Cohort 21 - Session 1 - 9-30-2025

And we will go ahead and get started. Okay, so welcome everybody to day 1 of cohort 21 of How to Build a Business in a Weekend with AI. Sorry, it's not the weekend, it's only Tuesday, but, you know, we can pretend and we'll still, we'll still learn how to build fast and do amazing things. And I'm excited to get to know each of you. This is actually, in my opinion, like one of the, one of the best group sizes. I've had some where I have 16 or 20 people. And I've had some where it's like 3, you know, or, you know, small groups where I'm training a company. It's just like 2 or 3 people and 6 hours or 7 hours with that small of a group gets really intense, right? So this group is good, right? You can kind of relax and listen to other people. I'm going to be asking you all to share your screens at some points during this course and it'll be really fun. But, you know, just quick intro from me and then we'll do quick intros round robin here. So, I live in West Mountain, Utah. West Mountain, for those of you that don't know, is about 20 minutes south of Provo, Utah. I can see the Y over there on the mountain. So, I'm not too far away from civilization, but I am far enough away that it feels like I live in the country and I have 5 acres and horses and all that good stuff. So, I enjoy it out here. And my background is in technology, in sales, in entrepreneurship. I founded a company in 2016 for which I raised about $13 million and ended up selling that in the beginning of 2023. And since then, really been doing AI stuff. And I founded this company officially on March 1st of this year, but have been kind of unofficially doing AI things for a while, for a long time, for a couple years, and doing it a lot. And I've learned how to do some cool things, and hopefully I can share some of those things with you today. So I will share more about my background and story and business endeavors and things like that as we go throughout the course. I'm just sharing things that I've noticed. So I'll, uh, I'll just kind of stop there for the moment and we'll just go— I want to do round robin intros. We'll start with Tyson and then Laura and then Mike. And anyway, I'm just going down my list. Um, it might seem random to you, but, um, want to hear your name. Obviously we can see it, but we'd love to hear you say it just in case we're pronouncing it wrong. And then where are you from? Where do you live currently? What do you do? Okay, so name, Where do you live? What do you do? And then lastly, whether or not you know how to code, right? And the answer of no is perfectly acceptable. I just want to know, right? Is anybody here like, we already know Laura is a little bit, so want to understand that just a tiny bit. And each of you have about 31 seconds as your time limit. So Tyson, go for it.

Okay.

Uh, Tyson Thacker, uh, live in Alpine, Utah. Uh, co-founder of Boss Retirement Solutions and Advisors. We've got 8 offices here locally, 1 in Boise, 2 in Washington. Ryan is the other co-founder. And we help pre-retirees and retirees to figure out their plan to retire. Was introduced to you through Brad Parker, who's been through your course. I don't have any idea how to code.

Awesome. Well, neither do I, so we're, we're on equal footing there. I used to live about 3 blocks north of Alpine Elementary. Okay, so I used to be, I used to be right in there in the Alpine whatever. I remember what the heck might was Alpine Third. I don't remember. Anyway, just right in that area. So awesome, cool. Well, nice to meet you, Tyson. Laura, go ahead.

Hi, so Laura Giacerta, and I live in Charlevoix, Michigan. So if you— it's kind of like very north of Michigan. And then I work with Make Wellness. I'm an entrepreneur helping the company distribute all-natural peptides to the world. And then I do not know how to code. So I have an engineering software background, but I don't know how to code. It was more of engineering software for, you know, engineering services and software.

Got it. That's very cool. Tell for just a second, because I actually am just curious, give us the 30-second, you know, pitch on Make and the peptides. I actually really want to try them, but I haven't actually bought them yet. And but tell me, tell me why I should.

Well, peptides are the future of wellness. They are naturally in our bodies. It's the language of how our cells communicate. But as we get older, they kind of shut down. And for the last 30 years, we've had synthetic injectable peptides, but these are the first-to-market all-natural oral peptides. So we market them through an affiliate business model. A lot of physicians and gym owners and social influencers, entrepreneurs are helping us market their brand.

Very cool.

They are game changers. So they help to build muscle and energy, lose visceral fat. Just a really exciting brand to be part of. I've been in the direct selling space affiliate for 18 years after, after I left engineering software and had a big global business, but then found Make and been with them since last July.

That's awesome. Were you with Nu Skin before?

No, I was with a skincare dermatology company before. And then, but I've heard of Nu Skin all those years. But, and I mean, and the company's like, I'm sure Truman shared with you, like we had a plan to do $10 million our first year and we're projecting $100 million. So it is like in 18 years as a top global leader, my last company, in just 3 months, my business is much bigger than in 18 years. So it's Definitely. We're, you know, because of the whole wellness and peptides, GLP-1s are the big buzz right now. But people don't want those side effects. And so we're, we're just really excited to be here.

You know what? So I'm diabetic and sorry, guys, for the little aside here, but I'm diabetic. And so I actually am prescribed a GLP-1, right? A very low dosage, right? Not like the crazy, you know, whatever. But But it's a miracle worker for diabetics. It's, yeah, it's amazing. I'm a big, big believer. Oh, someone's crawling behind you. I love that. We saw you. Yeah, they saw you.

It's my husband. He was sitting here. And then, um, the other thing too is we're seeing people's A1C reduce. So you definitely want to check it out.

I know, I need to check it out. Thank you so much, Laura. That's awesome. So Truman Hunt, guys, by the way, is— he used to be my I don't know, he's my bishop and stake president, good neighbor for a long time, like just an amazing guy. But he also was the founder and CEO of Nu Skin, which is a multi-billion dollar company here in Utah. And anyway, so good stuff. Awesome. Thank you.

Did you know that he's our chairman?

Yeah, yeah, yeah, yeah. So, so that's awesome. All right, Ryan, you're next.

Well, great to be with everybody. Tyson and I are brothers. We shared a bedroom growing up. So of course, not anymore, though.

Yeah, of course we have to live by each other.

I can see his rooftop from my deck, back deck. Live in Alpine, Utah. We are serial entrepreneurs. We love different trends, and AI is the one thing that we will not delegate to anyone on our leadership team. And we're just super excited about what it can do. Laura, love that. Why not $100 million? Instead of $10 million. We think it's a game changer and we're just here to learn and, and grow and see what we can do to expand this capacity.

So cool. Tell me really quick and sorry, guys, I know I said 31 seconds to people, but I just, I love to hear and learn. Tell me why you say you won't delegate AI to anyone else on the team. It's just so important that you just feel like we need to know this too, or like expand on that a little bit.

I remember I was at BYU I remember the first time getting on the internet in the business school at BYU, and I just knew that this was going to be a game changer on the internet. Now we can run our entire businesses off of our phone.

Right.

So, and we think this is way bigger. So we believe we want to be alert, curious, and responsive and resourceful. And with such a big change that's happening right now with AI, we just don't want to delegate it. We want to be in the driver's seat on the front row, watching what's happening. And Tyson and I, as brothers and also business partners, we have the ability to fight and conquer. And so it's really fun with this.

Very cool. How big is Voss? How big is your—

We have just under 100 employees. We manage about $4 billion in assets and only for—

What's that? That's awesome.

And we are growing. We're growing exponentially right now. So open up a couple of new markets as well, which we think AI can help us with.

I love it. Um, well, really excited to have both of you, Ryan and Tyson. That's awesome. All right, uh, Mike. Oh, you just muted me. There we go.

Yeah, uh, good morning everyone.

My name is Mike Chung.

I live, uh, in Edmonton, Alberta, the 51st state, and, uh, uh, I'm a retired safety officer and trainer from, uh, General Motors, and I have no idea how to code And I'm doing this because my wife is sick and tired of me hanging around the house. She wants me to do something and get out of the house. So here I am.

Well, here you are. You're going to sit at your computer out of the house, I guess.

Yes, I'm going to go in the garage.

It'll be good. No, I love it. I love it. That's fantastic. Well, welcome from Canada. I love Canada. I actually really love— I'm a whitewater kayaker and there's good rivers all over Canada. There's beautiful ones. Whitewater. There's literally a place, guys, in Canada that's called whitewater region. And it's beautiful and just filled with beautiful, amazing rivers. So the Ottawa River up there. So absolutely. Yeah. Thank you, Mike. All right. Let's go to Jamie. Yeah. Hey, good morning.

Jamie Evans here. I'm located in St. George, Utah at the moment. Also live in Boulder, Colorado. I've been in tech for 25 years working for— from Seagate, running cloud disaster recovery services. My last venture was with Data Direct Networks, running the data center operations for research and development. I've also founded 3 businesses in the hemp and wellness space: Hampton Farms, MicroSol, and Canna Comforts. So I've got a little bit of Python experience and scripting, by no means an expert, but I have, you know, fiddled around with Replit and some other things. So I'm I'm here to kind of hone my skills and learn as much as I can and, um, try to make a difference in the world.

I love it. Well, thank you, Jamie. That's fantastic. Um, great that you have a little bit of experience on in, in actual coding and in Replit, which isn't actual coding, it's vibe coding. And I can do it, um, and I'm not a developer. I don't know how to do it. So, all right, we'll go Ben, then we'll finish with Danny.

Excellent. Everyone excited to be here? I'm Ben Lowell. I am somewhat of a serial entrepreneur as well, but on a very micro scale. My first business I started was a window cleaning business when I was 16, and I picked it back up again when I became a junior high Spanish teacher because I needed to make more money. So, I've been in various companies. The last 10 years, I've been various roles within 4 different tech learning companies, learning tech companies. And currently I do some contract work for a couple of companies, one out of Switzerland and one here locally in Utah. So excited to be here.

And we don't have to hide it. Ben just was hired yesterday as the chief of staff for GenAIPI. So super excited to have him here. And we're growing fast over here. Excited to share with you guys more about what we're doing. But anyway, I'm like, Ben, you got to come take this course so you can learn how to do some of the stuff I talk about all the time.

So yeah, I'm here in Layton, Utah, and I do not have a background in coding. So I've dabbled a little bit, but I have almost no experience. So I'm excited to learn. Yeah, excited to be here with all y'all.

Amazing, Ben. All right, Danny, last but certainly not least.

Hey friends, my name is Danny Warner out of Orem, Utah. I've been— I'm just a sales guy. I've been doing the residential alarms thing for 12 years. I took a break with that, went back to school and became an RN. As soon as I graduated from that, I found myself in solar and I've been doing that for the past 6 years. So Hearing about this opportunity and what, what is this technology, it's just kind of always been something that's been on, as it is for everybody, CEOs across the world, uh, all have this like, how do we leverage this? And I've had that same question. And when the answer to that question was, we'll teach others how to do that, learn how to master the thing yourself, I've, I got, I've gotten really excited about it. So, uh, goal here is to help learn, you know, the system to be able to share that with others and leverage what I could do to help others. So excited to be here and learn. I don't have a ton of experience outside of making— I've made some pretty amazing memes that like have great effect on the social medias. No, but I think I'm capable of more. So for that, I've come.

I love it. Danny is also— has also just joined the GenAI PI team in that sales role that he is so very, very good at. And so excited to have two internal guys today. That I just met in the first, you know, in the last, or, or just brought on, I guess, in the last week. So, um, we are all brand new here. Um, and anyway, so this is, this is awesome, guys. This is a great, very strong group, actually. Um, you know, sometimes I have people that struggle. They're like, how do I minimize the window? And, you know, they don't know how to do stuff with computers so much. And even then, even then, this is what's critical, that the specific lady, she's a stay-at-home mom. She was probably about, I don't know, 65 years old, and just, just kind of your, I don't know, super nice person, had this baking business, and she, she had a nice social presence with about 30,000 or 40,000 followers, but she didn't have a website. She didn't have a way to monetize that in the way that she wanted to. And by the end of this course, she had a fully functioning website ready to launch. And so taking someone who just couldn't even hardly like use the computer well, she knew how to use her phone and post pictures of stuff. You know, she made an AI-powered baking assistant that could go and could like help do all these fun things. And so And then I've had very, very technical developers on here that come through and just like, man, I had no idea this was possible. My whole world has changed. And so, so I hope, you know, these next 8 hours or so, 7 hours and 39 minutes left, will, will help you guys accomplish all of your goals, right? I mean, from, from the Ryan and Tyson, you guys are, are going to do some really, really cool stuff that's going to open up your minds for probably things like customer dashboards and the ability to maybe automate some of your, your communication to people, to Mike up there in Canada. You know, we're going to create some— we're going to create a hobby for you that gets you into the garage and, and makes you some money too and pays for the hobby as well. Down to, down to Jamie, it sounds like you're really experienced down there, and I'm excited, I'm excited to, to have a serial entrepreneur like you on here that, um, you know, obviously is in the tech world. I think you're going to get this stuff really, really fast. And Laura, we're going to help you sell more peptides. Okay. And maybe that's not what you want to do. Maybe you're doing something else. Now, one of the things I asked you to do in preparation for this is to come with an idea. Okay. So let's talk about that for just a second. If you have an idea, amazing, right? And that could be an internal tool, maybe for the, you know, the Thackers, right? Something that's, you know, I'm building something that we're going to use internally, whatever, right? Or maybe you just want to do a, you know, a personal brand website, right? A personal brand website is really, really important. It's really fun. I think, I think everybody in the world should have RyanThacker.com, TysonThacker.com, JamieEvans.com, LauraGiacerta.com. You guys should have your own website. And it's just about you. What do you do? What do you like? What's good? Whatever, right? But it helps with SEO. It helps with getting your message out to the world. You know, even if people do end up on BossRetirementSolutions.com or whatever that is, you know, having other places to get there and figure out who you are. People like to do business with people. And, you know, one person who I really, really respect their opinion. He swears a lot, but his opinion is almost always exactly spot on is Gary Vaynerchuk. He is so brilliant when it comes to marketing. And I've followed a lot. I mean, I've turned GenAIPI into a $10 million business in the last 7 months. Okay. I've done that because I've followed a bunch of these principles, right? Along with a lot of other things. But honestly, this was a huge impact, a huge factor.. And so, we'll talk more about marketing tomorrow a little bit, but a big thing. And one of the things he said in a video just a few days ago, and he's been saying it and I've been hearing it, I've been doing, I've been practicing it for a while is the future of the world is personal brands, right? It's not going to be these big, huge corporate brands like it has been. These personal brands, those, the Logan Paul, the people that are all these big YouTube people, they are going to become the new big, $1 billion enterprises is these people, right? And even if you're not going to be a billion, that's fine. You can still have your personal brand be worth $10 million, be worth $1 million, right? What if, what if your personal brand is worth a million bucks and you're monetizing that right now on top of everything else you're doing, right? So come up with ideas, come up with fun things. If you still can't think of an idea, that's fine. Just send me a little private DM and I'll, and I'll help you out. But, uh, but anyway, okay, so to get started today, guys, um, you know, it sounds like most of us are not developers here, and so we're going to start off with a developing activity. Again, I'm not a developer, and so don't worry that you're not either. We're going to do some really fun stuff. To do this, I'm going to share my screen here and we're going to just dive in. For now, I want you to just watch for just a minute. I'm going to X and going to Super Grok here. You know what? I actually am going to do fast because I don't need it to think that hard. First of all, let me just ask this just to make sure. Has anybody here not used an LLM before, like ChatGPT or Grok or anything? I've had some people be like, oh yeah, I never have. I'm like, okay, well, we need to back up a little bit. But since you all have, okay, we can skip that. Grok is one of my favorite LLMs. It's just really smart. It thinks differently than some other models. Probably Elon's putting some of his autistic brain in there or something. I'm not sure what's going on, but But it does some cool things. The other thing that SuperGrok does really well is it can search the internet. It can search X, right? It can search what people are saying about something, right? So if you want to do market research on something, instead of having to go spend a month doing research, you can just jump on SuperGrok and say, hey, what are people saying about this topic, right? Find me 200 tweets on this and get me some market sentiment analysis, right? What if I do this, right? And you can just kind of bounce through and do some really, really fun things. For now, we're going to use it as a coding tool or partner. We're going to start off here and we're just going to say, using p5.js, create me a mind-blowing visualization and give me the code in a way that I can deploy it. On p5js.org. Okay, I can copy. Okay, so I'm going to copy this prompt. I'm going to drop it in the chat. Just leave it in the chat for a minute. You guys will use it in a second. Okay, so basically I just said, okay, hey, just give me a mind-blowing visualization, right? p5.js is what's called a front-end language, which means it's just display, right? Everything that we see on the screen is front-end. Okay, my hand right here is being rendered because of front-end technology that allows a video to come through, a video feed to come come through and be displayed right here in this video chat. Okay, so writes me a bunch of code. I don't need to know what it means. It's just there. There is this cool little button here that says copy, right? You can just copy that. And then it says, hey, head over to p5.js and paste this. Okay, so it just gives me a link. I'll come over here. I don't want to donate right now, so please— thank you for asking. Okay, we're going to delete all this stuff right here. So I just Command Command A to delete it all, and then Command V to paste it, and I'm going to push play. Okay, boom, I've created something. I have created fire. Okay, now, um, this is, this is, you know, obviously, you know, it's— oh, this is interesting. That's interesting how that kind of went together. I don't know if I'm trying to click. Yeah, I guess it's like coming to my mouse somehow. Every once in a while it just goes to the mouse. It's interesting. Wrap around, add mouse interaction. I don't know what that means, but there is some mouse interactions. That's fun. This, my friends, is a bad prompt. Why? Can anyone tell me why they think maybe it's a bad prompt?

There's no specificity.

Yeah, exactly. Yeah, I'm just— I mean, what if I, what if I hired a developer, right? I, I go and I pay $120,000, I hired a developer for my company, and then I just say, hey, go build me a mind-blowing visualization. They'll be like, you know, and then that's it. That's all the instructions you get. I'm going back in my office, you go in your office, come back to me when you're done, right? That developer is going to say, okay, well, I mean, I guess I'll just try something, right? And they'll go and they'll build this cool thing, and then they come back to me and they show it to me and say, here you go, Jon. I'll say Well, what the heck? This isn't anything like what I wanted. Well, you didn't tell me what you wanted, right? So you got to, you got to be, you got to treat your AIs like humans, right? In terms of you got to help them understand what you mean. Okay, so let's try now maybe a little bit better, a better one. Okay, so let's say now using— okay, actually just go say now build me a 2D animated model of the solar system. All right, cool. So what's interesting is I didn't say this time p5.js. Okay, it already knows. It understands from the context of our conversation that's what we're doing, right? And so I can just ask it, hey, build me this, and it knows what's going on. And then this is actually a slightly better prompt because I said, all right, I need a model of the solar system. I need it to be animated. And the solar system is actually pretty detailed. It's a pretty detailed element, right? Now if I'm a developer, I'm going to have to go and I have to look up, all right, what are the planets again, right? How many are there and how far are they from the sun and what is their relative speed? All these different things and their orbits, all this stuff. How are we going to figure all that out? And so normal developer is going to take some time to go do a bunch of research to figure all this out. AI just knows it, right? And so that's incredible. We're already saving time and doing some pretty cool stuff here. So we'll come in, we'll delete this one. And we'll click play. And here we go. Hey, I have an animated model of the solar system, nice and labeled and working pretty well. Okay. It's not interactive. I didn't ask it to do anything that way. Right. It's just there. It's just working. And so I'm actually kind of happy with this. So let's go back here one more time and say, um, add a, uh, slowly rotating, rotating. Sorry. I like to get my typing right. Slow rotating, uh, starry background. With twinkling stars. When I click, have a shooting star shoot out from the cursor with a brilliant and vibrant tail. Okay, perfect. All I'm doing now here is iterating. I'm just talking to it. One thing that a lot of people get really worried about with AI is that they have to say things just right. Right? They have to know how to prompt and have to know how to do all these, you know, weird formatting things. And the reality is that might have been true a couple years ago, right? 3, 4 years ago, 2 years ago even maybe. But in the last year, and really in the last 6 months, a big, big focus has been placed on allowing anybody to use AI. And so all you have to do is just talk, and AI interprets what you mean, builds its own prompt internally behind the scenes, and then processes it and does it all so fast that we can't even, you know, we can't even hardly notice that it's happening. Okay. So we're going to click play now. So now we've got our thing. We've got— okay, these stars are rotating, right? And they are slightly pulsing. You can see some of them pulsing on and off. See that? So that's cool. And then let's see if we can click. Hey, I got a shooting star. Awesome. So I built something that works. Okay, now you guys get to do it. All right. So you can take my prompt that I gave you Okay, in the chat, which is using p5.js, it says create me a mind-blowing visualization and give me the code in a way, blah, blah, blah, right? Replace mind-blowing visualization with whatever it is you want to build, right? Maybe you want to build a Pac-Man clone or a Tetris clone or whatever, some fun game, or maybe you want a visualization of the Earth and population decay or whatever you can think of, right? Maybe don't, don't focus so much on this one. On a business thing, right? Like you could build a dashboard for your retirement clients, but probably not for this one. Okay, we'll do that in a minute. Right now I want you to do something fun. Okay, so come up with something, build a little game with, you know, say, hey, dinosaurs chasing me, you know, and I have to run away with the keyboard controls or whatever, right? So give it as detailed a description as you want. Drop it into Grok. You can do it in ChatGPT as well. I'm using Grok here just to kind of bounce back and forth between a couple different models. Um, but yeah, go for it. And I'm going to keep iterating a little bit on mine. And this is a time when you can ask questions. Okay, so if you just say, hey, I don't know what to do, or what do I do here? And then as soon as you're done, as soon as somebody has something working, um, well, we'll ask you to share your screen and show us what you got.

