotebookLM. Unfortunately, because of size limitations, I had to break it down into 22 separate markdown files and upload them one at a time. And again, was able to ask some interesting questions against the content of Ask Leo. Claude Code is letting me do that without the upload. It's all happening here on my PC. I will say that it is absolutely a non-trivial install. It is not something that you just download something and install it, and then all of a sudden you've got Claude Code.

Leo [00:20:31]:
You've got to set up a couple of prerequisites in Windows itself. You need to install the Windows Standard Linux Library. Essentially, you're installing a version of Ubuntu Linux in Windows so that when you're actually using the tool, you're actually using it in what boils down to a Linux Bash shell. So it's a command line tool, but it assumes essentially the Linux interfaces. And it works. It's, it's actually, you know, kind of cool. So I had a couple of different scenarios that were fun to play with. I use Obsidian as my note-taking app.

Leo [00:21:15]:
It's just a big pile of random stuff, daily notes about what I've been doing, random ideas about articles that I need to write, half-written articles, both for Ask Leo and for my personal blog and so forth. It's just a big pile of stuff, as is probably common for a lot of people's Notion.

Gary [00:21:38]:
Sure, yeah, for me definitely.

Leo [00:21:40]:
The reason I absolutely love Obsidian is that its native storage format is Markdown. Markdown is a text format. These are all text files, and the Markdown part of it is that, you know, there's some very simple typing techniques you use to do things like headers and bold and tables and, you know, all the kind of formatting stuff that even if you don't render it, if you just look at the plain text file, it still makes sense. It still is readable. When you actually render it, then yes, you get the pretty formatting and the tables and all that kind of stuff. But the bottom line is it's fundamentally text, which makes it ideal for consuming by other applications like Claude Code. So I did that. And the way I did that is I said, hey, Claude, My Obsidian vault is in this folder on my PC.

Leo [00:22:37]:
That's all I needed to do because then I could say, okay, summarize it. Tell me what kind of stuff I'm writing about. You know, tell me what's, what's going on in that vault. What's, what's common, what's not. And it actually did a very, very interesting job of identifying random things. In the vault that I probably would not have recognized. The big thing it's good at is themes. Themes or consistencies or things that are, you know, pop up a lot throughout a lot of the content there.

Leo [00:23:12]:
So that was interesting. I can now ask random questions of my vault. My vault, by the way, also contains whenever I highlight something in Kindle, I have that hooked up to Readwise so that the highlights show up in Readwise, and I have Readwise hooked up to Obsidian. So all of those highlights now also show up in my Obsidian vault. So it's a great way for me to snag a quote while I'm reading and then know that the next day it's gonna show up in, in, in Obsidian for me to operate on. So all that stuff is there. Okay, so that was an interesting one. Like I said, I can now ask questions against my vault.

Leo [00:23:53]:
You know, I knew I wrote about this, where's the quote for that, that kind of stuff. Then I decided, you know, I have this personal blog. I've been blogging for probably 15, 18 years, something like that. There's a bunch of articles out there. I wonder if I could point it at the blog. So I did. I said, okay, Claude, go out to leo.notenboom.org and summarize what you find, which is— those aren't the exact words, but that's pretty close to what I did. And it did an okay job.

Leo [00:24:33]:
The problem is that it didn't list, it didn't take the time to read absolutely every post on the site, which was honestly disappointing, right? I mean, I would have at least liked it to have said so, but it picked some random selection of articles and gave me a summary of that. Okay, then I got kind of a light bulb moment. I use WordPress for my blog. It has an export function that I have never used because it exports in a very dense XML format that I don't know is used anywhere by anyone. Yeah. However, I wonder, you know, Claude can figure this stuff out.

Gary [00:25:28]:
Yeah, definitely.

Leo [00:25:29]:
So I downloaded my entire blog, all of the articles. I guess it's since 2004, so it's been 22 years, into this one file, this big old.xml file. And I told Claude, okay, give me that. Give me an analysis of what's going on in there. It did a couple of really interesting things. It realized that this was the content of the online blog that I had just asked it to process. And it said, oh, great. Now, I mean, it literally said something to the effect of now I can paint a much more complete picture of what is going on in that blog.

