Cohort 20 - Day 1 - 9-9-2025

Okay. All right, guys. Well, welcome to day 1 of 2 of this AI business class. Really excited to have all of you on here. There's a couple faces here that I recognize, old neighbors and friends, a couple people on here. And I think Abby, you're probably calling from the farthest away. Up in Canada, am I right? I think you're— anybody else farther than Canada? We had someone from Africa last week. No, 2 weeks ago, 2 weeks ago, so that they would have won. But anyway, um, so that's cool, awesome. We've got an international crew here, lots of good people. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11. I think everybody's here. Welcome, John. Um, although, John, your camera seems to be doing something weird. That's just a little like image of a camera or something. Um, okay guys, so let's go ahead and dive in. Um, right now it's 9 AM Utah time and we will go until 1 PM. And I, for some reason, like never need to take breaks. Okay. And so I've had people complain before, like, can we like slow down or stop for a second? And anyway, um, here we are, we're going to make it happen. We're going to go fast. Um, One good thing is this is being recorded. And so because it's being recorded, if you're like, man, I don't know if I caught that thing or that process or that workflow or whatever, it's here, right? If you have to step away for a second, it's here, right? I'll do my best to catch up. I also want you, even though there's, you know, there's 11 or 12 people on here, I want you guys to feel like you can, you can jump in and, and ask questions live. Say, John, hold up, I didn't understand what you said. Just interrupt me. Okay. There's going to be plenty of interaction. We're going to have a few times where I'll say, okay, everybody, cameras on. Let's, let's discuss this topic. And we'll just kind of put down our pencils for a minute and chat about life and AI and everything that's happening with it. There we go, John. We can see your beautiful face now. And Cole, welcome. Looks like we got you as well. We've got a full house now. So, and then other times, guys, you know, you're welcome to have your camera on or off, whatever you want. Sometimes I'm going to ask you to share your screen to show, you know, what you just built with AI. And we're going to do some really cool stuff. So to start off, guys, I really want to do just a quick intro. And because we have 12 people, everyone gets like a 28-second limit. Okay? So we're going to keep it really short and sweet. I want to know your name, where you're from, and I want to know what you do. Okay, and then I'm going to add one bonus question. I want to know whether or not you know how to code. Okay, whether or not you've ever written a line of code, right? And so it's just like, yep, I know how to code, or no, I don't. And if you do, tell us what you know how to do. I like to know who the coders are in the room so that I can, you know, if there's no coders in the room, then I can fake a bunch of stuff and you guys won't even know it. Um, no, I'm just kidding. So, uh, but anyway, let's start with this. I'm just going to go in the order that I'm seeing it on my screen here. So we'll start with Richard, who wears the wrong colors in my opinion, but that's okay. We still love him. He's an at Utah guy, but I'm excited for the game this year. You guys might beat us this year. So we'll see. We'll talk more about that later. Richard, go ahead. Then we'll go to Jake. Fair enough. Richard Fairchild. I am not a coder and I oversee the Trademarks and Licensing Office at the University of Utah. So I work with Under Armour and 400 other manufacturers you've never heard of, along with Walmart, Target, and all the big retailers you've heard of. I love it. And you're based where? You live where? Oh, sorry. Based in Utah. Lehi. I live in Lehi. Work at— work in Salt Lake. Work up in Salt Lake. I love it. Awesome. I'm not actually one of those Utah haters. I just like to razz people a little bit. Um, yeah, I intentionally did, uh, wear non-Utah stuff because I knew that there was a couple of those, uh, fans from SUU, I think, down south. I forget the name of the school, some little school, whatever. Is that a Houston Astros hat at least that I'm seeing? It, it is, it is. So probably another reason to hate me too. Well, you know what, I'm from Houston. I grew up in Sugar Land, and so I'm an Astros fan. I, I grew up going to the Astrodome and then eventually Minute Maid Park and Anyway, so I love it. I'm from LaPorte.

So nice.

Very cool. Very, very cool. All right, Jake, sorry, we took longer than 28 seconds.

I'm Jake Merrill, live in Lehigh, Utah, from Idaho.

And I have a— I work in the swimming pool industry, have a couple small businesses in the swimming pool world, and I do not code. Very cool. That's a modest intro. He's got a great swimming pool company. I know that. All right, we'll go Nicholas and then Carrie.

Hey, good morning, guys.

So I'm Nick Kerlou. I've spent 20 years in aerospace and defense. I'm in operations. I'm in between jobs right now, and I don't know how to code. Awesome. And you're based where? Sorry.

Sorry about that.

In Connecticut. You're in Connecticut. Awesome. Very cool. East Coaster. Love it. Carrie.

Hi everyone, I'm Carrie. Sorry, what were the questions again? You want to know what I do?

Where are you from? What do you do? Do you know how to code?

I am from Utah, actually lived in Houston for a while. Nice. Can move back here. And I am a recreational therapist. I work at the Huntsman Mental Health Institute. And I don't know how to code. However, I've built apps before and just on the Microsoft Office platform, and then I can be YouTube certified in most things. So I'm very confident that I can figure it out. My husband codes, he works in aerospace, and I've looked over his shoulder many times.

And yeah, so you've been around it. I like that. That's good. That's kind of me too. I don't know how to code, but I've been around it a lot. And so I understand a lot of what's going on. So that's awesome. You're gonna do great.

I'm pretty sure that I can write a basic program. I also work in VR doing student courses for virtual reality with the Salt Lake City School District.

Virtual courses, so you teach, so it's just like it's 360 video that's seen through VR or what?

It's, we have some, we have some virtual reality modules that we run as just part of the school curriculum for middle schools in Salt Lake City. Got it. And it was developed by my husband's company. And so I was just— Pardon me?

What's the company?

Uplift Aerospace.

Oh, it's Uplift. Okay, awesome. Very cool. That's fantastic. So I tell— Yeah, your husband probably knows who I am. I think I know. Does your husband own the company?

Yeah, he does. He's the CEO.

Tell me his name.

Josh Haynes.

Josh.

Yeah.

Yeah, I'm actually pretty sure I recognize your name as well.

Yeah, so I was the founder of and CEO of Seek, and we were big-time augmented reality space for, you know, 8 years, right? Big, big company, raised $13 million for that, and I sold out a couple years ago. But, um, that's awesome. So very cool, fellow XR enthusiast. I love it.

Yeah, yeah.

All right, Matt Schneifarth. Did I get that right?

Wow.

Yeah. Well pronounced. Uh, so Matt Schleifarth, not too many people know that last name. Good job. Uh, from St. Louis, Missouri originally. Um, live, uh, in Park City now. Um, uh, involved in kind of a variety of investing and, uh, small business ownership activities. I don't code. Um, I'm interested to kind of learn though. Very cool. Well, we won't be learning how to code today. We'll just be learning how to teach or how to watch AI code, right? And make sure that it doesn't do anything crazy. So we'll be doing some fun stuff there. All right, Ryan, then Ethan.

Sorry, did you say Ryan?

Yeah. All right. My name is Ryan Moody and I live here in Lehigh, Utah. I have been working in cybersecurity for probably the last 20 years.

And I've been in sales for about the last 27 years.

And so I also consider myself an entrepreneur. I do not have coding experience, but I'm super excited to learn more. Very cool. I love it. Ryan used to live about 55 feet from me, so— Yep, stone's throw. We had one house in between us. So anyway, good to see you, Ryan. Ethan, I know your dad and we met on a plane. So, this is crazy. So, Brad Parker's dad owns a company called Private Auto, really awesome startup that's in here in Utah that basically allows people to sell cars to each other through basically an app that just is super slick and super amazing. But I was going to Cabo about a month ago on a flight down there and literally, 3 rows in front of me, I was like, "Brad?" Anyway, and Brad was sitting there just a couple of rows up. Had his family and his son Ethan here. So I've only met Ethan for a total of about, you know, 60 seconds. But Ethan, go ahead and introduce yourself. I'm Ethan Parker. I am 20 years old, so I just recently started college and stuff, but I've had a little bit of experience working with my dad, doing some marketing stuff. I'm based out of— I'm in Alpine, Utah, but currently living down in Provo for college and stuff. But I'm actually really excited about this. My brothers both did this class and and it's been really good for them. So I'm really excited. Yeah, excited to have you. And you just got back from mission from where? Where were you at? Yeah, I just served for 2 years in Brazil. Love it. Very, very cool. Well, that's awesome. Really excited to have you, Ethan. Josh. Bottom of the mains. There you go.

Very cool. Yeah, congrats, Ethan. Josh Butler. I wrapped up some time at a company that was just acquired in May. 20-year veteran of ed tech, education technology. I live in Salt Lake City.

What company? What edtech company?

Company called GoReact, authentic assessment, video-based assessment, was acquired in May. And I saw my exit and I thought, now it's time to start something on my own. And so, yeah, not a coder, but I play one. I've sold a lot of things. I've built products with teams, built teams around it, seen all the work that goes into it and all the requirements to really know what you're doing and the problems you're solving. And if I could do that on my own, or at least as a proof of concept, that would be fantastic. That's why I'm here to learn and appreciate you letting me in late, Jon.

Yeah, we got Josh in like an hour ago. So really excited about that. He slipped under the wire here. All right, Jon, also Schleifarth. I didn't realize there were two Schleifarth. Also Schleifarth.

Yeah, we're brothers.

I'm a criminal defense attorney in St. Louis, Missouri. And I, um, I'm not a coder, but I, I see this as an opportunity to do a lot, both for my law firm and potentially create a software that, that other lawyers can use as well. Very cool. I love it. Well, you'll have some software working about 3 hours from now, so I'm excited for you. Cool. Uh, Cole. Yeah, I'm Cole. I'm actually in a part of Ethan Parker's ward. So we go to church together. And so yeah, we just had a fellow member, he referred this, this teaching or this learning opportunity to me. And so I'm also from Alpine, Utah. I'm 20 years old. I'm staying at BYU, just returned from a mission from China, Hong Kong, just like 3 weeks ago. Cantonese or Mandarin speaking? Cantonese speaking. Got it. I served in Taiwan, Mandarin speaking. Oh, sweet. That's awesome. So I'm super excited about this program. I love that, Cole. Very excited to have you. Hong Kong is a very special mission.

That's—

there's some cool stuff going on there. Well, very awesome. Yeah, I didn't, didn't realize we're going to have all these cool returned missionaries here. That's fun. That brings a fun little angle. You guys have been off your mission for like a month and you're ready to start a business. I love it. Great example to me. And last but certainly not least, Abby.

Um, I, I am a coder, or I was anyways. I haven't worked as an engineer for a couple years. Um, my best achievement would definitely be starting the mobile projects at Rivian and architecting those apps. I have a patent for that, which is pretty cool. It's like a top achievement of a career before changing, um, that I went to product management. So I'm in Canada, um, right now I am trying to start my own business, actually, just because I've been really struggling the last few years with my health. I actually got long COVID and it's just been a nightmare ever since then. So yeah, what are the other questions? What am I doing? All right, there we go.

I mean, I'm a huge Rivian fan. I own the SUV, the R1S, and I've had it since basically when it came out. And it is my favorite vehicle. I've ever owned by far. I had a Tesla Model X Plaid, right, which is, you know, awesome. But I'm, I'm in, I'm a kayaker and I like to mountain bike and I like to go in the mountains, and I can't do that in a Tesla. Like, it's not an SUV. It said it was an SUV when I ordered it. It wasn't when I got it. It was crazy fast, don't get me wrong, but the Rivian still goes 0 to 60 in 3 seconds flat. And I think the software works great, Abby. So thank you for building that and doing amazing stuff.

Um, you know, if you, if you, um, get it dirty kayaking, did you know that most of the interiors you can just spray it with a hose? Like, it's waterproof inside.

I didn't know that. I do throw wet kayaks in it all the time.

Like, like the cab though, the inside, just everything.

I can just spray down the seats even?

Not, not all of the interiors, but most of them. Yeah.

All right, all right. Well, that's cool. That's good to know. I probably wouldn't—

Hey Jon, I think you missed Nick.

Oh, well, no, I— there's two Nicks. I think Nick is Nick Pro, right? Yeah, yeah. Oh, gotcha. Got one for audio, one for video. So, but yes, thank you for keeping me, keeping me honest here. But guys, yeah, this is a fantastic group. We've got young people just back from their mission, ready to tackle life. We've got Abby, who designed one of the coolest products I've ever, or at least a huge part of designing the coolest products ever. We've got law firms, we've got, you know, amazing entrepreneurs that lived in my old neighborhood up in Lehigh. And everybody in between. And so, I'm really excited to meet all of you by the end of this. Hopefully, we get to know each other better. And yeah, so let's dive right in here, guys. To start, we are going to— so, I'm excited that Abby's a developer. She'll keep me honest. That's good. Let me just do my intro really quick. So, I live in West Mountain, Utah, which is about, I don't know, 15 minutes, 20 minutes south of Provo. Most people don't know where it is unless you've actually come here to like pick fruit in one of the farms, right? There's just a bunch of ranches and farms out here. And it's a beautiful spot, beautiful view. Maybe I'll show you guys the view later. The sun's kind of like glaring at us right now, so you wouldn't be able to see it very well, but I can see the Y out of my window, so I'm not that far from a lot of you guys. And, you know, I started this course kind of on accident back in February of this year. I, you know, I post on LinkedIn every day and we'll talk more about social and marketing later. How to use AI to do that kind of stuff. But I, I just kind of gave myself a challenge just randomly as I've been using AI a lot, been working with companies, doing all kinds of stuff. And I was like, I wonder if I could build a new piece of software and get my first customer in 2 days, right? Can I start Thursday and get my first customer by Friday? And I ended up failing. It took me 3 days to make it and I got my first customer after 4 days. So a $15,000 customer. But I built it. It took me about 35 hours to do the process I'm about to take you guys through in 8 hours, right? And so obviously we're not going to be able to do everything, but we're going to go over at a high level basically everything I did and a lot more than that. More, you know, AI has changed a lot since, since February, of course. Lots of new cool models and programs and nanobananas and all kinds of other things in between. That'll be really fun to look at. But who here knows what Nanobanana is? Anybody? Yeah, 2. Yeah, you guys have no idea what I'm talking about. That's cool. Except for a couple of people who are, you know, paying attention a little bit. But it's impossible to stay on top of everything, right? There's 1,000 programs, AI programs I've never touched, right? It's just impossible, right? And so you find the ones that work and I've been able to really put together a process and I've now taught— this is the 20th cohort of this course that I've taught. So I what I built was this company called GenAIPI, the General AI Proficiency Institute. And I'd already been doing some consulting and things like that, but I decided to turn it into kind of, actually my idea was to turn it into more of a consumer product, right? Where you would, you would just kind of land on the site, you would take kind of an AI IQ test, uh, you'd get a score. And then if your score was bad, then I'd try to sell you a course on how to get better at AI. And if your score was good, I'd sell you a certification. So that you could say, hey, I'm AI certified by the General AI Proficiency Institute, and you could throw that on your, you know, your LinkedIn, your resume, whatever. And so that was kind of the idea. And then, you know, quickly realized that there was a lower hanging fruit, I guess, in terms of, you know, revenue, which was working with businesses to, to help them. And, and not only that, working with tons of people like you guys on this call, entrepreneurs, people that say, I want to build something. In the same way that you did that, that sprint. How did you build a business in 3 days, in a weekend? Right. And, and so people said, hey, will you put on a course? And I said, I mean, I guess I wasn't planning on doing this. And I just kind of threw it up for, you know, $1,000, $1,250 and, and filled a cohort. 16 people signed up, boom, really fast. And, and then the next week it filled up again. And so here we are. I did— I looked yesterday, you guys, or I've now had 171 people take this specific course. You do the math. It's a lot of money that just kind of came because I just kind of started working on it and I've been able to do that. And then, actually, I'd say probably about 30% of my revenue comes from this. Most of it comes from working with businesses, consulting, training, and really enabling companies to do just incredible things. And I even have a client or two on this call. I'm so excited to be working with you guys and doing some really fun stuff. I want it to be clear, my goal with this course is not like— I don't know if you guys have ever taken another— I've taken a bunch of courses online and most of the time, they're like, "Okay. Now that you've done this, now, you can pay $10,000 to be part of my inner circle and you can do this." That's not really why I'm doing this. My goal really is for you guys to be like, "I now get it. I'm going to go do it." And if you want to start a business where you're doing what I'm doing, go do that and I will support you. I've had so many students come out and be like, "I'm going to go help other companies." do this too. And so it's just, it's growing. There's about 8 billion people or so that need to learn this, right? And so there's a huge market, right? If you, if you take a few people and you want to go teach them, I'm not going to feel threatened. I'm going to feel very excited and I will support you. I will help you launch your business. I will talk about you on LinkedIn and I will help you get customers. Okay. This is, this is one of the most important things that has ever happened on this planet from a technology perspective, from a business perspective, from a societal perspective. We're going to talk more about that later. It's big. It's big time, right? And understanding the enormity of what's happening is a huge opportunity and a warning, right? And you have to put both of those. Really smart people are looking at both sides, not just taking one or the other. Some are like, yeah, I can't do anything. It's not going to do whatever. And other people are like, AI is everything. And the reality is it's probably both. It's probably some of both. And so, we're going to talk about that. And I really look forward to hearing from you guys too. That's a fun group discussion we'll have in about an hour and a half. So, right now, We're going to all build a video game. Okay. How many people have built a video game before? Anybody? Ethan, that's all right. Carrie, a little bit. Yeah. Abby, probably like in your sleep. I don't know. We're going to build a game that's kind of like, I don't know, Atari style. We'll say it's about that quality, that difficulty. But here's the cool thing, guys. By the time you're done with this course, you guys should be able to confidently say, you know what, I think I could probably build a plane in my backyard. That works. I think I could probably do it, right? And it's not because you know how to build a plane, it's because you know how to work with AI to have it guide you through the process to do that, right? And building a plane is a lot harder than code, right? Code can just— you ask it to do something, it just whips it up. Okay, I asked you guys to sign up for ChatGPT, paid version of that. On the $20 version, it's fine. If you're on the pro version, even better. That's cool. That's $200 a month. And so might not be for everybody, and that's totally fine. Grok is another great tool. We'll talk about why. We're going to start with that one right now and then Replit. Replit is, in my opinion, the best— it's called integrated development environment or IDE that's out there. We won't get into that for another about 30 or 40 minutes. But today we're going to start with Grok. I'm going to share my screen and we're just going to dive in here. Let me close out all my embarrassing windows here. I'm just kidding. All right. Well, I'm only half kidding. Okay. By the way, you'll— I'm doing pretty good right now. I only have about 30 tabs open between, you know, a few different, a few different browsers. But by the end of this course, I'm going to have like 100 open. We're going to start with x.com. Okay, so go ahead and just watch for a minute here. And then I'll let you— I'll set you guys loose. Okay. So a lot of people ask, you know, why use ChatGPT? Why use Groq? We'll talk about that a little bit more later. What we're going to do right now is I'm actually going to not have him do this. I'm going to say let's go fast because that's totally fine for this. We're going to say build me a mind-blowing visualization using p5.js and show me where I can deploy it. p5js.org. All right. So p5.js, guys, is what is called a front-end language. Okay. Many of you might not know what that means, but that's okay. It's very simple. Um, uh, front-end language is what you can see, right? You're building with a language that you can see. We're not using a backend database. You got frontend and backend. Frontend is what we see here on the screen. Backend is in the database. It's all the code and logic that happens behind the scenes. Although logic can happen in the frontend too. And that's what we're going to do here. Right? So, All I did was give it one tiny little sentence and say, hey, just give me a visualization using p5.js. I told it what language I want. We'll go ahead and click copy here. I should be able to just click on this editor.p5js.org, that'll open in a new window. I'm going to go ahead and we're going to get rid of this software thing. It's cool. I'm not going to donate right now, but we're going to go like this and paste and we're going to click play. Okay. All right. I click play here and we had an error. Okay. There's some sort of issue here. Okay. This happens 9 times out of 10. You're going to have an error when you first paste in your code. So I'm going to do is I'm just going to copy this and I'm just going to give it to this. I'm just going to give it the code or the error, right? All I did was copy it and paste it. Okay. So it's going to figure this out. Here's the corrected code and we'll make this work. I could click think harder if I wanted it to do something for some reason, but for this type of a thing, this is a very basic thing for AI to do. We'll talk about how AI thinks a little bit later. For now, we—

Jon, can I just ask a quick question?

Go ahead.

You're using Groq within X. Yep. Do— I just went to groq.com. Like, that's— is that fine or does it need— there's no lost functionality?

No, you actually have a little bit more functionality on Grok.com, right? I just do it on X because I like to have the X premium, whatever. Anyway, so that I have my, you know, check my blue checkmark. But no, no other reason.

Got it.

Yeah.

Virtue signaling.

