Leo [00:00:26]: And we're back. So last— well, I want to say last week, but that's wrong since we skipped a week. Two weeks ago, I was all excited about some of the things that I had been doing with Claude. And I intend to continue that because I've done a couple of things since then that I also found really interesting and really cool in some really geeky ways. So, one of the things that I know you and I both face is when our server starts misbehaving, our web server starts misbehaving, and we're trying to figure out why. Very often we'll see that, you know, all of a sudden our server that has a load of you know, 1 on a, on whatever scale you're measuring, is suddenly maxed out for reasons that we don't understand. And we know that it's not because we suddenly have that many people actually interested in our site. It's the old, remember the old slash dot effect, right? If you got mentioned on slash dot, then within minutes your server would go down. Leo [00:01:39]: That doesn't happen. It's never happened to me, but it's also one of those things that I don't think happens as frequently anymore for various reasons. Now, the approach that I would have taken in the past was to, you know, take a look at some server reports, take a look at the status from the web server software itself. I tried something different this time. I downloaded the raw server log. For those who care about such things, I am, you know, I run the Apache web server, which is basically, I think, what like probably 90% of websites run these days, if not that, then NGINX. And I just downloaded the raw log. It was like, I don't know, 150, 200 megabytes in size. Leo [00:02:30]: Since I have Claude code running on my machine, things like upload sizes or those kinds of limitations were not really in play. And I just said, hey, Claude, analyze this log. No preprocessing, no filtering, no other qualifications. Just here's a log. What can you tell me about it? And it was fascinating. The number of things I didn't have to do. I didn't have to filter out anything. I didn't even have to tell it that it was an Apache status log. Leo [00:03:07]: It just understood what was in there. Now, this is cloud code running on my machine, so it took a little while, like, like maybe 10 or 15 minutes, and it came up with a prioritized list of who's hitting the server, how hard they're hitting the server, followed by a prioritized list of things you might consider doing. Now, what cracked me up is that I was my own worst enemy. The number one cause of the load on my server was my own processes. If when you're running WordPress, there's a thing called WP-Cron, and it actually kicks off every time somebody visits the website. So it has kind of this magnifying effect on any kind of load that happens. The alternative to that is to disable this automatic WP-Cron and then set up your own timer to fire it off at a more reasonable rate, like once every 5 minutes. Once every 5 minutes is plenty, not every single call. Leo [00:04:22]: So it told me, you know, 10% of the hits on your website are this WP-Cron call. And the call is coming from inside the house. Right. So, so I, and it's what, again, probably like you, I manage enough web servers that, you know, yeah, I should do this. But apparently for that particular site on that particular server, I had overlooked it and we had reached this point of threshold where all of a sudden it was becoming significant. I went in, I did the thing, I, you know, set up cron on its own timer. Every 5 minutes, like I said. And sure enough, web server became significantly more manageable. Leo [00:05:07]: It also gave me a list of suspect IPs and IP ranges for things that were hitting the server particularly hard. One of the things that I think all servers these days are hitting, are getting hit by pretty hard, are spiders and AI scrapers. And sure enough, here's an IP address that has hit your website like 500 times in the last hour. And it's been asking for things that don't exist, which is a classic probe. It's looking for vulnerabilities on your website. You might want to consider blocking that IP address and this IP address and this IP range over here. It was awesome. It was very, very impressive. Leo [00:05:55]: So I was doing that on essentially a effectively a client's site. And I said, you know, I should do this on my own site on askleo.com. And sure enough, turned up a list of suspect IP addresses that really had no business behaving the way they were behaving. I added them to the firewall as blocks, and all of a sudden my server load drops by a significant amount. So it's something that it's really simple to do. With Claude Code and just tell me about what's in this log. And it's probably something that I'll do periodically because, of course, as we know, the attack changes all the time and there's always, you know, some new player coming on the, um, on the scene. But I just thought it was a fascinating and easy way to identify some of the things that in the past have been difficult to identify. Gary [00:06:49]: Yeah, um, I've used it Many times for exactly that. Leo [00:06:52]: Really? Gary [00:06:53]: But not recently. Well, not exactly. I mean, going through the logs, figuring out what it is, not messing around, because boy, yeah, I used to mess around with trying to figure out what was in there. Leo [00:07:05]: Right. Gary [00:07:06]: But just giving it a chunk and just saying, what's going on here? Sometimes with the information that I've got, like things are running slow or I'm getting this error or something's not working. But sometimes just with like, here's the log. So yeah, it's really useful. Strangely, maybe it's because I got the last time I did that, I got things working so smoothly. I haven't had to do it for a long time. I mean, I think it's a combination of things. I mean, maybe I did, you know, crush a few bugs with what I was doing back then. Maybe it's because I've simplified things and maybe also because stuff's getting better. Gary [00:07:46]: Stuff's definitely getting better. I think I talked before about the last time I moved servers, which is something you have to do. I didn't move server companies. You get to the point where it says your current server can't support newer things. You migrate over. Yeah. So the last time I was like, oh, I have to do that. It seemed to be a longer gap. Gary [00:08:06]: I forgot everything I needed to know to be able to do that. So I kind of start from scratch, set aside some days, and then an hour later I was done. I was like, what? I'm sure that should have taken longer, but everything seems to be working okay because things have gotten better. The training has gotten better. The docs have gotten better. The interfaces have gotten better. Yep. Leo [00:08:27]: The— so I'm in exactly that spot right now with askleo.com. It's running on a currently unsupported version of PHP, which I can't upgrade in place because the Linux version that I'm running is too old for the newer version. Gary [00:08:44]: Yeah. Yeah, that's exactly the issue. Leo [00:08:47]: I've spun up a replacement server. I'm out at AWS, so it's just an easy thing to say, make another one. This time with, you know, current versions of Ubuntu and the default up-to-date versions of PHP. What I found myself running into was that just like you, it's been years since I migrated these servers. They've been just chugging along, which means, of course, I forgot everything I did 4 or 5 years ago. Gary [00:09:19]: Exactly. Leo [00:09:20]: So I would migrate the stuff over and it wouldn't work, as in it wouldn't come up. And I basically— and then when I finally did get it to come up, it was slow as molasses, even though the server itself was probably slightly more powerful than the one we're running on., so an iterative process with, again, Claude, I think is what I was using. And it said, okay, well, here are some common things that can affect site performance across a server move. And it said, oh yeah, you probably tweaked your database configuration on the old server. You might want to replicate that tweaking on the new server. Because the default configuration for these databases assumes you have a tiny server, right? So you don't necessarily end up making use of all the resources on the new machine. I said, yep, okay, I should do that. And oh, another thing that sometimes