Ep. 10: Are We in an AI Bubble?

[00:00:00] Foreign. [00:00:05] Welcome to ChatGPT Curious, a podcast for people who are, well, curious about ChatGPT. I'm, um, your host, Dr. Shantae Cofield, also known as the Maestro, and I created this show to explore what ChatGPT actually is. Really, though, are the files in the computer, how to use it, and what it 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 10 of ChatGPT Curious. I, uh, am your grateful host, the Maestro, and today we are talking about AI Bubbles. More specifically, answering the question, are we in an AI bubble? Very short, sweet, simple answer. Absolutely, yes, we are. We can all see it. We are. So why am I dedicating a full episode to this? Well, like the title of this podcast, I'm curious. I personally wanted to do a little digging, uh, and put some numbers and stats to the feelings and offer up and think about what, you know, this bubble and the bubble bursting could mean. So here we are. Uh, but real quick, before we get into that, um, because I want to encourage more of this behavior, I want to share a question from m. The audience, the good homie, Rachel. She reached out and asked, is it better to use ChatGPT or what if I'm using Gemini? I don't really know why I'm using Gemini, but I am. So should. Should I switch? For those of you. For those of you who don't know, I'm having a tough time with the words today. For those of you who don't know, Gemini is Google's LLM, right? LLM stands for Large Language Model. Uh, my answer, use whatever you like. They all kind of do the same thing. [00:01:50] And the average user, I honestly don't know that they'll actually notice a difference. You know, when. When everything first came out, I was very bullish on Chat GPT, uh, because it was objectively better. [00:02:02] Um, but as is with the case with all things, the competitors catch up. So why did I actually make this podcast Chat GPT Curious? Because it's the. It's the most known. I gave myself permission to talk about other things. [00:02:14] It's also the model that I use. But, um, I gave myself permission to talk about other things AI related, but I wanted people to. When people. When you say LLM, uh, or no one says that, when you say AI or like Chatbot or, you know, people just know Chat GPT. It's kind of like kinesio tape. I worked for rock tape many moons ago, but that is a type of kinesiology tape. But the number, the first person, my first to market, they typically get the biggest market share, and people know what it is. And so if you're trying to educate the masses, it's easiest to use the. The household name, which is Chachi PT. Um, but like I said, I was bullish on ChatGPT because it was. It was objectively better. Um, but as is the case with all things, the competitors catch up. Um, if you have, you know, a specific use case, you might prefer one model more than the other. Or maybe if you're using a bunch, maybe like, I like this for this and this for that. That's cool. Go ahead. Um, but at this point in time, if someone was, like, just starting out now, it's pretty much, you know, it's pretty much vibes. Like, what do you like better? Okay, cool. Uh, and just go with that. Because they can all, like, do all the things. Uh, for example, Right, Right. We think about Google Home versus Alexa. Like, I'm a Google home person. This is a Google home house. Because I had, like, a million things, uh, coming in. I. I think back actually to my friend Caroline. She's always been into tech. She actually does WordPress, uh, websites, and she had a Google Chromecast. And this was like a zillion years ago. Like, like 2014. And I, like, legit couldn't understand what it was. I was. It was felt like the files are in the computer. I was like, the TVs in the TV, the phones and the TV. Like, what the. The computer's in the T. I didn't understand what it was. Right? And, you know, look how far we've all come. But, um, it's the same thing, right? They all do the same and just, Just pick the one that you like. So use whatever you want. And if you have any other questions, please hit me up. I would love to answer them if I can. And if I can't answer them, I'll make something up. I'll use Chat GPT and make up an answer. No. Uh, but it sparks conversation. And if you're wondering, I'm guessing someone else is wondering as well. Um, which means I can run the rabbit hole. Maybe I make a new episode. We'll see. Um, but either way, DM me at the Movement Maestro or text me. 310-737-2345. It will be green. It's my sideline. It is me. All right, so back to this AI bubble, right, that