Cohort 14 - Day 2 - 6-25-2025

Jon Cheney demonstrates AI agents that automate tasks like scraping LinkedIn and generating websites, then explores the Model Context Protocol for connecting AI to external tools. He contrasts "surf" (leverage existing tech) versus "swim" (build it yourself) strategies, using Netflix and a senior-care automation business as examples of future-proof models.
Chapters
Record. But yeah, pretty awesome. And we'll welcome guys to day 2. Um, like I said yesterday, same will be true today except maybe even faster. Um, we're going to go over— there's about 10 platforms I want to show you today and play with. And, and, uh, obviously we'll do some more, uh, progress on our, on our little projects we're working on. And, and one thing I want you guys to, to also know is, you know, I'm sure you've worked some on your project, but don't be afraid to to start a new project, right? Or even if you, you know, you started this one and you're like, you know, it's just not exactly as I want it, but now you kind of know what you're doing. It's actually sometimes faster to just start over again and, and just knock out something new that just has exactly kind of what you're looking for. I've done that many, many times and, and I always create a better version the second time. I have this one project that I worked on for a client that I started back in like January of this year. And I worked on it for probably about 35 to 40 hours, like a good, a good amount of time. And something weird just happened, and I could not get the project to fix itself. And I'm like, you've got to be kidding me, like I've done so much work on this. And, um, I was like, you know what, I'm just gonna, I'm just gonna rebuild it really quick. And I knew exactly what I needed, I knew exactly what problems there were going to be, I knew exactly where, where I needed to focus. And I whipped out the whole thing in like 3 hours, right? So once you know what you're doing, you can like really direct things fast. So the goal here is, is to help you understand how to do it once and then, you know, you just spend time, spend time working on it. So anyway, how many people stayed up late working on ideas? David, you know, one cool thing, guys, Replit does have an app. Got an iOS app. I don't know if they have Android. They might. I actually haven't tried. I'm an iOS guy, so, um, but they have an app and you can talk to the agent, you can check your database, you can redeploy, you can make fast changes, right? And so, I mean, you can connect APIs, you can, you can do everything that you can do here on your desktop or, or your laptop, um, you can do, you can do, you know, on the fly. So I recommend downloading that, um, signing in, having access to that. And it's fun even just like, yeah, I have an idea, right? You just like start something, you know, while you're while you're, you know, going to the bathroom or sitting at church waiting for church to start or whatever. You're like, oh yeah, that's kind of a cool idea. I wonder what that would look like and just whip up something really fast. So anyway, pretty cool. There's a couple other things about Replit that I want to show you guys today that I think are pretty cool stuff that I haven't played a lot with, but, but I want you guys to just be aware of them and I want to play with them more. So we'll probably start off in there. But before I, before I dive into any of the course content today. Um, any, any questions or, you know, roadblocks you guys run into, or just thoughts on yesterday where you're just like, man, I couldn't stop thinking about X? No.
So my, my question was, so are we gonna— because I'm, I'm trying to remember the pattern of how we went to set everything up, right? Because you're using a lot of different language. Um, are we going to have access to those so that like because I've got some more ideas and want to work through some of those other ones.
Yeah, totally. So, so number one, you will have, um, you know, access to the recordings of these. I did not get it uploaded yesterday. I really, you know, I finished up and had to go focus on my, on my 8-year-old. We didn't get home till like 9:30. We stayed at the— we went to the Provo Beach Resort and stayed till it closed. So that was a lot of fun.
Nice.
But anyway, um, yeah, I luckily didn't hurt myself on the FlowRider. Which is good.
It's actually amazing how deceiving that thing is when you're on it.
Oh yeah.
You know, when you fall, you think you're like falling into an ocean, but like—
but then you're just like, okay, I'm okay. My kids though, my 8-year-old, they like, they'll tumble. That water is powerful. But, but yeah, anyway, so you'll have recordings. I'll get those. I'll definitely get those done. I'll get both of the recordings up today. And in the follow-up email that comes after today, you'll get not only the recording, but you'll also get a bunch of guides in that Google Drive folder. Folder that I've, that I've shared and showed you guys a little bit, but it has like step-by-step, like how to build a business in 3 days type stuff. And so it goes through like, okay, you're going to start here, you're going to build a strategy, you're going to do this, you're going to copy these prompts, you're going to do all that stuff. So it's very, you know, very detailed and kind of the process we've gone through to get into that. The reality is you can just jump into Replit and just say, hey, this is what I want to build, right? And then give it more details later, or, you know, pull in a, a business plan or something like that later to give it more information. But, um, but it's not that hard. You, you, your basic strategy is start the project, add a database, right? Ask, ask Replit to add the database so that you're now storing data correctly. And then from there, you know, just kind of go nuts. Start to figure out what kind of APIs you might need, what other services, what do you want the user experience to be like, what do you want to happen. Just talk to Replit, just talk to the developer. So, um, Hopefully you get a bit more comfortable with that today, Brandon. Um, Hunter, anything, anything from your end that's on your mind?
Um, you kind of addressed it at the beginning. I was like, as I was kind of tinkering around last night, I just like had a few other ideas come to mind and like this initial idea, I was like, yeah, like it'd be cool, but I think idea X, Y, or Z would be better, a better use of my time, right? So I was kind of like, do I just like scrap this and be like, I'll work on it at some other time and move on? Or do I finish it? But like I said, you, you kind of answered that for me, so.
Yeah, all of the above is the right answer, right? Just whatever you want to work on, work on it, right? And, um, you know, I, I used to play— I, I don't anymore because, you know, I just, I don't know, I'm not interested in it as much anymore. But I used to play World of Warcraft when I was like a teenager and like kind of young college student, whatever. And I would, I would start a character and I'd go to like level 30 or something, and then I'd be like, I want to try a hunter. And then I'd like go and I'd do this and I'd get it like level 28 and then be like, I want to try a Druid, you know. And I would— so I had what they would call alt-itis, right, where I wouldn't ever like take my character all the way through to level 60 and then work on endgame stuff and all the raids. And anyway, I don't even know what the top levels are now. But, um, but anyway, I, I warn you against that just a little bit, right? Obviously go do whatever you want, but it's really, really helpful. And, and obviously you start getting value from it when you finish that project all the way to the end and you're saying, all right, my goal is to get a customer for this product, right? Like I've got this idea or I've got this whatever. Or, you know, maybe David, you're building something internally like, hey, I want, I want to get other people on my team using this product, which means I need to get it good enough so that I'm happy to share this with other people and I feel good. And anyway, so take those projects all the way through to completion, you know, get a couple of them done, get people in them. As fast as possible. Um, but at the same time, you know, dive in. You know, if you've got a cool idea like I'm working on, I'll show you guys just a few of the things I'm working on right now because I think it's maybe just helpful for you guys to see here. Let's minimize some things, make sure Do Not Disturb is on. All right, we're good. So here's my TAD agent, still working on that. I literally haven't touched my computer since we hung up yesterday. So, um, but yeah, the other apps that I'm working, like things that I'm Like, I've got a bunch of kind of like fun stuff that I've just kind of worked on. One of them is Cougar Companion. This is a cool one. I've actually come back to work on this quite a bit. I think it's pretty neat. Um, you know, I'm a big BYU fan, huge, huge, you know, basketball and football fan. I'm, I'm at every single game usually. I've been, you know, a significant sponsor for, for many years. And, um, anyway, so I built this, I built this cool thing, um, called Cougar IQ. I haven't even, I haven't even talked about most of these things, um, in public, but, but I basically, you know, there's no upcoming games yet, whatever, because, you know, we're kind of in the offseason right now. But, but you can go through and you can like talk to Cosmo AI and you can ask like, hey, you know, what's the schedule? What's, what's the first football game this year? Right? Obviously I could Google that, um, but like checking ESPN data, searching the web, all these different things. And yeah, see, this isn't working. I've been tuning this AI. It should be able to talk and answer that question. I have it connected to Perplexity to do those. But one thing that I did work on a lot is this past games section here where it just goes back just 3 seasons here. And you can see, you know, the win rate and the losses. You can look at basketball, but you can go to any of the, like, let's go to the bowl game here, Colorado Buffaloes, you know, view stats., and then I can see, you know, all the different, all the plays, box score. Um, I can see the team stats and all the players, all this stuff. And I pulled all of this from the ESPN API, which is free actually. And as part of the ESPN API, um, they also include this, this, uh, this thing here. And so it automatically created all these highlights and all this stuff for me. And I, you know, I built the UI with Replit and having it do that and kind of worked on it, but all these videos are just here. These plays, whatever. And really quickly, you know, I can obviously go full screen if I want to watch the plays, and I can even download the video if I want, right? So it's just all that just comes from the API for free, like all this data. And so I'm just pulling it. So I built this really, really fun game so I can go back and forth between football and basketball and, and all these different things. So that's a fun, a fun app. What would be your hope for this? Jon Cheney. You know, there's a few things I really wanted to get the AI piece working. The idea was to get the, you know, the Cosmo AI working and then some social feed, but just like create this awesome— my idea was like, like when I'm listening to BYU games, you know, you know, Greg Rabel is like legendary in terms of his ability to just make stuff feel awesome and be awesome. And he pulls all kinds of crazy stats out. Like just, you're just like, how did you even like come up with that stat? Right? And so I wanted to kind of create an equivalent of that, but in an AI form. That did it live, right? So when you're listening to a game, you can like, this is, you can have this pulled up and you're seeing cool stats and it's like, hey, this play just happened. Oh, did you know that this, right? And oh, hey, that, that's, you know, you know, 1 million rushing yards. It was just hit, you know, in the history of BYU or whatever, you know, it just kind of finds cool stats and it just kind of comes up with that. And then like I said, I've been very connected to BYU, so I could take this. I would just probably take it over to the athletic office and sit down with my guys over there and be like, look what I built. How do you think this can be used? And they might say, oh, you know, we really like this piece. Can we throw that into our app? And, you know, I'll license it to them or something like that, right? Or at the very least, you know, I get better at building cool stuff and I've got a little tool for myself. You know, it costs me a total of probably, I don't know, $2 a month to run this thing. You know, same.
And so are there any other APIs you had to connect to or just ESPN?
Yeah. So for this one, yeah, I can just show you my secrets thing here. So here I've got OpenAI API, a Perplexity API, API. And then I have the ESPN API, which they didn't— you don't actually have an API key, and so it's not showed in this— it's not shown in the secret here, but, but it's, it's connected. If you just go to ESPN, it's actually kind of interesting, um, if you just Google ESPN API, it doesn't provide, uh, they don't have one, but people have like pulled it together for you, right? So like, here's the GitHub like public repository, and like, yeah, here's where it is, here's all the things you can get. It's not like an official API, but it's just there. Anyone can just access it and pull data from it. Pretty cool. Again, I didn't have to understand all this. I just found it and I didn't even give it to Replit. I just said, hey, I want you to use the ESPN API and it's like, okay, and it went out and set it up. That's pretty cool. Anyway, that's a fun little project I'm working on. Let's go to a couple others here. Stuff like I'm actually going to, you know, so Outpost X, obviously it's a fun one that I showed you guys. Priority Tracker, Risk Calculator, Global Trek Tracker. That might be one that becomes something.
Is that your kayak thing?