happens is that, you know, you could, since you're probably running WordPress, you may want to look into something that's a persistent object cache. Leo [00:10:39]: Which speeds up PHP and its access. And said, well, yeah, I suppose I should do that. I'm surprised, but oh, look at that. I actually did do one of those on the old server and it's not present on the new server. And son of a gun, I turn it on the new server and boom, response times are back to where they should be. Rather than 10-second page loads, I'm back to, you know, like 1-second page loads. Even though, Even so, with all, I think all of the I's dotted and T's crossed, I could probably do this at any time. I'm setting aside a big block of time so that I can actually do the transfer and deal with any issues that pop up no matter what. Leo [00:11:26]: And hopefully it'll just be a quick upgrade and I'll be moving forward and happy again. But yes, that's the kind of stuff that I'm using AI for and Claude for quite a bit. It's actually done a very, very good job of reminding me of things to think about. Now, one of the other things I have, I don't know, we've talked about it, but I don't remember what you do. I use something that is equivalent to Zendesk. You know what Zendesk is? It's a support ticket server service. I have an open source version of that called Free Scout. It's installed on one of my servers. Leo [00:12:03]: And for the past couple of years, It's been collecting people's questions. It's a MySQL database. It has lots— all of the questions that come in get placed into that database, and that allows me to coordinate responses to those questions between myself and my assistants and so forth. It's actually a very, very nice system. From the user's perspective, it all looks like email, but from our perspective, it's just an interesting interface to play with things. Dawned on me that, gosh, there's a lot of interesting data in that database, like years of questions. Yeah, but analyzing it would be a real pain in the butt. Now, normally what I would have done, said, okay, great, it's a MySQL database. Leo [00:12:51]: I can fire up some MySQL queries to extract the information that's interesting. In other words, the questions and their responses., but that's too much work, right? I don't want to do that. I could do that, but I don't want to do that. I wonder if— so what I did is I simply did what's called a MySQL dump, which basically dumps the entire contents of the database to a text file. I downloaded that, pointed Claude at it, Claude Code specifically. And sure enough, it did this wonderful analysis of, oh yeah, here are the most common topics that people are asking about. Here are the questions that most often go unanswered. Here are where people are worried about things. Leo [00:13:42]: I didn't tell it what kind of database this was. I didn't tell it anything. I just said, here's a database. Analyze the contents. That one specifically, because Apache logs, sure, it would make sense that that would be part of its training, if nothing else, right? Depending on what kind of training stuff. But the fact that it just was able to read this really random database from, you know, this niche question and answer tool, that just blew me away. I just thought that was Amazing. And once again, it's one of those things where I need to remind myself that the data format no longer needs to be an obstacle. Leo [00:14:28]: If I can get it out in text, chances are I could run it against Claude Code or one of the other AIs and actually get some really, really interesting insights from what it finds. Gary [00:14:44]: Cool. Yeah, I've been playing around, starting to play around with Claude Code. It's interesting. Probably, I mean, I've been doing a lot of vibe coding, something maybe we'll talk about later, maybe next week. And Claude Code seemed like that there might be a few pain points in dealing with regular large language model vibe coding that Claude Code might solve. Being able to just update the code on the fly and stuff. But I'm kind of cautiously starting to dip my toe in with Claude. And actually just this last week, I started subscribing to Claude in addition to ChatGPT. Gary [00:15:22]: I think I had, well, no, I know I had before some time ago. I had a subscription mostly back in the day when Claude was known as the one that wrote better. And now it's starting to get a reputation as maybe the one that codes better too. Leo [00:15:38]: It writes code better, yep. Yeah. Gary [00:15:41]: So, you know, I'm trying to just, you know, diversify between ChatGPT and Claude. Leo [00:15:45]: I think I mentioned a couple of weeks ago, I found it fascinating that if you watch Claude do its thing while it's processing whatever I handed it, it's generating code, it's running code, it's generating some more code. I mean, it's just fascinating to watch. Yep. So, and yeah, like you, I did end up subscribing to Claude. Mostly so that I could get access to Claude Code in its current form. I'm actually letting ChatGPT go. It doesn't have the added value for me anymore. The big thing for a while was its ability to generate images, which, you know, through DALL-E. Leo [00:16:26]: But I'm finding that Gemini and Nano Banana actually do everything that I need. Once you kind of learn the nuances of how to prompt it for what you want. And as it turns out, I've got a subscription to Gemini because I've got a subscription to another Google service that includes, hey, you know, subscribe to this and we'll throw in Gemini for free or something like that. So I've got one of those bundles. So that's actually been pretty, pretty fascinating. All right. One more stupid Claude trick. Yeah. Leo [00:17:02]: I was writing an article in my personal blog about writing. It's basically tracing how I started as a writer over 50 years ago, as it turns out, which surprised even me, and, you know, some of the steps along the way. And it dawned on me as I was talking about something that I had done at a company that I worked at for a couple of years before Microsoft. I actually published a newsletter. I thought to myself, you know, I wonder, I suspect that those newsletters, hopefully I have copies of those newsletters in my basement in a box somewhere as one does. And sure enough, I was able to find them. I ended up scanning them all and along with some other information about this really small niche company that was doing Z80 processor data entry terminals. Gary [00:17:53]: Yeah. Leo [00:17:53]: And, um, And I just, I even went whole hog. I went in and uploaded all that stuff to Internet Archive. So while I was in that box, I ran across a journal. In fact, I'm showing it to Gary. You won't be able to see it if you're listening. It's actually a handwritten journal from, it starts in 1979. Um, I'm still— I refer to my now wife as fiancée to give it some, some context. Yeah, I was looking at it and it actually had some really interesting things to say. Leo [00:18:32]: I was surprised at, um, how lucid I was for a 22-year-old. I wonder, I thought, what if I scan this page? Okay. Scan the page, and I threw it at— I think it was Claude. And Claude, I asked it, okay, please OCR and return me the text on this page. And my handwriting has never been great. Um, it certainly hasn't gotten any better, but it certainly wasn't any good 50 years ago. And Claude just booped out. Yeah. Leo [00:19:10]: Here's the text on that page. Well, then I thought, got to thinking, as with most of my journaling efforts, it only lasted about, I don't know, maybe 30 pages over the course of 3 years. So it wasn't— I don't journal often, but I wonder. So I scanned each page in turn. So I've got the entire journal scanned. It's all my handwriting. And I just gave all of the images to Claude. And I said, okay, reconstruct this, make this, you know, give me OCR this and put it together. Leo [00:19:54]: And it just did. It just gave me the text of my 50-year-old journal, um, in, in text form. I've got it, um, a piece of it. I ended up One of the reasons I did this is I wanted to quote something that I said a long time ago with respect to writing, and I've included that in my, on my personal blog. So I'll add a link to the show notes here in a bit. But again, I just find it fascinating that take a picture of a handwritten page, throw it at AI, and boom, you've got