Cohort 22 - Session 1 - 10-22-2025

this computer. Hey, all right, well, welcome to day 1 of this how to build a business course, and we might do a little bit of AI stuff as well. And, uh, this is, uh, it's gonna be fun. I'm excited to, uh, to get to know each of you. I only know Kalima. I met Kalima yesterday, and, and he's telling me about some of the stuff he's up to, which is very impressive, and we'll have— he'll have a chance to share that a little bit. And I said, dude, you got to come, you got to come join this. And so And this is going to be awesome. It's going to be really fun, guys. So to get started, guys, I like to just kind of go around, do some intros. I want to get to know who you are, where you're from, and then I'll go last, and then we'll kind of pick up from there. But yeah, I said in the beginning, this is kind of a business class with a little bit of AI, and I really mean that, right? We're going to be using AI the whole time. We really will. But there's so many courses out there right now that'll teach you how to use AI, but then they don't think— and you have to do it from the beginning— they don't think about, okay, what am I actually going to do with this when I built something, right? How do I actually sell it? How do I actually get customers? How do I actually, you know, get, get real value out of this? Uh, AI has completely changed the game in terms of software development, entrepreneurship, startups, investment, you know, all kinds of things related to to this world, and we'll talk about all of that stuff. But really, you know, I'm an entrepreneur. That's what I've done, you know, for the better part of my career, and I love it. And I really, really respect anybody else who considers himself to be an entrepreneur. It's the hardest profession there is. It's the most exciting one also. And yeah, so let's— I'm excited to be in the trenches with you guys. And it looks like it is all guys today. Normally we've got a gal or two in here, but, um, today is going to be a bros club, so that'll be fun. Um, so that's good. But anyway, um, so let's start. So I want to hear your, your name, um, where you're from, um, meaning where you live now, right? If you want to add like originally from Houston, Texas, that's fine, you know, but, uh, but currently wherever you are. And then what do you do for work? Or, and, um, whether or not you know how to code. I'd love to know that, just yes or no, whether or not you know how to code. So, um, and everybody gets, I don't know, 38 seconds or so. So I'll just go in the order that I'm seeing it here. Cody, Justin, Morton, Jacob, Kalima. Okay, so Cody.

Um, my name is Cody Cokerhans. I am from, uh, Utah. I live in Lehi, so I'm by Jon somewhere. So congratulations on the BYU win. I was at the game. That was wild. I'm a Utah— University of Utah—

I'm a Ute.

So, um, anyway, but I, I respect greatness where you can find it.

You know what, Utah was freaking good, man. The fact that they legitimately had basically 4 turnovers, right, because turnovers are downs if you count those, and only lost by 3— like, you guys are good.

Utah, they, they, they, uh, I mean, it's a typical Utah curse. A couple weird calls and then can't convert. The office doesn't show up when you really need it.

Well, I'm grateful that we were on the other side of that curse, cuz we were there for about a decade as you know.

Yeah.

Yeah. Yep.

But it was, it was cool. I was over there with family and I had a good time.

Good stuff.

For work, right currently I am not working. But I have been working for, in finance. So I think Justin, you're in that world. I worked at Zions Bank and Mountain America doing sales enablement. And trying to bust sales operations into regional banks and credit unions, which is a very difficult task. So I've been spending 15 years doing that with various degrees of success, and it's been really rewarding and frustrating at the same time. But do I code? No. But I've been following you, Jon, for a long while and actually watched some of your videos too. Take to, to actually apply for some jobs. So I did, you did a Gamma video, you were comparing all these different slide decks and stuff like that. And I actually used a lot of your tools to, to create some stuff that, that would require me to code. So I created a full-on objection management app based on videos that I watched you do.

Yeah. Very cool. Did you use Replit? Did you use Replit to do it?

Yes, I did.

And I did it over an evening.

Yeah.

And I had 4 different scenarios that I would actually talk to and it would talk back. It took me the longest part of that whole process. Was creating a realistic voice to respond back to me. I didn't know how to like talk to AI to like, hey, I need it to sound more realistic. And then like the type of level of education they needed to speak at. So I learned a whole bunch of stuff actually just trying to like, how do I get what I'm wanting out of that app? So that's my experience so far with—

Fantastic. You're going to go— we're going to go deeper. Like those videos that I've done are just, you know, a taste of things that I cover in here. Obviously, we're going to go, you know, You're just seeing the tip of the iceberg there. We can go deep. We can go deep. That's awesome.

