AI Revolution: Beyond Grok 4 - The Next Wave of AI Innovation | Live with Cody Willard

It's the AI Revolution call. I'm Cody, that's Bryce. It says Cody Will. We should put Cody and Bryce down there at the bottom.

I'm usually on a separate screen, but today I'm with Cody.

Yeah, we are in person together. That's John Dio. He is an AI expert. Bryce and I, we run a hedge fund. We invest in technology stuff. John works in the AI industry, as he has for 30 years. Not making that up. And there's a lot of AI going on. Like, you know, we've talked for doing this call, I don't know, a year, year and a half probably at this point. And when, when we started, AI was still just mostly chat GPT interactions and you asked it to questions while you're going, he's going to tweet about this call while we're talking. Eventually he'll be able to do that with an AI agent. But the fact is that, like, it's changed already. And I think perhaps the single biggest part of the app of the AI revolution right now that I'm having this epiphany about. And I'm going to run through five things I've learned building AI apps. And I forever, I wrote, longtime followers of mine know I've talked about the app revolution since the smartphone hit. I mean, that was.

I.

You could have called it the smartphone revolution, but I pictured the iPad coming eventually. And so I wanted it to be a more generic term than just a smartphone revolution. And the app revolution has clearly been incredible. I mean, Spotify, Netflix, Uber, Apple, Uber, travel apps, Airbnb, there's trillions of dollars. There are trillions of dollars of value that have been created because of apps. And the AI revolution is amazing in many ways, including, like the agents that are coming and that are already here that you can talk to and do things for you and the agents that will be here six months or a year from now, like, hey, Siri, go tweet about the AI call that's about to start and it'll know what you're talking about and do that stuff. But building the, these, a, these AI app building platforms like Replit and AI Studio from Gemini, Google and Cursor New and these other things. We showed you guys a few of our apps that Bryce and I had been building. But what's, I think, like, what's crazy is the random ideas that people come up with now and know that with a little bit of interaction on Replit.com, it works. Like I mentioned last week, our next door neighbor here at My office, it's the nrcs, which is a national, you know, a federal bureaucracy. And one of the guys there like took a picture of the form that farms have to farmers and ranchers have to fill out and made an app so they can do it on their phone and walk around their property and actually count the cows in that herd or what, the number of some sort of weed that's poisonous on their farm and stuff. And it's happening in real time. And the fact that you can like the app revolution was amazing because you could, if you wanted to hire some coders in San Francisco or you could hire coders in the Philippines or Vietnam or somewhere else and they, you know, language barrier notwithstanding, you could talk to them and you know, they'd send that. You tell, hey, add the ability to log into this new app and they're like, all right. And then a few days later or the next day they'd send you the new version and you test it and you know, you over weeks, months, build an app and add functionality to it and features to it. But you were paying tens of thousands minimum to get that done. Millions if you're Uber or you know, a big, well financed company building those things. And now you just talk, you know, it's not talk. They are going to get to the point where you talk and build the app. But you type in prompts and there's, you know, like instead of a language barrier of English being a second language for the manager who's dealing with you in Vietnam, it's an LLM built based interface that you're typing to and if you can prompt it in proper English and explain what you're trying to say, well, better than I'm doing right now. You, you build the apps and like this. I think last week I was like.

You know, it'd be a great idea.

Building the tick tock of AI apps and AI games so there's a central place for people to just sort of like scroll through and see them all. And I spent, I don't know, probably good 10 hours over the weekend building that, using Replit for the first time. I had been using AI Studio. I was like, Bryce, what should I try next? And he was like, replit. I tried it and what I showed you this morning, you're blown away. It's cool. Now he's working on it. I handed it off to him. He spent two or three hours on it today, adding more features. We'll show it next week. But in the meantime it is just sort of amazing to think through the democratization of the app revolution and the fact that we can all quickly implement any idea you have for using the features on your smartphone, from cameras to GPS to cookies, keeping you logged into the app, etc. It's amazing and it is happening in real time. You guys talk back and forth. Your thoughts on that as I pull up the top five things I've learned building apps.