I just want to tell everybody that I'm having fun right now, so I'm glad. I hope you are having fun too.

Love it. Okay, I have one.

Got something working?

Yeah, I don't, you know, like I said, that's okay.

Let's see it.

Let's see it. So I just said, go ahead and share your screen. Perfect. I wanted a birthday game for my grandson who's turning 2 in October. So it's a happy birthday pop Mickey's.

Yep. And then do you just like click on them? You just click on the Mickey Mouses to pop them?

Well, you know, it's good to know that, but they're not.

Oh yeah, there you go. You just got to hit right now. There you go. That's it. Awesome. I love it. You could probably do this on mobile so they could just tap it too, right?

Yeah. Well, so now how do I like export this? Like how do I use it now?

Yeah, good question. So we'll, we'll get to that a little bit later, but basically you've saved this code. You technically could you could just save that. Well, you actually, you will have saved this conversation automatically. So you could just pull it back up and just click play again.

Or could I do share?

Yeah, yeah, you totally could share. Now sharing it would actually share the entire conversation, but then somebody could still go in and say, hey, I wanna run this code and just click play on it, right? When you're developing inside of Groq or inside of ChatGPT, it's not really built to share it externally, right? But you could copy it. You could copy the code. And so if you just click on the right there, just to the right where it says code, see that? Yep. Yep. So now you could, you could copy this and then take it somewhere else where you'd then be able to make it into like a website or something.

Right.

Which we'll be able to show you later. Okay. But this is not, this is great. Awesome. You got something working and you were able to share your screen properly. So that's, this is actually guys secretly a test to see if you actually know how to share your screen. So, uh, well done. Thank you. All right, who else? Who's next?

I've got one.

Go for it, Jamie. All right, Bitcoin something, real easy. And what is it supposed to do?

It just—

you just click and it does a boom, a Bitcoin boom?

That's it.

Yep. That's awesome. Hopefully this is good for my Bitcoin holdings, right? Click it more.

To the moon, to the moon, to the moon, the Bitcoin moon.

I love it. Okay, fantastic, Jamie. Um, well done. Okay, and you guys are starting to notice too, probably, um, p5.js is not a language, a front-end language that you would use to make, you know, Fortnite or Call of Duty or something that's really graphically impressive. It's supposed to be very simple, very simple shapes, simple things like that. You can use it to do some pretty fun things, but you would need to be able to import some different graphics, probably in a different interface from this. All right, who else has something ready?

I got a syntax error.

You got something?

I got a syntax error, I said.

Okay, perfect. That's good. If you get a syntax error, just copy that error. You'll see it down on the bottom left there and just give the error to Groq. You don't even have to say anything, just paste. There'll be a syntax error and it'll say, oh, sorry, I did this, this, and this. Let me fix it. Boom, and then it'll work.

I had an error too, and at the bottom right it said fix.

Nice. Well, yeah, that's because you actually did it inside of the Groq editor, and so it can fix it right there. When we're going over to this p5.js editor here, you have to do it a little bit. You did the easy way and actually probably the smarter way, Laura. But what I'm trying to do here—

I didn't even know what p5.js was. I didn't even know what p5.js was.

That's okay, that's great. Um, Danny, I just saw your thing, looked like you had a shooter thing, but was it not shooting properly?

No, it was, it was doing all right. Um, oh my, I'm unmuted. Okay, yeah, I can— I was just sharing it, but I just put in a Space Invaders, kind of like new school Tron version. So yeah, yeah, freestyled, and it's working.

And it's working.

I'm not like playing with it, but yeah, it's something.

Yeah, you know, and guys, the way that you— so if you spent 20 minutes, right, refining this and making it smarter and, you know, working on the controls, like I added, you know, over here on my end, I added, you know, the ability to have like a restart and I can, I can, I can, I'm supposed to be able to play. I'm supposed to be able to shoot. Yeah, it's not, it's not quite working. Oh yeah, it's got The front of the bus. Anyway, so I can, I can try to get a high score. So I've got waves and I've got a score. It's really hard. I need to make it easier. I need to ask it to make it easier. But anyway, bottom line is it's there. And, um, so you can turn it into like a fully functioning game. You know who really loves this kind of stuff is my kids, right? My kids are ages 8 to 13 and they just, you know, hey Dad, can I make an A again? Yeah, absolutely, go do that, right? It's way better than letting them sit down and play Fortnite or Roblox or something like that. You know, they get to use their brain, build something, be creative, and use AI at the same time, right? That's a win-win-win across the board there. All right, who hasn't shown something yet? Ryan?

So I decided to go through ChatGPT.

That's fine.

Great.

Took forever. I'll show you. Let me show you my screen. In fact, it just barely finished doing that first thing.

I don't know. Yeah, you probably have it just on the normal mode and it does a lot of hard thinking. You have— when you're doing like fast iterations, you want to change it to fast mode. And that's it.

So let me, let me share my screen and then I'll have you.

Yeah, now let's look at it.

Oh, where I go to fast mode on that. So see the screen here?

Okay. Yeah, just push that allow button so it gets out of the way there. There we go. All right.

So this, like, like the whole time we were talking and showing this was just—

it was building this amazing thing. That's cool. Let's try it. Go ahead and click copy code. Okay, and, uh, yep, there you go.

Great. And then just into the p5.js?

Yeah, yeah, p5.js right there. Yeah, just into the web editor is great. Yeah, um, start coding, bottom left there. Okay, and then go ahead. Yeah, get out of that little X thing. It wants you to donate also.

Let's see.

Top right. There you go. All right, perfect. Now, yeah, go ahead and select all and delete that.

So what was that command?

What command? Yeah, Command+A selects all. Yeah, go ahead and click, click in there first to see you select the— there you go. And then Command+A, delete, and then Command+V. All right, let's click play and see what happens. Uh, top left there, you— that play button. There we go. Okay, go and click Allow All so that we get rid of that little cookies thing.

Um, sorry, I'm just moving my, uh—

yeah, yeah, you're good. All right, the Zoom thing, I can see you guys on Zoom. Yeah, okay. All right, so we've got an error down here in the console. Okay, so go ahead and scroll up in that little thing there. There we go. Yeah, very top there. All right, failed to use program. All right, so go ahead and just copy that, that red part there. Okay, just try to select it, copy it, and then take it back over to, um, yeah, back over to there to ChatGPT. And then, um, in the top there where it says ChatGPT-5, click on that top. Yeah, just top, top left where it says ChatGPT-5 and a down arrow. Yeah, right there. And then go ahead and click Instant. Yeah, all right. And then go ahead and paste that error into the, into the chat there. Yep. And just click go. There we go. Okay, so it should come back basically immediately.

Copy code again.

Here's a fixed— hold on, go scroll down a little bit. Okay, just say this, just say give me the entire code again, but correct. Perfect. Got it. All right, so you got to wait when it says copy code, it actually is thinking, it's building, it's following its pattern matching, you know, process, and it has to finish. So you have to kind of watch it. Watch it go all the way down. Still thinking, still writing. Okay, now it's ready. Go ahead and click Copy Code right there and head back over. Same thing. Click Command+A and delete and Command+V. There you go. Let's try it again. See what happens. Go ahead and just click Play top left. Yeah, we got something. That's pretty cool looking. Very cool. What is this thing?

Some Grateful Dead.

Yeah. Awesome. Or Windows 93 screensaver. That's good, Ryan. I love it. That's pretty cool. Honestly, I think that's one of the coolest visualizations I've seen using P5.js, because like I said, P5.js isn't good at crazy stuff. And that's really cool. I really like that thing. Yeah, that's cool. Now here's a cool thing, guys. You could actually take this, this visualization, and you could grab this code later on when we're building a website or something, and you could make that the background of some little module or some little card on your website or something like that. P5.js is actually a very lightweight, you know, front-end kind of design and graphics engine. And so you could throw that on there and it would be just maybe behind your message of whatever whatever you're trying to do. Now, obviously it would have to be the right, you know, the right thing and you'd have to like it and, you know, but I'm just saying you could use that there. So that's pretty fun. All right. Thank you, Ryan. Well done. All right. Let's see who else hasn't gone. Tyson, you ready?

You bet.

And then Mike will bounce to you last.

Okay.

Okay, so let me get rid of this. I did a, um, video of the BYU-Utah football game last year.

Whoops. So let me get back on that. Where am I? Hey, you're like me. I have like 900 things open. I have so many.

Yeah, but now I don't know. I have 4. This is why I can give my brother a hard time.

Let me find it again.

Let me stop sharing for just a sec.

You're good.

Can I ask a question while he's looking for that? Go ahead, Mark. Explain again the difference between the ChatGPT and this other one that works.

Yeah, so the short answer is AI is not like browsers, right? Like, you can use Chrome or Safari or Internet Explorer, Edge, and if you go to foxnews.com on any of those or cnn.com on any of those, they will look the same, right? Those are browsers that are rendering some data and it's all the same, right? So it's not like you can ask the same question to Grok or to Claude or to ChatGPT or to whoever and get the same answer. They're all— they're called LLMs or large language models, and that's just the technical name for the actual AI itself. Okay. And really it has to do with a lot of things. It has to do with how it processes the data, what data it's trained on, you know, what instructions it has in terms of any biases or non-biases or telling it to not do whatever. Right. Like try to avoid this, try to avoid that. You know, one thing you might notice in some LLMs is, you know, every idea you give it is good, right? And sometimes that's good, sometimes that's bad, right? Sometimes you want to feel good about your idea, but sometimes you just want it to tell you if it's a bad idea, right? And so that's based on a lot of different things there. So this is pretty cool. We'll talk about that a little bit more, Laura, as we keep going. But Tyson, So is this basically every time there's a completed catch?

There's—

this was like the last 50 seconds of the game. So it just, it just shows what happened in those last 50, 60 seconds for BYU to win the game from those plays.

That was a crazy game, man. That was really crazy. So I am about to announce on LinkedIn as soon as I I think I'm going to do a press release, which means I'm going to have to wait another week to actually announce it. But GenAIPI is now the official AI partner of BYU Athletics. Oh, cool. Super cool. Really, it just gives me free tickets to all the games and lots of good stuff there. And actually, no, what we're doing is we're working with all the athletes. We're going to help them be ready for the AI-powered job market. So we'll be doing all these training sessions every week in October. Every Monday, I'll be meeting with all the students in big groups, and then be smaller groups, and I'm gonna make sure that I single out AJ so I can hang out with him. Uh, there you go, some fun stuff there. So anyway, um, well, that's cool. I love it. I'm a huge BYU fan, and seeing Jake Ritzlaff here, man, he was, he was awesome in that game. That was, that was really fun. All right, um, Colorado game was fun on Saturday too. And okay, last but not least, Mike, show us what you got. That's a cool one. That's a cool one, Tyson. I really like that one.

Mine's working too, so when Mike's done, I'll do mine.

Yeah. Oh, yeah. Yeah. Awesome. Okay.

You had the error, so I went retro there with Santa supposedly going down a chimney.

I love it. So he just goes down the chimney. This is—

yeah, it's not as sophisticated as Laura's where you can just click on, right?

Yeah, that's just it. I love it. Super easy. Create an animation with Santa going down a chimney. Love it. It's a cake. Piece of cake. All right, Ben, cool.

Um, I used to play this game on Atari called Joust, if you're familiar with it.

It was—

you get on a giant like ostrich bird and fly like you're jousting from the old, uh, the old, uh, Middle Ages. Anyway, so, uh, I just asked it to replicate it. So here's— you get a jump over it. Normally there's like lava below, so, right? Just try to avoid all of the things. If you miss, you can go backwards. Try to avoid the other birds flying at you. That's it.

You can go through the screen on one side.

That's pretty cool.

Yeah, yep, yep. So yeah, that's it. If you fall in the lava, it's over.

I mean, obviously you'd say, hey, make sure that if the character is sitting somewhere that it pushes things out so you can't just sit there the whole time and win, right?

Yes, right, exactly.

I just keep telling it to iterate, but this is great, right? So you've got a score that goes up every time one of those passes. And anyway, Guys, look, I don't know about you guys, but the first time I did this as a non-developer and I was able to just bring up a simple game or a simple animation or something, I was like, holy cow, what else can I do? I started pushing it and pushing it and pushing it until I built GenAIPI. I want to just take a second and tell the GenAIPI story in just 3 minutes. But I like to post on LinkedIn every day and I've been getting really good with vibe coding and so earlier this year, It was a Thursday morning. I just said, okay, what am I going to post on LinkedIn? I said, you know what, I'm going to try building something in public and I'm going to try building a piece of software that I'll start today and I'm going to see if I can get my first customer by the next day. And which is obviously like a crazy task, right? I'm going to start a new business, new software. I'm not a developer, whatever. And I'm going to have somebody paying me for the software the next day. Okay. So I posted that and got a bunch of people following me on LinkedIn and, you know, following the post and, you know, probably watching me crash and burn, wanting to watch me crash and burn, because that's more entertaining. But anyway, I just started doing it and, you know, started posting updates every couple hours, whatever. And by the end of the first day, I went till about 10:00 PM, just all day, went from about 9:00 AM to 10:00 PM. And I got pretty dang far, right? Like I worked out the whole business problem strategy, things that we're going to do here in a minute. And I ended up, you know, getting pretty dang far, but then I was like, okay, I've got a lot more to do. And my idea, my idea was simple. I wanted to create something in the AI space around training. I wanted, I wanted to help businesses and employees be more AI ready. I had been having lots of conversations with tons of business leaders in the area and, you know, smart people that are my neighbors, right? None of them are using it. Now, these are all a bunch of ranchers and, you know, construction company owners, whatever, call it more blue-collar stuff. But they just, most of them didn't even know what ChatGPT was. They'd never touched it, they never downloaded it. And I'm just like, this is scary, you guys need to learn this stuff. So I thought, okay, I'm going to make like the equivalent of an IQ test but for AI, right? So I'm going to see how good are you at AI, how well do you know it. And then if you score poorly, I'm going to offer you, you know, these training courses. And if you score well, then I'll offer you a certification, right? And my idea was, okay, I'm going to be the General AI Proficiency Institute, right? And so I'm going to be this, you know, genaipi.org, very, you know, official sounding, right? By the way, that's worked so well that I've had government organizations from Australia contact me and say, "Hey, can you help certify our people? You guys seem to be the standard." And I'm like, "Absolutely, I can," you know? And so, I've done that, which is awesome. So, you know, the ACT had its first test at some point and now they're a global standard. So, I knew the model I was going after. That's what I was going for to become the standard. So I built this thing out and, uh, Friday came along. I worked until about 3:00 AM Friday, right? Worked about 15 hours Friday. Could not quite finish it because I had some payment thing I hadn't set up. Everything was working except I had, it was still free. I hadn't set up Stripe yet and I obviously needed Stripe to be able to get some payments. And so Saturday I get up, had some random things, service projects and things like that and hanging out with kids and about 4:00 sat down by 3:00 AM Saturday night, Sunday morning I was done, right? It's finished, right? And so I clicked go on my Facebook ads and let it run all Sunday. And then Monday I get up, check it, and I had about 120 people that had clicked on the link to, you know, and landed on the site and nobody had purchased. And so I was like, what the heck? Like, I would, I would've guessed somebody would've purchased, right? It was pretty cheap. It was anyway, but nobody had done it. So I went in and I found out that I had a little Stripe error and I was like, ah, dang it. So I fixed it, took me about 10 minutes to fix it, republished it, and I, it was ready to go. And, and guys, the site that I built, by the way, right? I mean, most of you have used the site because you purchased something or you've been on the site, but it was this, right? This is what I was able to build, right? This free AI assessment, start the test, right? This was the idea. You're going to take a 12-question test and I'm going to funnel you towards either courses or certifications based on how, how you did, right? And so anyway, then I remembered, you know what? I'm a sales guy. And so I called up, I just started calling some people, messaged people on LinkedIn. And within about 15 minutes I had somebody. This is on Monday. Yeah, it was on Monday. So I started Thursday. This is 4 days later on Monday. I started messaging a few people and I got one person to say, hey, yeah. For those of you in the area, you might know Johnny Hanna or you've heard the name. He's the founder of Entrada and Homey and a couple other big things, right? But I was like, hey, you know, I've got this new AI thing. I'm going to see how smart you guys are. I want to test all your employees and I want to help train you guys, right? And he's like, sure, let's do it, right? And I had a $15,000 contract within 4 days of launching this new software business and training company, right? And, and that was the beginning. 6 weeks later, I had $180,000 of revenue. And, and I already, you know, jumped to the end earlier where 7 months in now, Jenny IPI is doing some big-time revenue. We're bringing really top-tier talent in and it's worth $10 million. And I've done a lot of revenue, right? And so, I mean, completely income-replacing revenue type stuff, right? In 2016, I founded a company called Seek and I raised $13 million, as I mentioned earlier. I am making way more money right now than I ever did in the 7, 8 years I had that venture-backed $100 million business.. It's just I'm in a different, I'm in a different spot now. And I was able to do all that, guys, as a non-developer solo, right? Now I'm finally to the point where it's like, okay, I've got so many customers, I need people. So I'm bringing people in to make that happen so that I can expand, right? The goal and idea was to say, how far can I take it by myself with AI? And, and I'm still doing a ton of things like, like, yes, last night I was working on, you know, new updates for the website and some new features and some new organization and better SEO. And different things like that. And I pushed some new updates to the site. So I'm still going to keep using all those great, you know, elements that allowed me to do what I did. But now we're going to have more people. And obviously I'm training Ben and Danny up to say, hey, you guys need to know how to do this stuff too, so that you guys can act. You can do the work of 10 men, right, by yourself, right? Or Laura, 10 women, really 5 women can do the work of 10 men. So yeah, 10 women would be really, really powerful for you. So anyway, bottom line is you can multiply yourself just you know, exponentially. And, and I want to, I want to point out something that's just so crazy here, guys. So I raised money for my last company Seek because I did not know how to develop and I wanted to build software. Simple as that. I had to, I had, I needed to pay somebody, right? I did have a co-founder that was technical and whatever, but we had to, we had to pay people to do things that even he couldn't do. So I went out and I raised $20,000, then $100,000, then $500,000, then a million. But it wasn't until we had about, we'd put about $500,000 into the business, into the software that we actually knew we had a business that was viable. Right. And how risky is that? Right. I mean, Ryan and Tyson, your job is to mitigate risk for people's finances. Right. And I know you have wealthy people that come to you and say, hey, I want to put $500,000 in this startup. And you probably say, okay, just know that it's probably going to zero. Right. I know you've said that to people. If you're okay doing that, and I think you are, your portfolio's big, you can do that, right? But maybe you should only put $250,000. You know, I don't really like you guys when you say that. I want, I want that investor money. No, I'm just kidding. Um, you should say that. You absolutely, absolutely should say that. And if you're an investor, sorry, if you're an entrepreneur asking for money, you need to say that too. Otherwise you can get sued. So, um, you gotta, you gotta make sure the risks are very, very clear, but it's very risky. It's very, very risky to invest in entrepreneurs. And as an entrepreneur, it's insane. Insanely stressful to take a bunch of people's hard-earned retirement money and get $13 million of it and then, hey, go turn it into something good, right? Don't give up or else everybody goes to zero, right? That's a stressful position to be in. In order to build everything I built with GenAIPI, it would have taken me back in, let's say, 2020, it would have taken me $3.2 million to build out all of the curriculum, the AI tests, all the courses, all of the coding, the entire platform, everything that I had. $3.2 million in 18 months. That's what it would have taken me to build it. I built it in 3 days and it cost me $140. So just let's just think about that for a second. Okay, let's do some, let's do some math. I'll share my screen and show my handy-dandy calculator here. $140. Divided by 3 point— whoops— 3.2 million. Okay, let's go like this and let's go minus 1. So we did it for 99.995625% cheaper. That's insane. Okay, and we did 3 days instead of what, 700 days, you know, 600 days. Okay, so we're doing it 99 99% faster— or sorry, 99— yeah, we're reducing the time by 99% and reducing the investment needed by 99%. Okay, so that's pretty crazy. That's pretty crazy, guys. We live in a whole new world. So this is Replit. We're going to dive into Replit here, and, uh, you don't need to get into Replit yet. We're not— I actually don't want you to get into Replit yet. We're going to build something the wrong way first, like we— like we— like I like to do. Okay, then we'll build it the right way and I want you guys to build it the right way. So you don't need to be in Replit. If you haven't set up your account yet, you could probably do that right now. But every time I come in here, I build something new. I come up with some new idea and, and I just kind of, I kind of build something. So I'm just kind of looking around my room to think about what I should build. I've built over 100 apps in here, so it's hard to think of. I like to think of new ideas. I want to try something that I haven't done before. So I like a lot of things. I like kayaking, and I like music, and I like movies. I like movie music. I wonder if we could do a music thing. I did a music, a sheet music app one time.