we are in. So what does bubble actually mean? A, uh, bubble, quite simply stated, is when the perceived and projected value of a, of a sector far outweighs the actual value. [00:04:39] All right, very simple. The, the bubble bursts when, you know, similar to the wizard of Oz, the curtain gets pulled back and folks really see what's going on. Right? Uh, we have been here before, folks. Let's, let's enter a little story time. [00:04:54] So let's go back to the mid-1990s. Gosh, I'm an, I'm born in the 80s, right? I'm an 85 baby. So I remember this mid-1990s, right? Internet is new. It's exciting. [00:05:05] Uh, there's investors, venture capitalists, public is thinking like, this could change everything. [00:05:11] Any company that had.com in the name, it would get money. Whether it was real or had like a real model or not, like, is getting money. Stock prices for Internet companies went way, way, way up. Uh, like they doubled and tripled in a day, right? [00:05:26] We move to the late 1990s, early 2000s. Companies are promising the rate. They're raising money by promising eyeballs, right? We're going to get website visitors. They're not promising. We're going to make a lot of money and you should invest in us because of that. They're like, no, we're going to get visitors and like, eventually they'll buy. Like, we're going to get a lot of visitors, right? Um, traditional valuation metrics said, we said, fuck those things, right? Earnings, profit margins, revenue stability that we're looking at website visitors, that's what we're going to get, right? So investors, they're piling in, you know, thinking they're going to make money. The stock prices go up, go up, go up. The nasdaq, right, it's very tech heavy. Uh, stock exchange, it quadrupled actually between 95 and 2000, right? This is just like massively inflated valuations for, for many of these companies. [00:06:15] And then 2000 to 2002, the crash, right? We, we're in the wizard of Oz and we're seeing behind the curtain, early 2000s reality setting in many dot com companies, they have no money left. They just burned through all the cash. Ain't got no money, no profit. [00:06:31] Some of the, the bigger, you know, high profile businesses, they collapsed and that started triggering a panic. Investors start dumping the shares. The Nasdaq fell almost 80% from its March 2000 peak to 2000. Two trillions of dollars of market value. Evaporated, because that shit's fake. Books. I. If you want to wake up and you put yourself in a bad mood, wake up and go learn about the stock market. [00:06:57] This Google, you don't have to Google it. We're on chat GPT. Chat GPT. [00:07:03] What shorting a stock means. [00:07:05] Just go read about that stuff. You've been in a bad mood in two seconds. Uh, but we see trillions of dollars of market value evaporates. Retirement savings, mutual funds, venture portfolios, they are wiped the fuck out. Layoffs, they just Sweeping through Silicon Valley, sweeping through. Startups are folding. [00:07:23] Amazon. [00:07:24] Amazon lost 90% of its stock value before it recovered 90%. Fun fact that we'll kind of circle back to later. It took nine years for Amazon to turn a profit. Nine years, actually, I'm just going to say it now. OpenAI, parent company of ChatGPT, it is not on track for that at all. At all. [00:07:45] So we've been here before, the writing's on the wall, the outcome here, right? Economic damage, people losing their jobs, retirements, investments, they slow down. It did slow down the economy, right? [00:07:57] Investor distrust after that. After the crash, people were more skeptical to invest in, in tech, but obviously, clearly that rebounds, right? But I just want to, I just want to paint the picture of we've been here before all, uh, right. Does this sound familiar? Because it should. We're literally seeing this happen again with AI to slap AI powered by AI on the name of something. And people are like, all right, give it money. You're going to go to the sky, going to the moon, right? Let's give you some numbers though, all right? Because just, just to, to go away from all being all about vibes, right? That's one of the reasons I made this episode is I want it to be actual data, uh, and, and some metrics. Um, but like I said in episode two where I talked about the environmental impact, there is like zero transparency with these companies. So as I, yes, I want to give data, but the data that's, that I'm presenting and that's out there and that you can find is like somewhat speculative and like kind of guessing and, you know, which to me means