Leo [00:26:15]:
Yeah. And it did. It gave me, you know, the top themes that I like to write about, how it's changed over time. You know, obviously things like average length and average frequency and all that kind of stuff. When my most prolific period periods were, all that kind of stuff. It was really, really interesting. And again, now I have this approach that will allow me to— okay, let's ask some just completely random questions against the contents of my personal blog. Okay, this is good.

Leo [00:26:54]:
It was actually pretty interesting and pretty impressive. My third scenario is since I already had those Markdown files that I had used for NotebookLM for all of the Ask Leo content since 2003, I said to Claude, okay, here they are in a folder. I mean, there's literally just a folder with Markdown files for each year of Ask Leo's content. Give me the summary for that. And this time I have a link to that summary in the show notes because it's absolutely fascinating. It is correctly identifying all sorts of interesting things about Ask Leo. It starts out you with, know, some simple metrics. Like there's 5,900 articles on Ask Leo.

Leo [00:27:59]:
It's like over 4 million words. The peak year apparently was 2012, although the peak year by, by the length of articles or by the number of words is 2025. It's identified the most common topics. It's showing the evolution of the themes on Ask Leo over time. I like this concept of the core commandments, Leo's core commandments that it identifies, the gospel of Ask Leo. But then it comes up with this really almost snarky comment at the end. If Ask Leo had 10 commandments, thou shalt back up would be the first, second, and third. Right.

Leo [00:28:37]:
It's, it's really, really interesting how it goes through and has analyzed very, very accurately exactly what's gone on in Ask Leo over the last 20-some-odd years. So yeah, like I said, it was not a trivial install by any means. It's not something for the faint of heart, but man, once you get it installed, there's some really, really interesting things you can do with it. I've only scratched the surface here because in all 3 of these cases, be it my Obsidian vault or my personal blog or the Ask Leo, all I've really asked it for is summaries. Right? I've said, you know, give me a summary of what's there, which honestly is what AI is really, really good at, giving it a bunch of content and then asking it to act on that content. What I'm wanting to do next is ask it some more, I don't know, generative questions like, okay, what are some topics that I'm missing? What are some things that I could, that I've only skimmed the surface on that I could go deeper on? You know, those kinds of questions that would help me perhaps decide what content I should be writing moving forward based on how successful some of this content has been in the past. But yeah, it's fun. I honestly, anybody that has a body of work, it's really worth taking that body of work, throwing it against something like Claude and having it just summarize.

Leo [00:30:19]:
You will come up with insights that you did not realize.

Gary [00:30:23]:
Yep, I've done similar things with my stuff because I have transcripts for all my videos and I've, you know, thrown a lot of that at it and then asked questions against that body of work and just even just the list of titles for questions such as you know, like, which video, you know, which topic should I revisit, you know, and that kind of thing, because it's got dates and the video titles that it could figure it out and be like, oh, you should probably, you know, go back to this, it's changed recently or something. It's been really cool. I wasn't even going to mention this, but this was something interesting I did this morning that took me 15 minutes, should have taken me 2 weeks. It's not Claude code, but I used ChatGPT and I woke up with this idea. Somebody asked me a and the— question I, I have this forum on my site. People could ask me a question and then I could respond and I could press a button and it sends them an email response or I press a different button and it and it sends posts a, you know, something to my site with the question and my answer. So then I build up, you know, this thing. People can ask me questions and then other people can read the answers.

Gary [00:31:37]:
But a lot of times people don't ask a very clear question or it's very specific. So I just hit the email button, email back. And the problem is every once in a while I get one and like, I just answered this. Like I think maybe a week ago the same person asked the question. Maybe they didn't check their email. Maybe they didn't see that there was a response posted, whatever. To figure that out is really hard because it's done in this form. I was smart enough that 10 years ago when I set this up, every time somebody asked me a question, it gets saved as a little text file in JSON format, which is like a pretty human-readable format, but it's getting curly quotes and commas and everything in there.

Gary [00:32:17]:
And then when I respond, it puts a text file up with like the first line's the title and the rest of it's my you response, know, to to the, the thing. And it saves all that in kind of this log of files. Which is now filling— it has thousands of text files in this directory. If I need to look something up like, did this person ask me this question like a week or two ago, or was it a similar question I should be pointing them to or whatever, I got to go into this directory of text files on my server and start poking at each one. Is it this one? No, was it this one? Whatever. So I thought, should— you know what, an easy thing to do is have a page that just compines all that information, throws it up into a web page, and I can search the web page. It's just a big web page. I was going to— okay, it's just going to be a loop.