No, I've got it. That's a different one. Trails. Yeah, I think the Global Trek Tracker was like Let's see what it says here. It's not actually called Global Trek Tracker, but let's see what it's— let's see what it shows here. I think it's an interactive world map. I think what it was doing is, is it's like a website that shows everywhere I've traveled in the world. Yeah.
And so I built it to pull up like the things to do, things not to do within that area that you travel.
Yeah, whatever I wanted to do, right? But, you know, I'm able to go in and, and like log places I've been. I actually haven't been to these places. These were dummy— this is dummy data. But what's cool here is I installed and set up, you know, a fully functioning map, right, that, you know, lets me zoom in and, and I can, you know, look at all these places. And anyway, pretty cool. So this uses something called OpenStreetMap. And so that's just an API. I have another one called Echoes app that I'm working on with this very similar idea. And Echoes app, I've actually worked on a decent amount. So, and I've got this nice, beautiful logo that I've worked on. And we'll log in here. This was built with Replit. This was one of the first projects I built with Replit. And this is what really, you know, opened up my eyes to what's possible. And the idea was the same, right? I wanted to, I wanted to work on, I wanted to be able to, you know, record places I've been, visited, whatever, and then be able to, you know, view. Let's see, let's go to, let's see if I've got anything here. Yellowstone Park. Yeah, these are just fake images, right? I'm just— finally it works, right? You know, so I have the ability to add images. I can add a memory, an image, video, audio, journal or whatever, right? I can click to upload and, you know, it does all this cool stuff. I've got a profile that shows all of my, you know, how many countries I've visited, cities I've been. I built an achievement system, my top memories, right? When, you know, someone can go in and they can like memories. I can have a follower system. So it's built as a social network, right? So I can— here's my wife's profile. She hasn't done anything on it. So but I set up a profile for her. And so, you know, anyway, I can message people. Actually, I would be able to— I don't have that one set up yet, but there is a messaging system that will be able to do it. But basically the idea was, hey, I want to see where my friends have traveled. And then if I'm going to be traveling there, I can be like, oh, hey, so-and-so, you know, hey, tell me, tell me what's going on. David, like, I've seen you've been to Italy. I'm going to Italy. Tell me what I should— what should I do? Right. Because it's just good. Your friends, it's better than going to like TripAdvisor and getting random, you know, anonymized or maybe bot-produced, you know, and stuff like that. So anyway, kind of a cool idea.
Like, a thought came to mind, like, for, for reservations, I've always found it difficult because you've either got to go to OpenTable, Resy, uh, what's the other one, Yelp, you know, to like book a reservation. Do you think those have APIs that you could just connect all three and then under one under one system.
Absolutely. Yeah. Yeah. Lots of those things have, have those APIs. Yeah. Let's see here. What's this? Personal Catalyst. Oh, yeah. This is an AI life coach, I think. So this was one. Well, let me, let me just double check. So I remember what this is exactly. Yeah, this was a pretty cool idea. So basically I wanted to be able to track, like, you know, I have like entrepreneurship, family and relationships, music and composition, adventure and outdoors. Those are like categories and things that I'm focusing on. I've got my farm, I've got my health and fitness, writing and publishing, spirituality, church, AI and tech, personal growth. And basically my— the purpose here is to like say, this is what I've worked on this week, this is what I'm doing. And it's like, okay, You know, you haven't done anything on, you know, this project for a while, or you haven't, you know, you haven't gone on a date with your wife.
Where's your spiritual?
Where's the spiritual one? Let's look at that one.
Yeah, let's go look at it. I don't even know if this is, if this is actually working. I only have two masteries, your personal study and church participation, and the grades are— I don't think they're actually calculated. I think they put this here, but I would come in here and the idea is to say, okay, this week I went to the temple and it was awesome. What went really well, what could be improved, whatever. Then AI will supposedly analyze that and then add that element right here. It will add that element into my score and recalculate it and then say, okay, you went to the temple here, you've been focusing on that recently.. And then, and then, yeah, here's, here's the rating updates, right? So I'm currently at a, at a B. Hey, this reflection indicates a positive experience with temple attendance showing engagement, appreciation. However, the reflection lacks depth in terms of personal study insights of specific learnings. Elevating the grade reflects good momentum and potential for deeper exploration and understanding. So now I'm going from B to a B+. Man, you know, anyway, so kind of cool, right? Kind of a cool idea. Now it hasn't updated the rating yet, so it's not finished yet. But it's kind of a fun idea to be able to just keep track of like, okay, here's what I'm doing. Here's what I've worked on this week in this thing. And then AI says, great, here's your, here's your coaching, right? And I'm going to help you do even better. Hey, make sure in your reflections you go deeper and you make sure you cover this. And this is kind of one faceted or I don't know.
So, so APIs, does every app have one and does every website have one?
No, no. And there's a lot. And today, so today we're at some of what we're going to do is we're going to spend some time looking at what do you do if there isn't an API? Right. And what, what would—
like, I, I kind of feel like there's got to be some regulation coming on this. Like, why would ESPN just continue to allow, uh, you know, access to their, their API for free?
You know, they've been doing that for a long time, and APIs, APIs have nothing to do with AI, right? APIs have been around since, you know, 25 years ago, um, and, you know, since computers—
but just to be able to link to somebody else's information that they've gathered for free and so that you could then you know, license it to another place because you've brought it under an umbrella.
Yep.
You know, there has to be like— I feel like there's got to be a change in the industry for something like that, right?
Like, no, I don't think so. I think it's going to— it's just not part of their business model. That's not how they make money. They make money through advertising. That's how ESPN makes their money. Right. And so, you know, they, you know, they get— they get all these broadcasting rights and all these different things. Right. But then once the data is there, it's just data. And so they want to be helpful and they want to, you know, be a leader in providing sports data so they could charge for it, but then other people charge for it. So maybe I want to go to the CBS Sports API and maybe CBS does charge, or maybe CBS doesn't charge, and so they competitively just have to do the same thing that CBS does, right? I don't, I don't know that answer, right? Yeah, probably a good discussion to have with, um, with the, you know, ChatGPT or something like that, just like, hey, why are these APIs free? But But no, I don't think that's going to change anytime soon. I bet 10 years from now ESPN's data is still free. I bet more and more data will become free.
Well, according to whatever— who was it that was speaking on the video yesterday?
Was it the Google? Yeah. Eric. Eric.
Eric.
Yeah.
He's— what is he saying? 5 years, we're done.
Yeah. Well, everything, everything is going to be free.
It's like at that point, middle class is gone and it's either extreme wealth or extreme poverty, right?
Like that's very likely an outcome that happens. I think the middle class will shrink dramatically because of what's happening here. The last thing I'll show, guys, this is a personal website. So right now, if you go to JonCheney.com, you know, it's just, you know, it's just my piano music, right? And I sell lots of piano music every year. You know, I've got, I've got, you know, you know, lots of hymn arrangements and Christmas music, piano books and and all these, all these things. And so that's, that's awesome. But, you know, I made this site using Shopify and I probably published it for the first time in like 2013. And I don't really think I've updated it since then, like, except for like minor little things, publishing new music. I'll just add new products to the page or whatever, right? And so, you know, and I perform and I do all these things, but what I'm doing is I'm, you know, a lot of people these days, they're— they have a really powerful, like, personal brand website. And that personal website then pushes people out to all of the different things that, that you do. Right. And so I'm kind of working on some ideas here, and I want it to be more than just like, here's everything about Jon. I want people to land here and then have stuff to do and learn. And, and it's all about— so the kind of message I'm wanting to push out into the world is all about creativity. I believe that, that creativity is humanity's greatest gift, right? Creation. Is our—
is our— I'm just sorry, I'm just playing a little sample here, Carol of the Bells.
And so, yeah, yeah, have fun listening to that. But that's great. You know, I've got, you know, music and artistry, entrepreneurship, faith and meaning, adventure. And so I want people to be able to come here and, and, you know, they can buy my music and all that stuff that way. And I'll make it very easy for them to find that. But, but anyway, just working on kind of a look at this cool effect, this like bubble that's like moving around this image. It's kind of fun. And so, you know, 4 Pillars of Creation. So I'm working on a personal brand website that's, you know, I still probably need another 20 to 30 hours on this thing before I'm going to feel really excited to launch this. And I'm going to connect it up to a bunch of APIs. So one of the things that I'm going to do here, just so you guys know, just this is— I think it's just helpful to kind of hear me talk about some of this stuff because I've had to figure it out. But Shopify has an API. And one of the cool things about Shopify is it handles all the payments, all the, you know, product management and the images and the SEO, all this crazy stuff that's there, right? You can do so many things through Shopify that they've, you know, spent billions of dollars creating this amazing, this amazing, you know, tool. And so instead of me having to come in and recreate an entire, you know, content management system and store and e-commerce capabilities, I can just API some of that stuff in and then focus more on the creative side over here and add other, you know, additional features, right? And so when I click on shop, for example, now I don't have it set up yet, but when I click on shop, it would, it would just pull up something that, you know, it's built and hosted in Replit, but then it's going to be, you know, pulling, pulling products from the Shopify API, right? So So anyway, that's kind of fun. And then, you know, some of these other things, like, I don't even know if I— right. So, so anyway, that's kind of fun. And then, you know, some of these other things, like, I don't even know if I've worked on all this stuff. I'm trying to work on kind of a dual navigation thing here because they do a lot of stuff with business and entrepreneurship, consulting, resources, success. Yeah, I don't even know. I mean, I haven't even, like, done all this stuff. And, and right now it's only going down to— I mean, I guess those are pulling up pages. That's cool. That's cool. I've actually done a little bit of work here. Anyway, so many things, so many things to work on.
What's up, Hunter? So through that API with Shopify, for example, if someone were to like purchase a piece of sheet music on your Jon Cheney website right here, would it just link them back to Shopify? Or could they like actually purchase it on the website?
Yeah, they would never leave this, right? All of that would happen in the backend, right? It's just kind of behind the scenes. It's connected to Shopify's kind of auto deliverability of everything and all that good stuff, right? Gotcha. And, you know, the tracking, like, for example, I don't want to have to build, like, when you, Shopify has, like, tracking, it has insurance, it has, you know, it's all these things that I don't want to have to go build all of that stuff. And this all ties back into a conversation we're going to have later today about surfing versus swimming. And that's going to be really fun. A couple other things I want to point out here, guys, inside of Replit. When you go to build a new, a new app, right? You've got this Create App button. You can jump in here and you can start talking. You can also use— there's different templates you can use to start with that, whether build the latest in AI or websites, graphics. You can start with some cool templates. I have not actually gone in and done this yet, but there's a lot of really cool things that you can— What's paint with code? US.
What's paint with code on graphics? You have to know how to code, or will it code for you?
I don't think so. Yeah, I have no idea though. I don't, I don't know what that would be and done this yet, but, but there's, you know, a lot of really, really cool things that you can— What's paint with code? US.
What's paint with code on graphics? You have to know how to code?
Or will it code for me? I think so. Yeah, I have no idea though. I don't know what that would be. We can click on it, see what happens. It's like a graphics thing. Oh yeah, so it's like p5.js. Yeah, yeah, so it's just, it's going to be a bunch of like front-end, you know, things like p5.js. That's what you used yesterday, right? Making it really easy to do a bunch of kind of visually, you know, whatever engaging stuff. So templates is pretty cool. You can also import code or design. This is pretty cool. So, and you just jump and click on it, see what happens. It's like, oh yeah. So it's like P5.js. Yeah.