that the numbers are probably worse than, than we see. But all right, some numbers here we're, we're just going to do for OpenAI and Anthropic, right? OpenAI is the company that owns, um, chat, GPT and Anthropic is Claude's parent company. So OpenAI annual revenue between 12 to 13 billion. So there's like a Bunch of, bunch of sources. But we'll go with 12 to 13 billion. [00:09:19] Billion with a B. How much? They're bringing in losses in 2024 alone. [00:09:26] 5 billion. [00:09:27] Expected losses for 2025 this year, 8 to 9 billion. [00:09:33] Losing. Not profit losing. [00:09:37] OpenAI is projected to lose $115 billion through 2029 before profitability. [00:09:48] Like what? [00:09:51] And yet their valuation is $300 billion. [00:09:58] Right. So this uh, is when I say they're projected to lose that, that's like you know, all together. But from start to 2029, losing $115 billion before profitability. [00:10:08] Yet they are valued at 300 billion. [00:10:13] How, what's, what's going on? Anthropic revenue, 5 billion. [00:10:18] They're doing a little bit better here. Right. So uh, they're making less money. Right? Revenue is 5 billion. A smaller company losses in 2024, 5.6 billion. But that's projected to decrease going down to about 3 billion in 2025. So that's this year valuation. 100, $183 billion. [00:10:36] I don't want you guys to fall asleep with me reading the numbers here. So if we just zoom out because I also want to bring in two of the big, big bad boys, right? We bring in Google and Meta. Right. OpenAI and Anthropic, they are AI specific companies. If we bring in other place that do have AI, but they're not AI specific like Google. Right. Their parent company is Alphabet and Meta, right. We start to see even bigger numbers and we kind of see like who survives this. [00:11:01] Google or Alphabet has a market cap of $2.3 trillion. Just actually, just so you know, I valuation for the other ones because they're private companies. Alphabet, uh, and Meta, they're public. So you say market cap, okay. Uh, but Alphabet is a market cap of $2.3 trillion. Meta has a market cap of $1.3 trillion. The AI sectors of each of these businesses often isn't broken out. But no matter what the, the, the parent company remains profitable because their main product isn't AI. So they, they're actually. Meta is actually negative in the negative for their AI sector. But it doesn't matter because they make so much fucking money. Both of them, Both Google and Meta. How? Via, uh, ads. That is that like I'm not sure if people know that, but like that's how they fucking make money. [00:11:51] Ads. And it's a gazillion dollars, right? $2.3 trillion for Alphabet, that's Google. 1.3 trillion for, for Meta, I think it's Worth noting just because, you know the connection here of Circle back to episode two where I talked about the data centers and everyone's like, the data centers and AI and they like are blaming things like open AI and they're blaming chat GPT and it's like a significant portion of data center usage. And this is of all the data centers we already. Because we already had the data centers, we're making new ones. We already had them. A significant portion of data center Usage is for recommender 3 systems. That's for ads, right? For getting recommended the next video, the next reel in your. On um, on Instagram. It's for getting recommended the ad that's most suited to you that you're likely. Most likely to buy something from. So it's just worth worth noting. Right? But how does this happen? All right, how does it happen? We go back to the two numbers from before. Open AI anthropic, right? These companies are losing billions of dollars every year. OpenAI 8 to 9 billion dollars this year. It's going to use but it has a valuation of 300 billion. [00:12:58] How? [00:12:59] How? [00:13:00] Well, my friends, greed. That is the answer. Greed. It's at the fucking heart of everything, all the problems. It's greed. If you're a finance bro listening to this, what are you doing? I. That's. I don't have that person in my audience, but I guess, hello, uh, if you're a finance bro listening to this, I am sure that, you know, there's a good chance that maybe you disagree that it's not actually how this happens and that's fine. Uh, but it is. I agreed investors buy in. My investors invest the money in hopes to be able to sell later at a higher valuation and make money. They're not doing this like I'm gonna help humans and move us forward. Those are like imma make money. Uh, imma, um, get that bag. That's what you're saying right? [00:13:42] As I'm sitting