That was pretty cool. What is Replit used for?

I am about to show you. That's a great question, Maura.

I mean, is it a website builder, or?

Yes, amongst many, many other things. Yeah, it's a coder, it's a developer, right? It's an AI-powered developer. And so a developer can build anything. You can build websites, it can build games. You see here, general games, agents, automation. It can do all kinds of stuff. Okay, um, let's see here. I've got an idea. I actually was talking to my dad about this the other day. So someday I'm gonna die, right? Um, Ryan and Tyson plan for that all the time. So, for other people, not for yourselves. I mean, hopefully, you prepare for yourself too. But anyway, one of the things that I'm worried about when I die and my dad's worried about when he dies is just like access to systems, like passwords to bank accounts and how to access my crypto accounts and how to find all the email addresses and how to log into the Utah State LLC registrations to be able to just all the different stuff that's going on with my life. So basically something that, that, you know, I'm going to have access to, my wife's going to have access to, probably anyone else I would want to have that, maybe my oldest son or my, my dad or, you know, just maybe be even, you know, maybe the password and the information would be in my will or something like that. Right. That just kind of allows them to say, okay, you have access to this system and then it's going to help you get into everything else. So obviously we're going to have to care about security. We definitely don't want this to get hacked because it's going to have passwords and information and ways to get in there. So that'll be a critical component here. But anyway, I want to build a system that will help my loved ones be able to access systems, or actually, yeah, systems and accounts after I die. Bank accounts, social accounts, email, business accounts, and Many other things. Passwords. Yeah. Yeah. Obviously the passwords would be, I think that would, yeah. Bank accounts.

Yeah.

Yeah. It will, we'll say that it will need to be very secure because there will be many, there will be many usernames and passwords tied to sensitive information. Let's build it. Okay, so I'm just going to say let's go, right? One thing I want you to notice as I'm typing here, guys, I'm very casual in how I type often, right? Doing that on purpose. I'm just talking just like I would to a developer. And so again, what Replit can do here is really incredible. It's really incredible. I appreciate you entrusting me with this sensitive and important project. Okay, well, I'm glad you do, right? Um, okay, hey, this is crazy. Before I create a plan, I need to clarify a few key aspects to ensure the system meets your needs. Access trigger mechanism— how should your loved one gain access after a specified period of inactivity on your part? It's just a wait, like, oh, you haven't logged in in a while, you must be dead. Uh, no, that's kind of weird. Um, access trigger should just be a username and password or auth code from verified phone number, authorized phone numbers, or something like that. Um, recipient configuration— how would you like to organize who gets access to what? Um, Yeah, different access levels would be nice. Storage preferences, I'm comfortable with cloud-based storage. Also, yeah, we'll just say that for now. Go. Now we're going to just start building, no plan to approve. Understand. Start building. We're going to click refresh here. All right, hopefully it's working. Probably is. Um, so now this is going to take a minute, guys. This is going to take a minute to, to do this, but It used to take, in order to build just an MVP of this, it would have taken easily, like something this complex with security, things like that, with all the different types of, you know, access that I'm, I'm telling it it needs, it would have taken probably somewhere between $20,000 to $50,000 depending on how intricate you wanted your MVP to be. And that would be paying, you know, some dev shop somewhere, maybe here or maybe out in Ukraine or whatever, right? Um, and I've worked with a lot of these types of dev shops out there. Like, this is easily $20,000 to $50,000. It could be $100,000 depending on how fully functional you wanted it to be. To get it to where it's production-ready, customer-ready, it's at least a quarter million bucks. Okay. And we're going to build it here for probably $2, and I've already done all the work I need to do on it. Okay. Now, one of the things that, um, one of the things that I, I didn't do well here, you know what? Replit's been having this weird problem. And I think it might just be on my computer. Because I actually haven't seen it do this on other systems. Your app is having trouble. I wonder what that means. Yeah, it's totally working. It's just like a display issue. I don't know why it's having that problem. Replit. Error. Yeah, that looks like it's probably my problem. Yeah. Temporary browser cache. Yeah, I probably just need to restart my computer. All right. So, so anyway, guys, to be able to build something like this with just a few sentences is, is crazy. Now, while this is working, we're going to, we're going to have a little show and tell from Mike. All right, Mike, um, it's actually less of a show-and-tell, just to tell. I want you to help us understand, Mike. Just tell us, um, how you get ready in the morning. Just walk us through. You wake up and then you're ready to walk out the door. What happens in between? What's your routine?

Oh, okay.

First thing I do is, um, I'd go— I hit the stairs, do the stairs, do a morning workout. And, uh, grab coffee, breakfast, and then I'm out the door.

Okay. All right, let's zoom in on, um, getting breakfast. Okay, tell us, tell us how you get breakfast. Like, what do you normally eat? Tell us step by step what you're doing, right? You're getting your coffee, you're getting your cereal, you're getting your milk. What is it? Walk us through it step by step.

All right, um, well, just go to the coffee machine, make the coffee, um, pop bread into the toaster and go through that sequence. And then within 2 minutes, I'm having breakfast.

Okay. All right. Very good. So what I asked Mike, which is— thank you, Mike. Right. So what I asked Mike was very simple. I said, how do you get ready in the morning? Right. And so that's kind of what we would call a top-level task, right? That is the task. That's the overarching goal. Then underneath that, there was some automatic, you know, you know, subtasks that he's just like, oh yeah, my brain's going to do that. I'm going to go, I'm going to work out on the stairs. You're right. Immediately went to the stairs, right? I'm going to work out. I'm going to get my coffee. I'm going to get my breakfast. You're probably going to brush your teeth and go to the bathroom and shower and whatever those things are, right? You've got all these different things. And then each of those tasks, right? When you're in the shower, you're going to, you're going to wash your hair and wash your face. You're going to wash, right? And so you're, each of those subtasks, even, even when you're washing your hair, right? That has subtasks. Okay, I got to pick my shampoo or my conditioner, and I have to, you know, reach out and I got to put it in my hand, and then I'm going to lather it, I'm going to let it sit, and then I'm going to do this, right? And so we have this what's called in our brain, we do it really, really well. Um, let's just try this, see if it works. Okay, so we have our brain works really well at doing something called chain of thought reasoning. Chain of thought reasoning means we take a high-level task and we break it down, right? Or COT reasoning. AI has only been able to do this for, I'm going to say, about a year, maybe not even quite that long. This is kind of the newest capability that AI has, and it has enabled what are called agents. Okay. An agent is able to break down a task and then, you know, break it down into smaller tasks. And then execute on each of those tasks and then sub-subtasks and sub-sub-subtasks all the way. And as it finishes those, it reports back and it tells the agent what's going on and how it all connects together. And then it's able to go and build something, right? So when I'm asking it, hey, build me a cloud-based website that, you know, is a secure vault storage for all this sensitive information, okay, it's going to say, okay, great. Well, I guess we're going to need a website probably, and we're probably going to need a login and we're going to need a place to store social passwords. We're going to need a place to store business accounts and we're going to need different types of users. And so it's able to think through and reason and process and say, okay, I'm going to need all these things. It's going to build a huge task list, which it did over here. Right? Here's the initial prototype. It's plan, right? Whoops. I didn't mean to do that. Here's its plan. And then it will— luckily the agent keeps going even if you refresh it. So no big deal. But it sees all this stuff here and it's going to start building out a plan to make that happen. Right? Right now it's generating design guidelines. And so it's just trying to think through what is this going to look like? Okay. And so that chain of thought reasoning is very, very powerful. Now, in Replit here, we're doing this for software. Okay. There are other agents out there that can do all kinds of things, right, that are more general. For example, you've probably at this point talked, you've spoken with an agent on the phone, right? An AI agent on the phone. Okay. Now, There have been voice-powered phone trees for a while, but an AI agent is a little bit different where you can actually have a conversation with it and say, hey, question, what do you do? And it can talk back to you like a human. Maybe you may or may not have spoken to an AI agent, but they're on the phone. There are also AI agents that are just more general task doers. You can ask it to go research something. You can say, hey, I need you to go find me 20 new prospects for retirement, you know, people that are probably approaching retirement or might be good prospects. Go, you know, go search my LinkedIn and find this. And it can actually navigate and use your LinkedIn and act like a human typing things in, clicking on things, researching, reading, checking something with a Google search, whatever. It can do all these different tasks with the overarching task of, hey, I'm going to deliver a list to you of 20 qualified prospects, right? Or maybe semi-qualified prospects. You can then reach out to and see if it's going to work. Right now we can see it's starting to do some things here. It's generating a few design guidelines, it's created a server, it's creating the schema and some encryption stuff. If I click on this, I can actually see it is written code. It's writing code right here. If you're a developer and you've ever set up a project like this, again, I'm not a developer, but I've been around developers a lot as having been a tech CEO for for decades. I've, I've been around this kind of technology and I've seen it, and so I understand how hard it is to just even set up a project, right? Here's 166 lines of code and all these different things—users, beneficiaries, accounts—and all these fun things, right? It's almost like, it's almost like life insurance, right? It's kind of how this is setting it up, which makes sense, right? It's a death kind of thing, so it makes sense as to why I would do that. But it's already written over 1,000 lines of code, right? Just sitting here while we're talking about chain of thought reasoning, right? In the last 5 minutes, it's been able to do that. And now it's on a roll. It's just, okay, I need a database, I need schema, I need to figure out the storage and the encryption. I need to figure all these things out and I'm going to make this happen. Okay. And in just a minute, we'll be able to see what it actually is thinking through, right? It might take— we might take another few minutes and that's fine. We've got plenty to talk about. So I'm going to pause here for a second and I want to just open it up for a question or two if, if needed. If not, we can keep going. But on agents. And, you know, this is at the very beginning. You're going to understand them much better by tomorrow. But any questions on this, what's happening or how we would use this?

For you, how often are you just to kind of spell check what you're doing or asking like, hey, I've put in this, these prompts, what are things that I might be missing that are like, I mean, it's almost like a question behind what is the creation of it? Like what are like help me look at security flaws or something more specific to develop out for somebody based on a request? Yeah.

So this is a really good question. And, and it's, it's a question that a good leader would ask, right? I'm going to answer it by saying, well, what if you actually had hired a developer, right? And you told them, hey, I want to build this thing. And then they come back and they say, hey, here's this thing I built for you, right? You would immediately say, okay, is this secure? Are we going to have data leaks? Are we going to get hacked? Is it following, you know, ADA guidelines, American Disability Act guidelines? Is it following GDPR? Are we following laws or is there anything like, you know, those types of things you're going to ask that, that developer, you would in the exact same way ask Replit and you can absolutely do that. And it knows all the laws, it knows all the securities, it knows all the protocols for how you're supposed to do those things, right? Maybe you need SOC 2 compliance or maybe you need ISO 97001 compliance or whatever it is. There's a lot of different compliance type things out there. And even if you don't know what those compliance— all these things I'm saying out loud, you're like, I am not going to remember anything Jon just said. That's fine, right? You can say, hey, I'm sure there are some compliance issues we need to worry about here. Can you run me through what you think matters most for this app, right? Oh yeah, I think you need this, this, and this. Great. Let's fix all that. Is there anything else? Oh yeah, this and this. Okay, great. Let's fix that. It's fixed now. Is there anything else we need to worry about? Oh yeah, you might want to do this, right? And you just keep doing that. So it's like, you know what, you are enterprise ready. You are enterprise-grade security. You are enterprise-grade, you know, compliant. You are ready to launch this thing to the public.

Okay, quick, quick follow-up question on that. Does it make more sense to do that throughout that process or have kind of what your finalized version is and then retool it based on other stuff like I just mentioned, like, is it— it's vibing, I guess, is the way it's defined.

But is there a best practices for that? I would say both probably, but more intense in the end, right? I would say in the beginning, remember, we need to consider security throughout, right? And I already said that to it in the initial prompt. I was like, remember, this is sensitive data. We got to make sure we build a secure storage, right? Right. And so because you said that in the beginning, it's going to remember as you go through.. But in the end, you're gonna say, all right, now we're gonna test this. And now I want you to try to hack this. And then you can take the system and give it to another AI and say, hack my system, find holes, break it. Right? And so as long as you're just thinking to ask those types of things, you know, some guys, sometimes this won't matter, guys. If you're building your personal brand website, like, okay, like there's no passwords or logins. It's just like, hey, here's who Jon Cheney is, right? Here's who Ryan Thacker is, right? And here's all all the cool things he does. Here's his companies, here's his links, here's whatever, here's his socials, you know, whatever. But, uh, but if you're storing data, especially for clients, right, there's some liability there. And so you just have to be smart enough to say, okay, have I, have I shielded myself, right? And have I done every reasonable effort to do that? Because really, you know, there's— if the Chinese hackers decide to target you they might get past you, right? But as long as you have done everything reasonably required by law and every reasonable effort and you can show that you did, which you'll be able to because it's in all the code in here anyway, you know, your liability will be greatly reduced, if not eliminated. So even if there is a breach right now, if there is a breach, you also should know there's laws. You got to tell people, right? You got to, you got to report it, right? There's all kinds of things you got to figure out. Probably you got to report it to the police. In some cases as well. But again, I don't want to scare you guys away from this. I really don't want to scare you away from doing these kinds of things. It's not that hard to get hardened, right? My, my GenAI API, like, I do store some sensitive data in here. If I go into the admin, I've got some email addresses and things of people, but not a lot. I'm not storing anybody's credit card. I'm not storing anybody's, you know, Social Security number or even their address, right? I have their name or their username, their name and their email. That's really it. Okay. And so I'll show you guys my admin later. But one thing that I do have here, right, guys, is things like my privacy policy, right? That's important. You got to have that. I've got terms of service, right? I've got cookie settings that allow all kinds of things to work, right? My— in order for, you know, Facebook pixel tracking and other things related to purchases to work well, you actually have to have cookies. And there's, there's actually laws around having some of those things. And so look at this, we comply with GDPR, CCPA, and other privacy regulations, right? Here's our privacy policy. And I've, I've spent easily 6 hours on this system, right? Just on this system, making sure that it works right and that it enables things. I've spent real time on it because I do know, I have been a CEO of, you know, big companies that are handling Walmart data, and like, I have to make sure that it's good, right? And so my developers had to worry about that, but I know that I have to focus on security. So security is important, right? We're getting a little ahead of ourselves a little bit on that, but it's important. It's very important. We want to make sure we're doing it right. Okay, this is still thinking here. It's— may I ask another follow-up question? Yeah, yeah, go ahead.

So, um, Ryan and I have written 4 books. We, we have a radio show, we have a TV show. We've got a lot of intellectual property. Yeah. And, um, So I'm sure you've had every time you do this class this question, how do you protect your IP when you're doing things like this where you're creating apps or you're throwing some of that intellectual property out on the web into AI?

How do you protect that? You can't. Straight up, you can't. If you can see it, it's not protected. That's for any image, any video. If you can hear it for music, any audio, it's going to be there. It's, it's absolutely going to be there. Okay. And that's okay. That's okay. Because now look, if you were to take your entire book, all your books, and upload it to ChatGPT and then talk to it about it, I can tell you this right now, ChatGPT is not stealing it all of a sudden, right? Um, you know, we'll be in— we'll work in ChatGPT in a second. I'll show you how to do some data controls and privacy things so that you— so it like forces to not even like save your conversations. But what it— oh, the only thing it looks for, Tyson, is patterns. That's all it's looking for, right? It's not going to say, oh hey, I'm going to steal all of the Thackers, you know, proprietary, whatever, and now I'm going to give this data out for free to everybody. It's not like that, right? Your data though is important because your IP is important because of how it's packaged and how people access it. They're not accessing it through ChatGPT. Even if they say, hey, tell me about what this book says, right? It might have some sort of synopsis or something, right? But if I want access to your book, I gotta buy your book and I gotta open it up and I'm gonna read it. And it's synthesized and organized and put to— put together in a way. Your IP is, is the, is the complete package. It's not the individual page and the data might be on that page. And that's really how you have to think about copyrights and things like that, right? Um, at this point you can't— you know, the cat's out of the bag, or we'll say the mouse is out of the bag. Mickey Mouse is well known by every AI. Why? Because it can see it everywhere, even, even well outside of copyrighted domains and places where it might show up. It's been used so much in pop culture and other things and places and other, you know, just it's all over the place. Everybody knows what Mickey Mouse looks like. And so AI is able to then, you know, figure that out very quickly because it's seen that pattern so many times. And so even the court cases are coming out right now that are saying, you know what, that's not copyright infringement, that's not trademark infringement. It's not because it's able to, like a human, because it's seen it, it remembers it. The fact that it's a digital memory doesn't change the fact that it's seen it. Now it remembers it and it knows it and you can't do anything about it. Now, in the same way that if I use Mickey Mouse as my mascot for Gen AI, I'm probably going to get sued by Disney, right? And so how I use it or how the AI uses it and actually then puts it into production, into commercial use, that's where the copyright infringement starts to take place again. And those laws are still being worked out. But being able to generate a picture of Mickey Mouse right now and show my kids, there is no copyright infringement right there. So really fun side, really fun discussion around IP. Keep asking your question. I'm going to start registering here while we talk.

So my next question around that is, let's, you know, we're the largest independent financial services company in the country. So we have a lot of competitors that want to always know what we're doing. Could they say something like, go out and find all the radio shows of the Thackers and mimic their, their radio show calls to action and, and different things like that that would help me to, to compete with them better, or, you know, different things like that.

Absolutely. Yeah. Yeah. I mean, it has access to a lot of that kind of stuff. It may or may not have access to everything, right? But you obviously could copy and paste transcripts in and things like that and say, all right, how do I compete with this? How do I make this work? How do I make this better? How do I stand out more? But yeah, everyone has access to that immediately, right? But again, your value in being the largest of what you are is, is being that. If I want to compete with you, I don't care if I have word for word, everything you are saying, your principles, your secrets, your hacks, your shortcuts, I am light years behind you because you have a huge brand and that is your IP, your brand and your customers and your goodwill and, you know, everything that people know about you. That's, that's where it all comes from. So, um, all right. So guys, I tried to register here. We now have Secure Vault. It's awesome. Um, and, and it, and I tried to register, I got a little error. And so all I did, was I screenshotted it, right? And it just said, I just said, here's like, this happened when I tried to register. Okay, this is, by the way, exactly what I would do if I had a developer. I would be sitting there with my developer and I'd say, hey, look what happens when I try to register and be like, oh, let me fix that for you, right? So he's going to check it and figure it out, right? Replit's going to go, um, help me out here. Okay, so let's see if we can get inside and see what's actually happening here. It's actually kind of interesting. I kind of asked it for more of a visual preview, and instead it gave me like a, a full-on, um, let's try again. Try again. Oh, it needs an encryption key. Um, here, we'll just call it, um, I don't know. So sometimes it'll ask you for something, it'll ask you for a little question, and you just, you just give it what it asks for. Okay. All right. It's running without errors now. Now let's make sure that's set up. See if I can create an account here. Yay, we're in. Okay. All right. So guys, I just want you to take a second and consider what I just did in a few minutes, mostly while paying attention to you guys having a discussion. Okay, this is crazy. Like, it's crazy to me still that it can do this. I can come through and I can add bank accounts and social media accounts and email accounts and business and legal and personal and add beneficiaries and activity logs and settings. Hey, the settings page doesn't work, but everything else works, right? And it's— and by the way, it's not that it doesn't work, it's that it hasn't been built yet. That's all it is, right? It just, it didn't go that far, and that's fine. I really just wanted to see like a preview of what's going on here. Okay, but I'm telling you guys, we're, we're $30,000 in, right? Like we should be, we should be $30,000 in. And right now we are, you know, the last action was 71 cents. I don't know how much this has cost me so far, but it's probably somewhere around $3. Okay, $3 and And I've got— I mean, it's free. It might as well be free. I could go on a walk around my street and find that much money on the ground, probably, right? Just coins that were thrown out. Um, it's basically free. Okay, but activity log— you use— I actually like that it's like a really, a really in-depth activity log. Um, that could be a really good security thing, so I can really see what's happening in the system. If anything weird happens, we could probably put some sort of a what's called an oracle or something that's watching everything to say, hey, if anything weird happens, send me an alert, you know, that kind of thing. So, um, what do you guys think about that? What— I mean, what, what are your thoughts on just watching something like that come to life almost instantly?

That's crazy. It's awesome. Jon, I've got a question. Um, do you have any experience taking any of these proprietary apps and loading them on, say, like Raspberry Pi?