Yeah.
So it's just, it's going to be a bunch of like front end, you know, things like P5.js. That's what you used yesterday. Right. Making it really easy to do a bunch of kind of visually, you know, whatever engaging stuff. So templates is pretty on Google or jump on GitHub and look around and you can find, yes, this is so cool. I love that. That is such just like a. Shot over the bow. So Lovable is a competitor to Replit. So lovable.dev looks very much like Replit, right? Pretty cool. You can also import code or design. This is pretty cool. So, and you just jump— hey, ask Lovable to create a landing page for my whatever, right? So you can have it come in here and you can just like, yeah, build me a weather app, right? And, and it will, you know, we'll listen to you and we'll build that. And we'll do it on Google or jump on GitHub and look around, and you can find— yes, this is so cool. I love that. That is such just like a shot over the bow. So Lovable is a competitor to Replit. So lovable.dev looks very much like Replit, right? It's just, you know, very, very much like, um, Replit. But the thing that it doesn't do well is, um, is all the backend stuff, connecting it to databases and the APIs. Like, you can do it. I, I want to clarify, it can do it, but it doesn't do it as easily as Replit. For somebody like Hunter, David, or Brandon or Jon, right? Where we're not developers, we don't really know what's going on behind the scenes. We're just like, I just want to like push buttons and have you do it for me. You know, that's the kind of experience I'm looking for. And hey, ask Bubble to create a landing page for my whatever, right? So you can have it come in here and you can just like, yeah, build me a weather app, right? And, and it will, you know, we'll listen to you and we'll build that and we'll do it just, you know, very, very much like Replit.. But the thing that it doesn't do well, trying to work with and build. Now the reality is, can I build with Levable? Yeah, I actually probably am smart enough to do it, but, but I don't use it because I like to do this. But it looks like you can just import from Levable directly into here. So that's pretty cool. This is also super cool. Figma designs convert is, is all the backend stuff connecting it to databases and the API. Convert your designs into live apps using Replit Agent. That is big time. Yeah, I think you can do it. I, I wanna clarify it can do it, but it doesn't do it as easy, right? You can have, so, so I have this, this is pretty cool. Let's see here. Let's see if I can pull this up. I wonder if I've got it. Let's see. Gen API Echoes.
I have two. Sorry, what folder was that under on the left-hand side in Replit?
It was under, yeah, it's just easily as Replit for somebody like Hunter, David, or Brandon or Jon, right? It's on the main. Yeah. When you're going, when you're, yeah, sorry. It's great. It's on home. Import code or design. Right, where we're not developers, we don't really know what's going on behind the scenes. We're just like, I just want to like push buttons and have you do it for me. You know, that's the kind of experience I'm looking for. And right. All right. I know I have Echo Zap here. Where is it? Right there. It's okay. Okay. So I've got somewhere here. Figma. Oh, I know I have a Figma design somewhere. Docs login, it goes that wireframe trying to work with and build. Now the reality is, can I build with Lovable? Yeah, I actually probably am smart enough to do it, but, um, but I don't use it because I like to do this. But it looks like you can just import from Lovable directly into here, so that's pretty cool. This is also super cool— Figma designs convert. Maybe it's in, maybe it's in this other Oh, where is it? Where is it? Where is it, man? I swear I've got this. Convert your designs into live apps using Replit Agent. That is big time. Somewhere— why is it not pulling this up? That doesn't make sense to me, but I just want to pull up a, uh, a Figma design here. That's, that's, uh, that's really nice. Now I haven't actually, I haven't actually tried this, but I thought it'd be cool to just like click on a button. I'm pretty dang sure I've already signed in here. I think maybe I just deleted all my, um, cookies or cache or something like that. Yeah, okay. Awesome. I'm not crazy. So this, uh, this Echoes app, right, that I've been working on, I was just showing you guys a minute ago, um, where, you know, this is the kind of the travel log. What is my password? I think it's just password. Yeah, because it's just a test user right now. You guys are welcome to log in and play with it or create your own account and just mess with this. But, um, but anyway, this was kind of the design. And then, you know, when I'm adding a new, like, memory or something like that. You know, this is kind of this. And so I had somebody— I actually, actually had somebody go, um, on Fiverr. I went on Fiverr and I said, hey, this is kind of my idea, this is what I want to do. And I paid him like $200 to whip up some, you know, much nicer designs, right? And so here's, you know, there's your welcome back, there's your— I don't know if I'm a fan of the map itself. My wife says this looks kind of like the Friend And I probably agree with her, so I probably would just say don't follow that. But look how nice this Create New Echo is compared to what I have, right? I mean, this is like way better. You know, this, this, all this information about pyramids and here's the title. This is pretty cool stuff, right? So I can take this and I'm pretty dang sure I can just copy this now and drag it into Replit. Figma design. Yeah, I'd have to log in with Figma and then I just point it at my Yeah, my different thing, right? And then select the frame you want to build in Replit, copy the frame URL and paste it. Boom. So yeah, I mean, being able to go from designs over to, over to this and Bolt. Bolt, that's so cool. I think it's bolt.new. Yeah, is another competitor to Replit. So now it's just saying, yep, you can start wherever you want, but you're going to end up in Replit. Replit's awesome, guys. I really think that Replit is going to be at the top of the game. I also saw yesterday, so GitHub is Razor. Now this is interesting and important to note. If you do import through GitHub, so I saw one yesterday, or I screenshotted the post. Ow, my wrist. My wrist hurts. I did the Flow Rider and I did one little weird tweak. I got surgery on my arm like 6 weeks ago or so, and I probably shouldn't have been doing the Flow Rider. But anyway, so there's a— here it is right here. Okay, so Google released or made their Gemini 2.5 agent available to everybody. And so I can copy this, copy link, should be able to copy that from my phone and paste it right here. Yep, there it is. And we're going to go—
actually, it's Quick Start.
What's Gemini? I don't know. Yeah, so Gemini is Google's AI model, right? Like ChatGPT and OpenAI. You know, ChatGPT is OpenAI's AI model, and then you've got, you know, Anthropic has Claude, right? It's just the names of their models. But Gemini 2.5 is awesome. And so I can import from GitHub I just push this button right here. I actually haven't tried this, but supposedly you just jump in and then I saved my post. By the way, guys, the place where I'm finding all of this stuff and learning, because that's one thing that a lot of people wonder, like, how do you learn all this stuff? It's all on Instagram. It's almost all on Instagram, right? And so I follow a bunch of amazing pages that put stuff out like this. We're like, hey, we found this thing and this is awesome, whatever. And, um, and so yeah, here's all posts. I want to find this. I want to just kind of walk through this really quick with you guys so you can see what's, see what's happening here. Um, right here, right here. Okay, so I do go there, then I would go into the commands, and you set up your Gemini API. I'm not going to do this all the way because you have to have a Gemini API key, and I'm not going to go find that right now. But you basically just follow the instructions on this random Instagram post, you compile all this, and all of a sudden you have your own agent that you can then go in and change. But this is an important thing. When you import a GitHub project, it disables the agent. So the normal agent that you talk to, you can't talk to it. You can talk to the assistant, but it's much slower and much less capable. And so you're probably not going to want to build huge projects unless you really know what you're doing with that. But And I'll show the assistant a little bit more. But anyway, this is kind of cool. I think it's fun. Um, I might play with that a little bit more later. So you can import from a lot of different places to come in here and start. You're not always starting from ground zero. You can start from a template, you can start from some other idea. Um, and I— yeah, I really want to— I really want to go try that, um, and see if I can get an agent up and working, because that's actually what I'm trying to build right now. I'm trying to build this, um this TAD agent thing, right? And can I use some of what that has over here? So anyway, guys, I know a lot of that possibly went over your head, um, but hopefully not. Hopefully you just get the idea that it's okay. Look, there's a lot of ways that I can import and start. And the way that you learn is I, I see something on Instagram or wherever, on X. I see a lot on Instagram and X, not as much on LinkedIn, um, or Facebook or other places where I hang out, but but really Instagram and X, which I don't spend a lot of time on X, but when I go on X, there's lots of good AI content. There's people who'll be like, hey, so-and-so just released this model. Here are 10 ways you can use it. And I'll read through, oh, hey, I wanna, I wanna, I wanna just cache those ideas in my head. So when I'm like, oh yeah, there is that thing that does this thing, you know, then I, I go try it. And whenever I see something that I really love and I'm like, that sounds really cool, I will go try it right then and there. I will find the website and I will go try and figure it out. Okay, so let's, um, let's spend the next little bit here working on some more advanced agents, um, like how to use them for yourselves instead of having to build them. Okay, um, let's close out Echoes here. We don't need that. Um, and we're going to go into a cool platform called Manus. Anybody here heard of Manus? Manus is awesome. Manus is like, Manus is like, um, and you know what, I'm gonna come to Manus second, but I would say Manus is like Replit for anything almost, right? Um, it can, it can automate anything. But I'm actually going to go in here, we're going to start with operator.chatgpt.com. So this is pretty cool. Um, in order to have access to Operator, which is still in early access right now, you have to, um, you have to be on the ChatGPT Pro subscription. Which is $200 a month. And so, you know, do that if you want or don't do it. But, I use, I use several pro versions of the, of the AIs that are out there because, because I don't want the limits and I want to try all their newest stuff because that's my job. My job is to know what the latest and greatest is. I spend probably, I actually have, I actually need to do the math. It's in my, on my like to-do list, but I probably spend somewhere between $800 to $1,200 a month. On AI subscriptions. Okay, so it sounds like a lot, but, you know, when you consider that I teach, you know, roughly 50 students or so a month in this type of a course, right? This is a really abnormal course with just 3 people, but normally I've got 15 people and I do that 3 times a month, right? So I, you know, I do that and then I, you know, work with businesses and I'll charge them $50 grand or $70 grand or whatever it is to go in and do like big consulting jobs with them, $1,000 a month to have all the tools that I need and all of that is easy, easy decision. Anyway, operator allows you to or allows the AI to do things for you. By do things, I mean to click into whatever. Let's just give a quick example here. Go— no, head to my website, jennaipei.org, and take the Mastery Predictor tests. The free predictor tests. Don't ask me anything until you have finished the test. Yeah, sometimes it'll be like, okay, great, which answer do you want me to select? Like, no, I want you to go take the test for me, right? Sometimes it's like, I don't like taking tests. We'll see. Sometimes this works and sometimes it doesn't. I haven't figured out why. But basically it's opening up a browser for me. It's typing that in. It's actually parsing the screen and reading. It's basically taking a screenshot, analyzing what it needs to do, choosing what it needs to do, and then doing that thing. All in the blink of an eye. Here it's looking at my screen and saying, okay, where can I take the Mastery Predictor test? Well, it's probably this. Oh, that's interesting. Why is it saying clicking next question now? Why can I not see it? I should be able to see what it's doing. Retrying button click for progress. View in browser. This is not doing what it normally does. Retrying button click for progress. It should, it should be clicking on this thing right here. And then moving forward instead of actually just sitting here. But it's attempting to activate next question button. Weird. I'm going to take control for a second and see what's happening here. It's like the display is off or something. Yeah, it scrolled down even. All right, I want to go here.
Start, Jon.