Exactly. Yeah. I just want to feel important. So. All right, so here's my mind-blowing visualization. This is kind of— I don't know if I call it mind-blowing, but I mean, it is something we created, something. And, you know, we can look at it. I don't know what is happening here. It's kind of, it's kind of interesting at the very least, right? I don't know what it is. But anyway, so guys, what I would do, what I want to, I want to make this clear. This right here is called a bad prompt. Okay? And there's a lot of reasons for that. One of the primary reasons is that it has almost no instructions, right? Imagine I hired you, Richard, right? And you're on my team now and, you know, I've hired you as my COO, right? And I say, okay, Richard, welcome to the team. You know, go put together some reports. That's, that's the totality of instructions I give you. You'd be like, okay, what kind of reports? And I'm like, sorry, I can't talk right now. I got to go. So you'd go put together some reports, you try to dig into the books and figure stuff out, and you could probably make something valuable. And then you'd come back and be like, here, what do you think? And I'd be like, this is not what I wanted. Right? And so, so many people give instructions to AI that feels like that, where you're just like, hey, do something, but you don't fill in the gaps. And so because, you know, I didn't fill in the gaps when I gave Richard that instruction, he has to fill it in. Because I didn't give Grok, you know, more instructions here, it just has to be like, all right, well, I guess let me try and make something mind-blowing. I don't even know what that means, right? Now, sometimes I get something that's really cool by doing that. I could use this same prompt over and over and over again. All of you could copy the same prompt and you would all get different results. Okay? And, and it's because it's just going to think, it's going to just reason and go forward and figure something out.

Okay?

So let's see if we can make a little bit better prompt. All right. So let's try this. Let's say, now build me an interactive Now let's just say an animated model of the solar system. Make it use a neon theme, add some stars to the background that twinkle, and have the entire background slowly rotate so it gives the scene Life. All right, let's just start there. Let's just start there. So now this is actually a pretty detailed description, even though it's only a couple sentences or a few sentences long. When you say the solar system, right? Again, if I was going to Abby and I said, Abby, please go make me, you know, you know, a model of the solar system that works. We're going to have this be the background of the new Rivian app. Okay. She would have to do a bunch of research, right? And she'd have to figure out all this stuff. Let's define the planets and figure out their distance and speed and relative, all this stuff. Because she probably doesn't know that off the top of her head. Maybe you do, Abby, but, but I don't, right? I don't know that stuff. And so I'd have to go do that. But AI can do that research immediately and then apply it and then put all this whole thing together really, really fast. Right. And so then I gave it, okay, here's the color, here's the idea, this is what we're doing. I didn't tell it 2D or 3D, so it's going to have to kind of, you know, figure that out. So I'm going to delete this and start over. We'll paste what I just copied there. And here we go. I have now a working version of the solar system, a working model of the solar system. Pretty dang cool, right? So this is a little bit better prompt. It's now working. It's not interactive yet, but it looks pretty cool. Like they even put a ring around— is that Jupiter? Not Jupiter. Uranus. Am I saying it wrong? It's Saturn. Saturn has the ring. My brain is not working right now. I'm not an AI, so I get things wrong. But let's take this one step further, guys. Okay? And then I'm going to set you guys loose. Now, you can obviously make this more interactive. I'm going to do something simple. Make, uh, let's see, um, this is great. I want to just kind of let it know that I like what it's doing, so let's keep it the same. When I click the mouse, it should shoot, uh, vibrant shooting stars out of the cursor that are different colors each time. The shooting star should have a long and beautiful tail. I can iterate on this just by talking. Oh no, we got an error. That's okay. We just copy and we just do it again. This happens sometimes. We're going to refresh it, see if we can get it to wake up. Paste, go. Groq is broken on me. I don't know why. Please open a new conversation with a different request. I've never had it actually just die on me like this. Let's try for fun. Let's try going to auto. And if this doesn't work, then we're not going to do it. Okay. Bottom line is I could do this. Theoretically, this should work and I should be able to then just paste the new code and start making it interactive. So I want you guys to now take your own idea, come up with something. I'm going to give you the prompt, the very first prompt. Okay. And I'm going to debug. Well, I'm going to see if I can get Groq to work again. I highly doubt it's actually shut down, so it should work fine. But I'm going to drop this prompt into the chat here for everybody. And yes, Abby, I do love Arc Browser. I am somebody who needs to use it more. I have it right here. It's sitting right here, but I just— I'm not using it anyway. Mind-blowing. Okay, so take my prompt and instead of mind-blowing visualization, replace it with, you know, a clone of Snake or a clone of whatever, or make a game where I'm running from dinosaurs or make Whatever you want, some model of something. Okay, come up with something cool. I'm going to ask some of you guys to share your screens. Okay, so just go ahead and dive in and feel free to ask me questions as you do this. Okay, by the way, if you're in Groq, if you're working right on groq.com, or if you're not in Groq and you're in ChatGPT or whatever, there's also the ability instead of having to copy and paste the code over, you can run the code right inside of Groq as well. That's another option, right? But I want to show that even if you don't have that and you're using an LLM or you're just getting code, You just say, where do I do this, right? I happen to know it's p5js.org where I can just copy paste this, but you just do what it says. Just listen to the AI, copy paste.

Okay, so Jon, if I press the play button in Groq, I, by the way, I'm just a brand new user to Groq a few days ago. So that's why I'm asking all these questions. If I press the play button in Groq, I just at the very bottom get like a green check mark.

Okay, let's see it. Can we— does it?

Yeah, does it actually work for you?

Let's see, it should, it's supposed to. Yeah, yeah, that was the video one. You have to click the share button and click entire screen or window or whatever you want to do.

Oh, it's been a long time since I've been in Zoom. Sorry.

This exercise is secretly one to help you, you know, build some confidence with AI building your first thing and two, secretly to see if everybody knows how to share their screen.

Right here you see the The green check mark on the bottom right. Yeah.

Yeah. Can you make your window larger? Because you're kind of in the split screen right now. I'm just wondering if somehow it's just hiding it.

Oh, no, I just put it in split screen.

Oh, you just barely did that. Yeah.

Does it run down here for you?

It's supposed to run. It would run in a window that would open to the right. So you click the Run button up there.

Interesting. Yeah.

Try refreshing for fun.

Did you see how when I press the button, it does process— it processes something and then it gives me a little check.

Are we just working off your first basic prompt and then customizing it? Yeah, use the first prompt and just, just change mind-blowing visualization to whatever you want, right? So like a model of a zoo or a model of a pool or whatever it is. Abby, Describe that problem to Croc. See what Croc—

describe the problem. All right, I'll do that. It is working on the other side, of course.

Yeah, I saw that you did have something running there. Yep. And then whenever somebody has something running, please, please share. Just speak up. Doesn't have to be perfect, doesn't have to be amazing. This is not the type— p5.js is a very, very low-level, you know, graphics engine. It's not— you're not making Call of Duty on here or Fortnite. You're going to be You're gonna be making something that looks more like an Atari or a Nintendo game. We can still create some pretty fun stuff. Like, you create a working Tetris model or something like that. There's a lot of fun things you can do.

I'm on p5.js, and I'm not seeing, like, where to paste in the code. Oh, sorry, lower left-hand corner, start coding. Okay, sorry, I figured it out.

Awesome. Now you're good. Yeah, ask the questions as they come. No bad questions in here, guys, really. Like, this is, this is new to everyone. This was new to me a couple years ago, and I had to figure it all out. Hey, John, I'm getting a— I got two different prompts. One is to pay some stuff into the index, and the other is the paste into the sketch. So do I have to do both? And I don't see where— yeah, yeah. So I got the other. When you did the prompt, did you ask it specifically to do it on p5js.org? Like was in my— yeah, my prompt. Yeah, say, say, I need you to just type, I need you to format this so that I can paste it into p5js.org. It's interesting. I did that and, and then it, uh, gave me an error. Then when I did the error code and put it in, that's what it gave me, the two different, uh— yeah, just, just say I need it in one, one spot. I need to be able to put it in p5.js there. Yeah, what it's probably trying to do is set up like an entire web page for you. You don't need that because p5.js takes care of that little element. Are you coming, Ethan? You got something? Yeah, yeah. And I'm making a game where a dinosaur goes around eating things. It's working. That's fun. Let's see it.

I'm just doing an edit on this one.

I'll share my screen. Yeah, you're good, Richard. Keep working that. If, if we can't figure it out in the next minute or two, um, we'll have you share your screen. We'll debug it.

I can show you something. I don't know if it's awesome, but it's a start.

Yeah, see what you got. Let's hear—

oh, let me— I'm going to do my share issues as well. Where is that? Ping type. I'm on a new screen too, so that's going to be fun.

See here.

I'm gonna have a problem sharing for some reason too. Give me a second.

Yeah, you might have to reset your— you might have to like allow some system preferences and then close Zoom and rejoin. Yeah, that's okay. That's why we do this, guys. Yep, that happens. So go ahead and, you know, do that. I'll let you back in as soon as you click. You got it. Carrie, how you feeling there? You got something working? Sorry, I know I'm making you bounce between screens.

That's all right. I am making a game where I can swat flies.

Where you can swap what?

Swat flies with a fly swatter.

Oh, you can swat? I thought you'd swap. No, um, swat flies. That's a good one. That's a good one. That would be— I hate flies. I hate them. They are the bane of my existence, especially out here in Westmont. We have 5 acres and, and, you know, we've got horses there, we've got horses across the street. We have a horse, but it's— we keep it, we keep it about a mile away. But anyway, there's just lots of flies around here, and so I've learned about all the cool fly traps.

Um, anyway, I'll have to pick your brain about that some other time. I have chickens and they create a lot of flies too.

Yeah, yeah, we have about 50 chickens as well, and that's that happens.

So I do not have 50. I have 11, but I live in the city, so I'm not allowed to have 11.

I know you're not allowed to have 11. So we— when we lived in Traverse Mountain, up kind of by Jake and Ryan, these guys, um, and I think Richard— but anyway, uh, we had— my wife had, I think, 10 or 11 or 12 or something in that same kind of range. You're only allowed to have, I think, 6 in that area.

Yeah, you know, Rebels Without Causes.

That's right. That's right. All right, Richard, how you, how you doing there? Did you get past the— No, but now I'm trying to go back and I made it too complicated, probably. So Guys, don't, don't stress over this, okay? You're doing great. It's okay if we're waiting here for a few minutes. I want you guys to take 5 minutes and just get it right and just say, okay, I can build something by just talking back and forth to the AI, right? Don't think about the technical side. Think about— act like Abby went from being a developer and now she's a product person, right? And an entrepreneur. But, but when you're a product person You just, you just say, this is what I want the experience to be, right? You just talk about the experience in the same way you would talk to, you know, just another human, right? Hey, I want it to do this. Hey, when this happens, do this, or this happened and it broke, right? Just talk to it and you can, you can iterate. How you doing, Nick? You got something to share? Yeah, I'm a car guy, so I decided to just take your, um, visualization and make it about cars instead. Well, let's see it. See what we got.

I use ChatGPT instead.

Just totally fine. Totally fine to see if, uh, you know, the kind of the coding would be longer or more exhaustive. But, uh, can you see my screen? I can. How do they decide where to go? Is it— does it follow your mouse? So I did, I did not get into manipulating the cars yet. I just wanted to kind of random. Yeah. ChatGPT would respond the same way.

Yep. So I'll keep playing with it.

Very cool. I love it. Awesome. Good.

Thanks for sharing.

Yeah, that's perfect. You passed A+. Ryan, how you doing? I can show mine. I mean, I tried to Hold on, let me see here. That's right, let's see it. I tried to add more detail, but it wasn't really— That's okay, that's okay. Don't try to perfect here, right here, guys. I want you to just get something running, right? That works. Yeah. Okay, here we go. Oh gosh. This is my first time using Zoom on this computer. I think it's wanting me to like quit it and reopen it so I can allow it to— So go ahead and quit it, enable the setting or whatever, and yeah, and get it to do that. I'll put you back in. Okay, awesome. All right, yeah, this will be much more important, guys, when we're sharing our real projects and I actually want to see what you're doing, then you don't have to quit and come back later. Okay, um, who's got something to show that hasn't done it yet?

Uh, I have a quick tip. Uh, I was getting errors and then I looked at it and I realized they had given me HTML. So anyone that's not, um, proficient with coding languages, if you're getting a random error, go back and say, are you sure that's the right format? Because when I did, it was like, oh, you're right, I gave you an HTML file, and then they gave me the right one.

You want to post the game into the p5.js online editor or just do preview? You can just do preview if you're— yeah, if you're just in, in Grok or in ChatGPT or whatever, you should be able to just click preview and it should pop up. So you don't have to use the p5.js editor. Right. I just wanted to show that as an example where you can do that.

I'm ready to share mine whenever.

Let's do it, Carrie.

Cool. Ethan, I'm sorry to jump in front of you. Is that okay?

You're good.

I'll go right after you.

Sorry. I'm glad that we're doing this little— I don't know if this is going to— oh, I just have to go through settings. Hold on.

That's what we— that's why we do this, guys. I'm telling you, like, this is, this is the secret thing. All right, Ethan, go for it, and then Carrie, when you come back, we'll get you in.

Cool.

All right, let me know if you guys can see.

Yeah, we can see this green screen with— are those birds? Are they dogs? Are they— what is this? It's supposed to be dogs. I told it to be dogs, and it's supposed to be like a dinosaur, but I can control it with my keys, and it like counts my coins up top. Like, I get a coin each time I pass over one. The time you eat a dog. Yeah. Is it an endless scroll or is it just you just trying to pick up everything on the— see, nope. Yeah, it's not an endless scroll. It stops right there. They kind of reproduce. Yeah, they do, they do spawn again. That's cool. Um, that's awesome, right? So I love what you did, right? You get, you get the A+ as well because you added interaction and you added a score and those types of things. And so I love it. I think it's super cool. Now, obviously, guys, if you're gonna publish some sort of a game like this, you would— you'd probably not do it with p5.js. You'd probably use something where you can upload some images or some 3D models or whatever it is that you're trying to do, some sprites, right? Maybe some animations, things like that. And by the way, Grok and ChatGPT, they're all capable of even creating sprites. A sprite, by the way, guys, is like a little animation where you have maybe like 4 or 5 kind of frames of a thing where it's like you see, you know, an action. It looks like a sword is moving or somebody's cheering or, you know, it's kind of a loop, right? It's like a, it's like a very early version of a GIF, right? Put a bunch of these little sprites in a row and then you just put them in sequence and that creates an animation. So anyway, um, really cool. Love the interactivity. Richard, how are you coming there? I did get something but not functional, it doesn't look like, which is, which is fine. Um, and then I'm trying to do an ChatGPT instead, see if you can get it over there. Yeah, yeah, I broke Grok. I tried it a couple more times, it will not listen to me on my, on my thing. So, you know, it is what it is. But, um, all right, who hasn't gone? I won't make everybody go, this is kind of an optional thing. But, um, who's— Carrie, are you— did you get your setting right, or—

I'm having a hard time changing my settings to allow.

You'll have to, you'll have to quit Zoom. You'll have to save it and then quit Zoom and rejoin. That's the only way to do it.

Okay. I was just saying that I need to go into my security settings and allow my system settings.

Yep, that's exactly right. So you'll click allow and then it will say you'll have to restart Zoom for these settings to take effect.

Oh, okay. I haven't even been able to find the allow button yet.

So I'm doing the same thing. I'm trying to— I've gone in and out and actually jumped in. Usually I can enable this and no problem, but for some reason today it's being funny.

Yeah, I don't normally use Zoom. I usually use Microsoft Teams.

I hate Teams.

So yeah, it's not great, but you're used to that one.

Yeah, I get it.

Okay, I'll be back.

You found it. Awesome.

Very cool.

Matt, it sounded like you were about— you were ready to show something. Yeah, let me share the screen here. Okay, so I wanted to create a video game that, uh, got my son excited to learn how to play tennis.

Mm-hmm.

Um, right now I'm not moving the racket, but it actually did work a minute ago. Um, yeah, you probably have to, uh, click on the green area. If you click on the green, it probably will— will— oh, there we go. Okay. Yeah, yeah. So he actually came in while I was in here and he loves it. So it worked. Yeah, that's cool. You basically made a little Pong, but it's tennis. Yeah, yeah.

Star. This was my absolute—

gives him encouraging remarks. So it's cool. Yeah, that's pretty cool. I love it. Awesome. Also interactive, um, simple game. But, but again, guys, as non-developers, I don't know about you guys, but this should feel pretty fun, pretty cool, right? And the more you do that, like, it's really easy. Like, we've had a couple of weird errors this morning and we're trying to figure out how share screens and stuff. But usually you can do this stuff, you can whip something up really, really fast. Like my kids know how to do this. My kids are ages 8 to 12 and they'll be like sitting in church and be like, dad, can I make like a temple matching game using AI? And you know, they won't use p5.js, they'll use, you know, JavaScript and other things and import images and stuff like this. And let's sit there on my phone and they'll like create this cool, you know, church game or something, right? And, and I'll, you know, whatever, I'll let 'em do it. My son right now is working on a full-blown thing inside of Replit. He's building a website where you can come and you can make 3D comics of dinosaurs with bubbles and chats and things, and you can then publish your comics and stuff. Unbelievable. Creating real software that creates real value. So cool. All right.

Do you mind, Jon, if I share that just to make sure I can?

Oh yeah, do it now, Carrie. Welcome back. Yeah.

Awesome. Hold on.

Now that you work through that.

All right. Can everyone see that?

Yeah, it says it started, but I can't see it quite yet. There we go.

Now I can see it.

Yeah.

So I've got my little fly swatter and I can swat the flies.

Nice.

Chase them around. If I hit the vase, it falls over.

That's awesome. Are you using your arrow keys? Is that how you're moving it? Or WASD or something?

Or I'm using the trackpad.

Oh, it's the trackpad. Okay, got it.

Yeah, there you go.

I love it. So you made working software. I love it. Very cool. I got mine as well. Let's see it.

All right.

You see my screen? Yeah. This is a shark chasing fish game. It looks like you have to hit, hit it like from— oh no, it's working on the side too. For a second I thought you had to maybe hit it forward or backwards or something. That's cool. Yeah, it's kind of just like an endless game, like, uh, you know, a time. Yeah, it's just like— yeah, I didn't give it too much. I didn't give it enough parameters, but it's still something. No, it's awesome, right? I mean, it's 115 lines of code and it works. It's code that works. Jake, how are you doing there?

Good.

I couldn't get mine to work on the P5.js, so I popped it into Replit and that seemed to work. Let's see it. Yeah, Jake's jumping ahead. And he has a little bit of an advantage because I've showed him some of these things. But, but let's see what you built.

Okay.

And Replit, guys, by the way, we're about to go into Replit. It's much, much more advanced and it's— it can do incredible things which are really fun.

So I just built a little car game. This is supposed to be a Corvette in Las Vegas. Dodging other cars, but it's pretty rudimentary.

That's okay, rudimentary is fine. Um, awesome, it looks like it's fully functional, no problemo. I've seen some pretty amazing games created with Replit. Replit can do like, you know, it, it really is focused on kind of more, um, I don't know, like if you want to make a really good game, guys, you're going to use something like Unreal Engine or you're going to use, you know, a real, you know, game, game engine. But this is awesome, this is really fun. Well done, Jake. Okay, um, all right guys, is there anybody that has not shared that would like to share?

I'm trying to get mine still set, Jon, and it is— I've done this hundreds of times to enable screen sharing, for some reason just doesn't want to work today.

Josh, will you share your screen really quick and maybe we can debug together? Sometimes the power of 12 people looking at your screen helps.

Well, I'm— that's the point, is I actually can't share it at all in any way.

I was just— man, I wonder if there's—

there it is. Okay, let me see. I'm good.

I'm good.

You got it working?

Yeah, I'll be back in just a second.

Yay!

Awesome.

Cole, what you got? Yeah, it's kind of different, but I had this like math equation game with like the water cycle. Maybe I could show it real quick. Let's see it. Let's see it. It sounds awesome. Um, oh yeah, I'm having trouble as well.

Great.

Yeah, go ahead and push all those buttons and let it give you the permission. All right, Cole's on. Holes on it. All right, Josh, let's see what you got.

All right, lucky me, finally. Okay, so this is a, a video game. These are supposed to be snakes boxing in a boxing ring, which is ironic, um, and as you— as they're fighting, they're supposed to punch every 3 seconds. And if you click on one, your favorite, it'll grow every time you click on it. So they'll kind of go after each other.

Yeah, yeah.

That's it.

That's— they just kind of fight each other and you just kind of pick who the winner is. Yeah, yeah, that's, that's random and kind of cool.

Well, just whatever came to mind.

It's something, it's something working, right? Um, the snake is punching. You've got a snake punch function.

So those are supposed to be boxing gloves too.