editable text, which again, just blows me away. Gary [00:20:33]: Wow, I, uh, I'm gonna have to try that. I was just looking while you're talking, and I've got a handwritten journal when I drove cross-country through Australia in 1996. Okay, and it's handwritten journal, and I scanned it in at some point. Oh cool, okay, 2011 I scanned it in, and it is, uh, yeah, my handwriting is horrible The whole thing is here scanned in many, many, many pages. Uh-huh. Hmm. Leo [00:21:02]: All right. It's an interesting project. I mean, yeah, you know, there's, there's other things that I could do this to, but there's just really no point. But this whole, this whole journal where I start journaling when I'm 22 is just kind of fascinating to me. And I had to, I decided I just wanted to read through it all and see if this exercise would work. And it did. Gary [00:21:25]: Cool. Leo [00:21:25]: So next topic. Yeah. Speaking of age, prove that you are not a minor. Prove that you are of age. That turns out to be a very, very difficult problem. Many, many sites simply rely on your honesty, right? Say, are you 18 or over? Sure, I am. Okay. Yep. Leo [00:21:50]: Of course. You're in. But of course, that hasn't been good enough. And I think it's Great Britain that actually has been spearheading and in at least one case enforcing the requirement that some sites verify the age of their users so as not to accidentally expose those younger users to inappropriate content. Um, Discord is one, Reddit is another, and indeed, I mean, there's plenty of age-inappropriate content on both of them. Uh, but, um, the whole concept of how you go about proving that someone is a certain age is fraught with peril. Um, You've got some great points here. I'll let you run down those, but I just find this— I'm not sure that this is really a solvable problem. Gary [00:22:51]: Yeah, yeah, I know what you mean. It's not— so it's not only is it not a solvable problem, but it is— it's so intricate that even I was trying to come up with, you know, a simple solution and— or it's simple, like, way of looking at it, and I kept running into things I hadn't thought about before. Before, like an article that came up that was like a group of articles, not that I want to actually put a link to anything, but about the negative aspects of the whole thing. So yeah, so you have to do age verification. So how do you do age verification? Well, you have to get some sort of positive identification of who you are. And you could do a lot of things for that. You could upload a driver's license, For instance, you can enter in identifying information about yourself that links then to government databases that can then verify your age. A lot of stuff like that. Gary [00:23:52]: All of those have the big problem, hey, what about privacy? Leo [00:23:56]: I was gonna say, you're uploading sensitive information. Gary [00:23:58]: Yeah, if you're like a parent and you're worried about your kid, it's easy to look at it from that direction., but if you're like 35 or 55 or whatever it is and you just wanna read a forum on tech coding tips or something, and it's asking you to prove your age, and to do that you have to actually identify yourself, and you're like, no, I wasn't up for that. I don't want my full identity to be stored with this website that I just discovered. So that's like a big problem. It's like it's not just identifying the 16-year-old, it's identifying the 55-year-old too, because you don't know what anybody's age was before they identified themselves. So on the one hand, you've got privacy advocates and lots of people that are, you know, want to keep their privacy online, not necessarily because there's any real reason for it. It's just that on the other end, there's a website that they don't know if they can trust. Right. Yeah. Gary [00:25:02]: You run into codingtips.com or whatever, and you're like, oh, I have to give them my identity in order to go further. I don't know this site well enough to give them that information. I've been taught for years now that I shouldn't be sacrificing that information to just anybody. Then it's, yeah, there's the problem and it's got nothing to do with kids because you're not a kid and the site doesn't have kids' content, but now the laws all say that, you know, all this has to happen. So you've got that— the issue of privacy on the one hand butting up with, you know, protecting kids online. On the other hand, you can't just verify the age of the kids. You have to verify everybody's age in order to know Who are kids, right? And there are other aspects of it. Like, one interesting one I came across was political speech. Gary [00:26:05]: In Australia, right, it's been in the news a lot. They passed a new social media law saying no social media for kids. I don't know if it's 16 and under or under 16. Yeah, yeah, at 16. I don't know if it's include 16 or not. But one of the things that had never even occurred to me to think about was people hitting back there with, well, they've just eliminated kids from the political, you know, political speech there. Leo [00:26:30]: Right. Gary [00:26:31]: Because there's no political speech going on outside of social media online. I mean, political speech online is all on social media. So now if you're 16 or 15, you can't participate in that anymore. You can't organize anymore. And there are certainly plenty of kids in high school, of high school age, that are participating in protests, that wanna make their voices heard, and wanna stay informed as they're getting close to the age where they can vote. And suddenly they're cut off from that. So their freedom of speech has been, you know, pinched in the name of protecting them from, things they shouldn't see or being influenced by people, you've now cut them off completely from politics, which is not necessarily something you want to have happen. So an unforeseen circumstance there, and probably something that will mean it would never— something like a 16-age cutoff for social media will probably never happen in the United States. Gary [00:27:32]: It may not last in Australia. Leo [00:27:36]: So it will be interesting to see what happens down under, just to see if it does last. Gary [00:27:40]: Right. I mean, you know, they could always, you know, with a much lower, smaller population and stuff, they could always probably maybe find ways to still continue it. But I don't see it happening in the United States and probably and maybe not in Europe. I don't know. Leo [00:27:55]: But I mean, it is certainly a problem. It's kind of sort of happening in Europe, right? I mean, not the EU per se, but it's Great Britain that's having this trouble. I think they're actually suing Reddit. For, well, yeah, not having done the right thing. Gary [00:28:10]: Well, yeah, they're the, the, the, the stuff going on there is that those companies, Reddit and Discord, right, they— it's not that they weren't— they were doing it, but they were doing it using a company that people considered was violating their privacy. Ah, okay, right. So, and of course, on Reddit and Discord, you're gonna have a lot of people that are very concerned about their privacy., but then they were forced to say, well, we need to go and figure out like what to do with, you know, we can't have kids on anymore, so we need to figure it out. So the problem is there is they used a company that had a good idea. And there's other companies I'm sure is like, well, every single site can't do age verification. Like that's ridiculous. You can't, in order to be online, have your age verified by 100, 200, 300 different sites and probably 100, 200 different ways. Leo [00:29:00]: Yeah. Gary [00:29:00]: So we're gonna have, I mean, yeah, nobody's gonna do that. So we're gonna have like one way you can verify and then companies can come to us for a solution. So then if you verify with, through that company while you're trying to log onto Reddit, now you find you go to Discord and Discord's like, oh, you're good. 