Gary [00:33:03]:
It's going to go through, it's going to date. I said, you know what? Let me throw this to ChatGPT. I said, give me a PHP page that just gives me a summary of all these questions and answers, and these the are two types of files, and here are 10 samples. Of these. I just threw— gave it 10 example text files. And then I— and I gave it literally 3 paragraphs of description of what I wanted. And ChatGPT said, you know, on it, and then started thinking. And then I went and did something else.

Gary [00:33:40]:
And 5 minutes later, I looked back and there was an entire PHP script there. So I uploaded that to the site, ran it for the first time. And what I saw was incredible. First of all, it looked beautiful. White text on a black background, little boxes containing everything with like rounded corners, titles in a different, you know, nice bolded font, little rules with lines, little tags showing different things for everything, a search box at the top that allowed me to search this entire thing. And it took took the JSON files and my text files, figured out what those were, combined all the information. And it was like every question, what they asked, all the information provided in a nice format. Here's my answer.

Gary [00:34:29]:
Did I answer by email? Did I answer by post? This whole thing was incredible. And I asked one more thing. I said, well, let me instead of 7 days, I said, give me the last 7 days. I said, allow me to just give me 7 days and then have a number at the top. And I could change the number. And it would fetch 30 days or whatever. I said, sure, and 5 minutes later generated an update to it. I updated it.

Gary [00:34:51]:
It looks better than anything I could have generated in 2 weeks. I was just blown away. It's way more than I asked for, way more than I asked for. I mean, there's CSS in here, there's tons of HTML and JavaScript for the searching is all in there. It's just incredible. I just was blown away. It's like I almost decided to I could wait till Friday and dig into this and see what I could create a day. But instead, in 15 minutes this morning, I got way past what I needed.

Gary [00:35:20]:
And now I have this little bookmark when I go to and I could see like this complete history of all the forum stuff that I've done. And it's going to be so useful.

Leo [00:35:30]:
You are now officially a Vibe Coder.

Gary [00:35:32]:
Yeah, and I've done tons of Vibe Coding. It's never gone this smoothly. Usually I'm asking for smaller stuff. Like something that's like 10 lines. And then I get it back and it's not quite right. And I go back and forth and I eventually get it right. I've never been like, this is a vague description of what I want. I guess the key is I didn't really want something extremely specific.

Gary [00:35:50]:
I was like, I just want something better than I have now, which is nothing. And it just delivered this huge, like, here's an entire— I've created an entire software project for you that does what you need. Anyway, it's really, really cool. And it's going to save me a lot of time and make my job make what I do better.

Leo [00:36:07]:
Nice, nice. But it's not— yeah, so two things. One, um, as you were talking, I was thinking, okay, fine, the other thing, the other approach to your problem would be to do essentially what this— the equivalent of what I just did. Yeah, download all of that information or upload it into a chat— into an LLM, and then just query against the LLM, you know, show me the— show me the ones that, that match this criteria. Something I forgot to mention about Claude, about Claude Code, is that I started by saying it, I never, I haven't used it to write code. But what's interesting is watching it work because it's actually kind of chatty as it's processing the various steps and doing what it's doing. Claude is writing Python. Internally to process the files that I've given to it.

Leo [00:37:05]:
And apparently, you know, taking the output of whatever the Python is, whatever those scripts are doing. I found that part also absolutely fascinating. And it wasn't just one, it was like, okay, we're going to write some Python to do this. And then we're going to write some Python for this analysis and do some different Python for that analysis. And apparently, I've never seen the Python. I have no idea if it's preserved or not. All I see are the results. But again, yeah, Claude code, it may not be generating code for you, but apparently it's more than happy to generate code for itself.

Gary [00:37:41]:
Yeah, cool. But of course we have to balance the, oh, AI is so cool with the, oh no, AI is going to destroy us all story. So I, like a lot of other people, I watched the Bad Bunny concert this weekend, which was really good.

Leo [00:38:03]:
Which had some kind of a game at beginning and end?