I don't. I've never tried to do that. I, yeah, I've had other people pull pranks on me mostly with Raspberry Pis. But yeah, just building little, little things to mess with me. I had, I had some developers that loved to do that and they would like just hide stuff and make things happen where I was like, what is happening here anyway? So, but I don't know. I have never tried to actually do that. But the cool thing is AI knows how to do it, right? You could easily build something here and grab the code And one of the things I want to share, I'm going to skip way ahead here, but it's not that far ahead, right? But I'm just going to show you how this works. There is— where is the tab? Do, do, do, do, do. Sorry, they just barely changed or removed a There used to be— oh, here we go. Here it is. Okay. Goodness gracious. Okay. So here's all the files, right? The files for this are all— this is all the code that has been written. Okay. And so there's— we'll go down here. We'll go to— yeah, we'll go to like source and components and UI. Like there's code for every single element that's being used here. All the UI command, collapsible checkboxes, charts, carousels, everything's here. And, you know, all these different hooks and pages and, you know, accounts.tsx. And how does this work? And The beneficiaries page. Point is, it's all code. It's all there. You can actually take this and you can export it very easily into whatever. If you want to get more advanced or maybe pass it off to another developer, you can obviously share your Replit build with them, or you can connect to something that's called GitHub. GitHub is a repository, a coding repository, where you can then push that code out. And I'll just go to GitHub's website. This is, this is like a billion software engineers use this, right? I don't know if there's a billion, but there's 100 million people use it. And, and then you can push from GitHub to anywhere else. You can see here's all the code. I mean, this is— there's even AI inside of GitHub that you can use, but it's not anywhere close to as powerful as Replit. Okay, but every big company in the world uses GitHub, right? Every small company in the world uses GitHub. This is how developers work. And so the reason I'm saying this is you don't have to understand how to use it. You don't have to. But you do need to understand that if you build something in Replit, it's not stuck in Replit. You can actually push it out and give it to developers and let them work in the ways that they know how to work, right? And so you can, you can whip up something that is, you know, again, we, we've now built this, this working preview. How much better would it be to give your developers that you hire a fully functioning preview of something instead of just like, I've got this idea for this thing, but hey, I put it together, make it look like this. Here's the code, even here's the design, here's where the buttons are, you know, all that stuff. And then, and then, you know, from CEO idea to developer, you know, you don't have any loss of translation. It's actually like, hey, this is what I want to happen and here's the exact way to make that happen. Okay. All right, guys. By the way, I should say this. We're an hour and a half in. We've got 2 and a half more hours. I will sit here like the whole time. Like, we will take like a 5-minute break, maybe a 10-minute break at some point. But if you need to get up and go to the bathroom, grab a snack, this is a— this is an open class. I'm not a mean professor. You can, you can do whatever you want. So go ahead and do that if you need to. This is being recorded. So if you happen to miss something, you can always go back and be like, oh, what was that little thing? I want to watch that again. And you'll be able to go back later and pause and be like, okay, let me catch up with Jon. On. Okay, so all right, time for you guys to get busy again. Now we're going to dive in to— we're not going to get into Replit yet. Okay. And here's what I'm going to say, and this is going to be a little bit different. Well, actually, it's, it's, it's about the same as what we did with Groq, right? This is actually a bad prompt. The prompt and the process I just took you through here is— this is only 3 sentences, and then I answered a couple questions here. 3 questions. That's it. All right. So remember that chain of thought reasoning process that we talked about, okay? Where you say, hey, build me this storage vault, right? For, for, you know, when I die. Okay. And here's a couple details. Well, what it's going to have to do is it's going to have to build all these subtasks. Right? And as it's building all those subtasks, for every answer that you haven't given it, it is going to give you its answer, right? Does that make sense? It is going to fill in the blanks on its own because it's smart and it's capable, right? How many of you have ever had a conversation with ChatGPT that goes something like this? Hey, I've got this really good idea for this new marketing thing, or I've got this really good idea for church talk. I don't know, whatever, right? Or help me write this. And it's like, okay, great, there's your church talk. You're like, what do you think? Oh, this is good, tweak this. Okay, I've done this. Would you like me to add something here? Yes. Right? Would you do this? Yes. Okay. And then all of a sudden, you know, you've got this church talk or you've got this idea or whatever, and all you've done is you've asked it one question and you said yes, thanks, good, sure, you know? And that's all the input that came out of you, the human. Right? You've let AI completely take over your task of writing a talk or a speech or a post or an idea or a report or an email or whatever it is, right? It's just done all this work for you and it's taken you out of the task completely, right? That's why some people are like, oh my gosh, I'm going to be replaced by AI, I'm going to be replaced by machines, right? And we probably will be in a lot of things, and we'll talk about that later. But Humans are still far more creative and capable than AI is, right? Ryan, you have way more domain knowledge than all of AI does, right? Because you know your customers, you've been there, you've lived it, you have this experience, you can feel, you, you understand things in a way that AI just can't. And so if you're creating this new marketing initiative on how to get more customers, it, it needs to know your thoughts, right? You don't want to just you know, let AI do the work. Okay, but you got to know how to do that. Okay, so we're going to go ahead and we're going to jump into ChatGPT here now. Um, so ChatGPT, guys, is my— um, I, I love ChatGPT. It's, it's absolutely the best overall platform, um, for, you know, general tasks. Although it is not showing— there we go. I was thinking really hard. I've got lots of projects and lots of, lots of data and lots of conversations Okay, now while we're in here for a second, let's just talk about some of the things that matter. So number one, um, we've already kind of seen this and many of you probably know it. If I were to go, um, into, uh, let's say the top 5 stock picks, okay, I asked it, I said, hey, give me 5 stock picks that are really good buys right now, right? And, and I actually love doing this kind of stuff. I went in and I bought exactly what it told me to do. I just put $1,000 in, right? You know, $200 in 5 stocks, right? But I like trying this, just testing it to see like, is it right, right? And I'll go back and I'll check, and I have these different portfolios and little lists, right? And so it said, hey, go do Sunrun. It's been beat down. Bumble, Upstart, whatever. Okay, so it gave me a few of these different things. And Rivian— I actually love Rivian, so I was happy that it told me to buy that one. I have a Rivian and it's the best car I've ever owned. So anyway, that's my little Rivian plug. But if I come here and like, okay, um, How are these doing since you gave me this suggestion? Okay, so it'll go and it'll do its research. Now unfortunately, I just set off a pro research grade. This is probably gonna take like 5 minutes and I didn't mean to do that, but whatever. Here, you know what we're gonna do? We're gonna stop it and we're gonna ask it again. Copy. And we're gonna go with instant. Please do the research really quick. Searching the web, instant. Here's how we're doing. Okay, so let's see. That's how it is today. That's how they're doing now. Okay, there we go. Trading around here. I don't have the exact entry prices since I didn't lock in those as official buys. Got it. Well, it doesn't know, so I could, I could say this was on— I wonder what day this was actually on. I didn't actually check initially. I'm actually kind of wondering if I can see details for a conversation, like when I actually had the conversation. Okay, let's assume it was 30 days ago. We'll just do that because I think that's probably about right. Okay, here's where they are. Okay, so I'm down 1.38%, couldn't Locate this one down 1.4%. So a couple others are down. For some reason it's having a problem giving me that. But the bottom line is what I'm trying to show here is that when I go into a past conversation that I had, I can, um, I can just pick up where I left off, right? I don't have to like say, hey, remember we were talking about these things, right? I could just start talking to it. So it remembers the context of a conversation. Now, if I go into a new chat here and I say, what were those 5 stocks I was asking about? Okay. All right. So this is really interesting. It says September 3rd. Now, about 30 days ago, I was kind of right. But here's what's interesting. These are totally wrong. These are not the ones that it suggested to me. Okay. Are you sure those are the ones you suggested? Yeah, that's really interesting. And it's saying it again. And I'm pretty dang sure that I did not get asked about this. I remember my conversations. I remember what I talk about here. Yeah, it's totally wrong. I wanted to make sure it picked the winners. I guess. Yeah.

I mean, yeah. Rewriting history.

ChatGPT is learning quickly. Totally, totally. Now, we're not going to get too much into the technical details of this, but what I do want to say is that ChatGPT, while it says it remembers your past conversations, it doesn't. What it does remember is high-level topics, and I'm going to show you what it actually does remember. Okay, so you can click on personalization and then go into data controls. Nope, nope. I think, yeah, archived chats, delete all chats. Hold on a second. Where is the memory? It keeps changing this thing. Here's the memory. Okay. Personalization memory. Okay. So it remembers very high-level things. It remembers that I wanted to build a 60-post, but this is a past replica. I was building something, right? It remembers that I was launching a new album. I'm a pianist, right? And so, I was launching a thing for It Is Well With My Soul here. And, it knows other things about me, right? It knows that I'm building a business here. It knows that I'm fluent in Spanish and Chinese. It knows I'm a member of the church. It knows that I'm a class 5 whitewater kayaker. It knows that I founded this company. It knows that I'm CEO of this and whatever, right? It knows I'm a father of 4 children. It knows that I live in West Mountain. Okay. This might seem scary to some of you and that's okay if it does. We could talk about that in a second. But for me, I want it to know who I am. Okay? It knows who I am anyway. The internet knows where I live, that it knows these things. I'm not worried about that. Okay? But because it knows, for example, that I'm a member of the Church of Jesus Christ of Latter-day Saints, if I ask it a question that's religious, it's not going to refer to the Torah or the Koran right now. I might have a question about the Koran. Hey, can you teach me about this? What do Muslims believe here? What do Jews believe on this thing? Right. That's totally fine. Now, if it knows I'm a Jew or if it knows I'm a Muslim, then it's maybe going to more default to those types of things. And so because, because it knows what I am, I then have to teach it less and give it less context on every single question. Right. And models are getting better and better at having what they call the memory, right, where it knows who you are. And if it knows who you are, it will be able to do things better. Now, this is— let's, let's get back to business here for a second. Right. If it knows, okay, you're doing, you got $4 million, we'll use the Thackers, right? We got $4 billion AUM and we have, you know, 6,800 clients and we have, you know, this is the average size of each one. And, you know, geographically we're here, here, and here. And we have offices here and, you know, we have 115 employees and, you know, this is where we're struggling and these are the things we're working on. These are our goals for the next year and whatever, right? If it knows all that stuff, now when you start asking it questions about how do we do better, Okay. You don't have to worry, Tyson and Ryan, about that data that I just said going to the public. It won't. It won't go to the public. And not only that, you can do data controls and you can say improve the model for everyone. No, off. That means none of what you say will go out. Okay. You're not training the model. Okay. So you can give it all this amazing data and then it keeps it secure and stored encrypted on your account. And then you can ask it questions and say, all right, what are you considering all of our goals? What should we do here? Right? So that's amazing. So you can have this system-level information, you can have, um, conversation-level information. And then my favorite thing in here is projects. Man, I'm a project guy. I love it. We're going to call this Data Vaults. Okay. And so in a project, it's going to remember everything at the system level, and then it will remember all of the chats that are in here. Okay? It will be able to pull from all those chats, all that data, right? So I have, for example, a General AI Proficiency Institute project. And inside of this project, I have hundreds and hundreds of things, right? All, you know, going, going back all the way to February 27th, which is when I said, hey, I'm going to see if I can build this this thing, February 27th, right there. Okay, AI aptitude test platform. This was my original idea right here, my original chat, and boom. Okay, I want to, I want to have an AI IQ test. Help me build this, right? Let's see if we can figure this out. That was what I started Thursday morning, and now here I am months later, and I have all of this cool data inside of this. And I can add files even. I can upload code or documents, PDFs, images, whatever I want. To give it all this information so that it has the context of my business, right? So maybe you already have all kinds of things like that, or you have these Excel documents and reports or, you know, Q1 profits, whatever. You could upload that kind of stuff and it's all there and then it can reference it. Okay. So we're going to go back to my other project here. Okay. And we're going to start on Data Vault. Now what I'm going to do here is I'm actually, because I already wrote it, I'm going to take the prompt that I started with up here. And yeah, whatever. We'll just copy this whole thing here and I'm just going to, I'm going to paste it in. Help me. Yeah. All right. We'll just say let's, let's build it. Instance is fine. Okay. So AI is action-oriented. Okay. So when you tell it something like this, it's just going to be like, okay, I'm going to start building it, right? So I'm going to start writing some code or might start building some plans and all kinds of some great stuff, right? So here, I'll give you a full practical blueprint you can start building, right? Here's your goals and constraints and high-level architecture, and here's how you should do it. Okay, again, this is where AI is saying, all right, I'm smarter than you, I'm going to take over. Okay, and you have to say, okay, hold on, slow down there, bucko. We need you— I, I need you to, to work with me on this. Okay, so we're gonna, we're gonna use a process where we flip the script a little bit here. Okay. So we're going to let it just keep typing, but I'm going to, I'm going to, I'm going to pretty much ignore what it said here. It said, okay, thanks. Thanks for all of that. But I want you to help me build a plan. I want you to interview me one question at a time about this. Idea and help me build a business plan. I want you to ask me about the name of the company, the branding, the USP, the unique selling proposition, the, you know, marketing plan. Marketing plan— whoops— monetization strategy, and anything else— whoops— you need to know to help me build out a solid plan. Okay, so this is really important. I'm going to copy this, I'm going to paste in the chat so you guys can grab that as well. You'll use it in a minute. Okay, and then I'm going to click go. Okay, and so now what it's going to do here, guys, is instead of just like, again, I could have said, hey, I like your plan, let's do it, right? There's so much data that it just presented to me here, right? Practical steps. Let's do all this stuff. Let's— final notes and warnings. Here's what you got to worry about. Here's all this stuff. Okay, whatever. Right. What I really want is I want it again. I want to be the boss. I want it to pull information out of my head and then I want it to help me organize it. Right. That's important. Okay. So first, what do you want to call this company? We can explore brand names to communicate trust, security, legacy, continuity, family, or protection depending on the vibe you want. Um, family and trust are good, good values here. We want to take care of our family. We want them to be able to take care of things. So it's going to help me come up with a couple names. Cool. Family Trust Vault, Legacy Pass. Family Trust kind of has this like trust, like a legal document trust, you know, thing. And so I don't know if I want to confuse it with that. Um, Continuum is kind of cool. We're going to go with Continuum, even though I know that's, um, it's probably already a company, but it doesn't matter, right? You know what? Yeah, we're gonna go Continuum just for fun. Okay, um, let's go with Continuum. We want everything to continue. After we die. Okay, we want people to have access to our stuff. So what do you see as the unique selling proposition for Continuum? This, this, this, and this. Okay, so I don't need you guys to watch me build out my fake business plan. What I want you guys to do is to go ahead and do this yourselves. Make sure you switch to ChatGPT-5. So build a project, make sure you switch to Instant. You could certainly switch to Auto or Pro or Thinking later, right, if you wanted to really dig into your idea later. For now, for this exercise, I want you just Instant so you get fast things. Go through maybe 10 questions or so. All right. You'll start by just telling it. You're going to just say, hey, I want to build XYZ, okay? A personal brand website. I want to build this customer dashboard. I want to build this calculator that helps me do this, whatever. Okay. And then, and then just say go. It'll give you some, a bunch of stuff like this and then say, all right, slow down. Here's what I want you to do. And you'll copy, you'll paste that, you know, thing that I gave you in the chat. Okay. And so right now it's 10:49. I want you to go until 11 o'clock. Okay. Just work on this for the next 11 minutes. You have to build an entire business plan in 11 minutes. Okay. Right now all you're doing is answering questions one at a time.

Just get through as many as you can. Can I ask you a quick question really quick on ChatGPT? So ChatGPT, when you start a new chat every time. So I guess I just want to know, like, how do I learn how to do like that? Projects. I just have been using the same chat until it doesn't let me use it anymore. And then I start—

Yeah, no, you always want to start a new chat with a new topic. And, but you also want to use this projects tool, right? Just click new project.

Okay. And I like that you can put in files with those projects. So, and are you limited as to how many projects you can have?

I have like 60.

Then how do you see them all?

Just by like right there in that section? See more. And unfortunately it's going to take a minute minute to load all of this, but if I sit here for about a minute, it will, it'll show me this huge list.

So, and so how do you, you have projects and then you have chats. So what's the diff?

I guess so a chat, just think of it this way. A chat is a, or sorry, a chat is just a single topic conversation, right? A project is a collection of chats. Okay. That's it, right? That are all surrounding that one topic and the project can remember everything that's been said in those That's perfect collection.

So anyway. And then maybe not now, but later, I'd like to know like this instant and those dropdowns. I never even saw those different options before. So I, you know, like in what scenarios would you not wanna use instant every single time, right?

If you want it to do a bunch of research, right? Instant, it's gonna use different models. It's gonna think more. About the problem. It's going to consider more variations. Imagine you ask it a question and it says, okay, here's my quick 3 answers. Okay, but actually, maybe I should consider 300 answers, right? And then I'm going to narrow it down to the best 3, and then I'm going to narrow it down to the best one, and then I'm going to present that to you, and I'm going to present you with options, right? So instant doesn't mean it's faster. You want it to think. Yes, instant's going to be faster because it's going to consider it very shallowly. Is that a word? It's gonna be very shallow in its analysis, right? If you wanted to go to deep analysis, um, then you're gonna, you're gonna have it be on, on Pro or Thinking or something like that. Yeah, got it. Yeah, good questions. All right, 9 minutes, guys. Go ahead and just talk to your AI. And we'll take a break here, guys, in about 20 minutes. Okay, just a few-minute break. By the way, guys, if you— if it asks you a question— so it just asked me like, who's your initial target customer? Gave me some options, and I was like, where do you think the easiest entry point is, right? And so I can— you can ask it questions and it will still resume the, the interview, right? So Hey guys, take another maybe minute or so. Hey guys, so go ahead and finish up. You probably got through, you know, 5, 10, 15 questions, maybe only 4 questions. Maybe you've only done a few so far, but that's okay. Now, if you're building a real business, you know, and you might be right now, and so I understand, but for this, for this exercise, we're going to kind of end this exercise right here. You can go back and do this later. I do this every day. Every day I will have an AI conversation about like, hey, I need to work on this thing. Can you interview me one question at a time so we can really get, you know, figure out how to say this and figure out how to, how to, you know, approach this problem or whatever it is, right? You know, I'll give it, you know, 40 pages of legal documents. I'll say, hey, I need you to, I need you to walk me through this. And, you know, one page at a time, help me, you know, ask me things you need to know so that you can give me good advice on this, right? So all kinds of ways to use this interview me one question at a time tactic. Okay. But from here, we're now just going to say, let's, you know, when I built GenAI API, by the way, I spent easily 3 hours on this step. Just FYI, I spent a long time working through, thinking through. Okay. Strategy matters. Okay. All right. So please build me a 2,000-word business plan, um, taking into account everything we've discussed. Use the Canvas tool for this copy. Okay, so I'm going to give you guys this in the chat as well. Go ahead and do it on yours. Okay, so what it's going to do here is it's going to go ahead and synthesize. Now it's going to organize all of the different questions and things that you, you know, that you went through and answered And then it's gonna build it in what's called the canvas here in ChatGPT. And canvas is just like a Google Doc or a Word doc or whatever, or a Pages doc if you're one of those weirdos that uses Pages. But it will, you know, allows you to then go and copy and you can even edit this document or download the document. But it builds this document that's inside of your, inside of your chat, right? That you can, you can mess with. Okay, super cool. Built me a 2,000-word business plan and, and I've got a really nice thing. Now I can go in if I want to, I can edit it. And but, but like, it built me this beautiful executive summary, right? This is a beautiful— this is a secure, family-focused digital inheritance platform that ensures loved ones can access critical accounts, information, and instructions when the account holder passes away, right? Here's the problem and opportunity. Here's the solution. Here's the market and target audience. Here's the USP and competitive advantage. Here's the business model, marketing strategy, product roadmap, vision and brand positioning. All these different things, guys. Here's, here's what's critical, by the way. I don't need it to say word count at the end of this, so I'm just going to delete that. That's not important. But here's the cool thing. In a minute, we're going to take this and a couple other things. We're not done yet, and we're going to give this to Replit. Right? So now instead of giving it a 3-sentence prompt, we're giving it a 2,000-word prompt. And now when it has to go in with its chain of thought reasoning and fill in the gaps, it's going to fill in the gaps with our thoughts, with our strategies, with our ideas, with our understanding of the market that we're going after. Okay, and so when it has to answer a question, it says, oh, that was answered, let me just drop that in. Oh, that was answered, let me drop that in. Right, and so this is a much, much more complete, um, prompt is what we'll say. Okay, this is a better thing. Like, again, back to the developer. If you were to give a developer this business plan instead of the 3-sentence prompt, that developer is going to have way more answers and they're going to make it much closer to what you are trying to figure out and build there. Okay. All right. So next step, we're not done with this. Okay. This is a good business plan. We're going to see if we can make it any better. Okay. So what we're going to do here is we're going to copy this business plan. Okay. So go ahead and click copy on your business plan and then we're going to come over to Groq. And we're going to start a new conversation. Okay. And we're going to say, please review this business plan. Search X for posts related to this. Find competitors and do a complete analysis on how I can make this business plan stronger. Don't focus so much on the technical details of the business plan document itself, but focus on the ideas and the positioning and go-to-market strategies. Okay, so then I'm going to go Shift+Return, Return, right? And you, when you, if you hold Shift while you do it, then it won't just send the thing off. And I'm going to paste my entire business plan. Okay, so here in Grok, so I've copied that thing from ChatGPT. I'm going to come over to Grok, a different model that thinks differently, that searches differently, that can, that has other, you know, capabilities. Okay, and I'm actually going to switch it to experts. Put that intro back into the, uh, yeah, yeah, I will give you this.

