Cohort 24 - Day 1 - January 2026

Okay. Um, all right, so we'll go ahead and jump in, guys. Um, I know we might be missing— I think, I think we're just missing Soraya at this point. Um, let me just double check. I always need to do a little bit of roll here. So Dave's not here. Soraya's not here. Eve Kelly, who's that?

That's me.

Evelyn. Evelyn right there. Okay, got it. Sorry, I'm just looking at my email and I'm just seeing email addresses. And then do we have Matt? Matt. Oh, you know what? I think Matt's the one that— that's another one that I had to adjust. Uh, okay, it looks like Sarai just messaged me. She said she's getting on, so that's good. And then who's Lit Fix?

That's me.

Okay, that's Patrick. Perfect. And you did respond to me. Thank you for doing that. And then Spencer at Recyclops.

That's correct.

Yes. Um, that's awesome. I've heard lots about Recyclops. That's really cool. Excited to have you here. We'll talk about all your stuff in a minute. And then Tanya. Okay. And, uh, so Tanya is, um, actually our new CMO at GenAIPi, at my company here. And so I have anybody that works for me is required to take this course. Just because I want them to, I want them to get up to speed with AI and this course is, there's no other course out there that I've seen that gets anywhere close to what we're going to tackle over the next 2 days. And I really just want to make sure that we, we kill it. Hey, Soraya, welcome.

Oh, hey.

Hi.

I just shared with everybody that you used to be my babysitter. So anyway. Yeah, sort of.

I don't know.

Way back, way back in the day.

Yeah, 2013, we were in the same company.

Yeah.

Yeah. Anyway, so, guys, there's lots of cool things we're going to be going over the next couple of days. It's bright and early. Anybody here Pacific time?

I am.

You are. Okay. So 7:00 for you. Anybody else outside of Utah? Where is everybody? Well, we'll go over that in a second. But anyway, so it's not crazy early at 8, but we're going to go 8 to 12 today. And then tomorrow we might adjust the schedule just a little bit. I'll talk to you guys about that later. I have a flight that leaves at 1:45 and I'm an hour from the airport. And so we got to either start earlier a little bit tomorrow or we do a bonus day with you guys. So we do 3 hours tomorrow instead of 4, and then I'll do like a bonus 2-hour session for you guys, give you guys an extra 2 hours as a follow-up, which actually might be cool. Do that in like a week. So anyway, we'll figure that out. We'll figure that out. I scheduled that flight. I have to be in Houston by tomorrow night. So, well, we'll get started. So this is being recorded, as you know. And so there will be times in this course where you're like, holy cow, you're going so fast. Right? And You might— I'm going to try to help you guys keep up. This is a very interactive, hands-on course. I sent you guys an email so that you can sign up for ChatGPT, and many of you probably were already, but Replit, maybe you weren't signed up for. Hope you've done that because we're going to be inside these platforms building. I'm going to be asking you guys to share your screens so we can look at it and critique you and make fun of you. No, just kidding. There's no making fun of. There's absolutely no shame in learning, right? Learning is awesome and you're going to get it wrong and that's okay. I get it wrong still. But we're going to work together through some of these things and we're going to really focus on a lot of really fun stuff. The purpose of this course is the name of it describes it well, obviously, How to Build a Business in a Weekend with AI. But the importance is actually in the name, right? In, in order, right? So how to build a business is first, right? So this is an entrepreneurship, like how do you start a business? Now, if you're taking this course because you want to get better at your existing job and you're not necessarily taking it because you want to start a business, that's okay. There's a thing called intrapreneurship as well, where you're building new things and ideas and things inside of your existing company or just wanting to get better at doing your existing job. That's— you're going to get that out of this, right? How do— when you build a business The reason why this is a really fun topic to teach AI around is because when you build a business, you have to do brainstorming and ideation and verification and research and marketing and building products and finance. You have to do all— you have to touch everything that you do in business anyway. You just have to do all of it as the entrepreneur if you're starting a business. But anyway, so this is an entrepreneurship course with HyperSpeed attached to it in a weekend with AI, right? And then obviously it's an AI course and, and I think many of you are taking it so that you can become very proficient in AI. And, and that really is the idea is to be able to come out of this course saying, all right, I know more about AI now than 99% of people around me or 99.9% of people around me. And that's ultimately why I started this, this company. It wasn't actually to do these courses. This course wasn't actually in my business plan when I started this. I just, had a bunch of people ask, and I've really enjoyed doing it. This is, as I mentioned in the email, the 24th cohort. So I've done this a lot of times, and I've had— and those are just this how to build a business— and I've had anywhere from 5 or 6 people up to, you know, maybe 15 in this course. And then I work with businesses, and I've had 50 people, 100 people. I've taught in front of 100,000 people these same, these same principles, you know, working with Tony Robbins and Dean Graziosi and other great people. To teach these same principles that we'll be going over. We're going to hit probably 15 AI platforms over the next 2 days. Right. I know we talked about ChatGPT and maybe Groq and, and Replit, talk about how we use those, but we're going to go over a lot of them and it's going to be a lot of fun. So anyway, bottom line is it's being recorded. You'll be able to go back and, and, you know, watch something if there's some, you know, process or something where you're like, man, I really want to do that again, but I don't really remember how Jon did it. Get it. I will also provide you with, after the course, you'll get this big long guide that kind of shows you all the different pieces of software and tools that you use during the course and that I talk about, and probably even a few that I won't even hit on, but lots of resources for you in a nice big organized folder that you'll all get to have. So, with that being said, I know that's just kind of a little bit of an intro, and that was kind of just about the course. I'll talk about me in a minute here, but I want to go and I want to get to know you guys. So we're just going to do it in the order that I'm seeing it here on my screen. So we'll go Spencer and then we'll go Tanya. And, uh, and what I want to hear from your intro is just your name, where you live, um, what you do for work. And then I'd like to know whether or not you're a developer, right? Even if, you know, maybe you're like, I've vibecoded before, maybe you're not quite a developer, but anyway, so just tell me, tell me those things. Um, Spencer, go ahead.

Sure, yeah, uh, good to meet you guys. Um, I, let's see, I live here in Utah, Eagle Mountain, Utah. Um, and, uh, you know, we love it out here. It's a great, great part of the state. Um, I, let's see, I forgot what questions you asked. I, I mentioned I, so I work for a company called Recyclops, and so I'm, I'm the one-man technical team at Recyclops. So we're, we're a small company And what does the Cyclops do?

Share with the class. Sure.

Yeah. So Recyclops, we're split kind of into two parts. So half of our business and how we got started was out of BYU providing recycling initially to apartment complexes. But we, we quickly expanded to kind of more rural and remote areas that don't have your traditional city-sponsored program. So we use DoorDash model. We contract drivers. They go pick up the recycling and put it in their truck or Suburban or whatever they have. Put those bags of recycling there, and then they haul it off to the same facility that, you know, Waste Management or Ace Recycling would. The other half of our business, which is rapidly growing and what we're primarily focused on these days, is helping large brands who either have large volumes of material they want to keep out of the landfill and/or just kind of really hard, odd material that they don't quite know what to do with. They want to keep it out of the landfill, but they don't know how to do it right. So we come in and help them with that. And so yeah, we used to have an engineering team. I am a developer. We downsized a few years ago. And so for the last few years, it's been just me. And so, you know, my intent or goal in doing this course is, is we're growing. I'd love to keep our engineering team small.

As small as possible.

And so obviously love to see how we could compound that.

Yeah, that's awesome. That's awesome. Well, wonderful, wonderful. Thank you for, thank you for sharing. Tanya.

My name is Tanya and I live here in Utah. And what I do now, I'm a CMO for GenAI PI. And traditionally I help people build their companies. I've owned a few of my own and over the last 10 to 15 years, I think technically about 17 years. I've done fractional consulting for many, many companies. In fact, Spencer, I remember when you guys came up through BYU. I worked at BYU for about 5 years in the entrepreneurship program there. So good job, by the way. And that's me.

How long have you been part of the Recyclabs team, Spencer?

So I've been with Ryan from the beginning, joined full-time 5 or 6 years ago though.

So you're a founder then?

No, I mean, a year into it, he needed some help. So I worked nights and weekends. So he's, he's still the sole founder, but I came shortly after that. Yeah.

Well, in my company, I would call you part of the founding team, at least.

Correct. Yes.

That's awesome. Very cool. Well, I love it. Thank you, Spencer and Tanya. Patrick.

My name is Pat. I'm not sure if it's allowed, but I'm not from Utah. I'm in San Diego. And I don't, I haven't worked for like 30 years. I do, I make little websites of my own that sometimes sell for like $50,000 or something or sometimes don't and just bring advertising income. So my general, my expertise for the last 30 years is probably just building really great things from a user's perspective. Not being fancy about advertising, not being fancy about developing per se, just more about creating really high quality things about subjects that are useful to people. And over the years I've seen the, again, for before Google was doing this, I was doing websites before Google existed and Basically, I've just always kind of traveled and I haven't tried to set down roots or make billions of dollars. But now I'm more in that mode. I'm getting old and I want to get something solid beneath me. And I've really never worked with anything where a customer pays me something. It's always been I get money either through affiliate income or through advertising income. And there's a million useful ways to do things. But the problem I've encountered over my lifetime is mostly whatever I've learned has gone out of favor by the time I need it the next time. And I've almost stopped wanting to learn things or because it really doesn't seem worth it to me. But now with AI, that landscape has changed. So perhaps I could more easily, like I don't have to learn Python to interact with a particular API anymore. Where I don't have to do. So basically, I'm trying to learn the landscape of what actually is possible and to filter through my head and figure out what I can do with it rather than having a specific business ideal that I'm like, I need this by 3 days. So that's my angle. I don't know.

So you're a fast, you're a fast developer, right? Um, focus on the website side of things. How much, um, vibe coding have you done?

Really not much because, because I, I, I have purposely over many years kept things simple, so simple that like, I don't even use JavaScript in my websites.

Yeah.

And I've made like a lot of money and I've sold a lot of crazy, I've made a lot of crazy tools with APIs, but I've kept it extraordinarily simple. And for example, or a better way to say that is like, I wouldn't use like jQuery, I would use vanilla JavaScript if I did use it. Or I don't use like CSS-bound, I don't use frameworks, I just use the actual CSS and things like that. Yeah, but I just prefer the simplicity of that because often things are over. And again, my main frustration with web work has been learning something, you know, spending months just, you know, brute forcing my way through something because I'm dumb. And then, you know, 2 years later it stops working and then I have to learn 3 new languages to figure out what's wrong. And I think that era is gone now with AI, but I— so I'm just seeing—

so very cool. Well, um, welcome, um, Pat. Excited to have you here. Um, this will be fun. This will be really fun for you. Evelyn, and then Tommy, and then Soraya.

Okay, so I am Evelyn. I— Jon spoke at one of my marketing classes earlier, I guess later last year in like November, and then he explained the course and I emailed him after and I just asked what kind of got into a little more and then he offered that I could take his course. I, so I am a student at UVU. My dad is like, an entrepreneur and he just had, he's had some very successful exits, but like as I've seen him create more companies, his, like all he does is like build things with AI and he like pushes it a lot on me and it's like very interesting to learn. And like over Christmas break, like he's not a nerd, but he just was like, okay, your job, like when you're bored, you have to make me this app for our family or whatever. And so I've done a little bit of live coding. Um, but I haven't really gone on to like all the AI platforms, so I'm excited to learn about that. Um, but yeah, I'm just a student right now at UVU studying marketing at the moment.

So what platform did he have you use to build the app?

I just used Lovable. It wasn't like anything crazy.

Yeah, but it was good. Lovable is amazing, guys. It's a great platform. I personally like Replit better, and there's a few reasons we'll go over that, right? But Lovable and Replit are very similar. I think Replit is more complete. But Lovable is awesome. And if you've done that, you're going to jump right in. It's going to be really fun with Replit. So excited to have you, Evelyn. Good to see you again. Tommy.

Hi.

Yes, I'm Tommy. I'm living here in Utah. I'm in Provo. I go to BYU right now as well. Right now I'm also working on a software company. And I don't know any coding at all. I've only kind of taught myself through AI. On how to code and how to build this software I'm trying to create. And I kind of hit a wall on as much as I could create without, you know, external resources. And so tell us, tell us about it. Yeah, so it's a software for contractors to basically create their bidding process within a couple of minutes instead of, you them having to measure out everything, creating the design. So the AI will create a Zillow-like model. And so then the customers can look at it directly and they can see exactly what it looks like. And then based on your location, it'll, it'll list out the expenses and how much it should cost. And then the contractor will be able to bid instantly through that.

What is the roadblock you've run into?

So creating the actual 3D modeling is the hardest part. So I've created everything else and everything else is functional on it. But creating an accurate model that can either be taken off of like a picture or, or even just measurements. It's just the AI isn't able to code that part or doesn't understand it because it doesn't exist currently. And so I'm trying to—

Have you tried using an MCP with Blender?

I have not.

Okay, that's going to be your next step. We'll talk about that later. So luckily my startup that I— so I, back in 2016, raised $13 million, built a tech company, sold it, that kind of thing, right? Um, it was called Seek, and we were a 3D company, right? That's all we did, 3D models, and we did it for the biggest companies in the world. So I actually know a lot about that piece, Tommy. So I'm excited to be able to, um, you know, share that with you. Uh, well, thank you. Um, good to have you, Tommy. And Soraya, last but not least.

Hi, I'm Soraya Hillam, and I live in Provo as well, but I'm a little bit older. I, after I was working with the company that I was with Jon with, um, I was asked if I would be interested in working for a small legal publishing company, and they wanted an IT person, and I didn't have very much experience in IT, but they said that they had people who said they knew what to do with computers, and it didn't work, and so they'd rather have someone who could figure out what was going on and do it. So I came on, I had to learn how to be able— they had like 10 different websites that I was putting in together and different social media accounts that I was trying to merge because my boss would just start up new ones here and there and everywhere. And so I got the websites down to 2 websites, and we grew— we've grown our YouTube channel from about 26 people, 26 subscribers, to about 36,000 at this point. My boss has been— he's a lawyer, but he speaks at different conventions and things where he sells products where you can do your own legal work, where you can be able to set up your own business legal formalities and be able to do your estate planning like wills and trusts and other things, or at least be educated enough so that if you were talking to a lawyer, you'd know what kind of things they needed to have put in there and be able to watch and see if things are done correctly. And so he teaches a lot about things. He's more of an educator than a lawyer at this point. And he's semi-retired. He stopped doing actual wills and trusts. On his own, but he still enjoys the education portion enough that we are able to put out YouTube videos almost once a week, usually on different legal terms or on different things to help people do it. He also has tax aspects because he works with someone who used to work for the IRS who does, who does like care for different small companies to be able to do any of their auditing or make sure that they're taking advantage of whatever legal tax breaks there are. So they primarily work with a lot of real estate companies, people, because that's who had the big conferences that my boss would speak at for a long time. And both he and the tax guy have their own properties. So they do a lot of real estate type things, but they also work with any kind of small businesses or people who want estate planning. And my boss is like, AI is the new thing. We've got to do AI. And I'm like, but how do you want to use AI? And he's like, well, we just have to use AI. So when I saw Jonathan putting this on his social media, like, I'm teaching this thing, I'm like, this is perfect because our website, because I did a massive renovation and putting it together 10 years ago, is now very outdated. And need. And I, it's been because we're working with a new marketing group who's been help, who's been really pushing things out on Facebook and other places where I'm not doing as much of the marketing, but, and I'm more tech support in the background with whatever, you know, just integrating whatever my boss is doing with this marketing group. And so AI is kind of a way to be able to be, well, what do we do going forward as my boss is semi-retired but still wants to be doing his stuff? And can AI help me just rebuild out the website and integrate with our different social media platforms and such in a way that's more seamless? And that can keep him going even as he needs to slow down since he's over 70 now.

Very cool. Very cool. So, no, that's awesome. I love it, guys. I love the broad range of applications here, right? Some of you are trying to build your own things. Some of you are in school still. Some of you have been with companies for a while. And some of you just want to figure out how to use AI broader, right? It's awesome.

Right?

This is— you're in the right spot, right? I can't think of anything today, any better type of, or any better topic you could focus on today than to get yourself up to speed. And I hope that this, this, this course, the purpose of it, or I hope the result of it is that you start to think about the world differently, right? You think you'll understand things in a little bit different way. The speed at which you can accomplish things with with AI, how you can, whether you're developing, you know, for Recyclops or Tommy building something brand new, launching a business, or even just looking at, you know, your job, your, you know, your industry, like, like Soraya, just to pick on your industry a little bit. Like, I don't need to go to a lawyer anymore to make a will. I don't, right? Now, does that mean that lawyers that focus on estate planning are worthless? Absolutely not., but maybe I don't have to pay them $12,000 for a will, but maybe I'll pay, maybe I'll pay 'em $1,500 to review the will that I made with AI and just be like, hey, what am I missing here? Right? And so the lawyer still gets paid their, you know, $750 an hour or whatever. Um, I, I get peace of mind that, hey, this is actually done right. Because even though the AI knows a lot, you still have to kind of fact check it. And since I'm not a lawyer, it's hard for me to fact check it. Now, I, I think I can, I'm pretty dang good at AI and I can use different AIs to bounce off each other and stuff like that. We'll talk about some of those things and ideas. But the bottom line is you can just approach your whole life differently when you understand how to use AI really proficiently. And also, again, like I just mentioned, when not to use it. So, a little bit about me, guys. So, I live in West Mountain, Utah, which is currently— I can't really see the Y, but normally I can see the Y over on the mountain. And Pat, since you're not local to Utah here, the rest— I'm in Utah, so my network's a lot here, and a lot of people see me., and that's why we get a lot of Utah people in this course. Um, but I have had people from, from India and from all over Europe and China. Like I've had a lot of people take this course from all over the place. So this one's a little bit more local, it looks like. Um, but anyway, just south, uh, south Utah County. Um, live on a 5-acre farm. Um, you know, I, I said I, I started a company, sold it. Um, I sold it 3 years ago in January, so it was my 3-year anniversary from from being— from graduating out of my company Seek that I founded, which was an augmented reality company and really focused on, you know, manage the SaaS company, managed 3D assets for companies like Walmart and Lego and Nestlé and Amazon and lots of other big awesome companies. And so we had a, we had a great company there and learned a lot, learned a lot. Ended up raising $13 million for that company because I'm not a developer, and I had to hire smart people like Spencer to come in and build my product. Because back in 2016, we didn't have AI. We didn't have those, those, you know, benefits. And honestly, I'm super jealous. I'm super jealous of entrepreneurs starting today. And, you know, I mean, I'm kind of jealous because I'm doing it now too. I was able to use AI to, to start what I'm doing now with GenAIPI. Basically, when I, when I sold my company in 2016, I said, hey, you know what, I want to just start kind of helping people and, you know, see what else is out there. And, and I started doing a lot of consulting, and I've always just been very interested in technology, right? Next-gen stuff, augmented reality, um, you know, obviously AI, Web3, the metaverse. Like, if you look at my desk, which might actually— my office is kind of a mess. I'm glad you can't see that way. Um, but I, I just have like devices. I've got headphones hanging off my, you know, Apple Vision Pro and I've got this cool ball that lets you look into it. Kind of like the, you can like put images on it, kind of like the ball in Lord of the Rings. What's that thing called? Anyway. Palantír. Yes. That. Thank you, Sariah. So anyway, I just have always loved tech and that just kind of led me to, you know, being really interested in playing around with AI a lot. Started getting really good, starting working with companies, helping them implement it more and more. And I'd be hired, I'd get hired for, you know, $20,000 or $10,000 or $75,000 for some big companies. Or I did, you know, long, longer engagements, things like that. And eventually said, you know what, I'm just going to start a company here where I help, where I help people learn AI. Like there's just so many people that, that need to know it. Like smart business owners that are doing, they're killing it. They're making $20 million a year and they've never even touched ChatGPT, right? They just know, they know nothing about it.. And I'm just like, man, you're, you're gonna miss out if you don't jump on this right now. And so that, that was the kind of bonus for starting this. I ended up giving myself a challenge to see if I could build some software around AI proficiency tests and AI training and courses and things like that. That was all gonna be automated. It wasn't gonna involve any live stuff. And anyway, I ended up starting Thursday, spent about 10, 12 hours a day, Thursday, Friday, Saturday, finished it. In a weekend, launched. And by the next Tuesday, I had my first $15,000 contract for that. And, you know, 6 months later, had $1 million, you know, of contracts and was able to really just do this, this process of entrepreneurship so many times faster and for so much less money. Guys, it cost me a couple hundred bucks to launch that. Couple hundred bucks as opposed to millions and millions of dollars that I had to do back in 2016. 2016 with, you know, when I did it the traditional way, right, where I had to, where I had to hire developers and, you know, work through things and I couldn't write, you know, 50,000 lines of code in a weekend.