Sorry, quick side note. A credit on Manus, would that equate to like one question?
No, I have no idea what it is. Right. Like, I'll ask it to do something and be like, that'll cost you 340 credits. I'm like, I don't know why it does.
Oh, wow.
It's just arbitrary. It's just kind of random. It's— I don't think it's arbitrary. I think it tries to estimate the difficulty of the task and then it just says, here's what you're going to do. Right. Okay. So now I'm going to click finish up and I'm going to tell it started The test, finish it, resume. Yeah, why is it not displaying where it actually, where it actually was? This is not, not working. This is showing live what it's supposed to be doing here, guys. Now I will, um, there it is. Now it's showing that again. That's good. Maybe, maybe it just had that display issue there for a second. Perfect. It's a stupid question. I don't like that question. I need to edit that in my database. But anyway, bottom line is it's now taking the test for me. It's reading it. It's considering it. Yep, 5 cents is correct there. Awesome. Next. And so now it's going through and it's actually doing something for me, right? And so this, you can start to kind of think ahead as to how you can use this. So I asked Manus to work on something like this the other day, and I wasn't able to get it to do it. It still has some limitations and it can't get past bot protection, right? So there's some sites that are like, hey, we don't want bots on our site and it just won't let you in. And so I had the other day, I posted this on LinkedIn. You may have seen me post it, but I, I woke up and I had 66 orders of sheet music on my piano website. And I was like, that— I mean, cool, but I don't think that's right. Like, that, that doesn't seem like— and the thing is, they had all purchased the exact same thing. It was a Christmas piece of music. It was only a $3 piece of sheet music, basically the cheapest thing that was on my site. And, um, and they ran, ran all those orders through, and, and they were— they all had the same address, but they had different email, email addresses and different names. And all of the credit cards worked, right? They went through and they were, they were doing it from the US and from the UK. And I was like, all right, whatever. So by the end of the day, I had like $350 of, you know, revenue from, from this thing. They'd done 105 orders and, and I was like, oh my gosh, this sucks. And so I talked to Shopify and they're like, yeah, you know what, it's probably fraud. You're going to need to go through and cancel each one. If you don't cancel it and then, you know, people report it as fraud and you get chargebacks, It'll be $15 per charge, right? What the heck? Like, this isn't even my fault. I'm not even doing anything. I'm just like, I just have a website up. Someone's coming and doing this to me. Why am I gonna have to pay for that? And they're like, yeah, yeah, it sucks, but you probably should go through and cancel them all unless you want to be at risk of that. I was like, okay, whatever. Then I started doing it. I was like, man, this is super tedious and repetitive. And I'm like, I wonder if I could just get Operator or Manus to do this for me. And unfortunately, um, Manus could not get past their bot verification. Even when I took over, I couldn't get past it. The IP address or whatever it was just would not log into Shopify. I couldn't get it to do it. But you could. That's gonna be a temporary limitation. As AI gets, you know, better at, at, you know, gaining access to stuff like this, you will be able to just say, hey, you know, do something for me. Right? So right now, obviously, it's taking a test, which isn't really helping us. But you can see that it's reading. It's, it's probably getting the right answers. I haven't really been following along because I've been talking, but it's now submitting all these questions. It's almost done with a 12-question test, and I could have been taking a shower and I finished my test at the same time. So obviously you can think about what this means for something like homework. And, you know, I can— another cool thing is this computer session here, I can log in with like my Delta credentials or log in to whatever system I want, right? And it's a secure session and it will retain my, my session control, right? So if I go to log into Delta and then I come back, you know, 3 days later, it will remember like this was my own computer, right? It stores those cookies in those sessions and things like that. So I can, I can say, hey, book me a flight to New York for tomorrow morning, right? You know, just get me off. You know, I need, I need to take off by— I need to land, you know, before 3 p.m., so get me there. Right? And it could just buy that plane ticket for you, right? So as these things get better and better, um, it will, um, you know, open up lots more opportunities for things that it can just autonomously do for you. And so that's part of what I'm trying to build right now in, you know, with my project, my TAD agent, is I want, I want to connect a system like this to something that's fully automated that only checks, you know, right, like right here it's asking for some feedback. Hey, you need a username, email, password, you know, please provide that stuff. Um, you know, I can either log in on the browser or I can just give it that information and I could say, hey, why don't you just go ahead and log in for me, right? So, or register for a new account. So anyway, um, pretty cool operator. Operator's neat. I'm a fan. Um, it is— you cannot API into this yet. There's no way to API into this. So there is that limitation right now. Now, Manus allows you to do very, very complex things, but, you know, as long as that site isn't blocked. Here's— this is where I was. Hey, how to refund over 90 customers on Shopify, right? I just said, hey, you know, I've got all these people, right? This is when I had 90. It ended up being 105. But hey, I want you to do this. It asked me a couple of questions and I need you to just kind of go through and do this. And then it could not login for me. It could not figure it out. But I want to open up one thing here. We'll go to manas.im right here. And one of the cool things that Manas does to kind of give you ideas is it, uh, it allows you to kind of view a replay of different things that people have done with this, right? So, um, but this is your Pro account? No, this is— we're in a totally different system now called Manas. Right.
So, well, yeah, but isn't there—
there's subscriptions to Manus. There are subscriptions to Manus. I am not paying for Pro on Manus. I pay for the normal one and I get tons of credits because they're just trying to— they give you all kinds of fun.
Yeah, because I'm on the free one and it's not showing anything that you're showing.
Yeah, well, what I'm— yeah, so what I'm in right now, if you just go to Manus, I am without signing up for it. This is on the main home page. But yeah, here's an interactive space exploration site, but basically with one, uh, okay, 3D Earth for, uh, for me, cool. I don't want to know the impact of climate change on our Earth. Build 3D Earth with CC Impact. I don't know, let's just see what this is. So it shows you a replay of the request, right, for what other people have done. You can obviously try your own things and try to build cool stuff, but this gives you an idea of what is actually happening. So Yeah, build a 3D earth for me. That's all it says here. And so now it's going to do a bunch of research. It's going to figure it out. It builds out, you know, an entire task list of all the things over here, and then it just gets to work for you. Just a bunch of agents that get to work and start to do things, and it moves from task to task to task going through all of these different things until it's done. And then I can skip to the results and then it's completed it. And then I should be able to click on this and see what it built. Here's the website, 3D Earth Visualization with Climate Change Impacts. So I don't know what it's supposed to be showing. Oh, here it is. It's over on the right. That's right.
Your Earth is greener than it was 50 years ago.
Yeah, Earth rotation speed. I can speed it up. It's kind of weird. So the Earth is shifting fast, but the, but the weather is not. Um, cloud— oh yeah, I guess you would have to— you would want to match the, the clouds to the Earth's speed to make that feel right. So that's kind of cool. You can do that. You can change the cloud opacity. Let's, let's see the Earth a little bit better here. I'm going to do this. Let's stop Earth from spinning and let's stop the clouds from spinning as well. Show atmosphere. I don't know what that does. Sun intensity. Year 2024, 25. I wonder if this actually— yeah, it's— oh wait, hold on. Show climate effects. Yeah, I didn't actually do that. I was gonna— I was gonna be the problem. Let's see if, if this is changing, like ice or something here. Nope, I'm not seeing anything there. Oh, maybe I have to do this: ice cap melting, extreme weather, sea levels rise, temperature change. Yeah, so there's a few, there's a few adjustments, a few little things there. It's not really showing the ice cap melting like, you know, I would want it to be showing here. But anyway, kind of cool. So that's kind of a fun— I've never clicked on this before, so I'm just randomly exploring with you guys here. But again, all that took was one prompt. Colorful Tetris. This is cool. Start game. Yeah, I'm controlling that. Super cool. Well, that didn't work. It just kind of disappeared. Next piece. Next piece. Start game, pause, restart. Anyway, so it doesn't work perfectly, obviously, but being able to, to just like create an animation with p5.js. Other people are doing the same type of thing. Periodic Elements Sound App. View replay. Create an interactive web app that takes the electromagnetic frequency colors produced, converts them into the audible range by dividing the octave until the tones are in the piano range. Closest to middle C so the tones are pleasant. That's crazy. That's a pretty complex ask there, guys. He's asking for all kinds of audio analysis, JavaScript audio context oscillators. This person knew what they were asking for. So let's skip to the results and let's try it. There we go. Looks like you turn the sound on and it just stays on. Yeah, it's not that hard, I guess. This actually makes sense. It just does the audio frequency. 440 Hz. That's an A for those of you that are not musicians. Anyway, so that's cool. No, I love this. This is really neat. Very fun. So anyway, Manus allows you to do all kinds of cool stuff. I'm going to show you a couple of ways I've used it that I think are even, even more powerful than some of what's going on there. I really wish this Shopify thing was going to work because it was going to be a cool post on LinkedIn. I was going to be like, look at what I did, and I saved myself, you know, 2 hours of time. Turned out I didn't save myself 2 hours of time. I added an hour of time trying to get this to work and still had to do the other thing. So that was a fail. But, but anyway, so I received some information about some gold, a gold mine, and somebody's like, hey, could you just like run this through your AI and analyze it and see if this is, you know, accurate information? And so I ran a big analysis on that. I won't share it because it's technically under NDA, but It's, it was pretty cool to go through. They handed me a folder of like 150 documents and I was like, okay, cool, let's throw it in here and just analyze all this, tell me what you think. And then it pulled it all together. So here's, here's a pretty cool one. Extracting HOA data, right? But did you ask me to do that, David?
Yeah, you did that for me.
It was crazy. So I'll show you exactly what happened. This is, this is pretty neat. So this was your— David reached out to me, he's like, hey, you know, I kind of want to take your class and see if I can solve this problem, right? And this was what you— this is what you typed to me in the LinkedIn message. You're like, hey, if you go there, you know, I just want to pull everything down, right? And so then I added, all right, I want to— I want something that can pull all this data into a spreadsheet, eliminate duplicates, provide a long and detailed CSV of all this information. Right? Actually, I think you might have given me all the way through here. Right? And I just added these couple lines. I just copy pasted from LinkedIn, did this, and then said, get to work. And then it took it about— Manus took about 30 minutes. And, you know, so much data. It just kind of went session by session by session by session. And when you get to the end, so it reached its context limit, right? Couldn't continue. But what it can do is it can take everything that it's done so far. It can inherit the compressed context and continue the task. So all I did was push that button, David, and it opened the second one here and it did it again, right? Went through all the way through until it reached the limit and then did it one more time and one more time. Finally it was done and it kicked out this, this CSV document. HOA name, registration date, HOA number, address, city, state, Primary contact email, chair phone, chair email, and then boom, it's done.
Right?
So that's how I accomplished it, David. That's what I did. All I did was copy your prompt into Madison, and then I had to kind of babysit it for maybe like 2 hours, but I didn't sit here and just watch it. I just would come back and be like, oh, inherit and continue, inherit and continue.
And then finally it's like, boom, it's done.