Gary [00:38:06]:
I didn't watch any of that, whatever that was. know, But, you and then there was a commercial. I didn't see it live, but I heard about it afterwards. Fascinating commercial. Commercial from Ring, which is Amazon. Ring makes those doorbell cams, right? And a variety of products around them. And they had a— I can't believe so many people looked at this ad and they didn't see this coming because the ad basically was an ad saying, oh, your Ring doorbells, we can now use all of those to help find lost dogs. With the idea being like, you lost your dog, then we can ping all the Ring cameras in your neighborhood.

Gary [00:38:51]:
They could look out onto the street and see dogs wandering by themselves and match it to your dog and help rescue your dog, you know? And they show— and at the point I'm watching, after I heard there was a thing about it, you got to see this video, this commercial, and I watched it. And then when they show the video of all the Ring doorbell cams, like sending out these little waves into the street, Like they're watching everything that's going on. I was like, oh, how did they not know people were going to react negatively to this? And at this point, it's only been 2 days and you could search for, you know, Ring doorbell Super Bowl commercial. And it's an endless scroll of news articles and blog posts about how bad this is for privacy and stuff. Basically using your Ring cameras to surveil your neighborhood and dressing it up as, oh, we're just gonna use it to find lost dogs. When of course everybody's already like, doesn't Ring already have like a whole thing going on with the government using it for surveillance for ICE and, you know, with Flock and all that. And so everybody was already thinking like on the edge with Ring and they put this commercial out there, and it looks like it's a disaster for them, which good. They— I've seen a bunch of stuff.

Gary [00:40:14]:
You, of course, showed a really interesting video about it.

Leo [00:40:20]:
So, yeah, I follow We Rate Dogs.

Gary [00:40:22]:
We Rate Dogs. I've seen that. Yeah, yeah. We Rate Dogs.

Leo [00:40:23]:
Yeah, I love their stuff.

Gary [00:40:26]:
Yeah.

Leo [00:40:26]:
But he did a specific video about this, about this. Yeah, it was good. And he nails all the issues.

Gary [00:40:34]:
Yep. I know. Yeah.

Leo [00:40:37]:
The thing that surprised me, he threw out a statistic. I'm going to assume it's at least close to being legitimate in that, you know, supposedly the Ring network will find one dog a day.

Gary [00:40:50]:
Yeah.

Leo [00:40:51]:
Which is apparently 0.04% of the dogs lost on any given day.

Gary [00:40:57]:
Yeah.

Leo [00:40:58]:
It's like this really, really tiny number. Now I get it. If it's your dog that gets found, great. But yikes.

Gary [00:41:06]:
Well, and also that number, I mean, you think about it, if you're coming up with a statistic and that's the biggest number you can come up with, you have to balance it with how many of those dogs would've been found regardless.

Leo [00:41:16]:
There's that.

Gary [00:41:17]:
You know, there's always a thing. It's like whenever the police say, oh, we put these rewards out and the rewards have, we spent this amount of money on rewards to catch criminals, and this is how many criminals have been caught. Except that. That the tips they get are only a piece of the information. Right. And they don't take into account that they may, may just have found the guy later in the day. Right. Without the tip.

Gary [00:41:40]:
Right. It's the same thing here. It's like how many of the claims, 500-some dogs in the last year. It's like how many of those 500-some dogs would've been found anyway? Right. Almost within the same amount of time without the Right. camera.

Leo [00:41:52]:
Right.

Gary [00:41:53]:
So the network of cameras.

Leo [00:41:56]:
We Rate Dogs. I actually don't know the name of the guy behind it. I mean, everybody recognizes him, of course. Anyway, his point is very simple. If you want to get your dog back, if they're lost, make sure they're chipped. It's that simple. You don't need a Ring camera to do it.

Gary [00:42:16]:
Yeah, you don't need. And you know, the thing about Ring is, of course, well, okay. I'm sure it was on this show. We've been doing— how long have we been doing this now?

Leo [00:42:26]:
It's been a long time.

Gary [00:42:27]:
Long, long time now. It's incredible. But episode 260. So, I mean, and we've taken plenty of weeks off here and there. It's got to be 7, 8 years, something like that.

Leo [00:42:39]:
Uh, let's see, the first episode is, uh, from 2017. Okay, so we're coming up on 9 years, right?