So I love it, I love it. Okay, well, thanks for sharing. Glad we got the screen sharing working. That's fantastic. Okay. And Cole. Okay. Sweet. Okay. Do you see my screen? Yeah. Click allow there maybe. Yeah. Yep. We can see it. Let's see what you got. Let me put this full screen real quick. It's kind of— so basically you got to put in the equation before, uh, the moon sets, otherwise a new equation will come up. But once you hit 5, then you should be able to win the game. Oh shoot, that's a little harder. Okay, wait, let me try to get this real quick. Oh, there we go. Yeah, and then you win, dude. That's amazing. See, that's a fully functioning game with, you know, a goal you're supposed to get. It's pretty fun. It's pretty cool, pretty creative. Um, all right guys, I think we're good. Anybody, anybody else want to share? I think we're probably— we're good, but you know, if you want to speak now or forever hold your peace. Um, we're right on schedule, guys. Um, 9:55.

I am curious, Jon, what is, what are the most interesting games that you've seen people come up with?

You know, I've seen, I've seen some like really good Tetris, you know, clones I think is really cool. I've seen some Space Invader games that are pretty fun. I've seen, I'm trying to think what's some of the more really interesting ones. Well, one that I made actually, here's a cool story and this is a fun party trick that you guys now have available to you, but I was sitting there talking to this girl that I used to date. And she's like, she's not a cousin. She's like the best friend of my cousin. So she's at like all party things. And we dated for like, we went on like 5 dates, right? But we were really good friends anyway. Long story short, we're sitting there like 3 months ago. And, you know, just a bunch of us talking about whatever. And, and I was talking about some of this AI stuff. And she's like, that's really cool. I need to take your course. I really want to build this game. And I was like, Tell me about the game, right? So, she's just talking. I'm just like sitting there kind of typing with my phone, just kind of like, just listening, just kind of letting my fingers do their thing. And then I just click go. And she had basically said, hey, I want to build like a Flappy Bird game, right? A Flappy Bird game, but with bees and you go and you collect these things and pollen. And then when this happens, that happens, whatever, anyway. And so, we're sitting there literally for like 2 minutes while she's talking. And then I'm like, kind of like this. And I hand her my phone and the game's working. Right? And she's like, what is this? How did you do this? Where did this come from? Did somebody else make this? I was like, no, I just did it while you were talking to me, right? And so anyway, you can do some pretty cool stuff. I mean, being able to whip up a solution in seconds is amazing. You know, we're going to dive into Replit now here. And, you know, I've had the experience, guys, of raising you know, a lot of money. You know, we raised $13 million for one of my companies. I raised $3, $3.5 million for another company. I'm in the process of raising literally $2.5 billion for another company. But I've had to raise money for these other software companies because I don't know how to code. I don't know how to do this. And so, I had to pay people. And when we wanted to whip up even a simple game or a simple, you know, and we actually were working on some game stuff back in 2016 with Seek. Which I could talk about more later. But, but if I wanted a new feature, if I wanted a button moved, if I wanted something, right, it was, all right, well, let us add it to the list and, you know, we'll get back to you in a few weeks, right? Or, hey, we have this idea for something and like, okay, well, give us a couple of weeks to spin it up and see what we can do. Now, literally during the meeting in which you have the idea, you can have a working prototype. Like now you can have it ready to go right now. And it's just incredible what's possible.. So, um, you guys are all now developers, which is pretty cool. You've all done it. You've all built something that works, turns on. It might not be perfect. And by the way, guys, I handicapped you. I gave you guys a not ideal coding environment, right? ChatGPT is not the best coding environment. Groq is not the best coding environment. I had you use a language, p5.js, that's kind of finicky and kind of weird and does some weird stuff. I handicapped you for sure. And, and that was on purpose, right? And so we're gonna go from there. We're gonna now go into something that's a lot cooler, right? And so this is another time where you just kind of sit back and, and watch for a second here. We're going to dive into Replit and I'm going to see if I can find Replit and we're going to go home. They just changed like the layout of Replit. And so it's kind of— anyway, I'm— it's going to take me a second to get used to it. Okay. So one of the things that I like to do when I teach this course every time is I like to come up with a brand new idea, something I've never done before. Um, and so I want to build a business, like I'm gonna, I'm gonna build like an actual business that could be viable, that could be making money by, by Friday.

Okay.

Um, so I just got to think about what it is that, that I wanna, that I wanna do. Um, let's, let's think here, let's think here. We've got a lot of interests. I'm a musician, as you can see, I've got a piano sitting right here and If you look a little more to the right, there's a little saxophone. Anyway, um, and I love kayaking and I love movies and I love— by the way, speaking of movies, I just finished The Paper last night, which is The Office kind of spin-off thing. It's good. I finished all 10 episodes even though they only came out like 2 days ago. Um, I, I finished— I finished TV shows fast. Um, all right, I gotta think here, guys. I need a business. I live on a farm. What is something that causes me problems? Okay. Yeah, that's a pretty good idea. Okay, so I'm going to build a, um, like a kind of a chore tracker, but like more advanced, right? So one of the things that's really tough is when I have— I have 5 acres here and we just have a lot to take care of, lots of just things, right? For example, I have to change the filter. I have to clean out the filter on my irrigation. And I have to, you know, and I have to do that once maybe a week, right? And I have to— I have to make sure that I, that I, you know, push out all the, all the mud out of some of the pipes. And I have to do that maybe once every 3 months. And I have to do— you know, there's all these little things, but it's really easy to say like, man, when was the last time I did that? Right? And so I want to be able to keep track of it and create something that's like, okay, it's green, meaning I did it, it's good. It's yellow, like, hey, it's starting to get bad. And red, like, you need to go do this thing right now. Like, you haven't cleaned your garage. In 6 months and it's disgusting. Go do it, you know. So that's, that's kind of the idea. I've had that idea for a bit and I'm going to build it today. So let's see here, build me a chore tracker using React. Chore tracker using React. I have lots of chores to do around the house and farm. And I want it to notify me when I need to. No, I want an easy visualization of the current status of all chores around the house and farm. Let's use a color system. Green means it's been done. Recently yellow means it's time to get to work, and red means I'm way past, past due. Let's use a color system that, what do you call it, decays over time based on the setting of the chore or the task. And how often it needs to be done. Okay. All right. We're just going to start with that. Okay. So here we are. We're going to build a new app. And I'll tell you now, guys, that this is what I would call a bad prompt still. Okay. I usually start with a bad prompt just so that you can see what kind of things you can do with a bad prompt and then how much better it can be with a good prompt. Right? A really good prompt for Replit is going to be about 5 pages long. Okay? And we're going to be able to whip up that kind of strategy in 5 pages and all that here, here in a little bit. We'll show you how to do that. But Replit is still quite powerful. It might build something just amazing. Right? And, you know, the way that Replit works, we're going to initially get what's called a visual preview. So it's just going to kind of think, okay, what— how does this work? And it's going to start building something, and then it's going to really get to work. So let's just go approve plan, and it's going to start designing this visual preview. Okay, now while it's thinking, I want to talk a little bit about how this AI is working, right? And how you can, if you understand this, it's really easy to be like, oh, I understand kind of what it's doing now and how it works. So I want, we're going to have to do, we're going to pick on someone here. We're going to pick on Ryan. Ryan, I want you to share with the class, please. How do you get ready in the morning? You wake up. Suddenly, I was awake, right? Then, then what, right? What do you do? What's your routine to get ready?

Well, for me, it's usually I'm getting up and then working out. After workout, then shower, get ready for work, eat breakfast. That's kind of the— and then start working. That's usually a typical day.

The Moody household. Yeah, I love it. Okay, very simple. Let's zoom in on one thing that I know you didn't mention specifically. You said get ready, but in, in there, I'm sure you do it every morning. Tell us how you brush your teeth. Walk us through step by step how you brush your teeth.

Uh, put the toothpaste on the toothbrush, pick up the toothbrush, put the toothpaste on the toothbrush, put that toothbrush under the water to get it wet. And then start brushing my teeth in a circular motion from left to right.

Uh, do that for roughly 2 to 3 minutes and then spit out whatever's in my mouth, rinse, wash off the toothbrush, and turn off the water. Amazing. Okay, perfect. Thank you. Awesome. Very simple. So what you're doing there and what this AI in front of us is doing— by the way, isn't this amazing, guys? Isn't this crazy, like, that it can just whip up something like this, like I can use this today. I could say, hey family, we're going to start using this chore tracker, right? I probably have to integrate a bunch of different chores and figure out all that stuff. But there's even fast ways I can probably figure out how to input a lot of things really quickly. So, but anyway, what's happening here, guys, is doing something called chain of thought reasoning. Okay, so chain of thought reasoning means you take an overarching task, right? Like get ready. Okay. And then you break it down into, all right, well, I'm going to work out, I'm going to brush my teeth, I'm going to go to the bathroom, I'm going to shower, I'm going to eat breakfast. I'm gonna do whatever these things, say my prayers, read my scriptures, whatever it is that I do in the morning. Okay. I've got all these different things. Get ready. Okay. And so then when you start to dive into those, you can say, all right, let's take get ready. All right. Well, get ready is going to be, I got to shower. I got to find some clothes. I got to brush my teeth. I got to do these things, right? I'm going to floss. I'm going to use mouthwash. I'm going to brush my hair. I'm going to do whatever, do my makeup. There's all these different things that can break. And then you even get down smaller to that, right? I'm going to do my makeup. Well, doing your makeup might be 10 more small tasks, right? And doing all this stuff, right? And so we're taking larger tasks and breaking them down into smaller things and going out, and then we, we complete those smaller tasks and that then helps us finish this task. And then we're done with that. Then we move on to the next task and then we finish all the subtasks and then we come back. And so that's what it's doing here when we say, okay, hey, I want you to build me an app that is a, you know, a chore tracker. Well, it's going to say, well, most apps need some header. We need a menu, we need settings, we're going to need someplace to add chores and we're going to need to do this and there's going to be some status something here and we're going to be able to add new chores here. So we'll put a little plus button down here. So it's just thinking through, okay, what do I think I need for this? Now here's a visual preview of that. Now I'm going to go back in and I'm going to start building. What it's doing here is it's building all the different technologies that are running this, the schema that tells your app how does this all fit together and where are we going to put storage and routes, where do the buttons go and how does this all happen. Now we're going to do the index, which is your homepage. It's going to create all these crazy things and there's even a nice file system here that shows all the different files that have been created. You can jump in and start to look at those things. You could click on it and read the code if you want. If you're a coder, I don't know. I don't know what all this stuff means. I kind of do because the green tells me what this function does, right? I need to get all the chores, right? I'm going to get a single chore. We'll talk about GETs and APIs and stuff later. We don't need to worry about any of that stuff right now. But bottom line is these green, these green parts of the Markdown are basically saying, hey, this is what it does. This is an actual code. This is just kind of a label, right? Which is really useful, right? If I were to hand this over to Abby later and say, hey Abby, will you help me build this? And finish this app, she would be like, okay, so what is it doing here? Right? And it would be able to quickly understand. She'd be able to understand because the code is, you know, it's been, it's been detailed a little bit. Okay. So none of this stuff works, by the way. If I try to click on any of these things, nothing will work. And so what it's done is it's created the visual preview and then just kind of pausing and then it's creating a ton of stuff behind the scenes. And in, I don't know, 5 or 10 minutes, it's going to say, hey, I have finished. Okay, I finished. Here it is. And I can start playing with it. I can add chores. I can do all this cool stuff. All right, my understanding now is depending on the complexity of what you give Replit, it can work for up to 2 hours by itself. Now that is a brand new function. The most I've ever seen it work by itself is about 15 minutes before. I am in V3 of Replit, so my Replit might look a little bit different than yours. V3 hits the streets tomorrow, so you guys, I'm in a beta program with Replit and I'm part of their community kind of managing, management stuff. I do some cool stuff with them. Um, I've met the CEO of Replit. He's super awesome. Replit's about a $3.5 billion company. They're growing super fast. In my opinion, they are the best option for all of these. There's, there are a couple others. There's one called lovable.dev that's pretty popular. Some people like it. I don't like it because for, for what I, you know, a beginner like me, right, I, I don't know how to code, and you have to do just a little bit more manual configuration with Lovable. You have to connect it to outside things Replit has everything internal, but you've got Lovable, you've got Bolt. Bolt is another one, right? Hey, type your idea, right? Build me a weather app, right? It's going to be a very similar, you know, type of functionality. There's another one that a lot of people like too. It's kind of newer called Base44. Again, it's simpler. I think this is simpler than Replit, but not as powerful. And I think Replit is more than anybody can learn it. Anybody can do it. It's so, so easy. Because the types of things I'm asking it to do, I told it React simply because I wanted it to be a specific type of app. You don't even have to do that. It's probably going to choose a React app for you anyway. But I basically just told it, here's what I want to do. It looks like we've got a working version now. These buttons work. Vacuum the living room, or let's clean the bathroom. I'm going to click Check. Great job getting that done. Now, one of the things that's weird is, it, um, none of the colors are showing right now. Okay. And so I could go through and I could fix this, you know, let's, let's do this. Let's do this really quick. I'll just show, we'll go, we'll go ahead just a tiny bit here and see if we can make this work. But I'm gonna say, okay, um, I'm just gonna give a screenshot of, I'm gonna say, um, I see no colors. Okay. So it's got some CSS, it has something that's, you know,, wrong with it that's not letting the colors actually show up. Okay, but yeah, this is amazing. It's set up with CRUD operations, responsive homepage, color-coded status. Pretty amazing. I'm going to just click on this and see if we've got a database. Nope, no database. So I wonder where it's actually storing that data. It's not going to work very well long term, but we'll worry about that stuff later. Um, all right, so it's trying to figure out— it's going to check this, but here's the cool thing, guys, right? This is my coding language. I see no colors, right? Anybody can do that, right? This is not complex. My 8-year-old can do this. My 12-year-old is doing this. My 11-year-old is doing this. My daughter can use this as well. They can build stuff that works completely because they just say, hey, I clicked on this button and it didn't work, or I clicked on this and it did work, and whatever. Anyway, so, um, it's going to keep working on this and it's going to keep doing it. Now here's, here's the one thing that I don't love about this. And this is where a lot of people get things wrong with trying to— when trying to build with AI. They just dive in and they don't think to plan first. And so let's say I want to make this an app that I sell in the App Store, right? And I want to make money with it, okay? Because I have lots of farming neighbors, right? Whatever. Even if you're just, you know, Carrie with her chickens at home, she needs to remember when was the last time I fed the chickens, when was the last time I cleaned the chicken coop out? When was the last time I switched the hay? When was the last time I did whatever, right? There's a lot of different things that have to be done there. When was the last time I switched out the fly trap? So, um, so maybe we can sell this to a lot of people. But the problem is right now, this— the AI only built this for functionality. It did not build it to drive revenue. Okay, so I want you to think like an entrepreneur now. An entrepreneur it's going to build value in a business that is actually, that has a goal of making money. Maybe that's not the goal always, but, you know, it probably is. It probably is supposed to be. And, and so because we haven't told it that, right, it hasn't, there's nowhere to sign up for anything. There's no premium features. There's nothing driving us towards a purchase. It's just, hey, it works. And that's cool. Maybe that is what you need sometimes. But if you actually want to say, hey, I have my first paying customer by Friday, well, we need to give it a plan that actually involves money. Okay. Still not seeing anything. I'm sure you have some conflicting CSS overriding something. Figure it out. CSS, guys, is like the part of the code that dictates colors and fonts and sizes. And kind of, you know, that kind of stuff. Okay, we're gonna do a little bit of coding school later, like a 5-minute coding school where you just will teach you a few, a few terms just so you know that. Okay, but for now, guys, we are going to kind of scrap this project and we're gonna go into ChatGPT. Time to, time to really get started. Okay, so let's dive into ChatGPT. Okay, so go ahead and just watch here for a minute, guys. We're gonna— I'm gonna show you just a few things that are important. ChatGPT is a, um, it's a very powerful tool. I use ChatGPT to do all kinds of things, right, from research to, uh, you know, writing documents, to, you know, to, you know, helping me come up with ideas, and, and just talking to it about whatever I'm working on at the moment, right? But its overarching functionality for me is it's my project manager. Okay. My— it's my project manager. It's where I'm going to come back and they have this really neat thing right here called projects, right? And I have all these different projects here and you can go down and see dozens and dozens and dozens of projects that I've created here. It's going to take probably 30 seconds to load because there's so many of them there. But anyway, I have tons of projects here, like my General AI Proficiency Institute project. Okay. This is, you know, where I started back on February 27th, I believe. Yeah, here it is. This was my very first chat ever. I started this project, told it what I was trying to do, and then everything I've done since then is in here. And so when I can go into— I can talk to it and it remembers what's going on, right? So let's talk about that for a minute, guys. There's different levels of kind of memory, right? Or context., right, that happens inside of an AI system. Not all of them are— not all AI systems are built the same. Um, some of them have the ability to look back at other chats and some don't and whatever. ChatGPT theoretically has the ability to look back at your chats, but, but not really, right? When you— if you try to go back and you look at how many chats I have, I have, I don't know, 1,000 chats, and there's no way— some of them are super long— there's no way that its context limit, that the amount of data it can understand at once, can include everything, right? And so there's a couple things that they do to kind of make it feel like that, but really, you know, that are kind of cheating a little bit. And sorry, this dang Zoom annotate button is in my way. I don't want this button here. Okay, so I can go into, but let's just do this. Let's say, Let's see here, what's something? Yeah. Let's learn about dev versus production databases. I've asked it a question about something, teach me about the difference between these different things and I was just actually trying to use this, but I could say, all right, make this easier to understand. Let's use instant, so we just Go fast. All right, make this easier to understand. Okay, so now it's going to help you. Let's strip it away and let's figure this out. Okay, so the cool thing about this is what I'm trying to demonstrate here is I didn't have to tell it what I was talking about, right? That might seem obvious, but it'll make more sense in a second here, right? I was able to go into an old conversation and I can just pick up where I left off. I don't have to be like, hey, okay, remember we were talking about databases and all this? No, I could just pick up right where I left off. Super cool. And so now I can learn about databases. Okay. A second level of memory here that's inside of ChatGPT, we're going to click on Settings and we're going to go to Personalization and we're going to go to Memories. Okay. The Memories, this is a bunch of different things that it remembers about me, right? And things I've talked about. Okay. For example, Let's see if we can go down here. Um, all right, I'm a, I'm a husband and father. I live on 5 acres in West Mountain. I'm fluent in Spanish and Chinese. I'm a class 5 kayaker. I remember the church. I did this, I did that. I use Replit, whatever. I'm a pianist. So it understands these things, and because it understands— let's, let's use the religion part as an example— because it understands I am a member of the Church of Jesus Christ of Latter-day Saints, if I ask ChatGPT a religious question, it's not going to give me an answer that maybe a Jew would be looking for. And if I was a Jew using ChatGPT, I wouldn't want an answer from the Church of Jesus Christ of Latter-day Saints. I want a Jewish answer, right? I want something— I want it to understand me, right? You know, one of the things— there's a great, a great guy that, that I've become really good friends with over the last few months. His name is Dean Graziosi, one of the best marketers, maybe the best marketer in the world, right? Works with Tony Robbins. He and him do, you know, Mastermind.com together and a bunch of other things. Anyway, really, really awesome guy. But he and Tony are connected at the hip. They do everything together. And, you know, he talks about how in several of his books— he's an author, right? And he talks in several of his books about how there's so much bad advice out there, right? Because, you know, if I ask Ethan for advice, right, it's not that Ethan's going to give me bad advice necessarily, but Ethan doesn't understand who I am, right? He's half my age. I don't speak Portuguese. I haven't lived in Brazil. I'm not currently going to BYU, right? We do have a similarity in that, yeah, we're both members of the same church. And so that's good. I could probably still get advice. I'm not saying that Ethan's going to give me bad advice, but the bottom line is, um, because Ethan doesn't understand everything about me, I have to just consider that. I have to take that into calculation. With AI, especially when you just start using one of these platforms, it's not going to understand you very well. And so you got to talk to it about yourself, help it understand who you are, right? And the more you do that, the more context you give it, the better it's going to guide you. Okay. So because I've been working with it for a long time, it understands me, has all these cool memories and can go back. And, and so then the third level of memory, guys, is inside of the project itself, right? And so if I want to do something, if I need a new contract, Right. Or I need something. Hey, you know, create, create, create new, a new privacy policy for me or my website, right? Boom. Right. It already knows everything we've done. It knows the one that it wrote, you know, months ago. I could obviously point it if I was going to do a better, a better whatever I'd say, you know, that's, that's at genaipi.org and I'd come down to the privacy policy here and I'd grab this and I'd show it the link. Here's the current one, right? And boom, now, now it understands everything about the project. Here's the current one. Let's make it better, right? Okay, cool, right? So it will then, it will then do that, right? And I don't know why it's better and, you know, or an update and we can figure out what's going on. And then I could ask it right after it's done writing this out for me, I could say, why is this better? Explain to me, like, do I really need to replace this or is my current one okay? Okay, right? And so, I can just ask it questions, but the point is it understands the context. So, you've got individual conversation context and memory. You've got project-level memory that groups some stuff together so it really understands what you're working on. So, if I'm a business owner, right? If I'm Jake, right? And I own Infinity Pools, I'm going to be saying, okay, I need a project that's for Infinity Pools. And I might even inside of Infinity Pool, you know what? I don't know. It depends on how you want to organize it, but I'd be like, okay, everything infinity pools related I need inside of a project, right? So that it just starts to understand and just build stuff out, right? So conversation, project, and then the platform as a whole has this memory about who you are. You can also, you can also, guys, when you create a new, when you do a new chat, you can, there's a, there's a fun button somewhere. They used to have a secret button. Maybe they've gotten rid of it right now, but you can come in here And you can go settings— sorry, I didn't mean to do that. I meant to say customize. You can also customize your GPT to say, okay, here's what I do, what kind of personality should it have, what traits should ChatGPT have, anything else you want to know or should know about you. And you can— so you can give it, you know, some more, some more context inside of this as well. I like to, to just keep it, keep that kind of blank. And then I give each project, and I give the whole system— it knows all my memories— but then I give each project kind of its own contexts and things like that. So anyway, memory and context matter. So what we're gonna do now is we're gonna create a new project. Okay? So I'm gonna create a new one. I'm gonna call this Chore Tracker. Okay? Create project. And then I'm gonna go back and I'm just gonna grab my prompt that I used to start this thing. Right? And I'm just gonna copy this into here. Paste, and I'm going to say, I want to build a chore tracker. This is what I do. All right, we're gonna do instant for a second. And all right, so I'm just going to give it the same prompt I gave Replit. Okay, so it's going to help map it out, and it's going to do a bunch of stuff, and I'm going to kind of ignore it for a second. It's just going to write some things. Then what I'm going to do, guys, is I'm going to say, all right, let's change this to auto. Now, it might take a little bit longer to think about a few things, but we're going to go into something, into a thing. This is where you guys are going to start to jump in here in just a second. Just watch this and then I'm going to give you guys about 15 minutes to start building this out. Okay. So we need to build that 5-page prompt, right? We need to build the really detailed instructions about the why, about our business model. What's the purpose of what we're doing here? How are we going to make all this happen? Okay. And the way that we do that, guys, it's really easy. I want to, I want to just go off on this for a second. It's really easy, guys, to, to let AI take over. AI is very action-oriented. It wants to do stuff for you. That's cool. That's a great employee. But you're the boss. Okay. Just like we did that, those bad prompts where the AI has to fill in the gaps. You want to make sure that you don't let it fill in the gaps, right? Jake's got his business. Ryan's got his business, right? Matt and Jon, they've got theirs. Nick's got theirs. You guys all are going to have your own things that you're doing here, and you want it to be Carrie's idea. You want it to be Cole's idea, right? You don't want it to be, hey, you know what? I want to build this thing. And then it says, oh, that's a great idea. Why don't you do this, this, and this? And you're like, that's cool. Let's do that. And then it's like, should I do this next? Yeah, build that. Oh, that's cool. Do that. And all of a sudden, all you've done is say, yes, great, I like it, let's do that. And boom, it's built something and it works and it's great, but it's not yours anymore. It's ChatGPT's or Grok's or Anthropic's or whatever model you happen to use to have that discussion with it. So we want to make sure that we understand that AI is trying to push good stuff at us, and that's great. But there's ways to say, hey, let's flip the script. Let's have it ask us questions, right? And then we can ask it to organize our thoughts, right? That's really the more, the more effective way to feel like, wow, I'm really building something here. Okay. Creation is fun, guys. Creation is how— is one of the primary gateways to joy. Talk about that a little bit more. It's— I'm writing a book on that. Okay, so here's what we can do. Let's see. I want you to interview me one question at a time about this business idea. I want you to ask me about the branding, the colors, the, the USP. That's the unique selling proposition. Why is this unique? The monetization strategy, the, you know, core functions, the mission and vision, and anything else, anything else, sorry, you think we need to know? No, anything else you need to know in order to help me create a great business, man. Okay, so I'm going to drop this into the prompt— or sorry, into the chat. You guys can use that in a second. But what's going to happen is this, and give me about 30 more seconds, then you guys are off to the races here. Okay, so basically you're going to say, here's my idea, maybe it's just I want to build a chore tracker, right? Maybe give it a little bit more instructions than that, maybe 3 or 4 sentences about what you're trying to build. Okay. And so, but what it's going to do is it's going to kick off this process where it starts interviewing me. All right, so here's the first one, branding. What do you want this to be called? You know, how do you want people to feel when they use it, right? This, you know, I live on a 5-acre farm and I want it to be primarily for people like me that have tons of things to keep track of. I want it to be professional but also fun. Let's think fun and user-friendly. Let's think like Apple on this one.