That's crazy. Crazy, crazy. So Madison's awesome, guys, for pulling data, doing repetitive tasks. You know, I took a couple screenshots of things I saw people using, using Mannus for. I saw I used Mannus to give me 100 LinkedIn profiles that had 10+ characteristics we were interested in for our product. And it actually worked really well. I reached out to those 100 people, got 2 calls with major players for one of our portcos out of that approach. So this is obviously a an investor, a VC or something, right? And they, you know, they spent some time finding, you know, using AI to find people that wanted to do that. So research, right? Research is— Manus is good at it, right? It can go out. It can really— it doesn't just find the information like, like, you know, Grok can do. And Grok can't really go look inside of a LinkedIn profile. The cool thing about Manus is it can— you can give it your username and password for LinkedIn and just tell it to log in and get to work, and it will just do stuff for you, right? And that's pretty cool. Um, last thing I'll share here, guys, it— with Manus is this bid preparation. So, um, for anybody that's ever worked in construction, you know that one of the hardest things is preparing bids, right? One of the things that's just really, really time-consuming is opening up, uh, like a, uh, what do you call it, construction plans. I'm going to pull this one up, this paste and text. This is a real example that I used to train this thing. But you look at this like, this is really, really difficult. You have to go in and, you know, normally you have to, you know, read through all of these and do these measurements and like do this process called a takeoff. And you go through all of the different pages where you've got so many measurements and like, okay, this is how much You know, 5-inch slab we're going to need, and this is how much pipe we're going to need here, and these are the trenches we're going to need. And then there's obviously a cost tied to each one of those calculations. Then you put together a nice, beautiful bid, right? And so I basically said, okay, I want you to— so the first, my first, my first task attempt at this was, okay, just do the takeoff. Here's the file. You know, prepare the bid, right? Comes through, gives me all this stuff. I send it over to my neighbor who, you know, I was kind of working with on this, on this idea. And he's like, this is way off. Like, it's not, it's not close. I was like, okay, well let me, let me work on, let me, let me try this again. And so I said, neighbor, send me a completed bid on this project. Like you did one, a correct one. I say, yeah, yeah. So send me the bid. I was like, great. So then I gave to Manus, I said, all right, here is the— I don't even know if it's even here, but let's see here. Let's go up top here. Yeah, I don't even know where it is, but it basically said, here is the— let's see if I can find it. Yeah, I was like, here's the Pace and Tech construction plans. The other documents attached were created through a standard and thorough construction takeoff and was done by humans manually. Right. So I basically said, I want you to— I want you to build some— I want you to try to do this on your own and then compare your results to the actual results that I've given you. And, and then I want you to keep iterating, testing and iterating until you have a working product. I want you to build some software for me. Right? And so that it kept going and working on it, and then it got to a point where it said, okay, I'm way off. Look at this— 90% off, 96% off, 99% off. Like, it wasn't even close, right? And then it kept working on it, right? And it got closer and closer and closer until finally it said, I got it. I now know how this works. I know how to make this happen. And then what it gave me was a series of Python files .py files, which is just coding, it's just code. I said, okay, well, I don't know how to use Python files, I don't know how to put that together, but I do know somebody who does know how to put a bunch of Python files and code together. I dragged all of that over here into Replit. Let's see if I can find this one here. Plan Measure right here. So I said, okay, um, I want you to— let me just go to the top here. I said this solution allows you to automatically extract measurements from construction plans and generate takeoff data similar to the manual process. Here's whatever, and I just copied this over from Manus, right? Here's everything that there is. There's an enhanced measurement tool, there's a user interface, all this stuff. So what I want you to do is I want you to just build this, right? And so it took all of those Python files, all those PYs, right? That all came created from Manus, and it created a user interface for me with all of that. So now I can use all these Python tools. So now I just go through, I create a new project, I upload the PDF. I do have to calibrate the scale manually. Measurement tools, automated takeoff, export results. And so now My, you know, my neighbor, it takes about 20 hours manually to do a takeoff of that, of that kind of quality and scope. It now takes 20 minutes. So that's crazy. So that's a really good example of taking, you know, Manus, right? One, one thing to say, hey, help me work through a problem and figure out a solution and a good way to build this. Great. Now I still don't know how to build this with whatever you give me. Now let's take it over and put it into something where I can build a nice user interface and publish it. And Replit can do that, right? So combining those two tools together becomes an incredibly powerful workflow. All right guys, so Manus operator, Replit, questions? What questions do you have? Is this, is this clicking for you? I know you haven't gone and tried it yet, but how's—
talk to me about it.
That's awesome.
But Brandon, to capitalize on this, you had mentioned if we have ChatGPT Pro, it would transfer over.
What would transfer over?
To get a Manus Pro account?
No, no, no, no. Manus and ChatGPT are completely unrelated.
Okay.
So what you're doing here though with this and transferring it over to Replit, do we need an upgraded Manas account?
Nope. No, I mean, I mean, you need, you need one that's, you know, a paid version to be able to do stuff, right? But like, so my, my, sorry, this is my, let's see here, where's my settings? Manage subscription. Here we go. So I am on this plan right now, $39 a month. Yeah, right. And so that gives me unlimited access to chat mode, advanced mode, and agent mode, 2 concurrent tasks, right? I could upgrade to $200 and then I can do 5 concurrent tasks. I can also do scheduled tasks. Yeah, cool.
That's pretty cool. Could you give another example of when we would cross-reference these two?
I mean, it's just anytime you need to think through a process, right? And have it kind of do a process that requires output, a lot of output. So like ChatGPT, you're probably going to be doing— you do get output, but it's usually just text. It's thought, it's strategy and stuff like that, right? Manus might do a huge research project and then you say, hey, you know what? I need, I need a huge research project and then I need you to, you know, create a presentation for me on this. Right, so ChatGPT can't do that. ChatGPT can't go and build stuff for you, um, whereas Manus is going to be able to build a website for you. Build— yeah, here's another— here's another really basic example of Manus. Um, let's get out of here. And, um, I had it create a biography of me, right? And so I just said, hey, go figure out who I am. This is all the information I gave it. Right. And then I think it probably gave it a little bit more. I also said, hey, Jon also founded this GenAI company here. You can find his music at joncheney.com. Here's his LinkedIn, right? So I just gave it some information and then it went through and found all this information and then created a website with my biography. Right. And it's actually— it's a pretty long biography. I wanted it to be a 2,000-word biography. And, you know, it should be like a good magazine article kind of length. And I actually haven't shared this with anybody, haven't published anywhere or anything like that. But I've read through it. I just kind of wanted to test it, right? And it is accurate everywhere. I did not tell it that I was born and raised in Houston, Texas until the age of 12. I didn't tell it that. But, you know, age of 8, he was already developing his business acumen by selling door-to-door. I didn't tell it that, but I probably did it in some interview sometime that was on a podcast somewhere, right? And it went and found that. This is accurate. Everything is accurate. Hey, moved to, moved to Paraguay, right? His dad called him mission president. Like, it's crazy. Like, it, it, I didn't tell it any of this stuff. It knows, hey, I'm an, I'm fluent English, Spanish, Mandarin, right? International stuff. You know, returning from Taiwan from a mission. It, it knows all this stuff. Then here's my AI expertise and it goes through. I read through the whole thing and did not find, you know how people worry about like, you know, hallucination and all this stuff? Not one false thing here, right? Jon's commitment to quality in his musical pursuits is evident in his recording practices. He records at Counterpoint Studios, a world-class recording facility in Salt Lake City, using a Fazioli grand piano. Yeah, that's true, right? I don't know where it found that. JonCheney.com recording. Ah, that's where it found it. I said it right there. I said the The Fazioli right there recording at Counterpoint in Salt Lake City. So that's where it pulled that from. Amazing. Good job. So it just pulls little things that, you know, YouTube, Instagram, Facebook, TikTok uses the handle @chenipiano. Correct. All correct. And so the way that I learn how to use this, guys, is I just kind of sit back and I'm, I just test it. I just try it and I'm like, okay, I'm super impressed. And it, and then it, you know, through my LinkedIn here, my music website, here's my genaiapi.org. It added these little links. It's— and this isn't a website that I would probably ever publish. I might, I might share this for fun on like LinkedIn and be like, hey, look what LinkedIn— or let's say, look what Manus did for me, right? Um, I don't know if anyone would ever read it, but the, the cool thing here is look what it can do, and then how can you use that? How can you apply that, right? So anyway, pretty neat, pretty neat stuff. Um, any other questions on Manus or, or any of this operator stuff? There are other systems out there, guys, that are working on this. This is kind of the next phase. This is what a lot of AI people are working on right now, is trying to solve for being able to click through and do things for humans, right? Um, there's a lot of privacy concerns though. Um, you know, I'm giving it my passwords. My logins, my access to my Google Drive, you know, like, whatever. I'm just giving it stuff because I want it to do stuff for me. And in order for it to do stuff for me, it needs my information, right? If you give, if you give an AI permission to go, or you ask an agent to go buy you some, um, my logins, my access to my Google Drive, you know, like Whatever. I'm just giving it stuff because I want it to do stuff for me. And in order for it to do some concert tickets and then text your friends about it, okay, and say, hey, I bought tickets for us, this is what we're doing, text them the information. Let's think through what kind of information it needs to know about. Do stuff for me, it needs my information, right? If you give, if you give an AI permission to go, or you ask an agent to go buy you some First of all, it needs to know where we live, right? Or at least the area we live. It doesn't necessarily need to know our address, although actually that's not true. It is going to need to know our address when it uses our credit card. Some concert tickets and then text your friends about it. Okay. And say, hey, I bought tickets for us. This is what we're doing. Text them the information. Let's think through what kind of information it needs to know about. It puts in the billing address, but it's going to need to know our address. It's going to need to know our credit card numbers. First of all. Needs to know where we live, right? Or at least the area we live, doesn't necessarily remember. And expiration date and CVV codes. Gonna need to know our name. It's gonna need to know our address. Although actually that's not true. It is gonna need to know our address when it uses our credit card. It's gonna need to know what kind of music we like. It's gonna need to know how much money we have to spend on this. It's gonna need to know the phone puts in the billing address, but it's gonna need to know our address. It's going to need to know our credit card number. And expiration date and CVV codes. It's going to need to know our name. It's going to need to know what kind of music we like. It's going to need to know how much money we have to spend on this. It's going to need to know the phone numbers of the people we're going to text. It's going to need access to our texting system or our, you know, whether that's— and I don't think we can API into iMessage. So you're probably going to have to use something like WhatsApp or Signal or some other system like that. You're going to have to give it command access to be able to send text messages for you, right? You're going to have to, you know, just numbers of the people we're going to text. It's going to need to kind of lower all these amazing barriers to privacy that have been created over the last 20 years, right?
As people are like, what would stop somebody to go on and say, hey, Jon Cheney, he's my best friend, and offered to buy me some tickets? Just put it on his account.
Well, it wouldn't have access to be able to do something like that. And here's the other thing, at least according to the developers, Manus and all of these conversations and things are private, right? They are encrypted. And so people can't just go in and read them. They can't share them. That data doesn't just go out there somewhere. And so, so no, nobody would be able to do that.
But But so what is the privacy concern? Just that you gave something else your information, personal information.