'Cause we use the same company. It's the right idea. But the problem is in this case, that company wasn't doing enough to protect privacy. According to the people that use Reddit and Discord, which now, you know, it's caused a problem for them. Leo [00:29:34]: Right. Gary [00:29:35]: And one of the kind of solutions to this that's been talked about is, all right, let's have the big guys come in and do this. Like, forget about having these little companies. Let's have mainly the two biggest, but we could expand that to three easily, Google and Apple, right? The two companies that own the mobile phone operating systems. Have them come in and basically you— because you probably, if you're on an iPhone, you probably have an Apple ID and you probably have an Apple account and all of that. And it's very common to log in to apps, just say, yeah, yeah, I don't want to create a new account, just use my Apple ID for my account. And it's the same thing for people on Android. It's very common to go and say, I'm not going to create a new account, just log in with Google, hit the login with Google button. Button. Gary [00:30:23]: And the third one is there's also login with Facebook. It's very common. Matter of fact, the three of those appear together, you know, those three buttons you see everywhere. So you could include Facebook too. You could say let these big companies handle it and then let all the little companies be able to use that, which is what they're doing now for logins, but have that overflow into age verification as well. So you not only as a small site like yours and mine, we could potentially just, you know, use those for logins. Now, problem with that is then we have to give up our own personal login, like our own login systems, because right now typically it's like, hey, log in with Apple, log in with Google, log in with Facebook, or create an account, and you can create an account with your own email address. These sites will have to say, yeah, that's not an option anymore. Leo [00:31:10]: Well, do they though? I mean, I was thinking about that as you were talking about it, and I could easily see the age verification requirement, insisting that you have a Google account or an Apple account or whatever. But that could be like just for age verification. It might be associated with your login account. Not that that isn't a scary thing in and of itself, right? But for example, on our sites, I could see somebody logging into my site with their Ask Leo account, with their ask, you know, the whatever email address they use. And then at some point we have to say, well, we need to verify your age, please. Login with or use one of these third parties, Google, Apple, Facebook, for me to get your age token, right? And maybe it's a one-time thing and I just store that in your record and we're done. But it doesn't necessarily mean switching over the entire login infrastructure. The part that concerns me the most is the loss of anonymity, right? Yeah, sure. Leo [00:32:15]: I mean, the implication as I understand it is that you have to be logged in in order to see the content from these age-gated sites. Again, regardless of how old you are. That, like I said, removes anonymity completely. Gary [00:32:31]: That's frustrating at best. Oh yeah. And also these companies, I haven't read it. I'm sure they're not going to have spokespeople come out and talk about it, But why would they wanna do this? Like, they're just gonna be punching bags, right? Because there are gonna be people on one side saying, oh, now you're violating our privacy, and people on the other side saying, you're not doing enough to protect children. And with no financial gain, right? They'll be like, oh, okay, so we're gonna have a team of like 50 people working on this from now on, and all they're gonna get is grief from both sides, and we'll have a few government officials and a few activists all being happy with us, and the rest of the world not liking it. It almost seems like it should be, and this kind of thing rarely happens, but some sort of consortium that should do it, like the Unicode Consortium, like some sort of nonprofit international organization that handles this, and the idea is a bunch of big companies, Apple, Google, Facebook, Microsoft, and all that, blindly give a bunch of money, a drop like change in the couch compared to what these companies make, But give it just like they do with Unicode Consortium and other things, and say, here you go, you've got this budget. It's an international organization, it's got some bylaws, it's outside of government control, but it's listening to what the governments are asking. The governments are saying, well, this is what we would like, and they say, well, we're gonna come up with a solution that hopefully will make you happy so that you don't have to bother the tech companies and the small business people and websites and stuff, and it's our little mission to do, to do that. Gary [00:34:12]: That would be an interesting way to go ahead and do it, and we certainly have plenty of examples of things like that existing. They would still get a lot of grief, but they could shrug off the grief because they're not like a company like Apple or Google where they've got reputations and all this. It's like the Unicode Consortium takes tons of grief from people who don't like the emojis, who don't think that they're the right emojis, who don't understand why there are emojis, you know, and all this, and I'm sure that the actual people that work there probably say, Eh, it doesn't matter. It doesn't affect us whether you like it or not. We still are trying to work hard to make a nice, you know, to unify the character sets worldwide. Leo [00:34:54]: And it's our little mission. Gary [00:34:56]: Why didn't you include my favorite emoji? Yeah. Leo [00:35:01]: So yeah, it's really, it's tough. Can you think of an approach to age verification that works. Because my biggest thing about most of the age verification approaches that I keep hearing is that they're almost as safe as just asking the person to answer, are you 18 or older? Right? I could upload somebody else's driver's license. Yep. You could log in with somebody else's account. Yep, exactly. Um, I just— Gary [00:35:34]: that's why I started off by saying I'm not sure this is a solvable problem. Yeah, no, you're right, 'cause if you're anonymous when you're online, whether companies like it or not, because you could simply have somebody else create the account for you. I mean, they're putting themselves on the line to do it, but it certainly is the case where parents create accounts, social media accounts themselves, and then hand them over to their kids, Right. That could still be done. I mean, I guess the idea is you try to make it harder, right? So you say, okay, well, a kid still needs to have one parent, you know, on their side or one adult on their side. But I mean, kids get, kids get vapes, you know, kids get cigarettes and kids buy alcohol and stuff because there are adults that'll hang out or adults that will take money for it, right? So if you make it hard enough for a kid to get a social media account, eventually it'll be worth something. And eventually you'll be able to find somebody, give them $50, and they will create an account and then hand it over to you. I mean, I guess you could restrict people to one account at a time, right? So you could have like You know, if an adult does that on Facebook, they can't get a second Facebook account, right? Leo [00:37:03]: 'Cause their ID's been used. But all they need to do is fire up another email address, right? Gary [00:37:08]: I mean— Well, no, but if, well, the idea is that let's say, like, let's say Facebook says, okay, Facebook is supposed to be one account per person. Yes. Right? You're not supposed to, for a long time, you're not supposed to be able to open up a second account. Is you are a person, you have one account. That's it. And they've set it up so to make it really difficult. They've done a good job at this because I've heard of people getting kicked off Facebook and they can't get back on. They can't create another account with another email address. Gary [00:37:36]: It figures out who they are during the creation process and they're stuck. Mostly people dealing with Instagram, right, are stuck. And so there are ways that they can limit that. So the idea is that if you create your Facebook account, and you verify that it's you doing whatever Facebook does for that. I don't even know. My Facebook account dates back to the beginning, right? But then if you sell that for $50, so somebody could have an account on TikTok, and in the verification process on TikTok, one of the choices is verify with Facebook. Verify