Okay.

So I didn't actually answer one of the questions, right? I didn't ask it what it should be called. I want to see if it catches that. It gives us a solid tone for branding. Yeah, I didn't do that. I didn't— it didn't go back and ask that, right? So I can also say, you know, so it asked me a question here, colors and visual identity, all this stuff, right? You didn't follow up. On the name. Do you have a good name before— or a good name idea before we dive into the next question?

Okay.

All right, so it's gonna give me a few ideas. Chorley, Task Haven, Farm Flow. All right, Chorley sounds steady. Homestead. We're gonna try Chorley. We're just gonna go— let's go with shortly. Perfect. Okay, so it does that. Then it's going to say, all right, now let's go back to colors and visual identity, right? So just remember where it is. It's really smart. It just comes right back to, you know, that part of the interview. Okay, so I want you guys to now take— it's 10:28, we'll give you till 10:43. Okay, you're all going to take 15 minutes. You're going to build an entire business plan. This is what I was talking to Abby about a little bit. We're going to do it in 15 minutes instead of— what did it— how long did they have you work on it, Abby?

Uh, well, I didn't because I left in the first week, but it was 8 weeks.

An 8-week business plan strategy. Yeah, we're doing it 15 minutes, guys. Okay, so that's all you get to build your strategy. Um, you know, get, get to work. Okay, so start a project. Start a project, give it the prompt, and, and then, and then give it this interview thing.

Okay, Carrie, are we just doing a project that we can think of off the top of our head right now?

Yeah, if you don't have— yeah, so as you saw in my email that I sent out last night, right, come prepared with some sort of business idea. If you don't have a business idea, I highly recommend building like a personal brand website for yourself, right? Show off who you are, what you're about. You know, maybe you're writing books, maybe you're working on a business, maybe you're a musician, maybe you're all those things and you want to just talk about— here's all the things. Hey, I want to just write, make a cool website about about Carrie, right? I think that's a really cool thing. I think every person on the planet should have a great personal brand website. So if you can't think of something, if you don't have a specific business idea, then do that. And by the way, even a personal brand website, guys, can lead to money. It can lead to speaking opportunities. It can lead to interview requests. It can lead to, um, jobs, right? It can lead to, you know, pushing— maybe you're, again, you're a musician, you're pushing that out. You're an author, you're pushing this out, you're creating something, everyone's creating something, right? It can give you a place where you can actually push that out. Okay, so go ahead and do that if you don't have a specific idea right now. If— or you could have a conversation with ChatGPT and be like, I want to come up with a business idea, what's a cool idea? Take 5 minutes to do that. But all right, 14 more minutes. You'll probably get through 10 or 15 questions in your interview. And I'm just going to keep my screen sharing and I'll just keep typing, but you guys do your own thing and feel free to jump in and ask questions as you have them. Another about 5 minutes here, guys. Hey guys, take another 2 minutes or so. All right guys, so, um, go ahead and finish up. Um, when I, when I really do this part of my business, right, um, depending on how complex what you're doing is, right, my— this Chorley app is, is pretty simple, right? It's not a very complex thing. But like when I was building GenAIPI, I had to figure out obviously the name and the branding, and I had to figure out the, um I had to figure out this, um, really complex series of, okay, we've got an IQ test, and so I have to create a bunch of questions for that, right? And I have to figure out what— how should the AI IQ test like really work so that it actually, you know, tells you whether or not you're good at AI, right? Is it really accurate? Then I had to create a bunch of training courses, and I had to create certification tests, which is kind of like the, the, the intro test but like on steroids. And then I had to build all the, you know, business model around it and everything. It was really complex. I spent probably 3 hours on this step alone. Okay. And so we're only giving 15 minutes. It's fine. You can dive deeper and deeper and deeper. But, you know, I got through, I don't know, maybe 7 or 8, maybe 10 questions here. And, and so now I'm, you know, it's saying, hey, do you want to do this? It might still be asking about random questions and it'll keep asking you questions for a while if you don't let it, if you don't stop it. Right. So what we're going to do now is we're going to have it build it into a business plan for us. And this is what we're going to do. We're going to say, please create a 2,000-word business plan based on all of these responses and this discussion. Use the— oops, use— sorry, guys, I'm like perfectionist or OCD on the, on the Getting everything spelled right. Use the canvas tool to put this together. Okay, so canvas is a new thing you guys are going to see here in a second. Some of you may have used it before. I just dropped that into the meeting chat so you can use this. But basically what it's going to do is it's going to create a document for us right here. So you see this, it's kind of like creating a Google Doc or a Word doc or whatever, right? You can actually go in and edit it. You can download the document as soon as it's done, but it creates this really nicely formatted, you know, full-on document, right? And so it's got my executive summary. My problem statement. Here's the solution. Here's the unique value proposition. Here's what we're working on. Here's some of the core features for V1. You know, some of them divided between this and pro, right? Because I've got a— I put a marketing— or sorry, a monetization plan together. Market analysis, monetization strategy, free plus pro. Here's my marketing strategy. So guys, it's so important to understand for the AI or for your developer, even if you're just giving this to a developer, right? You're not even using AI. If you give them this type of a plan for them to go create something, they're going to create something way closer to what you're trying to do, right? Most of the time people are just like, go build this feature and, you know, figure it out, right? But if you actually sit down and you think through what does this mean in the, you know, the grand scheme of things, then, you know, you can get much, much better results. Okay. So I can go in, I can edit it. If I don't like something here, I can go on. Like, for example, I don't need the team. I don't need the team in this. That's not going to help me build something awesome here. I don't know why it's not letting me edit this. I should be able to click here. There we go. There we go. So we're just going to get rid of that. I've got the conclusion, gross margins, financial projections. Even keeping financial projections in the plan is pretty cool because this is what I'm like, okay, here's how I can probably do this. I can get to 5,000 users, convert this, I can keep growing. It understands that and it's going to build solutions and things and little features, little decisions that it makes as it's building to be like, okay, we need to help people convert from free to pro. That's part of the business plan. That's part of the monetization strategy. That's part of the financial projections. In order to do that, we're going to have to do things that convert people here. And so We're going to put a little button here that makes that happen. Okay, so go ahead and put that prompt in, guys. Create your business plan and then it's done, right? Once you get a business plan, that's great. Now we're going to do something. We're going to do something else with it. Okay, so let's just start with this. Also, I don't need it to tell me my word count in the business plan. That's a useless piece of information. So we'll just delete that anyway. Doesn't have to be perfect. Doesn't have to be perfect, right? So that takes you guys, you know, about 60 seconds to do that. Okay, then you can copy this. You can copy the whole business plan, click copy. Then what we're going to do, guys, is we're going to go back to Groq. Okay, now Groq is really interesting. It's kind of unique in the, in the LLM space because it has a few extra capabilities. One of the main capabilities that it has as compared to something like Anthropic or or ChatGPT is it can search all of X. It can search tweets. It can search what people are talking about, right? And so this can give you actual real-time market sentiment and analysis on what you're doing without having to go, you know, do a bunch of research. Now, obviously, you could— you can do research as well, but when you're trying to build a business in a weekend and you're moving fast, Grok can do some amazing stuff. Okay, so we're gonna do is we're gonna do this. We're gonna say, um, I want you to evaluate this business plan, not so much for the business plan and structure— whoops— itself, but more for the idea and how you think the market will accept it. Where are the strengths? What are the weaknesses? Be sure to do extensive research from articles and from user comments on X related to these topics. Okay, then you're just going to go And return, return. You go Shift, return, return. If you hold Shift down, then it won't actually send this. Then we're going to paste the business plan. Okay, don't worry if it has all these, you know, formatting things. It's okay. And then we're going to make sure— I'm going to have it think hard. I'm going to click on Expert. Okay, and then I'm going to say go. Okay, so I'll give you guys this prompt here.

Copy. Chat, do I have to be logged into X to be able to access the articles and comments and things?

Nope. Nope. You can use Groq. You can use SuperGroq. You can use— yeah, you can just use that. I just happen to be using the x.com interface. But yeah. Okay. No problem there. Jon, is X better at this particular— or is Groq better at this particular step? Or are we just trying out both platforms? No, Groq is better for this step. It can do research. It can do research because ChatGPT can't look at X, for example, right? And I really love the real rawness of X, right? What are people saying about chores and chore apps and farms and whatever? I don't know, like that. I want, I want real-time, what are people saying out there in the world about this topic? Okay. And so that's why we're going to choose X for this. Now you don't have to. You can use ChatGPT to do this. You can use Perplexity. You can use all kinds of different platforms out there to do this. But there's another thing that we're doing here, guys, and I hope you're already working on it. Don't just sit and, you know, listen to me. Just go ahead and do this. Thing so it can be thinking. Um, but not all LLMs are created equal, okay? What I mean by that is, you know, you can go between Chrome and Safari and Microsoft Edge and you can find the same, you know, Google answers, right? You go to msn.com or cnn.com or foxnews.com, you're gonna see the same website. On Chrome, Safari, or Edge. LLMs are not like that. Going and asking Grok about something is different than asking ChatGPT about something. And so part of what we're trying to do here is not only the research part, we're trying to get another model, a different LLM, to opine on our plan. We want it to think, we want it to think through, hey, what is this doing? Is this good, right? And Grok 4 is brilliant, right? It is a super, super powerful model. And so if you're— Jon— oh, go ahead. Sorry to interrupt. I'm, I'm a dummy, so I don't know what an LLM is. Yeah, sorry. An LLM is a large language model. Yeah, it's what— it's what powers the, you know, the system, right? And so it's just a model, right? It's a, it's a model that's been trained on a certain number of parameters. ChatGPT is something like 1.76 trillion parameters, something crazy, some crazy number, right? And so they keep— each model has more and more parameters that it's trained on, but it's basically how much text has it been trained on, how many patterns has it been trained on. But that's what the LLM is. It's the model that, you know, it's the flux capacitor of AI. If the flux capacitor enables time travel, the LLM enables ChatGPT to work, right? We're engaging with— there's a prompt sitting on top of a model, a database, basically the brain. It's the brain behind it all. Okay, so going to different brains that think differently, right? If I take my business idea and I run it through my head and then I take it to you, Richard, you're going to have different ideas because you have different experience, you have different training, you have different lens to look through it, right? So That's another reason why we're going to go to a couple different models and kind of bounce back and forth, get feedback from this on the same topic from different people, right? Different, different things, right? So I can now look at my, you know, thing. Here's the overall assessment. This is pretty good, right? Here's some strengths. Here's some weaknesses. You know, chore tracking is pretty crowded with a few things out there. Here's, you know, you can probably do this better, whatever. Anyway. So, so what I'm going to do after— now obviously if I'm building a business, I'm going to take time, I'm going to read through this, I'm going to really think, and okay, how can I, how can I really make this happen? Okay, so bottom line is I am going to click copy on this. So it searched 57 web pages and found 11 posts to support what it's talking about here, which is pretty cool. Um, there's got to be some, there's got to be some big things that make this work. So I'm going to click copy, I'm going to copy the feedback that I got from this from Grok, and I'm going to go back to ChatGPT and I'm going to say, please create a new canvas called Improved Business Plan. Use the feedback below to improve the plan. And I'm going to paste it and I'm just going to go. Now, if I was really building my business, I might take time to read through and consider how much, you know, maybe I agree with that, maybe I don't agree with that, whatever, right? But it's going to improve improve the business plan, and, and it's going to create something better here. Oh, you know what, this is not what I want it to do here. I don't like what it's doing here.

You can create a canvas yourself.

Yeah, yeah, you just push the button, open canvas. Yeah. Okay. This is— this is not what I asked. I want you to create a new business plan based on all the new data combined. With everything we already discussed, focusing, keeping my ideas front and center. Create a new canvas with— and call it Improved Chorley Business Plan. Okay, so guys, I'll give you, um, why is it asking to fill a code? Is that— Jon, do you mind putting that prompt into the chat? I am.

Yeah, I was gonna ask the same thing.

You got it, guys. Yeah, that's what I was going to do, but I did not like— yeah, and I'm gonna, I'm gonna actually improve the prompt before I send it here. Um, we'll get it for you. Matt, are you still here or did you have to roll? Yeah, Matt, I just saw your thing. I know you have to sign in. Yeah, it's all being recorded. You'll be able to come back and do that. You can catch up later. You can jump back in in an hour if you want, whatever, you know. So yeah, go, go take care of family. You're good. Awesome. Thank you. Use feedback below to improve the plan. And then you paste your plan, but we're going to say, please create an entirely new business plan, keeping in mind my priorities discussed, but improving it with this feedback. Okay, that should be a pretty good prompt for you guys to use. And then just paste the feedback you get from Groq. But the idea really is to say, okay, can we, can we just kind of refine this idea a little bit? And this is a very, very fast way to go through, get some market feedback, see what's going on. Who are my competitors? What's going on here? How do I really stand out? Okay, so guys, go ahead and take about 2 minutes to catch up to this spot here where I am, and then we're going to do— we're going to create one more document here in a second, and then we'll dive into Replit. All right, guys, so let's, let's keep going here. So you should have a business plan now that's improved. It's been, it's been iterated upon. And now we're going to— so now we have a business plan. That's great. We're going to use this document in a minute. Now we're going to create another document. We're going to call this the features list. Okay, so I want you to create a, um, a new canvas called Features List. Okay, this should be a simple document of all front-end and back-end features that'll be needed as we create this app. Do not specify the technologies that will be used for each, just list out the features. Okay, so we'll do that. I'll give you guys that as well, so go ahead and copy and paste that. So what that's going to do is now, hey, we know we need user authentication and we need a dashboard and we need chore management, right? So for me, I need all these different things, and now it's kind of, kind of, kind of give me a list of things. And we're going to give this list to Replit so that it really keeps track. Um, hey Jon, process.

Yeah, go ahead. Um, I'm curious why you, uh, choose to use ChatGPT for all these rather than Groq, because I noticed the other day when I first was trying Groq it has project folders. It also can reference, um, chats if you tell them the name of the thread, like separate from the chat that you're currently in. And, um, I thought that I saw that it does documents also, like a canvas.

Yeah, the reality is you can use whatever tool you want. Lots of these tools, like you can also use Anthropic Claude, has great project folders and things like that as well, right? I just happen to use ChatGPT. For me, ChatGPT is the better interface because of a lot of different things that it does behind the scenes. Grok is incredibly powerful, but they've— the projects feature didn't exist until like a couple months ago. And so I've been doing it for a long time. And so I'm just used to using ChatGPT. It's really all it is. But use what you're comfortable with, right?

If you're like, you know what I'm saying? No, no, that's the situation I'm in also because I've been using it for my business since last February, so a year and a half at this point. I'm like, I'm just curious, like, if I am like good or should I—

No, I think you're good. And I think, yeah, I mean, like, they're, they're all of these models are competing with each other, right? They're like, oh, we did this, now we have this feature, now we have this feature, you know? And so you're going to find a lot of overlap between them. But yeah, it's a great question. It's a great question. There's no crazy reason why I do it. It's just that I've been using this the longest and this is where my, this is where my stuff is, right? So, um, okay. So now guys, we have 2 documents. Now we're going to do one more thing here. Last thing here before we go into Replit, we're going to say please write a 3-sentence description of what we are building. Very simple.

Okay.

All right. Um, yeah, Matt, you can be using 4.0 if you want. Um, it's totally fine. Like GPT-5, like, is absolutely better. A lot of people use 4.0 more. I don't know, 4.0 is going to be a little bit faster. And so with what you're doing right here, you're going to be just fine. If you were doing really complex PhD-level math, I might say, yeah, let's switch to 5. But for creating a business plan, 4.0 is going to be more than capable. Okay. All right, guys. So now we have 3 assets. We have a 3-sentence description of what we're building. We have a business plan and we have a feature list. So now we're going to give those assets to Replit. So go ahead and pull up Replit. Now we're going to copy this 3-sentence description here and I'm going to— wherever the heck my Replit is. Here it is. Okay, so I'm going to scrap my old project. It couldn't even figure out how to get the colors right. So whatever. We're going to do this right here. We're going to go paste. So don't just click start chat or click enter after we paste the 3-sentence description. We're going to do a couple more things. So I'm going to paste this, then I'm going to go back into here and I'm going to grab my feature list. I'm going to click copy. I'm going to come in here. I'm going to paste it. When I paste a really long piece of text, it's going to look like an attachment. It's just pasted text right here. Okay, then I'm going to go back. I'm going to grab my improved business plan, the latest one that I created. Okay, I'm going to copy that. And I'm going to paste it. So my prompt should look like this. I'm going to add one more little thing. Please be sure— and I'll be sure to reference the attachments as well for more detailed plans. Okay, I'll give you this little prompt as well if you want it, but this is what your plan should look like. Okay, so I'll leave this just sitting here. If you get to this point, guys, you've got your 3 sentences, you've got this little 'be sure to reference the attachments,' you've pasted your, your feature list, and you've pasted your business plan, go ahead and click Start Chat.