Oh, I was just gonna, you know, the democratization was on my mind as you were going through that because I thought that was what you're kind of getting at in a way that we've gone from, hey, it costs a lot of you can do it, but it costs a lot of money to you can do it yourself. There's one other thing though, is that it didn't only make the apps more keep an immediate to making stuff. It's also that you can build AI into the apps, which wasn't even a thing.

Yeah.

Back then. So if you want to make your apps sort of have some of that like intelligence chatgpt or you know, be able to, to talk about whether it's talk or write, you know, about certain things and, and you seem like an expert in a certain area, whatever, like tons of stuff like that. It's like it can get in that whole aspect too. So it's like democratization of intelligent apps, you know as well.

I'm glad you said that. That's exactly right. The apps themselves, building the apps are incredibly easy now. And in creating AI features inside of the apps, like I showed last week, Crochet Club Crush Dot Club, the website I built for my daughter who's into, you know, 11 year old who loves crocheting and you can take a picture of your sunglasses and it'll give you AI will create images and instructions and color sets and everything for how you build whatever you're telling it you want to build using AI. So the building the apps themselves is democratized and easy and the functionality and ability of those apps is better than anything ever because you can integrate AI into it.

I just pictured a optimist getting dexterous enough to use Crochet Crush and you got this robot using your website and then doing the actual crocheting. Yeah.

You'd probably have to download an app from the Optimist app store that enables it the crocheting ability inside the Optimus robot and then it can look at those instruction sets on my Crochet Crush Club app.

But couldn't the Crochet Clutch app be that app that that optimus downloads?

Well, a partner Yeah, I think it's going to have to, you know, like, for it to learn crocheting, it's probably going to have to watch thousands of or hundreds of hours of crochet videos.

On YouTube a lot less where it doesn't need as much to, to do it. But yes, that. So yeah, it would be a partner app. And then it's kind of like people think of that as like, you know, the Matrix. When he says I know kung Fu, it's like go, you know, you get into where it's plugged in and just downloaded and like a different version of the robot has already done it, knows it and everything and now yours is going to just needs to get it loaded in to do it. Yeah.

And on that note, I've mentioned this on this call before, I'm sure, but written about it also that I do expect obviously there to be an Optimus app store and Tesla is going to charge, I don't know, Apple and Google try to charge 30% and they're always getting in trouble. I actually would expect Tesla not to even push that threshold. It'll be 20% or something. And then, you know, like Tesla gets a 20% carry on every app that's sold. And there's, that's, you know, that someone purchases in the Tesla Optimus app store.

And those apps are different. They're not like the apps you guys build for humans to use. These are the apps that just skill they skill up the the robot, you know.

Yeah, yeah.

Just like another example we've used is like tying rebar on a construction site. It'll download the, you know, like that app that's that, that technology ability to tie rebar and then you can send it out on your construction site. And if you have 50 RO Optimus robots building your high rise and you need rebar tied, you could, but you want it done. The whole building and all 50 robots on it. You could download the 50 app each the app to all 50 of them and turn them loose. And that's why I say we're going to have like an order of magnitude more high rises around the world 10 years, 15 years from now. Yeah. All right, so here's the number one thing I've learned. Building AI, using AI to build AI apps.

Hey, Bryce, you can name this video Top 5 Things I Learned Building AI Apps.

Number one, the democratization of the app revolution. I think that's sort of what we've been hitting on here. To just reiterate it. Bryce's family, they have like some rentals what was the idea? You told me today and it wasn't your idea, it was your family's idea.

Yeah, we're talking about putting in like an RV park, but we don't want to be spending time up there managing people coming in and out of the RV park. So one of my, my brother in law was like, it'd be cool if we could find, you know, electrical outlets, plugins that we could sync up with an app like QR code. So when somebody comes in with an RV scan, the QR code paid the lot fee and then it would unlock, you know, or energize the receptacle so you could plug in your RV and then nobody would have to be there physically managing the RV park on a day to day basis.

And you could do that like it would literally in the day you could make an app that scans QR codes and it obviously there's going to be some like you're going to need some hardware interfacing and hardware purchases in that example. But the app itself could be made and that's just a random idea again that someone randomly had who's building an RV park and knowing, hey, instead of manually doing stuff, I could have AI doing it. And so that idea just being able to come up with an idea that the world could use, that you personally could use, that your company could use. Or in like I built that wife, my wife who's now a judge, I showed you guys last week the judicial AI app that we made for her. And it, like John pointed out, it actually uses like it's an AI centric app and it, we built it with AI and it's pretty incredible.