Right.

And again, I'm not a developer, guys. I, now I have been the CEO of a tech company for a decade. And so I've been around a lot of developers. And so I kind of understand how it all fits together. And my goal today will be to, and over the next 2 days will be to teach you guys a little bit of that structure, how code works. At least how the structure works, like a front end and a back end. Very simple things like that. Some of you will be like, okay, I know this. It'll be very easy. But, and we won't spend a lot of time. We'll spend maybe 5 to 10 minutes just going over some concepts here that really will help you figure out how to better tackle AI. Okay? All right. So guys, we're right on schedule. 30 minutes in. Intros are done. Let's dive in. What we want, what I want to do first is we're going to dive into, I'm going to use Grok. You're welcome to use ChatGPT if you want. Want, you can use Grok. Um, uh, first you just watch me for a minute and then I'll set you guys loose. Um, so let's, let's jump into this. Let me open up a, um, um, let's see here. Let's get something pulled up here. No, I'm not going to use grok.com. We'll just go to x and Super Grok. So is X where you would host something for a more advanced section of Groq? Nah, no. In fact, using Groq on X is probably one of the weakest places to use it. You can actually go to like Groq or Super Groq. What is it?

Groq.com? I don't remember.

Groq.com brings up a prompt. Yeah, yeah, yeah. No, Groq.com works, right? And then, but I'm, I'm, I subscribe to like X Premium or Plus or whatever it is, right? Anyway, that's where I'll use it a lot of the time. But yeah, Groq is actually not a great place to kind of work on projects. ChatGPT is going to be a little bit better with some tools that we'll go over, or Claude if that's something that you like. And then Replit obviously is going to be a place where you're building more powerful things. But you'll see that a way that I teach, a pattern that I use as I'm teaching is I'll usually show you an okay way to do things, or even a bad way to do things. Then I'll talk about why it's bad, and then we'll do it the right way. And I found that that's been a very efficient and effective way to teach over my entire career. I'm a pianist and a gymnast and a kayaker, and in all of those different things I've gotten to very good levels. And I, and I've taught a lot. I've taught a lot of piano, used to teach gymnastics a bunch, about a decade, used to coach, and and I've run kayaking schools on the Provo River, actually, and I've been kayaking a lot, actually, out in California too. More northern— there's not a lot of rivers in southern Utah, but northern has a lot of beautiful ones. So anyway, I love to teach. It's something that I just get a lot of joy out of. The amount of money I'm making from teaching you 6 people today is way less than the rest of the things that I do for my job. And so I do this, and I'm going to keep teaching these once a month because I just really enjoy it. I love these 8 hours that I get to spend with you guys. You guys are going to get 9 hours out of me because I have to switch the schedule up. But anyway, okay, so here we go. Let's share screen. Share desktop 1. Okay, so guys, I'm just on x.com here. And so what we're going to do is we're going to do just a little coding exercise, right? Some of you are coders, some of you are not. Some of you are developers, some of you are not. And that's okay. It doesn't matter. You'll all be able to do this. You can do this again on Groq. You can do it on ChatGPT. Almost any platform will let you do it. I'm going to use Groq today. I'm just going to show you a couple of things that I work on here. So we're going to build with a fun front-end language. Front-end means that it's It's just display. It's just a display language. There's going to be no databases, no APIs, no talking to other systems or anything like that. It's just what we see on the screen is what we get. It's just a very visual language. I'm going to say, make me a mind-blowing visualization using p5.js and show me where I can deploy it. Yeah, let's just see that. Let's see what it says here. It's going to think here for a second. What's p5.js? p5.js is again a frontend language. It's just a display language. It's actually a type of JavaScript. It allows you to just display things in a certain way and it's a fun one. Anyway, let's see what it says. It's going to enable 3D. All right. Here's a mind-blowing p5.js visualization. So I just asked it. All right. It's asking me to do this. Okay. So first of all, it's asking me to do it in an online editor. More on that below. It'll tell me what to do, but it's actually writing the code. And by the way, when it's, when it's running, I can click this copy button right here. You can't do it until it's done writing. It's actually thinking about what the next word should be all the way down. That's how it works. Okay. It wants me to do something crazy here. I'd rather I'd rather do it. Yeah, give me the code in a way that I can use it on this link right here, editor.p5js.org. Okay. So, so guys, some, so like if you're in ChatGPT doing this, or if you're in SuperGrok, you can actually run the code right inside of SuperGrok or in ChatGPT. Okay. But in case you're working in some language that can't run right in the browser for whatever reason like that, you can actually ask it to tell you a place, right? And so here editor.p5js.org is a way to do that. So it writes this code and I'm going to click copy and I'm going to click on this link here and it's going to give me all this. I'm just going to delete everything I see. So I'll just Command+A, delete it all, Command+V, paste it. And I'm just going to click play and we'll see what we get here. Okay. Hold on. Allow all.

Okay.

It looks like I have an error. Okay. Syntax error identifier B has already been declared. All right. So there's a problem. So what I'm going to do here is I'm just going to copy this and I'm going to come back over here and I'm just going to paste that error. Okay. All I did is here's the syntax error. Error you're seeing is this. Great. You're trying to do this. Quick fix, clean it up. Okay. So give me the entire code again so I can just copy it. Okay. So what it did is it just gave me a little snippet of it. Okay, seeing this error, guys, is part of the plan here. I want you guys to be able to see how easy it is to just take an error, copy it, tell it what's going on, replace it, try again. Okay, click play. Hey, I got a mind-blowing visualization. Okay, I don't know what it's doing or what it's supposed to do. It looks like I can zoom in on it. That's a fractal. Yeah, yeah, it's just a fun fractal. Um, And it looks like since it is a fractal, I should be able to endlessly zoom in on it, except the zoom here is— it's literally backwards. When I scroll down, it zooms in instead of up, which is weird. And it's not centered. So again, it's not very good. But why is it not very good? Why do you guys think this isn't very good? Not specific. Yeah, exactly, Pat. I just basically said do something, but I gave it no instructions. Like, imagine I hire you, Pat, and I say I need a website, and that's the end of my instructions. I'm just like, hey, build me a website. Be like, okay, cool, but like, what's it about? And like, what style do you want? And why am I doing this? And are you going to pay me? And like, you know, I— there's just no instructions, there's no details right now. Fractal, for those of you that don't know, you probably all do 'cause you're all smart. It is an endless pattern. I will be able to zoom in like this forever. As I zoom in, we'll see this same pattern go forever and ever. And that's why the computer's actually struggling a lot to build it. So that's actually kind of fun. So it's actually cool. And I could zoom in and take another 20 minutes doing this, but I'm not going to. Okay? So let's give it a better, let's give it better instructions. Okay? So we'll stop playing that. We'll save the memory of the computer. So now I'm going to say, Build me— let's build me a— now, now build me an interactive model of the solar system. Okay. Now this is a little bit better prompt because now I've given it quite a bit of detail. Okay. Number one, it knows already because we're in the same conversation, it knows that it wants me to do it using p5.js. So I don't have to specify that again. Number 2, I want it to be an interactive model. Okay? So that now I didn't tell it exactly how I wanted it to be interactive, but maybe it's, you can click on it and you can see things and information, whatever. I don't know. Okay? And then the solar system is actually quite a bit of detail, right? And now, right now it's doing a bunch of, of research on, hey, what is the solar system? How far is Mars from this? And how, what, what are the relative orbit rates? And all these different things, right? And so there's actually quite a bit of information in this. I can still make it stronger for sure, but this is going to be kind of, we'll say this is our level 2 prompt, right? This is the level 1 where we're just like, eh, it's not that great. We messed up a couple times, had to fix a couple things. And then now, hey, now we're getting into something that's a little bit more specific, a little bit more useful maybe. And here you go. It did all this research. See, this is amazing, guys. Obviously you guys understand the power of AI. When I first started teaching this course, there were a lot of people that didn't even, had never even used AI before. And so they're like, holy cow, how did it know all the planets? Obviously you guys have been around it a little bit longer now, but AI can instantly do that research. And quite frankly, it's probably doing more research than it needs. I could turn this to fast. It knows this information already. So anyway, very cool. So paste the code. Let's see what happens. Let's, so I'm just going to click this copy button. Come back over here again. I'm going to replace this. Paste and play. And we'll see what we get.

Okay.

So, first of all, we're looking at it from the side. Okay. That's cool. I can at least do that. I only see— okay. There are some planets in there actually. They're tiny, but they're black, right? So, I got some feedback for it. I can't zoom in and out. I can't zoom in and out and I can't see the planets very well. Although I think it's kind of cool that it does have the light coming from the center here. Right? Although it's a black sun. So let's give it a little bit of feedback. The sun is black. I also can't see most of the planets. I'd like to be able to zoom in and out also. Okay? So I can just talk to it and just tell it in very Layman's terms, very basic, whatever. Right? There's also some other things I can rotate the view. I can speed up simulation. I didn't read that. I can toggle the orbit paths on and off with O, and then I can reset the speed with R. So again, I didn't give it very specific details about how I wanted it to, to do this. And so I kind of just have to take what it gives me. Obviously I can get very specific and build something really nice. Okay. Now we've got a full-on, full-on thing. Now I want to be able to zoom in and out still or rotate. Now I can't rotate it. Oh, hey, it looks like something's off here. Okay, anyway, I should be able to do the plus or let's see, maybe slow down. Yeah, I just slowed down the orbit a little bit. Now I could speed it back up. Now we're going really fast. Anyway, so I've built something that's kind of okay in my opinion. I can click R to reset the speed. There we go. That worked. O removes orbit paths. Now I just see this all moving. I wish I could, you know, kind of move it around a little bit more. That's okay. So guys, this is a very, very basic demo, but it's your turn. You guys all get to do this now. So I want you to go into Grok. I want you to use p5.js. Your first prompt should be something along the lines of make me a mind-blowing visualization. I'm going to drop this into the chat here, and I'm actually going to change it a little bit. Editor.p5js. Hold on. And I'm going to say, and format it, format it so that I can use it at editor.p5js.org. Okay. So I just, if you look at the chat, you can copy my prompt and paste it into, onto your screen. Okay. And, Um, but change mind-blowing visualization to whatever you want. Okay. Um, change it to, uh, Pac-Man, make me a Pac-Man clone or show me a model of this thing or make a little guy running away from dinosaurs or make Flappy Bird, whatever the heck you guys want. Make a little game, make something fun. Um, and then you'll get to share it. So you guys get. Like 3 minutes. 3 minutes to build something, ready to go. If you have questions, ask me. That's what I'm here for. As soon as one of you has something running, just say, I'm ready.

I'll have you share your screen and you wanna have the, the first setup that's not amazing yet first.

Yeah. Whenever you're ready to share it. Whenever you're ready to share it. Yeah. Yeah. It might take you an iteration or two. And yeah, I want you to do that. Right. I want you to be like, hey, add this, add that. I've got something at least. Let's see it.

Yeah, go ahead and share your screen. So make something that like follows along with whatever you're doing with the motion of the mouse. Cool. I just said for an amazing visual interaction visualization.

That works, that works. You wanna go ahead and share your screen, Sarah?

Oh, I thought I did.

Sorry. No, it's okay. I don't see it at least. Um, there we go.

Share.

There you go.

Forgot to, I forgot to hit share apparently. All good. Okay, cool. So you're, you click and then, and then it'll follow your mouse. Yeah, that's cool. You know, this is a really fun thing to do with your kids too, by the way, right? Like, my kids know that, you know, they're not allowed to play video games during the week, but if they come to me and they say, hey, Dad, can I use AI to make a game? I will say yes every time, right? Totally different, you know, style of using your brain and all that good stuff. So anyway. All right, who else? Sarai has beaten all of you so far.

Not a competition really, but— well, I really just did what, um, I'm a little bit slow on the getting tabs up and going back and forth, and I, um, I've actually recently switched to Apple, so I'm also kind of— you're new. Yeah, so, so like what takes probably everybody 3 seconds, I'm gonna have to It might take me a little bit longer. But okay, so I just copied yours.

But, but now I'm going to put something a little bit different, something fancy on it.

Fancy, fancy-ish.

Not even that. I have a little robot going here with my mouse, but it— let's see it, Tanya.

Make it more realistic. But it's— by the way, guys, this is not meant to be visually stunning, okay? This is not what you would use. This is not the 3D graphics engine you would use to create Fortnite, right? Or Call of Duty. This is used to make Pac-Man, you know? And, and sometimes it'll be like a block and a circle, or like a sphere, right? Like, it'll do very rudimentary shapes to make something happen. It's just kind of a fun front-end light language that you can do some really fun things with. It's actually— you can— I've seen some pretty neat things made with it.

But anyway, okay, hold on, it keeps following my mouse, so I can't get over here and share.

Hold on. Yeah, I went faster by using a second computer, which is why I'm in a different place.

Okay, let's see.

Can you guys—

you guys see my screen here?

Yes, we can.

My little— my little guy went away. Hold on.

There he is. So what is he supposed to be?

What's he supposed to do? He was supposed to be a Superman robot.

Oh, nice. Okay, I see the Superman. I see the cape moving. It's waving in the air a little bit.

Yeah, so I was trying to get him to go horizontal and then also a little more realistic.

But yeah, that's again, I think personally, I think it's impressive that it was able to create this little thing that kind of morphs this blob that makes it look like it's a waving cape. I think that's pretty creative. AI is pretty smart. AI is pretty smart. I love it. It's actually kind of cool. And it's very unique too, Tanya, right? It's not like something that anyone else would've visualized or whatever. Like you could use this almost as a mascot for something, right? Yeah. Because it's so unique. You would recognize that immediately. Be like, oh, that's different than anything I've ever seen. Character design is something that really fascinates me. So anyway, thanks, Tanya. Okay, cool. By the way, guys, this is secretly an exercise to see if you actually can share your screen. Because we're going to be doing that later. There we go. All right. Who else? Who else is ready? Spencer, I know you've got something there. This is easy.

This is a piece of cake for you. I told it to build the old Asteroids game. Asteroids. Oh, yeah.

Excuse me.

Um, so I'll show what it's rendering. It's, it's not, uh, it's having a hard time doing the key inputs, but that's all right. So let's see, I got a lot of windows open.

Let me find the right one. Yeah, all good. You and me both. I have so many things open. In fact, I should close some of these. This is This is ridiculous.

My computer is going to crash. I do giggle, John, every time you share your screen. I'm like, how does he know where he is?

Yeah. Not only do I have lots of windows, then I have like 40 tabs open on each, you know, instance of Chrome in each profile.

And I don't know. Okay. So, let's see the— see if this is better.

Here to play. Okay, awesome. Oh, nice. So you're like blowing up these things. It looks like it like breaks it down into smaller pieces.

That just makes it harder. Yeah, I, yeah, I think it can move it.

Then I just built this.

He's taking us through different types of, um, yeah, I mean, so it took 4 or 5 iterations, but, uh, pretty quickly got to the right spot.

Wow, wow, that's really cool. Again, guys, very simple, um, very simple, you know, to build this kind of a game or whatever, right? But, but so powerful to just understand that, hey, I can just put— I can just whip something up and put something together, um, almost instantly. So that's really cool. I love it. Um, okay, who have we not seen yet?

I can go now. You see me, but I haven't, um I, I wasn't doing any of that. I wasn't, um, I, I said like, make me a realistic dog, is all I said. I'm excited to see that dog. And, um, I'll try to share my screen again. I, go ahead, Pat. I'm very not literate with some of this. Um, I see the share, and it might actually So it might take me longer because I'm not sure what it's—

it's asking me for different things. Um, you mentioned that you had just switched to Mac.

What's that?

Did you just switch from a PC? Yeah, relatively recently, but I haven't done anything like what we're doing for 10 years or something. So it's just being on a Zoom call or something.

So are you trying to—

I'm trying to share my screen and it's asking me to open something and then—

Yeah, here's what you're going to have to do. Yeah. So what you'll have to do is you have to do what it says. You click that and it will ask you to quit Zoom and you have to rejoin the call. So you got to do that because we're going to need you to share your screen later, too. Again, that's the secret reason behind this exercise.

Okay. Well, I'm going to go ahead and—

Yeah, go ahead and do your reset. We'll have Evelyn demo while you just do that and come back.

Oh, wait, I'm seeing my—

Let me try something. Oh, hey, you got it. You got it. Let me see it now. Click Allow.

Yeah, just click Allow.

All right. Yeah, we're good. We're good. We can see your screen. Awesome.

Okay. And then where is the thing that I— oh, here it is.

That's all I came up with.

I love it. So while you all were making asteroids, I was trying to open Windows and like—

that's okay. Yeah, so I love it, I love it. Very cool, very cool. All right, awesome. Thank you. So, so look guys, it doesn't— this is not supposed to be, um, this is not supposed to be a crazy, a crazy thing here, right? Even just getting something running is good, right? You got past that. Evelyn, and then I think Tommy will be last.

Okay. Okay. I just did— I just asked it to— here we go. Just a volcano erupting and birds flying around it.

So nice. They kind of look like fish, birds, or—

Yeah, I tried. I know I was thinking about taking the birds off.

They're kind of concerning, but— Or like, what is that, a TIE fighter?

Yeah, I mean, more like a bat.

I don't know. Yeah, I don't know what it is, but it looks awesome. Cool. But you've got something functioning. You're— you told it to do something, it created something that actually works. Guys, I just want to be clear on this. It doesn't matter that it looks crappy or whatever. The fact that you're able to render something and it actually functions and represents a volcano, like That used to be something that only really good developers could do. To even put this together, it would take hours upon hours to put this kind of a thing together and think through it and get this stuff all right. There's almost 200 lines of code there. Now you can do it in a moment, in one sentence basically. Sometimes a one sentence prompt, boom, you've got something going. By the way, we're using the least powerful way to do this right now. We're going to jump in a second here into one of the most powerful ways to do this. You'll see how amazing we can make something happen here. Great job, Evelyn. You can stop sharing and then we'll go to Tommy last but definitely not least.

Let's see.

Can you see it? I think.

Yeah.

Yeah, we see this crazy horse flipping, but it kind of ended up being a little weird.