with Facebook, I'm really this guy, even though I'm gonna use a cool alias here at TikTok., and then it says, great, you're in. But now that adult that really has the Facebook account cannot create another TikTok account, right? Like they're stuck. Gary [00:38:28]: So they're selling for $50 their access to TikTok. They're not selling like, oh, I can repeat this 100 times and have 100 kids using my ID. So you could get that far on it. Um, you know, one account per human and then really Then you could also, you've got government ID stuff. Obviously everybody in the United States has a unique Social Security number that you're not supposed to use for anything except for Social Security, except that we do use it for lots of things besides just— All over the place, yes. All over the place, so that ship has sailed, plus it's being used for things with really no pushback in lots of cases, and in the case of protecting children, it would have a lot of pushback. It's like, oh, if that's all it takes to protect children, who cares that our Social Security numbers are now being used once again as ID numbers? Leo [00:39:24]: And heck, most kids don't even have Social Security numbers, but— Gary [00:39:27]: I don't know if that's kinda true anymore, but then that would work, because it's like, if you're not supposed to be on something, and you don't have a Social Security number, then it matches. You can get a Social Security number as a, I mean, as an infant. Leo [00:39:41]: Oh, I understand that. Gary [00:39:42]: I'm just saying that most don't. Yeah, it's— yeah, I don't know what the current pro— like, I mean, obviously to get your first job you do, right? But there may be other processes happen earlier that you may need to actually get your, your Social Security number. Um, I don't know, a question for ChatGPT perhaps. But the, the, uh, but yeah, I can't think of a perfect way. I can just think of a way of making it such a pain to, to fake it that it becomes difficult. Leo [00:40:19]: You know, there's, there's that. Um, my other concern is that whatever process they come up with finally, uh, or processes if each company ends up doing their own thing it's going to act as a disincentive to use that service at all. Anybody that is concerned about their privacy is going to say, no, you can't have this. At which point they're going to say, well, that means that either you can't have an account or your account is forever locked as a teen account, right? As an underage account. As I understand it, that's the default. That's the, the approach that some of these services are taking, that if you, if you're not willing to verify your age as being over 16 or 18, then we have to treat you as if you're under. Gary [00:41:07]: Yeah. What, okay. What happens if, so what happens if there was an organization you could log into it, create an account and prove your, your age, right? Right. Through various things, right? Through what you could say, oh, just use the Colorado or Washington State database. Sure. There's my driver's license and all that. You could do it through a federal thing. You could do it. Gary [00:41:30]: There are a variety of different ways you could prove it. Some of them free. Some of them cost, right? Right. Yeah, so you can choose from a menu, and once you have that, you can now got an account with a password, and it's your age is there, you know, right? And then you go to another site, and then you log in, you create your account there, and it says prove your age, and you can— it asks you to log in, authenticate with that service, right? Leo [00:41:59]: You authenticate, but that service doesn't tell, doesn't share any other information rather than your age, right? Gary [00:42:09]: It doesn't even have to share that. It kind of does because it changes. So it's the idea being like, well, what if you're— what if, you know, you can't just be a green check mark or a red X? Sure you can, because— Leo [00:42:23]: yeah, but then depends on the question you're asking, right? If the service you're trying to sign into that requires verification All it needs to say is, is this person older than 18? Gary [00:42:34]: Yeah, but then you would have to— so then if you were younger than 18, you would have to reauthenticate at some point saying, oh, it's a— Leo [00:42:43]: but no, they wouldn't automatically know. The third-party service, its job, its one job is to store your birth date, right? But authenticate Yeah. So yes, every time it gets asked, there's a possibility that the answer changes because you're getting older. Gary [00:43:04]: Right. But— So the idea is you would have to then, so if you went in and it said no, right? You said, oh, it's because I turn 18 next month. Right. You would have to have the functionality on that site. And this is something sites are bad at. You would have to have the functionality to go back and say, oh, You still have me as a no, but my birthday is today. Leo [00:43:25]: Right. Gary [00:43:26]: Recheck that because it's going to tell you yes. So as long— that's one way to do it. The other way to do it is just pass a birth— like it could be a birthday, but it could be the kind of thing where you could specify, oh, I want my exact birthday because that's going to be important soon. Or you could just say, no, just within a year. Right. Right. So I was born in this year. That's all you need to know because that puts me way past any minimum, and we're good. Gary [00:43:55]: And then it records, okay, this person was born in this year. We know that. We're set. And then it doesn't get any other information, and the connection between your account at the one site and your account at the site that checks is broken at that point. Leo [00:44:10]: Like they have— Gary [00:44:11]: Once you've become old enough. Yeah, it's just sending you back. You're basically, you've gone off. You've given it and it shoots back this little, you know, it sends a token out, then you add to the token, oh, here's my account information, and then it sends back the token with the age or a checkmark, and then it doesn't know what information you gave to that other site. It just knows that it verified you, and it could, so at that point, the idea would be that you can't fake it. You can't, you could only use somebody else's, but that would only be work one time. Like somebody would sell you that, then they've lost it. So somebody couldn't get rich doing it. Gary [00:44:53]: They could be like, here are the top 10 services. I'm gonna do it. I'm not gonna use any of these top 10 services. I'll sell each for $50. I've made $500. I've made the sum total of all I could ever make doing this scam, and it's over. And one person was able to slip past. And that's it. Gary [00:45:12]: Not everybody slipped past or everybody used this technique. It's just like a few people slipped past and it's still illegal. You know, it's still like get caught, get banned from the service forever from both ends. Oh, right. Leo [00:45:26]: Of course. So there's still consequences. As we know, just because it's illegal doesn't mean people aren't going to do it. Sure. The concern I still have is that in your scenario, you've got a third party, uh, who is charged with validating, verifying, confirming your age. Everybody has to trust them. Yeah, completely. And I'm not convinced. Leo [00:45:57]: I mean, it's— that's essentially the same argument as saying, well, Apple and Microsoft and Google should do it. Granted, we all have reasons not to trust those companies, but it's no different than this random third party. Gary [00:46:09]: Yeah, well, they could be— if it's one, like, consortium that is funded by all these companies, right, they could still be open to audits by the companies that give the money and by governments, right? So, and they probably would, as part of their charter, get audited on a regular basis. And that actually would be like the main point of the whole company would be to have everything to be so squeaky clean because half of what we do is this age verification. The other half of the thing that we do is submit ourselves to audits, you know, and the audits just need to be like, oh, everything is so perfectly clear that this audit took no time at all. And, you know, we'll see you again 3 months or 1 year from now. Leo [00:46:59]: And the auditor— Gary [00:47:00]: see, so do we