Cool.

Very cool, Cody. What part of Lehigh are you in?

Right next to Primary Children's.

Okay. Northwest. Yep. Yeah, I was in Travers Mountain for 5 years prior to where I am now, but—

Oh, cool.

Very cool. All right, Justin. Hey, good morning, all.

Justin Markey. So I live in Wilmington, North Carolina. So on the other side of the coast near the beach. I just moved here about a year ago. Prior to that, I was in Charlotte, North Carolina for about 13, 14 years. And if you don't know the difference It's about 4 hours apart. That's all the way to the west side of the state. And then before that, I was originally from New Jersey, New York area. For career, I've traditionally been in financial services. Most recently was at Citi National Bank as the head of change management. I've transitioned out of that role. Part of that impetus to transition out is the lack of wanting to get into AI space when there's low-hanging fruit to leverage that. If you've never been in financial services, it's moving mountains to get things done. I'm sure in other industries it is as well. Jon, I came across your post about the, the actress, that the AI actress really spawned additional interest. Prior to joining this, I had actually done kind of, I'll call it half a boot camp, trying to learn a lot more from the Triple Ten sessions, if you're familiar with that. However, seeing that that was a little more remedial than I really expected because I'm trying to leverage this from both creating a business perspective as well as if I do go back to corporate, I want to be the Tony Stark. So I just give me all the tools and I can put it all together and let's fix shit in an easier way.

That's it. That's fantastic. So Charlotte is one place where I try to get to at least once a year, sometimes twice, because I'm a big-time whitewater kayaker and the U.S. National Whitewater Center.

Oh, you've been there?

Yeah, it's great.

Yeah, many times. Yeah, that is, that is like Disneyland to me. Um, yeah, for those of you don't know, we're, we're taking just like a 30-second detour on this and we're going to show you what it looks like. It's so cool. Whitewater.org, you know, it's gotten cooler every year, which is great. No, it keeps getting better for sure. But yeah, it's— let's see if I can find— let me just do this. USNWC map. Basically, it's this Disneyland. Yeah, this is actually the whole thing here. It's massive, massive, massive. But you've got this huge river system right here and then all these mountain biking trails. It covers, you know, hundreds and hundreds of acres. But, but yeah, it's all— yeah, we'll go to the Whitewater Center here. There's a— there's like an amazing picture somewhere here that's going to show like what it actually looks like. This is just like water. I don't want to see water. I want to see like actually the kayaking and stuff. But anyway, it's man-made water. You can see that there. Yeah, there you go. There's a little bit there. But it's very, very cool if you want to take your family or something on a cool, cool trip. Go there. Go there. So I love North Carolina. I'm a big fan. So anyway, all right. I think next is— I don't know who I said next, Jacob or Morton, but we'll go— we'll go Jacob because that's the next one on my list at this point. I get the list is changing.

So sounds good.

I'm Jacob Boyd. I'm from a little town in Florida called Bunnell. So we're kind of by Daytona Beach area.

Cool.

I right now, I— so me and my wife, we own two different businesses. One is like a land clearing business and the other one, we live on a cattle ranch and so we sell beef. So we like have a butcher. We take the cattle to the butcher, sell them. And so we sell beef as well. So those are the two things. And then also I went to school for marketing, and so I'm kind of backing out of the land clearing a little bit and trying to get more into marketing. And I was just kind of— the way I heard about this is I was, uh, honestly just kind of sitting praying one day about what I should do because I don't really want to do the land clearing that much anymore. And, uh, one of my buddies, Brantley Brumley, he, he ended up, uh, he popped up on my feed on Facebook, and he was saying something about AI. So I just messaged him, and he was like, you gotta do this class, and sent me all the stuff.

So I was like, all right, that's good.

Uh, but yeah, so, um, no coding experience, no AI experience, nothing really. Um, yeah, so just wanting to kind of get into something that I can— more what I'm trying to do is get into something I can do from home so that I can spend more time with my kids while they're young. That's the main, main thing that I wanted to do.

So how many kids? What ages are they? 3.

My oldest is 11. Then we have a 5-year-old and a 3-year-old.

Yeah, man, that's the age you want to hang out with them. I'm right there with you. I'm my minor. I have 4 kids, really close ages, 8 to 13. And yeah, they're just awesome at these ages. So that's awesome. Well, very cool. Yeah, I've been out to Daytona. Um, it's cool. It's a cool spot. I got to go there with a, with a NASCAR driver and, uh, I was like down on the, in the pit, like with them and everything. It's really fun. Kind of a cool experience. So anyway, good stuff. Um, good to have you, Jacob. And yeah, Brantley's, Brantley's amazing. Brantley is, was my first hire for this company, for Jenny IPI.