Done. I, I just, I had some errors with ChatGPT, so I'm actually quite a few steps behind. But can I just verify, after you do the 2,000-word business plan, you take that to Groq and you ask it to evaluate, and then you go back to ChatGPT?

That was the flow, right?

Yep.

Yeah, you just copy the feedback it gives you and then paste that back over. Yeah.

Thanks.

Yeah. I'm showing my improved business plan, but I'm not showing the features list. Did I miss a step? Did you, did you ask it for the features list? You know, I think I must have missed it. Yeah. Yeah. So right there in the chat, you'll see I want you to create a new canvas called features list. You'll see that in the group chat, just maybe 3, 3 or 4 messages back there. Yeah, just ask it to do that. It'll, it'll whip it up in 30 seconds. Guys, I just added one more little thing here. Just trust me on this one. Um, replace the 'be sure to reference the attachments as well' with the new thing I just pasted into the chat. Um, so just something related to if you have logins for your system, which you probably will. If you don't have logins, you do not need to do this. Like, if you're making, you know, your own personal brand website, Carrie, maybe people aren't going to come to your website and create a user ID and whatever. But if you guys have a user ID system that's in mind and that's going to be important for you, like mine definitely is, you're going to log in, you're going to have your farm, you're going to have your chores, you're going to have your house, whatever. So there's going to be a login system.

If you have a login system— actually decided to make a competing app of yours.

Oh, nice. Cool, that's awesome. Totally fine. Let's do it. Let's go, let's go, Carrie.

I'll take you on.

We'll see. We'll see who wins.

What's yours called?

Uh, this is early 2.0. What?

Sorry, early 2.0.

Yeah, mine's a Quack Track. Quack Track. I love it, I love it.

Very cool. It's about getting your ducks in a row, and all your tasks are ducks.

Yeah, see, you're gonna win already. You're beating me on branding.

Right. That's awesome. Uh, what was your background again? You, you, uh, you've got some good, uh, I don't know how to say that, but like visualized ideas or metaphorical ideas. Um, my background is recreational therapy, and so I do a lot of experiential education. I basically use recreation as a tool for therapy and help people learn coping skills and communication skills and all that. Using experiences. Very cool. Yeah, and I have ADHD and I work very much in visual. I need to see things and interact with things, otherwise it's too hard for me to focus.

All right, guys, so, so you should have your 3 sentences, your past business plan, your feature list, and then this little thing here, right? Okay, cool. Cole, thank you. So, so what's happening here, we're going to click start chat. Okay, go ahead and click start chat and it will start your project. This is going to take a few minutes. We're going to do our— normally take a 15-minute break. We're going to do a shortened break and it's going to be like 11 or 12 minutes. So it's 11:08 right now. We'll go to 11:20. Everybody stand up, go outside, go to the bathroom, get a drink, grab a snack, whatever it is. Take a minute, let your brain kind of reset. We'll come back, we'll have a good conversation while we're building stuff out. But we're, we're doing great on, on time, guys. This is awesome. A quick question, this may be dumb, but I run into it a lot. Um, before I hit the start chat button, if it— do I need to put like a space between paragraphs, or do you just plug it all in when you're— um, like, you don't put in my— you don't have to. Yeah, I mean, I, I do because just for me visually I want to— I want to do that because when I press enter, it does start chat. Yeah, yeah, yeah. In order to push enter, you have to hold shift down. I said that earlier. Yeah, you go shift enter or shift return, and that won't actually send the chat. Okay, that's what I needed. Thanks. Yep. And then guys, when you see— if you can see on my screen right here, you also want to click approve plan. Okay, so it'll kind of start for a second, and then it'll just approve plan.

Okay.

All right, so I'm gonna go ahead and turn off camera, stuff like that, guys. Um, I'll leave my screen on so you can just kind of watch it going, but, um, we'll see you guys 11 minutes now.

Abby, whereabouts are you from? Uh, I'm in the West Coast. Uh, Vancouver would be the closest big city. Cool. I've got a lot of family that lives up in that area. Yeah. Um, I'm actually on Vancouver Island, if you're more familiar with the area. Yeah, in, in Comox Valley.

Well, let's see, where are they? Abby, I was born in Lethbridge, Alberta, Canada.

Oh, no way!

Yeah, it's amazing. Yeah, Mom is from Cardston, so, uh, Canadian born.

Yeah. What is—

what's your mom's maiden name?

Williams.

Sorry, Williams. Mike Williams. Williams. Yeah, yeah. My mom's maiden name is Olson. She's also from Cardston. I'm sure. And my dad's from Saskatchewan.

They would know each other. That's, that's how small cards—

I'm sure they would. Yeah, I'm sure they would. Um, so are you from Vancouver or are you living there now? I'm in Comox Valley, a small town called Cumberland. I was in Vancouver a lot of my life at different times, but also other places on the Sunshine Coast like Gibsons and Powell River and stuff like that. I've also lived in Manitoba, but from, from BC. I consider myself West Coast born and raised, just not specific to one town, so we moved around a lot. Mm-hmm. Amazing. It's a very big place, and, um, BC is just absolutely gorgeous. I would love to live there someday. I'm quite lucky where we live right now because we're basically in the woods. It's great. Oh, it's beautiful. So it's starting to get cold up there. Yeah, it's kind of gray today. We've also just had some nearby forest fires coming in, so it's hard to know if it's gray because of the weather or because of the smoke. It's hard to know if it's gray. I love how you said that. Yeah. All right. So I'm also a few steps behind, so I'm just catching up. What were the documents that were pasted in? Was it the, the last two that we made? It was the, um, I'm just doing that step as well. It's the, uh, gosh, what What is it called? The feature list and the improved business plan and the improved business plan.

Okay, thanks. All right, guys, welcome back. Hey, hey, Jon, can I jump in real quick? And I failed to name my business. So is it possible to type in the name and add it, how do I get it to update? Yeah, well, I just need to re-sync all that. No, no, no, you'll be able to, you'll be able to change all kinds of stuff here, right? This is okay. This is the very early rough draft, right? We'll get to a final draft. Um, you can change things ongoing, you know, tons and tons of times. It's just as simple as saying change the name to this, right? Um, all right guys, so, um, all right, we're— before we, before we look at our projects, right? I mean, I'm sure you guys are all looking at them, right? Anybody have stuff that looks cool where you're like, wow, this is awesome? Yeah, so it's pretty, it's pretty, it's pretty fun to see this, your ideas come to life, right? Um, but what we're gonna do right now, guys, we're gonna take a second, we're gonna watch a quick 2-minute video Um, this is right here. See if I get this right. Okay, this isn't perfect. I had a nice cut-up version of this, but it's not. And so I'm sorry, I want it to be— can I view? I zoom in. Zoom in and zoom in. All right, cool. So let me see if I can share the screen now and we'll watch this little video. So this is of a guy named Eric Schmidt, um, chairman of Google, former chairman of Google. Um, worth— this guy's worth like $20 billion, super smart, very connected. Um, and he— I want to say that he said this in mid to late April, like April 16th or something like that of this year. Um, so, you know, what he's saying is already outdated a little bit, but still very applicable. So we're going to watch it and we'll just have a little discussion here. Okay.

We believe as an industry that in the next 1 year, the vast majority of programmers will be replaced by AI programmers.

We also believe that within 1 year, you will have graduate-level mathematicians that are at the tippy-top of graduate math programs. So that's 1 year.

Okay.

What happens in 2 years? Well, I've just told you about reasoning, and I 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, and Anthropic and so forth, is that they're now somewhere around 10 or 20% of the code that they're developing in their research programs is being generated by the computer.

That's called recursive self-improvement is the technical term. So what happens when this thing starts to scale?

Well, a lot. One way to say this is that within 3 to 5 years, we'll have what is called general intelligence. AGI, which can be defined as a system that is as smart as the smartest mathematician, physicist, you know, artist, writer, thinker, politician. I call this, by the way, the San Francisco Consensus, because everyone who believes this is in San Francisco. It may be the water. What happens when every single one of us has the equivalent of the smartest human on every problem in our pocket? But the reason I want to make the point here is that in the next year or two, this foundation is being locked in and it's not— we're not going to stop it. It's much more interesting after that because remember, the computers are now doing self-improvement. They're learning how to plan and they don't have to listen to us anymore. We call that superintelligence or ASI, artificial superintelligence. And this is the theory that there will be computers that are smarter than the sum of humans.

The San Francisco consensus is this occurs within 6 years, just based on scaling.

This path is not understood in our society. There's no language for what happens with the arrival of this. That's why it's underhyped. People do not understand what happens when you have intelligence at this level, which is largely free. All right, guys, what are your thoughts?

I have a lot of thoughts. Do share. Well, I think I— so I'm a, a mother to young kids. You know, I have a 7 and 6-year-old, and, um, it's just interesting because right now we are living in a time where their school experience is going to be vastly different from anything that I have experienced or even imagined, you know. And so I believe and feel strongly that learning about AI and doing, you know, this course and and also exploring and diving in and not being timid about my exploration of AI is extremely important because who's going to teach our little ones this if we don't teach our little ones this? And so, I mean, yeah, just from like a family standpoint, I think it's really important to be diving in and jumping in and not— and, and just embracing what is, because it's not going away, it's not changing. It's now recreating the foundation of our society, and we have to go along with it.

Yeah, you know, one of the things that worries me when I look at the education system is that our kids don't go through an education system that's very different than what we did.

It's— yeah, that is so mind-blowing, especially because I'm working in VR in that sector, in the education sector. And it's like, how are we still like writing on a whiteboard with the same textbooks that have maybe been like revisited once or twice since I went to school over 20 years ago? It's scary. How is that the reality?

Yeah, Carrie and Jon, you just, just didn't know, like the company that just, I was just wrapped up with, which just acquired, was kind of the leading innovator when it comes to educator preparation in the space. They're working with over 300 educator prep programs. I ran all the partnerships. We scaled and supported hundreds of thousands of educators pre-licensure. And just the notion of being able to use simple tools to be able to actually scale and provide kind of innovative ways to learn was just difficult for these programs. It was just like, we can't go that fast. We can't understand how things are changing. We can't prepare even teachers to teach the way they should now. We're, we're too steeped in what we used to do. It's wild to think how teacher educators are not equipped to transform the classroom either.

Absolutely. And to give you an example, we just went through an adoption process for writing a new textbook for schools in Florida to incorporate virtual reality into their schools. And the whole adoption process is crazy itself. But the— when we finished writing the textbook and we finished writing our curriculum and we sent it off, they came back and said there's no possible way that we can even consider this because of the virtual reality component to it. There has never been any type of curriculum written and adopted by a national, by anyone that incorporates VR. So they had to rewrite their entire way that they even look at text, like textbooks and things like that. That just happened just a couple months ago. In this latest round of adoptions, which started in March.

Yeah, guys, I hear, I hear teachers, or I hear stories of teachers, you know, students say, you know, that their teacher says if you use AI for this, you fail. Yeah. And that's the dumbest thing that teacher could do for that student, right? We are entering into a new world, guys, that is so dramatically different today than it was even 2 years ago, guys. We just built— we all, all of us just built a new piece of software in like an hour, okay? In order for me to get to where my Replit project is sitting right now, right, a couple years ago, a few years ago in 2020, right, would have taken me $50,000 would have taken me 6 months to find a team that can do it and put it all together and all this stuff. And I just did it for like $3 and 20 minutes. You got to do the math on that, right? 6 months to 20 minutes, we're talking 99.99% reduction. You know, $50,000 to $3, it's 99.999. We're improving things at scales by several 2 or 3 or 4 orders of magnitude in some industries. And nothing has ever hit the world like that before. The internet was big. Mobile was big. Electricity was big. Factories were big. The Industrial Revolution was big. The wheel was big. Nothing comes close to what AI is doing and how it is changing the world right now. Every single industry, every single job will be flipped on its head. Right? Even ones where you don't think it's going to flip you on the head, like plumbing, for example, right? Plumbing is probably one of the safer jobs right now, right? It's hard to get to teach a robot how to go and change a toilet, right? It'll be able to do it eventually, but we're a ways out right now. In the future, though, it will not be. But if you're a programmer, if you're a paralegal, if you're in marketing, if you're in sales, if you're in creative things, any creative industry, Yeah, you are now the low end of the capabilities, right? And compared to what it can do, the amount it can produce and the quality at which it can produce is just absolutely amazing. Now, does that mean that we're all doomed? No, I don't think so. I really don't believe so. I believe, and I am doing it right now, I am making tons of money, guys, with AI, right? Helping and working with people and building things. And I'm not just training, right? I do other things with AI and I'm making money in ways that I wanted to do. I can put more music out with my, you know, I write music and, you know, I can do that better. I can write my books faster. I can, not that I have AI write my books, but just the processes around it, I can do it faster. I can whip up a new idea faster. I, you know, when I was, when I first discovered Replit, I had gotten a quote from a Ukrainian dev shop that I'd worked with in the past to build this cool travel social network app thing that I was— I had this idea for, right? And they gave me a quote, it was $106,000. And I was like, cool, let's go, right? This was for an iOS version, Android version, and a web version. Like, this is going to be really awesome, this is going to be so cool. It's kind of just the MVP, very basic stuff. Then I saw this post on LinkedIn of somebody that I, you know, knew, and they said, I just saw somebody build in like 5 minutes. And I was like, what does that mean, right? And I clicked on this Replit thing and I opened it up and I just was like, oh my gosh, this is insane. And I ended up— I took the plan that the Ukrainian dev shop had, you know, put in front of me and I just handed it to Replit and said, build this. And, you know, I don't know, 20 hours later I had it working, right? As a non-developer, that was the first time ever I had really built software where I was just like, Oh my goodness, this is crazy, guys. I had to go raise $13 million for Seek, my company that started in 2016, because I couldn't even prove if my idea was viable. I could— I didn't even know if I was going to have any customers until I had spent $500,000 of other people's money. And all of a sudden I'm stuck. Not that it was bad that I was disappointed with that, but now I'm like, I'm committed, right? Like, I got to go through with this. This now. I can't just be like, okay, well, it works, that's cool, but I don't really want to do this, I want to go do something else. You know, no, I've taken people's money, now I have to get them a return. I got to figure out how to go take this through all the way to the end. And by the way, that's crazy hard, and it's going to take 5 to 10 years to do it. Like, you have to be very, very committed to that. What AI has done for entrepreneurship, and then I hope you guys grasp just after the first couple hours, first few hours today, and what we'll get through tomorrow is that the risk, the risk is usually twofold: time and money, right? It takes a long time to build up a startup idea, and it takes a lot of money to do it, right? Even if it's just your own money, right? Your own time, your own resources. You could be spending that time doing something else, right? It's opportunity cost. But time and money are the big things. And now you can prove an idea by the end of today all of you could have something working that someone could pay for, right? Carrie, you could have your ducks in a row, quack quack, whatever it's called, working to a degree where you could then go get customers paying you by later this week or maybe by tomorrow, right? Or you could talk to the customers. They might say, you know what, we don't like this, we use this and look at this app, it's way better than what you're trying to do. You're like, okay, cool. Right? Or you're like, you know what, mine's actually different and it's better. I'm going to go after this. Or I don't care that there's, you know, another competitor out there. That means there's a market. I'm gonna go, I'm gonna go make this happen. This is awesome, right? Bottom line is you all have— you've all but eliminated the risk. So when I built GenAIPI, I put together a, um— I wanna, I wanna share this. I think this is worth pulling, pulling up the— pulling up this piece here. So let's go to here. Come on, Google Drive, wake up. There we go. Okay, so let's look at this and let's look at this. So if I had built, um, GenAI API in 2020, all right, it would have taken me this much money and this much time, right? I went through and I did a big analysis of all the things that I had to create, all the different pieces of software, front end, back end, deployment, cloud infrastructure, taxes, paying salespeople, paying salaries, all this different stuff. It would've taken me 18 months and $3.2 million. It just would've, right? You can go down, break it down category by category. I'm only paying $100,000 to my developers, QA, you know, half-time on a project manager. I mean, heck, I only have 0.1, you know, 10% time from one executive. It would be more expensive than this. I think this is on the low end, okay, of the types of things you would need to do in order to go out and build this out. And it would have taken 18 months and taken $3.2 million. I did it in 3 days and it took me less than $300. Okay, I mean, you just— you can't— that's the difference in time. It thinks that's a phone number. It's not, but 0.00009375, right? Like, we, we're talking about just a dramatic, dramatic difference in time. 99.99065% cheaper, right? And I did in 3 days instead of 18 months, right? So 365 times 1.5, 54, so 547.53 divided by 547. 5. Okay, and we'll do the same thing here. Whoops, there, minus 1. Okay, 99.45% shorter time. Okay, building a business today is an entirely different equation than what it was just a couple years ago. You can build something and have customers almost instantly without anything. You have no excuses anymore. You just got— now, the one thing where a lot of people will fail, okay, um, and people have failed, take this course and they want to go start a business, is they don't follow through. They don't finish it. They just stop at this kind of rough draft. They don't actually finish. They don't go launch it. And because they don't launch it, because they don't get it to a point where like, hey, this is awesome, they tell no one about it. And a very easy way to get no sales is to not tell anybody about it. Right? But you have to not be afraid. We'll talk about marketing and sales tomorrow. You gotta not be afraid. You gotta post about it on social. You gotta post about it on the— you gotta get on the news. You gotta have articles written about you. You gotta do what— whatever, right? There's so many ways that we could talk about tomorrow about how to get this out in front of people where people are paying money for you or to you right now, right away, to turn this into a real business in a weekend, right? You really can do it. I could turn Chorley if I decided to. I guarantee you I could have my first users and even paid users by next week. No question. I 100% believe I could do that. Right? And Carrie, you can do the same thing with your business. It's the same one I'm building, right? You're just gonna build it better and it looks cooler, but you can—

I'm gonna take all your customers, Jon Cheney.

Take 'em, take 'em, run with it. I'll give you all my code. I don't have time to build Chorley. I might use it internally. I don't know. Or maybe I'll check yours out and see if I can use, you know, I'll pay you $6 a month or whatever. It is, right? So, but, but again, guys, I just, I, I want to open it. I've been talking for a little bit here, and I, I shouldn't, um, be talking so much. Any more comments on this idea of artificial general intelligence, artificial superintelligence, and how it is going to affect our society? Anything else that you guys want to share or jump in and that's on your mind? I recently listened to, uh, to a, a podcast about, uh, NVIDIA. And the, the one of the things I took note of is if you are training one single speech recognition model, it takes more than— more math computations than all the grains of sand in the world. That's how complex it is. And how quickly it's doing it.

Yeah.

It's crazy. All the grains of the sand in the world. That's how many computations it is doing. It's, it's really quite difficult, actually, guys, I think for our human brains to comprehend that. Like, it's really tough. Jon, did you want to say something? I saw you pop up for a second. I'm not sure if that was an accident. Oh yeah, I was just going to say I think the opportunities are incredible if we're on kind of the front end of this wave. Uh, you know, it's changing. I think every industry will be affected in significant ways. You know, Jon, you're a lawyer, right? Yeah. I will only work with lawyers now that are okay working with me and my AI, right? And so that means that I'm showing up with a rough draft of most documents or what I think is close to a final draft. And then I'll give it to you and I'll say, hey, I'm going to pay you, but now I'm going to pay you, you know, $2,000 instead of $20,000 for the same document potentially. Right. And that's a whole different world.

Right.

And, you know, but if you can adapt to that and say, okay, I'm going to— I'm not going to shun this. I want to get my billing hours in. I'm going to sit here and write this manually and use my templates and all this stuff. You know, you can do that. You're just going to lose to other people that don't, right? And so you just have to say, okay, what is my competition doing, right? Like, what are they going to do to beat me? And then make sure that you stay ahead. You stay ahead. And so I'm glad that you're here, Jon. I think there's some really, really cool things that can be done in the legal space to really stand out. Guys, most money in the world, and Jon's a perfect example of this. Is made because there's a knowledge gap, right? I'm going to go to Jon because he knows the law better than I do, right? He knows how to navigate this. There's also some licensing issues, right? I understand that with doctors and lawyers and some things. There's some things that our society has set up that kind of put some of that in place. So, even if I do, you know, write my own document on something like maybe I do a response to a lawsuit, I can't file that. I have to go hire a lawyer to go and respond. And, you know, there's, there's order, right? So there's that. But, but the, but the reality is my wife is an amazing gardener, like top 0.1% in the world. She's amazing. Because of that, she can charge money to teach people how to garden, right? You go to a doctor because he knows how to take your appendix out when it bursts. You go to, you know, whatever, every, whatever you can think of, right? You're going somewhere because they can do that thing better than They have the systems in place. They can do it better than you. And so because you guys are all now applying AI to whatever it is you're doing, you're going to have this knowledge gap. And I can't tell you how far ahead you already are just sitting here for a few hours, how far ahead you are of most of the world. They just— most people have no clue what's possible. And so when you show up and you can whip up a Flappy Bird game in 2 minutes while they're talking to you, telling you the idea,, or you can whip up a new business, or this chore app that you're making, Carrie, you could launch this and people would be like, man, how long have you been working on this? And you can say, I've been working on it for like 6 hours. Or you can be like, yeah, you know, I've been thinking about it for a while, whatever. And people would believe it took you 6 months to build this and that you're now an amazing developer, and, or you paid someone to do it. And like, man, you've worked a lot on this. You can make something that just looks incredible in so little time. Um, All right, guys, let's continue on.