Gary [00:42:48]:
9 years. So at some point early on I talked about it and I'm thinking probably at least one of the other earlier hosts or friends that had talked about it, perhaps you as well, about how Ring did have a system where you could either opt in or you had to opt out of over various times that you could basically— that your local police have access to your Ring camera. Um, and the local police could then use it to solve crimes. They would have a network that they could tap into and say, well, you know, there's 50,000 Ring cameras, but we have access to 10,000 of them, and we can now use the information from them to find a criminal or whatever. So that's— and we talked about that. I talked about how that's just— I'm never getting a Ring camera. And, you know, the whole like, you're becoming the surveillance network. You know, it's like if— it's like 1984, except that Big Brother is asking you to actually install the camera and give them access instead of, you know, them doing it.

Gary [00:43:49]:
So, but in addition to that, Flock has come up, and of course Flock is in the news all the time here, 'cause I live in Denver. Denver's the corporate home to Flock, and so it's talked about a lot, or not Flock, we're the corporate home to Palantir, which is also connected. It's all this big network of surveillance and stuff. But the reason Flock is in the news here is because our, they basically, wanted to install cameras everywhere. They want the city to buy their cameras and install them. And the city council said, absolutely not, we're not doing any of that. And the mayor stepped in and said, we're gonna do it anyway, 'cause there's like a budget number. Like if it's under this number, the mayor can just approve it himself.

Gary [00:44:33]:
And if it's above this number city, so he just went and said, what can you give us that's like $1 below this number? And he went for, so we have flock cameras here,, and people are upset about it. And the city council was like unanimously like, no, none of us want that. So, so we've got a lot of stuff going on with these like cameras and Ring and Flock. There is a way they intersect. Um, everybody claims they're innocent. Like Ring says basically, for instance, just taking ICE for instance, Ring says we have no, uh, we don't have anything to do with ICE., and Flock says, we don't have anything to do with ICE, but both of those companies do things with local police, and local police, in cases, have been caught giving information or access to ICE. So it's like bringing Flock saying, oh, we don't have anything to do with that, it's the people that are using our stuff that are doing it. We're just enabling it.

Gary [00:45:34]:
So there's a lot of that stuff, and it's interesting, I think there's a silver lining here, which is why I wanted to bring it up. The silver lining is the reaction was so negative against the ad.

Leo [00:45:44]:
Um.

Gary [00:45:46]:
That'S a good thing because what they were hoping is that people would say, oh wow, dogs, we can help rescue dogs. This is good. It's the same thing that Flockhammers did in the past with stolen cars and stolen catalytic converters.

Leo [00:46:03]:
Right?

Gary [00:46:03]:
They said, "We're gonna catch these guys." That's actually what they keep talking about here in Denver, is they keep saying, "This will help keep down car And, thefts." um, 'cause people are like, "We hate, you know, too many cars get stolen." Everybody hates that. And they're like, "We have a solution. Flock has a solution. We'll do these cameras. It'll help stop car thefts." And then people are like, "Oh, we're not falling for that. What you really want to do is track us." You're just using the, you know, like, you guys want car theft stopped, right? Well, we'll do that as the way to get in. And so it's the same thing here. We'll help find the lost— the poor lost puppy dogs.

Gary [00:46:40]:
Look at their faces. Listen to the Sarah McLachlan music. You know, we'll help. We'll help that. We don't— why don't you like us?

Leo [00:46:48]:
So a couple of things. It's episode 149. We've talked about Ring. Um, and I have to, you know, come clean, be honest. Um, we have Ring doorbells. Ah, um, yeah. However, um, I don't know if it's rationalization or justification. Um, the house is, you know, we're sitting on acreage, so it's not like we're actually actively able to monitor our neighbors or even the street for that matter.

Leo [00:47:21]:
Basically all we can see is what's coming up and down on our driveway, which honestly I'm absolutely okay with. Um, but, uh, it's also the case where, uh, a lot of this extended surveillance is something that as a Ring user you have to opt in to. Um, and then of course trust that they're, they're, that they're respecting your opt-in or opt-out Um, and of course then there's also the possibility of things like court orders being used to force disclosure of things that have been recorded. Um, but it's, it's interesting. The, my take on that commercial when I saw it was twofold. On one hand, it's actually pretty easy to think that they are just trying to sell more Ring doorbells. I also wonder if What they're really trying to do is get more existing Ring customers like me to feel good about opting into the extended surveillance.