Oh really?

Wow. Yeah. He, he works for me now and he's, I would, I wouldn't even say he works for me. He works with me. He's awesome. He's really good. Yeah. All right, Morten. Yeah.

Hey guys, I'm Morten and I am from Denmark, Copenhagen, Denmark. That's on the other side of the pond in Europe. And yeah, I lived there recently or currently. I actually, I lived in Denver, Colorado for 3 years. My wife is from a smaller town called Greeley, Colorado, if anyone knows that. So we lived there from 2008 to 2011. So that was, yeah, fun and I'm sort of visiting every year, of course. So currently I'm working as a sales director for a, like a scale-up company that we actually also deal with AI, but just in a different way. So we make AI CAD software. So that's computer animated design software. So an example is that you can actually take your smartphone and you can scan the ear if you're an audiologist. And then based on the scan, our software can output a, let's say, an earplug device or a hearing aid that fits exactly in your ear. We also make dental products that fit, you know, if you need a night guard or splint or a new crown or whatever, we can do that too based on AI. I do not know how to code, so I'm working with all these smart engineers, but I have no idea what they're putting up on their screen every day.

So many questions. I have so many questions. Your app, your app that scans the ear.

Yes.

Using— is it using the front camera, the dot matrix on the front camera, or what is it using?

We, we, we want to use the back camera because that's much easier so that they can actually see, you know, the idea is that even like the, the patient can do it themselves, or if you're at a concert, you want— we want actually people to be able to scan and then basically print an earplug right there.

Just right there. Yeah, that's cool.

Yeah, I mean, that's obviously taking it, you know, that's the vision. But yeah, so it's super difficult apparently to do the back camera and that's what the engineers are working on right now. But yes, we can also use the front camera. That's just a bit more tricky and it requires another person to be scanning your ear kind of. Yeah.

So the front camera has And I'm sure your engineers know this already. So I've made a whole career doing this stuff. Exactly. Okay. Yeah. Before, before AI, I was in AR and I was in augmented reality. And so I was all about 3D models and 3D visualizations and all kinds of stuff. Built a $100 million company, right? So I'll talk about that more later. But the point is, we did a lot with the dot matrix camera, the same thing that scans the face for Face ID and stuff like that. Okay, that matrix is actually a lot more fine details. And so I don't know if they're making it— so don't use the camera itself, use the dot matrix functionality, and you actually can get way faster, better results that are more fine-tuned. So I don't know, they probably already know that. If they don't, well, they'd be saying, yeah, have you tried this?

So yeah, that's, that's good. I appreciate that, Jon. I'll let them know, and then I sound really clever.

Yeah, then you'll be like, dude, like, what, are you using the stuff that use the Face ID anyway?

It's called.matrix.

Anyway, I am not a technical person, but I know technical stuff. I love technical stuff. It's the technology. I just love it. So, but again, I'll talk about me in a second. Kalima, last but not least, and with the coolest background for sure.

Thank you. I tore the insulation out of my basement, so it looks a little more professional. My name is Kalima. I'm originally from Honolulu, Hawaii. Currently living in Utah in a small little town called Eagle Mountain, which is far away and that makes it cheaper. So that's why we're here. I previously worked in all kinds of operations at Tesla. I worked in finance ops, I've done sales ops, marketing ops. Really enjoyed it, but have kind of made a shift recently into becoming a social media personality. So I make content on Instagram about ancestry and genealogy that has like blossomed in the last 6 months to 200,000 followers. Um, so part of why I'm here is to try and figure out how I can provide it for more people, provide the services that I, I make content on. I cannot code, uh, but I have abuse Claude to no end to try and make things work. And it has worked so far. Um, I don't know, sometimes it takes 2 or 3 iterations, but, um, yeah, not, not coding in the typical way, but definitely have tried to use AI to do it.

I love it. I love it. Yeah. Vibe coding, vibe coding is, um, it counts. It counts. It works. You can build stuff that where you push the button and it does what you want it to do. Right. We'll have some fun discussions about that. And well, that's awesome. Yeah. Kalimah told me about everything he's doing. He showed me— you guys got to look up his Instagram channel. How many— how many views do you have over that over the last 6 months?