here and doing the outline for this episode and do my research. I'm like how the do these people like just keep getting money? Like. [00:13:49] And it is great but like what are some of the nuances here? What are some of the ways and uh, some examples here. I'm going to give you like the top three here. One is government contracts, right? The government does invest in some of these things and that also beefs up the juices the, the market there. [00:14:04] Investors, they, some of them have a stake in the profits as well. Like a direct stake where. So the example here is Microsoft has poured in over $13 billion to OpenAI. And they've structured it so that they get 49 of profits until a cap is hit. So it's like, hey, we're going to put money in. Like, when you make a profit, we get a lot of it, right? They. They're directly invested in it. Again, greed though, right? We're making. How much money can I make here? [00:14:29] Uh, the. The investors, right, having a, uh, a stake in the profits here. And they're being told, right, hey, if you invest in this, this technology is going to decrease your. It's going to make more money. How? It's going to decrease your expenses by getting rid of employees. That's whack. And also increase your profit by increasing productivity. [00:14:50] This is not true. It's not. We see it's not happening, right? But this is. They have a stake and this is what they're being told. So they're like, all right, cool, I'm gonna do it. Make more money. [00:14:57] Uh, and then the big thing, honestly, what keeps people investing in this is everyone wants to be a part of the company that achieves AGI, right? That's artificial general intelligence. I have talked about that. Um, if you go to the episode I did, what is AI Talked about that, but. Right. Artificial general intelligence is Skynet My. For those of you that familiar with Terminator, it is that this, like, sentient all knowing can do all the function as. As well as a human does that. That type of, type of AI we are far from that. [00:15:26] Right? And these owners, the owners at Open AI, Sam Altman, right? They talk a good game. [00:15:32] They talk a real good game. It's giving Theranos. Does anyone remember that? That was that blood testing company that was founded by, uh, Elizabeth Holmes. That was fake. And it was like the investors didn't know and they were like, this sounds great. We're gonna make so much money. Great, here's more money. And then she was like, oh, it's fake. All right, so there's a little. Maybe a little difference there because Serenosa is fake and like, AI is real. I guess this exists, but like a AGI, we are not there. There's a bit of a fake there, but I'm gonna show my age here, right, and talk, uh, about jumping the shark because. [00:16:06] And that's how that is. What's showing my age is this reference of jumping the shark. So Open AI recently entered into a 300 billion with a B. $300 billion contract with Oracle. Okay. Oracle is a U.S. tech giant. They're best known for uh, their database software. Um, but today it's mainly a cloud and enterprise computing company. It rents out data center servers and uh, AI infrastructure to power other companies applications. [00:16:35] So OpenAI enters into a $300 billion contract with them. They say we're going to pay you $300 billion, right? OpenAI says to Oracle, in my opinion this is Fonzie jumping that shark. So if you don't know what I'm talking about, there's a show called Happy Days old show in it. We all know the Fonz, right? Fonzie literally jumps over a shark on his motorcycle. [00:17:00] So when we say someone or something has jumped the shark, right? It's, it's, it typically refers to like when a show attempts to use a flashly stunt or some like that to keep people interested in. And when it does that it very much reveals that its best days are likely behind it. Right? So the Oracle spectacle and this just happened, this is why I'm, you know, reporting on this. [00:17:21] OpenAI signs a 300 billion dollar contract with Oracle over the next five years, right? This actually catapulted the, the owner of Oracle, his name is Larry Ellison. [00:17:31] This catapulted him above Elon, that guy in net worth for a little bit, right? That's wild. Not network net worth for a little bit. [00:17:40] So on the one hand we have OpenAI, uh, and you know they're using this contract to be like, look like this is how they get more money. Look, we've secured the infrastructure for GPT 6, 7, 8. You know, for the next five years we're going to get to uh, AGI. This contract right here is going to help us get to AGI. In reality they are bragging about a giant fucking build. [00:18:04] A 300 billion dollar bill over five years. This isn't an unprecedented number. This is jumping the shark. Why? The meta Google cloud deal is 10 billion. 