I've got a quick question.

Um, it's a little off topic, but at this, at this stage, like where I'm at, I've got a lot of ideas racing around my mind of like businesses that I could start up and ideas that I do have. What is the best way, the fastest way for you to kind of just sift through those ideas and kind of find out what the winners are just early on this process? Because like, it's, it's nice that I can like build a full-on website, build a full-on app. Or a software this fast. Um, but what— how would you recommend really deciding which one to go with? So that's a great question, and I will take a minute to answer it. Um, I think it's fantastic, actually. Guys, startups suck. They're so hard. Even when you can whip up the MVP really fast, they are just hard. Right? And so, you have to be able to push through the hard times. And so, the number one thing that I'm looking for in a startup idea for me is, am I passionate about this? Do I really care about this? Am I okay to wake up every day thinking about this thing? Right? If I was going to start a new business around kayaking, I'd be pumped every day. And I am working on something with kayaking. I'm working on this $2.5 billion resort out in Tennessee. It's a kayaking thing. But anyway, talk about that later. But I am excited about kayaking. I'm excited about AI. When I wake up and I'm like, okay, oh well, I wonder what's happening with AI. Like, I want to get on my social feed and see if there's anything new, see what's happening with that. Oh hey, there's a new update with this. That's cool. I'm excited about it, right? So number one, guys, it's got to be something you're interested in, right? Because if you don't, you're gonna run into the problems where you make no money for 3 months and everything's, you know, nothing's working right. And, you know, you go, you're, you're heading to BYU, you're going to be dating, you're going to be going to school, you're going to be whatever, and maybe you're going to get married, you know, who knows? Maybe it'll take you a few years, whatever. Maybe it'll take you 3 months. Point is, you're going to be distracted. And so, you got to have something that's going to pull you back in, right? You got to be interested in it. So, interest is number 1 for me. Number 2, I'm then going to look at, let's say I have 3 ideas where I'm like, you know what? I would really be happy if I, you know, with any of these 3 ideas, I could wake up every day and focus on this and I'd love it. So, let's say you have that. Now, I'm looking at what's called TAM, right? What's your total addressable market, right? What is the opportunity, right? Is this a billion-dollar idea? Is it a $100 million idea? Is it a $100,000 idea? And I'm going to go after that. And I'm also going to ask some questions about how hard is this idea, right? Because if it's a really complex thing, but it's a billion-dollar idea, okay, that's cool. But can you do it? Can you really get through it? What are the competitors? Competitors look like. By the way, guys, I don't care about competitors. I actually want there to be competitors in my space because it means people are willing to pay for a product, right? If you see a big whatever, Ticketmaster, and you want to build a ticketing platform, go after, go after them, right? Figure out what you don't like about them and beat them. Beat them at their own game. Okay, so don't worry about competitors, but, but you got to consider all those things. Um, the reality is, Ethan, you can sit down and you can You can actually ask ChatGPT to interview you one question at a time, just like we did earlier, and say, hey, I've got these 3 ideas. I want you to really pull out of me, help me analyze this. What do I really want? Ask me questions so we can figure out which idea I should really go for and which one has the most financial upside, which one is going to get me there with the least investment, which is the one that's going to get me there where, you know, to be cash flow positive the fastest, whatever. There's a bunch of different calculus in there that's related to how much effort, time, and money it's going to take to get it to a place where you're like, hey, this is actually my job now. I can do this right, even if it's on the side while you're going to school. I think it's awesome. You're at BYU and keep going. I was an entrepreneur at BYU and I took 8 years to graduate, 4 years standard plus 2 years of admission plus 2 years of doing some entrepreneur stuff, entrepreneurship stuff in there. So anyway, I think, I think that's probably enough said on that, Ethan. But, you know, just find something you're interested in that has a good opportunity. And then here's the cool thing. Maybe you do have those 3 ideas and you can't really narrow it down. They're all kind of equal. Well, build all 3 of them, put them in front of some customers and just see what happens. And one of them is just going to like feel natural. It's just going to go, right? And just go with the flow.

Okay.

Good stuff, guys. All right. So we're going to dive back into Replit here.

Go ahead, Abby. Everything you said was so good. I just want to add something from because I'm actually from like the product world. Everything Jon said, 100%. When you get through all of that, the last thing you can do is ask ChatGPT or some AI to go on Reddit and see which of your ideas has the most number of people talking about it as a problem that can be solved for them.

And that can help you have some idea about which one has the biggest market available. I love it. Absolutely, Abby. Yeah.

And that, that's Reddit is— if, if you're looking for Ethan, I can tell you're kind of like the kind of person that's like, wants to research this stuff. So you wanna look for terms like product, finding product market fit, and you wanna, the other term would be, um, solving customer problems, something like that. That should find you everything you need to get deep into this topic.

Some of the best companies, thank you, are, were created by people that had a problem and they wanted to solve the problem for themselves.

Right?

And you're like, man, I love this. And I'm sure other people have the same problem, so I'm going to do that, right? But yes, absolutely. All of this. Thank you, Abby.

Thank you, Carrie. You said finding product market fit. What was the other thing you said? Solving customer problems. Just kind of reword of that. People who don't know the full term, we'll just talk about solving customer problems. I'm just writing that down in our chat so I have a record of it.

Perfect. Thank you. Yeah, guys. There are, there are so many cool things you can do. Like, I wish I had time to really dive in deep into my conversations that I've had with, with chat about, about some of the things that I've done with GenAI API. But I'm really digging in. I'm like, what is, what, where is the gap in the market? Right? Like, what, what is not done? That's not what, what has not been solved yet, but it's a real problem, right? And that's just other ways of approaching this exact thing, right? Um, but you can— it can come up and say, hey, yeah, this competitor does this, this competitor does this, this competitor does this, and this is kind of an open place that no one's really solved yet. And you're like, awesome, I would love to solve that. I'll include these other solutions, but then I'm going to solve this too, and now I'm going to have the most complete solution on the market, right? So I love it, guys. Um, really fascinating. I love startups. I really love startups. I love talking about ideas and how to, how to make stuff come to life. Okay. All right. So let's dive back into Replit, guys. Understanding how to use Replit is, you know, and use it well is, is a skill. It takes a little bit of work, takes a little bit of just practice. Understanding some tech things are going to help you. And so I have this folder right here that you can see on the screen. Software tools overviews, advanced workflows and things. You guys are going to get access to this folder. You'll be able to look at it. One of the more useful folders I think is this software and tools thing, right? Someone asked me earlier, might have been Rich, I don't remember who it was, said, hey, what are the, what are all the pieces of software you use to build this, right? And so here's, you know, this is a great document that has, you know, here's tools for planning and building, website development, hosting, branding, you know, sales, marketing, coding, backend. We're going to go over all of these in this platform, but— or sorry, in this, in this course. But, but that, um, that document's really nice. Um, we're going to talk really quickly about just coding at a very high level, understanding some words. And this will take about 3 minutes, but I want you to just pay attention so you have a few words that you can just kind of pull out. I think it's really helpful. So we already talked about frontend versus backend earlier, right? Frontend is what you can see, backend is what's happening kind of behind the scenes and the servers and the database bases. And there's a lot of things that that makes up, right? But as an example, right, few things you'll hear: HTML, CSS, JavaScript, React. You don't need to know what all those things do. You just need to know those are some languages that you might see there, right? An example would be like a login form where users, you know, you're on Netflix and you're logging in, right, where you type your password and everything. That's all front-end stuff, okay? But then when you click a button, well, there's going to be an API. There's a connection that is talking to the back end that's it's going to say, hey, we're going to check these login credentials against the database and say, hey, does this combination exist? Abby Jackson, password, green, green hair. Sorry, I saw your green hair, so I went with that, which I love. But anyway, so it's going to check that database and say, hey, yes, we've got this correct Abby and green connection. Yep, that's great. We're going to let her in to Netflix, or nope, that's a bad password, so we're going to it's gonna send a message back, and the frontend is gonna display and say, wrong password, try again, right? So, so the backend is where all this stuff is stored, right? It handles authentication, communicates with databases, manages business logic, etc. You don't really need to understand a ton about that. But you do need to understand that there are lots of things talking to each other. Okay, we'll talk about that more in a second here. So databases, all you need to really know here, guys, is we're going to be using a database inside of Replit, which is really cool. You don't have to go somewhere else to figure out a database. But a database, guys, is basically like a table. It's an Excel document, right? It's a bunch of cells, you know, you've got rows, you've got columns, and it just says like, okay, user ID, password, last login, you know, whatever. It's just all this information and how it kind of talks to each other. And understanding that will make you much, much better. Understanding how your database works will make you much better. And I'll show you how to do that. It's really, really easy. It's just logic. It's like putting together a puzzle. APIs are amazing. APIs make the world go round. An API, or an Application Programming Interface, basically allows different applications to talk to each other. For example, I'm a whitewater kayaker, and one of the things you need to kayak in a river is water. Sometimes rivers are dry, sometimes they're too low, maybe they're too high. But you need to be able to check the level of the river, right? And so I have a few options of how I could figure that out. One, I could maybe ask my buddy that lives by the river, hey, how's the river doing? Oh, it's perfect. Come, come jump in. Okay. That's one way. Two, I could go and set up my own sensors out there and, and say, okay, you know, I'm going to put this little thing here. I'm going to add a little cellular thing. And so it'll ping me whenever the river's at the right level, or I'll be able to see what it's at. Well, that's kind of tough., or 3, I could say, hey, you know what? I bet someone else has done that. And yes, the US Geological Survey has done that. The USGS has done that. So there are already sensors in every river in America basically, and I can check the level really quick using their, their software, right? They also have what's called an API, so I can get in there and I can actually pull that data. So if I wanted to make an app for myself, that's Jon's Rivers, right? And it pings me every time my rivers are right at the right level, okay, then I would be able to just plug into that API, pull that data, and I can do cool stuff. I could plug into ESPN's API and I could create my own sport app. I could create whatever I want, right? There's all kinds of things you can do. APIs— there are millions of APIs out there that basically connect one thing to another, right? A button to a database, a database to a report, you know, it's, it's all these different things. Now APIs, they do a few things here: GET, POST, PUT, DELETE. Okay, this is how the entire internet works. Okay, it's very, very simple. Let's say I go to Josh's Facebook profile, right? We're friends now, we're connected. What it's going to do is it's going to get, or it's going to retrieve, all of his latest posts, right? All the latest things he's put on there. Maybe he did that yesterday, maybe he posted something. How many of you guys have posted something to social media? All of you. All of you have done that. Post is actually a technical term that comes from API. We're going to post, we're going to put new stuff in the database. That's then going to be part, it's going to be tied to Josh's profile, and it's going to say he posted it at this time and this place, whatever. Then the display is going to say, when it's thinking about how is it going to retrieve the data, it's going to say, on the Facebook wall, it's gonna do reverse chronological order, right? So we're gonna see the latest here, and then we go back in time as we scroll down. Okay? And so that's just some instructions. You're gonna get data and you're gonna display it like this. And here's the formatting. And here, you know, that's, that's how it's actually working. Okay? You can also, you know, update something. Oh, I accidentally had a typo. I need to edit that comment. I'm gonna, I'm gonna do that update, save. Okay, boom. We just did a PUT, a little PUT thing, right? We're gonna update that data. Or, you know what? I probably shouldn't have said that. Said that, I'm going to delete it, right? Might take down later. Here we go. All right, so I'm going to delete that. So GET, POST, PUT, DELETE— that is how the entire internet works. Every news organization, every ad that shows up for you, every— everything. It's getting the ad, it's posting something when you make a comment or you push— put something out there, you put a new blog, um, it's updating something, it's deleting something. Whenever you click that reset password button, right? You're kicking off a PUT function, right, that's going to go through and it's going to update that password when you meet all the qualifications for it. Okay, you can also, um, so this is something similar to this, is called CRUD— create, read, update, delete. Let's say you're inside of your system in Replit, right, and you create, you know, whatever. We'll look at this. Maybe there's an admin section here in Chorley where I can manage users, right? And so in order to do that, I, I say, hey, you know what, I need an admin that allows me to create, read, update, or delete users. I want to manage users. I could just say, give me an admin function for users with full CRUD functionality. CRUD is a great little thing that just, that means I can add a new user, I can see the users, I can update a user, I can edit it, or I can delete the user. Very useful little thing there. Other than that, authentication and security, we got token-based. I've asked, my system to do what's called a JWT token, which is a JSON Web Token, which basically just means it's, it's a very simple, solid way to set up your, your auth. You can also set up login, login with Google, login with Facebook if you want to do that. A little more complex, we don't need to worry about that today. Okay, that's all we're going to do there. No questions on that for now, um, but I just— you can go back and review that. You can take another course, you can ask ChatGPT to teach you about it whatever. But bottom line is, understanding some of those things will help you understand what you're doing here. So here is how I now work. This might be a complete app, but it's probably not a complete app. Almost always when it goes in and it builds all this stuff out, right, it's only done 5 minutes of work. You know, it's not building the whole thing, it's not quite ready. I mean, there's a lot of things that built for me, right? But the way for me to check it, especially as a non-developer, is I just have to start using it. I'm going to try to sign up. Let's just try this. We'll go Chinnipiano, that's my username. I usually do— I'm going to use all real information here. Create account. Let's see if this works. Hey, that's pretty cool. We'll save that password. I logged in, it worked. Many times when I first try to log in, it won't work. Then I have to say, hey, I tried to log in, it didn't work. But hey, It got me in. This is amazing. I'm on the pro version, it looks like right now. Let's try adding a chore. Okay, so let's say, yeah, clean chicken coop. We'll just use this example here. Clean chicken coop. Do it well. And then I need to do it. Okay, I need to create a new zone. That's kind of cool. I might do that in a minute. I need to clean the chicken coop. Let's say monthly. Okay, we're going to create a new zone. Zone name, chicken coop. Okay, and create chore. Wow, I'm shocked, guys, that that worked. Like, I really am shocked that that worked. That's so cool. I need to check something really quick here. Um, one of the things we need to look at— we, we talked about a database, right? Do we have a database? We do have a database. Okay, this is amazing. We've got chores, zones, session. I don't know what session is. I think that might— that's interesting. That's just my session., and then users, right? So we've got my, my, my email, my username, my password. Now this is cool. I want you guys to look at this database, okay? It's not that complicated to understand. The ID is just going to be a randomly generated thing. It's just kind of a unique ID that the system uses. Then I have my username, my email, my user— that's, that's going to be my communication— and my password. This was not my password. I did not put in C5C964192. This is called a hashed password. Okay, it's going to automatically kind of obfuscate it, right, so that you're not seeing other people's passwords in your database. That's what you want. Okay, I've got my name, I've got when I was created, and then I'm connected to the chores and the zones thing here. So that's pretty cool, right? Then I can look at my chores, right? Title, description, frequency, zone ID, right? It's completed, false, last completed, created at. Okay. I start to understand how these things fit together by just reading the columns. There's nothing fancy about this. This is just like, okay, look, there's a session and this is actually a pretty well-built app already. Let's see, already complete. I want to edit this. I want to be able to edit. It won't let me edit this, so I clicked Edit. I want to be able to edit, for example, when I last completed it. Okay. So what I'm going to do here is I have this, there's, there's another button. So you've seen me before, you've seen me do some like screenshots, right? Or you can do that and you can drag it in and tell it something. You can also use this little tool right here that says select element. Okay. Or element selector, it might be called if you're still on the last version of Replit. Okay. But when I click select element, I can click right here in this card and I can say, um, I tried. Clicking the edit button in the menu, and it— in a little dot menu, I don't know what to call that thing anyway— and it didn't do anything. I'd like to be able to edit the last completion for a task, especially when I'm just adding it. We probably should add this functionality to the add chore, um, it's called modal as well, right? Okay, so there's several things that are going to happen here when I do this. Number one, I, I selected it so it knew exactly what I was talking about. It's not talking about this thing or this thing or this thing. It knows I'm I'm talking about this card and so it knows, okay, I can look right at the code right there. Then it's going to find the little dot, dot, dot menu. It's going to make sure that— so I asked it to do several things. I clicked edit, nothing happens. We need to make sure something happens. I also asked it to be able to allow me to edit the last completion for something. Then I also said, let's also change this modal so that we can put the last time it was done. Um, I also— it says enter new zone name here. Um, the chicken coop is there. That's cool. It did save it. I was going to say we probably need a menu so that we can select a past location, but it's already there. Although it was the, the, you, the, the experience was just a little bit off. So guys, I'm in V3 of Replit. Replit is getting better and better and better and better. If you're in V2 today, you might, you know, you might have to work a little bit more. It might not need it quite as much as mine has been done already. Um, like, this is, this is pretty cool. Fresh, too soon, all locations, chicken coop, all locations. Yeah, that's— this is crazy. I'm really impressed at how much it's done already, right? But it's going through, it's fixing some things, it's trying to figure out what's happening. It made some changes and then it found a problem and now it's fixing some errors, and I don't even need to understand that. But basically what you're going to do, guys, is when I was building GenAIPI— okay, let's just pull that up for a second while this continues to work. Um, I had to, I had to go through and every single thing had to work in order for me to put this in front of a customer and ask them to give me money. I had to, I had to make sure there weren't what I like to call paper cuts, right? A paper cut would be, okay, I click on start free assessment and then I click not now and there's an error that pops up, right? Or I click start free assessment, I click start the test and something doesn't work, right? But that, those buttons both work. That's great. Okay. There's a little sign up directly here. Okay. And let's go here. Okay. Now it wants me to sign in for that. Now I can go do that. Okay, that, that, that works. If I click this, this looks like this is just kind of a basic thing here, right? But if I click Explore Live Courses, that's going to work. Okay. If I go down to here and I click on, you know, the blog, does that work? Can I go to the privacy policy? Does that exist? Right. Is there an actual email that works? Can I click on something, you know, specific, you know, whatever. Right. Bottom line is, guys, there's a bunch of different places in your app where things can go wrong. And if I click on— if I have a user come here and say, oh, I am an enterprise, I'm going to explore solutions, and I click on that and it doesn't work, or it goes to a 404, some error or whatever, they're going to be like, oh, whoa, whoa, whoa, whoa. Their antenna is going to go up and they're going to say, this— I don't know if I'm going to trust this company because their website doesn't even work, right? I can't even click on a button and expect it to to happen. Uh-oh. Let's see. There's a problem. That's not great. Right? Let's start with assessment. Okay. Actually, that's not a problem. Ready to activate your versus app potential. Let's start with assessment. Right? Start with assessment.

Right?

