Ep. 41: Who Is All This AI Actually For?

[00:00:05] Welcome to Prompting Curiosity, a podcast for the AI curious. No coding background required. I'm your host, Dr. Shantae Cofield, also known as the Maestro, and I created this show to explore what these AI tools actually are really, though, are the files in the computer, how to use them, and what they might mean for how we think, work, create, and move through life. Whether you're skeptical, intrigued, or already experimenting, you're in the right place. All that I ask is that you stay curious. All right, let's get into it. [00:00:38] Hello, hello, hello, my curious people, and welcome to episode 41 of, uh, prompting Curiosity. I'm your grateful host, the Maestro, and today I'm going to do a brief. It's actually not so brief. I look at the outline that I have. It's not so brief, but I'm gonna do a rundown of, uh, the big AI advances over the past few weeks and look at them through the lens of who is all of this AI actually for? And do you even need to care? Right. These new models, they keep being rolled out and every week there's a press release about some new something or other. [00:01:13] And I am generally at least superficially aware of these big events and these changes in the AI space, largely because I am on threads a lot. Um, and also, um, and I should say publicly, preferably because, uh, I subscribe to a handful of AI focused newsletters, namely via Substack. I had a person who don't like Substack, but that's where these people are. So I read them on there. Um, so I'm going to include those in the show notes as well. I'll list them out here. I'll speak to them, but, um, I'll name them. But I figured I'd share that with you folks as well for any of you that may be just, you know, interested. So my, my, my current favorites, my top five, I think I have five on here, is number one, Bennett. And these are in no particular order. It's just like they were in this order in my inbox. So that's what they listed out like that. [00:01:58] So, number one, Benedict Evans newsletter. Uh, number two, Handy AI by Jake Handy. Number three, uh, Andy Maley's or Masley M. I'm not sure how he pronounces it. His blog. Um, he doesn't seem to have a name for it. [00:02:11] Number, um, four. And this probably should be number one because it's my friend. Uh, it's called Future proof your career with AI and it's by my good friend K. He. [00:02:19] And then lastly, Confluence. Um, it's the name of the blog is Confluence. AI Leadership and Communication by cra. So I will link all those in the show notes if you want to check them out yourself, you don't have to because I'm going to tell you about them. Tell you what the things, you know, that I say abreast of. Abreast with. I don't know what the right word there is, um, but those are my favorites. So I have spent the past, you know, few episodes doing more how tos and they've been more technical episodes, uh, and more technical topics. So I figured I'd just let today's episode be kind of an easy. Just listen, take it in State of the Union, State of the AI Union, if you will and share my two pennies on what I think is actually worth being curious about and also kind of the bigger picture here. So I'm going to kind of start with the conclusion, the bigger picture which is uh, that all of this AI activity that we're seeing, the model drops, the product launches, the pricing shifts, it is all for builders, coders and developers. The average user definitely, in my humble opinion, doesn't need to concern themselves with any of these advances, nor do I think they'll even notice the change. It is nice to, you know, be aware of it. Especially you know, I spoke about Mythos before. We still don't even know what the going on with that, but I think it is nice to at least just be aware. But at the same time like we can't really do anything about it. Uh, and I, I just think it is nice to have your finger on the pulse of these things. [00:03:40] Um, at least be like peripherally aware. But in our day to day, our immediate lives, our immediate usage, uh, we typically, you and I won't, won't really notice this change. However, I shouldn't say you or I, for the very average user that's just like, you know, asking basic questions, they likely will not ever notice a change. For aspiring vibe coders and curious folks like yourself, uh, some of these advances might be interesting and you actually might feel the difference. [00:04:12] But I don't think that any of these advances are actually necessary. So you know, there is a lot of talk and there continue to be, continues to be a lot of pushback against AI And I agree with it, right? We don't need all these data centers. We don't need bigger, badder, you know, more AI. We don't need it. What we have now is great and it really is like a dick swinging competition. Dick Measuring competition of like, who's got the bigger model? And you're like, for what? [00:04:37] Who's got the bigger data center for why? What are we doing? Like, uh, we don't need that. Right? [00:04:45] Uh, the models are already incredibly capable and they are fully able to meet our, our standards and rather meet our. [00:04:53] The question to be answered and the question that was posed in the title of this episode, who is all this AI actually for? Again, it is for builders, coders and developers. That's who's going to notice it. I can't say for certain that these, that these builders and these coders and developers actually need it, want it, demand more of it. I