Gary [00:48:32]:
Yeah, well, no, that's— I think that's probably the main thing.

Leo [00:48:36]:
What I— what probably won't happen— it would have been interesting to see if it would have happened— is if if that commercial aired and it did not get this wave of negative feedback, was Amazon planning on reaching out to Ring owners and saying, hey, you know, we've got this feature, it'll help you find your dog if you just turn it on. Um, you know, the fact that it does a bunch of other things, well, fine, but it'll find your dog. Um, but now of course with that, with this wave of negative feedback, they'd be be even more ill-advised to basically try and follow up and push any further on Ring for some time. I suspect that what they really need, um, besides changing some policies— but my guess from a publicity point of view is that they're probably going to go quiet on Ring for a while just because, yeah, that's the only way they really have to deal with this situation unless they actually want to start kneecapping their, their product.

Gary [00:49:40]:
Yep. No, um, that's, that's, uh, it's interesting. The whole thing's really interesting. I'm glad that we do seem to have a good pushback against surveillance, whether it's the city council here in Denver and the general population, whether it's people seeing this commercial and all of that. I mean, it is what's keeping us from slipping further and further into a surveillance state, is the fact that we've got, uh, a lot of people just feeling icky about it.

Leo [00:50:03]:
Yep. And—.

Gary [00:50:06]:
But the battle is between people feeling icky and the companies saying, oh, basically it's for finding stolen cars. Basically it's for finding lost dogs. You know, don't mind the fine print. Don't mind that 5 years from now we might get a lucrative contract from some government agency to do something else with it. It's technically okay because the fine print said it was. Right now we're starting from a good you place, know. Anyway, enough about that.

Leo [00:50:37]:
Yep.

Gary [00:50:37]:
Yeah, could, yeah, Um, we uh.

Leo [00:50:41]:
Why.

Gary [00:50:41]:
Don'T we move on to what was cool this week?

Leo [00:50:46]:
Yep. Well, so I basically did my this is cool.

Gary [00:50:50]:
Yeah, you did.

Leo [00:50:51]:
Yeah, stuff that I've talked about. I'll just re-highlight archive.org. Everybody should make use of of the, the resource that it is, and if you can give them Give them a buck or two. They're good people. Yeah.

Gary [00:51:01]:
Cool. So I've got— yeah, I've got an interesting one, a book. So usually when I try to pick a new book, like I'll read, you know, I'll read nonfiction, science fiction, a lot of memoirs and humor books and all sorts of stuff. And when I look for a new book, usually I have a bunch on my wishlist and then I'll usually check out what other people are reading. And then I also like to look and see what's popular. Just because every once in a while I like to, like, read what's really popular now to kind of, like, get the feeling of what people like. And so after a while, you know, sometimes I'll see what's popular and I'll say, no, that's not for me. But in this case, I kept seeing the same book at the top of the bestseller list, at least for Audible, over and over and over again for months, constantly.

Gary [00:51:51]:
And eventually you get— there's so many good reviews that I was like, this isn't really for me. I don't think so, but it's so popular. I just want to read it just so I know what people are talking about.

Leo [00:52:03]:
They wore you down.

Gary [00:52:04]:
So I went and my last Audible listen was Dungeon Crawler Carl. Have you heard of— I have.

Leo [00:52:13]:
This is the first I've heard of it.

Gary [00:52:14]:
Really?

Leo [00:52:15]:
Yes.

Gary [00:52:16]:
I've never seen a number of reviews number so high as this book. It's incredible. So it's a really— and you know, the thing is, of course, I thoroughly enjoyed it. Of course, I got towards the end and I started thinking, is this the first in a series? And usually when I start to think that, I'm like, please let it don't be the first in a series. But this case, I was like, I kind of hope it is the first in the series because I am getting towards the end and I want to read more of it. It's just a— it's a really cool book where basically I'm not quite sure how the whole world's going to play out, but it starts with the world ending, which is how my favorite book starts out. Hitchhiker's Guide to the Galaxy starts with the world ending. And like that too, this book has a lot of humor in it.