10 billion over six years. Even if we don't even, you know, you don't look outwards, we just say okay, I'm gonna put the blinders on and just look at our inside of our own company. [00:18:27] I said earlier, Open AI they're bringing in like 12 billion a year. Maybe I think it's a lie, but maybe $300 billion is 25 times their current revenue. What? [00:18:40] What? [00:18:41] Come on now we are jumping the shark, right? When the biggest cloud deal in history is worth more than most tech companies make in decades. [00:18:50] That's jumping the shark, right? It's a bubble that is in my opinion jumping the shark. All right, so are open AI's best days behind it. Who's to say? Uh, we are definitely in a bubble. Uh, the day that this podcast, not this episode, but this entire podcast. ChatGPT curious they did this, launched that was what, August 7th. [00:19:10] That was the day that GPT5 was rolled out. And that was largely and widely considered underwhelming as. [00:19:18] Right. I said it in the episodes that you likely wouldn't notice a difference between four and five. And the, the difference that people noticed was they're like, hey, this is mean. I've lost my friend. It was like it was a negative difference that they noticed. [00:19:29] Right. Uh, and so just for your own knowledge, what, what actually happened and why the, the flop of the rollout is pretty significant is that making the mar. The model larger stopped working as a way to make the model better. Because until then that's what they had done. That it kind of defied the, the laws that were out there of just like how big this thing could be and like whether or not it would get better. [00:19:53] And they just, they were like, it's going to pour m more money into it. And from the earlier mod, earlier models that worked, this surprised them and it actually worked. And if, if any of you, you know, use chat GPT in 2018, uh, excuse me, if you use it in 2022, right, when uh, GPT 3.5 rolled out, like, you know, like that was, that was way, way better than, than four. Right? A huge difference there. Uh, and they didn't need to like all they were doing is making the model bigger. But what happened with between 4 and 5 or is that. And we actually had like 4.1 and 5 is that they started to have to lean on these like obscure benchmarks and tests that the average user like didn't notice, didn't notice that these things or just didn't know what the, these benchmarks are on, really understand them or really care about them. Right. I reported, I felt like, I reported uh, on the, the rollout of five and like I watched the announcement and these things and I'm like what are these numbers even? What are they talking about? Like, all right, it seems like I m need to do some cool stuff, but not like, you know, leap years and light years, whatever. Better than the previous model. Right. And um, so when something is better, it's self evident. You don't have to lean on and start going to these obscure metrics that, that you've never heard of before. [00:21:16] To me that speaks to hey, some of our, our better days, they are behind us, right? [00:21:24] But flip side, there are a lot of really smart people in the world and in the tech world and in the AI world. And I do think that continued progress is possible, but the fact that there are all these different people is also what I think will slow the progress that's made towards the race for AGI. Right. Uh, what do I mean by that is that there's lots of different people and lots of different companies and lots of different fields. And I think that it's going to take all of them like working together for us to get to AGI. And I think you just heard the problem. All of us working together, these ain't trying to do that. I think the people actually working on it are, but the companies and they're, you know, each wanting to make the most money and keep their investors and, and turn the profit and be the first to reach AGI. [00:22:15] We saw that OpenAI, right? Go back to episode one and I spoke about OpenAI was called Open because it must be open source. And then they were like, wait a minute, we're closing this off, it's just going to be us, right? Uh, not good. [00:22:29] So, uh, I'm gonna throw a name out here, a little resource here. Um, and the first person that I heard talk about this idea of, of kind