think that the demand is being, is coming from the top, not from the users, right? It's coming from people who want to make more money and standing to make more money. And they're just like, we're going to have the race to AGI. And that means that at some point AGI is artificial, ah, general intelligence. So like, you know, think Skynet. [00:05:29] Um, you know, it's, it's all about a play for, for money, right? For this thing may be able to do this in the future, which means that it'll make a lot of money because it's going to replace workers and this and the other thing. And that's who's driving the growth, the unnecessary growth of, of of these models. But who is this act? Is this AI actually for like who's able to use these better quote, uh, unquote models? It's builders, coders and developers. Right? So let's, let's start with an AI use overview, right? Because this episode is large, I want it to be like, you know, state of the state of the AI union. So as it relates to the AI use overview, daily active users, um, and this is across, um, across the different platforms we have chat GPT very much in the lead with 45.3% of the market. Gemini 25.2%. Grok at 15.2% and then co pilot Perplexity, Deep Seek and Claude collectively splitting the remaining 14 scent. Right? You see, like maybe it feels like more people are using Claude because that's, you know, that's the world that I'm in. But I'm fairly fully aware that like it is not the, the top, the top dog by any means. [00:06:40] Um, and then as it relates to the average person and that's why my, you know, originally this podcast was named Chat GPT Curious because I know it's a household name, the average person using that you and I, we're not average. We're. We're coding things, we're building things. We're looking to help it be. Help you help looking to have it help us be more productive so that we can go and, you know, go touch grass. Uh, but that's the breakdown of the average US mobile app usage, right? Um, so as it relates to like actual numbers, like, okay, 48.3%. What does that mean? [00:07:10] ChatGPT apparently hit 900 million weekly active users as of February 2026. [00:07:17] Google's Gemini hit 750 million monthly users, which is making, which makes it the fastest growing platform right now. [00:07:26] Which I'm like, what? Who the is using Gemini? I know Rachel's using it. If you listen to this, Rachel, love you. Uh, but I'm like, is the worst. [00:07:32] I, I had high hopes for it and I'm like, it just doesn't do what I wanted to do. [00:07:37] But Claude comes in at the bottom with about 18.9 million monthly active web users, which is small by comparison. So if you're like me and Claude is your daily driver, welcome to the small club. The small kids club. Right. [00:07:52] Please note. Right. I want to highlight there that some of the numbers are reported in weekly average users. Some are. The numbers are reported in monthly average users. [00:08:02] It's impossible to find like just a graph, a chart with these, with these numbers, comparing them and these. This is intentional, right? These companies stay being intentionally shady and confusing with their reporting, right? It is intentional. This way we're like, well, we have this many users maybe, and then this percentage, maybe like everything is soft and fudged, right? I tried to dig in and uh, and find data about paid versus free users. And honestly, it's like impossible to find anything. These companies don't disclose anything. I don't really understand how they don't have to, but they don't. And, uh, so to me, this lack of transparency continues to be fully worth flagging and should keep users like us, uh, skeptical about all of these companies and also very prepared for pretty much anything to happen. [00:08:49] But most likely what will happen is something directly out of Uber's playbook. I have this again and again and again and again. And it is coming to fruition, right? When I say something out of Uber's playbook, I mean, increased prices. We have heavily subsidized initially, very low pricing. You get addicted, you get hooked. You're like, this is the best. And then they take away the subsidies and they pass on those costs to the user. And it's. We see significant increases in these prices. We. If you take an Uber or Lyft, you've seen it happen. [00:09:19] There are already starting to be rumblings and rumors about Anthropic removing Claude code from their 20 Pro tier, at least for new users. Which would mean that in order to use Claude code, and this is like, if you want to be able to vibe code things, that's what you're using, you'd need to be on at least the max 100amonth plans 5 times as expensive. [00:09:42] And what we're seeing is open AI people are switching back on opening. I don't, honestly folks, I don't know what happened with the, uh, the whole resistant unsubscribe movement as led and championed by Scott Galloway. Am I surprised? No, I'm not surprised. He's a, he's a multi millionaire and you know, he got on board for a little bit but like this doesn't actually really affect him. Come on now. So it has fizzled. I'm not surprised. But people are moving over, going back to coders, at least. I don't think coders ever actually left and stopped using OpenAI. I think the average person was like, yeah, I can get my needs met from Claude. It says I'm just asking it random question, basic questions. But most people that are in the coding space, many people in the coding space, they're using both, right? They