Gary [00:53:03]:
So the world ends and this guy is thrown into basically an action role-playing game, like, but for real. He's really in the game. With other humans. It's like a reality TV show meets playing Dungeons Dragons. It's really fascinating because you're basically listening to a Dungeons Dragons scenario, but it's being told by the main character. There's even like you hear all the stuff you hear in RPG games. Every time he gets an achievement, there's a voice saying, "New achievement!" Then it's like it describes the item. And, you know, that he gets and like how many points it has and all this stuff.

Gary [00:53:50]:
It's really fascinating, very easy to listen to, very entertaining. And so, yeah, I'm just like, I kind of thought I would enjoy it because if it's so— how could it be so popular if it's not good? And but I really am kind of enjoying it. I'm probably— I just finished the first book and I'm probably going to really quickly go into the second book. There's a whole bunch of them.

Leo [00:54:14]:
There's at least 5 from what I can see.

Gary [00:54:16]:
Yeah, it's going to be interesting to see where it goes, but it certainly is just a fun read and it takes you out of the, you know, your regular everyday.

Leo [00:54:28]:
Yes. Which, which is very valuable these days for sure.

Gary [00:54:31]:
Yes, definitely.

Leo [00:54:32]:
Cool. So in terms of self-promotion, the article I'm going to point people at this week is Managing Windows File Explorer's Navigation Pane. Asking. It's askleo.com/188995. It's weird. We've had this discussion before too. There are so many things that you and I take for granted. We just, we just use what we use because it's like second nature.

Leo [00:54:59]:
You don't even think about it. And I got a question a couple of weeks ago about, hey, you know what? It would be nice if somebody explained how to do this with Windows File Explorer. I said, you know, that would make sense. I should do that. And I did, obviously. And it has been surprisingly popular, especially the video of me showing how to tame what is essentially a fairly confusing left-hand pane in Windows File Explorer. I'm also getting the usual pushback that says, as is typical for YouTube, you should use Total Commander, you should use Explorer 2, you should use all these alternatives to Windows File Explorer, which is fine. I mean, I use one myself, but just being able to have some people feel better about being able to control what's in that pane in Windows File Explorer turned out to be surprisingly positive.

Leo [00:56:02]:
So yeah, managing Windows File Explorer's navigation pane.

Gary [00:56:06]:
Cool. I, of course, the big news from the last 2 weeks for Apple was the release of Apple Creator Studio, a subscription service. I've done a bunch of videos on it, but the one I will point out is the one called How to Keep Using Pages, Numbers, and Keynote if You Don't Want Apple Creator Studio, because The thing that people keep getting wrong or keep freaking out about is that Pages, Numbers, and Keynote, which were completely free before, they're still completely free. But now there's these extra features which aren't really that useful that you get if you have Apple Creator Studio. You wouldn't get Apple Creator Studio for these. You would get Apple Creator Studio for like Final Cut Pro or Logic Pro or one of you know, the, the other big apps that already cost a bunch of money.. And then you get these additional things that kind of come along with it. But if you don't want Apple Creator Studio, because you don't use any of those big apps, you can keep using Pages, Numbers, and Keynote just fine.

Gary [00:57:10]:
And that you're not really missing out on anything. So I have a video on basically, look, it still works just fine. It's still backwards compatible. It still just does everything it did. You haven't— nothing has been taken away from you. It's all still there. And you could just ignore the Apple Creator Studio stuff. And I even show some ways to kind of hide some of the stuff away so you won't see it as much.

Gary [00:57:34]:
So that's my biggest video of the week and probably the one that I would tell people to watch the most.

Leo [00:57:41]:
It's funny how people will— they're so eager to jump to conclusions. And oh yeah, no, in situations like this, you know, there's something evil. If they can even think there's something nefarious going on, that's the conclusion they're going to jump to.

Gary [00:57:57]:
Well, the biggest, like, one of the biggest responses I get from this video and from other things I've done is, well, okay, everything you could, could have done before in Pages, Numbers, and Keynote, you can still do now for free, but for how much longer? Oh, when's Apple going to put that stuff behind the payroll too? And I'm like, well, that's not new. They could have done that last year, 5 years ago, anytime. Any, any software product, any product of any kind, anybody could do anything with you. That was always a possibility. It's not like there was some guarantee that that was never going to happen until now. And now all of a sudden it's in question. It's like it was always an if, right? So it's nothing new. But if you look at it, it's like it was free before they made a big change.