Let me check. It'll give me Windows. I think it's just over 20 million.

Yeah, that's crazy. So 20 million views over the last 6 months, 200,000 followers from basically nothing. Right. Brand new channel. Doing— he like dresses up in like chainmail. Do you have that there with you? You gotta show the class. Hoping I wouldn't have to bring it up, but yeah, he like talks about it and like takes you back, like your ancestry back to, you know, Denmark where most of us came from. Um, you know, that's the old world you live in. You live in a cool spot, Morten. But, um, but yeah, he's— Kalima has— I think he told me yesterday 40,000 to 50,000 DMs.

Yeah.

And he has no offer for them. He has— we like— so, so we're gonna build a business today for Kalima. It's gonna be awesome. Um, you guys will build your own things, right? But, um, well, very good. Well, so again, I'm Jon Cheney, and I, um, I'm an entrepreneur. I'm a whitewater kayaker. I love long walks on the beach. Um, no, I love whitewater kayaking. I'm also a musician, if you can't tell. I've got 3 pianos in this room alone. 1, 2, 3 guitars right there and a rain stick and a bunch of microphones and other things. And, and I've got my cool little studio here. So, so anyway, and I live on a farm, 5 acres. I live out here in West Mountain, just on the south side of Utah Lake, which is just kind of head down to Payson. And instead of going into Payson, you turn west and come up on the mountain here. And I've just got, I've got a little 5-acre ranch that— so I sold my company back in 2023 and and my wife said, let's move out a little bit. And so, um, so we moved out, just kind of a little bit quieter place. I have a tractor, which, you know, makes me feel kind of cool. But, um, I don't have any cows yet, but I, I do plan on building, like, fencing off my north kind of acre and a half here, and we're gonna get, we're gonna get maybe, maybe one or two. I hear they like friends, so we'll grab, we'll probably grab two, I think. Yeah, anyway, so, uh, so anyway, I love, love, love that. But My whole career, I've been, I've been an entrepreneur. I've worked, I graduated from BYU and then worked for another startup for a bit. And then for about 3 years, we got acquired. I was there for about a year, year and a half, maybe, and I left to do some of my own stuff. And in 2016, I founded a company that was my first kind of big, biggish company, raised $13 million for that, took it up to $130 million valuation was our peak. We ended up selling closer to about $40 million. Right? So, nice, right? But didn't go all the way down to zero, which is really fun. Good. It's really, you know, entrepreneurship is hard, right? It's just, you know, just like a stock, right? It goes up and down. Gold goes up and down. Bitcoin goes up and down. Companies go up and down. Startups go up and down. And most startups fail. They really do. And the reason why a lot of startups fail is because they run out of one of two resources: time or money. That's basically it. There aren't many other reasons why a startup fails. They just, they ran out of the time that their, that their money, that their runway had. They, they, they run out of money because maybe they never even raised it. They weren't able to do it. They weren't able to get it off the ground. They weren't able to raise money to build software. They weren't able to do what, what they wanted. And, um, and I believe that AI dramatically reduces the need for either of those resources. Um, I believe that AI levels the playing field in a way that is world-changing, economy-changing. It used to take— let me give an example of Jenny API, and I'll show you. You know, you guys gone on my website to sign up for the course, and you've felt it, whatever. And it's a good website, it works, right? You know, credit card process happened. Could it be better? Probably, sure. But it works, it does, it does the job. I'm able to sell, I'm able to do what I need to do. And, uh, And it's growing extremely quickly. But what's cool is I vibe-coded that whole thing, right? I built the whole thing in just a weekend. And I started Thursday morning, took me about 35 hours to get it all done. And by the time I was done, I was ready to launch. And a couple of days later, I had my first $15,000 contract, right? And then 6 weeks later, I had $200,000 of revenue. And it just kept growing, right? And at this point, So, I started Jenny IPI officially March 1st, the day I launched it. February 27th, I started. February 28th, March 1st, boom. And I've now raised money at a $10 million valuation. Here we are just 6 months later, right? And it's going to be a $50 million valuation by probably March. And I'll share with you guys why. There's some really crazy cool stuff happening. But the bottom line is It took me 3 days. I wrote over 100,000 lines of code as a non-developer, and it cost me maybe $130 of credits, right? AI credits, right? Almost anyone— not almost anyone— anyone can raise $130. If you can't raise $130, sign up for DoorDash and go deliver some food for like 2 days. Boom, you've got your seed funding., right? Anyone can do it, right? Anybody watch Ryan Trahan? Yeah, a couple of you guys. I don't know. So, Ryan Trahan does all these challenges where he tries to like go across the United States and he has to