don't, they use this for this thing, use the other one for the other thing. [00:10:29] And so what we're seeing is that folks are going back to using, uh, OpenAI more, they're using Codex more. And I am not surprised, right? All that these companies care about is money. And I, I feel like that, feel that to be the truth, especially for OpenAI, right? So the people come back, ChatGPT. ChatGPT is keeping their prices low for the coding sector to get people back and they'll vase it at some point. It's going to go up at some point. It's just a matter of when and how those costs get passed on. But I am not surprised that as we're seeing Anthropic, which I think has been more, you know, fiscally financially responsible of all the companies as we see them start to tighten the reins on things. I mean like, yo, these costs are out of hand. Like, of course this is gonna happen at some point we see them doing this and you know, now people that were loving these, you know, insane usages, insane amounts of, of token usage, they're like, I'm going back to this cheaper thing over here. I'm going over to open AI and OpenAI will gladly take their money. And at some point, you know, uh, when, I don't know, but at some point it's gonna be like, yo, we're not making any money. And they have to generate that income somehow. But as it currently stands, OpenAI has not yet unbundled, um, things. Um, we know that they did begin testing ads in the free and go tiers back in early February. [00:11:47] Um, but all that to say we should expect changes in pricing and changes in user experience. It is only a matter of time. [00:11:56] So next up, let's move on to the advances and the updates. All these rollouts. Again, I just want to use this episode to like, hey, here's what's going on in the world of AI. If you care, here it is. If you don't also find. Uh, but M, let's start with Claude, um, because Anthropic has been busy. So Anthropic released a new flagship model, open. Excuse m. Me open. Wow. Opus 4.7 on April 16th. And that got mixed reviews. Apparently it has stronger coding, better, uh, better reasoning, uh, and built for longer, more autonomous tasks. Um, but there's also a cost issue that's worth naming as they change the tokenizer and it burns up to 25% more tokens per task. Translation, it is more capable, but way more expensive model. And maybe it's not more capable. This is, uh, you know, based on the benchmarks and the testing, more capable, but from what I'm seeing from users, they're not liking it and it's costing them more. Have I used it? No, I have not touched it. I actually haven't used OPUS at all ever. Sonnet4.6 is my daily driver and that is what I use for Vibe coding. So for those of you that have no idea what the fuck I'm talking about, yes, you can select what model you use with Claude. ChatGPT used to have this. I spoke about it in the early days of this, this podcast, and they had all the different models out there, uh, and then they simplified things considerably in early 2026 and now they have GPT 5.3 and it's instant thinking and pro. But if you're on one of the paid plans, ChatGPT uses an auto switcher, uh, an auto switching router, and it picks instant or thinking based on the complexity of the query and it routes it for you. Um, so going back and forth between the models on Claude, if you want, my suggestion is just stay with Sonnet 4.6. Stay with the cheapest model. Again, it is perfectly capable of what we need to to do. [00:13:52] We are not super high tech advanced coders. [00:13:56] Yes, I have heard that the trade off is that Sonnet uses fewer. Sonnet is uh, in theory can use fewer tokens because it's not as token hungry. But because it's not as smart of a model, there could be more back and forth because there's more mistakes made which ultimately could lead it uh, to using more tokens. But we know that right off the bat Opus is way more token hungry. Like yes, I may get it right on the first try maybe, but it's going to use a lot of tokens. So my suggestion and I have been more than able to build all the stuff that I talked to you about in those episodes Simply using Sonnet 4.6 which is what is the default when you are using Claude. Okay, uh, next, with, with anthropic updates, Claude Design, uh, Cloud Design launched on April 17th and I have yet to use that. But uh, it did make a splash and it also got a fair amount of complaints as is to be expected, right? Because these things are just doing complex math. Like people want this to like look amazing, be perfect. And I think people maybe should start realizing how incredible humans are because they are able to do these things well instead of being like oh, the computer can't do it and it's like, yeah. Cause it's just doing m fucking math. But quad design lets you create visuals that can go straight to Canva, PDF or PowerPoint. It is built for people who are not starting from a design tool and they need to get uh, you know, from an idea to something visual quickly. [00:15:23] I haven't used it. I know about a lot of the launches. I by no means rush in to use or experiment with a lot of the, the updates and launches because like I said, what's already there works great. That is how I live my life. I'm like, does this thing work cool? I don't if it ain't broke. I uh, don't need to go use this new because that thing's probably going to break. I hate updating my computer, I hate updating my phone because all the updates always break one Next one Next we got for the uh, also in