Gary [00:58:43]:
And after the big change, it's still free. Why would they suddenly do that? You know, go and say, oh, we're going to— we took 3 steps this way. Now we're going to go 50 steps that way. You know, they're not going to. It's like, it doesn't seem to be the most like the Occam's razor kind of thing for what happens next is not suddenly after going to so much effort to keep everything that was free still free that they would then totally do a 180 on that. Yeah. That's— does it kind of make sense?

Leo [00:59:16]:
So for some reason it reminds me of the, when Windows 10 was first announced. They were starting to— the phrase software as a service was being used. And what they were describing at the time was the fact that, well, Windows 10 and now Windows 11, they'll get updates, more or less continuous updates. It's software that you will get as a service. Everybody— well, I shouldn't say everybody. This huge crowd of people took that to mean they're going to start charging. Yep. And here we are, what, 10 years later, still waiting.

Gary [00:59:51]:
Still waiting. Well, and the funny thing is, is that people, uh, so many people have been saying, oh, they're, they've started charging for Pages, Numbers, and Keynote. Again, not true. It's free.

Leo [00:59:59]:
Right. Right.

Gary [01:00:00]:
But the funny thing was, is that originally Apple charged for Pages, Numbers, and Keynote. You had to buy them.

Leo [01:00:05]:
That's right.

Gary [01:00:06]:
They were the iWorks, oh, they were the iWorks suite, and then they were individual buys, and then Apple made them free. So you're afraid they're going to start charging for them now. They actually did charge for them in the past. Then after they got to be big enough, had enough features, they actually made them free. They're not going to go back to that. And if they did, it wouldn't be they started charging for them. It's that they're charging for them again because they were. And of course, Apple used to charge for the operating system updates too, right? Used to have to pay for those up to some point.

Gary [01:00:36]:
So a lot of people saying, oh, Apple wants It's going to put everything behind a paywall. It's like, well, no, but they actually used to have all the software updates behind a paywall and then they made them free. And guess what? All this stuff that people are talking about, still free. But they still haven't taken anything away. They haven't done any of that. So a lot of people speculating basically pessimistically for the worst. Like, what's— can we take this news and make it go in the worst direction as far as we can go?

Leo [01:01:05]:
It's part of, part of my job, part of your job, I'm sure. Part of my job that I dislike is just, you know, the fact that I'm having to constantly tell people, no, that's not what they're doing. That's not what they're thinking. That's not— doesn't make sense. But of course, people want to believe the worst.

Gary [01:01:21]:
So I think I'm actually— I mean, I've done— I did 3 videos on Apple Creator Studio, one on like, is it worth it? Like a price comparison. Is the $13 a month worth it? One on backwards compatibility because people had questions about like, if I create a Numbers document in the new Numbers, will my friend with old Numbers be able to use it and stuff? Or will I be able to get the— if I let the subscription lapse, will I still be able to open my document, right? So, I did a video that basically is like all that. You can do— it's fine. You know, you could let your subscription lapse. You can give the file to your friend that has last year's version. It's all fine. And then I did you this, know, what if you don't want the subscription? How can you keep using Pages, Numbers, and Keynote? And I got so many comments just like you're talking about. And I think I've decided, well, from this point on, I'm just doing comment— doing videos on individual features.

Gary [01:02:12]:
Like, here's how to use this feature of Apple Creator Studio to do something. I'm not going to take any more comments on like, is it worth it? Apple's gone, you know, gone crazy or evil or whatever. Like, This is if the video is about Numbers Magic Fill, one of the new features, then it's basically if you want to talk about something else, nope, comment deleted. This is a video about Numbers Magic Fill. It's not your platform for talking about how you don't like software subscriptions.

Leo [01:02:43]:
Right.

Gary [01:02:44]:
So, you know, yeah.

Leo [01:02:46]:
Interesting times.

Gary [01:02:47]:
Yeah.

Leo [01:02:48]:
All righty. Well, I think that pretty much wraps us up for yet another week. Week. As always, thank you very much for hanging around, listening, doing whatever. Uh, we will see you here again real soon. Take care, everyone. Bye-bye.

Gary [01:03:02]:
Bye.