So yeah, it's somehow through a floor of sorts. That's awesome. But you're— I love that you're playing with 3D models again, right? That's the thing you're trying to work on, trying to figure this out. Exactly. Yeah, it's good to work on it. So yeah, that's, that's awesome. I don't know why it's spinning. Mouse wheel, zoom speed, zoom factor. Anyway, you could ask it. You could be like, why is it spinning so much, right? So obviously understanding some of those things. Very cool. Very cool. All right, guys, we're done with this, with this portion, and we're still right on schedule, done before 9. What's fun with this smaller group is I really get to involve you guys more. If I'm teaching 30 people, I can't have 30 people all, you know, share their screen. It won't work in this timeframe. And so I love, I love being able to do that. We won't all be able to share everything we're doing, but at least, you know, sometimes I'll just be like, all right, Tommy, show me what you got. Or, hey, we're working on a problem. Ryan, let's work through your thing. So, um, so anyway, thank you guys so much for sharing. Um, so guys, look, this is not how I would build something that I'm really trying to, you know, make awesome and share with the world. Not going to use p5.js. That language isn't super strong. Um, it's only front-end based. It's, uh, obviously very rudimentary in terms of some of its capabilities and display, you know, options there, all that stuff. And so, um, what we're going to do now, guys, is we're going to jump into Replit and I'm going to, um, I'll share my screen. Um, and you'll just watch for a minute here and then we're going to just do a little bit of, a little bit of discussion here. So, um, okay. This can be fun. This is where it gets really fun. And I think I'm going to just use this Replit home here. Good. No, you know what? I'm going to use my other, I'll use my other profile. replit.com. So guys, I used Replit when I built GenAIPI, right? And so my entire, everything that you've interacted with with GenAIPI has all been done. You know, it's all coming out of Replit. If you, you know, when you signed up for the class, when you got an email, if you've explored the site, everything is done inside of Replit., right? And, uh, so Replit's been a great partner. Um, they're, they, they've, uh, become a partner. I didn't know them when I first started, but then I started doing all these things, talking about it, and they're like, hey, we love that you're like doing crazy stuff and talking about all kinds of things. But I have, I have literally built hundreds of apps now, hundreds of websites and apps and different things, um, that are, that are really cool, right? Um, and, and I can, I can show off some of these later, but what I like to do every single time I do this course is I come up with a new business idea and I do it on the spot. So I have to think of something right now and I usually just look around my room to see if anything inspires me. Interesting. I was just looking at my candles. I got this huge candle for Christmas that's like massive from Costco type thing that'll burn for like, I don't know, it looks like it'll burn for like a month. I love candles. I always have candles. I usually have candles burning in here, but I'm not gonna start a candle startup. I could, I could, I guess. I want to do something that could be useful for me. What am I working on right now? Part of why I do this, guys, is I want to show that you can make money money with basically any idea you can imagine. You can just come up with something by looking at around your room and start something that will actually be useful. Um, I'm gonna do, uh, New Year's resolutions tracker, which actually I'm going to make a little broader, which is going to be a goal tracker. And we're going to make it the best goal tracker ever. Okay, so I'm going to start with design because I actually like this new feature, this design feature wasn't here in Reboot before, but I'm just going to say, um, I want to build a web app that helps me track my goals, but I want it to be better than every other goal tracking app in the world. Make it visually stunning and exciting. And make it super useful. Okay. All right. So again, back to say Spencer, our resident developer here who's really building awesome stuff. If I gave you this as your, you know, instructions to build an app, you'd be like, okay, you know, that's something, but I'm going to have to answer a lot of questions here, right? Like I'm going to have to fill in a lot of gaps. The same is true of an AI, right? The AI is doing something called chain of thought reasoning right now. Chain of thought reasoning is something that everybody does every single day. You just don't realize it. We're going to pick on Evelyn for a second here. Evelyn, will you teach us? And you probably saw me do this in your last class, but we're going to have you do it. So teach us how you brush your teeth in the morning?

Okay, um, I grab my toothbrush and then I grab my toothpaste and then— you're gonna do the thing where I have— okay, anyways, and then I put the toothpaste on my toothbrush and then I bring my toothbrush up to my mouth into where my teeth are and then I brush my teeth. For 2 minutes, and then I, I spit into the sink, and then I brush my teeth a little longer, and then I finish.

Okay, awesome, perfect, thank you. So what I asked her to do was just tell us how she brushes, brushes her teeth. Now when she goes to do all those things, she's not actually thinking about, okay, you know, now I have to do this, now I have to do this, now I have to do this. She's just going through this motion, this higher level task of Brushing her teeth. And actually, that is a task that fits inside of a bigger thing, which is getting ready for the day, right? I get ready for the day. When I'm getting ready, I'm— I need to, you know, I have to wake up. I'm gonna go to the bathroom. Maybe I'm gonna say my prayers, brush my teeth, eat breakfast, work out, you know, whatever all these things are. Okay, but each one of those things then has subtasks and smaller subtasks. So Evelyn just broke down very aptly how you would actually brush your teeth. And so in order to brush your teeth, you have to go through each of these steps. Now, when you're teaching an AI, like a robot, for example, how to do something, you do have to be specific, right? You do have to be specific. Now, you don't completely have to be specific, partly because the robot now has access to AI that understands all those steps already.

It already—

if you were to ask ChatGPT, teach me how to brush my teeth, it would be able to step by step tell you how to do that because it's already been able to absorb that knowledge from what it's been taught. Okay? But here, this, this thing was, was able to say, okay, I've got a high-level task of I need to build a web app. It's going to be goal tracking. It needs to be better than everything else and make it stunning and exciting and make it super useful. Okay, so here's my current focus. Here's different things. And yeah, look, it's done, right? I can click on my dashboard. Maybe the dashboard doesn't work yet. Habits. I asked it to just do a design, so it's not really functional yet. But to build something this custom this fast in what took maybe 60 seconds, right? As long as it took for everyone to tell us how to brush brush your teeth. That's how long it took to build this thing.

I think that's pretty amazing. Okay. So I have a quick question. Yeah, go ahead. Is the, design tab in the, in Replit, that's specifically what it's for is for mockups such as this?

So yes and no. You can absolutely use it for just a mockup if you want. You can stop here. But you can also click this little button here, click to convert to a fully functional application. So you can start, I like to start at the design tab and it's kind of a little bit of a technical reason why, but it's not that complicated. Inside of Replit, there are agents and subagents. One of the subagents is a design one, but currently it's only available before you start. So you can change the design after you get going, but the really powerful one like this is, is only available before you start, then you convert it, then you start bringing everything to life. Okay. So I could obviously click now and convert it. I could also say, hey, you know what, change the color from, yeah, change the theme from purple to royal blue. I'm a BYU fan. Okay. And you know, you can, you can iterate a little bit here and then at some point you can, you know, you can change that. Obviously we can make it a setting. For anybody to be able to set it to whatever they want. It could be their school color or their favorite color or whatever. We can make that an option. I'm going to have it just change the colors to blue and that should change a lot of what's going on here. Just kind of all these purple hues we're going to change here. So revising color variables. And again, it built a lot of code here. Hold on. Let's see. What it actually did here. Client script server shared components. It's probably just here. But yeah, here's kind of the basic stuff. Index route static storage schema. So it wrote— it had to write all these different things, right? Yeah, public index HTML, meta names, Twitter cards, source. It's got all these different things that it has. Now, a lot of it, it's pulling in because it just has kind of a template it's starting from probably. But then it goes in obviously and it's changing it a lot, right? So now it's blue. Okay. Go Cougars. Look at that. Go Cougars. I've updated the theme. So that's very fun. It was able to give me all that information and that's super useful for me now. Okay. So be limitless. Design your future. Stop wishing, start building. Transform your aspirations into tangible reality with focus and clarity, right? I can add a new Goal here. Let's see if this actually works. I want to, you know, dunk a basketball on a 10-foot rim. Whoops. It's actually one of my goals this year. Health and wellness. Create goal. I'm turning 40 this year, and so I have to do things that make me not feel 40. So anyway, it looks like, by the way, I don't see it. I don't see my new goal that I just added. So obviously there's no, there's no, uh, storage, there's no database, there's none of those things, right? This is just a design at this point. I can also, by the way, click over here and I can say, what does this look like on mobile? Okay. Limitless. The, the type, it's kind of a little crowded up here. Couple problems there, but whatever. Anyway, I'm not gonna get into any of these, you know, into all of this, these details. I'm not gonna finish this because it's really just a mockup at this point. Okay. Now again, guys, this is the wrong way to do it. Okay? This is what most people do though. Most people, when they wanna start their new app, they just come right into Replit and they get busy and they're missing so much. They're missing so much. And this is why they fail. They build this cool thing, right? Like this goal tracker. And then they're like, okay, I've launched, but I have no customers. No one's buying this. I haven't really figured out if customers want this or anything. They haven't, They haven't gone through the correct process of entrepreneurship. When you build a business, you have to validate, you have to follow, you know, there's a, a lean canvas, there's methods, there's ways to do these types of things. And AI can guide you through that process, but it can't do it unless you allow it to, unless you ask it to. Okay? So we're gonna leave Replit for a minute. I just wanted to show you guys and just be like, wow, this is awesome. Like, look how cool that is., but now we're going to go into ChatGPT and we're going to spend some time in there. Okay, so let's zip over to my chat. I was making treasure pictures yesterday. This is Midjourney. We'll get to that later. So guys, I have this. I started a company back in 2016, Seek, but the company that kind of, that I started actually earlier in the year was a company called Treasure Canyon. And so I started this thing where I basically just wanted to, uh, hide— I would hide treasure in the mountains and then put on events and people would go do it. And we had thousands of people come to these events and we made a good amount of money. And then we decided to convert that over to an augmented reality treasure hunt. And that's where my company Seek came from. And we kept pivoting until we eventually landed on a SaaS platform. But anyway, I started this— and this is just kind of an aside, guys— but I started this Facebook page called Treasure Map. And I would just like post random things about treasure on it. And somehow I have like 30,000 followers on it now. And people like, I just literally make pictures with AI and post them on there and I'll get like 1,200 likes on it. Like, it's crazy. It's ridiculous. Anyway, so it's just kind of a fun thing I do. Most of my followers are in the Philippines and they comment in Tagalog and I can't understand it. So anyway, again, all on the side. So welcome to ChatGPT. How many people here have not used ChatGPT? Okay, that's what I expected. A year ago when I started this course, a lot of people said I've never used it, right? So I had to teach a lot about it. Okay, now I'm not going to spend a lot of time teaching about how to use ChatGPT, but I'm going to talk about just a couple things here. We're going to talk about a word called context, and context is very, very important when working with an AI. Context is basically— it's the same as any context, right? If I'm talking to Tanya, about, you know, I just start talking about something, ask her a question, she'll be like, wait, hold on, what are you talking about? I have no context here, right? Like, you have to kind of set the stage, right? If you do that, then, hey, we're going to communicate a lot better. We'll be seeing eye to eye. So there's different levels of context. So for example, I've got, um, let's see here what this is. Um, yeah, yeah. So I saw somebody yesterday that asked ChatGPT, like, create an image based on how I treat you, consider all of our interactions together. So I copied their prompt and I just pasted it. Apparently thinks I'm a girl and that I like anime, which neither of which are even close to true. And then I was like, why did you make this? Right? This is the pattern of how you act. And I was like, but do you think I'm a girl? Do you think that I drink coffee? I asked this question specifically because it knows that I'm a member of the Church of Jesus Christ of Latter-day Saints and I don't drink coffee. Right? It knows that. I'll show you how I know it knows that. Right? And so anyway, it was— I'd had a fun conversation with it. But the point is, this conversation I had yesterday, I can come back here and say, okay, make one that is a little more exciting and focuses on adventure. Okay? So I don't have to tell it again. I don't have to give it more context. I can just pick up the conversation right where I left off. It has context of the conversation itself. So that's the first level of context. The second level is at the system level. The system actually knows things about me, knows who Jon Cheney is. It knows a lot about me. It knows that I am a member of the Church of Jesus Christ of Latter-day Saints. It knows that I love to kayak. It knows that I live on a farm. It knows that I'm the CEO of GenAIPI, right? And so let me go to, let me, sorry, I'm trying to move these. Your faces around here. Okay, I'm going to come here into— we've got personalization or settings, right? Both are usable. But personalization, basically in here you can tell it some things about you. You can do all kinds of stuff. You can say, hey, do I want you to reference saved memory, stuff like that. You also have data controls where you can improve the model for everyone, where you can let ChatGPT and OpenAI learn more about how you talk to it and stuff like that. It's not going to actually take your personal data and put it in there. It's just going to learn about how you act with it. Okay. But, but anyway, inside of these, inside of here, you have what's called memories. Okay. And it has memories about all kinds of different things. So, for example, yeah, there was a loft transaction. Things that I sold here. That's how I sold my company to Loft. It's actually talking about some older businesses that I made here. It says that I'm fluent in Spanish and Chinese. I've visited over 40 countries. I'm a kayaker. Right now, these are important things, right? Even like what religion I am, right? If I'm Jewish and I ask it a religious question, I would rather it know that I'm Jewish than, you know, a Muslim. I'm a Muslim or a Christian or whatever it is, right? I want it to know this because then it's going to put— it's going to respond to me in the context of what it knows about me. And so you can actually teach it things about you. A lot of times what I'll do is I'll come in and I'll say, hey, Chat, I want you to just know this about me. Save this to your memory. You can actually tell it that and it'll just add it into the memory. Okay? And that will help it be better. So if you're, you know, tell it what your job is. Tell it how old you are, tell it where you live, tell it the types of things you like, things you don't like, whatever, whatever it is that you want to go over, it will, it will make that happen, right?

Now I'm in a tent working on—

where did you, where did you access that information? Yep, right here in settings or in personalization, right? There's a couple places you can access that, okay? And then you'll get down to memory, okay? And then manage memories, okay? So, these are your saved memories. I can also say, okay, notifications for this Automation are currently disabled. Okay. I don't think I want that. I actually did want this. This was good. Once this text saved as their personal strategy guide, building scalable autonomous, this thing. Yeah. See, this is a fake one. We're gonna delete that anyway. So I, lots of times I'll be talking about stuff like, for example, this goal, you know, thing that we're doing. So we've got two levels of context. We've got a conversation, I can just talk to it and say, hey, and it's going to make an image for me. It's going to do whatever I ask it to do. And then 2, we've got the system level. And then 3, and this is where, this is one thing I really love that ChatGPT can do really well is projects. Projects will consider everything that you have inside of that project as you keep working forward, right? So I have the General AI Proficiency Institute. This is the project that I started about a year ago on February 27th, right? So I can literally scroll back everything that I've ever talked about. Everything I've ever done using ChatGPT is in here, right? This was my first idea, my first discussion about, hey, I want to start a business, right? This was it. I wanted to create a website that checks proficiency levels based on an intelligence test that shows the aptitude for an individual to succeed in a world with AI, right? Give it a couple more instructions and So I really started out just, I wanted to build a test. That was the idea. And then it got bigger and bigger than that. Right. But this is the entire process of how I did that. The conversation, how I decided to call it GenAIPI. I want the tagline to be good. I want to do this. This is what I'm going to work on. Right. But now I can go in there at any time and I can just say, hey, give me a new contract for X. Right. In fact, I did that. I just did an, I needed a, I had, we were hiring some sales, some new sales contractors that, hey, give me a new, you know, give me a new version of this. Give me the latest. And boom, I've got a contract ready to go for my sales contractors. Okay? So I can do that and it knows what my company is called. It knows what I'm doing. It knows the types of things and what my, what my products are.

All that kind of stuff.

Okay?

So, when you create these, if you're like, you're doing a contract or something, are you making sure that it has all the information it needs to first?

Well, yes. If I'm asking it to create a contract, well, here, I'll show you. I'll just look at this exact, at this exact conversation here. Okay. I asked that, I said, have I had you put together a sales contractor agreement yet in this project? I didn't remember if I'd had it do it. And so I just said, don't go into detail, just search or chat. And it's like, nope, you haven't. Okay, great. Let's create one really quickly. Here's going to be the, you know, commission, 30% here, 10% on this. And, you know, you know, this is in the in the state of Utah. I'm just kind of reminding it of a few things, right? Generally, I have Proficiency Institute Limited Co. is the legal name of the LLC. Payments will happen 5th and the 20th. You'll get paid after cash is in the door for any applicable accounts, et cetera, et cetera. So anyway, I had it make it, right? That's great. And then I had to tweak it a little bit. And, you know, I, I, I read it like I know what I'm doing. I know how to make a contract like this, but AI can do it much faster. And so then I scan through it. Okay, I like this. Hey, all right, now this is good to go. Give me this in a canvas. So this puts it in kind of a document that I can get in and I can actually edit. We'll be doing this later too, right? I can actually type here and change it. And so, um, so anyway, that was my process, but in less than 3 minutes, probably I did that whole thing, maybe 4 minutes, right? To build, to build that whole thing. So I give it a few details, but then it knows, okay, you're in the state of Utah. Okay, great. So it's gonna, it's gonna put any clauses that might be required in in the state of Utah for a contractor agreement, right? A 1099 agreement. And it knows what the name of the company is and it knows this and that and whatever I told it, the pay structure. So I just gave it a little bit of information and then it can go fast. Okay. So for today, we're going to call this Goal App. Okay. That's going to be my new project. I'm going to create a project. And then what I'm going to do is I'm just going to take the same, um, right here. Nope. That's the wrong one. Hey, come to our cool event. February 3rd, if you're here in Utah, at DreamWalk in Orem at, at University Place. I'll share more about that later, but, um, we're doing a fun event with Replit. Actually, we're hosting it together with them.

Okay.

I'm gonna come in here and I'm gonna grab this. I'm gonna copy my same prompt I used in Replit and we're gonna start from there. Um, I think this is not it. I, again, here I am now. I'm lost. Here we go. All right. Goal app paste. Okay. We'll use Replit to do this. Don't go into any detail yet. Just ask me what's next. Okay. So if I were to just tell it that, it would like literally start building it and building code. And I don't need it to do that right now. Okay? So, okay, perfect. This has the potential to be exceptional. I'm locked in on Okay, world-class goal tracking, visually stunning, actually useful, built on Replit. We're going to details. Okay, so here's what we're going to do now. AI is so smart that if you use it incorrectly, you can let it do all your work for you and you will lose your brain and your ability to think creatively. And quite frankly, the things you do and build won't even be yours. They won't, they won't reflect Tanya's, you know, viewpoints of the world or Evelyn's or Soraya's. And, and I want to hear about what you ladies have to say. I don't really care what AI has to say. I really don't. Right. I need AI to help me accomplish things, but I don't want some AI to just sit there and tell me about, you know, whatever it is. Like, I, I scroll past anything on LinkedIn, on Facebook, whatever. I can recognize AI writing in an instant, right? Tanya sent me something the other day that was a marketing thing. And, and I was like, this was written with AI, wasn't it? And she's like, yep, totally fine. Right. But, but I can recognize it. Right. It wasn't a bad thing, Tanya, for sure. But, but the point is, um, there's a time to use AI and let it write for you. And there's a time when we want to hear from you. You. We want to hear from what you're about. Now, can AI help you organize your thoughts? Absolutely, right? Can it help you take the things that you think and believe and whatever and help you put it in a way that's just better and easier to explain or communicate to the people that you're talking to? Totally, right? So we're going to do that now. But the way that I like to do that is I like to flip the script, right? Instead of asking AI questions, I'm going to have it ask me questions. Okay? So here we go. I want you to interview me one question at a time about this business idea. Let's figure out the name of the business, the colors, the vibe, the USP, that's the unique selling proposition, okay, so why am I buying this from you, why am I gonna use your goal tracker instead of somebody else's, The USP, the monetization plan, the marketing plan, etc. And anything else you need to know to help me build an amazing and achievable business plan. Okay, so now it's going to start asking me questions. And it knows why we're doing that. And this is going to be awesome. Okay. Yeah. And I also want to point out, I want you to interview me one question at a time. If I didn't put this one question at a time qualifier here, it would just give me 10 questions. And you know what we're really bad at doing, guys? Answering 10 questions at once. We really are. Who here likes tests? I don't really, right? What about conversations? I love conversations. Conversations are easy and natural. And so you, if you ask it to ask you one question at a time, it turns this into a conversation that's very easy to do. Okay, so question 1, the core why. When this app is working perfectly, what specific change does it create in your life? Finish this sentence as concretely as you can. If I use this app consistently for 6 to 12 months, my life would be different because I would be accomplishing the goals that I'm setting and I would feel more, um, I want to say accomplished, which I'm going to do, which I know is bad English because I used that already, but I'm just going to say it because I don't care. Okay, that's honest. We need to sharpen it. Feel more accomplished is the symptom, not the transformation. Okay, you want an app that actually helps you follow through, not just plan. So goals stop being aspirational and start being completed, which compounds into real confidence and momentum. Okay, see, obviously it said it way better than I did, and that's cool. We did that on purpose to let it do that. The failure mode, right? Okay, why do most goals tracking apps fail for you personally? Tracking becomes a chore. Okay, so now it's going to help say that a little better. Goal tracking fails when it feels like work about the work, a maintenance task that drains energy instead of creating it. I agree. Okay, that's already pointing towards a strong USP. Must counter— if tracking didn't feel like a chore, what would it feel like instead? Right. Progress. Powerful. Energized. Those are things that keep me going. Okay. Gives more than it takes. Agency. Control. Confidence. Visible momentum. Okay. So it's going to keep asking me questions. Now, I don't need you guys to sit here and watch me do this. This is where you guys get to jump in. So the next thing we're going what we're going to do here is I want you guys to take my prompt. Okay, I want you to build a web app and just say web app that whatever. Okay, whatever it is you want to build. Maybe it's some internal tools, Soraya, for you guys. Or maybe it's your whole business that you're wanting to build, Tommy, or whatever. Maybe it's just a specific thing that helps you build the models. I don't know, whatever it is. Okay. And then you can say, we'll use Replit to do this. Don't go into any detail yet. Just ask me what's next. Okay. And then, and then you'll do this. I want to interview me too. So I'm going to copy and paste these prompts. So go ahead and change that prompt that I just pasted to work for you. Start your project. By the way, if you don't have an idea, speak up now and I can help you come up with one really fast. I don't really have an idea. Pat, you, what do you like to do outside of building websites? Nothing. You just sit and build websites all day? Do you pretty much, do you game?