the launch streak from Claude, Claude for Word, which rounds out the PowerPoint and Excel trifecta. And we have a rebuilt cloud code desktop app, uh, which supports multi session support, um, which supports multiple sessions running concurrently. [00:16:10] The thing that's kind of annoying here and I don't Know if you can hear it in my voice is like, last, just last episode, I did a whole episode on Cloud Code Desktop. And in the week between me recording that, publishing it, and now this one, they already changed the desktop. It's like a subtle change, uh, functionality for what we do. We won't notice the difference, but I'm just like, this is. It feels very much like one of the reasons I left. Uh, I was very happy to walk away from my Instagram intensive course because things change so fast. You're like, oh, another update. Another update that is at the heart of why I'm doing this episode because I don't think you need to care about a lot of these updates. Right? Uh, you can just stick with what you're using. It is plenty good. And for those of you that do care, you know, here's me, me, uh, speaking to these, these updates. But I think something that we can take away from these rollouts, namely the, you know, Claude for Word, um, uh, the cloud design, is that this speaks to Anthropic's efforts to make their tools more accessible, more useful and easier to use to be used by the average individual. Right. Anthropic's wheelhouse is definitely coding and they go enterprise, which is why they have such a small, you know, market share. They're, they're splitting that 14. [00:17:23] But they are clearly making efforts to appeal with, to folks with less technical background. Um, okay, let's move on over to the, the big dog open AI. Right, chachi pizza. So OpenAI has built and launched what they are calling a super app Spoiler. It is giving Apple, you know, an Apple new release and all the folks with, with green text being like, we've been had that, we've been had that update that camera that folding that, whatever. We've had all that. [00:17:54] I have yet to try, uh, OpenAI's new super app. Remember, I am all cloud all the time, baby. I, I put my prompt where my, my money, where my prompt is. Uh, but based on my digging, uh, it appears to be OpenAI's version of Claude Desktop, which again, if you don't know what I'm talking about, Cloud Desktop. That was literally last week's episode. Episode 40. You can go give it a listen. But, um, it appears to be OpenAI's version of Cloud Desktop with fewer rules. I'm not surprised. OpenAI is like just go crazy. [00:18:23] Like it's not, it's not good in my opinion. And a bit more power. And they do have an upgraded, they do have upgraded image generation Again, cloud doesn't do image generation, although they have this new design thing, so they can do it in that regard. Um, but just like making pictures and things, uh, that is OpenAI. [00:18:42] Uh, so this super app, you access it via the Codex desktop. Codex is OpenAI's version of Cloud code. That's all. Uh, so you go to the desktop and it's currently only available for Mac. Um, and what it does is it bundles ChatGPT, Codex and their Atlas browser. I also did an episode on their Atlas browser, um, as well. But it bundles all of those into a single desktop environment. It is literally the same thing that Claude has come out with. Um, and so with this new quote unquote super app, Codex can now connect to your apps, it can control your desktop, it can edit presentations, it can run multiple AI agents in the background simultaneously and it can browse the web. Again, it is largely just open AI's version of Cloud Desktop with a bigger PR moment. Uh, right. The other big rollout though that I think is worth noting, uh, from OpenAI, because this is different than, very different than cloud, is GPT2 image 2, which dropped on April 21st and is open. [00:19:43] Uh, and OpenAI is calling it a new era of image generation. [00:19:49] Lots of positive reviews on this one, uh, from users and you know, for my nerds in the audience, uh, that is because it uses what uses, uh, an auto, uses autoregressive models instead of diffusion based models or a diffusion based approach. [00:20:05] Um, and that's what powered Dall E. Remember that D A L E. And Sora, remember Sora? I, um, did an episode on that as well and talked about diffusion. Um, is a little bit different between Dall E, Dolly. I don't pronounce it a little bit different between Dall E and Sora. Um, but grossly different from what GPT Image 2 uses, right, that auto regressive model. And that's why it's like so good. [00:20:28] Um, it's. It. To me, I think it's, it's worth noting and taking a step back, right, because it's easy to get in the, get lost in the weeds. Although I think you folks listening to this, we probably have similar brains here, but people get lost in the weeds and they're just like, you know, they get so, uh, so willing to clap for these upgrades and it's just like a human could do this on the first try. [00:20:49] Like the fact that these companies rolled out these models that weren't good at image generation and they were bad, objectively bad. [00:20:59] And now that they are good. People are like, look, it's amazing. And it's like, okay. And also a, ah, human could always do this. I think it's just worth like putting that into perspective that there is clapping going on for what a human is expected to do, right? And you're expected to draw five fingers and you're expected to spell that word