And chat, it's funny, Jon, I was, as I was answering questions in ChatGPT about, I, I asked, I said, before we continue, will you share with me competitors to this plant? Like, what are the competitors out there?

So it's kind of interesting you're already thinking there. That's excellent. I love it. Very, very good. So yeah, So paste this little prompt that I gave you and then paste your business plan and then click go. Right. And then we'll sit here and we will let it think. Now, while it is thinking, because it might take 2 minutes, might take 5 minutes, might take 7 minutes, most of them probably be done pretty fast here. Okay. We're going to take a 10-minute break. Okay. It's 11:05. See you at 11:15. Okay. Awesome. Okay. See you guys in a second. I'm just going to turn off my audio and video, but Um, but we'll still be here, okay? All right guys, welcome back. It's, uh, raining pretty hard where I am. What's it like where you guys are? Storm's coming down.

Not yet.

Been in and out, but for sure it's been pretty, pretty heavy up here too. Yeah, I get crazy. I get crazy wind where I live. It was really, really windy a minute ago and then all of a sudden the rain's here. So anyway, it looks like it's coming from the north.

So yeah, rain, rain in Alpine is, is pretty hard.

Yeah. I loved Alpine weather. It's like this own little climate, you know, it's kind of tucked up in the COVID there and it's just nice. I love Alpine. It's a beautiful spot.

Okay. Because you needed to find some farming area, kind of. Yeah, we can't.

Yeah. So we were in Traverse Mountain for the 5 years prior to being here. And, you know, it's nice up there and we had a good house, but it was a little cramped for my wife and she wanted to spread out a little bit. She grew up in Eastern Europe and anyway, so she, she had a little bit of a desire to go back to her roots a little bit. She actually grew up in an orphanage until she was 13. Wow. And they, you know, they farmed. And, you know, when she was 5 and younger, before she was in an orphanage, they, you know, they grew everything and all that. And then her parents, you know, beat her, and her dad tried to kill her mom and all kinds of crazy stuff. So, so they ended up in an orphanage. Wow. To bring the mood down a little bit. But anyway, so my wife's got a really, really interesting story. She ran away from home with her 4 siblings when she was 6 years old, actually. Amazing. Wow. And she's a survivor. She's amazing. So, okay, so, all right, so maybe you guys have been reading your stuff a little bit. I haven't read mine yet, but basically what happens is, you know, we did this, we did this little kind of, hey, let's look at the market analysis and I can look here. It looks like there's a couple other companies out there that are working on this and trying to do this. And how am I going to be better? How are we going to do that? Right. So how can we strengthen the core ideas? Here's the idea of a secure digital vault, context, instructions, right? Incorporate some other things. Here's how we can refine the positioning, optimize go-to-market strategies, whatever. Right. So now I can copy this, right? Copy this text. And I'm going to come back to ChatGPT and I'm going to say, please incorporate this feedback and create a new and improved business plan canvas, taking everything into account, but remembering my— whoops— remembering my answers as the most important. Okay. All right. So put that in the chat. I will. Yep, that, that will go in the chat. So then I'm just going to paste this, that whole overview, everything that it did there. That's great. So I'm going to send that. I will give you guys the little— I'll give you guys the little thing. Where'd it go?

Would you say the primary reason for putting this through is just having access to like what is Live Sentiment and X-Access kind of, or just additional data kind of a thing?

Yeah, it's more data, and I'm, you know, I'm bouncing my, my plan off of you know, more than one LLM. That's the idea here, right? I do this all the time with all kinds of things, right? Even if I'm just asking a question, right? Like maybe I'm worried about some political bias or something like that. Like, you know, what really happened with 9/11? I don't know. You know, you could probably ask that question to Grok and you could ask that question to ChatGPT and you could ask that question to Claude and you're going to get different answers, different research or whatever, right? If you wanted to go in and, you know, be a conspiracy theorist. But even if it's not conspiracy theorist type stuff, right? Or it might even not be politically motivated. Maybe you just want to come up with new ideas on something and you want to ask several LLMs, hey, what do you think? What do you think? What do you think? And then you can take the feedback and then bounce them off of each other. Hey, this is what ChatGPT said. This is what Grok said. This is what this said, right? And what do you think? How can we make this better? And so by just bouncing it off of multiple people, right, in this case LLMs, you're going to get a more complete analysis. Right. And so, so I've now, you know, in, in just a few minutes, basically, you know, we took 10, 15 minutes to build the, to go through the interview process. And then if you were just doing this fast, it would take you another maybe 3 or 4 minutes to copy over to Groq, get some feedback, copy back over to ChatGPT. And so in 15 or 20 minutes, you've now gotten some really good feedback. You've got a continuum business plan or my business plan here, new and improved, a better executive summary, a better definition of the problem, the opportunity, the solution, the market, you know, competitors, things like that. It's all here now. Okay. So we want to try and beat all of these people out there. Okay. All right. Very cool. So let's, let's do one last thing and then we're going to jump into Replit. Okay. So let's— so guys, after you've done— you've got your new and improved business plan. We're going to say, I am ready to take this into Replit to start building. Please build another canvas entitled Feature Lists. This should be a comprehensive list of all front-end and back-end features needed to build this product. Do not specify which technologies need to be used for each, except for login, which should use a standard JWT. Auth system. Okay, don't worry about what that means right now, but use it. Okay, which should be— you should use the same GPT Auth system. Okay, that's all I need. Okay, so I'll have it build this and it's going to build me a new canvas, a new document here. Okay, I'll give you guys the prompt there in the chat. And now I've got a feature list. Here's every single thing that needs to be built. Okay, front-end, back-end settings, everything that has to be done. JWT authentication system. Very cool. Yeah, global compliance, GDPR compliance, data exports, SOC 2, GDPR, CCPA. Yeah, mine's complex. That's why I'm not actually going to build this, probably. I mean, it's a cool idea, but man, there's other systems out there that are doing things like this, and I don't need to compete with them, and I have too many things I'm doing already. But it's still fun, it's still fun to build. Okay, cool. So I'll give you guys 1 minute to finish that task. It should be already done, right? But, but basically what we're going to do from here, guys, is we're going to now build a Replit prompt. Okay. So we're going to open up Replit and we're going to go to home. Okay. Where it just says, hey, what do you want to make? Right. And we're going to then bring this stuff in. So I want to say first over here in ChatGPT, I'm going to say give me a 3-sentence description of what we are building. Okay, I'll give you guys that prompt. Super easy, but there it is, it's in the chat. Okay, so Continuum is a secure digital inheritance and continuity platform that helps families and entrepreneurs pass on accounts, crypto assets, and important instructions with ease and peace of mind. It combines bank-grade security with an Apple-like family-friendly design, ensuring loved ones not only receive access but also context they need to act confidently. Unlike password managers or estate lawyers, Continuum is a warm, human-centered guide that transforms digital chaos into clarity and continuity. All right, pretty good. So I like that. I'm going to take my 3-sentence thing and I'm going to come over here. I'm going to say— I'm going to paste it into this spot right here. Or maybe I am. Come on, let me copy, copy and paste. Why is it not pasting? Okay, we're going to click this copy button. All right, I don't know why it was causing me problems there. Okay, then we're going to— so don't click start chat yet. Okay, you can hit Shift+Return, Return, or Shift+Enter, right? Because we're going to add a little bit more. I'm going to say, please follow the attached guidelines closely. Okay. I'll give you guys that little thing you're going to add to that. So you're gonna have your 3 sentences and then please follow the attached guidelines closely. Then we're going to come in here and we're going to copy the feature list and we're going to paste it into Replit. Then you're going to grab your new and improved business plan and you're going to paste it.

Okay, so you're both the list and the business plan.

Yep, you want the business plan and the list. Okay, so your prompt in Replit should look like mine does. You got 3 sentences. Man, mine is copy— I wrote it all with markdown. It's really dumb. Not that it matters that much, but just Cleaning mine up. You don't actually have to do this, but I'm just doing it while I wait for you guys. But yeah, 3 sentences, then this 1 sentence, please follow the attached guidelines closely. And then you've got the business plan and your feature list. Okay.

As soon as my feature list pasted is a little box, is that perfect?

That's exactly what you want. Yeah. The feature list and the business plan will look like little attachments basically.

Where did you get the feature list?

Sorry, I was working on the— Yeah, me too.

Yeah, so the feature list, go into ChatGPT and give it the prompt that I gave you before, the one that says I am ready to take this into Replit to start building. Okay, the chat there, just copy that, paste it, and it will give you a new canvas document really quick. So we'll wait, I'll just wait here for a second and while You guys look that up. I'm going to look something else up that we're going to do while we—

and after we do that, then we can just go ahead and hit start chat.

Yeah, yeah, yeah, yeah. Once you're done with that, go ahead and click, click start. That's totally fine.

So I already ran that 3 description, 3 sentences description stuff, but I didn't put, hey, I'm ready to put this into Replit.

What was the kind of the process? That was for the feature list, Danny. Okay.

Did you do the feature list? I did all of this stuff. Just was my last thing I put in was la la la.

Yeah. Yeah. You should be good if you did it in just like the business, the business, like final business version. You'd have the business plan and then you have a separate document called feature list. Did you get that?

I didn't do the feature list, I guess. And so bless my heart, how do I do that if I did not do that?

Just go into the chat. In the, the, the Zoom chat here and grab the 'I am ready to take this into Replit, start building.' Oh, bless my heart, there it is. Just copy that one and tell ChatGPT and then grab it, right? You can give it to Replit later, it's totally fine. No big deal. Yeah, there we go.

Awesome. Now mine says I've created a feature list based on your requests. If everything looks good, we can start creating. Do you want to build the entire app or start with design?

Just say no. You want to just copy that feature list and take it over to Replit.

Yeah, I did. I'm in Replit now. Okay. After I put the feature list and the business plan in that first question. The 3 questions, it's now saying I've created a feature list based on your request. If everything looks good, we can start creating.

How do you want to continue? Yeah, just click continue.

Which one?

Build the entire app or start with— Go ahead and do build the entire app. Wait. Yeah. Yeah, let it go. Let it loose.

Hey, question for you. When I'm not getting like those little boxes, am I just doing the copy text thing, or am I doing like a send share conversation?

I'm bringing it over to Replit. You should copy the text. Yeah, you should copy the text from the document.

So if I copy that text and I put it over here and I'm okay, there it went.

So we're good. All you had to do was ask me about it. And then since I was looking, it worked. Yeah, I did the actual text, but now it has a little box, so we're good. Yeah, awesome. So yeah, you should have two kind of attachment boxes plus your three-sentence description and then go. Plus the please follow the attached guidelines closely, right? That— yeah, you wanted to make sure it has that piece. Man, why is mine having so much trouble?

Weird.

So your Google, your Chrome browser is out of date. I don't know, man. Yeah. Relaunch to update. I'm sure it's something to do with that.

So I have a brand new— I swear that happens to me. It's always Chrome.

I have a new laptop sitting there in a box that I need to open and I just haven't done it yet. Nice. This guy's only about 4 years old, but And it's an M1, like it's a good fast computer and it really is quite nice. 64 gigs of RAM, like it's good. But, um, I don't know, I've worked it pretty hard. I probably just need to update it. All right guys, so, um, you're all now building cool stuff. Okay, does anybody not have their— is Replit not building yet for anyone? Are you guys all rolling? Okay. All right. I love it. So while we're sitting here waiting, we're going to watch a movie, kind of. It's like a 2-minute clip. So this is Eric Schmidt. He is the former CEO or chairman of Google. Super, super rich billionaire and pretty smart and very plugged in. Now, this video was— this little interview or whatever it was, he's on stage talking here. This was on April 19th of this year, so this information is almost 6 months out of date, but still we're going to listen to it and we're going to talk about it. We're going to have— this is, this is our group discussion part of the course. So here we go.

We believe as an industry that in the next 1 year, the vast majority of programmers will be replaced by AI programmers. We also believe that within 1 year, you will have graduate-level mathematicians that are at the tippy top of graduate math programs. So that's 1 year. Okay. What happens in 2 years? Well, I've just told you about reasoning and I've told you about programming and I told you about math. Programming plus math are the basis of sort of our whole digital world. So the evidence and the claims from the research groups in OpenAI and Anthropic and so forth is that they're now somewhere around 10% or 20% of the code that they're developing in their research programs is being generated by the computer. That's called recursive self-improvement is the technical term. So, what happens when this thing starts to scale? Well, a lot. One way to say this is that within 3 to 5 years, we'll have what is called general intelligence, AGI. Which can be defined as a system that is as smart as the smartest mathematician, physicist, you know, artist, writer, thinker, politician. I call this, by the way, the San Francisco Consensus, because everyone who believes this is in San Francisco. It may be the water. What happens when every single one of us has the equivalent of the smartest human on every problem in our pocket? But the reason I want to make the point here is that in the next year or two, this foundation is being locked in and we're not going to stop it. It gets much more interesting after that because remember, the computers are now doing self-improvement. They're learning how to plan and they don't have to listen to us anymore. We call that superintelligence or ASI, artificial superintelligence. And this is the theory that there will be computers that are smarter than the sum of humans. The San Francisco consensus is this occurs within 6 years, just based on scaling. This path is not understood in our society. There's no language for what happens with the arrival of this. That's why it's underhyped. People do not understand what happens when you have intelligence at this level, which is largely free.

All right, guys, what do you think?

Believable. It's democratizing everything, right? Like leveling the playing field in every country, every place, right? If anybody has access to a smartphone, or which most countries do, right, you're going to have access to the smartest, most capable computer. And then everybody has equal access in theory, right, to this machine or this capability. And so the potential is that everybody— again, leveling the playing field is the best way I can think about it, but like across nations, across ethnicities and financial socioeconomic positions, right, in the world. Like a lot of that starts to change and it already has. We've seen it. Technology has changed those things, access to technology.

But I'm going to push back on you, Ben, a little bit. I'd love to hear it. For you and for me, I agree completely with what you're saying. Why? Because we're entrepreneurs. That are highly motivated, highly intelligent, very capable. We're wealthy enough to have access to everything we need to in order to basically bring anything to life we can imagine. Okay. What about for, you know, the 190 people in 2 out of 200, you know, call center people that just got replaced by an AI agent that can do their job way better for a tenth of the price? What are those 190 customer service agents that, you know, needed that to support their family do? UBI. Yeah, it's, it's absolutely an option that has to be considered. I generally am against UBI, right? Just from a political standpoint. Right. And I have my reasons for that. We don't need to really get into that, but I don't know another option. Right. It's hard for me to think of a better option than UBI. It's really, really tough. And you're talking about universal basic income. Yeah, right. Yeah. Yeah. What does it mean for society, right? For individuals and entrepreneurs and everybody on this call right now? This is awesome, right? I heard everybody, almost everybody say this is awesome. Like, so cool. I can't believe we're learning this and this is capable now. And now, you know, I can automate all the stuff we don't want to— we don't want to have to do. I don't have to hire developers anymore. Okay, well, what if you're a developer? Developer, right? Is that actually affecting developers today? Well, let me say this. I raised $13 million and hired literally 50 developers for my startup 2 startups ago. For my last one that I raised $3 million for, I, I had 6 developers on staff. For this one that I am now cranking with, I have zero developers on staff. Is that affecting developers? Absolutely it is. What do you do if you're a developer and all of a sudden AI can do your job? What do you do if you're a paralegal? All of a sudden AI can do your job, right? Plumbers are still safe for a while. Robots aren't quite dexterous enough to, you know, do that, but eventually they will be. You can give it 10 years, right? They will absolutely be able to come in and switch a toilet out, right? Um, today they're okay. So what do you guys think? What does this mean? What does this mean for society in general? What does it mean for you as entrepreneurs? Why is it good? Why is it bad? All of the above.

So, Ben, I'm not saying it's time to make changes. Oh, yeah. Oh, go ahead. I just mentioned it's— timing is important. Speed of adoption is important. This proves that people are all thinking about it and, you know, get in on it before the next guy does.

Kind of what it's just— it's cruising.

Yeah. You know, when Truman was in Dallas, he showed a video of this guy who was talking. I'll have to see where that— to get that video for tomorrow. But he was sharing about how every— this guy in the video— of every 22-year-old that's just graduating from college should be really excited because of AI and how they can create businesses and, you know, right, so inexpensively, right? And, and right from the start of their career. And what you just said kind of Yes, it's great for all of us entrepreneurs, but to help educate people that are losing their jobs because of AI and how they can leverage AI to, you know, to do something on their own is simply as we're all going to do or you've done or these 22-year-olds can do. And there's a lot of people out there that are afraid of the AI. They're afraid of something new, right? Like, oh, they're scary, so they don't want to even try it. It's kind of like the internet. When the internet first came out or cell phones first came out, people didn't just embrace it and try to learn it and try to go with it. They were just like, you know, just so afraid of it. And so I guess one question for you, Jon, this is, this is awesome. I've already learned so much and I know tomorrow is going to be even better. But for, for us taking this class, Do you, do you suggest us taking it over and over? Do you have a continuous learning program that we can do? Because I know like I'm gonna want more of this, but not only for myself, but I wanna be part of that process of helping educate some of those people that with a little bit of handholding could embrace AI and find out what they can do for themselves in this huge world of change that we're going to be experiencing?

Yeah, it's a good question, Laura. And it's certainly one that I've had a lot of people ask. You know, when I started GenAIPI, this course was not on the— it wasn't in my strategy document. This was kind of after the fact. People said, hey, you built this business and now you have $200,000 revenue so fast. And can you teach us how to do that, please? Like, I want to know how to do that. And I was like, sure. And I just kind of started doing some courses. Then I was like, this is kind of an AI course, the business, whatever, I'll just throw it on the website. And so it just kind of became part of the business. And so my goal wasn't to be like this course guru that, you know, has the next business and now you can join my inner circle and, you know, all that kind of stuff. And I just wanted to provide this class and just give it as a taking off point. But I've had lots of people say, hey, will you do an advanced course? And And I now have enough people where I definitely can do an advanced course. I actually sent an email out a couple of weeks ago and got a bunch of people said, yeah, I'd love to do an advanced course. So yeah, that would mean at least one next level course, right? Where you're just kind of, we're pushing farther, we're pushing deeper, we're talking about these things in more depth. But the reality is AI changes so much every month, right? Like it's crazy, right? If you go a year without looking at updates and without being up to date, today, like, you are, you are 10 years behind, right? As compared to what it used to be. The speed of improvement is mind-boggling. It's hard, you know, I'm in it every day and it's hard for me to keep up with everything. It really is like, it's a full-time job to keep up with it. And, um, so I, I offer what you're saying, Laura, to businesses, right? Um, you know, like, so for like Ryan and, you know, Tyson, like, hey, you want me to work with you guys and, you know, be your fractional chief AI officer, great, let's do it, let's talk, and I can help get this across your whole company. I don't really have the equivalent of that offering for individuals, and I'm trying to figure that out. I really am active.

Well, do you, do you do—

how often do you do this class?

Once a month at this point. So, because I, I would even like to even think about taking it once a month just because I appreciate the hand-holding. Okay, I'm going to show you how to— because I don't think I would have been able to do what I've already done today had it not been, okay, do this next and do that, you know.

I mean, that, that hand workflow. Yeah. I'm also working, um, I'm also more working on something that might, might be even better, right? Obviously come to the course over and over. I'll give you discount codes. You can come cheaper every time, right? So, but, but bottom line is, um, I'm working also on some prerecorded recorded modules and things like that, right? Where you'd be able to say, okay, I'm going to go and I'm going to learn about this topic. Or I'm going to, you know, tomorrow we're going to do fun stuff. We're going to do images. We're going to do video stuff. We're going to do marketing and sales. And we're going to do all kinds of fun things that focus on other elements and just introduce you to other, you know, we're going to do voice AI. We're going to do some API stuff. We're going to do all kinds of really fun things. And so I am coming out with more content. There is also, I will say this, there is a community on Facebook. That is everybody who has taken this course and it's decently active, right? And so there's about 100 people in it right now. You know, I've had several hundred people take this course and about a third or so have joined that. But the people that are joining it are like, hey, did you see this thing came out? Or, hey, I'm working on my project, I can't figure this out. And like 10 people will jump in and say, here's how you do it, here's how I fixed it, you know. So there's a good support community that's growing that again, I kind of just set it up because I was like, hey, it'd be great to have a place where people can ask questions.. And so I've used that.