have to trust the auditors? Well, that's why you have multiple auditors from multiple places, right? So they all have exposure into the information that's being collected by this third party. Yeah, but yeah, I mean, that's part of it. But, you know, maybe not. I mean, you walk into a building, you don't— we're not walking in with like, I'm going to download all your data. You're walking in hands-free with like, I just need to see what's going on. You can't really extract it out of there, and maybe even multiple auditors. Yeah, there are ways to do that. You have, I mean, audits for like this, you have multiple auditors from different, you know, today the auditor from Apple, the auditor from Sweden, and the auditor from a citizens community group of Western Canada, right, is all— they're all going to be here and they're all going to be going through the audit together. Gary [00:47:55]: No one's going to try to pull anything, right, because it's these three. And next month we've got, you know, Britain and Australia and, uh, Google is going to be here, you know, and, and all of that. And it— yeah, I don't know. I think, I think it can be done. Leo [00:48:15]: But will that path be chosen? Probably not. Yeah, my guess is it'll be something else that, yeah, doesn't make as much technical sense. This is one of those things where there are a lot of strong arguments on both sides, and unfortunately, because it's a political issue, rarely does technology get properly designed as a result of politics. Gary [00:48:38]: Yeah, that's true. So that's true. That's— oh yeah. Leo [00:48:44]: Yeah, we'll see. Watch this space. Gary [00:48:46]: But only if you're over 18. Yeah. So let's, let's talk about something that's not AI. It's something that's come up with for me a lot in the last month, and that's software subscriptions, which we've talked about before. The reason it's come up a lot recently is Apple, of course, created a new software subscription service. Called Apple Creator Studio, right? And of course, people are comparing it to Adobe subscriptions, to Microsoft's Office subscription, to many other subscriptions out there. And people have a lot of opinions. People have a lot of strong opinions about software subscriptions. Gary [00:49:30]: Yes, they do. And it's, to me, you know, I hate the subject. Because it's hard to convince somebody who hates software subscriptions to feel okay about them. But on the other hand, as somebody who's a software developer, it's not hearing the hate. The hate is always directed as in those greedy software developers, those greedy companies, you know, that want more money, that want us to keep paying, that want us to be locked into their systems and all of that. But software has been sold as products Since mostly the '80s, I mean, even before that, but the '80s, the software industry kind of started to pick up, and I entered the software industry in the '80s with little things that I made and would try to sell, you know, in magazines to schools, you know, that kind of thing. In the '90s, I entered it in a bigger way, and actually, I did sell at least one very professional software business product in the early '90s. I experienced a lot of different ways to sell stuff. Gary [00:50:34]: I had what people called shareware. I had some shareware games in the early 2000s. The thing that a lot of those have in common is that they had the selling software as a product kind of thing where you create a piece of software, you say you could buy this for $40, and then it's yours. You install it and we're done, like our monitoring transaction. It's finished. People used that for a while and it seemed to make sense because software was very new. It's the easiest way to sell things. It's just one-time fee. Gary [00:51:13]: Here it is. However, it was not sustainable. It took a while for software developers to figure that out. One reason is because technology moves so fast. I mean, if you sold a cool piece of software that ran on the Commodore 64 in 1986, you didn't care that much about like, oh, people will be using it 20 years from now because people wouldn't even be using it 20 months from now, right? They would already moved on to like another computer type or whatever. Technology moved really rapidly then. So either the software you made then was so out of date that Nobody was buying it anymore after a while because things had moved on, or things had moved on and your software was so good that you came up with a new version of it that ran on the newer hardware with the newer operating system, and the new version of it you were able to sell as a new product. That kind of is where the industry went for a while was we can still charge one-time fees because we're going to come up with version 2, version 3, version 4, and all of that, and each one is going to cost. Gary [00:52:21]: The feeling was you were still just buying it once and you weren't committed to buying it again, but in actuality, you were now on a subscription. You just could opt out at any time, which you can do with the subscription. You just could say, I don't want version 3, I'm fine with version 2. The problem is that your version 2 may not work on the newer operating system or the new computer you just bought.. So you may be forced to buy a new one. Sometimes there were upgrade discounts. If you already owned version 2, you can get version 3 half price, that kind of thing. But it was still unsustainable. Gary [00:52:57]: You still had tons of companies that would have a hit piece of software. They'd sell a lot in the first year, and then it would drop off considerably because they had saturated their customer base. They couldn't reach more people that needed their software. And they made all their money in that first year and now they couldn't make any money in the second year because they didn't have any new customers. It wasn't a consumable like buying a piece of food. You didn't need more. You could just keep using the same one you bought. Yet they still had to support it and if they wanted there to be, to continue developing that as a living, making a living, they had to find another way to come out with a new version or whatever. Gary [00:53:40]: Make another piece of software that they could sell once, and that would get them through another year, that kind of thing. It was just not very sustainable for software developers, and that was true even for big companies that seemed to be like, well, we're not— this isn't like a 3-person team, this is a big company like Microsoft, like Apple, and all that. You still have a team that still needs to justify itself, and if they sell a million copies of their piece of software the first year but 100,000 copies the second year because everybody bought it in that first year, there were going to be layoffs, right? People couldn't keep working on that software because there was nothing new to do. You've made all your sales. Now it's just maintenance moving forward. So software subscriptions kind of became inevitable as the only way to really have a sustainable industry where people could build software and maintain it, not have to pull out tricks like there's a new version out this year and you have to pay again. And continue to support it and all that and stick with the same piece of software too, not say, well, I've moved on from that because I can't make any more money with that. I have to make another piece of software that does something different. Gary [00:54:54]: So now you've got that, but then of course every time people switch a product from one-time sale to subscription, people get angry. And it still kind of makes sense when you have a new piece of software to sell it as a one-time fee because you don't even know if it's going to work. You don't know if enough people are going to buy it. So you say, "Hey, this— okay, I have a new app. It's $30. Buy it." Then all of a sudden, it's a hit, and now you're like, "Oh, now I'm going to have to switch to subscription if I want to keep it going in the long run." That's my take on software subscriptions. Is people were already doing that because they were paying for new versions. Software subscriptions are just— it's just what's needed for there to be good software out there, and they also have their upsides. Gary [00:55:44]: So Apple Creator Studio, for instance, the main things, the