right. And now that the computer's doing it, it's like, yay. It's amazing. But for what it's worth, a lot of the, the kinks have been worked out and the words that get spelled and things getting spelled and your ability to change just like the color of a shirt is significantly improved. Uh, and yeah, that is worth noting for the, the uh, rollout of GPT image 2. Um, but circling back, speaking about Sora, OpenAI did opt to shut the app down, thank goodness. Due to unsustainable computing costs and the web and app experiences. That app, it will close down officially on April 26th. Right, so that is the update, the news, the state of affairs for both Anthropic and OpenAI. [00:22:07] On to the other two players, another two players, which would be Google and Meta. I'm going to keep this section incredibly brief because there really isn't much to say about them. Um, but it does support the premise of this episode, which is that all of these AI updates are largely going to affect coders and developers and builders. All right, so Google dropped Gemini 3.1 in February and Meta dropped Llama for Maverick, both of which have been written to be or said to be competitive with GPT4.5, um, on all of the coding and reasoning benchmarks that folks like you and me don't actually understand or care about. [00:22:44] Google did also release a native desktop app for Gemini, uh, which again currently only available for Mac. So we see that they're all doing the same shit, right? All these. All these players, they're doing the same. [00:22:57] Something that is worth noting, that is different and I think is worth caring about. [00:23:02] Notebook LM now lets you export the slides. Wow, I like lost my voice there. Now lets you export the. [00:23:09] Um, I think maybe it's because I said export. Weird export. You know, you say a word too many times and you're like, it just sounds. This is a sound. I think that happened. You talked about that in um. What is that? So, uh, with the mustache. The guy with the mustache and he's a soccer coach. And uh, Ted. Ted. That Ted Lasso. Um, but either way, something worth caring about. Notebook alum now lets you export the slides as fully editable PowerPoint slide, which is actually a pretty dope upgrade. Pretty dope update here. So if you don't know, Google Notebook LM is one of my. It's actually legit, one of my favorite AI tools and is definitely worth checking out if you haven't already. I did a full breakdown of this in episode 25, so that was quite a bit of time ago, a few months ago. Um, but I will link that. Um, it's. I think it's a fascinating tool. It's a phenomenal tool. [00:23:58] You want to. You can literally learn anything, understand anything. So definitely worth checking that out. So lastly here, not gonna go into Grok because Elon and Trump. Uh, so that will round out the information portion of this AI State of the Union. What I really want to. What I really wanted to drive home with all of this and then kind of preventing this information and hopefully I didn't like bore you to death. But what I really wanted to drive home is, is like a handful of, of main points. First one, all of this AI and all these updates are largely for coders, developers and builders. Average folks. Doesn't impact them. Um, and we likely will not notice a difference. [00:24:41] Second, as is always the case, the big players are all doing the same thing. [00:24:48] They have to. It's like if this thing can do well, we got to be able to do it. If they're doing it, we got to be able to do it. I think this is a very bad business model. I think your best bet is to just keep your eyes on your own paper and just be the best at that. But yeah, no, I, that's, that's my take. Um, the big players are going to do what the big players are going to do and everybody's trying to do the same. So to that end, which LLM you use is largely just a matter of personal preference. [00:25:14] And we know many folks often use multiple models. [00:25:18] Number three, these companies are exactly that. They are companies. [00:25:23] They are running a business, which means that they are first and foremost, foremost concerned with money. [00:25:32] That's their main objective, their main goal. Despite whatever lies they may say, they care about money. So bottom line is the bottom dollar, top dollar, whatever the fuck the, the phrases. So get ready to start feeling that like I don't want to be like a doomsday prepper, but like that the pricing is going to have to get passed on. The, the. The money side is going to have to get passed on in some way, shape or form. When I first spoke about uh, when chat GBT rolled out in what, 20, 22, 2023, when. Whatever. It was November. Um, one of my biggest concerns that I. That I spoke about in that episode was that it would be almost like a de. Democratization of things. I think that AI is amazing because it can democratize things, right? I literally have this whole podcast automated workflow that I use, and, you know, I'm building all these personal tools, and it's costing me 20amonth. [00:26:31] That's phenomenal. Uh, in terms of access for people, that's incredible. [00:26:36] But as these prices start going up, who can afford them will change? [00:26:43] And that is, you know, that is something. Definitely something worth thinking about and at least thinking about. [00:26:49] And that is a. Definitely a concern to me. [00:26:52] You know, the flip side of the argument is that. And I said before, I am willing to pay more because