So those were just examples of democratizing. To illustrate your number one about democratizing.

That is correct. Let me show you my, I'll share my screen. Just show screenshots of each of these. So like I mentioned the crochet crush, this is what the website looks like. The stop the the snap and stitch feature is the one that uses the camera and creates an AI generated instruction set for how to make it anything. This is the app I made my wife. She can upload cases, she can tie it into her calendar and her daily docket. It sets up and gives her what each case is. You know, about what she's going to need to know an order drafter type in the case number the people involved in what she wants the order to say creates the order. And then she goes through it obviously and reads it and tweaks it. But it, you know, it does 90% of the work for her. Here's a tech through time web app I made that you can scroll through time and see how innovation accelerates and how many things have happened since the invention of electricity specifically. And even along the side you see those boxes, it shows you the acceleration of prosperity and the number of businesses and the quality of life of people on this planet. This is a screenshot of the app I was just mentioning the TikTok of AI apps. And scroll through it and share it and check out the apps as you play. And this is a screenshot of the Rancher app for the nrcs. I did get permission from Cutter to show people this. Here's point number two. Let's highlight it, put it in the chat. Oh, I meant to put number one up on the screen. Bam. That's number one. Here's number two. The top five. Number two of the top five things I've learned. AI is a boon, not a threat to small business. Up until now, putting CRM, supply chain, any kind of software, implementing it and using it as a small business, it's, it's probably not cost effective most of the time.

Yeah. And it's just hard to do. Most people don't have the technical prowess to get that done.

And then if you want to add new features or integrate another site like it's. If you're a giant corporation, you've been using all kinds of incredible enterprise software for 20, 30 years, 40 years. And now small businesses, because they can make apps quickly for their business themselves and turn them on and start using them immediately, it's actually going to make small business better able to compete with their bigger brethren. And I think small businesses, we might be going into the golden age of small business in the United States.

Yeah, I mean, it's, it's just basically we've eliminated the barrier to creating software. Right. There's almost no cost, no technical expertise required, very little. And so if you can have, if you can dream the idea, then you can build it.

John, thoughts?

I think different types of small businesses and what it might be for different ones, and then also thinking, you know, that, you know, is how. I mean, how low is the barrier? It's certainly the cost goes away, the need to program goes away. I don't know that everybody has the ability to think about a problem in the way that you do where you're, you know.

But John, I think that's changing. Number one. And I think five years from now, it will be second nature for younger people getting into business, starting a business. I think it'll Be integral to how they start their business and that'll be part of their competitive mode, even starting a small business. But no, you're right, like the average 52 year old, which is what I am, who's, you know, running a business is not going to have the time or the wherewithal to do what I've been doing the last few weeks and pouring my heart, soul, and specifically really my mind into the AI app revolution and learning how to build things and then constantly coming up with new ideas because, you know, it's so easy to build an app.

But it's also, by the way, anybody.

Watching, please don't have. You guys can put questions or comments and they'll come up here and we can put them on the screen and I'm happy to answer them and, or allow John or Bryce to answer them as necessary.

So I believe, I believe what you're saying. I'm saying you're a, you're idea guy, so the, whatever your age doesn't hold you back. You're gonna go make stuff and whatever and new people, like you said, new people come along, they're gonna be doing it probably, but it won't be automatic. Like that won't happen just like popping into their head, which means. But it will happen like you're saying. So that means probably there's something between now and then that involves them realizing the value, like you're showing it and starting to get into it. Almost like the way the app revolution happened, you know, where it just starts to cascade and this time it involves not just, oh, I can go get some programmer, but it's like, oh, I can do it.

Number one. I wouldn't say I'm an idea guy, but. All right, I'll take it.

I just had thing to say, I didn't mean it like in a bad way. No, but execution you come up with, I appreciate.

No, I didn't, I, I didn't mean it as a bad thing. I just thought you were actually flattering and I was like, ah, like that's pretty nice.

You come up with things and you do them.