of uniting, you know, uniting the field and having this hybrid approach. [00:22:38] And um, this is what he is actually known for as well, um, is Gary Marcus. He's a cognitive scientist, he's an American psychologist, he's an author, he's um, a professor Emmett emeritus. I didn't know what the that meant. I'm gonna look it up. It means like, you don't teach anymore, um, but of uh, uh, psychology and neural science at nyu, right? So the dude's dumps. And uh, I enjoy, I enjoy what he says, but he speaks about hybrid AI approaches and he calls them neuro symbolic systems. And I was like, this makes sense to me, right? A neuro symbolic system is going to combine neural network, which is an LLM like ChatGPT, which does what? Pattern matching. It's math. And it's going to combine that and a symbolic reasoning system which uses rules and logic. [00:23:20] And together these things could lead to AGI. And it's like, okay, this makes sense why? And the best example that he always gives, LLMs can't play chess, right? They just be making up. They don't like, remember, remember the rules. And it's just like this, why the can't you play chess? [00:23:35] Like this should be simple here. And as I was going uh, through the, the episode and doing the outline for the episode of what. What, uh, is AI? [00:23:43] I started going in the rabbit hole of, like, cognition and understanding and logic and reasoning and thinking. Because we use these terms and, you know, AI companies and OpenAI and everybody, they use these terms. And it's like, but it's not thinking, it's not reasoning. Where are you saying that? But it's not actually doing that. And then it's like, well, what does that actually mean to be thinking and reasoning? Uh, and so it's cool to have, you know, be exposed to Gary Marcus and, oh, that's his field of work, right? A cognitive scientist and an American psychologist. And him being like, yeah, we have to combine these things. And I'm like, okay, now we're getting somewhere. This makes sense. All right, but this is a topic for another episode. Uh, and when I know way more about that shit. But, uh, my point here is that we are in a bubble. And the shit we're currently doing and that combined with human greed, may be what stops us from achieving AGI skynet anytime soon. Which probably isn't a bad thing. Like, we are not ready. We need more controls around things. Gary Marcus talks about this. Um, but as related to this episode, yes, we're in a bubble. So what about if and when this bubble bursts? Like, what happens? What happens if and when the AI bubble burst? Who stands to lose? Well, it's the same that happened darina.com right. Companies, they're going to fail. Investors going to lose money, employees going to get laid off, worthless stock options offers taken back. Consumers going to lose out. Products may go away, they may degrade, the prices may go up. I told y', all, I'm telling y'. All. OpenAI is going to raise its prices at some point. It's going to introduce a different tier. It's going to throttle the free tier. It has to do something. It's bleeding money. [00:25:19] Services may start disappearing. The infrastructure providers will lose money. Ordinary people. And this is kind of what, something that I was thinking about, and I was like, oh, I wonder what this looks like. But then we can just look backwards. But your 401k and your IRA, right, we know that Big Tech is, uh, the backbone of the S&P 500 and the NASDAQ. And, you know, if AI valuations deflate, so do those indexes. So that means the 401k balance will dip, right? This is obviously a bigger issue for, uh, realistically, theoretically, it's a bigger issue for younger people. That kind of have more of their 401k and such in, uh, stocks, ideally. Older workers. No, old. As you get older, like me, uh, things go toward, they start to rebalance. They automatically balance towards bonds. But, uh, so that means not catastrophic if you're diversified long term, but painful in the short run. Um, perhaps I just lost you there, but these are things I think about. I'm curious, right? And for some of you who are maybe like, oh, yeah, that'd be interesting. Um, but y', all, I'm no Nostradamus, right? I'm not here to incite chaos. I am just here to be curious. And the numbers, they tell a story, right? They tell a story. And the story. We are absolutely in an AI bubble. When it will burst, I cannot say what will happen. I cannot say. But we can use CHAT GPT to ask what happened during the dot com bubble. And, you know, maybe we'll get a good idea of that outcome. So, last thing, my