Do you watch movies?

Do you, I mean, uh, I'm sorry. Do you, no, I don't.

No.

Do you like to hike?

Do you like to hike? I love learning.

I love learning about things. So, okay. You love learning about things. Is there anything specific you love learning about? You like learning about tech, you love learning about travel, other cultures.

Um, hey, actually, I have an idea on that. So, um, you mentioned Philippines earlier. I used to live in Philippines, um, and I'm going to be moving to Southeast Asia, and I'm trying to decide between countries for various visa because I'm, I'm old enough now I could get retirement visa places. Nice. Which was not an option 10 years ago when I was living in the Philippines. So, for example, I think I'm going to do something about let's compare retirement options for Southeast Asian countries.

Amazing.

That's actually a super specific, useful app that I bet you thousands and thousands of people. Yeah, no, it's a great— it would be a great website eventually. So I love it.

It's one of my websites. So copy my prompt. Yeah, copy my prompt here. I want to build a web app that helps people that are considering retiring in other countries identify the advantages and disadvantages of whatever. And then here's why it's going to be different. Here's why I want to do it. And then, yeah, so just kind of modify my prompt a little bit so it says those things. That's perfect. So I'm going to keep doing my questionnaire here, guys. If you have questions, interrupt me and just speak up. By the way, guys, I didn't say this. Um, let's go for another minutes, 7, 8 minutes, 9:38, at 9:38 my time on my screen, we'll, we'll be done.

We'll be done with this, this, uh, exercise. So your hope is that we're just continuing with the interview that they're saying?

Yep. Yeah, just interview, interview through maybe 10, 12 questions, you know, I don't know, something like that. Um, just kind of answer questions. It's really fun actually, guys, when you're talking about your idea and having ChatGPT interview you, it becomes— I don't know, it's fun for me. It's energizing to talk about my idea and think deeper about these ideas.

So Does it matter if you didn't start it as a project? Are you able to save midway, have this whatever interaction you're having as a project if there's a particularly good back and forth that you really enjoyed or thought was a good idea?

Yes. So if you look right here on my screen, you'll see your chats right here. You can click this little 3 dots. Okay. And you can move to project. Yep.

So you can add something to a project very easily. I also forgot to make it a project. So that's—

That's okay. For me as well. Yeah. So you can do that.

You can do that after So when you're moving it to a project, could you be able to move it to an existing project or a new one at that time?

You can create a new one. Yeah, if you want to create a new one, then create a new one and then add it to that one, right? Okay. But if you want to add it to an existing one, you can do that too. Yeah. Okay.

And full disclosure, my boss wanted to watch and see what was going on, so he's been—

Awesome.

In the background. I'm not the boss, I'm the rookie.

Fair enough.

All right, guys, you should be pretty close to being done. Now, look, this exercise can be done for hours. You could spend 2 hours sitting here talking through something. And by the way, it's not just for, hey, I want to build a piece of software, right? You can do this all the time. If you're sitting there, you need to put together a new marketing plan, or, you know, we're, um, Tanya and I have worked on I'm putting together this fun event that we're gonna do for Utah Tech Week in a few weeks. And instead of, you know, telling ChatGPT all these things, you could just say, hey, interview me about this activity or this event and help me make sure that I don't miss anything. What are the things I need to make sure I do? How can I say this better? How can we make sure it's more inviting here? You know, and, and, and it will just help you organize all that. And then, okay, great. Put that all into a brief. Done. Right? And you send it over and it's, and it's a beautiful thing that's, that's pulled. This is the important thing. It's pulled the information out of your brain. It's organized it a little bit better. It suggested better ways to say things and unlocked ideas, but it's still you. You're the creator. You're in charge. And that's super important. Okay. So I know you might not be completely done with this, but, um, but anyway, I want you to at least have gotten through, you know, the monetization, what the name is, that kind of stuff. Right. So, um, Let's see here. I'm gonna, I'm gonna kind of do a cheat here. I want you to figure this out. Let's see here. I want you to choose for me based on how we can tackle the market in the best way considering everything else that's out there. Okay. So sometimes you can just say, hey, you know what, what is the market doing out there right now? Where does, where's the hole? Where's the gap? How can we make in this, you know, the monetization wedge is actually a perfect, a perfect thing here, right? Here's how we can do this. Okay. So paid, it's going to do these things. It's avoids this. Um, we got to define the user, um, high-performing individuals. Of all types. Okay, identity signal, opt into identity. Okay, great. People who don't follow through. Sorry, I just need to get to one thing. Okay, um, Let's come up with a name now. By the way, my idea changed quite a bit. I started talking to it and it was like, how do you want to make this different? I said, what if this just became something that helped you track like all the different things you do in your life so that it's not a goal that's just like a one-time thing, but a constant ongoing tracking of your life. So you like level up in life, which I've had this idea for a long time actually. So I'm excited to kind of work on this. Just like when you're in a video game and you get like, you know, you level up from level 4 to level 5 and now you have higher strength or higher intellect or whatever it is, right? In this, every time you play the piano for an hour or, you know, read a book or whatever it is, right? You level up your intellect, you level up your music game, abilities or whatever it is. Okay. Um, okay. Let's go with, we're just going to call it hero. Let's just call it hero. Let's just call it hero. Okay. Okay. So you'll be able to make your own kind of person that represents you, and then you'll have all your different traits and you'll set goals and track your life. Okay. All right. So I could keep, it will keep asking me questions forever. This is the last thing, what I want you guys to do now. I want you to fill in the rest of the blanks with what you think will be powerful and best meet our needs. I want you to now build a business plan that is 1,500 words and put it into a canvas for me. Okay, perfect. So we'll, um, I'll give you guys this thing here. I want you guys to do this right now too. Okay, so go ahead. I just dropped it in the chat. Um, just copy paste that and it's gonna build a business plan for you, right? So you've got your one line promise. Hero helps you level up your real life, right? Hero's premium AI powered life progression platform that turns real world actions into visible compounding growth. It's not a goal tracker. It's not a habit app. It's going to do all these things. Okay, so I've got a vision. I've got a target market. Entrepreneurs and builders, executives and professionals, creators and artists. It's not for these type of people. Here's the core product concept. Here's the input system. You'll be able to just tap things, be able to talk to it and say, I kayak the Provo River for 2 hours.

Okay.

And then it will automatically and intelligently assign things. Okay, so now we've got a progress and motivation system. We've got streaks and XP and levels and trait growth curves, goal contribution, AI differentiation. Here's how we're going to start making money. Got a social layer as well. Visual and brand design, design principles. Let's see here. AI differentiation, progress, motivation. Yeah, monetization strategy. We didn't actually give it a price. That's okay. That's okay. We could, we could go to market strategy. Now guys, here's what's really important about this. What we're gonna do now in a minute after we do it, we're, we're not quite done yet, but in a minute we're gonna take this back over to Replit and we're gonna give it this business plan instead of just a basic prompt, right? We're giving it the why behind everything, right? And so Spencer, as an example, you're working on Recyclops. Okay. Let's say you're kind of rebuilding something, you're building a new feature. If the AI doesn't have the complete context of the entire company, it doesn't understand, oh, this is a company that's already up, it already has customers, we're already doing this, here's why we're doing it, here's how much it costs, here's how we fit into the market. If it doesn't have all that information, it just has to assume, it just has to guess and fill in the gaps and fill in the blanks, right? But if you give it all of that information in the context of your next task or your next feature or your next function, then it's going to be way stronger, way more powerful as it starts to, you know, build other stuff. You can have it build 10,000 lines of code for you in a much better way. Okay. So, um, all right. Created a 1,500-word production-ready business plan. I've got now, and it's in this canvas thing, right? We use this canvas tool. What that allows me to do is I can download this document. It's a PDF. I can click edit. I can just copy it and take it over into Replit. We're not going to take it into Replit quite yet. Okay. What we're going to do here is I'm going to copy this. I'm going to copy this right now. And I'm going to take it over into Groq. Okay. So I'm going to go back to x.com here. And I'm going to go into my Groq and I'm going to paste my business plan. Okay, now I'm going to hold Shift, Return, or Enter, right, so that I don't send the message. I'm going to say, give me feedback on this business plan. Don't focus on the quality of the plan as much as you focus on the business idea and potential. Okay. I want to know what people out in the market think about this idea. Okay. Copy. I will drop this for you guys here. Now, why is this cool, guys? Why is this more useful here on X? X. So I just dropped that into the chat. You guys can do that. So copy that little, give me feedback on your business plan, copy, paste it into Grok. We're using Grok as our market research plan. ChatGPT, Claude, Manus, whatever, any of these, Replit, none of these systems have access to X. And X has billions of tweets that go out every single day talking about everything about your product, right? Anything you wanna learn about, people are are talking about it on X, right? You cannot come up with a business idea that isn't being talked about on X. And so you can do automatic research, market research with X. Okay. So now it's going to say, all right, this is, this is a good idea. Here's what you're trying to do. Here's, you know, what's going on. $2,200 after averaging for game changer, planner product to be fun. So it looks like there are a few other apps out there that are doing this. Here's some critiques and limitations. Here's the business potential. Here's some opportunities, risks, and challenges. Okay, so basically what I've done is— now obviously if I'm building a real business for myself, guys, I'm gonna take time and I'm gonna dive in deep and I'm gonna read this, I'm gonna ask it questions, I'm gonna follow up, I'm gonna ask it for, you know, for its sources and what, you know, what are articles might have some of those things. You know, it gives me some opportunity to do some things. But here's what I'm going to do here. I'm going to copy what Grok just told me. So just copy the text and I'm going to come back to ChatGPT and I'm going to say, please consider this feedback and update the business plan to make it stronger. Given these data points. Don't, don't forget everything we discussed, as those are my, my preferences and what makes it mine. Okay, then I'm going to go Shift+Return+Return. I'm going to paste everything I just got from Grok. And I'm going to go back here. So now what I'm doing, guys, is I'm bouncing back and forth between two AIs, getting its opinion on different things and having it improve my business plan. And so now it's revised, it's strengthened, and it's based on actual market research, right? It's not just my opinion. It's not just one model's opinion, right? And this is kind of like a model, guys, is kind of like getting your news from Fox News or CNN. Okay? There are actual biases inside of models. Okay? You can look at the same Donald Trump did this, and one side thinks he's a hero and one side thinks he's a villain, right? There are biases inside of different AIs. And so you need to, if you're really trying to build something that's going to be broad and really you're putting it out there, it's very, very smart to get 1, 2, or maybe even 3 different opinions from different AIs to see how they're training and how their algorithms work. It's not, it's not like going between, you know, Chrome and Safari, right? Those are two browsers that look at the same thing. ChatGPT is completely different brain than Grok. It thinks differently, it outputs differently, it's trained to do different things.

Question for you really quick. What did you do? I'm a little behind on the Grok. What did you, when you put it back into ChatGPT, what did you say?

Yeah, I might not have. Did I not paste that one? Yeah, my bad. My bad. Um, let me, let me find it. Please consider the feedback. Yeah, I didn't copy and paste this one. Okay. So guys, I know this takes a second and I know you might be getting a little bit behind, so I'm just gonna take 2 or 3 minutes, catch up and read and read some of your stuff, right? Read what it's talking about. So you should have taken your business plan over to Grok, asked it for some feedback. Read the feedback, copy the feedback, bring it back to ChatGPT, and then let it build you a nice new business plan. And now you should have a pretty dang good— you should have a pretty dang good plan, right? You can read through your plan, see what kind of things that changed. So real market validation, clear acknowledgement why they succeed and where they fail, sharper defense of our premium positioning. Explicit safeguards against infantilizing, tedious, and novelty burnout critiques. So we're trying to figure out, okay, how do we really make this something that people are going to love? So it took all the negatives that it found in the research and said, hey, let's make sure we, we actively try to avoid these negatives, right? That's cool. That's making us stronger already. Great company that came out of BYU was called Snap. And, uh, or sorry, not Snap.

Snap.

I said that wrong. Scan, not Snap. Snap didn't come out of BYU, but the connection will be there in a second. Scan was created by a guy named Garrett Gee. Um, and he learned how to code, right? It's back in like 2011, 2012. And as one of his projects, he said, hey, you know what? I'm going to make a QR code scanner. This is before QR code scanners were kind of embedded into your camera. You had to have a specific specific app, and he just said, you know what, I'm just going to look at all the QR code scanners that are out there, and I'm going to make my QR code scanner a little faster. If it takes 1 sec— if the fastest, best app out there takes 1 second to load, mine's going to load in 0.9 seconds or 0.7 seconds. And if the best scanner scans in 2 seconds, I'm going to make mine scan in half a second. And if mine— if this one has this font, I'm going to make mine just a little bit more readable, or whatever it is. He just said, I'm going to make every aspect of my app better than all the best things of all the other apps. And just incrementally, like 10% better, just has to be better. Right. And then he, you know, released it, got millions and millions of users and ended up selling it to Snapchat for $50 million. Okay. So, um, you don't have to do something new for a business, right? There might be, you might in your research find that someone else is doing what you were talking about, Pat. There might already be great websites that help you retire in another country, but yours is going to be better. And you can ask the AI to help you figure out why. Okay. All right, guys. So here we are. We're, by the way, guys, we're just a few minutes away from break time. Okay. First break of the day. I know I can go for like 4 hours straight and just keep talking and talking and talking. I know your brains are probably a little bit scrambled. We're not doing anything crazy hard, but we'll take a break here in a minute. Okay. So we've got our business plan. I'm actually going to edit it. And get rid of this revised and strengthened thing here. Hero business plan. Okay, good. All right. So what I'm gonna do is now I'm going to ask it for one more thing. So we've got a business plan built in a canvas. Now build me a feature, a, sorry, a complete list of every front end and Backend feature required to make this app work. It's— I don't want you to specify the technologies we use, for example, you know, JavaScript or Postgres, whatever. Just build me a complete list. Put this in a canvas. Okay. So I will give you guys this as well. So now it's going to think through, okay, I need to build out an app for me on this hero app, right? This document enumerates every major front-end feature. So great. We're going to need account creation, login, logout. Password reset, onboarding flow, intro narrative, initial character creation, selection of focus areas. All these, these are all screens and features that are going to have to be built in order for my business, my app to actually work. Okay. So it's building all the front end and then it will probably shift to the backend. Yeah. Front end user-facing stuff first. And then There's a lot of front-end stuff. This is a big, a big app.

Notifications and updates, visualization.

Trying to make it the best ever. Yeah. User and identity management. So now this is back-end. I need to be able to, as an admin, control user accounts, overall level calculations, XP thresholds, all these different things. So this is just— it's not the details of of all of these things. It's not all the code and all the calculations and all that kind of stuff. We just need to figure out exactly what needs to be done. And then there's a huge thing. Now guys, look, if you were building this type of a, you know, we'd call this a PRD, right? Um, right where you're, it's a requirement kind of document, um, that allows you to say, okay, what are all the things that this app needs to have? This is what I would hand to Spencer or to my developer and say, go. If I handed a PRD that had 30 different things in it with multiple 10 things in each one, they'd be like, okay, hold on, slow down. Right. Um, so anyway, pretty crazy, pretty crazy stuff. All right. So I've got all this cool stuff. Um, now I can, and you know what it says, final note. I don't want my final note in my canvas. I want, this is actually a pretty detailed one too. This is more detailed built in than other ones I've done. I'm just gonna delete all this stuff at the bottom.

I don't want— what exactly is a canvas?

Is it just like a document? Yeah, it's like, it's like their word for a Google Doc, right? You know, it's just kind of a document that's editable inside of this. Okay. All right. So I've, I've, so now I have two documents in here. I have my business plan and I have my feature list. Okay. Now you can ask it to create other documents. Maybe you wanted to create a branding document, a branding guide. For me, I'm probably going to ask it in a minute to put together all the calculations for how levels are calculated, right? Just so that we can have ChatGPT think through that. And I can just hand that to Replit. Now Replit could figure that out too. But for the most part, most apps, I'm just going to be like, okay, I've got a feature list and I've got a, and I've got a bit business plan. Okay. So what I'm going to do now is now we get to go into Replit. We get to go back to Replit now. Okay. So I'm going to come back over here and I'm going to start a totally new, we're just going to abandon this one. Okay. And I'm actually not going to use design. I'm just going to go right into app for this one. We could do design later. Um, I'm changing my mind on the fly here. I'm actually going to do design. Yeah, I'm going to do design. So what I'm gonna give it now is a few things. So describe the idea you want to design. Now, obviously, we're gonna use ChatGPT to help us here.

Okay.

Where is my— What are the huge differences between the two and why did you choose design?

Design is just gonna make it look prettier.

That's all.

Design is just gonna be a better initial design. That's it. It's the only difference. Okay. I'm gonna be able to do everything else. It's gonna be just fine. Okay. But here's what I'm gonna do. I'm gonna ask ChatGPT inside of my project, inside of my conversation. I'm just gonna say, give me a 3-sentence description of what we are building. Okay. So now it's just gonna gimme a quick— Hero is an AI-powered progression platform that turns this into this. Okay. Fantastic. So I'm going to copy this and I'm gonna take it over into Replit and I'm gonna paste it. Okay. 3 sentences. That's all you need. Okay. So I'm gonna give you guys this. I'll give you this, uh, prompt if you want it as well. Give me a 3-sentence description of what we are building. I'm sure you can type it, but you know, you can copy paste as well. So there's that. Then I'm gonna go up and I'm gonna grab my 2 documents. Okay. So I'm just gonna come up here to the top of my canvas of this feature list. I'm gonna copy it. I'm gonna go into Replit again and I'm gonna paste it. Okay. Shift, enter, enter. Don't push start. Don't let it go without doing this. Hold shift and it will, it'll let you do that. I'm just gonna paste it. And when it pastes, it's gonna paste like this. It's gonna look like an attachment. You won't see all the text because it's a lot of text.

Okay.

Is that what the shift shift does? Is it? If you hold shift, Shift, it makes it so you don't, so you don't actually send the message to the AI. Yeah. You just hold Shift and then hit Enter, Enter, Enter, or Return, Return, whatever. Okay. So then I'm going to grab my business plan as well. So now I've got a business plan and a feature list. And I'm going to say, please be sure be sure to follow all guidelines in the attached documents. Okay. And then, okay, so you guys should have a 3-sentence description. Be sure to follow all guidelines in the attached documents. And then you should attach your feature list and your business plan. And then when you, when you're there, you can hit start. I'm gonna ask my ChatGPT to do one other thing for me. Please put together a canvas that details, um, that, that, uh, let's see here, gives a detailed overview of how the level and skills system works. I could obviously spend a lot of time working through calculations and how all these things work, but for what I'm trying to do now, I just wanted to figure it out. This is amazing. I love working with AI.

It just blows my mind sometimes. If you want to have audio for it, I mean, you might want to be able to read, they might want to be able to hear it.

But they can do it. Okay, very cool. Alright, this is just about done. It's modifying my document here. So close. Done. Okay, sweet. Alright, sorry guys, I had to do that extra document, but this is something you might have to do on individual projects where you say, alright, I need to add just a little bit extra information here. So I'm gonna go ahead and click Start, and I've attached all these things. Okay, so once you click Start, you can sit and watch if you want, but now we're gonna take a break. Okay, it's 10:01, we're gonna go until 10— actually, it's 10:02 now. We're gonna go to 10:15. Okay, take a 13-minute break, get up, walk around, get some water, go to the bathroom, whatever.