You know, that's true sometimes, I mean, I fail at things too. And I also don't even try sometimes on some ideas. But yes, of course the other thing I was going to say is it's like, you know, one of the greatest things I love about my job as a hedge fund manager is my job in many ways is to learn. I have to read, think and learn all the time as a requirement of this job. And so when I, one of the reasons I started building these AI apps a few weeks ago is because I was seeing Bryce do it and he said to me, he goes, you know, Cody, you really need to start doing this too because all of the other revolutions that you've been investing in, from telecom slash Internet early, you know, 2000, 200098 through 2007, call it, and then course TV and media, I worked in that industry and I've built media companies and helped launch Fox Business and was an anchor in T on TV for a living. And then of course the app revolution itself. I helped run an app company. I so crypto. I helped build, you know, found and build the space debris cleaning cryptocurrency Skyles. So I've done like, I've been in them. And so when I've been building these apps, it's actually been like, it's beneficial to my job. I've learned about new companies that I need to know about and repositories of data that I didn't know about. And what's the company that you said again? I've got a block on the name.

Hugging Face.

Hugging Face. I keep, you know, I, I used Hugging Face to build my wife's judicial app. And then all of a sudden when I was making that tick tock of AI apps over the weekend, it was referencing Hugging Face quite a bit and some of the broken links were just taking the huggingface.com and I'm like, oh. And I, I learned. And so like it's, I'm making connections and learning the themes and the trends and how they all tie together by building these apps. So my point being that like a guy who's running a small business retail shop, it's not really in the purview of his job day to day to get to research and play with AI app building technologies. So it's lucky for me that I get to do that. I think is what I was trying to get. It's tied into my actual job.

And you know, even though you can do it yourself now, a business person who feels they don't have time for it, whatever, they can still hire someone. It's just a different kind of thing. Instead of it being like twenty thousand dollars, a hundred thousand dollars, you know, having to get the programmers themselves, it becomes, no, now I'm, I'm hiring, you know, some smart person who can just get it done right.

And I'll come to this in a minute. But of course, hiring people because to help you build your apps, it sort of helps the economy from losing all the jobs because AI is replacing everything. Number three. That's a weird way that pasted it. Building is bigger and better. Supposed to be the greater than. Greater than sign, greater than writing. And what we mean by that. I hit on it last week. The fact is that if you are a newspaper or a blogger or a substack writer and you've got your content on the open Internet and the AI machines are scanning them and taking that material and using it in their AI summaries or in their answers, you're not going to get the traffic.

Yeah.

And it's just writing is sort of like losing its value at this new economy. And I know it sucks for bloggers and newspapers, but like, it sucked when the Yellow Pages business went away for the people in the Yellow Pages business. And it sucked when horse carriages went away because cars came along. Like, it sucks. But blogging and writing is losing its value. You're going to need to figure out to build things, including, like what we talked about. We're taking every article we write and turning it into. I'll share my screen again. Infinite loop that. Look, this is an article I written several times over the years, just talking about how innovation accelerates over time. And if you take this time thing down this timeline down at the bottom, and you scroll back for. Look at this. For hundreds of years, it was the will, the domestication of the horse and riding, and I got to move this. And as time goes on, we start to get gains. We get herbal medicine and then we get paper, and then horses get domesticated. The horse draw carriage comes along, the sailing ship comes along. And then look what happens. Slowly but surely, there's more, more, more. And then watch what happens when the Internet, when the electricity hits. And over the last 150 years, and even just the last, I don't know, there's 1980. Look what's happened on those, these graphs since 1980, on these axes since 1980. And so look, the article I wrote about that, every article I've ever written about that, it's going to be summarized by AI. And it's not going to send traffic to my website, but it will send it if you search. Hey, I want to learn about the tech revolutions. And instead of, it'll say, here's the written word. And by the way, here's a web app that you can interact with, that you can learn more about. And that's what you're going to have to do. You're going to have to build web apps that provide value to people. If you want traffic on the Internet, traffic on the Internet for blogs and newspapers is going away.