I've just used a Facebook group for that. So here, here's a business idea, John, for you. First of all, I'd like to do the class and the videos and the Facebook community. Yeah. But I know, you know, John Maxwell, right? He's a leadership coach. Yeah, of course. And I'm actually a certified John Maxwell speaker and coach with his program. But one day I went to his first, he just did this kind of a Super Saturday kind of a thing. And he said, started doing a coaching program and a, you know, coaching and a training program out of his content. So something where you may not be able to do classes one-on-one, but you could, you know, have people that are trainers like that to kind of extend your reach to get the training content. You create the content, and I think you could replicate that.

Absolutely. And so there is actually something happening behind the scenes that I can share with you guys at this point. Um, it will be officially public next week, but a few months ago, uh, I was really taking off, and I, I'm pretty good at making noise about my stuff, and that's part of why it's taken off a lot. And, um, anyway, Tony Robbins found me, and, and he said, hey, uh, I and Dean Graziosi are putting together a, you know, an AI, you know, curriculum. And we're going to have, you know, our— we'll have our events, you know, and we'll have, you know, we're writing a book and we're going to have all these courses and workshops and different things. Right. And they said, we are— we have— we're great at business. We have a huge, you know, massive community, global community. And so we can get the word out. But We don't know AI like you do, Jon, and you have built this amazing program with GenAIPI in such a fast time. We want everyone to know what you know. Can you help us make sure we're doing our curriculum right and everything, right? And so that launches, well, the first event, the kickoff event is November 6th, 7th, and 8th, but marketing for that event begins on October 6th next week. And so I'll be able to talk that, I'll announce that on there. But yeah, it's been really awesome, right? I'm actually a co-owner with Tony and Dean on that. Congratulations. So, pretty cool, pretty cool that, you know, I turned it into a $10 million company and that I got to get connected with these amazing, amazing people that I can now just text about whatever I need to text about. It's kind of fun. It's kind of fun to be chatting with those guys. They're amazing. But anyway, so, That, that might be also some of the solution that could help you, right? Is, you know, attend those events. They're going to be really, really great. I'll be on stage with Tony and Dean for those events. It'll be really cool. And it'll be broadcast. It's actually a free event. November 6th, 7th, and 8th is all free.

And then yeah, I'll be in Cancun, unfortunately.

Well, I'm sure we, I think if you upgrade to VIP, which is I think a dollar literally, then you, you can get a recording of it. So anyway, a dollar for your class or no, it's a dollar for their, for their event that they're going to do there, the November 6th, 7th, and 8th, right? Well, you're in Cancun. We'll be working hard and you can come back and watch it later. Okay. So anyway, but yeah, anyway, guys, you know, AI is changing the world extremely fast and, and, and Laura is asking the right questions here, right? And I know that felt like just a little bit of a side conversation, but really it's important. How do you, how do you learn it? How do you position yourself properly? And then how do you keep up with it? Right? That's, that's the, that's the, you know, million-dollar question is how do you say, okay, we're here now and we're good, right? When you guys come out of this course, you're going to be, you know, compared to the general public, you're going to be like way, way up here. Okay. You're way ahead. And with that knowledge gap, you're going to look like a Superman, a superwoman, right? For, for the things that you can do. Right. And some of you are probably looking at your, at your Replit screen right now and saying, wow, I just built this. Like, this is crazy. Right. And, and that's awesome. So anyway, yeah, mine is still, mine is still working. So hopefully yours is further along. It can take a while to build all this stuff out. And so that's okay. We have other things we work on while we, while we wait for Replit. But, but it's good to think about these things, guys. We're going to talk more about this. We'll have another kind of group discussion session or section of the course tomorrow that we'll probably kick off with tomorrow. Called Surf, Don't Swim. It's a really fun discussion that you want to really think about.

Okay. Any other quick question? Were we supposed to run the build the entire app thing?

Forgive me. Yeah, yeah, yeah.

Go ahead. Bless my heart.

All right. Start building. Yeah, you're good. Yeah. You know what? I just looked at mine and I just clicked it too. So you're right even with me, Danny. I was too enthralled in my amazing discussion with you guys. So. But yeah, any more comments or questions before we move on from that?

Anyone didn't get to say what they wanted to say. My, my app crashed and said it had some bugs, so it said click to debug. So I clicked it and it now says an agent is working on it. Got it.

So am I good just to keep watching that? Share your screen really quick, Ryan. Just want to double check. Yeah, occasionally that happens, guys. And it's usually— yeah, it's usually kind of right at the beginning. It's going to— but before it, you know, builds everything, right? Yeah, it's right here.

The app failed to run. Please debug it.

I can fix it. Perfect. Okay, so, but, um, while you're sharing your screen, we're gonna let— we're gonna use your thing as a little, uh, show and tell. So right there it says autonomy levels down in the bottom left, right? And you can— yeah, you can just push X out of that, right? But let's click on the little button that it's showing there, the little settings button. Yeah, down there, the little kind of setting over to the right. Okay, right side. Sorry, not that one.

Yeah, that one right there. Yeah, yeah.

So agent tools. There you go. Okay. So right now it's set on autonomy level high. Now I want you guys to know the way that you're paying for this. Okay. Now, um, each of you have in your first month, you'll have $25 of credits. And I think because you used my, my, uh, you know, referral code, I think maybe you get an extra $10. So awesome. You got $35 of credits. Those credits are used as the agent works. Okay. Remember I said that I built GenAIPI for a about $140, right? Which means it went through my month and then I had to pay another $90 or so to get that thing up and running. And so the autonomy level, you know, why would you not always have it at max? Hey, there you go, Ryan, you're ready to rock. Okay. But and we'll look at that in a minute. But keep this open for just a second. The autonomy level basically says, okay, I'm really smart. I know what I'm doing. You're probably going to have autonomy level at low. Which means you're going to say, do this. Okay, now do this. Now do this. You can put it all the way to max. Now max will just be a little bit more expensive. It'll use more credits, but max can run for 200 minutes on its own working and writing, right? And at the speed that AI is able to think and do, it can write in that 200 minutes, it can write 120,000 lines of code, an entire enterprise-level app. I will bounce between max, high, medium, and low based on what I'm doing. And that is something you're just going to figure out over time. Okay. As you're like, okay, I need to do just like a quick, precise change. I don't need to change the text on this thing. You're going to have it on low, right? But if you're like, hey, I need you to build a whole new system that, you know, takes these leads. It looks like you have a lead system, right? But it takes these leads and builds an automation that automatically kicks it out to an email that then does this. And then, you know, whatever. Whatever. Anyway, it has all these multi-step things. I might kick it over to Max and say, I want you to build everything necessary for this. Go. But you can see kind of what it does, right? Plans and completes new work independently. And you can see the different checkboxes of what it does if you click the different levels of autonomy. Right now, while we're working on things, yeah, so if you click on that, it can work for a long time, right? But click on medium, right? Medium is probably where you want to hang out for now.

And how do I change that if it's still in the working, if it's still thinking kind of?

You just wait.

Yeah, you just wait.

Just wait. As soon as, yeah, as soon as you get to here to this screen, that's when you can start playing with these tools. But yeah, why don't you go ahead and show us your cool new site since we're, since we're looking at your thing, Ryan. Let's see what, see what it looks like. So tell us what it's about and you get to check it out and show it to us while you discover it.

Yeah, so it, one of our constraints that we have is is advisors as we grow. And so, uh, want to build an advisor system. This is the Elite Advisor, uh, ecosystem here. Qualified leads plus coaching, uh, that goes. So, um, this is interesting. It's built out an AUM by advisor, average ROI 30 days, um, 3 different levels of coaching that can happen with that qualified lead delivery. Elite Performance or Performance Analytics. Here's what the different levels are: $1,000 a month, $6,000 a month, or $15,000 a year. On the Mastermind side, there's your full registration. That's pretty cool.

Yeah, to submit the application, whatever. It's, it's pretty cool, right? Well, now look, this is a first version, right? Um, you know, you've built all this in, uh, in, in basically I would say an hour of work with all the strategy work you put into it and then sitting around waiting for it, letting it do its thing and maybe fixing a bug or two. It's working now, right? It's doing some pretty dang cool stuff. And you can see right now it's working on, on adding actually a database. Um, and you can see at the top of that thing right there where the chat, it says 0 out of 6. You see that right now? Yeah. So that it's working on 6 tasks right now and it hasn't even completed step number 1. So It's going to be working on this for a minute. You know, Replit used to not be able to work this long. When I first started teaching Replit, it could work for 2 minutes on its own. That's it. And V2 changed that to 20 minutes, and now V3 is at 200 minutes, right? And so sometimes you'll ask it to do something, you're just like, well, I guess I'm going to do something else for a while. I'm going to go make some chili and turn on an episode of The Office or whatever you want to do. Really, you're going to probably multitask and go to other tabs and do do other cool stuff while this is just building in the background. Awesome. Very cool. We can stop sharing there. Perfect. Was there a question? It sounded like someone maybe was going to ask a question. I am getting pretty far. It looks like my preview will be available soon for me, but it's not quite there yet. Laura, will you share your screen really quick? I just want to double-check. Check something on yours. Just make sure you're in the right stage of the process.

Sure you are, but just want to double-check.

You might have taken a break. Yeah, Laura may have stepped away for a second. Either— uh, that's right, Jon.

Somehow I missed the, the sign-up stuff for Replit. I'm in, obviously, but I did the free account. What you said, there was some like credit or something that you can get for the—

yeah, it might be too late. It might be too late. You might not be able to use a code. Oh, that's fine. Just sign up.

It's $10 and, you know, it's what it is. You do the, you do the middle option though? The core.

Core. $25 per month. Yeah. Yep. Core will work. Yeah. Perfect.

Awesome.

And, and by the way, Ben, just a side note, we will have GenAIPI Replit accounts, but I recommend you have your own personal account on the side that you can use for other, you can use for your own personal projects. I don't know about you, but I have tons of personal projects that I'm just wanting to work on. I want to build something really quick and throw something up. I mean, heck, you can do it.

You can literally show it for this one. I'm excited. I've wanted to do this for like 8 years and it's like, what are you building?

What are you building?

Show us what you're building. Well, it's not, maybe they're not ready to show you, but I can show you the page.

I mean, like I can, it's already, yeah, let's see. So I like doing a little show and tell on this at this stage.

I had this idea a long time ago of just a home maintenance app as a homeowner. Homeowner that I'm not a Mr. Fix-It kind of guy, and I lose track of things. I just— my dad's really good at it. I was terrible at it. So there were things I just didn't know, like, hey, you need to change your air filter, you need to do these things for the house, you need to— and I, I had a business idea around like smoke detectors and, and replacing batteries and replacing outdated and broken detectors, and I just, I never fully did it. I started it, but I never got there. Yeah, this basically just created me an app that is, uh, it's— I called it Honeydew. I thought that was funny. It's great. It was one of the cool ideas, but friendly home maintenance assistance that's helping you prevent small tasks from becoming costly problems.

Like, oh, that was a pretty cool thing. Isn't that cool right now, by the way, as you watch on the left over there how, how it's, it's working and testing it for you? Yeah, that is brand new as of like 2 weeks ago, guys.

I'm shocked at how fast it is, how quickly it's iterating. And I just can't believe that seriously within, you know, a couple of hours, I'll have an app that if I felt like I wanted to, I could probably just go launch in the App Store and be ready to roll. I'm sure there's a little bit more testing involved, but like, there's a little bit more to that.

And, you know, we won't go in this course into all the details you need to know about how to actually launch something in the App Store. But if you have that question, ask it in the Facebook group or, or just ask ask me. I mean, obviously you've got good access to me, Ben, but, um, there's also great YouTube videos out there that will take you through. You just say Replit to App Store and there will be several videos that show you how to do it. You'll have to use another system, but I've heard— I haven't tried it, but I've heard that it takes about 20 minutes in terms of the technical process. You also— one of the things you have to do if you are publishing an app to the App Store, you have to have a valid business. You got to have an EIN. You have to get all that. You have to have a DUNS number, right? And so a Dun Bradstreet number. And so if you don't have those kind of— there's a few setup things you're going to have to have legally if you do want to actually push an app to the App Store.

But assuming you have those things in place— The app is ready to deploy whether or not I have all of the other, yeah, like you said, administrative tasks that need to be completed first. But yeah, it's crazy. I'm just—

I'm so shocked.

I'm not a coder.

I'm not a builder, but it's awesome. Anybody else want to show what they have so far? Jamie, let's see. Yeah, looking at other people's apps, guys, is really helpful because then we talk about little features and things and questions, and it just kind of helps you get comfortable around what's going on here.

Do—

did that come up? It does.

Yeah, we can see it. Looks awesome. So basically be a structure management, services-driven deployment of AI clusters and things like that. One of my last— at DDN, I was working with XAI and deploying 250-node clusters. Yeah. So I— what I'd like to do is build a services-driven organization to go help some of these smaller businesses, even some enterprises, and help them kind of dial in I love it.

So cool.

I could use that too, Jamie. Excellent. I could use a discount code to your peptides.

I take peptides daily.

Let's barter.

Yeah.

And Jon, this, when I did this, it automatically asked me for private and public keys for Stripe. Yeah, perfect. So I went and got those, put them in there. So I Thank you.

This is excellent. All right. So Jamie is stepping ahead to the next phase of what we're going to talk about here, which is APIs. We're going to talk a little bit about just some technical stuff for a minute. Okay. And by the way, guys, I know a couple of you want to show what you have. I want to see it. We're going to take a break from show and tell for a second and we'll come back. I do want to keep seeing it. Okay. Now, guys, the way— what you're going to do from here on out with Replit, and we're not going to work in Replit a ton tomorrow. We work, we work some.. But today's really a Replit day. Tomorrow's going to be Replit and a lot of other systems day. But all you're going to do is just talk to it. Okay? Pretend your developer brings you this product and you're like, hey, this button doesn't work. Hey, I can't read the text on this. I need to change this. Hey, you know, whatever, right? And you can screenshot it, you can click on things, and I'll show you a couple tools on how to do that, right? But when you're talking to Replets, it's nice to know what exactly you are talking about. Um, I need to sign in here. Uh, build a business. Okay. So this is a folder you're going to get access to at the end of tomorrow when you get the follow-up email. Okay. And in here there's a software and tools thing, and there is a couple of things, but we're going to go over coding quick start guide, right? So some of you, this will be very, very basic. Like Jamie, you can just like tune out, you're good here. But you'll, you'll get all this stuff pretty easy. I'm just going to go over a couple of terms and concepts that will help you better understand what you're doing in Replit. You don't have to read all of this and you will have access to this document, but just kind of at the high level, front end versus back end. Okay. Front end is everything you can see. And I mentioned this earlier when I was talking about how p5.js is a front end technology, right? It's just displaying explaining something, right? And versus the backend, the backend is behind the scenes, right? So as an example, a login form where users enter their email and password, right? If I go to Netflix or something, I go to sign in, that form where I put my username and my password and then a button to click go, right? That's all the frontend, right? That's what you can see. And then behind the scenes, right, it's going to check those users' login credentials against the database. And then it's going to return an authentication token or say, sorry, it's not, you know, that, that didn't work. And then it would display on the front end, hey, wrong username and password, and then you'd have to try again. But if it gets that authentication token back, then it lets you in, right? So when you're logging into something, you're already using the front end and the back end. When you're using Facebook, you're using the front end to see stuff, which is displaying things that are stored in a database. Okay, so databases are basically Excel documents. That's the best way to think of it. It's a table, right? You've got, you've got rows and columns. Okay. And so those rows and columns, you know, sit on what are called tables. And each table does something different. Like you might have a users table, a database might have 100 tables, right? Or 5 or 10 tables. Okay. But one of your tables might be users. And in that you'll say, okay, what's the user ID? What's the password? What's the, you know, when was the user created? When's the last time they logged in? Are they an admin? Are they whatever, right? So you have all these little, you know, pieces of information about that user, right? Maybe what's their birthdate? What's their last name? What's their first name? What's their profile picture? What's the whatever, you'll have just information inside of each of those. Okay. So basically just structures data. Okay, and I'll show what that looks like. I might be able to show a database here. I don't know if there is a database in here yet. Hey, there is, look, here's a user's role for my system. Okay. And it's already asking for my Stripe stuff here. So I'll click continue as well. We'll go over that in a minute. But here you've got your ID or your email, first name, last name, profile image, Stripe customer ID, Stripe subscription, subscription plan, encryption key, check-in frequency. So these are all the things that sit on this. Now there's no rows because I have no users in my system yet. Right. I made a user in my old system, in my demo one that we did earlier, but I have I have not set this one up yet, right? Um, I'm not a huge fan of the design here, so we'll work on that in a minute. But the database is up and running. So that is the database. The database is where you store stuff, right? Now, in a database, you can also query— you can ask it questions, right? So maybe you want to know analytics. How many users do I have? How many of my users are subscribers? How many of my users have logged in this month, right? You can get data out of databases, right? Okay. APIs. So Jamie was just talking about Stripe and you just saw me click something on there that was related to an API as well. An API is basically something that allows you to connect one piece of software to another one. That's it. It's just a connector, right? And it might be between a button and a database or a function and some data or a live feed or whatever. But it's just a bunch of data that's out there for different reasons, and I can connect to it. So for example, I like to kayak, okay? One of the things I need to know before I get in the car with my kayak and drive up to the river is I need to know if there's water in the river. So I have a few ways to do that. One, I could say, hey, you know, anybody, anybody for today, right? Is it on, right? Sometimes you need that because there's no, there's no sensor or anything, right? And so, okay, Great. Yeah, the river's going. It's good. It's at 4 feet. Come on out. Right. So sometimes you have to do that too. I could get a sensor. I could grab a Raspberry Pi, right? Like Jamie was talking about earlier, Raspberry Pi, just a little, you know, device. And I could, you know, make sure it's waterproof and, you know, add a little cellular thing and put a little sensor in the river that tells me what the river's at. And then I could go check my own sensors or I have it, you know, send me a cellular signal. You know, that's pretty tough. That's pretty complicated, right? Or 3, I can say, hey, you know what, I bet somebody else has put some river data out there. And in fact, the USGS has, right? The US Geological Survey has, right? And so I could say USGS river data, right? And so I can look at the current water data for this. And I can even say USGS river data API, right? And they even have this ability to say, okay, what is the latest value And so I can grab this API and it's going to give me a bunch of data. And I know this sounds confusing, guys, but it really is quite simple. But I can grab this data and then use it for me. I can, and all I have to do, by the way, is I have to say, hey, Replit, I want to be able to know if these 10 rivers or what the latest river level is. And I want to get an email or a text anytime they hit a certain level. Okay, Replit would then say, okay, great, I know we need a USGS API key, and it would go figure it out, right? It would just go figure all that stuff out. Okay, so APIs connect you to other data, right? ESPN has an API. You want to know scores, you want to see plays, the data is there. There's weather APIs. How does a weather app work? Well, you've got this front end that shows the weather stuff, And then how does all the data get populated? It goes to an API where there's weather data and it grabs it. You can build your own weather app, right? You, all of you could build your own weather app in 10 minutes. Okay. Super easy. Now, inside of that, inside of an API, this is basically how the entire internet works, right? GET, POST, PUT, DELETE. This is what happens inside of databases and APIs. All right, when I click on Ryan's Facebook page, okay, it's going to have his latest posts and the latest things he ate because he takes pictures of everything he eats and puts it on the internet. Um, and it's going to show all that stuff because it's pulling from a database, right? Why? Because he posted to those things. How many of you guys ever posted, right? Yeah, all of us have posted to social media. That's actually just a technical term, right? It's actually sending new data to a database, right? We're sending data somewhere, okay? Then we're going to get the data that's been posted, right? We're going to retrieve data. Or maybe, you know, you post something and you say, you know what, I have a typo in there, right? And I'm going to edit that, I'm going to update it, right? Or maybe I forgot my password, so I need to do a PUT request, and I'm going to update my password to something else, okay? You guys have all done that. Or maybe I posted that and I want to delete it now. I'm just going to delete that. I'm going to delete my account, I'm going to delete that comment, I'm going to delete that post, delete that picture, whatever it is. But this is how the entire internet works, GET, POST, PUT, DELETE. Just knowing that those concepts are there. This is another good thing to remember is CRUD. Create, Read, Update, Delete. Let's say you want to add— I think it was Ryan, you were just showing your thing about Hey, we want to help advisors. We have a blog, right? And you put a nice blog on there that would be good SEO content so people can find out about it. So you could say, hey, Replit, will you build me a full blog function publishing functionality with CRUD functionality, right? I need a, I need a blog system with full CRUD functionality. Now you can create a blog, you can read the blog, you can edit or update that blog, or you can delete that blog. And it would give you an admin that lets you do do all of those things. Or you have a users system. Hey, I need full CRUD functionality for the users. I need to create users and be able to see them and be able to update them. I need to be able to delete them. Okay, authentication. This just means you log in. We don't need to really talk about this too much other than I mentioned JWT earlier. Replit has a weird authentication system. This is your login system. Okay, I don't like Replit Auth, so I just say use a JWT. That's right here. Use a JWT auth system. That's all you need to know. You don't need to know why it works or how it works or what it is. That's all we need to talk about for now. That's all we're going to worry about right now on that. But now that you know some of those things, now you can come in here and you can start saying, okay, how do we improve our app? Let's take this to the next level. It looks like it's working as a check application features work well. Confirm main feature. I don't know why it says 3 out of 5. It looks like it's still working, but it also doesn't look like it's working, so I'm gonna refresh just to double-check here. Yeah, 5 minutes of work, $1.87. Love it. Okay, so the home page looks terrible. I want you to use, um, expert-level design principles and use apple.com as a reference, make the site shine. Can you put that in the chat too? Oh yeah, sure. Yeah, I can drop that in there. But yeah, it's, uh, chat paste, go. Okay. So, um, if you don't like how it looks, you know, you can go through and you can be nitpicky and say, all right, change this to this, change this to that, you know, whatever. That's one way to do it. Um, sometimes I'm just like, hey, you know what, I really love the design of apple.com, or I love the design of whatever. Okay, so I'm gonna, I'm gonna try the functionality though. Start Your Legacy. Why did it build with Replit Auth, man? I don't want it to build with Replit Auth, but whatever, we're just gonna Move forward for now. Ah, stupid robot. Okay, so now we'll let it redesign the homepage. Now I'm inside. This is pretty cool. Look at that, 256-bit encrypted, and it's even like animated so you feel safe while you're inside here. That's pretty cool. Good afternoon, Jon. Add account. Now remember what my thing looked like earlier? It was pretty good, right? I think this is way better. This is pretty cool, right? My Vault. I've got all these different, you know, levels here— executors, check-ins, dashboard. Got a database, I can look at the database and make sure Vault entries, executors, check-in sessions. So a lot of this stuff is built right now. The way that I'm going to find out, if I don't know what I know to go in and look at all these things because I have experience doing it, it's easy. All I'm going to do is I'm just gonna, I'm just gonna start testing things. All right, I want to add an account, I'm going to add a bank account. Okay, bank account Chase. My username is JonIsRich— not really. Okay, password, we're going to call it, um, don't steal my info. Okay, and then this is going to be at chase.com. Instructions, um, Don't spend it all in one place. Okay, save entry. Let's see what happens. Please, please save entry. All right. Hey, something doesn't work. Okay, that means something hasn't been built yet. And actually, we don't even know if something doesn't work. We think something doesn't work. So I'm actually going to click on my data here and I'm going to see vault entry. Uh, no, I don't want to save that. Check-ins, executors, release requests, sessions, users, vault entries. Yeah, it looks like it doesn't even have— I mean, maybe that's supposed to be a vault entry. I'm not sure, but it's just not working, right? All right, I just tried adding—