Okay, we'll see you guys in a minute. So I'm going to put the audio and video back on.

I'll take All right, guys, welcome back. Oh, I just got a text message from stake presidency to meet with me and my wife, and they're redoing the bishopric right now. That's not good. I'm probably just going to be a primary teacher. Um, so much time for that too. Goodness. No, you know what's cool is it's a counselor in the stake presidency, which means I'm not going to be the bishop. That's good. I don't really have time for that. Um, anyway, I, it could be nothing. It could be nothing. But anyway, I just know that that's what they're doing right now. That's fun. All right, guys. How's your apps? How are your apps looking?

Are they pretty?

Trying to make them actionable means that I'm not through all of the different parts yet. Yeah. But are you sitting— does your screen look like mine right now? You're in Replit.

It's built the first visual version of it. No, I just got my business plan. It took a little while to put it together because I was still going back and forth on stuff that could actually be— that could work right now.

So I'm behind. This is part of why this is recorded. And so some of you might be a little behind. Does anybody have something cool they want to show that looks awesome?

Um, actually, I like mine. I— so my roommate is trying to sell her lease right now. And in Utah, like in college, it actually feels impossible. Yeah. Especially with people going on missions and stuff. And so like before I came out of the course, I was like thinking of like, okay, what would be like a good website or like app to build? And I was thinking like right now in Utah, all you can really do is like sell your lease through like Facebook groups and like, sure, random group meetings, like just like so random. So I was like, what if you create an app like Zillow or like Airbnb, but it was just for like college-age students in like Utah County, I guess? Yeah.

Um, Let me share my screen. So cool. Let me click stop on mine and I would love to see yours. It's a great idea and it's also very hyper-localized and specific, right? That's a great way to start, right? If you tried to do this like, hey, nationwide, list your thing, right? No one's going to use it. But if you can get like really good critical mass in Utah County and then, you know, expand to University of Utah and then expand south to St.

George and, you know, it's really cool. Yeah, it was kind of interesting. It was like, um, It's just, it's such an interesting, like, area because of all the missionaries and stuff that, like, leases are just always, like, either always constantly in flux, up for grabs.

Yeah, yeah, for sure.

Um, let's see it. Okay, this is— yeah, so this is kind of what it looks like.

And then I haven't looked at, like, the app version of it yet, but yeah. And how are you gonna make money?

Are you gonna just charge people, like, 10 bucks to list it or something, or— Yeah, that's kind of what I'm thinking. Well, you'll have two customers. You'll have the customers that are buying a lease and the customers that are selling. And so I feel like with college students, like, subscriptions are always so scary because we always forget to cancel them and then we're like, bruh. So I kind of like researched on ChatGPT, like, what's the best way to market to like a college-age student? And I think it would just be like a one-time payment. And I don't know yet if it's like you list it and it's a payment because I'm like, how would I get if I wanted to charge for like just being on the app to find a listing.

So I'm not really sure yet, but my opinion is lower the barriers, right? Yeah, lower the barriers completely. And so for users, that's going to be, you know, the people are looking for it. You don't want them to have to pay to go look, right? But I think you also want to avoid spam and like not real listings and stuff like that. So if you just charge something like $9.99, $9.95 to list, right, or whatever it is, right, or $19.95 for a premium listing, right? Maybe a premium listing gets you higher or whatever, stuff like that. You have the ability to You put more pictures, but I would just charge on the listing side.

That's just my opinion. Okay. Yeah, no, that's a great idea because it feels right. Yeah, I agree. Like you don't, it's like you don't even know what you're wanting to expect. Also chat came up with the idea like you could do, I don't know how this would work, but that was kind of interesting. Like if someone did sell a listing on there that you could get like a bonus off of that, like for being listed on your website.

But I don't know, lots of different ideas on that. Got it. Yeah. No, awesome. Very cool. And yeah, you've got something that looks good. It looks professional. It's clean. Like if I was a college student, I landed here. No problem. No problem. I would, I would trust this. Okay. We're going to talk about that a little bit more in a minute too. Very cool. All right. Anybody else want to show off their, their V1? We've got, you know, we're going to, we're going to dive in and make it a lot more V2-ish, but.

Mine is basically just a Basically just a home page.

It doesn't show much at all.

Okay, that's all right.

I don't mind sharing it if you want. Let's see. Let's see it, Pat. I'm interested in yours because we've been following along with your idea too. I don't know where it is.

Yeah, right there on that other tab.

Can you guys see my screen? We can.

Yeah.

If you just go to the right of your kind of at the top of the screen there, just click another tab where it says Retire Southeast.

Yeah, yep, there you go.

So it's just that, that's all it is. Nice, nice, I love it. Okay, then you've got Start Assessment. What happens if you click Start Assessment?

Does that do anything? I have not tried anything. Oh, there you go. So yeah, the concept was turning into kind of like what, what, um, what your options really are, not kind of what, um, for example, you might want to retire in Vietnam, but I, I personally know that the visas are only 3 months at a time. They don't have a retirement visa, so you have to go out of the country every 3 months, which can be nice. But yeah, it's like forced travel, forced travel, forced, uh, resets. But it's, um, but you can still get 12-month leases and stuff. But, but my point is, is that kind of rather than be the get-all, like 'Oh yes, here's everything about Vietnam or Thailand or whatever it is. You know, what is realistic for your situation?' So it mentioned kind of like a decision engine, and I went with that. I don't think I'd do that in the end, but it's nice that they have a little assessment here.

Well, I like it. I actually personally love things that have an assessment, right? They're also smart for saying, 'How young are you?' Right? It's going after It's, it's being nice to the old people, right? How young are you instead of how old are you? Right. Right. So you answer the question and you go through and then you kind of qualify.

You have to qualify yourself to use this. There you go. It's actually that has stuff. So yeah, what is, what is your income? Let's say $5,000 or something.

Yeah. And sure, liquid savings, say $50,000, $100,000.

There you go. And just Wow. So impressive. It's already—

I didn't do anything to do any of that. Yeah, that's so cool. I love it. Very cool. Thanks for sharing. All right, guys. Anybody else want to share? You don't have to, right?

But if you're just like, I want to show this, how do I end my screen share? I got a question just for you. If you all be kind to an old man. When I think I'm screen sharing, I would assume that when I look, I shouldn't be able to see my face. But there I'm seeing my face and apparently I'm still screen sharing.

No, you're not. I'm controlling you a little bit.

So I stopped your screen sharing.

Yeah, but, but while I was screen sharing, I still saw the thumbnail where I could see— you can still see your face. Yeah, it does show your face.

And that you can— how do you know? How do you know that you're— how do you know you're sharing your screen in that scenario?

Well, with Zoom, you have to start it, right?

You have to start it manually. Like, I can't—

there's no indication on my screen saying— yeah, normally there's just kind of like a little green outline. It's just kind of like a faint outline around the, around the edge that shows like what it's, what it's showing.

So anyway, on my Mac, the part of the screen that I'm sharing on the very top, it'll be like currently sharing screen.

It's in like a red box by the browser. Okay. Thanks guys. I'm sorry. No, you're good. You're good. All right guys. So I'm going to share my screen again. We'll jump back in there and yeah, it says participants can now see your screen. I can see that at the top and I see this little, your screen sharing thing. With the controls. So, so here's my kind of level 42, my kind of hero of my life. I can, you know, if I, if I go kayaking, right, it adds to my strength skill or trait or whatever, right? And I can do other things and I have goals that I can set, um, you know, reach piano grade 5, do this, write 50,000 words because I'm a writer, right? Whatever. And so that would be towards my intellect, right? Coding, anyway, writing, intellect. Yeah. So that's really cool. So none of this stuff really is put together. Because again, this is just kind of our visual mockup. This is the design. We're still in that phase right now. I actually am curious just because I always like to see what this looks like on mobile. It looks like it did pretty good. It's got a decent, it's got a decent balance here. I like the feel of it already. So not bad, not bad. Okay, so now what I'm gonna do is I'm going to click to convert to a fully functional application, right? So I I can click that right here and I'm going to say, all right, send additional comments to agent. Be sure to look at the attachments that I gave you in the initial, um, app request, app design request, as you put this all together. Now I'm going to add, I'm going to give you guys this this prompt. Also, do not create a Replit login, just create a JWT, a standard JWT login. Okay, copy. Yes. All right, so I'm gonna give you guys this prompt. You guys can do the same thing. Most of your systems will require some sort of a login. You need a login in order to save like your user name or your emails or your progress and what level you actually are in my app, right? Or if your, your application or your, all of your details that you do, Patrick, for yours, okay? Evelyn, yours is obvious, right? You're going to have a profile as a seller or as a buyer, and you're going to need to be able to save that. And so what you need the system to create is a database, and it needs to make all the buttons work, and it needs to save the information in, again, what's called a database. That is a place where all of your information is going it's going to be stored so that if I log out of here and I log in on another computer, my information is all there because it's pointing at a database instead of just trying to save it in the memory, which is what it's doing right now. When you first build an app, usually it's going to build something called memory storage. So if I do things, if I add, hey, I like this, I like this, I like that, it might save temporarily. But if I were to click Refresh, all that information will be gone, or if I go to a new browser, all that information's gone. What's a JWT login? Yeah, JWT is JSON Web Token is what that stands for. And yeah, it's basically just a, a standard login. You're gonna have a username and password. That's as simple as I could say it, right? But Replit has its own login system and I hate it. It's the only thing that I don't like about Replit.

It's just weird. So I always say continue with Google for like 5 minutes now. For what? What do you mean? I was trying to be able to build something, put the stuff into Replit like you said. But it wanted me to sign in. So, I tried signing in with Google.

And it's still thinking. Well, refresh it.

Did you sign up for Replit?

I had not ahead of time. Okay. Do I need to do that before I do this? Yes, yes, absolutely. Otherwise, you won't have an account and you won't be in the database and it won't save anything.

Right.

Yeah.

It was trying to— when I tried to start, it tried to prompt me to get one, but then it's just continuing and continuing. I'll take out the prompt and then I'll try to create account first then.

You've got all the prompts in the chat so you can catch up. Now, while we're sitting here, guys, we're going to watch a little— while this is working, okay, we're going to watch a little video really quick.

Eric Schmidt.

Can I ask a technical question real quick?

Yeah, go ahead. Go ahead, Spencer. On the database side, is that just a database within Replit and how easily accessible is that data?

Those types of things. Great question. So yes, there is right now. So inside of Replit, you've got this little plus button, you got all kinds of cool stuff. This is why I personally love Replit. Like if you build something inside of, you know, Cursor or Lovable or any of these other systems, you usually have to go connect up to your own database, maybe Supabase or, you know, AWS or whatever the heck you want to use. And you could totally do that. And by the way, you can do that with Replit as well, but there is a built-in database, and it's actually just a Neon database. It's just using the Neon system, but it automatically connects everything for you. And so I've got my database, I can click on here, I can add tables, I've got data, I can do— I can change things on it. Um, I can, I can change things on it just by telling it, hey, update the database to do this, this, and this. And then when I do publish it, it will create a mirrored production database. So, so then you'll have two databases that you can manage, and I can go in and I can, you know, I can edit an entry here. If I wanted to add a record, I could do that, and— or I can edit records. I can do all kinds of stuff.

Yeah, so you have control. But sorry, I have a question too. So let's say I do make something, I have a database, um, I go to publish it on a web host I'm choosing, and it only has an option for your basic database. Can't even think of it right now, but, but can you, can you op, can you make it whatever database you want and tell it to do that?

Or so in Replit, you can tell it to do whatever the heck you want. And the way that you would manage that is you're gonna have, you have something in here, we're getting a little bit ahead of ourselves, but it's okay. We're kind of, we're kind of right at the edge where we can talk about this. We actually have this thing called secrets, right? It's gonna store sensitive information like API keys securely in your app, right? And so you have the ability ability to come in here and say, all right, here's my session secret. And then in a minute here, we'll actually have our database secrets as well. Right. And so you can link different things to it and you can really just click here and say, hey, add this to the secrets. Hey, I want to change the database to be pointing at this one. And then you'd give it all the information about the database and APIs and whatever keys or credentials you need to enter. It won't allow it to, to store it in this secrets tab.

I assumed you could do whatever, but I, I was just curious.

It's very flexible. Very, very flexible to be able to connect to anything. Like later on I might connect to Stripe and then connect into, you know, Twilio or something for text messages or whatever. But you can do all of that. You can connect all these APIs and different systems in here very easily. And you'll manage a lot of that through the Secrets tab. Go ahead, Tanya. Do you have a question?

Sorry, you were waving at someone. You were waving.

Oh no, I was, I was putting my hands up because it's just taking so long.

Got it. That's right.

Everything is very slow today for me. So yeah. So anyway, no, Replit, when you get into like where it's actually building the app out, this might take, this might take 10 or 15 minutes. Right. And so we're gonna let that just kind of work in the background for all of you guys. If you're still kind of working on it, Sariah, that's fine. Um, but I want you guys to pay attention now here for a moment. If you can, we're gonna watch this little video. It's kind of pixelated and it's from about this, this video and, and Eric Schmidt here, he's the, the former CEO. CEO of Google. He's chairman of Google, I guess, former chairman of Google, um, billionaire, um, very tied into Silicon Valley and everything that's going on.

He said this, I think, I want to say roughly April 22nd or something of last year.

Okay.

Um, sorry, really quick, can I ask a question? Um, when you are like creating an app or website or whatever, um, once you see how it is and you want to make changes, do you always go back to like AI or an AI platform and like put it in there and then sync get back into like whatever you're creating it on, or do you just like edit through that?

Yeah, we'll get into editing after we get all like the kind of first real working version up. But yes, I'm using like for genaipi.org, my website. Anytime I need to make a change, I open up my Replit project, I tell it to do something, and it does it. Yeah, so you're constantly using that AI to have it do it. Yeah. Got it. Okay. All right, guys, it's about 2 minutes long, 2 and a half minutes long. And yeah, just listen to what he says here, and then we'll have a little discussion here.

This is discussion time. 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 one year. Okay. What happens in two years? Well, I've just told you about reasoning, and I 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 there are now somewhere around 10 or 20% of the 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— I want to make the point here is that in the next year or two, this foundation is being locked in, and it's not— 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 There is a 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.

Okay. All right, guys, what do you think? What kind of thoughts come to mind as you listen to Eric talk here?

I'm actually kind of reminded of, like, whatever level the field is, everybody has access to it, and everybody has excellence or laziness. And I think that there's still— I guess I'm thinking of it as a competitive thing, but I never worry about like those broad statements of these large ideas because whatever occurs, I believe there's always going to be personal excellence and laziness.

And I think people with personal excellence will do better in whatever scenario than people with laziness. I love that answer. You're the first person who's ever said that. I've asked hundreds, if not thousands of people that exact question, and I think it is— that is the secret to success, guys. It's not based on some government-funded program that's helping you, or someone's figuring out some solution for universal basic income because no one's going to be able to work because AI is going to take our jobs. It's whatever. It's— look, We're all in the same system. We're all living on the same planet. We all have access to the same stuff. The real difference is what are you doing about it, right? Are you going to strive towards personal excellence? Are you going to sit back and try to let someone else take care of you? Because it's going to get really bad for those in the latter group. But anyway, excellent point, Pat.

Thank you. Other thoughts, guys? I was going to say something. I'm pretty sure, Jon, you said this. I'm like, Could— yeah, I think you did. But when you were teaching our class, you talked about how it's not going to ruin our— the food chain of people and AI. It'll just be a different type of the way society works. And my uncle the other day— I mean, in college, you have to take all the random classes. Right now, I'm taking an Excel class. And I was over at his house. And he was like, what are you working on, whatever? And I was like, Excel, whatever. He's pretty smart with AI and stuff. And he's like, well, in, you know, 1 month, 2 years, whatever, you're not going to have, you know, AI can all do that all or whatever. And I was like, it's so true. I think it's hard being a student knowing that like the future is just so like undecided and like even like professors, it's almost like I'm learning AI the same speed as my professors, which usually doesn't happen in college. You're not learning the same subject the same time you're professors are. And I was like, you're right. Like, that's probably true. But like, in 2 years, like, there's— I think there's going to be a big difference as to who learns and who knows how to learn and like use their own mind and who just use AI to like create a mind for them that's not even their own opinion, I guess. So I think that'll be kind of the difference. But it is so— it's so weird being a student right now and being like, yeah, like what—

What do I really have to pay attention to and what's not really sufficient for the future? But yeah, totally. Like, what, why am, why are you teaching me about, you know, how the Earth's crust works? Like, I don't need to know that in physical science, right? Like, I don't need to know that if I'm going to go be a marketer. Like, yeah, it's cool. It's good that we know how to learn and we learn how to critically think and those things. And I'll talk about that in a minute, but, but, uh, but you're right. It's very interesting. It's a very interesting time to be a student because you know more than your professor does.

At least you think you do, right? Yeah. It's, yeah, I was, I graduated like 2025 from high school, but during my, um, I graduated, I graduated from high school technically 2 years early and got my associate's. So I just went straight into like my bachelor's, but it's like crazy seeing my roommates that are doing like their basic, like English, you know, kind of like what you're saying. And I'm just like, This is in like especially classes that don't allow AI.

It's like if you're not allowing AI, like you're like, that is, it's just crazy. You're hurting us.

Yeah.

Like, yeah, it's not benefiting us. Yeah. When you get into the job market, you will be required to use AI. If you are not doing, if you don't know how to use it efficiently or effectively, you're gonna, you're gonna get behind. Right. I actually, I actually posted yesterday on LinkedIn. I'll pull it up just to read it so I get it right. But It was a teacher dealing with the same problem saying, hey, you know what, you know, how are we supposed to, how are we supposed to deal with AI? And, and they said, oh, here it is right here. Okay, so this was the post. I recently heard about a teacher who instead of trying to circumvent students using AI, which is impossible, she made assignments by going, ask ChatGPT to write a report on this subject and then research how and why it's wrong, right? And so that's cool, right? That allows you to really critically think. You're using the tool, you're knowing how to do that, right? I mean, those are the types of assignments where you should, you should be like, okay, this teacher gets it, right? They understand we need to know how to use AI in this, in this world.

So yeah, great points. Thanks, Evelyn. Jon and I— Jon, actually, when he first spoke to me about coming intelligent AI, he was talking about age demographics and what we've seen technologically advance over his lifetime and my lifetime. And the most interesting thing about what we just listened to is this concern about AI overtaking humans and what we're capable of. And I think as you've, if you study the history for younger demographics, or if you've ridden the waves with our older demographics. You can see, um, even, even if you go back historically, like the Romans creating, um, ways to get rid of their waste actually advanced their society, right? It didn't take away from it. Societies that didn't keep, keep up had the plague and all sorts of things, right? So the whole idea is the advancement of technology. You don't necessarily always have to be on the edge of it. You do need to be an early adapter and evolve with it. But the internet improved our lives. It didn't take away from it.

Right.

And there's— it has started to level the playing field in a lot of areas where a lot of people don't have either the exposure to it or aptitude for something or have missed something in their life where they don't have advantages that other people may or may not have. And AI, just like I said, levels that playing field, but also helps you to adapt a lot quicker. And it's those people that don't adapt or are unable to learn or adapt or be exposed to it that are more so than not concerned about it taking away from us. Anything we know about humans is we always rise up. We always rise up and we always help mitigate most risks to our group, right? So all— I, I love the idea that people are talking about this and where it could go and the risks of it, but I also love that we understand that so we know how to implement it responsibly and also to our benefit. So when I listen to these types of conversations, for me, I'm always so fascinated by, ooh, that could happen. So how do we pivot that and iterate for our own success? So I, I just, I love all of this, this new, the acceleration of growth that is happening with AI. It's really, really cool. To your point, Evelyn, what you were saying about already being halfway through your bachelor's while your friends are still doing English 101. You know, I did the same thing in my education, and I was sitting there thinking, oh my goodness, why aren't people just like so passionate and eating this stuff up and, you know, growing exponentially? And I think that's the— AI is what is making that huge difference right now. Such a cool tool. It's like what Jon said, it blows my mind what we're able to do now. All these things that I used to imagine, I could just— I could be like, I'm imagining this, and it helps me make it come to fruition and execute on it.