Exactly. Yeah. We're just kind of— we're trusting that this one program, Manus, or this one company, ChatGPT or OpenAI, has their security figured out well enough that it's not going to get hacked and somehow all of your stuff gets out. Right. Because there's all these different layers of security in every single platform out there. And now we're, we're just saying like, let's lower all those, these barriers and give you, Mr. Assistant, everything, right? Because we want to, because we don't want to have to go do that busy work. We want, we want it to be able to just kind of, you know, get to work for us. So anyway, kind of crazy. Something that will be figured out and worked on over time. And, you know, there will be secure ways to connect assistants and all that type of stuff. People are working on that right now. But it's a real concern, right? We're going to let them inside of our Trello. We're going to let them inside of our financial documents. We're going to let them inside of our whatever so that we don't have to do it right, because we don't want to do that busywork. When we know AI can fill out our taxes for us, we don't want to do it anymore. So we're just going to give it this folder and be like, do my taxes right. Or you can just pay your accountant and keep them, keep them employed for a little while longer. But they're not going to have a job for a whole lot longer. Okay, guys. So we're a little bit ahead in terms of voice. We did voice yesterday, which is pretty cool. We normally do that on day 2, but, but I wanted to help Brandon kind of integrate that and get that set up. Did you play with it more? Did you get it? Did you ever— were you ever able to place it?
I was so busy. Yeah, I never had a chance.
That's all right. That's all right. Yeah, just keep playing with that. I mentioned to you guys that there is a Facebook group where you're going to be able to get support, you know, kind of moving forward. There's lots of cool people there that are going to be willing to, you know, help you for free. And then if you want them to like do something for you, they'll, you know, you could definitely hire somebody out of this group. The group only has 76 members right now, but every single one of these people has been through this course. Okay. And so they know, they know a lot. They're very sharp. So I'll share this group right here. You guys can request to enter that and I will approve those member requests. And this group is growing quickly. Not everybody that has taken my course has joined it because not everybody uses Facebook or whatever. And Brandon, you might, you know, not be on there and that's fine. But yeah, I mean, it's, it's worth it to be on Facebook just for access to stuff like this, right? To be able to get help from other people. But if you have a question like Brandon or Jon, you know, I really, I really need help building something out. You know, I could probably, you know, point you towards somebody, one of these guys that posts a lot and does a lot. This guy Brantley Brumley. Yeah, he's pretty awesome. Pretty cool. Yeah, he posted this this morning. 11.ai for free use. I want to actually just see this. 11.ai. Come on. 11.ai. Just take me to that. 11.ai. That is such a good domain. Gosh, just thinking about my other thing. All right, experimental. Plan your day, research customers with Perplexity, manage linear tickets, and message your Slack team all with just your voice. MCP support. Guys, MCP is a great word I need to teach you, okay? Um, it's something, something protocol. What does MCP stand for? I always forget. I always forget these things. Stand for male chauvinistic pig. No, really? In the AI? Come on, dude. Model Context Protocol. Gosh, my brain's not working. I, I knew that one. But Model Context Protocol, guys, allows AI to connect to almost anything. So for example, there's an MCP that connects to um, what's it called, Unity, right? Or Unreal Engine, right? And so if you're trying to create a video game or something like Unreal and Unity, those are platforms that are used to make like AAA, like Call of Duty, Fortnite, like the best games in the world are built on those types of platforms. And they're very complex. I don't know how to use them. But now I can use AI to tell it what I want. Hey, set up an environment that allows me to set up a first-person shooter. It'll just do it for you, right? The AI will run the program and it connects via an MCP. So an MCP, guys, is an API specifically made for AI to connect to a program. That's what the MCP is. Okay. And so anyway, this is super cool. 11AI experimental. Definitely going to sign in here and see if this see if there's something obvious or something to play with. I think this is worth looking at really quick. Brand new, came out today. So this is Eleven, you know, ElevenLabs, and then just their experimental side of things that gives you a fully autonomous agent that you can, you can talk to with just your voice, which is kind of cool. Sometimes I want to type, like if I'm on a bus, I don't want to be using my voice if I'm, you know, on a plane or in somewhere public, right? I just don't want to have to do that. So hopefully they have the option to do both of those things. Come on, come on, 11.ai, you can do this. You can do this, or you can't. There we go. All right. Ultimate personal assistant MCP support. Okay, that's cool. This is how you refer to your assistant. Sure. Choose the voice for your assistant.
Jonathan. This voice covers— oh, that's amazing.
Shouts. Try it out today.
By repeating what students say.
So essentially, by repeating what's— I'm gonna do, I'm gonna do Alexandra. Uh, add to my voices. I reached my maximum amount. That's okay.
We'll use, um, is it Jonathan from Dungeons and Dragons?
Um, I think so. Yeah. Um, where are my voices? All right, are we cracking codes, solving mystery? Well, Burt Reynolds. I want, I want, I just want to play it. Just want to hear it. For the good old American life. For the money, for the glory. That's crazy. Anybody can use Burt Reynolds' voice. That is so crazy. We'll just say John and we'll just use Expressive John for now. We use Cohort 9, actually. Believe in possibility and tomorrow will hold wonders. Okay, that's perfect. We'll just use that. Create 11. I help you today. Add your first integration. Yep, I understand. So, all right, this is where I like to get to, right? Google Calendar, Hacker News, Linear, Notion, edit my notes and databases, Perplexity. I can search the web. I can connect Slack. Add integrations to your agent. Custom behavior, add integration. All right, yeah, so it looks like those are the ones that I can do right now. I could do a custom MCP server. Um, new custom MCP server. I don't know how to set up a custom MCP server, but guess who does? AI and Replit. And I could probably build one, right? Or I could probably find a list of a bunch of them that are out there, right? There's going to be different server types. There's— and it's basically, guys, you're just going to go look for this information. I don't know how to do this any more than any of you on the screen right now, okay? Really, I don't. I've never done this before. But if I was going to try and figure out how to connect some other thing, I would just go out and be like, okay, top MCP servers. Awesome. Here's a collection of servers for the model context protocol. Let's see what that is. Visual Bright Data, everything, fetch, file system, time, sequential thinking. I could go through and it looks like there's AWS stuff, Super cool documentation, cost analysis, AWS core. That's crazy. Being able to do all this stuff, guys, every single one of these is employees that you would formerly have to pay, right, to manage this kind of stuff, or, or a major function like bucket. Oh my gosh, that's crazy. I have paid full-time, paid $100,000 a year for someone to manage bucket. For my company. Cloudflare, same thing. Boom. Now the agent can do it and I can have a voice. I can talk to it. So boom, in my first search, first link, I was able to find tons of examples of how I might need to use this, right? And you guys might be like, I don't really know how to use some of this stuff, but like Email Inbox Zero, JetBrains, Kaboola. This will continue to grow. This will continue to grow. There's going to be so many different things. There's PayPal, right? You need something to take care of PayPal and all that stuff there. Perplexity, there's that. Playwright. I'm going to be using Playwright. So Playwright, guys, is a— it's like Operator or Manus where it can control the— it can control the computer itself. Super cool. So yeah, I can add that to cursor. I can install manually configurations. Now look, it's got all this crazy stuff here. I don't, I don't know. I don't know what I need here. Right. And so if I was going to try and set this up, I would come back in here and I would say, okay, server type SSE. And I would come back here and I would just search for server. Right. And okay, that's fine. I mean, maybe, maybe I know what this is. Maybe I don't. The reality is I'm probably not going to do that. I'm probably going to go like this. I'm going to copy it. I'm going to go into ChatGPT. I'm going to give a screenshot of this, what it's asking for. I'm going to say, is this information on this, on this, at this website? Can you help me figure out how to, how to set this custom MCP server up? And then I would just listen to it. Okay. And that's how I'd set it up. And now all of a sudden I have the ability to talk to ElevenLabs to get things done with Playwright. Right now, I don't know why I need to do that. Right. Maybe I don't need to do that.. But I'm just saying that's the process I would go through to figure out how to do this. This is, this is venturing into things that I just don't understand. It's way beyond my pay grade. But because I know how to ask my AI about it, I can blast through that barrier and I can figure it out. I absolutely can. And I think all of you guys are more than smart enough to figure out how to do these things. As well. Man, Airtable, Apple Notes. No way. That's crazy. That is crazy. I actually have a huge amount of content in my Apple Notes. What I'm surprised about is the word Apple. Like, I didn't know you could, you could connect into something like that. That's nuts. Calculator. Nice. So anyway, so many things here, and obviously this is just like a huge list. You could probably go through and, you know, that's like awesome MCP servers. But what else would you need to be able to do? Why would you need to be able to connect to this? Really, ultimately, you're going to probably, um, be using MCP servers, guys, to create an agent that does something for you, right? Where you can talk to the agent and then it does all the technical stuff for you. That's why you'd be using MCPs. Any questions on this, guys? I know it might not be perfectly clear yet, but Anyway, okay, just something I want to throw out. I want MCPs to be on your, um, on your radar. Okay, um, all right, let's, let's, um, have a little bit of a discussion here for a second, guys, and then we're going to dive into some, um, image generation, which is, which is a really fun side of, of, uh, AI. And we'll, we'll look at probably 5 different platforms to generate images with. How many of you guys have used ChatGPT to generate an image? Copilot. David? Yep. What about Midjourney? You guys ever used Midjourney? Okay. I have. Have you used anything besides ChatGPT? I guess that would be the better question. Copilot. You've used Copilot. Okay. Awesome. So, okay. Very cool. All right. We're going to have a little bit of discussion here, guys, because it's going to drive hopefully some of your thinking for the rest of the day. And, you know, as you, as you approach just business, this is kind of an AI slash general entrepreneurship thought discussion here. And it'll take probably 15 minutes, maybe 20 minutes to go through. But so we won't do any screen sharing for this, but I think it's really, really important. So we do that. Um, 11.ai, that is such a good domain. How many people here have never swam in the ocean? You guys all experienced that. All right, so you know how powerful it is. How much energy— how many calories do you think you burn, I guess, right, which is a measurement of energy, kcalories, in say 10 minutes of swimming in the ocean? And I'm not saying like you're exercising, you're working out, you're just like hanging out, jumping around, having fun.
10 minutes in the ocean.
50 calories. It's about 100. Yeah, it's a good guess, Brandon. 100 calories, you're swimming in the ocean, hanging out, having a good time. All right, so now how much energy relative to that do you think a tsunami creates? Is it 100 times? 1,000 times? Million times? 10 times? 10 times? What do you think, David?
No clue. 100 times.
Just throwing something out there.
What about you, Hunter? I'd go way more like 1,000, 10,000 times more.