let's see here— adding— to interact with that once you get the, you know, kind of the—

yeah, once you have your preview up. Preview up. Yeah. So notice I just did something cool. Yeah, I just did this button select. There's a little thing right here that lets you— it's your select element, right? So you can select anything you want, right? You can just grab exactly so that you don't have to be like, oh yeah, see that add entry button over here on this page? You can just be like, this button, right? So I just tried adding an entry with this button and, um, I was able to fill out the form, but I couldn't submit it. Make this work. It's as simple as that, guys, right? That is not technical. Anybody can do this. My 10-year-old can do this, and she does. Okay, so you can come in here, you can add all kinds of stuff. We can try these different things. Add a crypto wallet. See, add crypto wallet— by the way, this is, this is important, right? Um, how do you know whether or not something is working the way it should? Well, you guys have all used— how many apps do each of you have on your phone? Any of you have more than 10 apps on your phone? And if you have more than 300 apps on your phone— I do, I'm probably about there, I've got a lot— but the point is, you've used hundreds of apps. You know how it should work. When you click login, it should log you in. I know that sounds simple and obvious, but if you click something in your app and it doesn't do what you expected it to do, tell your developer, right? Super easy, right? So if I click add crypto wallet and it brings up 5 different things here, that's wrong. It just reset my app because it's editing. It's doing stuff right now. So, so it went away for a second. But if I add— if I click Add Crypto Wallet, it should immediately bring up the Add Crypto Wallet form, right? And so I should say that, hey, you should do this. This is what should— this is how it should work, right? And so I'm just going to go through and I'm going to touch everything. This is why it took me 35 hours to launch my business, right? This is why I couldn't do it in 2 days. Days, right? It took me 3 days. It took me about 35, 36 hours to get through building the entire first version of GenAI API. And what I was going for, guys, was simple. I had to make sure there were no paper cuts on my site, okay? Because especially if I'm dealing with something like Continuum where I'm dealing with sensitive data, if I click on something and it doesn't work exactly as I expect it to, you know, I'm going to be like, wait, is this like, can I trust this? Right? So when I'm on GenAIPI, when I'm building this out, you know, on week, week 1, day 1 of my business, I'm trying to figure this out. If you click on blog and it doesn't load the blog, well, that's a paper cut, right? If I click on one of the blogs and it doesn't actually load, well, that's a paper cut. If I click on x.com and nothing comes up, or this is share this article, hey, it works, right? But I had to make sure that every single thing, um, you know, works, right? Um, if I want to, you know, machine, right? Hey, that's right there. I saw machine right there, so I searched for that. Hey, it worked. Okay, but I have to click on all these things. Read more. Okay, does that work? Yes, it does. Great. Does this Instagram page load? The Instagram page, right? You got to go through and click every single button. Okay, and if it doesn't work You're not done. You're not ready to launch. Okay, if you want to launch and you want to get your customers in a weekend, right, by that next week, you got to make sure you go in and you do all this stuff. Okay. Um, all right, I don't want testing environments. I don't want to use the Stripe test environments. Let's just use live ones. It's asking for an API key. I'm going to show you guys how to do that in a minute.

Okay. But anyway, Jon, question for you. Can we— can— does it prioritize tasks? Like, can you be feeding it other things that it's to do following the prompt that it's working on, or is it best to let that run its thing?

Yeah, so if I go like this and say, when I click on this button, it shouldn't bring up the general list of accounts to add, it should just bring up the add crypto wallet dialog. It's called the dialog that, you know, went like a form, right? Anyway, to add wallet dialog. Okay, boom, right? So I give it that and then I say, okay, let's try add note. Okay. Same thing with each of these. Buttons. They should go directly to the task that was clicked on. Okay. And so what it'll do is it'll just create a queue. See that? And sometimes it starts doing something and you, this is kind of more in an advanced user thing when you know, like, hey, you're doing the wrong thing. Like, stop doing that. You can obviously stop it or you can just say, you can type something else to it. And then instead of having it sit in the queue, you can push this little button that says send message. And yeah, I'll just say send message because it's kind of related to that. And then it'll be like, okay, I'll go through and I'll fix all of these buttons at the same time. Right? So you can tell it something while it's going, or you can just say, hey, do this task, do this task, do this task, do this task. Right before we started the class, I was sitting here working on another thing for a client and, and I had 15 tasks lined up. And so for the first hour while we were sitting here working on stuff, my AI was off working, working, working, doing stuff, getting things done. Yeah.

So yeah, fill up the queue, let it go. Can I ask you another question? So if you've— let's, let's say like mine's starting to get done, but all this time that it's been working, if now I want to like, can you change this? Can you change this? Can you change this? Will it take as long as it did before, or is it a lot faster to give it back to you because it's generally just changing some things.

Something quick. Yeah, what it had to do just now to build your whole app is it was writing 30,000 lines of code, right? Most of the changes it's going to do here on out are going to be like 17 lines of code or 120 lines of code. It's going to be dramatically less. Now, you also can, as I was talking about earlier, as you can see on the screen here, these— this button right here, the agent tools that Ryan was clicking on for us earlier, and you can change the autonomy level. You can even turn off app testing if you don't want that. It'll go much faster if you want to be like, I'm— I want to be in hyper mode, change fast, change fast, change fast. Turn off app testing and go into low mode, and then just tell it, okay, do this, now do this, now do this, right? And so anyway, let's see what happens now. Add crypto wallet. Hey, boom, it works! Add account. Okay, now it brings up everything. Okay, add note should bring up that, right? If I click here, add note. Boom, Hey, I'm working. My system is now working really, really well. Invite executor. Executor invitation functionality will be implemented in a future update. Oh, nice. We haven't built that. We haven't built that functionality yet. That's kind of funny. Manage. Invite your first executor. Yeah, it hasn't. It just hasn't been built yet. It's just not there yet. So anyway, I can come back to my dashboard. I can then say, let's add account. Let's add my— whoops, not my business account. Let's do a bank account and we'll just try this again. Test. Tests, uh, https://jace.com, test, save entry, save entry. Okay, still not working.

You save entry button is doing nothing. Okay, so did you create a prompt that just says, um, test all functionality of the website step by step and have it do it for you instead of you going and doing it yourself?

You certainly could. Yeah, absolutely. Yeah. Especially if you have this, you know, this testing thing and be like, hey, go— I— what I would do is I would switch to Max and then I would say, I want you to go through every button on the entire website and make sure everything works as you expect it to work, right? Don't talk to me until you're done. You could totally do that, right? And it might cost $11 or something or $7 or whatever, right? But Your time, Tyson, I promise, is worth more than $11.

And you know that, right? So, so let's say you've got all of this interaction done and you kind of finalize, you know, that you like it. Then what? How do you get it to actually be something that you can publish?

Yep. Well, you'll just push this publish button up in the top right. We'll get there.

We'll get there eventually.

You'll help me. No, you're awesome. You're doing great, Laura. You're thinking exactly where you should be thinking, right? You're like, I'm so excited. I want to share this. I want I want to get my Mickey Mouse popper in front of my kids, my grandkids, whatever, right? So, um, that's not what I did. I know, I'm sure you didn't. Yeah, I'm just kidding, I'm just kidding. Um, yeah, I do want to see what you, what you have in a minute here. So, so right now anyway, it's working on some stuff, it's doing some cool things. I'm actually going to log out for a second and I'm going to look at the, at the new home page and see if it's any better. Okay, it's a little better, it's a little better. We're getting kind of some Applesque feeling here, right? We could probably upgrade it. It's still kind of bland, right? And these images are probably the problem. We probably need to put real people here instead of just big logos and icons and whatever. But this is definitely cleaner than the first version that you guys saw a few minutes ago. So it did work on that. It did finish that up. Okay. I am going to log in with my stupid Replit auth, which I just don't like, but that's okay.

By the way, it'll pick up right where you left off, like if it crashes.

Totally. Oh yeah, yeah, absolutely. If I close this out or go whatever and, you know, refresh the page and do, you know, or, you know, go to sleep and wake up the next day, it will leave— it will pick up exactly where I left off. Yeah. So if it crashes, don't feel bad to refresh it. Even if, even if the agent is working in the middle of a task, you can refresh the page and the agent's working in the background. Okay. All you're doing is refreshing the front end, right? The back end is still doing its job. Okay. All right. I'm going to try one more time adding an entry here. All right, so we're gonna say Chase, and then we're gonna say test, test, https://chase.com. We'll just click here, test, save entry. Yay! Hey, we have— your Vault entry has been securely saved, right? Now I'm, I'm curious what it means by securely saved. Did it do something to the database that's, um, you know, here's Chase, banking account encrypted data text instructions test. See, this is pretty cool. This is pretty cool, guys. This is really important, and I'm actually super impressed at it. And it might not be impressive to you, but I, I like it. It has— it saved a username and password to get into that bank account. But I, as the admin and the owner and the person who can go look at the database all I can see is this, right? It's all, it's, it's salted, it's encrypted, right? I can't access this no matter what, right? It's encrypted in a way that even I as the company owner can do that. And so I can very confidently say, hey, you can put your data in here, I can't see it. Now we're not done yet. I still would have to go in and say, okay, as I'm building this app out, I'd go in and say, okay, hey, If I was a hacker, how would I get into this? Could I get— if I could get access to the database, would I be able to decrypt this? Right. Or if they're like, no, it's 256-bit encryption, there's no way they're getting past it. Right. Unless they have quantum computer. Right. And that's another fun, you know, discussion topic. So, but for now, this thing is working. I was able to add a bank account. I have one vault entry right here. Right. So I can click on my vault. Theoretically, there's the account. Boom. Right. It's got instructions for the account. Right? And then could I edit it? Yes, I can't edit it yet. Right? And so I'm gonna say, hey, can I delete it? Nope. So I'll say the edit and delete buttons don't work. Okay? So I just go through and I just test it one thing at a time. Now, again, Tyson, I think it was one that asked the question. Yeah, I could have the AI go and test it for me. Okay? Now, let's say we're gonna now answer Laura's question. Right? Let's say, you know what? This thing works. I'm ready for my first test, you know, users. I'm going to send this, I'm going to put this in front of people and see what they, see what they think, see if they like it. Okay, so I'm going to go ahead and I'm just going to click this publish button in the top right here, and I'm going to be able to name it. Or this is, this is going to be a public-facing web address now. Instead of a.com, it's.replit.app. You can connect a.com later. Okay, for now it's going to just be this temporary domain,.replit.app, right? So I'll call this, you know, Continuum. There's no way this is available.

Yeah. Digital Continuum.

What was that button that you just pressed for the publish? Top right. Top right. Yep. And so awesome. Create production database upon publishing. Set up your production base of your current development data. That's fine. Publish. Okay.

Remember, the database is where your data is stored. You— what?

Sorry, you can edit it even after you publish it. Um, the domain, I don't know if you can.

You might be able to— not the domain name, I'm talking about the actual app itself.

Oh, of course, absolutely. Yeah, what you're doing is you're publishing the current state, right? So now this is— in a minute it's going to be published and I'll be able to visit. I could send you guys this domain. You could all test it. You could put in all your user passwords and stuff for me that I could go try and hack, right? I'm not going to ask you to do that, obviously. But, um, but, but then I can come back in here and I can change 100 things and it won't change on the public domain, right? The public one that I've done, because I haven't pushed it yet. I haven't published stuff yet, right? And so that allows you to then have what are called your production version. That's what's out in the public that everyone can see. And then your development version, right? That's what you're still working on, right? And so I, I'm almost constantly working on some new feature on GenAIPI. Like, hey, let's make this admin better, let's make this better, let's write a new blog, let's do this, whatever it is. And then, you know, every few days I'm going to publish as soon as I finish those features and I say this is tested, I've checked it, the security is good, whatever, it's ready to push. Okay, um, I can do really, really fast at this point.

Go ahead, go ahead. ChatGPT on the left side, you can manage your chats and your projects.

Does that same thing apply here? So, uh, yes and no. Um, so your chats, you have this history of your chats, but this is only your chats related to this app. You really should almost never have to go in here. Okay, although I will point out one thing. Let's say you go nuts and you just get in this weird loop and you break everything. You're just like, oh my gosh, we need to go back. You can come back and you can roll back to different checkpoints. 19 minutes ago here, you can see what was happening. You can load the preview of what it looked like in a past version before you broke everything and you can roll back your code. That's really important to know. But Laura, I think what you're asking is if you have other projects and things you're working on, this is all done now, we'll come back to it. But I'm going to leave, if I just go Home, or if I go to my apps, I can see all of the different apps that I've built, all of my different projects. If I wanted to, Danny's brother founded a company called Gig and I rebuilt the whole thing in 10 minutes. He spent millions of dollars building it and I was showing him this thing. Anyway, so I rebuilt this thing. This is a whole other project, The agent is right here. I can talk to this agent on this project, but this is the website that was built for that. Really nice website, actually. Pretty dang good. All I said was build a clone of gig.com for me that is fully functional and built in React. That's all I did to build this. Okay, now we're going to do something pretty cool here. We're going to go back to Digital Continuum to my other project here. Here we were inside of this. We just— so now the domain is public, so now I can click this. And anybody could go and visit this website right now. Okay. Now nobody will. It's not going to show up in Google yet. I have no SEO. I have no digital presence whatsoever. So people aren't just going to start showing up. But theoretically, everything should work exactly as it did over there. Now I got to do this. Replit. Whatever. Okay. We're going to get this happening. Okay. So now I'm in. Okay. Perfect. Then I should be able to add crypto wallet. It's all working. I should be able to see my entry of this Chase account. That's good. I can edit it. Oh, edit functionality hasn't gone live yet. We haven't fixed that yet. Right. But I can also sort by, you know, different— I can filter, I can test all these things. Just got to push every button. Okay. All right, guys. Now we're going to build a— so that's how you publish. When you want to connect, I will say this, when you do want to connect another domain, right? So let's say you want to— I want to create, you know, digitalcontinuum.com. Okay, well then I'll come over to domains up here, right? And this is here in the publishing tab. And I can use one of my domains that I've purchased. I can buy a domain in here. I personally am not going to buy a domain in Replit because it's still in beta and it doesn't allow me to do some things that I need to do. So I'm just going to manually connect from another registrar, right? So I'll say, you know, Digital Continuum. Com. I'm going to click next, and it's going to give me some information about how to connect it to like GoDaddy or Namecheap and things like that. I'm not going to cover that right now. It's pretty easy to do, and if you click this learn more button right here, it will walk you through it step by step. And there's a video even showing you how to do it. It's very, very easy. It sounds complex, but it's— you know what, I'll do this. I will go I'll do like a super, super fast demo of what it would look like. So you would go to GoDaddy, for example, you're buying something, you'll buy your domain, and once the domain ends up in your account, then you'll be able to click on My Products, right? So you'll have all of your cool domains that should show up here in My Products. Wait for it. Come on, internet. All right, so here's all my domains. I've got all these things, and you can see on all of these domains you have Manage, and then you have this thing called DNS. DNS is basically the instructions of where to point that domain. Okay, so let's say I want to use genaipi, which I'm not using, genaipi.ai. I use the.org, that's my main site, right? But let's say I wanted to turn on the.ai or the.com, right? Anyway, I would just click on DNS DNS, and that brings up the settings that I need to use. And so when we were over here in Replit, it said, hey, you add these records to your domain's DNS settings. So I'll get an A record. Okay, that's great. Let's go to the A record, add new record, choose an option, A name, Okay, so I'm going to put value. 34.111.179.208. I'm going to copy that and I'm going to paste it, right? And I'm done. Okay, then I'm going to add more records. I'm going to add one other one. I'm going to add the TXT record in the same way. And then I'll click over here on GoDaddy. I'll click Save, right? I'll click this Save All Records. And then over here I would click Link Domain. Okay, so I know I did that fast. This is recorded. Come back to it if you want to check that out later.. Okay, um, but that's how you would actually connect it to another domain. All right, okay, um, we're gonna add a useless feature, but it's going to be fun. Um, actually, no, we're not going to make a useless feature. Let's make it a fun feature. Let's, let's add a, uh, we're gonna add— we're gonna replace Tyson and Ryan in one fell swoop here with AI. We don't need them to tell us what to do. We're gonna get AI. We're gonna make an AI advisor, okay? That's gonna be integrated into this system, and you can only access it if you're a paying subscriber, okay? So it'll be $12 a month, and, and I'm, you know, obviously use at your own risk, but, um, that's, that's what we're gonna build here, okay? So I want to add an AI-powered estate and, yeah, estate planning advisor. Okay. You can only talk to the advisor when you're logged in and a paying user. For now, make sure that the I need to figure out what my— I can figure out what my user is. Uh, email, I don't know. Sure, email that the Jon— or that the user with the email jon@flowstateleadership.com is flagged as a paying user. Okay. All right, then. Let's use OpenAI for the advisor. And I want it— I want the advisor— whoops— to show up in the bottom right-hand corner of the screen. As a floating icon. The advisor's name should be Bob. Make sure the advisor is always on task and doesn't respond to questions unrelated to estate planning and related topics. Have it be very friendly and helpful. Also allow the advisor to teach. Be sure the advisor knows how to use every feature in the platform so that it can serve as a guide. To the user in all things related to Continuum. Okay. All right. So that, that's kind of a long request here, guys, but, um, and I'll even copy this and paste it into the chat just if you want to save it. Um, you could kind of see how I structured what we would call the prompt. While this is building out this thing, it's going to do some things. So first of all, it's going to ask me for an OpenAI API key. Okay, remember, an API connects things, and an API key is going to give us access to those things. Okay, so I already have, because I use Replit all the time, I already have an OpenAI key in my account. So I'm going to come in here and I'm going to open the Secrets Manager. So the Secrets Manager, this shows all the secrets and passwords and Stripe keys and things like that. Now, all obviously can't see it right here, but I have in my account— this is for the app, app secrets for Continuum. My account though has tons of keys for all the different types of APIs I've used to set different things up. Okay, so it's already saved in there. But if I didn't have it saved in there, it's going to ask me to go over to OpenAI to get an API key and then it's going to integrate it and that API key is going to connect to my account. I want to show you how this works exactly. You know, first of all, here it is. It asked me for the key. If I did not have one, then I would just follow the instructions here. You can get an API key from OpenAI by going here. Awesome. I just click there and then I'm going to log in or sign up and I'm going to get my key and I'm going to, you know, create new secret key, copy the key and paste it here. It's all you have to do. That's all you have to do to make an AI-powered assistant. Okay, I'm going to click continue. Okay. Now, while that's continuing and building out, I'm actually going to show you what that looks like. So OpenAI API, get into the API platform. To be clear, this is sometimes confusing. ChatGPT and OpenAI are not the same company. They have the same owners, right? But ChatGPT is actually built on top of OpenAI's API. I'm going to say that again. ChatGPT is an interface that is built on top of the brain, the OpenAI LLM. Okay, so I'm going to come in here, I'm going to log into the API platform, which is different than ChatGPT. Okay, and then I'm gonna look for my API keys, which they always— I always get lost in here. Maybe it's in my settings here. No, API reference. Oh, where are my keys? Where are my keys? Admin API keys. That's Yeah, this is just reference stuff. I need my actual API key dashboard. Maybe it's just on the dashboard here. Um, it says API keys. Here we go. Okay, perfect. So in here I would just say create new secret key and I would just call it, you know, um, fun. Okay, create secret key. There it is. There's my key, right? So you could copy this key and then take it over to Replit and paste it. That's all you have to do to install your— to give it the API key. The other thing you must do, this is important, if you are using an API key with something like ChatGPT or something like OpenAI or, or, uh, I don't know, some weather API or some whatever, right? It costs money, right? Now the amount of money it costs is insanely cheap, right? The amount of money we'll use in testing this API right now is not even equivalent to a penny. I will use a fraction of a fraction of a penny in doing all of my testing. If I gave it to all of you guys to test for an hour, I would spend a total of 2 cents, right? The APIs are very, very cheap, but you have to add, um, budget. You have to add a credit card. If you don't add a credit card, it's going to say the API is failing. The API is failing, the API is failing, and it won't tell you why. Okay. Make sure you add a credit card to your, to your account. Okay. So you'll have your settings and you'll go to billing and you just make sure you add— I have $20 of credit balance, right? Just sitting there. And if I get down low, then it will take it back up, right? When it gets down to $10, it will just charge me another $10 to take me back up to $20. So it looks like I've been— I've had it refreshed lately. Okay. Or recently. Okay. All right. So it's still working on this advisor thing. It's still on 0 out of 5. There's all kinds of stuff. It has to add chat feature, create AI, check user login and payment, check chat works properly. Okay. All right, cool. Well, it's making— it's making some stuff right now, so it's working on all these things. In a minute, we should see a floating little button here that will be my advisor. Right now we're going to get just— yeah, we're going— we're going pretty fast here, guys. Hold on, I have to pause. I know I just said a lot of crazy stuff. Questions? Is anybody confused or, or want, want to ask a question about anything I just said related to how to integrate an API into an app? Okay. All right. Either you guys are totally lost or I'm a really good teacher, so I'm not sure which one it is, but, um, we're going to continue on.