It's the coolest thing ever. I love it. Yeah, I'd like to say one more thing if I could.

Right.

It, it might be useful to everybody. I mean, I'm an old guy, but something that's unique about me, I would say, is that, um, I grew up homeless and I, I have a 4th grade education. I never had the opportunity to go to school. I never had the opportunity to, um, to do any of the things that you guys are discussing. And like, for example, my first website was made on a, in a free college lab. And like, you know, we're like a library, just I didn't have a computer. So, so like I couldn't even like save images or, and this is a long time ago, it was like 25 years ago or so. But like the, but my point is, is that the, one of the things that I've always noticed is that People don't seem to— people kind of assume, yeah, everybody goes to school, or everybody has a roof over their head, or everybody has a computer. And these things are not true for half the population of the Earth. And so it's very— so one of the wonderful things that I think is coming out of AI or this access to information is basically a lot of people with a lot of different perspectives are going to be able to see things. Whereas under the current system, or what's been happening, let's say, the last 20 years, it's mostly people who have things that are like rich things, like computers or like homes or, you know, or countries that are nice. And the world is full of other perspectives. Who have better opinions on a lot of things than we might. And so anyways, it's just, I love it.

I just love the opening of access to anything. So, yeah, I, yeah, it's really fun hearing you speak, Pat. You have a very unique background and it's fun to listen to some of the way that your brain kind of pulls things together. But access to anything is cool. That is cool, right? Like I, I'm a musician. Right? You can— I've got music stuff around me, I've got a piano behind me, I've got 3 pianos just in this room. But, um, I, I think it's really cool that I today can record a new piano album without having to go to a studio. And I used to have to go to a big expensive $2 million studio with a big $100,000 piano and, you know, all this amazing control room. By the way, I love going to studios, I think it's super fun, but I can get just as good quality now today here. The access has been lowered and now I can put out amazing content and whatever. Right now I have to fight against AI music, which we'll do more of that tomorrow. But, um, we'll do some of that tomorrow at least. But, uh, but anyway, yeah, lowering access is, is a, that's progress in the world. Now it is going to cause some problems, right? We've got Soraya's boss sitting there, or I guess underling, as he said.

He actually wants to make a comment, if that's okay.

Oh, I'd love to hear from him. I'd love to hear from him. He's a lawyer. I was about to talk about, hey, we don't quite, we don't have to go to lawyers as much as we used to because we can get at least fast things done. My whole life has been screw the lawyer industry.

But I'm really old, Pat, and I have seen technology come before the TV.

Up through everything.

And I can envision technologies.

I spoke on the internet back in the early '90s and told a big group in Vancouver, Washington— or Vancouver, Canada— that you'd have a.com on the back of your name, that you would sell more things over the internet. And they laughed at me so hard I had to quit speaking. And between AI and the blockchain and frankly, the self-driving car, you're going to see more change in the next 10 years than I have seen in my life. Yep. I agree with that. It's not slowing down. And guys, you can take, you know, you can be afraid of it and say the world is falling, the sky is falling, right? Or you can say, hey, look, we've been through this. We've been through the internet change. We've been through the industrial revolution. We've been— humans are very adaptable. Now, I will say though, I will say there are some changes happening right now, guys, that are going to cause pain. Right? For example, I also posted about this on LinkedIn maybe just yesterday. In 2024, 168,000 jobs on average were created every month in the United States. In 2025, 50,000 jobs were created every month. So went from 168,000 a month being created to 50,000 being created. Who knows what's gonna happen this year, right? Might fall. And I know there's other things that impact that. There's tariffs and new presidency and all kinds of, I mean, there's all kinds of things that affect that. I understand that. But the reality is AI is going to impact the job market and some of these lower level entry jobs, right? Like if I'm Evelyn, right now, I am a little bit nervous, right? Because I got to go out in the job market where it is cutthroat to try to get one of those jobs right now. If I'm Spencer and I'm saying, okay, how, you know, Eric Schmidt is saying that all developers are going to be replaced by AI, okay, great. The reality is though, Spencer, someone still has to manage the technology at a company, right? There are jobs. You just have to figure out how you fit in this new paradigm. It's all going to shift. But a lot of people are going to be scared and they're going to rebel and they're going to try to just go against it and they're going to not have a way to eat. And that's going to be tough. Our society is not really prepared for the speed that—

sorry, I don't know your name.

Soraya's boss.

Sorry, say that again.

Lee Phillips.

Lee. I have a stepson who got laid off the week before Christmas.

Larry H.

Miller laid off their entire loan processing division, and it's all AI now. Yep. It's real. It's out there, right? And there's, there's going to be companies that grow because of AI, and there's going to be companies that shrink and optimize because of AI. It's going to change, but it doesn't mean that there's nothing to do, right? The problem that Evelyn is tackling right now, the problem that Pat is tackling right now, Those are real solutions that people will pay for, right? I don't care if you have AI, you still need to find a place to live, right? And maybe Evelyn can add AI into this thing that it can help match you up. Hey, what do you, what are your, you know, preferences? Hey, based on all this, I think you're gonna love living in this spot cuz it's close to this and it's close to that gym and it's close to this, whatever. There's a pool here. I don't know, right? There's all these different things that you can, you can use AI to make your, your company stronger if you just know how to look. So anyway, guys, thank you for participating in this discussion. It's one that I encourage you to continue to have amongst your peers, right? Your family, your co— your colleagues, whatever, online, on LinkedIn, on Facebook, whatever. Talk about it and just say, look, I don't want to say good or bad, I just— let's talk about this. How do we deal with this? How to prepare for this? Hey, I thought about this. How are we gonna deal with that change? But unimportant employment is going to go up. It will. And so, um, my personal opinion, and I'm, I'm biased by my experience in life, is if you want to be safe in the job market, make your own job, right? And I don't mean you have to not work for someone. Like, for example, Tanya, you know, works for me, right? Works with me, right? Get into an an organization where you guys control your destiny, and if you do your job better, you're going to make more money for everybody, and you're going to make more money, right? I mean, there are ways to work inside of a company that where you're taking control of your, of your future, right? If Tanya, as CMO, as a marketing person, is able to generate more revenue for our company because of the things she's doing, using AI, by the way, to accelerate all that stuff, there's no way I'm gonna let her go. I don't want to have to deal with that. She's be awesome, and she's going to grow with us, and we're going to go for a long time. And I do expect that. But the point is, guys, the world is changing. The sky is not falling. There's so much more opportunity now because this is an entrepreneurship class. I'm going to zoom in just for a moment on this. When I built my company Seek back in 2016, I needed an app built. I had built these treasure months, right? And now I needed to be able to do Pokémon Go style treasure hunts. We want to do augmented reality treasure hunts. That was the idea in 2016. That was when— that was the summer Pokémon Go came out. And I said, oh, we should do this for that, right? So I had to go and I had to find a developer, a co-founder, and then we had to go get some money to hire a couple more developers to build out these different systems. And my first raise was $20,000. I got $10,000 from my dad, $10,000 from my co-founder's dad. And then about 3 months later, I had gone and I'd sold about $30,000 of contracts for people to advertise inside of our app. And I went, I raised $500,000 by the end of the year, right? So in the first 6 months, raised $500,000 and we built a crappy MVP. It worked, but it was just garbage. Okay. That was even the $20,000 is already a risk. Now I'm, I'm borrowing from family and, you know, they're expecting something out of it. Quite frankly, my dad was like, here, I, you know, here's this, give me some equity. And if it ever turns into something great, if not, whatever. By the time my dad was done, he had invested $500,000. I raised $13 million in total, um, for that, for that company and raised from a bunch of people. I raised from over 60 individuals and family offices and VCs and whatever across all that. It's a lot of pressure. It's a lot of fights. You know, everybody, even if, even if they're all excited about how the company's going, some people are still gonna argue about how much they should get and what, who should do what, who should be the CEO and who should do whatever. Right. And, um, it took us years, years and years and years to build up a product that really was capable of cash flowing positive. And, you know, where we then got offered, we actually got offered $60 million by Snapchat to be purchased and we said no to it and we ended up selling to another company for less than what Snapchat offered a couple of years later, which is kind of how it goes. Sometimes, but, uh, that's a different story for another time. But, but the point is, um, I built GenAIPI in a weekend for a couple hundred bucks. And, and no offense to you, Spencer, at all, at all. Like I didn't have to hire a developer. I still don't have a developer on staff. We as a company, GenAIPI, deliver software products to our customers all the time. And we have no developers on staff. Right? We have smart people that know their way around code, know their way around AI systems and know how to use AI to get the right answers and jobs done. But, but we're in a different world, right? We are in a different world and it's cool. It's exciting. It's fun. But the, the, the level field, or sorry, the playing field has been leveled literally back to Pat. You can be a homeless person and start a software business. Why? Because you can go work for Uber for a couple weeks and You know, I know it's hard. You gotta have a car at least, right? But you, but you gotta be able to find a way in a gig economy or something to go make $200 and then go to a library, sit down at a computer and build something, and then you gotta go sell it. Okay? So that's pretty awesome. Pretty amazing opportunities. Now the playing field is leveled. Anybody can come up with $200 and go start a business. Okay? So let's jump back into Replit. Okay. Dive back into making this something real. Where is my Replit one? Do, do, do, do, do. It's not this one, is it? Nope, it's not. Okay. I don't know where my— Oh, it was here. It was here.

Okay. All right.

So now I've got, I've, I've had my AI work. It worked for 7 minutes. So let's just look at this. I can click right here and it says during that time it wrote, it did 47 actions. It changed. 1,990 lines of code. It got rid of 190 and I paid $1.86 for that work. 2,000 lines of code for $1.86. That's amazing. So now I need to sign up for an account. Let's see if this works. So we'll just make my account Jon Cheney and we'll just use joncheney7, my normal email, and we'll just try password as my password. Hey, It worked. Awesome. Okay. That's cool. Now, if it didn't work, what would I do? Talk to the AI. Hey, I tried logging in or I tried signing up. It didn't work and fix it. Okay. But what you're trying to do now, guys, is you're trying to— so let's go back over to, to GenAI API really quick. Okay. When I launched this thing, my goal was to launch in a weekend,, which is a very, you know, tightened timeframe, right? But my goal was, okay, if someone lands on this site, they need to feel like it's real. They need to feel like they're not being scammed. They need to feel like there's real value here. And the first step to that, the very top level is, does every button work? That's it, right? Can I go to every page? Can I click on any button? If I click get certified, does it go somewhere?? If I click Live Courses, does it go somewhere? If I click on Blog, does it go somewhere? If I click on Instagram, is there a page there? Right? And so I have to do all those things and click through all those things and check and make sure that something's there if I, if I click on it. Because if I do— oh, you know what? This is actually not going to work. Yeah, this account doesn't exist. This is a problem. This is something we're like, oh wait, They have a fake X account on there. So I obviously need to fix this. This is a super easy fix. It's because we just changed from Gen AIP Institute to Gen AIP. We got the, we got the domain we wanted. Okay. But, but the point is guys, right now you have to go through your app and you just have to make sure that A, it does everything that it's supposed to do and B, that it's not broken. Right? Because as soon as someone clicks on something and it doesn't work, just a little antenna goes off and you're just like, Am I getting scammed, right? Everyone's afraid of getting scammed or engaging with a company that's not fully put together, whatever. But everything needs to work. If you click on something, it's got to work. Okay. AI leadership, schedule strategy call. Does this bring up a calendar? Yep, it does.

Fantastic.

That's amazing. And so I know that this is actually working now. Okay. So here, that's what I'm going to start doing in here. And now this is a brand new thing. And so obviously, you know, GenAI was all completed. What I'm going to do here is I'm just going to start saying, okay, let's see if I can do something. Let's go to Skills, create your first skill. My app crashed. Does this mean that the AI was dumb and it created a bad product? No, it doesn't. What it means is it hasn't built that yet. I'm just going to click this and I'm going to give it a little screenshot. I'm going to say I just clicked Create create your first skill and this happened. Okay. So remember your AI is the way that I treat them is I treat them like my employees, right? This is my developer. And so I'm going to talk to it just like I would if I hired Spencer, right? I'd be like, hey Spencer, I tried to click on this create your first skill thing and this happened. Oh, I know what's going on. I'll fix that. Right. And you know, he'll then, he'll then make it happen. I don't have to be technical. I don't have to understand how the internet works and I don't have to understand all these things. We're gonna talk a little bit about APIs and some stuff like that in a minute. But, but you don't have to be that, you don't have to be technical to do that. Now, I will tell you this, I have become far more technical over the past couple years in working with AI and working with code. Why? Because I'll actually sit here and I'll read this stuff. Oh, fix the crash was caused by an empty value in the dropdown. Okay, cool. Well, I, now I know that if there's some sort of empty array or something like that, then it's going to cause a problem for me. Okay. So now I'll go skills, create your first skill. Boom. Now I've got something here and it was probably this dropdown, none, intellect, spirit. Okay. So we're going to say piano.

Okay.

That's my skill name. Primary trait is, this is going to be probably a creativity. Activity 1, secondary trait optional. Um, it's kind of an intellect one. It helps me, helps me there. Create skill. I've got a skill. So now I was able to create a skill. It's showing up. It works. Super cool. Okay, let's add another skill. Let's just make sure this works well. So now I'll do kayaking. This will be a strength one and we'll have no secondary one. Awesome. Okay, so that Works. Now we've got these traits: intellect, spirit, creativity, strength. Now, um, I have intellect is contributing piano, kayaking is contributing to strength, piano is contributing to creativity. I have nothing for spirit yet. That's fine. Now I can't edit anything here. I'm sure at some point I'm going to want to be able to click on things and look at more details, but for now it works. So I'm just going to click through each of these pages, right? Reach piano grade 5. That's a, that's a goal it gave me. I'm going to create a new goal. I'm going to try. Okay. I can't create a new goal. I'm trying to create a new goal and it's not, it's not working. Okay. So I'm just going to say, um, I tried to create a new goal at, and I'm looking right here. This is the website slash goals, forward slash goals, right? So the repo.dev is just a temporary thing. Okay, whoops, Escape, it will do that. /goals and didn't work. Make all of the buttons on this page work with all associated functionality needed for it. Okay, so it's actually putting fake data here because I haven't made a goal to reach piano grade 5 or run a marathon or write 5 books. 50,000 words. I haven't, I haven't made those goals yet. Okay. So anyway, right now it's doing that. Now I can let it, I can just sit here. I can go over to social right here. Social module coming soon. Okay. That's good. Settings also coming soon. Right now it's really just focused on, on this stuff. So now here's a quick log. I want to do piano. Hey, I just worked for 60 minutes. Now there's obviously, this is not exactly how I want it to work. What if I only sat down at the piano for 20 20 minutes. Okay, what I probably want is like an activity and a duration or something like that, right? So maybe I could be like, okay, piano this log. Okay, but it's cool that I can be like, hey, I went kayaking. Boom. That's cool. I think that's a pretty neat functionality. Then I can decide, okay, what do I really want this to do? Okay, I've got a 1-day streak, it looks like. I don't know if that's real. Okay, just because it shows it could just be displayed. I don't know if it's actually calculating a streak in any way, shape, or form, right? My current status is I'm on level 1 for all these different things. So, all right, so here it says, I see the goals page is still using mock data. Let me rewrite it with full functionality. So now it's going to actually connect it to the database. It's going to build the right tables in the database. So that's all happening. And I can click on this and I can look at what it's doing. Decided on the goals page update. Restarted, reviewed the execution logs, took a screenshot. Now you can do it. Okay, so I should be able to go to the database now and come to overview or my data and look at goals. Now there's actual goals in here, or there's no goals in here because I didn't— I haven't done one. So let's create my first goal, and I'll say reach level 100 on piano. This means Everything. Okay, and then I'm going to link Piano to it. Okay, and then my target value is 100 here. And deadline, I'm going to accomplish this by February 1st, 2026. Okay, now a couple things I could change on here. Maybe I want my date selector to be a little easier. I had to manually enter that. Maybe if I click here, I want a calendar to come up. Oh yeah, a little calendar. Click here. I could add that or I could have something here. This gives me a little up and down thing here. There's probably other things I'm going to think of as I actually use it and do this. Now, if I click Create Goal, okay, I'm going to do this. I'm just giving it a screenshot. I'm showing it what I'm looking at. When I click Create Goal, nothing happens. Or appears, nothing happens. Okay. I'm actually curious if I go to the database here and I go overview my data and I click on goals. Yep. It's not actually, sometimes it will record it. And, and the UI, the user interface here that what I'm looking at the front end just isn't set up properly to close the, you know, this, this little modal that's sitting here. We need it to show me what's happening. We need to make sure it records in the database and, and the functionality is what I want it to do, right? So it says, hey, the error shows the deadline is being sent as a string instead of a date. Let me fix the backend to handle this. So it just made a mistake in figuring out, you know, how it should store that date data. Okay. So no big deal. That can be fixed. Okay. So fixed. Done. Create your first goal. Reach level 100 on piano. Test, whatever. We're going to go piano. Target value is 100. The date is 02/01/2026. Create goal. Boom. Okay, now it's working. Okay, so all I have to do is talk to it and just say, hey, I tried this, it didn't work. This is what I want it to do. Exactly how you would, as the CEO, tell that to your developer that you hire. Okay. Very, very simple. And then just rinse and repeat until everything works. One thing that you should never, you should always make sure you check, does logout work? Yep. Does login work? Does signup work? Um, anyway, let's build this social module right now. Um, we're going to do one thing that's important. This is actually a critical, a little critical thing that really sets apart, um, good Replit projects from others. And I actually worked with Replit to build this out. And they, they actually have done a little bit of that at this point. So right here, you have this, this little files tab. It's right up here next to publish up in the top right, this little thing. And these are all the files that you have in your system. Okay. So these are all, this is all your code, right? If you wanted to take this and move it over to another coding system or hand it to some developers outside of Replit and they wanted to keep building it, you can just give them this code and they can keep working with it. You can export this right to GitHub. You can export this to other systems that allow you to do all that. But I'm going to work inside of Replit here. But this little file right here, this replit.md file, MD is basically a text file called Markdown. Inside here are instructions. Replit has been taught to check the replit.md file for what your project is about, right? So it says, hey, here's the authentication system, and here's some credentials for testing. Here's the progression engine, and here's database schema. So here are some of the data that I gave it. It's not everything though. It's not the full business plan. It's not the full, you know, list, feature, feature list, all those things. It has some things, right? It's got some consistency rewards and level curves, and that's really cool. Okay. But it knows to look at the Replit MD. What we're going to do is we're going to add a few more files to that. So when we first started, when we all started, we added in this kind of 3-sentence description and said, here's what you want to do. And then we gave it 3 or 2 or 3 things, right? We gave it the business plan. Here's that. And then here's the feature list. And then here for me was all these level traits and math and how, my leveling system works.

Okay.

So I've got these things, but they're just sitting here in this attached assets folder and they're not really being referenced anywhere. So what we're going to do is we're going to create something. Um, I'm going to pull up my Google Drive here and we're going to look at this, how to build a business course folder for the first time. So in here, um, build a business in a weekend with AI. I've got, some flows and overviews and software and tools and things like that. One of the things in here is Replit.

Okay.