main 3 apps that you get are Final Cut Pro, Logic Pro, and Pixelmator Pro, and they were previously $400 $200, or was it $300, $200? Anyway, there were a couple hundred bucks each, and Pixelmator was like $50. One-time purchases, which the thing about software subscriptions is that there were a lot of people that just could never really use Final Cut Pro because they didn't have $300 to just get it. I mean, there was a trial and all that, but they couldn't get it. Now, $13 a month You know, people have $13 a month. If your plan is, I want to be a video editor, I don't have $300 now in addition to what it cost me to buy my new MacBook, but I want to start creating videos, then I could spend $13 now this month, next month, the month after that. Hopefully over time my business will grow and the $13 a month won't be a big deal. The $300 a month right up front at the beginning was a big deal. So there are one time, one time. Gary [00:56:54]: Yeah. So there are, you know, good things about it as well. Leo [00:56:59]: Complicating things. Yeah, go ahead. Well, if I had to characterize the biggest unstated objection, so I see this all the time, especially when Microsoft Office went to subscription. But your observation that It's always been a subscription model. It's just been in your hands rather than the company's, right? It's just been auto. It's just been a manual subscription rather than an automatic subscription. That I think highlights where people are really uncomfortable. They have no control. Leo [00:57:42]: Once you subscribe, You, you, you know, it's, it's the only way to get the product. And you will get the updates whether you want them or not. Yeah. And especially in the Microsoft sphere right now, that's a big deal because, of course, Microsoft is throwing Copilot and OneDrive and all this other stuff into these packages. People just wanted a word processor, right? They just wanted Word. And now all of this other stuff is coming in. Now, Microsoft, of course, has done pretty much what you just described with, with the Apple offering, and that is they've made the subscription very lucrative financially. If you actually were to do an apples-to-apples comparison, and I don't mean oranges to oranges comparison. Leo [00:58:36]: We'll do it that way. The Microsoft Office subscription is an amazing value because you could have 5 different people with it. You get all this storage, you get all these products. If you had to build that yourself using the one-off purchase plan, you'd be talking 10 times the cost. And yet that's not enough to make people happy. Gary [00:59:01]: Um, yeah, it's— they— yeah, go ahead. No, yeah, and there are, you know, there are ways to be smart about it. Um, but yeah, you're right. People, they think they— I guess they're maybe uncomfortable with the automatic thing. They're uncomfortable with the fact they've got so many automatic payments now. You know, there's Netflix and then there's their internet provider and then there's all this stuff. Yeah, it does make it a— I guess, but— and of course people don't like being bundled, but I think bundling has its upsides too. Like for me, I definitely use Pixelmator Pro all the time. Gary [00:59:48]: I don't use Final Cut Pro as much, but I do use it. Logic Pro I didn't even have, like I wasn't using at all, but I wanted Logic Pro. I just couldn't justify, right, buying it in the past because I was like, what am I gonna use it for? I'm gonna play around with it. I don't know. So now with the bundle, I'm excited that I have Logic Pro to play with. And I— if I didn't have Final Cut Pro, I feel the same way with Adobe's Creative Cloud, which I have for It's a big bundle that I only use a few things of, but I like the idea that I have Premiere. Adobe Premiere is like a big video editor. I, I, it's been several years since I've looked at it, but I, I kind of like the fact I could use it. Gary [01:00:33]: And once or twice a year I have to do something in Adobe Illustrator, you know, that would be a pain to do if I didn't have Adobe Illustrator. And I like it. So to me, it's not so much that When I think of, like, for me, the exact example would be Logic Pro as an audio tool and Pixelmator Pro, and it's like, okay, Pixelmator Pro is a no-brainer. I use it every single day. Logic Pro, it's definitely adding value to the subscription. It definitely makes it— and I think a lot for people like Microsoft, you might use Word every day, and that's why you've got the Office subscription. But Excel does come in handy. Ah, tax time. Gary [01:01:18]: I do like, I do like to go through my, you know, I could download the CSV from my bank and I open it up in Excel and I could do a few interesting calculations and it helps me make some informed decisions at tax time. And it's the one time I use it, but boy, I kind of, it's a nice bonus to have that. Leo [01:01:35]: It's funny, I'm slight, I'm in a slightly deeper situation with Microsoft Office. I use Excel every day. I use Word occasionally, and I use OneDrive some. Yeah, that's like maybe 33% of what's included in the bundle. I ignore all of the rest. I just, I just have no call for it. I suppose it's nice that it's there, but, you know, Am I really going to fire up PowerPoint just because I've got PowerPoint? I don't think so. You know, there are other ways to do, to do that kind of stuff. Leo [01:02:13]: The other, the, to go back just briefly to the cost thing, I have characterized Microsoft Office as the cheapest way to get a terabyte worth of online storage. In other words, if you want online storage, you can get a terabyte for $99 a year, I think it is. Oh, and by the way, they throw in all these applications as well that you can ignore if you choose to, but you've got online storage. Um, I don't know, it's tough. I, I, I get what you're saying about developers needing to be paid, uh, because I was one, but the counterargument if you will, is that open source alternatives suffer from roughly the same thing with less revenue, right? They're not doing a subscription model for the most part. They are doing repeated versions and innovation development hasn't stopped because they don't have a subscription model. Um, so I don't know. I don't know. Leo [01:03:30]: And it's also really, really hard to take a look at companies like Microsoft and Apple and Google, uh, with their record corporate revenue, saying, well, yeah, but they need the subscription models to stay afloat. Gary [01:03:49]: I had don't see it. Yeah, yeah, maybe. I mean, there are— I, I like to— I've been getting a lot of comparisons between, you know, Apple Creator Studio. Like, people say, oh, compare it to Microsoft, it's blah blah blah. I'm like, well, Microsoft's— the one you're comparing to is an office suite, right? 3 office apps. Apple's is 3 creative apps: video, audio, and image. And but Because Apple says we've got extra stuff for Pages, Numbers, and Keynote, people then tie it into Office. It's like no, Pages, Numbers, Keynote are still free on Mac. Gary [01:04:24]: You can't compare to Office, which has 3 centered around the flagships are 3 Office apps. The flagships of the Apple one are 3 creative apps. You could try to look at Adobe's., but Adobe's got like a ton. Leo [01:04:42]: There's like 20-some apps in there. Gary [01:04:43]: Oh yeah. It's amazing number of things. Yeah. Yep. And, and then it's a lot more expensive too. And then Canva has got a subscription which also overlaps some things that Apple's doing, but not quite like Canva. There's nothing like Final Cut Pro on Canva. You, the, the biggest value without a doubt on Apple Creator Studio is Final Cut Pro. Gary [01:05:06]: Mm-hmm. And there's nothing on in Canva like that. Like full-fledged video editor, you know, build a Hollywood movie app. So it's frustrating. I've just started, and what's one of the things that's frustrating for me is I've left the point where I'm doing videos about the Creator Studio. Like, is it, I did a video, is it worth it? Another one, it's like, you could still keep using Pages, Numbers, and Keynote, you know, without the Creator Studio stuff. I've left the topic of what is the Apple Creator Studio behind, and now I'm dealing with like, oh, here's what you can do with the Apple Creator Studio tools to help you with presentations. So it's using some of these tools in an app. Gary [01:05:52]: I'm still getting comments by people commenting on how they feel about software subscriptions. Of course. I've just started deleting them because if it's not on topic, I delete it. And it violates a few of my other rules as well. And so I'm like, no, this video was supposed to be about creating presentations. I'm not treating Apple Creator Studio as like, should you get it? Should you not get it? What are the advantages, disadvantages? Nothing. I'm like, we're building presentations and I'm showing you that using some of these tools. If you have feelings about software subscriptions, create a blog, create a Substack on software subscriptions and post them there. Leo [01:06:31]: Anyway, I don't think it's a problem that's going to go away anytime soon. Gary [01:06:34]: Um, yeah, I think, I think it's getting better. I, I think it will have to get better because they're not going anywhere. Software subscriptions are sticking around. Leo [01:06:42]: People are going to get more and more used to them, and the complainers are sticking around. I have to get more and more used to them. They're not going away. Um, and yeah, it's like I said, it's almost, almost goes back to what I was saying earlier about the other problem. I'm not sure it's a solvable problem. I don't think you're going make everybody happy. Gary [01:07:07]: Um, I— yeah, yeah. On that note, on that note, I should say that, well, two things are cool. One is that while we've been talking, my entire handwritten Australia journal has been transcribed from the handwriting. Is it halfway accurate? Uh, I think so. I mean, we've been talking, so I really haven't been able to look at it. I could read it, which I can't read my own handwriting, so that's kind of nice. Um, I got excited because I thought at the end that it actually totaled up how many miles I traveled during the trip. And then I thought, well, then I got mad because it's like I didn't ask for that. Gary [01:07:44]: I just said give me a trans— you know. And then I looked at the PDF and no, that was me in 1996 actually totaling up the miles. So it did stick to that. So I'll have to review it and now that it's readable, I'll have to read it. And see if it— Leo [01:08:00]: how many mistakes it made or if I can tell. But one of the things I've done along these same lines, and you might have some fun with the AI, what I do with a lot of my video transcripts is I'll go ahead and, you know, I'll grab a transcript from my video, be it YouTube captions or run it through one of the speech-to-text tools. And then I will ask AI, okay, this is a transcript. Don't change any of the important words, but clean it up. Yeah, yeah. And it does a really, really nice job. Oh, and since most transcripts like that come through as either blocks of text where the block, the breaks between blocks make no sense, or it's just one big monolithic block of text. So you know what, clean it up, add paragraph divisions where it makes sense, that kind of stuff. Leo [01:08:58]: And the results have been really, really impressive. So that might be a fun thing to do with this transcript, is— Gary [01:09:02]: or with this OCR, is, you know what, here's my original journal, clean it up a little. Yeah, or I might just ask it for information, like give me a bullet list of all the places I visited instead of making me search through them and stuff. I mean, if I'm gonna— I'm gonna try to have it draw me a map. Oh, neat. Of the route, and let's see what it comes up with there. But my real Ain't It Cool was, I have been watching The Knight of the Seven Kingdoms, the latest, you know, side story from Game of Thrones. Right. And quite enjoying it because to me it's, yeah, it's Game of Thrones, but it's also the opposite of Game of Thrones. Gary [01:09:43]: Because Game of Thrones throws you into this whole world where you're instantly hit with things you need to know about, like houses and families and politics, and you gotta keep it straight and all that. And then this story is fully a, you don't need to know anything. There's just a guy, and he wants to be a knight. Go. And I love stories like that where it's like, oh, I don't feel like I needed to, I, you know, I needed to do that. Leo [01:10:15]: So yeah, that's pretty cool. I'm simply going to echo that. Um, I've— Gary [01:10:18]: are you done with it? Have you watched the whole thing? Oh, uh, no, I've not. I'm not. Leo [01:10:23]: Um, I think I'm an episode, a couple episodes behind. So yeah. Oh, okay, cool. Gary [01:10:28]: Well then, I won't spoil anything. Leo [01:10:30]: Yeah, don't spoil anything. Better not. Um, but yeah, same thing where Game of Thrones itself is this multi-layered, multi-threaded political intrigue across continents. Um, like you said, this is a guy. This is the story of a guy. And, um, it's actually a very well-done story. And I am actually looking— I'm hoping they'll do another season. Uh, this season was what, only 6 episodes, I think? Yeah. Gary [01:10:56]: I want it to be longer. I want it— no, I think it— I don't know, I'll have to check to see. But yeah, um, self-promotion. Yeah. Uh, here, so the opposite of like AI and advanced software and all that is I did a video on just how useful it is to have little text files around. Something that you and I, because we've been using computers for such a long time, know. But, you know, on a Mac, you've got a little tool called TextEdit, and you can create little text files. And it's so useful. Gary [01:11:25]: You just put text files inside of folders for the stuff you're working on. They're kind of like notes, but instead of being in an app where they're off on their own in a journal or notes or whatever, They are with your files. Right. So you could leave notes for yourself telling you what these files are, notes for somebody else, notes to re— you know, whatever it is that you want for whatever purpose, they take up virtually no space and they're easy to make. Leo [01:11:53]: So people are responding good to that. One of the reasons that I use the note-taking system that I do, I use Obsidian for my note-taking. Is that its native storage format is text. So even if the tool goes away completely, you've got text files and they're readable because it's all stored in Markdown. Yep. The other thing I love about text is essentially a side effect of what I was talking about earlier. Anything can read it, including AI. So your little text files, your collection of text files for God only knows what, even if you lose track of them or can't find them or whatever. Leo [01:12:32]: Yeah, turn Claude code loose on your hard disk, tell it to locate all the text files and tell you something about them. My self-promotion is an article called The Nirvana Fallacy. It's askleo.com/179344. It's something that you and I have talked about before, except we haven't given it a name. I had a reader give it, give it a name. And that's the, the reaction that if something isn't perfect, it's crap. So if you're, the application you're using has some silly thing wrong with it that you really, really hate, it's not the thing that's wrong anymore. It's the entire application. Leo [01:13:20]: And you throw that baby out with the bathwater. The example that I end up using in that article is someone who was complaining that, I think it's File Explorer, wasn't displaying file sizes in his preferred notation. And it wasn't an option. Therefore, File Explorer was crap. And that, down that path, that path just leads to more and more frustration. Nothing is perfect. You gotta come to grips with that. Gary [01:13:53]: And you'll have a much, much happier life if you can tolerate a little bit of crap in applications. Interesting. I'm going to read through your article and look into more about this. Add it to my list of things like the XY problem. Sure. And other kind of ways that philosophy and technology kind of intersect. Leo [01:14:17]: It might make an interesting grouping. Yeah, yeah, this, um, of course, um, I link to the Wikipedia article that talks about the more general concept of the Nirvana fallacy. Of course, I'm looking at it specifically from a tech perspective. Yeah, yeah, definitely. It's interesting. Cool. Alrighty, I think we've done another week. Um, as always, thank you everyone for listening, and we will be back again real soon. Leo [01:14:44]: Take care, everyone. Bye-bye. Bye.