I use this thing a ton. [00:26:58] All right, a hundred dollars a month. Do I want to pay 100amonth? No. But, like, it's incredibly valuable to me, and it replaces other tools, places. It allows me to, you know, solve. Solve problems and get back some of my time, and it's worth it, but not everyone can say the same. [00:27:19] And so that, to me, does become problematic. And we just kind of. That chasm just gets widened, and, you know, the rich get richer. And one of the things that I was so excited about with AI and that it could bring more voices to this, to all the spaces, you know, that that becomes less of a possibility. [00:27:37] So there is that. [00:27:40] Get ready to start feeling that financial impact. Um, and I. I will definitely. You already know, I'll be speaking more to that and covering that more as it. As it unfolds. But that's. We're definitely there. All right, Last thing that I want you to take away from this episode is that folks like you and me, folks like you and I, we do not need better models right now. We would be best served to simply focus on most efficiently using what currently exists and really leaning into building what would call what we will call these, you know, personal helper tools, if Vibe coding is in the cards. All right, all right. I'm looking at the time. This episode is getting a little bit lengthy, so we're going to wrap it up there, uh, with the. The State of the Union stuff. We got one last section, which is how I used AI this week. If you're new here, each episode I share a quick example of how I used AI that week. [00:28:34] This week. Speaking of those personal helper tools, uh, I use Claude code to Vibe Code, a personal web app that allows me to Edit and save fillable PDFs directly inside of Google Drive. Fun fact. I also made said fillable PDF using Claude a few months ago. And it's all, you know, stylized exactly what I want it to look like. Uh, Bell branded. I love it. [00:28:58] So before, I was using a tool called Doc Hub, and that only gave me three free docs, and then it charges $16 a month to get more documents. And it has like a bunch of features that I don't need. And I'm like, I literally just need to be able to edit and save, like, right into the fillable PDF and then save it. I want to be able to do it directly inside of Drive. I don't have to open up some other thing. I'm gonna do it right there. [00:29:24] So I built it right, because Doc Hub allowed me to do it. So I'm like, it's clearly possible to have this. Uh, and so I talked to Claude code and we built it in about two hours. [00:29:33] Uh, integrating it with Google Drive was actually much easier this time because of the familiarity that I've gained through the other, you know, builds that I've had with my, you know, putting the name at M, my dash, my dashboard, uh, but just the auth that's needed and this connection with the API. I'm like, I understand what I am where I'm at right now and like, what I need to be doing. [00:29:50] Uh, and what's called the, uh, env Keys. Environmental. I think it's the word environmental keys. I'm, um, like, okay, I am understanding this. Um, but I use it a few times a week. A few times a week. It's exactly what I need. And it's free, which is a big win. So, you know, to me, this is also why I'm m, you know, willing to pay more for claw, because the things that I'm building end up being free, right? I deploy them via Vercel. I build them, um, using, uh, the framework that I use is next. Uh, js. If I need a database attached to it, I use Supabase. And those have generous, very generous free tiers, Superbase and Vercel. [00:30:26] Uh, so it's like, it doesn't cost me anything to run these things. It costs me something to build it. But once it's built, it's done. And then it's the maintaining portion, which is why, you know, it's a bit of a different topic here. But SAS is not dead, right? Service as a. What is that? Service as a. No. [00:30:43] Software as a service. That's what SAS stands for software as a service. [00:30:47] It's not dead. Why? Because you can build things, but then you have to maintain them and you have to debug them. I love this podcast automation that I have, but it breaks every now and then. I have to go back in and be like, what the fuck is wrong with this thing? [00:30:58] And I do need cloud code for that. I needed to debug these things. And so it's like, how much are you willing to have to. How much are you willing to pay to have the debugger on staff? [00:31:07] Uh, so that the things that you build, which is like kind of a one time thing, then you can actually maintain them. So, you know, that is how I use cloud code this week. [00:31:16] Those are my thoughts on all these AI rollouts and these updates and who are they for? And that is all that I have for you today. It's a longer episode. I'm grateful for you. Hopefully you found this episode helpful. If you did consider sharing it with somebody, you know who you was curious about the old AI. Don't forget I have a companion newsletter, the Curious Companion, that drops every Thursday and is basically the the podcast episode in text format. So if you prefer to read or just want a written record of the things, you can join the newsletter fam you head to prompting Curiosity.com forward/newsletter or check out the link in the show notes. As always, endlessly, endlessly, one more time. Endlessly appreciative for every single one of you. Until we chat again next Thursday. [00:32:05] Stay curious. [00:32:07] Sam.