This has a bunch of tips and tricks with Replit that will really help you. So right here, we're going to grab this docu— or this little, this little thing. Please use the attached business plan that's in currently in the attached assets folder that I just showed you. I'll show you that again. Okay. And the attached feature list in the same folder, create a read me.md file with the business plan and create a progress.md file with the feature list. Okay, so I'm asking it to create a couple documents that it's going to add. Okay, don't write any code, just create these documents. On the feature list, create a checkbox system that shows us what we have completed and what we still need to create. Please examine the app and make sure the progress.md file is up to date and add any items that might be missing from the list that we've created. Also, add references to these files and their purposes to the replit.md file. Okay. So this will make more sense after we put it in there. I'm going to drop this thing into the chat so you can use it in your project. But here's what I'm going to do. I'm going to paste this in. Now remember, I'm asking it, I'm saying, look, we've got these documents here. I want you to create MD files and drop them into the actual project here. Then I want you to put a reference to them inside of here so that when you're looking at your thing and you're trying to figure out what you need to do, you remember to look at my information, my business plan, my feature list, my plan that I worked hard with ChatGPT to create. I don't want you to create your thing. I want you to create my thing. And this gives it what I like to call bumper lanes, right? So that it stays on task and in the project. Okay? Now I'm going to say, um, there's also an additional file called hero level traits. Um, call this, um, levels.md. Um, and yeah, all this levels.md. Okay. So I have an extra one, so I'm giving it instructions for that one, but now all I have to do is say go. Okay. And what it will do is it will look at the attached assets because it knows where this folder is. It sees those documents, and then we'll sit here for, I don't know, maybe 60 seconds or so, and it should create some new MD files that will be nicely organized. It will have my entire business plan so that when it's— so that when we start building, for example, the Stripe integration to start setting up payment, it knows why people are paying, what types of systems or what types of things are tied to the premium subscription, right? And oh hey, we're going to need some AI set up and we're gonna need this and that and whatever. All these different elements here are going to be very, um, specifically detailed as they move from the attached assets into this markdown file. So this is probably the most technical step or one of the more technical steps you have to do. It's not that complicated, but the purpose is let the AI know what you're doing. Be the boss, be the CEO. You're the one in charge of the plan. Give it the the plan. If it doesn't have the plan, it can't do it. It can't follow it. And it won't keep following it just because you said it once. It's not like ChatGPT where it just remembers your chat because you might start a new chat. You might refresh and do things and start a new whatever. It's not going to be able to remember every single thing you've done. It's too much context. It's too much data. So what you have to do is you have to give it this high-level plan and then let it do everything in the context of that plan always.

Okay?

So you said having— you had an MD document. Is that a master or MD files? Is that master document file so that they know what mostly to refer to?

So MD stands for Markdown. It's just a format like a.docx or a.txt or, you know, whatever. It's just a format of text, right? That is commonly used in coding. Okay. It's just a simple way to format text. Okay. That's very low. Um, it's, it's very small in size, right? And it basically just follows a bunch of, you'll see if you look at, uh, if you look at the attached assets, assets like this, this is actually Markdown where you see all these like hashtags, all these pound signs in here. That means it's, it's saying what should be, you know, bolded and what should be italicized and what should be whatever. Right. And so it's just, it's just a way for the system to, to read texts. Okay. So it's just one type of, of system, but this Markdown file, again, just keeps you, keeps you in charge. So I'm gonna look at my README file here. Okay. Now it's got my business plan. Here it is. It's beautiful, right? I click on the MD. This is exactly what I pulled over from ChatGPT. My market context, my target market, who I'm going after, all these different things. Then I've got my levels document, right? The levels document. This is all the math behind my levels and how it all works. And then now it did consider these things in my first prompt and in my first prompts because I asked it to. I said, make sure you look at these documents. And then when I asked it to switch from Design mode to the normal mode, make sure you look at these documents. Then what I can do at, at any point through this, I could just say, let's work on the next feature. Look at the progress file, the progress.md file that it just made right here. We'll click on it in a second. Okay. Now I can see, all right, completed or not yet implemented. Okay. It looks like we've got account creation with email. We don't have social login yet. We don't have invite-only support. We don't have password reset recovery. We don't have any of those things. There's a lot of things need to be done, but now we have kind of a list of what's done and what's not. And it automatically created it by checking the whole system, right? By checking my list and then checking the system. Does this exist? Does this exist? Does this exist? Right. And it goes through and it just edits it and modifies it. And then if you're good at keeping track of your work, you'll ask it after you build things, you'll say, great, update the progress file, update the progress.md file. And then you'll constantly have this beautiful working document that shows you how much, how many things need to be done. And we can say, all right, you know what? I want to build, I want to build the, the internal admin system. Let's build the admin dashboard. All right. So I can see this stuff. Okay. Let's just do user management for now. Great. Right.

And so then it'll do everything that it needs to do to make that happen. Okay. One question, as you then, you know, you refine your business plan or you refine your feature list, right?

Yep.

Are you then going and regenerating one of those original documents and then from there saying, okay, update the markdown file that's attached?

Or, you know, what are the flow of steps you're doing to— Yeah, I, I will typically not go back to ChatGPT again after I'm in here. I'm typically just working in here because the, the AI you're working with in here is Claude 4.5. Like, it's brilliant. It's just as smart as ChatGPT for doing that type of a thing. And so, I will then just update the live document. So if I, if I'm going through discussion and I'm like, okay, this is a little bit of a change, a little bit of a pivot, I'll say, please update the progress and the README files to reflect this latest change or these latest, you know, this pivot.

Right.

And it will be like, great. And it'll, it'll edit where it needs to edit. And it's smart enough to do that. Awesome. Yeah.

Yeah.

I like to just kind of keep it all in here now once I'm, once I'm in. So now it's created these things. It did the README, the progress, the levels, and it updated The ReplitMD, which is the one that the system is telling it to look at, right? So now if I come into ReplitMD, it says, by the way, here's some documentation files. You've got these things. This is the purpose of each of these files, right? Developer reference, technical specification, complete feature checklist, business plan, vision, market strategy, monetization. Amazing. Now we're being a good boss. That's what it, that's what it really comes down to. If you're not a good boss, you can't get mad at your team, right? When it doesn't do what you ask it to do, right? If I ask Tanya to do something and she goes and does it, but I don't give her specifics, I might be like, hey, Tanya, that's not how I wanted it. And I'm sure she's already experienced this in the first month of working with me. Totally fine, right? It's because I didn't, I didn't give that instruction, but I'm not mad at her. I'm just like, okay, yeah, we didn't communicate. That's my fault. Let's make sure we're aligned and I'll be clearer on that thing next time.

It's just, he doesn't get mad. He is very, very diplomatic about this. And the example would be the one he used earlier about it's written by AI. He pointed this out and I was able to say, how much AI do you want me to use in this? Because we're, we're now in our name and it was so easy to clarify.

He's great. So the whole bad boss thing, just mull over that. He's an awesome boss. Yeah. I'm not, I don't, I don't think it's productive to do that. I think it's productive to communicate Well, that's really what I think it is. And if someone's doing something that I don't agree with, my first reaction, and this took training, by the way, I did not start this way. My first reaction now is, what did I not share? Or, oh, you know what? I didn't. I look at myself. What did I not say to get the result I was looking for? Oh, you know what? I wasn't clear on that. No big deal, dude. Here's what I want now going forward. Let's just adjust this. You're doing awesome. Keep going, right? I think that's a really productive way to be a leader, is to just, you know, it's always on you. If somebody— now, if I give an explicit instruction and that isn't followed, then I'm like, come on, man, I told you here, right? So anyway, but I still don't get mad. I still don't get mad. I just want things to get done. But anyway, so it says we're roughly 30% done. That's probably true, but, you know, 90% of the work's in the last 10%. To get this really where I want it to be, where it's gonna be just an awesome, amazing experience. I'm probably gonna have to take, take a good amount of time to make this happen. Okay, so guys, I just want to go over kind of a couple, a couple basic things that will help you. I'm sure a lot of you for the rest of today, or for a few more hours today, will keep working on your project and you keep doing it without me, and we'll come back tomorrow and you'll be like, yeah, I worked for like— I've had people say I worked I worked till midnight on my project and now it's done and I've spent 15 hours on it or whatever. Super cool. But you'll do a little bit better if you just understand a few things. So we're going to take a moment here in this folder again and we're going to just go over the coding quick start guide. This is very, very high level. Spencer, you can go to sleep probably. But it's just very high level to make sure you guys understand when the agent is talking to you or when you're asking it to do something, how it works. And this is going to be kind of one of the big things I want you to get out of here is APIs and how they work. Databases, APIs, all that stuff is going to be really, really awesome to help you start to get really creative about what you're doing. Okay. And to know you don't have to build everything. You can tie into other systems that other people have already built. Okay. So front end versus back end. We've already talked about this a lot. Front end is UI is basically everything you see. So, as we look at the screen right here that you see, this is Google Docs, and there's a bold button, and there's an italicize button, and there's text, and there's this thing down here, and I've got all these icons. I'm just— what I'm seeing on my screen, this is the, this is the front end, or the UI, the user interface, okay? It lets me click, and type, and interact, okay? Then, anytime I do something, something, though, anytime I click on something, typically it's going to start talking to the backend and say, hey, we're going to have to adjust something here. Right? So if I, if I type in this document, for example, right? It's actually updating a table that looks just like this one right here. Right? It's, it's like an Excel table. Right? And there's, you know, there's going to be a place that has a really long in Markdown, probably format, um, bunch of text, right? And so that's when I, when I click on that and I change something, it's changing it on the front end, it's displaying, but then it's actually updating and saving to the database every time I do that. Okay. And you can make it explicit to where you have to click save for it to update the database. Or you can say, every time I make a change, update the database. And you have to, there's some, there's some options. Optimization you could do to actually make that work well. But, but the point is you've got front end, you've got back end. The back end is where all of the calculations happen and where all the data is stored. Okay. That's really all you need to do. So here's, here's the really easy example that everyone can understand. You've got a login form and on the login form on the front end, that's where you type your email and password. And then on the back end, it's going to check when you click go, it's going to, it's going to receive some information and say, hey, Here's the Jon Cheney username and here's the password. Password, does this match an entry in the database? Yes, great, let them in. Give them a token, right? It's a JSON Web Token, right? That's what we're doing with this login, right? It's going to give them a token or a key to get in, okay? If it's not, hey, we're going to send back an error. And then the frontend would need to be able to understand that error to say, hey, You entered the wrong thing. Try it again. Right. Click here to reset your password or whatever. Right. Whatever, you know, pathway, whatever experience you want it to have, you can tell it to do that, but that's what's happening. There's communication happening between what you see on the front end and what actually happens in the backend. Okay. So the database that we've been kind of bouncing back and forth, this is where you're keeping all of the data and it's just like a table right here on our users page. We've got a demo account and we've got an email. Or my username, right? Here's my email. This is my password. Now it's hashed. This is good. This is a good security thing. It will automatically default to this. I can't actually see the password. It's encrypted. Okay. Or hashed is what it's called here, but I can see my username. If I wanted to change it, I could change it here. I can also build a front-end admin dashboard, maybe in settings or something. If I'm logged in as an admin where I could adjust my profile, profile, or my— or I could adjust Tanya's profile when she makes one in my system, right? I could go in and I could have an admin system that lets me do that, but, but I can always get into the database and change something manually. Okay, so that's really— I need to know— and we're using what's called a PostgreSQL database. You don't really need to know much about what that means, but that's what we're using, at least inside of Replit. It's a Neon PostgreSQL database, structured data. Okay, unstructured data is a little bit complicated. I'll let you guys learn about that. ChatGPT is a great teacher. Okay. So yeah, basically you've got tables, you've got relationships. Relationships are important. There are some times where you'll say, hey, let's go to a goal. It might connect to, yeah. So it connects to these other tables, the skills table and the users table. Okay. So it's going to be looking for those things or it's going to be connected to those things for whatever function it needs. Those relationships there. Okay, um, API. An API is a connector. That's all it is. Okay, let's say Pat, in his, in his, uh, hey, where are we gonna— where am I gonna retire, you know, app, right? Maybe one of the things that he thinks is important is the weather, right? What's the weather like in the Philippines versus Thailand, right, versus wherever. So So he could every day go and check the weather and do that, or maybe have ChatGPT put together a big description of something and put it in there. Or he could just tie into an existing weather API, right? And so he could just say, let's go to weatherapi.com. You just Google it or ask ChatGPT or whatever, or actually just ask, just say, hey, I want to show the weather on my app, right? And he would come in and he would sign up for this thing, right? Pricing first. I'm always going to look at pricing.

Okay.

And look how cool this is. This is a system that automatically has weather. If you want to make a weather app or if you just want weather in your solution for whatever reason, you can have up to 1 million calls per month. You can call the weather a million times, right? And get data from this API for $0 a month. That's crazy, right? And a lot of APIs are like this, that the lower end, and I know a million sounds like a lot, but the lower end of calls, the lower end of actions are usually free because they want you to be able to set it up and test it, whatever, without having to pay a lot of money. Okay? Now, if you, if you publish a weather app to the App Store and you want millions of users, well, you know, maybe you're gonna have to pay $35 a month and get 10 million. If you have 10 million calls per month on your weather app, you're making more than enough money to cover your $35. $5, I promise you. And if you're not, you have a problem with your business model. Okay. So APIs are amazing. They allow you to connect to things. So maybe I want to connect to, um, ESPN. ESPN actually has an API. I built this cool little thing called, uh, Cougars. The heck is it called? Cougars Dashboard. Uh, what is it? And my brain is like totally probably blanking right now. I can tell you right now. Home, open a new tab. Come on.

Okay, apps.

Cougar IQ. I think it's Cougar IQ. CougarIQ.com. Cougar's IQ. What the heck is it called?

Okay.

I'm gonna find it. Do, do, do, do, do. I just wanna show how cool this is and how much data is actually out there for free APIs, guys.

APIs are, man, this is like Replit's struggling. Are these all of the different apps that you've created in Replit?

Yeah, these are some of them. And it's struggling to pull the next page of apps. Yeah, it's like the server crashed on me or something.

I don't know how to fix this.

Refresh.

Let's try something. Replit. I've used APIs in a couple different ways that are pretty easy to illustrate. Google has a PageSpeed API, and I made a tool where you put in your website and it judges your page speed. Yeah, that's an example. Yeah, another one is eBay API, where you might have a website about, um, teacups and it shows what teacups are available to be sold or something. Yeah, it's amazing how many APIs are out there.

For anything. Yeah. The US government has them everywhere. I don't understand why I can't see this. Anyway, there is an ESPN API. By the way, if I look up ESPN API, There actually isn't a real one. It's a hidden API. And I wanted to be able to pull data in. Um, yeah, it's not Cougars IQ or Cougar IQ, or maybe it is. I just don't have the site live.

Okay.

Finally, this thing woke up. There we go. There we go. I just had, had a problem, but, um, but the ESPN API can pull any information on any game on any sport. Might be Cougar Companion. It's probably this one here. Let's just open this up and let's look at my publish. CougarIQ.com failed. Okay. Yeah, it failed. So it needs to, it needs to republish. Republish. Um, I could probably click on it here though. This will be the, the Replit one. Might take a second to spin up the servers because it's just been sitting there quite quiet. Yeah, here we go. So, so in here it's pulling from the free ESPN API. And so I can come in here and look at, you know, like football and I can say, all right, Southern Illinois, or let's see, Wyoming Cowboys, whatever. I click on this game and I can view the stats on it. And then I can come in here. I even go into like, I can look at the box score and team stats. And, and I built this fun system that lets you look at these different things. And then I can even look at highlights. It even pulls video, right? So I can just click play and that's pulling right from ESPN's database. They just have first and goal. They just have that video stored there, right? Boom, touchdown Cougars, right? So I made this cool BYU fan dashboard completely for free using all this ESPN API data. And now I don't have to go out and find all this stuff. Right? I don't have to manually do that. APIs allows you to programmatically bring in other data from other systems. Okay. You can even, and I know this is getting maybe ahead of ourselves a little bit, but you could even, if you wanted, um, build your own API from your own system. Right? So maybe I wanted to have this hero database or I wanted a GenAIPI API. So every time that I do, um, some studying on genaiapi.org, it automatically sends a signal over to my hero app that gives me progress towards my intellect because I learned about API or so I learned about AI, right? I could set that kind of a system up. You can, and all in order to do that, all you have to know is what's possible. And then you could say, hey, Replit, will you create an API that pulls data every time I do this and sends it over to this app?

Yep.

I'm, I'll get to work. Here's what you do. Add this, add this, add this code. And usually what they'll do, guys,, right? So like if we're using this weather API as an example, right? I'll log in here. 'Cause I actually do have an account. I've used this API before. But if you were to grab this API key, and by the way, I am doing a big no-no right now in showing you guys an API key.

Okay?

Now the good news is I don't actually pay for it. I don't have any, you know, credit card attached to this. And so whatever, but you never want to display your API key. If you have access to this, you can bill my account for as much as you want, right? At least if this was a paid account. Okay. But like I can, I can attach, and we might do it later. We might attach to an, you know, an OpenAI API key. So now I have AI pulling into my app and doing things. And so it will give you a unique code, a long string like this, and Replit will ask you, hey, what's your API key? And then you'll give it to it and it will set it up. Okay. So, really cool. All right. Now my cougariq.com should work.

Nope.

Not working. Still not working. Failed again. I don't know why it failed. Why is it failing?

I'll look at that later. It doesn't matter right now. So, um, it was, it was talking about buying a new domain.

Have you bought that domain yet? Oh yeah, absolutely. But I bought it on GoDaddy, right? Don't buy your domain through Replit. Pro tip. Don't buy it through Replit. The reason is simple. And I'm not going to explain it super, but, um, for Spencer, people that might look at it, you can't adjust your DNS settings. You can't do like complex DNS settings. So if you need to like connect it to a mail system or something like that, you can't. You're stuck with just like the basic, like deploying to a domain. You need to do it in some GoDaddy or Namecheap or whatever the heck you like, Squarespace, whatever you want to use. Used to be Google Domains was my favorite, but they sold to Squarespace, which was really stupid. Um, okay. So guys, I'm just going to keep, um, going through here and in my app, and I'm going to start adding in functionality. I'm going to start adding APIs. One of the things that I want my thing that tracks all of this, you know, the levels of my life, is I don't want to have to come in and quick log like kayaking every time I do it. Maybe, maybe I can connect it to my Apple Watch, and if I finish a kayaking workout on my Apple Watch, that sends it to Apple Health Apple Health syncs to this and boom, it shows up in my system. Right. And so I can connect and automate things via APIs. Okay. So just think through what is the experience you want to have and then tell your developer and then work through it. Right. You just have to be willing to try things that you haven't done before and you will learn, you'll, you'll run into a few blocks or whatever, but, but you'll figure it out. And if you have a question, You can ask me, you can ask your peers. One of the things that we'll do, and we'll do it right now because it's on my mind, we need to go to Facebook. You guys get to join a fun group I have on here called Build a Business with AI. Okay. So, all right.

Go in.

Now let's see, I don't want to invite, I just want— maybe I do want to invite. This is fine. Let's do this right here. All right, so I'll just drop this in the chat here, guys. You can click on that, request to join. The people that are in this, at least for now, um, are, are people that have taken this course. And not everybody, you know, I only have 121 members in here even though I've taught hundreds of people in this course, but not everybody has Facebook. Not everybody uses it. But this is a pretty fun spot to get help. If you need help, come in here, talk to stuff, talk to people. It hasn't been really any discussion this year because this is the first cohort of the year. And anyway, but people ask all kinds of questions like, hey, I'm working on this thing and how do I do this? What do you guys think about that? And other announcements, stuff like that. It's not, it's not a super active one. By the way, Tanya, I'm glad you're in that now. Cause this is a company thing. I forgot about this one. But anyway, it's a, it's a great group. It's a great place to ask for help. So if you're trying to figure out like, hey, I'm trying to work on this thing, I don't know how to get this API or how to do this thing, just ask. And there's lots of very, very helpful people in there. Okay. So guys, that's really all you need to know about APIs authentication. We did the JWT thing. You can also set up like a login with Facebook or login with Google. Here's the thing. And that's really kind of all you need to know. We could talk about hosting later.

That's all we need to cover in that for now. I have a question real quick I've been meaning to ask. Yeah, go ahead, Matt. Keep referring to different Google Docs.