And so that kind of brings us to a contact page. I actually don't love that. I don't love that, that workflow. Right? Schedule a live training consultation. It's okay. I think it's alright. I kind of would like to just go to a scheduling page. Right? And so maybe I need to update my enterprise AI solutions so that when they click schedule, it actually brings up a calendar that they can schedule on, right? That's going to feel a little bit better. Okay. And so even I have to go through my website and figure out, okay, what's going on? And so I can improve this. But the point is, on my website, at least everything works right no matter what I'm doing. Like, I can click on stuff. It's, you know, if I want to take the test now, it's going to make me sign in, whatever. If I click on these courses, if I want to get access, it's it's going to take me to here. I want to buy this, it's going to make me sign in, um, it's gonna make me sign up for an account. Oh yeah, it's the authentication required that I have to sign in and that will allow me, although I've already purchased those because, um, because I gave it to myself because, you know, I'm the boss. But anyway, bottom line is, guys, you got to get through each of these little tasks. Like, it has to go through, um, so let's see if I can even edit this now. Now that it's been working on it. Yeah, it's been working for a minute now. Considering options, task lists, add last completion date. Yeah, it's still, it's still working on stuff here. Okay. But I don't love that it just says already complete right when I put it in here and I don't have a date. And so I need it to do that. So it's going to, yeah, updated authentication and storage to support user session refactors using the, yeah, it did something crazy that, you know, it needed to do in order for that to happen. Right? So, um, now it's building the chore edit modal. So it's, it's probable that the chore edit modal doesn't even exist yet, right? So now it's going to figure out, okay, what needs to happen here. I need to create this that allows for all these things. I'm going to do all this stuff. It's really cool to watch it think as you, um, go through, guys, and you work on your different features. You just have to get it to where it works, right? And when it doesn't work Just tell it what doesn't happen and what you want it to do. You don't have to be technical. Really, you don't have to be technical. I know a few words just because I've spent time reading these things. I also have the advantage of I was a tech CEO for a decade. I had smart developers that I would ask questions to. I'd be like, what is Kubernetes and what does this mean and why do we have to Dockerize this and what does that mean? Weird stuff that unless you're in the room talking about those things, you probably won't understand and you don't need to. But Replit makes it really, really, really easy to do all this. One thing that I want to make sure you guys all do is have a, um, a database set up. So if you click on database and it says please set up a database, then you're going to want to, you're going to want to click that button and have it set up a database for you. Okay, you want your system to work off of a database. The reason is if you don't, it's going to create something that feels like it works but doesn't actually work. It's going to be using temporary or what's called memory storage. And so as soon as you refresh or turn off your computer or go to a different device or whatever and you try to get in and do something that you've saved, it'll all be gone. If you save it in an actual database, it'll be there wherever you access it from. Okay, so you need a database. Make sure you go in and do that. It's as simple as just clicking on this little plus button. Click on database, right? It'll be right in here and just click create a database. Okay, that's all you got to do. It knows the context of everything you need to have happen there. That's all you got to worry about for right now. By the end of day 1, guys, where we are right now, I like everyone to have their project up, their kind of first version up. I like you to have created your database, and I want you going in there and playing with your app like I am doing right now, where you're just saying, okay, I want to click on stuff and I want to see if it works, right? Let's click on the map view, see if that works. Nope, doesn't work yet. OK, so I'll tell it, hey, my map view doesn't work. OK, alright, I'm going to pause for a second. Any questions about what's going on here? Is anybody super lost or do you guys feel like you understand some of what's happening? Jon, I have kind of a side question. Yeah, go ahead. So I, when I was in ChatGPT, I had it create a landing page for me, which was awesome. And now I've got my page over in Replit and it looks pretty good, but I prefer to look more like the other one and I have the code for the other one. Is it— can I copy the code and tell it to create it more like that one? And it's not going to break anything. No, especially if you use exactly the words you just said. Say, hey, I need you to modify the homepage to look more like this. Here's the code that, you know, ChatGPT created for me. I want you to make sure it fits the project, but I want it to look like this. Okay. All right. It's a great thing. I typically don't try to build stuff outside of Replit simply because if you're doing something more complex than that, it can get confused because ChatGPT does not understand all of the code of your whole project, doesn't understand how it all fits together. But when you're at the very beginning and it's just like your homepage, no big deal. Okay.

Okay.

All right. Let's click edit, see if that— Yeah, no problem.

No problem. Other questions, guys? Jon, I'm, I'm, I think about a bit off more than I think As far as the requirements and everything like that, I'm actually still not able to pull up a preview of what I built and it says my app's not running and I've rolled back a couple of times and look at other ways and ask it to reconsider and it's just stopped. I wonder if I've just got myself into a corner.

That's okay. Why don't you share your screen? Let's get you fixed here. So here, okay, click— what happens when you click run over there on the right to preview your app? Nothing. Okay, so yeah, just say the app isn't running. I know you've said that already. I just want to see what happens here.

Go ahead and scroll up to the very top.

You see the original chat maybe, or is this good? Yeah, click on that and scroll to the top. Previous events. Yeah, keep going up. Okay. All right, that's fine. So go back to the other chat, which is going to be— you push that little button over there on the left top top left there. It's like a little like history. Yeah, it's the history thing. You just scrolled over it. Yeah. And then just say view chat or continue chat or yeah. Now that's the ended one. Maybe, yeah, maybe go back to the other one and say continue chat. Yeah. Scroll to the very bottom now and then just continue chat. There we go. There we go. Yeah, it's almost like you, you like deleted a lot of stuff that you've done. By the way, I will say this, guys. Sometimes I get into a weird loop and I'll just be like, all right, I'm just gonna start over. I'll start a new project and I'll copy paste everything back in, and in 5, 10 minutes I'm like way ahead of where I was, right? Yeah. Um, and so once you've done it a couple times, you'll see that I, I, I don't really run into this very often at this point. I'm pretty smooth now because I know what I'm looking I know what I'm going to mess up or what's— anyway, front end built successfully. Now let me start the full server with the API. Let's see if it actually— hey, it looks like it's actually pulling up. Preview be available soon. Agent is working. It's doing a server restart. It's interesting. So, Jon, Are you using Replit to make the edits? So you're not going back to ChatGPT or Grok and say, hey, we need to think more about this or that. You're making all those edits and updates in Replit itself. So Replit, guys, is using Claude 4.0, you know, in here. That's who you're talking to. That's what's underneath the hood. And it is every bit as smart as Grok or ChatGPT, right? You do not need to go to ChatGPT or Grok. We're done with ChatGPT for now. Okay? I use that as my ideas, organizing, creating some documents and things like that. And then I'm gonna put all that, I'm gonna hand it to Replit and say, all right, let's go. Yeah, it's still not working. Yeah. Honestly, Josh, I would start a new project. Copy all the stuff, just paste it back in, start a new project. That's, it's gonna be your easiest thing. I don't know how much work you've done on it. I mean, not much, obviously, cuz we're just sitting but something broke and I don't like to use the rollback feature a lot unless I really know what I'm rolling back.

I was just wondering how long do you persist until you just pull the cord and be like, okay, I'm going to go back and reboot and set it?

I'm going to do that if I'm working on GenAI API and something messes up and I've already put 400 hours into it, Well, yeah, I'm gonna fight through. I'm gonna figure out like, what did you do? I'll go back a little bit. And I'm doing it very modularly, right? I'm like, we're gonna work on this part. Let's— please don't break anything. Like, that's some of the more advanced stuff we'll talk about maybe tomorrow. But, but you have to, you have to really be careful when you start to get into really complex things because it'll just kind of do crazy stuff. It's getting smarter at not breaking stuff. Like, like, again, I'm I'm honestly shocked at how well this works right now.

Yeah. Right. Jon, is there a backup feature? Like if Josh wanted to keep trying, could he, like, is there a way to mark a backup? Like as a programmer, I would just revert to a previous commit, right? Is there a way rather than doing a rollback, is there a way for him to be like, okay, I'm gonna mark this at this time and then he'll do a whole bunch of things for like 2 hours and be like, ah, that's not working and go all the way back.

Yeah. So there's, I'm gonna say two primary ways to do that, Abby. One, if you look at my screen here, you can do, you can sync with Git, with GitHub. Oh, you can. Okay. Yeah. And so you can do that. You can pull, push, you know, whatever you need to do. You just go into your settings, connect, you know, create a repository, set that stuff up. And so guys, you don't need to worry about this right now, but bottom line is GitHub is a place where you can push your code to and have some version control, right? Which is great. Another thing is if I'm gonna try something crazy, Abby or Josh or whatever, if you're gonna like, okay, I'm here, this works. I really want to try something that's kind of different and interesting. I can do what's called a remix, right? And so I actually, wait, I don't want to, I don't wanna do that. I just wanna click here. Here we go. So right here I can click more options and I can click remix. And a remix will just create a copy of everything you have at that moment. And then you could— you're— it's a branch creating a branch, right? You're just saying, okay, I'm gonna go these two different directions here. One of them is gonna go this way, but it's actually create— when you do a remix, it's a whole new project, right? It creates a whole separate project for it. Um, and so, you know, maybe I help Carrie put something together for her thing and I'm like, hey Carrie, why don't you just take my project and go? Well, instead of having to let her into my project in my my, you know, Replit account or whatever, right? Which I can do. I can share and add people to it, but maybe I just want her to go do her own thing. I'll be like, hey, I'm going to remix this. You can take where I've taken this to and you can just go, right? It's all yours now. Okay. So that's really helpful, especially if you're doing stuff for clients, right? If you're working with a client, you're building something. If I'm building it in my own, you know, account, then I'm going to be able to push that over to somewhere else. Right. So anyway, I don't know what is going on here. It's doing some sort of testing here, but it— I feel like it can't do it well. I don't know why. Anyway, my edit works now. That's cool. Last completion date. Last completion date was August 5th. Update chore. Yeah, okay, so we got an error here. This is good. This happened when I clicked update chore. After entering a date. Okay, so I can give it that. I'm just going to tell it— let it go. It's kind of stuck in a weird loop, so I'm just going to let it do that. So guys, you just have to kind of— you just have to kind of work through. And this is where it takes— this is why it took me, you know, 3 days to knock this out, to knock out GenAIPI, is because I had to change the text and I had to change this and I pushed on that button, it didn't work, and I had to test the payment here and I had test this, and I had to get an email system set up. We're going to talk about APIs for just a second here, guys, um, um, and how to set one up. I'll let you try to venture off and do it. APIs are very, very easy. For some of you, this would be like, oh my gosh, this is awesome, so easy. Some of you might be like, okay, I'm gonna wait till tomorrow and see this a little bit more. But an API, remember, is we're connecting to another service. So one of the things that could be really useful for a chore app, maybe is what's the weather going to be? That might determine what I can do today. If it's raining outside or if it's snowing outside, maybe I can't work outside today, right? If it's going to be nice, 75 and sunny, yeah, let's, let's do some more stuff outside. So maybe I can make this thing a little bit smarter to say, hey, here's what you should do today based on the weather and your priorities and what's coming up, what's overdue, what's due soon, right? So let's see if we can add a weather API here. So it's going to consider options and it's going to work for a minute here. Um, man, I really don't like this annotate button. How do I get rid of the annotate thing from Zoom? All right. We're just gonna move it to that corner. All right. So there's, there's some really cool stuff here, guys, that I love in the new version. One of them is build one of them and you can plan and edit. This stuff is really, really awesome. It's it. You'll have it. You'll have access to it probably tomorrow when it goes live. Or maybe, maybe it doesn't go live for another week after that.

I'm not sure when it goes live, but it's going to be very, very soon. I like just got access in the last 10 minutes.

It wasn't there a moment ago. That's cool. I know they're rolling it out like right now. Like it's new, it's new stuff. But, but guys, because the agent is, um, it wants to do stuff.

Right?

It's action-oriented, right? If you ask— sometimes you want to ask it a question like, hey, how does this thing work? Or why is this thing not working right here? I'm trying to figure this out, right? What's wrong? And by the way, Josh, that's what I would have done next with you. I would have said, you keep trying to get this working, it's not working. I need you to explain to me why it's not working, right? Don't write any code, just tell me. Let's just have a conversation, right? And then it will actually take a step back and think and do stuff, right? And that works really, really well. Now, instead of having to like say, hey, don't run any code, you can switch the mode to plan, right? Ask questions, plan your work, right? And you can ask it to make a plan. And so you could say, okay, and Josh, you would have been able to say, you know, okay, now that you understand this, what's your plan to actually make this work? Tell me a plan that you know 100% will get the app running, right? And then boom, you can go into it and, and make that happen, right? And then you switch back to build and say, do the plan, right? But you can have the AI work with you and plan and do all that stuff. And that's why, Richard, you don't have to go back into ChatGPT or Groq to plan and stuff. You can do it all right here, right? Wanted you be. And, and, and the thing is ChatGPT, like I was saying this to Jake earlier, ChatGPT does not know, you know, you, we've already written, you know, 10,000 lines of code in here, right? There's all these different things and ChatGPT doesn't have access to this stuff. So it has no idea what it's even trying to do and what the context of everything is, right? And so that's why you really want to have your conversations and plan and build all inside of the platform. That's my advice.

Okay, so— Jon, a question for you, just maybe diversion for a second, but I think about this from a— I'm testing, I'm looking, I'm trying, making sure everything works, and also just the learning that happens as far as understanding behavior in the app once people are using it. Is there, is there tooling that's kind of new or early stage even for Replit that's really kind of helping you kind of analyze user pathways, experience, and optimizing that as well? Or is it pretty traditionally still the web analytics, the, you know, those things? How do you refine things once you have people in?

Maybe that's ahead of the— No, that is, you're, you're jumping about 8 steps ahead, but all the short answer is, um, Replit doesn't have built-in tools that are going to look at like heat maps of where people are clicking and where they're getting in trouble with stuff, right? You would install an API, you would install some sort of kit, right, an SDK or something that, that's going to track that, right? And there's, there's 50 different things you can use. Microsoft, right? I mean, there's amazing systems you can just drop in that then will show you exactly, you know, what the user is doing, right? Um, another thing that's really simple Josh, it's just release and tell people, say, hey, this is, this is, you know, still a work in progress. When you run into a problem, just tell me, right? Then you just take the problems, give it to Replit, take the problem, give it to Replit, take the problem, give it to Replit. There's the next version, right? And so just release early, release early so that you can, you know, start to get that, that feedback. That's cool. Thanks. Yeah, yeah, for sure. All right, guys. So we're going to build an API now and this is very, very easy. So the API I want is going to be something called— it's a weather API. API, right? Now inside of Replit, there's something called a Secrets Manager. Okay, so here's your secrets. It's right here when you click on that little plus, the tab button right here, you can click Secrets. And Secrets is going to show, all right, here are all the secrets related to your current app, right? Your database URL and some of these things that, you know, let your session start and actually give you access to come in and edit your app. All right, then you have a second tab here that's Account Secrets, right? So I've, I've made 100 apps inside of Replit, and I've used tons of different APIs, whether I'm using OpenAI or ElevenLabs or Mapbox or Cloudinary or Bright Data and tons more. And I even have some weather ones, right? I know about this Weather API key. It's a great one that's out there. Now let me show you what a weather API, what Weather API looks like. Okay, so here's weatherapi.com. When you're looking at some service and you're like, I want to integrate with this, right? I want you guys to just really know that this is not complex, right? Some people look at this and like, what the heck is JSON and XML and all that? Like, I'm lost already. Forget about that. All you need to know is this is an API, which means it's going to be a connector and it's going to have weather data and you're going to pull it in here and you don't have to understand it. Why? Because you can literally just copy this and hand it to Replit. You copy the thing and just say, figure out how to install this, and it will just do it.. Okay, but one thing I'm going to look at, I'm going to come over and I'm going to look at the pricing. Okay, I want to see, is this a good thing for me to work with? So most APIs will have some sort of free tier. Some will have, you know, then packages going up, right? So let's say I'm making a weather app or I just want to integrate weather into my chore app. Okay, so I can look here, I can scroll down, look how many calls per month I get with my free. A million calls per month. That's crazy. Right. So I can get so much data, you know, there's, by the time I hit my $7 a month minimum charge, you know, I can get 3 million. I'm going to be making money. I can afford the $7. I can afford the $7 a month or the $25. Maybe I want a 14-day forecast right from the get-go. Oh yeah, that's fine. I can use a 14-day trial and I can install that and boom. Right. I'll start with the free and I'll see if I like the data and how it's coming through in the format and does it work well? Is Replit able to build with it? You can look at whatever other features you want. You guys are all smart enough to just read through this and be like, okay, do I need to know, you know, sports API? I don't even know what that does, right? Sports API history with weather, I'm not sure what that does. Like, was it raining during the game? I'm not sure what that is, but sounds cool. Um, you know, commercial use. Hey, you can even use commercial use on the free one, right? Non-commercial use, of course. Are there annual payments available? SLAs? Um, yeah, SLA means service level agreement, like how often is it going to be up, uptime, 100% guaranteed. They are going to put a dedicated server on there for you, and you can pay for more if you need, right? Tide data, you don't get tide data here. Marine weather really starts to kick in, it looks like, at the, you know, at the paid version. So bottom line is you kind of just understand what you have. Then you're going to sign up, okay? So I'm going to log in, right? I've already signed up. You're going to log in. And here's my username and password. And then what it's going to give me, guys, is an API key. Okay, now I'm exposing my API key to you right here, and that's a no-no, right? Typically shouldn't do that. I'm not worried about you guys taking my weather API key. I'm on the free plan. If you want to take it, go for it, but it's easier for you to get your own. You don't want to use mine, you want to see your own data. Okay, so that's why I'm not really worried about this. But basically all I'm going to do is copy it and give it to Replit when it asks me. I'm going ahead, I'm getting ahead of myself. Okay, but right now I'm going to say I want you to add a weather widget to the logged-in dashboard that shows up— where do I want this? Yeah, that shows up in the nav menu. Maybe let's just do that. Shows up now. Maybe I don't want the nav menu. I'm trying to think about mobile. Maybe that's not bad. It's not bad. Yeah, in the nav menu next to the, to the left of the profile dropdown. Which is what I think this is.

Yeah.

Whoops, I don't want to do that. Unselect. Okay. And then I'm going to say let's use the weatherapi.com API. Okay. Um, yeah, let's just say that. I don't know what this new, like, error connecting to the browser— I'm not sure what's happening there. I don't know what it's doing. Feels like it's working really hard to get something that should be pretty simple to work, right? See if we can change the completion date to August 6th and update. Yeah, still getting the same error. I don't know why it's thinking so hard about that. But anyway, guys, all I had to do to implement this, uh, thing here— I'm just going to go ahead and tell it this API will fix to this other thing later. Right now I want to add the Weather API element. Okay, so I basically just told it that, and what it's going to do in a second here is it's going to say, okay, here's how you are going— you're going to do that, right? And it would tell you, go to weatherapi.com, get— create an account, create a free account, grab the API key, give it to me here, right? So let's just let this go for a second. Let's make sure we're not on high. Automatically test changes after making code changes. That's interesting. That's where it's getting hung up on some of what I'm doing here. I might want to turn that off for now and I'll just test it myself. I think it's cool. I think it's a cool new feature. I'm learning this, guys. Brand spanking new, right? Um, Sometimes when you get into really complex problems, guys, you can actually say, hey, use the higher power model. So it uses— this is again using Anthropic, um, is the company, right? Like OpenAI is the company, ChatGPT is the product. Anthropic is the company, Claude is the product here. It's the, it's the actual LLM, the brain, right? And there's different, there's different levels, right? So we're not using Opus, we're using what's called Sonnet. Okay, so Sonnet 4 is super smart, but Opus 4.1 is way smarter, but it's also much, much more expensive. Export pricing. So if you look at how much it costs, right, Claude Sonnet 4 is using, you know, you're talking $3 million— or sorry, not $3 million, $3 per million tokens. Okay, whereas Quad Opus 4.1 is 5 times as expensive. Okay, so you can switch to that higher powered model, but it will cost you a lot more. Right now you have inside of Replit, you have, you know, $25 or $30, whatever it is, of credits every month. And then as you go past that, you're going to get charged a little bit more every $10. Right. So again, to build GenAI API, it took me about $120 of credits, right? I spent money on a couple other things as I was building it out, but $120 of credits to build the fully functioning GenAI website. If I was using the high power model to do it, I would have spent $1,000. It's a lot more expensive, right? But we're still talking $1,000 even at the highest power model as compared to what I would have to pay Abby to build something for me. She's going to be way more expensive, right? She'll be $1,000 in the first 3 days. Right? For real, right? Like, that's just what developers are going to charge. And so I know that because I've paid them, right? It costs a lot of money to pay people. AI is still dramatically cheaper. So even if you did do that, I'm only going to use the higher-powered model, um, if I'm, uh, if I'm running into a problem that's really complex and it can't figure it out. So, and I'm probably going to use the higher-powered model to plan and then I'll use the lower-powered model to actually execute because the lower-powered model is like, it's still brilliant, right? It's more than capable of doing it, but sometimes it just needs to think through some weird puzzle.

So, all right.

And Jon, how can we tell how many credits we've already used? Yeah, there used to be a button right here. They've changed the UI, so I need to find it again. I need to figure out where it is. Is it account? It's probably not account. I mean, it might be. Let's just look at it. Let's see if that comes up here. Usage right here. So replit.com/usage. I got to it in account, but, but you can see how much you've spent, how much you've done. I have crazy amounts of credits and things like that, guys, because of some of the Replit stuff that I'm in, right? But, but you can see how much usage has been done and for what project, right? So I've spent a total so far on Chorley. I spent $5, right? Everything I've built so far has cost me $5. This month I spent $22 on GenAIPI. I'm working on some BYU stuff, $4.39, Built for Life. Anyway, some other things I've been working on, right? There's the chore tracker that was kind of my thing I tried earlier today, $1.81. And then there's a bunch of other data here which we won't get into today because I don't want to confuse you guys, but you can look at how, you know, how many credits you have left and what's going on. I personally think that the paying for the annual plan, if you're like, I know I'm going to use this, I'm going to be running and deploying everything through this, I like to just pay for an annual plan. It's way cheaper and you get extra credits for doing that. Anyway, let's see if we have— fix this. Let me add the weather widget. It's about to do the weatherapi.com API key. Sorry guys, I'm, I'm, I'm learning a little bit. It's going just a tiny bit slower than normal because it's got all these new features and it's kind of— I think it's good. I think it's good. What it's doing is it's doing more complete things for me instead of doing half actions that I then have to go clean up. It's doing complete actions, but those complete actions take a little bit longer. Okay, this is what I wanted to show. So to add the weather widget to your dashboard, using weatherapi.com, I need an API key. API keys are secure tokens. It tells you, it's like teaching you as it does this. Go here, sign up for a free account, supports 1 million calls per month. After signing up, you'll get an API key from your dashboard. Copy that and paste it right here. Now, it already chose mine from the account because I already have it in my secrets. I already had that one sitting right here. But that's it. Like any of you could do that without my guidance. You just read it, you paste it, and you click continue. And then it will say, would you like to add to your account secrets? And then click yes, right? And you'll save it and you'll have it then in your secrets later on. I've already done that, so it didn't ask me that, but that's it. That's all it takes to set up your API key. Okay. To now connect to another system. And now it's going to be pulling the data of where I am here in, and I could have it I could have it choose, maybe I need to be able to set the location of my home, right? So it's like, oh yeah, it's in West Mountain, or here's a GPS location or whatever it is so that I could set that properly. But you just have to think through what's your user experience, right? Do you want people to be able to have to manage multiple homes through your chore app, right? Do you want them to be able to do multiple things through whatever it is you're doing, right? So you just have to think through what is my user experience? Experience, and then tell it what you want it to do. Okay, and then you can say, all right, I want the weather app piece to always match not where I am, I want it to match the location of my home. So if I happen to be in Mexico on vacation and I'm wondering about some chore when it was done, it doesn't pull up the Mexico weather, right? It pulls up my home weather. So I can tell whoever, my friend or someone, hey, could you run over and, you know, fix the chickens? Like, oh yeah, it's sunny, it's good, You know, but if it's raining, like, oh, could you go close that door and make sure that something happens? Whatever. That could be a useful piece of information that just shows up right inside of your app that you're using to manage all of these to-do things around your farm, around your house, around whatever. Okay. All right. So this should be setting up here. It looks like it's creating something that should go right here. Yeah, between the add chore. Task summary. Yeah, implementing the weather widget, considering options, thinking.