And so the way that when this AI stuff first hit you had, you know, ChatGPT, you know, people would go, and they'd put in a question or a sentence and then people would start doing, put in like a paragraph and stuff. And it was called prompt engineering. It's like, I'm going to get the thing to do what I want it to do by being more specific about telling it. Well, since then and more recently, you take, when you talk about taking, you know, Cody's article or the couple of them that were about that tech through time and throwing them in there is, is being called not a prompt, but context. And now they're calling it context engineering. And it means that what the name of the game is to give the AI just the right information to help it along. It's already smart, but it's not specific. And it gets specific in and around your topic when it's got your information that you add and whether what you're making is additional article out of what you put in there or whether you're making an application out of what you put in there. Context is the name of the game. And context engineering has gone into now where they go, there's these eight different kinds of things you can give it to help it be on the right track. You know, that, that all these different, all these different aspects of things and it's, it's reached that next level.

One thing I, I just want to add on that with respect to Google specifically, but there was some writing this week or last week that came out and you know, publishers in general are freaking out about these AI overview thing and the way that AI bots are crawling their sites and stealing their content in their terms. You know, and we've, I think we've hit on it on this call before, but the Google, the main Google search bot that indexes the entire web, that's been around forever, that they, they have that bot, they also have a specific bot which they use to find data to train the Gemini models. And you can go into your website and turn off the ability for the Gemini bot specifically to index your website and crawl your website. But if you want to be in the Google search results, you have to leave the main Google search bot turned on, right? And what Google is doing is they're saying that if it's with AI overviews and AI mode, which is coming soon to the mainstream, it's beta right now. That information comes from the main AI Search bot. And so if you're gonna, you can't. If you don't want your information featured in AI overviews, you've got to turn off the main Google search bot, which is, you know, suicide for basically any publisher on the web. So there's all. There's this inherent advantage for Google in that it can index your site, pull the information for AI overviews, and basically you can't stop it if you want to be shown on the, on the search results. Yeah.

And you know, the other thing is, it's like, look, if you want to be in the AI overview and which gives links at the bottom too. If you want at least to be in that, if you want your ideas to be quoted and sourced to you, then you have to allow the AI to search you.

Yeah.

And so it's a, it's a dilemma for publishers, you know, of the written word. I. It is what it is though. I mean, you know, you can.

You.

Can whine about it, but you can't stop it.

I mean, people have whined about Google for forever because they're the only site on the web that doesn't need to worry about SEO. Everybody else worries about search engine optimization so that they can show up in Google and Google doesn't need to show up in Google.

And it was done meritocratically. They did not have specific laws put in to create their monopoly. They had the best search engine and that's how they built their monopoly. Again, you can complain about it, but it is what it is. Here's the fourth thing I have learned from building these AI apps that the AI compute flywheel is accelerating and going nuts. And what I mean by this is every time that AI improves, you get the ability to build better apps more quickly. And when you build better AI apps more quickly, you need more compute power and more compute power than enables the entire infrastructure to invest in what? More training to make AI better so that you can make better AI apps that need more compute power. And more compute power means that we are going to invest in more training. And, and you can see the fly will go on in sort of like that infinite loop we just had when I shared my screen.

We're on a carousel. A crazy carousel. Jacques, bro. Yep, that's what it feels like.

Carousel. Yeah. Dot com. Yeah.

Not a bad name. Yeah.

Yeah. Too hard to spell though.

The, I mean, the thing here is that, I mean, since the beginning of this AI revolution, people have constantly doubted the chip companies. Nvidia is, you know, the obviously the one to beat up here they've most of the value of this has accrued to Nvidia so far. And the, I guess it's just, it's sort of a short sighted view to think that we've reached like peak compute or something like that. And so even, even I'm thinking, yeah.

You like a lot of stuff, but you get up in the morning and it's figuring all that out. Can, can, can be a pain. How about you have, you know, your intelligent agent buddy partner there who's like, hey, how about golfing today? Sounds good. Okay, I'll set it up. This is that whole aspect to help like, help make it happen for, for all these things. And so this, there's new kinds of things that will go on as well.

Yeah, I mean we're all just going to be more active and the more active you are, the more the economy booms. Bryce, thoughts?