Maybe it's 4 billion is the answer. 4 billion times stronger. A tsunami is so much stronger than a human that there's just no chance in heck you're going to beat it. Okay. Now, what's interesting is when a tsunami is coming in, Surfers love to go out to the ocean, right? If you're, if you're swimming, you don't, you don't want to mess with a tsunami, right? But, and surfers will put on, you know, when they're going out to this, they're putting on helmets and they're putting on life jackets and they're like gearing up for like, because if they get, if they get dropped into the ocean and they get away from their board, like bad things can happen, but good surfers know what they're doing, but it actually creates really, really good surf conditions. Um, Except for, you know, the riptides and crazy stuff that happen when that tsunami is coming in. But if you're on the ground, if you're in the shallow end, if you're in the city that gets hit by the tsunami, all that stuff, like, you're going to get smashed against the rocks and you have a good chance of ending your life. So you probably want to avoid the tsunami. So, um, let's, let's bring this back into AI. So AI is the largest most impactful piece of technology and technological development that's ever happened in the history of planet Earth. There is nothing that has ever— that comes even close. The invention of electricity, the Industrial Revolution, computers, the internet, social media, mobile phones— nothing comes close. All of those things were impactful, right? Um, you know, we talked about this a little bit yesterday about, you know, you brought up my post about the, the farms, you know, farmers and that type of stuff, and what they must have thought if, you know, they knew 200 years from now no one's going to be farming, or one out of, you know, one out of 50 is going to be farming. Um, they'll be like, what is everybody going to do with their time, right? So obviously we already— we've already kind of gone over that, but, um, when a technology is first hitting the market, um, there's some very specific strategies that you can follow to help you win. To help you win in a big way. And so what we want to learn how to do here, guys, is we want to learn how to surf this tsunami instead of trying to swim in it. Right. And so let's talk about what surfing means. The definition of surfing in this analogy here is any time technology improves that is related to your business or around your business, does your business get stronger or does it get weaker? Right. So I'm going to say that again. Any time technology that your business is built around, right, either on top of or next to, does that technology improve? Right. Do you improve with it? Right. So you're surfing on top of it, right? Or are you swimming and it passes you up? Right. So what? So let's, let's get a couple examples. What is, what is Netflix's business model? Like, what kind of business? What's their core IP?
Anybody know?
Well, streaming, right? So they do streaming. Yeah. Did Netflix invent streaming? No. Nope. So that's not their core. That's not their core technology. They invented melon DVDs. They, they did. They did invent that. They did invent that concept. And that's awesome. Obviously don't do that anymore. And you're closer. You're closer, Brandon. Right. Keep, keep digging. Just think about Netflix for a second. Your experience. What is it that they do that's unique to them? And I'm going to give you a little hint. It's not only unique to them anymore because other people do it now. Right? Like Hulu and other people that, you know, other streaming services have copied this same, same model. But what is it that they really call their core and they still really do it better than everybody else?
Is it the fact that they're like an entertainment service that started their like own originals without having like a TV network to run it through?
So that is something original. You're thinking along the right lines of, of, you know, something that Netflix kind of did first. They did do that first. That's still not, that's still not their core technology. And I would argue that that is their— not I would argue, it's a fact that that is their least profitable division, right? Right. I think it's strategic, and they understand that they're choosing to lose a billion dollars a year so that they can make money elsewhere on that. But, but, uh, but no, that's not what I'm looking for. It's not the production studio. So we've said Netflix does streaming, their production studio. What else?
What else is unique? Connect you so that you don't leave.
Like, they want you to binge watch. Okay, so what do they, what do they do to make you, to help you binge watch other than the play next, you know, or the autoplay, right? But what do they do? What's the— no ads? Nope. Series. How many did— does anybody know how much content is on Netflix at any given moment? No. Roughly 30,000 pieces of content. Okay. How much of that content— like, I don't know, maybe you guys are like crazy binge watchers, right? I watch a lot of TV and stuff at night. Like, when I— once it's like 9 o'clock, I'm probably in my bed just watching something until I fall asleep at 10 or 10:30, right? I watch something every night. Sometimes I only last 10 minutes, sometimes I'll last 3 hours. But, um, But anyway, so I've watched a lot of content, but I guarantee you I have not even scrolled past more than 1,000 pieces of content, right? There's no way. So out of 30,000, I've only— I'm only aware of 1,000 pieces of content, right? This tiny little percentage of, of that, of that whole thing, right? So their core technology is actually a content suggestion engine. That's what Netflix really does. Right now they take advantage of the internet. That's a big wave. Media content production, right? No ads. You have the DVD delivery service, right, where you could do all this stuff. All of that kind of comes back to— but all of those things enable their core little, you know, content suggestion engine element to really work, right? But ultimately they are surfing on other people's technology, right? So let's ask a couple of questions to test if they're actually surfing. If the internet gets better, do they— does their service get better? Absolutely. Yeah. Yeah. You stream faster. Now I can stream 8K and 16K and 32K or whatever it goes up to, right? You just keep getting better that way. Okay. If, um, if Akamai that provides the actual streaming capabilities, if they get better and reduce, you know, lag or buffering or whatever, does that improve Netflix's experience? Absolutely it does. Right. Um, and now, now, now let's ask a couple questions about, about Akamai and the internet, right? Which are just two elements that they're, that they're, you know, to technologies that they're around and using. Is Netflix sending people to, you know, the bottom of the Atlantic Ocean to fix cables and make the internet, you know, work? No, definitely not. Are they investing money directly into Akamai, um, you know, to be able to get that to work? No. Do they potentially have an ownership stake in Akamai? Maybe. I don't know. I actually don't know. But, but they're not necessarily investing directly in that technology to make it better. Okay. So that technology improves without them having to do anything and their business just gets better. That is the definition of surfing. Okay. So let's examine another company. What does AT&T do? Or let me ask the question a little bit differently. Is AT&T a surfing or a swimming company?
I don't know. Probably. I mean, I don't know. I've never—
I haven't looked into the growth of AT&T and what they're actually doing. Just based on what you know about them, what does AT&T do? They provide— they provide, you know, I know they've got a lot of stuff. They got TV and internet stuff, but really at their core, they're a phone company. Yeah, I just—
I don't know. Probably. I mean, I don't know. I've never—
I haven't looked into the growth of AT&T and what they're actually doing just based on what you know about them. What does AT&T do? They provide, they provide, you know, I know they've got a lot of stuff, they got TV and internet stuff, but really at their core, they're a phone company.
Yeah. I just don't know if they're growing with the technology that's in place. You know, it's like, yeah, they've always been a cell phone provider.
Yep. Do they build their own towers? I have no idea. Yeah. Do they? And when, when the internet improves, when mobile internet goes from 3G to 4G to 5G to LTE to whatever, right?
All those different moves, if they're growing with the technology that's in place, you know, it's like, yeah, they've always been a cell phone provider.
Yep. Do they build their own towers? I have no idea. Yeah. Do they? And when, when the internet improves, when mobile internet goes from 3G to 4G to 5G to LTE to whatever, right? All those different moves. Are they spending billions of dollars every year investing into that infrastructure to make moves? Are they spending billions of dollars every year investing into that infrastructure to make it work better? Yeah, absolutely they are. Right. Netflix is not. Remember, Netflix is surfing on other people's technology. They're going to let AT&T do those investments. And now I can watch Netflix on my phone. It work better? Yeah, absolutely they are. Right. Netflix is not. Remember, Netflix is surfing on other people's technology. They're going to let AT&T do those investments, and now I can watch Netflix on my phone, right? That's crazy. That's awesome. AT&T, however, is a swimming company. They are not— they're building the technology. As technology builds up around them and improves around them, they have to build to keep up. Right? So think about this. Let's ask this question, right? That's crazy. That's awesome. Can you guys think of a technology out there right now that could put— Inti, however, is a swimming company. They are not— they're building the technology. As technology builds up around them and improves around them, they have to build to keep up. Right. So think about this. Let's ask this question.
AT&T out of business.
What's Elon Musk doing? Yeah, exactly. Skylink or whatever it is. Starlink.
Starlink.
Yeah. I mean, I know with my T-Mobile phone, because we were with Sprint, now it's T-Mobile.
I got a text just yesterday saying that Um, where's— can you guys think of a technology out there right now that could put AT&T out of business?
Is this that? Because it was just yesterday. Um, T-Mobile surprise, you'll soon have early access to T-Satellite. And it goes on and on.
But, um, yeah, I registered for that, um, like early in the spring just to What's Elon Musk doing?
Yeah, exactly. Skylink or whatever it is.
Starlink. Starlink.
Yeah. I mean, I like being on the— and so when every time I go up to Sundance because there's not good reception up there, it's just like you're now on Starlink.
Yeah, it's amazing. Yeah, you can be on satellite for text messaging and things like that. So, so yeah, obviously T-Mobile, AT&T, these guys are thinking about this. They're like, holy crap. Are we— we have an existential crisis here because Elon Musk—
I know with my T-Mobile phone, um, because we were with Sprint, now it's T-Mobile— I got a text just yesterday saying that, um, I know, where's this at because it was just yesterday, um, T-Mobile surprise.
Elon Musk could— and this is the type of thing that he does— he could be like, internet is free for the next month for the entire Earth. After that, it's $10 a month.
Boom. Now he owns everybody. Soon have early access to T-Satellite. And it goes on and on.
But he owns communication on planet Earth because who's going to say no to that? Why are you going to need a cell phone anymore? A cell phone number, AT&T, you pay $10 and you've got it. That is crazy, right? That is a crazy thing that he could do. And probably is aiming to do, honestly. Right. And so you then have all of the telecommunications platforms globally all of a sudden extinct. I can tell you right now, I am not investing in AT&T. That's not a place I would—
not a place I would put money. Yeah, I registered for that like early in the spring just to like be on the— and so every time I go up to Sundance because there's not good reception up there, it's just like you're now on Starlink.