Is that something we, we have access to or we will have access to? Yes. Yes, you will have access. I will email this to you, this folder. In fact, what I'll do now is I'll just, I'll just drop it in the chat here. You can, you guys can open it. You'll have access to it. But you'll get it in the, in the kind of post-course email tomorrow automatically. But, uh, yeah, anyway, so there's that, but I just dropped it in the, in the chat. So you guys have access to that. I'll also put that Facebook link again in there if you, you know, if you don't get it today. Okay. Um, but good question. Good question. Um, all right. Let's set up an API. That's really the last thing that, that I want. Let's actually do one here, here in this. Um, now let's teach a couple things too. I like this, like this. This is actually a good thing. So down here in the bottom left, this is where you're talking to the agent. There's a lot of extra cool buttons in here. Okay. This is fast, right? If you need something like really fast, like change the spelling on this to that, right? Or change the text from this to that, right? You can do that and it'll just do it really fast and it will cost almost nothing to do it. Okay. You also have all these other options, right? So you got the fast, you got the autonomous. I like to work in usually medium or high. Sometimes I go into max. Depends on if I'm working for a client or not. It costs more. It costs more to work, um, with the Max system. Um, but it will do everything. Long-running, hands-off building experience. It can literally work, guys, for 3 hours by itself. That's crazy. Okay. And so if you're like, hey, I need you to populate the data for, you know, 10,000 entries on whatever. And go find all this data and fill it all in. You could just let it get to work doing all that stuff for you.

Okay.

And so you can do that. If you do low, it's going to generate, execute. If you go medium, it reviews the latest code changes and fixes issues found. So it's going to be just a little more autonomous, make sure things are running better. High is going to expand the review scope and it costs a little bit more. The higher you go, the more it costs. Okay. Expand the review scope to the entire app. But look, for this 4-minute task here, it cost me $1.56. Okay. When I'm comparing those numbers, and so you guys, you guys signed up, maybe you signed up with my referral code, maybe you didn't, but either way, you've got about 25 credits, $25 of credits per month ongoing, right? Maybe you got an extra 10 if you sign up with my code, but, um, every time you use the agent, it's going to, you know, use some of those credits. If you get up to $25 and you want to do it more, then you're going to, you know, you're going to use, you're going to have to pay for that, right? When I built, just for reference sake, when I built GenAI API over a weekend, it cost me roughly $140 of credits. Okay. So again, I built an entire business for $140. Crazy. I'll pay that all day, any day, every day. Okay. And I actually usually like to keep it on higher max, right? That's just, it does more. It makes less mistakes because it takes into account more context. And that is why it costs more money. Okay? So, down here in the bottom left, well, there's a couple of things down here. One, this is a select element button. It's very, very useful. Okay? This select element button right here. I can click here and then I can come over here and I can tell the AI like, hey, I want you to work on this specific thing, right? Like this thing, right? Make it say this or change the color to this. And you can actually manually go in and even change some things. If you want. Okay, that element selector is very, very useful to tell the AI this is exactly what I want you to change. Okay. You can also click Attach and that's fine, right? Then you can click right here and you have the build mode and then you have the plan mode. The plan mode is new, right? Plan is awesome. They used to have to be a little more explicit and say, don't write any code, just listen to me and let's talk about this. And in fact, I have that, Yeah, please create a plan to build this. Don't write any code, just outline the steps. This is actually kind of old advice. You can still do it. I still do it sometimes, but you can go into plan mode and I can say I want to add in the first AI feature and connect the OpenAI API key. Look through the feature list and figure out a good place for us to start with AI. What should we build?

Okay.

So I'll ask it that. It's going to look at the code base. It's going to look at the feature list document. It knows the context of what OpenAI is and everything can do. The, again, the model that's underneath this guys is called Claude. Okay. Claude 4.5. Okay. And let's not augment code. Augment code is great. It's actually a great tool, but Claude is made by Anthropic. Okay, so Anthropic is— Anthropic has Claude, OpenAI has ChatGPT, X has Groq. Okay, so you've got these, these application level things, ChatGPT, Groq, Claude, and then you have models above it that power those things. Okay, so anyway, Claude is is made by Anthropic. It's very amazing. Very, very good tool. It's, man, I feel like, I don't know if it's my computer or what it is right now. All right. So proposed tasks, set up a connection to OpenAI, create backend for AI insights, create prompt for user analysis, add AI insight display card, make AI, mark AI features as done. All right. So what it's wanting me to do here. So this is monthly recaps pattern. Voice input, AI parse natural language, logging. Wants to do this, shows the value proposition. Nope, don't want this. I want to actually— whoops— actually first focus on the AI-powered logging. So again, I'm still in plan mode. But it's not going to start building until I give it permission to start building. Okay. So it's thinking for a second. It's going to say, all right, it's a great choice. Yeah, it always thinks you're right. Okay. Very cool. Practice piano for 45 minutes, felt focused. And so then the AI is going to parse that into actual structure. So this is what I want to do here. Smart text box on dashboard and go.

Okay.

Build it. Okay. So now I've asked it to do this. Now it needs to connect to OpenAI. In order to connect to OpenAI, it's going to need an API key from me. Now, it used to be that you had to go over into OpenAI, right? And I'd have to come over here into this and I'd go OpenAI API. I Google this and go into the API platform and log in. Man, I swear my internet is really struggling. Going very slow. Okay, by the way, guys, when you do log into OpenAI's API platform, it is different than ChatGPT. I want that to be really clear. OpenAI has two logins. They have a ChatGPT login, which is their product, their consumer product, and then they have the API platform, which is their developer product. Okay. ChatGPT is built on top of their own platform, their own API platform, which means you guys could actually rebuild ChatGPT in about 10 minutes. All you'd have to do is give it an OpenAI API, tell it to clone ChatGPT, and you'd be done. You'd have your own functioning, you know, ask Sariah instead of ask ChatGPT.

Okay.

So we'll log in here, continue with Google. I think it's this one. It's not the internet speed. I don't know what it is. Is. I'm probably just overloading it with too many things open. But anyway, guys, in— once you're inside here, inside the API platform, then you have to just get your own model. And they always make it really difficult to find this thing, to find your actual keys. There we go. API keys. It's in settings. But here you're going to have different API keys and it's not going to show me the secret key. In fact, it's so secret, that I can't even look at it again. If I lost my API key or it stops working, I just have to create a new one. So it's easy. Just go create new key. And you just say key 3, whatever, right? And you'll select a project. It could be whatever default project, create secret key. And it will give you a brand new key that you can then copy and pull into your project if you want to. Now, this is kind of the old way. You can still do this if you want to be very specific about your API key usage. But Replit now allows you to, okay, so here's this right here. Confirm OpenAI integration. Add OpenAI to your app using Replit's managed service. No API keys required. Charges are billed to your credits. So you can just have it use your existing credits in the system and you can just Click approve, right? Now, you don't have as much control over that if you do it this way, but it has made it a lot easier with at least some systems. They have something like 200 systems, 200 AI systems that are automatically just available here. If you want image creation, if you want video creation, if you want any, any type of AI, anything you can imagine, they have a Replit account that will just automatically set it up for you, right? I use it sometimes., but I like to know how to go in, grab my API key directly from OpenAI, use it and do whatever. Okay. And by the way, this key is super long. You can't even see the whole thing here. So you can't even steal that if you go back and watch this recording later on. Also, I don't need it. I'm just going to delete it. So we'll just delete this, revoke the key, and it is gone. Okay. All right, back to Replit and let's get this AI thing working here. Okay, so the OpenAI integration is now set up. It uses Replit AI integrations. Great. Now let me create the AI-powered action parsing endpoint and update the dashboard with the smart input feature. Now I'll create the AI and I can even see what they're doing here, right? Now I'll create the AI action Parsing endpoint and add the smart input to the dashboard, right? So this is just saying, look, I want to be able to just talk and say, hey, I just played the piano for 20 minutes. I worked on this song and whatever, right? And be like, great. And it'll, it'll take the important information from that and it will add it to this automatically without me having to manually go in and type, I played for 20 minutes and I did this. Anyway, that's the idea here is to lower the friction on recording progress towards goals. Okay, so now let me add the API function and update the dashboard with the smart AI input. We'll just wait a moment here. So we're almost done today, guys, about 15 minutes left. Probably finish up with just a few questions here. Let's do a quick poll on tomorrow. So I have to be at the airport tomorrow at 1:45.— or I have to leave. Sorry, the flight takes off there at 12:45. I need to leave my house not a second after 11:30, maybe 11:15 to be safe, just if there's traffic or a crash or something. So that, that means we could either start 30 minutes early tomorrow, or we can do a bonus session, I don't know, in a week or something like that.

What would you guys prefer? I like the idea of a bonus session so that we can be messing with this stuff and be able to come back with more information.

I think you'll get more value out of that. I agree. So that's cool. You guys get to do a little bit shorter session tomorrow, um, and I'm actually going to call it 45 minutes early. So we'll do 3 hours and 15 minutes tomorrow. We'll stop at 11:15, and I'll hop up, and that'll allow me to actually get in the car and zip off. My son and I are just doing a quick few days in Houston together. Just a father-son outing. We'll go to a Rockets game and eat some good food. I'm from Houston. I'm from Houston originally. So, so anyway, is there, is there a time that works better for you guys to— I know, you know, you have to kind of set aside a little bit of time. Do we want to aim for another, like, Tuesday, Wednesday kind of morning?

Is to go from, like, 9 to 11 or something on one of those on one of these upcoming weeks? Maybe next— I'm completely open.

You just tell me what time to be there, I'll be there. Okay, I'm just looking at my calendar here. What if we did the 20th, which is next Tuesday? You do Tuesday or Wednesday, so we'll just do a poll here for everybody's thing. It looks like Tuesday I actually have a little bit of stuff. I could do like 12 to 2, like a lunchtime thing.

12 to 2 Mountain Time next Tuesday. Does that work?

I actually am taking my mother to a doctor's appointment at 1:20.

Okay, that's all right.

That's why we're doing this right now. But Wednesday is open for me. I could do Wednesday. Wednesday is much better for me, actually, because I think— yeah, we've got Martin Luther King Jr. Day next Monday. And so Tuesday is kind of the and things are just kind of busier. What if we did 9:00 AM next Wednesday, 9 to 11? Yeah, that works for me. I'm fine. Yeah, same. So guys, we'll use the same link, same spot, and you guys will just get a bonus third session and you'll get to ask tons of questions. We'll get to go over probably some platform. Tomorrow is a little bit different than today. Today we were very heavily focused on let's build this business and kind of get it up and running. Tomorrow is All right, now let's look at everything else AI can do to then, and then we talk about how you funnel it back into your business and things like that. But we're gonna do image generation. We'll do voice, some AI voice stuff, AI music, AI video, AI agentic systems where you have systems that can do, cuz right now we're using actually an agentic system, right? But it's focused on developing, right? It's focused on building a website or building an app or something like that. Tomorrow We're going to look at agentic systems that can do anything for you, like, hey, cancel my Netflix, right? Or complete all these tasks for me and just get busy and just do a bunch of stuff, right? So teach you guys how to use agentic systems and some things like that. And it'll be fun. Be fun. We'll have a little bit more discussion too. We'll talk about a concept called surf, don't swim. So you can think about what that might mean. And yeah, tomorrow's a fun day. Tomorrow's a fun day of just like lots of stuff. Stuff. But having this third day of next Wednesday, and actually I'm going to put it in my calendar now so it doesn't get blocked out, or so, sorry, so that it is blocked out. Tomorrow starts at 8, correct? Bonus. Yeah, tomorrow's 8. We'll go 8 to 11:15. Yeah. Okay. Bonus session for cohort 24. You know, honestly, I think that the format we're doing by having a third kind of follow-up session a week later actually might be like really smart.

I might need to. Yeah, it seems like it would give so much time to play with things and then we could say, oh, I ran into this wall or whatever.

Totally. Yeah. Because you kind of learn, it's kind of like fire hose learning and then you can go like try stuff for a week and come back and be like, okay, questions, questions, questions. So we'll probably have a good, you know, half hour Q&A session. Um, questions about today, guys? We've got 10 minutes left.

Um, questions about anything we're working on, um, your specific project, maybe you're running into something. I have a really ultra-specific thing. I'm just curious, have you, have you ever worked with the eBay API or know a tool that works with that well?

That's one of the ones I'm interested in. Yeah, I don't have any personal experience with it, but I will say that I had two guys in a past course, maybe like 4 months ago, and they were using it a bunch and they said it was working really well for them. So they were using it with Replit? Yeah. Okay. Yeah. Any API, any API in the world will work with Replit because remember guys, Replit isn't a system by itself. It's just a system that helps you see and work with an AI agent that's writing code. Right. And so if there is an API that exists, Replit can take that and use it, right? And it can look up the documentation. And if you have like a custom API or something like that, you can, you can always ask for that documentation or, or ask ChatGPT for it. Or if you built your own, ask the AI to write documentation for it. The documentation related to an API, guys, usually looks something like this.

In fact, we'll go back to the ESPN API.

It's so awful. It's like millions of— it's a lot of information. Yeah, but, but documentation looks like this. So here's the ESPN one, right? Here's the domain, here are the purposes of each of these things, and then how do you use that? And there's just all these different, you know, things. This is just a random GitHub entry that somebody wrote. It's a bunch of data that I don't need to understand. I don't understand it. Well, I kind of understand it, but, but the reality is you don't need to understand it. All you need to know is that it exists and you just can go copy, go over to Replit and say paste, use this information. Here's the documentation. And then it knows exactly how to do it. It's learned, right? Okay. All right, cool. So now I technically have an AI that's working now. So now here's the smart log. Let's see if it works. Tell me what you did. I played the piano for, 90 minutes. Okay. Effort moderate, whatever. But it gives me a little thing to confirm the activity, log the activity.

Boom.

Piano practice, 90 minutes. That's so cool. I'm shocked, guys. Let me just say this. If I was teaching this class last April and I tried to build something that could do that, it would have taken me 30 minutes to get that right. Like it would have broken. It would have, because, because there were a couple things there. It had to, um, here, let's try something a little bit harder. Um, I, I did a, a kayaking workout on the Provo River. Um, I, I did the run, um, from Bridal Veil. No, I did the— I'm gonna make this even harder. I did the Bridal Veil run 3 times. Um, took me about 45 minutes each lap. Um, I'm really tired. All right, so now it's gonna— so I gave it some information about effort. I gave it a little bit harder information about how long I actually worked on it. Let's see, let's see what it says here. There you go. Look how cool that is! Kayaking, duration 135, 45 times 3, and effort intense. I'm really tired, right? Log activity. Sweet. I'm so impressed at that. I really am so impressed at how well it's doing that. Skills, here's my progress, all my XP's adding up. I'm starting to have a real, a real hero character here now. So that's cool. Um, I can now, if I want, I can publish this. I've got this kind of random thing here. Let's go into the publishing tab. This little tab up here lets you access a bunch of other stuff and you can just kind of slowly work your way through these things if you want. But if you go to publishing, it's the very top one here, I guess. You can change what this is called. So life_progression --flow_state jon.replit.app. Okay. I want to just call it hero. Now hero's unavailable. Not surprising. Hero app available. Okay. The reality is this doesn't really matter because, um, it's a.replit.app. Like I'm not really going to publish it as that. I might publish it, you know, just kind of a beta version or an MVP version or something like that. Um, I can make it private if I want to keep it inside of my team. Um, or I can make it public, right? But I click publish and this will then allow me to then share this with you. If I just share heroapp.replit.app, you guys can go and sign up for an account. Start using it. Okay. I can also, of course, like you saw in the other app, I can connect a domain to it. Right. And I would do that by going into manage and adding a domain. And then we'll, we'll go over that a little bit more tomorrow. But, um, in fact, we could probably do it really quick right now. So we don't have to do it tomorrow, but you would just go to like your GoDaddy account. Okay. So if I want to just go hero, there's no way hero.com is available, even though I'm going to look for it. Come on. Hero search domains. Now, even if I search domains on GoDaddy and hero.com isn't available, I could look around and see if there's something else that might be kind of like it, right?

hero.online, hero.space, hero.horse, hero.loans, hero.golftips, hero.pizza, hero.guide, hero.tech.

Um, I have to Attend class. You're good. Get out of here, Evan. We'll see you tomorrow.

Okay.

Thank you. Thank you. Thank you. So guys, you go in here, you find your place, right? Let's say I wanted to go with Hero App anyway. Let's say I bought it. Okay. And then gonna come in here and I actually am signed in on another profile, so we'll go over to a real one so we can actually look at it. But you would, you would, um, come into your products wherever you buy your domain. Okay, and you have to change what's called the DNS. Okay, um, man, this is going so slow. I'm gonna restart my computer before tomorrow. Um, but I have all these different things, but you can see DNS is like a major thing that's showing up here. So let's say I wanted to set up jennyapi.ai, which actually isn't connected to anything right now. I would just click on DNS, and there's a bunch of just records. You've got a type of a record and you have some values and all you're going to do. I know this feels complicated if you've never done it before, but it's really, really easy. You're just going to come in here and you're going to say domains and you're going to say connect your own domain and we'll say like heroapp.com. Next. And then I would grab now obviously heroapp.com I don't own, I own the genaipi one, but you would put the actual domain in and then you've got a type. A, here's your hostname, and here's your record. And you would come over into the GoDaddy settings and you would say, add new record, choose option A, @, and then you would put the value. You would copy the value from over here, right? Just copy that, the little copy button there for you, set up, and I'd paste it. Done, right? I'd add one more record. There's a TXT record and I click save. My domain is now connected. That's all it takes. DNS is just the settings that point where a, uh, where the internet goes, right? Where, what, what database it's looking at for that. Okay. Very simple. So, um, so anyway, guys, yeah, now I have, now I have a real app kind of, I've got, um, yeah, anyway, we're not going to keep this, that other domain. So, but now I can go to heroapp.replit.app. You guys could go try it if you want and go sign up for an account. I can sign in with my password. You guys even know my password. You go ahead and give me a bunch of skills. Um, but I should be able to sign in. Something's not working. Something's not working. Okay, I'm glad this happened. We've got 3 minutes to try to, try to do this. I don't know what's not working, but we're going to come in here, we're going to go to More Tools, we're going to go to Developer Tools. Failed to load with a response. Okay, so something's happening here that's not working right. So I'm just going to literally screenshot this. And I'm going to come over here. I'm going to drag this in. Okay. I published the app and then tried to log in, but it didn't work. Oh, you know what? I know what the problem is. I actually just remembered the problem. All databases. Yeah. There's no users on the production database yet. That's why I know the problem already. I forgot that. I'm glad I, I'm glad I made that mistake here. So when you publish, guys, you'll have a development database so you can continue to make changes back here. But the production database, this is where if you guys go to set up an account right now, now it's in here and it's, it's real data. It's live. It's out there. I don't want the real world to be messed up while I'm messing in my development environment. And so they have two different databases. So you kind of have to just kind of deal with that going back and forth. So I don't even need to ask it this question. I know exactly what's wrong. But what did happen was I tried— let's see, I published the app, tried logging in, and it failed. But I got no message as to why it failed. There is no user on the production database. Database. Let's make sure we set up proper errors and messages to the user if login fails. Perfect. So anyway, it was confusing to me as the user because I didn't know what was wrong, right? It's just failing to load something. I would click— I'm clicking sign in, sign in, and and it's not working. There's no message, there's no nothing here. So obviously that's not a good user experience. We need to make sure that it's working right. And guys, you know, you don't have to be a genius to understand a good user experience. How many of you guys have 10 apps on your phone? What about 100? What about 200? Okay, I don't know, I probably have 200 on my phone. I might have more than that. How many of you guys have used 200 websites throughout your life, like easily. You understand what a sign-in should do. You understand what happens when you click password reset. You understand a 2FA text message, right? And you can ask Replit to set up a 2FA. Hey, this is secure. I need to make sure that people can't get in and mess with my stuff, right? Bottom line is you understand what's good, what a good experience is, because you've experienced good things. So you could say, hey, I want my reset password to be the same What's the same flow as Netflix's reset password? And it would be like, great, let me go look at what Netflix does. And it would go test it and try it and then come back and build it and copy it. Okay. So just use, just use your, you guys are all smarter than me, right? In your, in your, your domains, use that knowledge, use AI to fill in the gaps, go do cool stuff. So thanks guys. We'll finish here. We'll see you guys tomorrow morning, bright and early, 8:00 AM. And it'll be a little bit shorter. Day tomorrow, which is good. Um, we'll still take a break in there, but, uh, should be

fun. Good luck out

there. See you guys. Thank you. Hey, thank