friends, before we wrap up, how I use ChatGPT this week. So for those of you who don't know each episode I include a section where I briefly discuss how I use ChatGPT that day that week. Um, and this time I use it. I used Agent mode. And, uh, I use Agent mode to see if ChatGPT could pull some rental car prices from me. Um, so remember, an agent is AI that can do more than just answer questions, right? It can actually execute tasks. [00:27:08] So I was looking to find a one way rental from Tempe, Arizona to where I live in Redondo. [00:27:15] Uh, I needed a one way. I'm going to pick up a piece, uh, of equipment. Uh, if you follow me on Instagram, you saw that I was looking for this piece of this like stretching equipment. And uh, my girl Sarah Asadorian had one. And she was like, it's yours. We have one in the closet. It's yours. But I gotta, I gotta go get it. And so I was like, okay, cool. I could fly there. It's like a six hour drive, but I could fly there and just drive back. And I was like, okay, cool, one way rental. But I couldn't find any. [00:27:38] Uh, so I put it in Agent mode. And to put it in Agent mode, uh, there's a little plus icon on the little on the, on the little on the left side of the input field of note. I didn't talk about this feature of Agent mode in last week's episode where I talked about helpful features because I don't think it's that helpful. And, uh, I hadn't used it yet, uh, and it was like medium helpful. Um, but I put in agent mode. I told it the dates that I wanted, I wanted one way and I asked for it to search all the sites. What was cool is like it actually creates like a little mini computer screen inside of the window and you can see what it's doing. It's like literally going to the website and it's like clicking the things and the little arrows moving and everything. It's like oh holy. [00:28:15] Unfortunately it did this for 21 minutes. [00:28:19] I just thought it regular rock. I just wasn't, I was like, I just want to see, right? I'm not going to use heat this week. I'm going to balance out the usage, right? Let it run for 21 minutes. Uh, and during that time it kind of broke I think and I had to start it again and I was like are you, like are you there? Uh, but after the 21 minutes the output basically said uh, you know, reported back what each of the sources had showed, whether there was availability and if there was the prices or if like a site error occurred or if it showed sold, ah, sold out. [00:28:47] Um, and then it like gave me a conclusion at the end and was like yes, there, there are SUVs available but they're very expensive. [00:28:53] Um, and that third party sites are a better way to give you options because um, they actually tend to list more of the one way rentals. So like that was helpful because there was a reason why when I went to the actual individual websites it kept saying sold out, sold out. Um, but you know, overall it's not the coolest use case but um, it does tie into the episode, this bubble, uh, in that one of the biggest talking points for LLM progress, for chatgpt progress, for OpenAI progress for a while was agentic AI uh use having agents. [00:29:28] And that talking point has been relegated to nothing more than a hidden feature in the prompt field. [00:29:35] Um, and the second way this ties into the episode is that I do believe right AA Wow, AI and LLMs, they have the potential to be really cool and really helpful. [00:29:47] But uh, we're still working on it. [00:29:49] So that is all for today. Hopefully you found this episode helpful. I, I actually really enjoyed this episode. I, I went into it being like I'll just like kind of throw it together and like, I don't know, I just want to look at some stuff. And I was like wow, I'm actually very interested in this. So hopefully you are as interested as I am. If not, thanks for listening. Anyway, if you found it helpful? Interesting. Consider leaving a reading or review. Thank you Marcia. That's the most recent review, uh, that was left. Y' all are the best. All of you leave these things. You're the best. Do not forget folks, I also have a companion newsletter that drops every Thursday that is basically the podcast episode in text format, written format. So if you prefer to read or you just want that written record, join the newsletter fam. You can head to chatgpt curious.com forward slash newsletter. Uh, or you can check out the link in the show notes. As always, my friends, endlessly, endlessly, endlessly appreciative for every single one of you. Until we chat again next Thursday, stay curious.