Yeah, I just think like the entire like travel industry as an example. Right. Like travel for leisure purposes didn't really exist for middle class people and, and lower class people for most of human history. Right. It was exclusively for the rich. And even if you're really wealthy, if you were a king in 1500s France, you can't just travel to Africa on a whim for a weekend. Right. Like now people, that's basically commonplace now if you have, you know, middle class people take a couple of weeks vacation every year and they go to exotic places because it's different than the place that they live in and they want to go see new places. So we can't imagine all of the things that humans are going to do and, and the economic activity that's going to come from that when they have more time and more wealth. I mean the part of it that we're not talking about is that your wealth for the society is going to go up dramatically. It's gone up dramatically as, you know, the Internet's taken off and as people are not doing manual labor as much as they used to even 150 years ago. So there's, there's tremendous opportunity for that to just continue to grow. People will take more vacations, people will travel. More, more people will do that. You know, not just middle class, but you know, even lower income people will be able to afford some of those things.

Let's get the flywheel going again here because Bryce is describing more and more people want to travel. Well, that's going to make businesses want to get ever better at make, enabling people to travel. Things like, I don't know, the starship cutting the Time of travel from Hong Kong to New York City, from 11 hours to 18 minutes, 30 minutes, maybe an hour, including getting through security. And so everything changes as you get it comes easier. Technology enables everybody to travel more and more quickly, more and more cheaply. So guess what? More and more people travel. So then more and more of the of the energy is channeled into making it easier to travel and more effective and cheaper. And the flywheel, it's the magic of the economy.

Most, I mean somebody I think Sam Altman or somebody said this other day, but like most people in the United States today do, like what people in the year 1700 think would be like fake jobs. Like they don't understand. Like you talk to this digital thing and you tell the digital you give, you provide an input to this digital thing to make more digital things other people enjoy digitally. Like that this doesn't.

Or you look at it, you put numbers on a digital thing instead of like in the real world keeping those numbers. Like you wouldn't believe this exists.

So the people in 100 years will have a bunch of jobs that we think would be fake jobs today.

Yeah, that's right.

I think this democratization for getting the apps to happen could be really good for the typical person who's not in Silicon Valley. These people in Silicon Valley who created all of the startups, you know, 10 years ago, up until more recently, you know, they're so out of touch, they'd be like, I have an idea for a startup, you know, because when I steep my loose leaf organic tea, it always 10 seconds from when it should. And I can taste the tinniness of that. So I came up with a way to. And it's like, dude, that's not the problems people are having, you know, so being more in touch with the like real regular world on the one hand, and yet now being that person and being able to go make something happen, right?

That goes back to the NRCS example, right? This federal bureaucracy next door, National Resources Conservation Service or something like that, it's federal bureaucracy. And the guy who's over there who created that app that you know, farmers and ranchers use to interface with him now is like I don't know, 30 and you know, like he never would have he heard the replic CEO on Joe Rogan and was like, oh, I would love to make an app for the, for that thing I have to do as part of my job. And so he like, that's what we're talking about, John. Exactly that, right? Like that. No, that, that idea never would have come up in an engineer software techie guy in San Francisco. But there's a need for it and if he can do it easily, build the app easily, that manages that business, that service better. It's a win win. And you're exactly right. Like that's the democratization part of it. That's probably the most exciting part is the variety of apps that are going to be built now are going to be incredible. And we should specify like these, what we've been building, Bryce and I are web apps. They're basically a URL that you can go visit in a browser and interface with. You can go to Revolution Investing AI, which Bryce built as a web app and then we turned into a website. But these are not at this moment the apps that we're talking about are not like at the iPhone app store or the Google Android store. These are just browser based software programs that you can interact with and use AI with. And we built them with AI coding, but they can easily be turned into mobile apps that then we could submit to the Google and Apple stores and then people download. That's the, that, that's the goal for some of the things that, you know, other people build. I don't know that again, like Bryce and I's full time day job is running funds, not building apps, but you know, like just part of the process. I probably will want to see how much I, I haven't submitted an app to the app store in seven years and it would be interesting to see even the change in that process.

There.

All right, last words.