So It's probably somewhere in between, probably somewhere in between.

Um, I look forward to rewatching it personally, so yeah, that's what my takeaway— yeah, you'll probably need to rewatch that as you try to do it. But again, guys, I want it to just be so simple, okay? When you're building in Replit, you can connect to tons of cool things, right? Here's an example. BYU— let's see here, uh, cougariq.com. Okay, I built cougariq.com. I'm a BYU fan. And, um, oh, you know what, it actually, it actually might have a little error, but it's okay. It still, it still should mostly work. But, but here I built this little past games feature where I can look at like the last few seasons of games, right? So if I wanted to go and look at the, you know, how we beat the— actually, let's go to something fun. We beat the Buffaloes here, right? Uh, we beat Colorado last year, so I'm going to view the stats.. And so I can pull this up. I've got, you know, box score, all these different things. It's pulling all this data. I've got team stats, total yards, efficiency, players, what was going on here, who did what. Um, and then I can even go to highlights, right? And I can, I can watch these, these videos. Every possession started right around the 50-yard line. Pretty cool, huh? Right? All I did with this was use the free ESPN API. And I just said, hey, I want you to tell me all the data that we can pull in. Oh, we have videos, we have all the scores. I was like, great, do it all, especially the videos. Let's make that really awesome, right? I can even come in here and say, hey, I want to download this highlight and post it to my Instagram. All right, hey, remember this cool play? Look, ESPN's API allows me to download this, this play. Isn't that crazy? That's so cool, right? And I can do it for any sport. Hey, let's watch the touchdowns. Feels good when a touchdown goes in. I just feel happier. So anyway, but bottom line is I can go through basketball games, I can go through different seasons. I could go back here to 2022. I go back to 2012 if I just built the button for it, right? But I can come in here and I can say, hey, you know what? I want to see this. I want to see— oh, okay. Somehow that didn't connect. Maybe this one didn't connect properly. I haven't actually tested this for a minute, but okay. SMU Mustangs. Maybe they didn't broadcast that one. Oh yeah. Anyway, but I can go in, I can look at these highlights for anything that's there. Maybe I didn't connect basketball. I swore I connected basketball. Let's go back one season and then go to a game that probably was recorded. We lost to BYU or to Utah. Yeah. I didn't connect basketball. Somehow that isn't working. Obviously, I can go into Replit and just be like, hey, update this please, and it would work. We're starting to get a little bit of progress here. You just ask the system how to do it. Hi, I'm Bob, your estate planning advisor. How can I help you today? Okay, so that's cool. It might not be working yet, but we can do this, right? Um, how can I prepare for death? It's kind of a deep question, so I'm curious. What it's going to do here, if it's going to answer the question related to, you know, in a way that's related to the estate planning and Continuum and what's going on here. Now it looks like nothing's happening at the moment, um, and so it looks like it's almost done with this too. Check AI chat works properly. Uh, okay, it's working. Verify backend front-end component integration authentication. Looks like the error is fixed. Let me call this to make sure this is working. Yeah, an early return. Architect found some issues, so it's still working on it. Okay, but we almost have this working, and I'm telling you guys, you could, you could actually go in and, um, you could rebuild ChatGPT on OpenAI's API. You can call it, you know, askjeeves.com, right? Um, and, and it would, you know, you could, you could build your own search engine, your own AI-powered search engine, and you could go charge money for it. On. And that's totally within the rights and the usage, you know, policies for OpenAI. Now, you're going to be competing against ChatGPT and all of their marketing and their market position and everything. But function— functionality-wise, you could be getting the exact same answers that ChatGPT gives. Well, mostly, mostly at a high level. There's, there's a little nuance there, and I'll talk about that more tomorrow. We'll talk about prompts how to really set that up. But anyway, so it is, it's going to work on this kind of stuff here. Backend is fixed, update the frontend component. I really want, I really want to talk to Bob and get him to do this. Another thing we could do, by the way, and let's look at the database. Let's go to the data here. Yeah, one thing I don't see here is chats. Right, what if I wanted to save my chats with my advisor so I could go back to past chats, right? So I could say, hey, let's, let's make it so that we can actually save chats. And so why, why would that be important, guys? Why would you want it to save chats if it's your estate advisor? Legal, a whole bunch of different reasons. So potentially legal, but I'm thinking more like I want good— I want a good product. Right? Yeah, I want— yeah, I want my estate advisor to remember who I am and what I've talked about, right? And so then it has context, it starts to become smarter and smarter about my situation, and I could tell it and give it access. I could say, hey, I want to give Bob— I want it to be able to see all the types of accounts I have, everything that I have here, everything that I have set up. And maybe I add more features later, right? Maybe I add, you know, trusts and wills and different things like that that could be managed and kept in here. And then I could keep asking it questions about it and it gets to know my whole system better and better and better. And then it becomes actually a pretty dang good advisor, right? Because it understands how it all fits together, right? So any questions on this, guys? I like to mess with the AI to see if I can get it to go off topic, or ask it kind of a weird question to see how it— if it brings it back to the conversation, things like that. And it's still, it's still working on things. So what I'm going to do is while it's still working on things, I'm just going to say this. It seems that the responses are not coming through. Chatbot. Good. And I'm just going to give it to it right now because I know it's still working on stuff, but I'm just going to make sure it knows that that's happening and then it needs to test that feature.

Okay. Would you say, would you say that it's necessary to go through and like be the one who's clicking, like kind of like the user experience that, that forward, you know, facing customers are going to be using, not having like, is there not a way to have AI kind of go through and test these things?

Is it necessary to go through what is potentially be a grip of, of, you know, links to verify that they're, they're working? Personally, I'm gonna do it myself. I mean, before I'm publishing, right? Like, yes, I might have AI do it first, right? Like we was talking about as Tyson earlier, right? You could certainly have AI go through and test everything and build everything and make sure it all works. But then you gotta go through and make sure it works in the way that you expect it, you the user expect it. And so you gotta be like, okay, And you have to start from the beginning. Okay, I'm on the home page. Would I click here? Yep, that's good. Okay, I'm ready to go. Let's get into this. This registration work. Did the registration send me an email? Okay, great. Let's check on that, right? By the way, I have not connected an email API to this yet, right? What if I forget my password? How am I going to get in? I got to set up some sort of an email API, right? By the way, there's a lot of them out there. You could just tell Replit, say, hey, I need a forgot password functionality. So we need to set up some sort of email and it'll say, great, I recommend you use this. Go sign up an account here. You get 100,000 free messages a month, you know, or 1,000 free messages a month. And then, you know, it's $0.01 per email sent after that, you know, whatever. Anyway, so by the way, guys, with GenAIPI, with all of my fees aside from Stripe, because Stripe takes, you know, a few percent, right? But aside from Stripe, all of my fees for SendGrid, my email stuff and my AI integrations and other APIs and everything, my total cost is less than $100 a month to run a multi-million-dollar business. It's unbelievably cheap. Don't be afraid of these APIs. They're amazing. If I wanted to connect a weather API to this, weather API allows you to, again, you could build your own weather feature. Maybe you just want to show the weather in your app because for whatever reason that matters. Okay, look at this. At $0 per month, you will get an API key that can do 1 million calls per month. That's huge. That's crazy, right? And so while you're in that testing growth phase, usually APIs are almost free, right? And then let's say you start to, you know, get some real traction and you're getting 3 million calls per month or 10 million calls per month. You can afford $35. You absolutely can afford $35 because you've got a few million people using your stuff, right? And hopefully you've set up some sort of monetization strategy, right? You're making money somehow. Okay. Um, all right. Okay guys, we've got about 10 minutes left today. I want to just leave it for questions on anything. We can go back and talk about Eric Schmidt's video and how we're all doomed or how we're all excited.

Um, and Or we can keep talking. Let's peek at Laura's, at Laura's thing.

She wanted to share that. We'll see if that's cool.

We can do that tomorrow.

It's still working actually in the background. Oh, is it? Yeah. Right on. So I'll just let it still continue to work, right, Jon? Just let it keep getting done.

And then once it's done, I can interact with what I want to change and then we'll look at it tomorrow. Yeah. Yeah, exactly. That's interesting.

400. Your organization must be verified to stream this model. Can you send us the Facebook group though? And one of the things, you know, I don't know if this is a good spin-off, you know, like we talked about the continuous learning, but I wouldn't— I'd be interested in doing a mastermind, like doing a where we meet once a month and somebody like you kind of walks us through a project and just gives us more hands-on. I'm, I'm a very much a visual learner.

And so it's, it's nice to be able to walk through projects together.

Um, Laura, are you thinking like, like almost like having a regular one-on-one coaching with somebody like this as you're, as you're working through a project?

Yeah. I mean, even if it's not one-on-one, more like, I think you're talking about group. Group to, you know, with like me, right?

Where you can ask questions like kind of like we're in a meeting right now. I just like— yeah, I just like a repeat of this. But, you know, office hours. Yeah, I'd be okay to do like this training class once a month because it dedicates me time without distractions as well as get some of the ideas that I have going where I could take some of the AI into my business. Um, but I think that I'm gonna need that dedicated— I just know that if I don't have a structured class like this or a mastermind or, uh, you know, a dedicated time, I won't do it because I just— like, today is month end in my business. Like, I told everybody, what was I thinking? Like, the busiest month of our, our, our business.

And it's like, so I've told them, okay, You know, after this time I'll be available, but until then I'm not. I'm busy today.

Yeah. Sorry. Sorry to take away your morning here. Oh, no, it's, it's been very valuable. And to be honest with you, month end, the last day of the month, first day of the month is, is the busiest. Sure.

But it was really a nice kind of like a reprieve for me as well. Yeah, absolutely. So I just sent the Facebook group to the chat there. Go ahead and click on that if you want to join it. And I'll just, you know, I'll just accept everybody that clicks on that and that'll come through here in a second. But this group, again, is, is really quite nice. A lot of really good people on here. You'll see Brantley is an amazing guy. George has been around for a long time. Rob, lots of good people in here. Kevin. Yeah, lots of great people doing, doing good stuff in here.

So awesome. Other questions? Comments, concerns, thoughts? Thank you, Jon. This has been amazing.

You're, you're a wonderful coach and teacher.

So thank you very much. Yeah, I hope it— As far as like the— sorry, go ahead. I totally interrupted. Go, Danny. Um, my thought was what, like, so do you have like a takeaway, like notes-wise, or like kind of the, the workflow for what we just kind of walked through to kind of just give it some reps and stuff, make it sticky with bringing it back and forth to stuff and Or are we just kind of gathering the concept for now because that's not in stone?

Yeah, I do. I absolutely do. You're going to get them tomorrow. Okay. So that'll be in the, in the Google Drive that I send you. It has like workflows and how to walk through and the process of how to do this stuff and all the different tools and tips and tricks on how to use those tools better and all that stuff. Right. And so we'll work on some of that stuff tomorrow. One of the things. Yeah. Yeah, one of the— yeah, so today what I want you to do, guys, if you're going to work more on this stuff, right? Laura is getting back to work, so she's not going to do this. But, you know, a lot of people will come back and be like, I stayed up till like midnight working on my project. That's fine. Just keep working at it and trying it and don't feel bad if you run into something, you get stuck, right? Come tomorrow with a question if you get stuck. But the reality is I got good at this just by trying it, right? I am no smarter than any of you guys, right? I just have more experience, right? I'm a whitewater kayaker. If you watched me kayak off of, off of huge, you know, Class 5, whatever, with a 20-foot waterfall at the end, I'd make it look easy. And you'd be like, how do I do that? I'd be like, well, you just got to spend time in your boat, right? And you're going to get used to how the water feels on the bottom of your boat and what happens when you tip over and what happens when you hit a rock and what happens when you do this. You just got to, you just got to go through it, right? And so, you know, I can't really, you know, shortcut you to the front. I've given you some shortcuts, right? I've definitely shown you what's possible.. But the reality is just play with it, just use it and mess with it and, you know, have fun, have fun, go build cool stuff, start a new project and try to build a quick game with your kids, whatever. And as you see this, be like, oh, that was easy. That was super cool. And then you'll just learn, you'll learn a few of the buttons. And there's other, by the way, there's other websites besides Replit out there. There's Lovable, there's Bolt, there's Base44, there's other systems. And some of them, some people like better than others. I have tried all of them and I absolutely 100% believe that Replit is the best, especially for people like us who are not developers but need a full powerful solution, right? There are other systems that are maybe a little simpler, but they get stuck at just, hey, you built a website, you can't actually build like a full-on application like we're building today. We're building applications. Now I have built websites too with Replit. Replit's awesome at doing those simpler things as well. Like, hey, I need a landing page for this you know, this idea. Okay, boom, done, launch. I can do that in an hour. And it's way faster than using even a template like Wix or Shopify or any of those types of things, right? Those things take days to configure and get right. And you have to write all the content out manually and it just, you're always switching between the templates and it's not working right. And it's a lot of work. And Replit, you just be like, I have an idea, launch the site, publish, you know? So I, I'm, I'm a Replit fan. I'm a big Replit fan.. But it is just one tool and you'll get bit like any tool. Like the first time you used Microsoft Word, you didn't know what all the buttons were. That's okay. Now you know how to bold things and italicize things because you've done it a few times. It's not much harder than that. And again, the cool thing about AI in general, and this is for today and tomorrow, is the language of AI is whatever language you speak. Just speak to it. If you don't know what's going on, I don't understand this. How do I make this work. Explain it to me like I'm 5. Yep. Okay, just talk to it, right? Um, and it will— it'll talk back to you, right? If you want to speak Spanish to it, you can do that too. If you want to speak Morse code to it, have at it. Elvish? Anybody? Any Elvish speakers?

Like Tolkien fans? Like crazy? Yeah, either. But I know there's— I have another question for you, Jon, about language. So I do a lot of Zoom, PowerPoint, or Canva presentations, and we're kind of launching into the Hispanic market, can I put it in ChatGPT, the presentation, and say, can you add to each slide the Spanish interpretation?

And would it do that? Because I can't confirm that the language is accurate. So, right. So yes, although I wouldn't personally use ChatGPT to do that, I would use either Manus. Or Gamma, and I will teach you both of those platforms tomorrow, right? Um, but yeah, there's some really, really cool things you can do there. You could also have Replit build a presentation. It's actually pretty amazing if you take an existing PDF or PowerPoint and just drop it into Replit and say, turn this into a really beautiful presentation, you know, a presentation site, right? Then it will still go through the slides, you can still click through, but it will like animate everything everything and bring these things and pull more data in. And then you could just simply say, add a, add a button on the site that switches the whole presentation to Spanish. Oh, wow. Just automatically translate it all and put that button and functionality in place. Like, it's, it's crazy. So here's a cool thing. So I mentioned this BYU, you know, thing on the BYU athletic part, whatever. I was sitting down with the head of the Built for Life program, which is basically the equivalent of the principal for all of the athletes, right? His job is to make sure you're getting good grades and you're doing well outside of sports. Okay. And so one of the things he has to do, in fact, I'll show this, I'll show this. It's pretty cool. One of the things he has to do is go through and check on the grades of everybody, right? So he gets all these reports and everything and he has to put together all these, you know, graphs and crazy stuff. And it takes him, he said it takes him about 20 hours a month to do this. Like a lot of work. He works on it constantly, like a few hours a week to get this done. And, and so let's see here. Where is it? It's one of these. It's the grade, grade dash. I think this is it right here. Okay. And so he said, hey, could we have Replit just like, you know, synthesize all this data and put it all together in dashboards that makes it easy? I was like, sure. And so I had him drag in one of his files. It's just a huge Excel document. Document with tons of data. And all of a sudden, he's got this dashboard here that it literally— I did this sitting in front of him in his office in 5 minutes, right? And all of a sudden, boom, here's your reports, right? Here's all the different test scores and all these different things. And here's the sports, you know, lowest scoring up to the highest scoring. Let's do, you know, let's just, you know, and do, do freshmen and females, right? Our freshmen females doing right. Here's all the, all the records. Here's how everybody's doing, right? It looks like basketball is the highest scoring, right? Anyway, point is, this is much more than just a, hey, let's build a piece of software for my business, right? It's, hey, I need to display something really quick and it's pretty complex data, right? He also that same day had a presentation to do about, you know, I don't know, whatever this is that comes up here. And this is actually slides. I'll show you. I'll show you what I was talking about here. This is— that's why I thought of this, this version here.

And I know we're right at 1:00. If you have to go—

will it publish into Canva or just like PowerPoint where you could just upload it? Well, if you wanted to use PowerPoint, but really this, they— this was built to just be its own standalone. Just to do it. You know, you play it, can you edit it, right? Yeah, of course you can edit it. You can go through and do whatever, right? But this was mental health and performance, understanding roles, responsibilities, and teamwork, and it was a boring, bland presentation, right? And it turned it into this, right? It built these things and like, here's how you can pull this stuff together here, guiding principles. And, you know, taking just any presentation and all of a sudden bringing it together, setting up all these different things, he was able to then put this together and Boom. Look how much prettier that looks. I promise you this thing was crappy, right? I think you might even have— this is what it was. I don't know if I can download this. Can I download?

That's okay.

Anyway, but it was, you know, it was just a very nice— it was your average PowerPoint, right? And it turned it into an interactive thing. And again, Laura, you could say, hey, you know, let's edit this and then let's do this and let's add this button to change the language. And we can do now Spanish, Chinese, Portuguese, and German. Boom. All you do is ask it and 10 minutes later you'll have all those languages with a dropdown menu to select the language. The entire presentation would be working perfectly in any of those languages. That's what it's capable of doing, right? So anyway, thank you guys for today. It's been awesome. Um, I will upload this and I will send you this recording in a couple hours. So it takes a minute to like download the whole thing and then upload back to YouTube and all that stuff. But I will get it over to you as soon as I can, and we'll see you tomorrow morning, same link, 9 o'clock. Thank you. Thank you.

Thanks, guys.

See ya. Bye.