Yeah, you can be on satellite for tech. My money's because they're not surfing, they're swimming, trying to keep up. And as other technology comes up, they don't improve. As Elon improves SpaceX and Starlink and all this stuff, AT&T doesn't naturally improve with it. They have to fight and try to keep up. Okay, let me show you one more example of surfing, text messaging and things like that. So, yeah, obviously T-Mobile, AT&T, these guys are thinking, hey, I'll share my screen here for a second. We'll go to Facebook here and I'm looking about this like, holy crap, are we— we have an existential crisis here because Elon Musk— my mom. Okay, so this is my mom. She is 63 years old and as boomer as they come. She's awesome. Amazing. Best mom in the world. She is terrible with technology. Aflac could— and this is the type of thing that he does— he could be like, internet is free for the next month for the entire Earth. After that, it's $10 a month. Absolutely terrible. Okay. And, you know, 6,300 likes on this video, 689 here. This is just 53,000 likes on that video. She probably posted all this stuff— let's see, that's June 22nd. Yeah, so just 3 days ago. Has— she hasn't. Boom. Now he owns everybody. He even posted today yet, but she posted one there, one there, one there. Man, 53,000 likes. That is crazy, guys. So she's doing about one a day. He owns communication on planet Earth because who's gonna say no to that? Looks like two a day, three a day on that one. And so 2 years ago, she had a few hundred connections on Facebook. And I said, Mom, why don't you just post all your stuff? She's a photographer, she likes to travel, and she loves to get wildlife photography. That's what she does, right? So why are you gonna need a cell phone anymore, a cell phone number, AT&T? You pay $10 and you've got it. That is crazy, right? That is a crazy thing that he could do and probably is. In Africa, she was just there like 2 weeks ago, and, um, I said just start posting. Right? Yeah, look at this one. 58,000 likes is crazy fat hippo. That thing is huge. I mean, to do honestly, right? And so you, you then have all of the telecommunication platforms globally all of a sudden extinct. I can tell you right now it's the pregnant hippo, right? Comments on it are pretty funny. So you guys can go explore, explore social media on my mom's page. Look at this. 650,000 followers now in just 2 years of posting every day. She is the ultimate surfer. I am not investing in AT&T. That's not a place I would— not a place I would put my money. It's because they're not surfing, they're swimming, trying to keep up. And as other technology comes up, because she is not investing in Facebook and they're content algorithms. But as their content algorithms get better, right, she gets better. As the internet improves and, and internet access improves and speed and all of this, all these different cool things that are happening with the internet out there, as all of that— they don't improve. As, as, as Elon improves SpaceX and Starlink improves, you know, Facebook will improve. And therefore, people that want to access and have the ability to link and all this stuff AT&T doesn't naturally improve with it. They have to fight and try to keep up. Okay, let me show you one more example of surfing. She's going to be to a million followers, I bet, by the end of the year because this stuff— okay, I'll share my screen here for a second. We'll go to Facebook here and I'll look— as you know, it's exponential, it grows, it grows crazy fast, especially because she's putting out good content that's getting 50,000. You know, she'll be like, yeah, Jon, I got 17,000 new followers yesterday. I'm like, what? Like, and I do social media, like, I, I try really hard. Look up my mom, okay? And I have, I have like, I don't know, 13,000 followers on Facebook and 20,000 on LinkedIn and 15,000 on Instagram. I have, you know, maybe 50,000 in total across all my platforms. But I'm like working really hard to do this. And she's just like, yeah, I'm just going to randomly post you know, she doesn't even hardly— so this is my mom. She is 63 years old and as boomer as they come. She's awesome, amazing, best mom in the world. She is terrible with technology. Absolutely terrible. Okay. And, you know, 6,300 likes on this. You know, she uses ChatGPT actually to create all of her, all of her hashtags and her little, her little things here. And she just says, yeah, make it simple, whatever. Can I show you how to do that? So she just copies and pastes a ChatGPT caption, uploads a video that's 15 or 20 seconds long, and she makes $15,000 to $20,000. Video 689 here. This is just 53,000 likes on that video. She probably posted all this stuff. Let's see, that's June 22nd. Yeah, so just 3 days ago. Has she— hasn't even posted today yet, but she posted one there, one there, one there. $1,000. A month from this. Okay, that's crazy. Okay, so man, 53,000 likes, that is crazy, guys. So she's doing about one a day. That is ultimate surfing. She is investing nothing, right? And as all of the things that she's built on improve, she improves. Now surfing does have obviously a risk to it, right? What if Facebook gets banned? Okay, well, her income's gone right now. I don't think Facebook's like 2 a day, 3 a day on that one. And so 2 years ago, she had a few hundred connections on Facebook, and I said, Mom, why don't you— it's at risk of doing that right now, but you know, what if the internet goes down? Well, yep, she has just post her stuff. She's a photographer, she likes to travel, and she loves to get wildlife photography. That's what she does, right? So she has no income, right? But we'd all be worried about bigger problems if the internet really went down. This is in Africa. She was just there like 2 weeks ago. And I said, just start posting. What if, what if she gets banned from Facebook because she posts something and people think it's offensive and then they kick her off the platform? Yeah, I mean, all of that is possible, right? Or what if you're surfing on a technology in a company and that company goes out of business, which happens, you know, you're working with us, right? Yeah. Look at this one. 58,000 likes. This crazy fat hippo. That thing is huge. Start up or something. Okay, so let's give another example. Okay, I'm giving a few examples here because I want— I'm going to give one more example here and I want you guys to just kind of really understand. And then it's the pregnant hippo, right? Comments on it are pretty funny. So you guys can go explore, explore social media on my mom's page. Look at this. We're going to apply it. I have a buddy named Jeremy Witt. He is the author of a book called Surf Don't Swim, which is where all this comes from. And 650,000 followers now in just 2 years of posting every day. And he, he owns 7 AI businesses right now. Several of those businesses are doing $1 to $1.5 million a year in profit. Okay. And the way that he's built all of this, she is the ultimate Surfer. This is, uh, this guy, this guy's been around the block. He, he had about a $125 million exit back in, in the dot-com era. And anyway, so he did, he did, because she is not investing in Facebook and their content algorithms, but as their content algorithms get better, right, she gets better. As the internet improves and, and internet access improves and speed and all of this, all these different cool things that are happening with the internet out there. As all of that improves, you know, Facebook will improve, and therefore people that want to access and have the ability to access it— she's going to be to a million followers, I bet, by the end of the year. Because this stuff did really well. He didn't own 100%, he owned about 20% or something. So he, he, you know, got his tens of millions back then and then has turned it into a lot more since then. And he was my chief strategy officer at my company Seek for many years. He's, he's very awesome. And so, um, he was an investor and then we brought him on full-time. And, uh, one of his businesses— so one of the things that he bought with some of his money is he, he purchased a senior, senior care facility in Kansas. You know, it's exponential, it grows, it grows crazy fast, especially because she's putting out good content that's getting 50,000. You know, she'll be like, yeah, Jon, I got 17,000 new followers yesterday. I'm like, what? Like, and I do social media like I, I try really hard, and I have, I have like, I don't know, 13,000 followers on Facebook and 20,000 on LinkedIn and 15,000 on Instagram. I have, you know, maybe 50,000 in total across all my platforms. The city, you know, this guy's smart, he's always thinking about the future, and he's like, you know what, people are always gonna— but I'm like working really hard to do this, and she's just like, yeah, I'm just gonna randomly post, you know. She doesn't even hardly, you know, she uses ChatGPT actually to create all of her, all of her hashtags. And then, you know, this is, this is future-proof, it's recession-proof, it's AI-proof, it's everything-proof. So I'm going to buy, I'm going to buy some stable, you know, some stable income, generate some little, little things here. And she just says, make it simple, whatever. Creating assets like this senior care center. And so we bought that. And, you know, one of the things that, that senior care centers will pay for 'Can I show you how to do that?' So she just copies and pastes a ChatGPT caption, uploads a video that's 15 or 20 seconds long, and she makes $15,000 to $20,000. Is they'll pay for people to come in and talk to their, their people, right? Because as, as you get older, your mind goes away. And if you don't use it, right, if you don't talk, if you don't talk to people, you're going to get Alzheimer's, you're going to get dementia, you're going to get those things. And then, you know, insurance has to deal with it. A month from this. Okay, that's crazy. Okay, so that is ultimate surfing. She is investing nothing, right, with all kinds of medicine and treatment and doctors and all kinds of things to take care of that. And so, and so, and as all of the things that she's built on improve, she improves. Now, insurance like UnitedHealthcare will actually pay for prevention of that, and so they'll pay for people to come into senior care centers and actually just talk to people. And so Jeremy says, well, what's the AI equivalent of that? How could I just take— surfing does have obviously a risk to it, right? What if Facebook gets banned? You know, take the people out of the equation. And Brandon already knows how to build this, right? Built a system with— okay, well, her income's gone right now. I don't think Facebook's at risk of doing that right now, but You know, what if the internet goes down? Well, yep, she has no income, right? But we'd all be worried about bigger problems if the internet really went down. ElevenLabs that calls these people at these senior care centers and just talks to them on the phone, and it remembers their conversations, built a database to manage all of the conversations or the conversations always. And so, you know, now the AI will automatically dial and just, you know, it'll automatically initiate it. What if, what if she gets banned from Facebook because she posts something and people think it's offensive and then they kick her off the platform? Yeah, I mean, all of that is possible, right? Or what if you're surfing on a technology in a company and that company goes out of business, which happens, you know, you're working with this clock on Thursday, whenever the appointment is, and just say, hey Betty, how are you doing? How's, how was Lucy's piano concert? And how was this and how was that, right? And just, it remembers, it will talk to you and remember that conversation. Startup or something. Okay, so let's give, uh, another example and, um, and, and pick it up whenever you want. You can talk to a girl, you can talk to a guy, you can set it up however you want. And, um, and so he's running this entire business where he now is making about— okay, I'm giving a few examples here because I want— I'm going to give one more example here and I want you guys to just kind of really understand it, then we're going to apply it. Um, I have a buddy named Jeremy Witt He is the— with this specific business, he's making close to $100,000 a month in profit with the author of a book called Surf Don't Swim, which is where all this comes from. And he, he owns 7 AI businesses right now. Several of those businesses are doing 1-to-1, 1 part-time person that works for him taking care of billing issues. And that's it. That's the only customer service element. Everything else is automated with AI. The scheduling of the appointments, the actual dialing, the actual conversations, the recording of the conversations, the reports of those conversations, the follow-up, $1.5 million a year in profit. Okay. And the way that he's built all of this is this guy, this guy has been around the block. He had about a $125 million exit back in the dot-com era. And anyway, so he did, he does, everything is all done automated by AI. Completely. And all he's done is built this little API web, this little network, right, of just like, okay, we've got ElevenLabs here and we've got this system here and we've got ChatGPT here and we've got these little systems. And all he has to do, all Jeremy has to do— really well, he didn't own 100%, he owned about 20% or something, so he, he, you know, got his tens of millions— value is maintain those connections, right? And if— here's the cool thing— if TwelveLabs comes out and it's a big competitor to ElevenLabs, and it's like then, and then it's turned it into a lot more since then, and He was my chief strategy officer at my company, Seek, for many years. He's, he's very awesome. And so he was an investor and then we brought him on full time and one of his businesses. So one of the things that he bought, like, hey, we're way better than ElevenLabs and their voices and, you know, use us instead. And Jeremy looks at it and tries it. He's like, actually, TwelveLabs is better than ElevenLabs. All he has to do is go with some of his money as he purchased a senior, senior care facility in Kansas. Go into his code or tell his AI to do it and swap the API key. Done. He's now upgraded his system to the new one. And yeah, I mean, he does have a risk of using somebody else's city. You know, this guy's smart. He's always thinking about the future, and he's like, you know what, people are always going to get old. They're going to be— you know, this is, this is future-proof, it's recession-proof, it's AI-proof, it's everything-proof. So I'm gonna buy, I'm gonna buy some stable, you know, some stable income generating service, but there's always someone else. So he just swaps in the new one. If that one goes out of business, he's creating assets like this senior care center. And so he bought that, pops in the new one, right? And so he's built this amazing, you know, little logic box that's producing $1 million of income for him with just that business. And he has 6 more kind of like this. Okay, so I'm going to share And, you know, one of the things that, that senior care centers will pay for is they'll pay for people to come in and talk to their, their people, right? Because as, as you get older, your mind goes away. And if you don't use it right, if you don't talk, if you don't talk to people, I'm going to apply this now and then we'll be done with this part of the conversation and then we'll jump into some image generation stuff, guys. So I have another company called Velocity Access. Velocity Access does, you're going to get Alzheimer's, you're going to get dementia, you're going to get those things. And then, you know, insurance has to do venue management software focusing on, you know, like Sundance, right? Like a complex resort that has ticketing and it has, you know, with all kinds of medicine and treatment and doctors and all kinds of things to take care of that. And so, and so insurance like UnitedHealthcare will actually pay for prevention of that. And so they'll pay for people to come into senior care centers and actually just talk to people. Front management, because they had a couple restaurants up there, it has— they have a concert venue, they have lodging, they have gift shops, they have rental shops, they've got the lift access. And so Jeremy says, well, what's the AI equivalent of that? How could I just take season passes, day passes, all of these different activities, right? And it's very complex. There's lots of different things to do. And so Velocity Access is software I started building a few years ago focuses on bringing all like 10 pieces of software, right? Toast and Ticketmaster and all these different pieces all into one. You know, take the people out of the equation and Brandon already knows how to build this, right? Built a system with 11 Labs that calls these, these people at these senior care centers and just talks to them on the phone and it remembers their conversations. Built a database to manage all of the conversations or the conversations always. And so, you know, master resort control software, venue control center.