Perfect. I have a task list. Hey Jon, I've lost my preview. I only have the chat.

Do you know how to get it back? Um, share your screen really quick. Um, that's kind of a weird one. There's like a layout button that I think you can click on. Uh, here, go ahead and do that again.

Go ahead and click share again. Sorry. Okay. When I click this, which is the preview, it opens in a new window.

Right, right, right. At the very top right. Yeah. Wait, no, that's— you're in the Arc browser, huh? I'm in Arc, yeah. Yeah. Well, you're in the plan phase. That's why. I'm in plan, oh.

Yeah, you're still in—

How do I get into build? That's not a button. It doesn't make sense. Why are you in the plan phase here? Have you already seen your app? Like, did you have your app—

it was up previously? Yeah, yeah, it's been working and stuff. I just—

I went to look at the usage, and when I came back, this is what I had. I have no idea. Um, click on that little down arrow. There should be There used to be like a little layout button. Which down arrow? Sorry. There's a lot of down arrows. Yeah. Yeah. Sorry. The top left there. Right, right next to Inhale Mine.

Yeah. That thing right there.

Create a CL—

CLUI.

No, I don't think it's that. Um, Abby, could it be on the right-hand side, upper right, just where there's a pane view or there's the filter?

Oh, yeah.

You're back again. You reset it just by, yeah, you reset it just by closing the project and opening it up.

Yeah.

I just exited it and came back. Awesome.

Okay.

You're up and running. I love it. Oh, good. Good job. Okay. So, all right. One thing that's important that just happened with, with Abby is the— she, she closed out of the agent. The agent doesn't actually stop working if you refresh the page. If something gets weird and you have to like go out and back in, like you're not going to lose any progress. Like it saves everything you've done. The agent will keep running if it's in the middle of a task. Weather. Okay, all right guys, um, look, I'm in San Francisco now. Check it out, we've got a weather app or weather widget, right? So we've added this new feature here. I'm going to show a couple other things really quick that just kind of while we're waiting for this thing to finish. You can also switch to mobile view. You can switch this little button right here that's right next to your address bar. You can say, hey, what does this look like on a phone? Okay. So it's obviously not great, right? Chorley is too close to the add chores, so we'll have to fix some formatting later. We'll do that later when we, when we get there. I can click on this. I think this is kind of just a little bit off. It is showing the weather here. That's cool, though. I can't see where I am or whatever, you know, but you're going to have to go through. You have to check these things on mobile and on desktop to make sure this stuff works. Okay, so, but it's good to just look at, look at all those different variations there. So you see now it's testing the app. I haven't had it successfully test the app well. That's okay. It's going to set up the browser, navigating. I've completed this. Anyway, kind of fun. Let's see if I can change a date successfully yet. Update chore. Oh yeah, because I interrupted it. I told it to kind of not finish that. It's just what's happening, I think, is it's trying to send some information that's not properly set up in the database. And so we'll let it fix that later. Um, but guys, we've got about 20 minutes left today. Um, what I want you to do the rest of today— some of you are just going to go back to work and do other things, and you won't touch this till tomorrow. That's fine. But some of you also will be like, no, I'm going to sit here, I'm going to work on this for another 10 hours. Like, I've had many people be like, I was up till 3 AM and That's fine if you do that. If you have a question, if something isn't coming up right, um, feel free to just message me and say, hey, I ran into this problem, what do I do here? You know, do you have an idea on this? I'm happy to help. I'm going to drop my cell phone in the chat here. Okay, some of you have it already, some of you don't, but that's my cell phone. Feel free to text me anytime. You guys are now part of my, part of my crew, so feel free to use that. Don't be afraid to text me and ask me a question about anything before and after the class, you know, during it, whatever. I'll get back to you when I get back to you, right? If I'm super busy, then I just won't see it. I'm on do not disturb often when I'm on meetings, but, but really guys, you're just going to keep pushing forward. You're going to say, what do I not like about this? Like, honestly, this, this nav menu now is kind of crazy. Like it's too busy in my mind. Right? So I'm going to have to work on that later. Maybe I get this map view and move this pro thing. I don't know. Got to figure out some of these things, but I need, I need the dashboard to be a little bit better. Profile settings doesn't work yet. Notifications don't work yet. Upgrade to Pro doesn't work yet. None of these buttons really work. Help and support. Let's see if sign out even works. Yep. That works. Let's see if I can log in. You always got to check that. See if login works. It should work because it, yeah, it was great. It pulled me in. Um, it pulled me in just fine. So that's good. But you just keep clicking through and playing with these. Like, I can do this with, with a, a card style, or I can switch to this list view here, right? Mark complete. That's interesting. It changed. Oh, that's cool. Actually, it looks like it actually did change the date, but there was an error probably just in the display.

Anyway, I'll figure that out later. Jon, what cues do you look at as you're testing, working through, and even scaling that you're like, I should probably get like a developer, architect, someone involved that I'm really maybe not going down a rabbit hole that I may regret later? How do you cue into that versus— because it's obviously there's this enterprise-grade thinking, but also just this proof of concept. Where do you find yourself thinking like, I'm going to get a I'm gonna get Abby in here and do some heavy lifting.

So to each his own on this one, or to her own. I have not brought in a developer on GenAI API, and I have, I've done close to $1 million in revenue with GenAI API. My, my business isn't crazy complicated, right? I mean, there are some things, there's a lot of users. I have hundreds of users. In my system right now. I have, you know, tests and I have certifications and courses. And so it's a little bit complex, but I just felt confident enough. And I'm someone who just likes to figure it out, right? I'm just going to dive in and figure it out. Some people are like, you know what, I'm probably confident enough, but I don't have the time. So I'm going to hire a developer to help me, you know, push through this, someone to manage it. And then sometimes you're like, you know what, this actually is going to be a very, very complex thing. I'm probably not bringing in a developer for almost anything because Replit can build very complex stuff. I'm probably not doing it until I have proven it with customers. I want customers to say I am willing to pay for this, and even here's money, right? I want to use your platform. And then on the side, you know, you got to keep it working, you know, to a certain extent. And then on the side you could say, hey, I want to build it, do that. But, but that avoids that question of is anyone gonna even pay for this, right? And so, I mean, with Chorley, with this type of an app, I don't think I'd ever bring a developer in, right? Personally. And, and here's the thing, we haven't optimized for scale. We haven't optimized for GDPR. We haven't optimized for ADA compliance. We haven't optimized for tons of things. You know what? It's not that hard. You just tell your developer, your Replit developer, and say, hey, we need to make sure this is ADA compliant. What do we need to do? Okay. And hey, we need to make sure this is compliant with GDPR. We need to make sure this is compliant with this. We need to make sure this is compliant with that, right? And you just work through it. You just got to do, you know, do those things. Again, maybe you have time, maybe you don't have time to do that, but, but Replit is capable of doing it, right? It really is. And then you just have to ask good questions, right? Hey, I know you set up this ADA thing. Can you do an analysis and tell me, you know, on a scale of 1 to 10, how ADA compliant are we now that you've done those things? Oh, you're at about a 7 out of 10. You still have these other things. Okay, great. Are those important? Are we going to get fined? Well, actually, no. If you're up to the 6 out of 10, then you're going to be good, right? So you're fine. There's a couple more things you could do. How hard are those things? Oh, that's pretty fast. Well, let's do those.

Great.

Now you're at a 9 out of 10. Just keep asking questions and following up. Another element that we'll tackle more a little bit tomorrow is cybersecurity, right? You got to be careful. You don't want to expose API keys. You don't want to expose people's passwords. You don't want a data breach. You don't want all these different things, right? So if you feel confident enough to kind of go and tackle those things, then great. If you just know to ask for those things, you can get by with, with a lot of what's happening here. You really can. Yeah. So So anyway, yeah, hopefully that, hopefully that answers that question. So right now it's just giving me San Francisco weather. There's some things I'd have to do. I'd want to say, hey, you know what, I want to make— I want to add another layer of data here. We need— so I'll say this, I'll say, oh, you know what, guys, I'm going to do that tomorrow. We're going to do one last thing here. This is one last thing that everyone needs to do, and I totally forgot it, and I apologize. I normally do this a little bit earlier in the project. We spent all this time building a beautiful business plan and a beautiful feature list. Okay, here's what we're going to do with it. Okay, I'm going to give you guys a, um, little thing here that's inside of one of these. Here's the paragraph. Okay, so I will drop this into the chat. Okay, basically what it's going to do is it's going to take— so right now if you look at the attached assets folder in your project. You have this pasted business plan and you have this pasted feature list. Okay, so now we're going to use the attached business plan in attached assets and the attached feature list in the same folder. Create a readme.md file with the business plan and create a progress.md file with the feature list. Don't run 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, an MD file, guys, is kind of a text file. It's called a markdown file. All right. But if you look here in your project, you're gonna find a replit.md file. And if you click on it, this shows, hey, this is what Chorley's about. Here's user preferences, here's system architecture, here's the front end, this is what we're using, here's the backend, here's authentication, this is what we're doing, right? Here's the database schema, here's the API design, right? Replit made this on its own. This is a document, guys, that exists because I've worked with Replit. Over the last 7, 8 months in building out really, really amazing best practices. Teaching the system or having support documents in the system that it can reference as it goes is incredibly powerful. So that's what we're creating here, right? So I'm going to go ahead and have it create these documents. Okay, it's going to create this README file, it's going to create a feature list, and then it's going to add them. Just, they're going to be a additional files that sit here, and then it's going to modify this document, this Replit document, to say, hey, there's also a README document, there's also a progress document. Okay, so then every time you push new chat right here, okay, this little new chat button with the agent, every time you click new chat, it is like hiring a brand new developer, right? If I hired Abby right now to come work for me for GenAIPI, It would take her a few weeks probably to really get up to speed on everything that's been built in the project. She's gonna have to look at all the code and how's it working and all the APIs. And she'd be like, this is garbage over here and I would change this. And she'd have all these ideas and things, right? And that's fine. That's what happens when a new developer comes into a project. They have to get up to speed. But instead of starting from scratch every time, you can say, hey, look, we're starting a new chat. I want you to look at the README, the progress, and the replit.md files, familiarize yourself with the project, and now let's work on this project. And so it says, okay, let me read for a minute. And then it's like, okay, I understand your project. What do you want me to do? You're getting it up to speed super fast instead of having it just randomly do something in a generic way that isn't as applicable to your project. That's all I'm going to say on that for now. I have something else I want to go deeper on, but we're going to, we're going to talk about prompts and system prompts tomorrow. So we'll, we'll do that. We'll do that later. But anyway, guys, you want to do this. All of you need to just copy paste that thing that I did. And what it will do is it will add these files. It should go, should go pretty quick. It's, it's still doing some research on me. Yeah. It's just going slower. It's just going slower than it used to. I don't know why. Um, oh hey, there it is. Oh wait, no, that's, that's just the work cost. Yeah, this work costs $1.44. I'm trying to figure out where the heck the, uh, the usage— there used to be a button that allowed you just quickly go and see how much you've spent on the project, how much you've spent in a chat, and all that type of stuff. But I don't know why they're making it harder to see that. Maybe they don't want you to see how much money you're spending with them. So, um, all right, okay, so it's going to do this. It should not take long. Okay, it's already working on the README file. Plan README. Let's click on it. Yeah, so the README file is done, right? This document right here that it created that looks beautiful now is exactly what ChatGPT created for me. Okay, but now this is all of the why. This is the— this is your bumper lanes of your project. Okay, now it knows the mission and the vision and the problem statement and solution and the core values and the market analysis and the monetization strategy. It now understands, okay, this is what we're going to do. This is why we've created this. And so as it goes to consider something like, like you could say, hey, let's set up the payment system now, please. Refer to the README file so you know, you know what we're trying to do here, right? So you can give it instructions to explicitly go and look at something so that it knows what it's, what it's doing here. Okay. Then it's created this progress file as well, and it's gone in and created this beautiful little checkbox system because I told it to do that in that, in that thing here. Okay. We haven't built the map zone yet. We haven't built notifications, haven't built multi-user support. Import, seasonal templates, AI voice input. We haven't done a lot of these things, okay? But we have created, we have created some features. Kind of all the basics are here. Chore management dashboard, kind of the free version is basically done. All the pro stuff has not been done, okay? Still got to work on some user settings. But now the AI knows, right? This is Replit knows what's going on, what it, what needs to happen, okay? That's the next thing. And then it also should have updated replit.md Let's just see if I can see it here. Yep, here it is. Project documentation. Okay, so it updated this replit.md file and said, hey, there's a business plan here and there's a progress.md file. There's development progress. Okay, this is— I actually am going to talk about the prompt. Replit itself is an application built on top of an LLM, an AI model. Okay, so it says, okay, here's how I want you to act, right? A prompt teaches the AI what it's supposed to do. Okay, whether it's a simple prompt like make me a picture or make me a business plan or whatever, right? Or it could be a really complex prompt that says, here, I want you to, I want you to do all of these things and act like this all the time. Okay, one of the little lines in its prompt is be sure to always refer to the replit.md file. Okay, and make sure you stay within the structure of the project. Okay, so if we add instructions to the replit.md file, then it will apply to the whole project. Okay, so we've given instructions about its progress and about the business plan. Okay. I hope that— does that make sense to you guys? Yep. That makes sense why that matters. What's the MD stand for? Markdown. Markdown. Just a type of file. Just a type of text file, basically. It has special formatting and, you know, but, but again, to say it one more time, you've got Replit. Replit has a prompt. The prompt teaches Replit how to act.

Okay.

Inside of that, it says, hey, I want you to refer to the replit.md file. I want you to refer to this thing. This is your project. And then inside of that, we've given it instructions to pay attention to the business plan and the feature list. Okay. Awesome. So we're just giving it this nice wrapped instruction to go and follow our plan. It's the bumper lanes for the project itself. Okay. All right, guys, I'm gonna stop saying complicated things. For the last couple minutes here. Let you guys ask a couple questions if you want before we finish up. Um, tomorrow is going to be a little bit different than today. We will do some more Replit stuff. I'll answer some questions. I'm sure you guys will be like, what do I do here? I ran into this problem, help with this. And again, you've got my phone number, ask me questions if you want. Another thing, I'm gonna give you another resource right now, um, which I'll give you again tomorrow, but there's a, uh, there's a Facebook group that is for anybody who has taken this course. Okay. And so there's a lot of really helpful people in there and they will help you. They will help you out as well. Right. So if you have questions, I don't want to share. I want to do chat. There we go. Okay. So there's a Facebook group. Go ahead and click on that. Ask to be added. You'll all get in. I'll let you in there. Okay. There's about 100 people in there right now.. And it's pretty useful. It's a pretty useful spot, right? There's a lot of people. Sometimes I'm not available to help, right? But other people that have done this course that are awesome, they've gone and built huge, amazing things, making money. They're building businesses. They're like, oh yeah, do this. And hey, let me jump on a call with you. And you know, they're very, very helpful group of people.

Okay?

But tomorrow's gonna be more, we're gonna, we're gonna kind of rapid fire through a bunch of different AI platforms. We're gonna do some more advanced agenting. Stuff, right? So Replit is really for building software. What about when you want to use an agent to do a big complex task that has nothing to do with software? You want it to go do some research or you set it up to go call some people and do some things. And, you know, there's some really amazing systems that we'll go over tomorrow for that. We'll do image generation tomorrow. We'll do video generation tomorrow. We'll learn how to pull those things into Replit. We'll learn some more advanced things around sales and marketing, how to tie in AI to automate some of that stuff, all at very high levels, but enough to make you dangerous to where you'd be able to say, okay, I remember what Jon was talking about. Let me, I'm going to go tackle this thing and AI should be able to guide you through it. So anyway, any questions before as we finish up today, guys?

There's just a couple, Jon.

If we need to build our own custom API integration, do you, Is it just as simple to plug it in? API for what? Let's say document drafting, like you give it a database of documents and you wanted to customize the document based on past documents with new information. Yeah. So yeah, absolutely. And you can build all of that in Replit, right? You're just going to explain to it, okay, I've got, I've got documents over here and you have to put it somewhere, right? It's going to be in a Box.com or Google Drive, whatever it is, whatever secure system you use, right? If you have an existing, you know, legal document, you know, repository, they might have an API and they might not, right? If they don't, then we have to work through that problem. We have to figure that out. Um, but everything you're talking about is very, is very doable, right? You just have to figure out how to build it. But yes, Repliq can build, can help you build your own APIs for you know, for your specific use cases. Cool. In between now and tomorrow, anything you want us to focus on specifically? And can we access the video between now and then? Thinking of my brother who wants to catch up. Yeah, definitely. So this is going to end in a couple of minutes here, and then it takes quite a while to process a 4-hour video, and then I have to upload that to YouTube a little bit. So I'm going to say somewhere around an hour No, let's see. I actually have to run to the doctor right now. I have a doctor appointment that's in 30 minutes. It's about 20 minutes away. So I have to leave like right after this. But as soon as I'm done with that, my next meeting's at 3. So before 3 o'clock, I will push publish on it. And by probably, let's say, 3:30 or 4 o'clock, you guys, or it'll be up. And what I'll do is I'll send an email to everybody, right? So that same kind of pre-course email thread we had. Had going. I'll just reply all to that and make sure you guys, you guys get on there. But you'll have it by 4:00 my time. So about 3 hours from now it should be up and running.

Jon, will the email also have the Google Drive access that you talked about?

I normally don't send that till tomorrow, but I'll send it today. That's fine. I'll do it. I'll drop it. Yeah, you'll get that. You'll get all that stuff. Normally I send the Facebook group and Google Drive and all that stuff, but I kind of give it— in fact, I can just give you the Google Drive right now while we're just on. Um, yeah, there's— you can get ahead if you want a little bit, but I'm not, I'm not worried about it. And this is nothing, nothing crazy in here, but there's a Google Drive with all the, all the different guides and things and, and some of the stuff we'll be going over tomorrow. But tomorrow's fun, guys. Tomorrow I think is, is more fun. Today is a lot of like your brain's overloading a little bit. You're just like, man, this is a lot. Like, I can do it, but it's a lot, right? I want you today to just get comfortable with it and try to dispel any thoughts you have in your mind of like, I'm not smart enough for this, or this is too technical for me. It's not. The language of AI is English or Spanish or Morse code or whatever the heck you want to use. If you speak Elvish, you can do that too because it understands it, right? And it can interpret that. You just have to tell it what you want like a human. Okay, if you get lost, Josh, in some weird loop like you did, right? Click the plan button and say, hey, you know what? Look, just let's just plan for a minute. Tell me what's going wrong. You're getting lost. Tell me what's going on. Just almost the same way you would talk to a developer that's confused, right? Like, hey, let's just take a step back. Give me a list. Show me what's happening here. Great. Okay, so now that you understand this, what do you think we could do to fix it? That's great. I agree. Go do the plan, right? Just work through it. Talk, talk to it like a human. Be kind. Use please, use thank you. Actually, Sam Altman would tell you not to do that because it costs a lot of money. But anyway, that's another fun topic. Okay, guys, we were right at 1 o'clock. I'll let you guys go on time. Thank you so much. I hope this was a fun first day. Tomorrow is going to be great. We're going to wrap it up and pull lots of things together and, and And, you know, you get bonus credits if you have your first customer by tomorrow.

Thanks, Jon. Take care. Good luck.

Thanks.