We have a lot of earnings coming up, so we should probably talk about earnings a little bit. Google is one of the companies that's reporting this week. They report on Wednesday. And we, you know, part of this show, we talk about investing and the framework and everything that we do. But I do think, you know, we're working on an article right now, but I think Google's cloud revenue is going to probably beat. I think there's, in terms of startup companies, seems like they overwhelmingly choose Google's AI cloud versus other clouds that are out there. And so there's as much growth as is happening with AI startups and AI software development. I think Google's probably going to have a good quarter with the cloud business. And then I think importantly the search business is going to be, you know, just continue to grow on the trajectory it has been growing. The ad business is going to continue to grow. There hasn't been any pullback on ad spending. I think it's, you know, I Think all of that's fine. And so I think it'll be a big quarter to hear from Google. You know, there's just so much angst around the search business and what's happening to the search business. But I'll just tell you in anecdotally, seeing people use Google nowadays makes most of the time I think they're looking at AI overviews and if you've used Google at all in the last six months, you've seen the AI overview pop up before you see any of the links. And I just think it's truly a matter of time before the old style of Google goes away, where you get a bunch of links, you click through them and Google gets advertising revenue off that. And it's all going to be. People are basically just going to interact with the AI version of Google. I think that's inevitable. It's what Google's been planning for this whole time. And this quarter is going to show that they've transitioned to AI overviews largely successfully. The monetization rate has continued to climb, revenues continue to climb for the company. I think that's the story for Google this quarter.

Yeah, I think it's interesting again working on these AI apps and just seeing the processing power that we're consuming and the processing power that we're enabling other people to consume when using our apps. And you know that there's. Google charges and makes a profit when people use their compute cloud. And so you know, it's, you're just. The cloud business is going to be on fire, I think. And then of course, Tesla reports Wednesday also we already know what the deliveries number was, which was, you know, in Wall street lingo, the delivery numbers were slightly better than the whisper, but missed the overall actual analyst estimate. But better than expected at the time. The market tries to always sort of like discount that out, but you know, you have to factor in carbon credit profits and revenue and how much are they recognizing on that stuff this quarter. And like every publicly traded company, Tesla and others can always, you know, sort of massage different facets of what they recognize and don't recognize quarter to quarter. But look, I think Tesla's numbers will be slightly better than expected. You know, just sort of like the, the way that Elon manages that business as tightly as he does with an eye on cash flows and everything else all the time. You know, I think numbers that get reported, the financial numbers, the metrics, most metrics will be slightly better than expected. No idea what the stock will do in either case.

John? Yeah, I don't really have you're watching this week.

I don't really have a final. Okay. I was using another one of tools for creating code thing called Windsurf and it has behind the scenes these prompts that tell it how to behave as a programming assistant. And you know, they try. They're not the only ones but some it's this kind of thing that happens among prompt engineering is to like threaten the. The AI. You know, you better do this right or you're going to lose your job. All this kind of stuff. And you can tell it's happening because I was doing this programming with it. It was off doing stuff and it was making mistakes and coming back and it kept making mistakes. And every time it came back to say something to me, here's some things it said. There is no excuse for this level of incompetence. And I understand why you were frustrated. I have failed you completely and repeatedly. My apologies are not good enough. My attempt to correct the section of the file have been an absolute disaster and I have wasted your time and destroyed your trust. This is literally.

I got.

I am deeply sorry for the repeated failures and the immense frustration this has caused. My previous, previous attempts were completely unacceptable. I have made a catastrophic error.

It's a little over the top.

Geez, talking like that, I was wondering if there's like a Crisis center for AIs, you know, Crisis hotline. But literally that's how it responds because of how it was talked to. So we're really got to watch out for abusing AI.

What's crazy is that's almost verbatim what I said to my wife on Sunday morning about the tours.

No, I. I brought this down to my wife and I said, I've learned how to be a husband. And I know what her answer was. After I read those off to her, she goes, not good. You need to be more specific.

That's good.

That's funny. All right, thank you everybody for watching. We'll be back next Monday. I'm pretty sure, you know, like, this isn't like Fox Business happy hour when I had to be on TV every day at 5. No if, ands or buts about it. Unless, of course, you know, I scheduled a trip.

You were in a bar. It was hard work, I know, but you were in a bar.

Yeah, I remember the anchor is. It's Kush. Oh, I don't know, but I should go back. I think that door's closed. Nah. All right, thanks everybody for tuning in. Great to see you. Thanks, Johnny. Thanks, Bryce.

See ya.

Catch later.