like reset every day and he has no money. And so, one of the ways he makes money and he has to make like a couple hundred bucks a day and he figures out tons of different ways to do it. But one of the things he does, he just does DoorDash in between all the things he's doing. And so, he just makes money. It's very easy to make enough money to start a business now. Now, you have to have an idea. You gotta execute on it well. You have to know how to sell. You have to do a lot of other things too, but AI can fill in those gaps. And that's really, really exciting. Entrepreneurship used to require, so for Seek, my company that I started back in 2016 and raised $13 million for, I had to raise money because I wanted to start something that was technology related. It was augmented reality, right? It was right when Pokémon GO came out and I wanted to build something that was kind of like Pokémon GO, location-based AR. With a map just like Pokémon Go, all that stuff. You go around, you'd open up a treasure chest, and it would have, you know, random stuff on it, randomized stuff in it. And you'd get like things, you'd get discounts or free stuff. And, you know, we had customers like Larry H. Miller and Swig and Goal Zero and Universal Pictures and Samsung and Six Flags. And, you know, we did millions of dollars of revenue having this thing where it was just kind of like a marketing spin on Pokémon Go. And it was a lot of fun, a lot of fun. But man, in order for me to to even prove that that business model was viable, I had to go raise $500,000 of other people's money and put it at risk and take on that stress and that responsibility and all that stuff that comes along with raising money. It's not easy, right? And when you take other people's money, if they don't like what you do, they might sue you or they might just be angry and cause lots of problems for you or You know, and I'm not saying that it was, that it was that bad. I'm just saying those are risks. Those are real risks that you take on when you take other people's money. Then I took $13 million, you know, over the years, you know, $2 million here, $1 million there, $5 million there. And, you know, those risks and everything just continue to go up. Right. How many, how many startups have you guys seen raise $10 million, $20 million, $50 million, $1 billion and fail? Happens all the time, right? It's a risky business. Startups are tough. AI allows you to tackle it in a way that, um, that is just basically— I, I want to say kind of like fast fashion, right? Where it's just like, yeah, it's cheap, it's whatever. Like, I just, I just need it to work, and let's see if it works. If it does, great. If it doesn't, oh well, you know, I've lost a couple hundred bucks, um, no big deal. And so you can iterate, you can test, you can pivot, you can say, hey, you know what, I don't need to do this anymore, I'm gonna go do this thing. And so that's fun. You can stay in control. You can bootstrap way better. So, all right, guys, we're going to kick this off today. We're going to go through over the next 2 days. We're going to go through probably 15 to 20 different AI platforms. Obviously, we won't be able to go deep into all of them. Some of them will just be like, hey, look what you can do. I want you to be aware of this. Watch what I do. Do this. And then some of them, I'm going to be like, all right, let's dig deep. Let's work hard. Let's, you know, pull out your fingers and let's type. Okay, so, um, you— if you got my email last night, you should have— you should have a Replit account, you should have a ChatGPT paid account. If you don't have a paid account, you're not going to be able to do some of the things we're going to be doing. And then Grok is, you know, if you have it, great. If you don't have it, it's not the end of the world. Um, but, but I like using multiple AIs, partly because I don't trust AI, right? And you have to understand where you trust it, and you have to be able to bounce things. I like bouncing things off of different AIs to say, hey, What do you think? What do you think about this? And then taking that and going back and forth. And I also really like Grok for market research, right? We're going to do fast market research, right? Because Grok is the only AI that has access to X, right? Which is where the internet speaks about everything, right? So you can do like really legitimate market research on your topic very, very quickly, right? You want to know about what people think about buying beef? You can do that. You want to know what they think about, you know, whitewater kayaking? Fantastic. You want to know what they think about genealogy? Great. It's all there. People are talking about it all over X, and, uh, and Grok can access it very quickly, analyze, see what, see what the market's talking about, what kind of opportunities there are in there. So anyway, um, all right, we're going to start today, guys, with, um, just, just watch my screen, I guess, for a minute here. Sometimes I'm going to say, hey, just watch, and then sometimes I'll say, let's do. Okay, so we're going to start today. I'm just going to bounce over to X, and I'm going to jump into Groq. Okay, so, um, some of you might have tried some of these things, and that's cool. If you have, great, this will be easy for you. If you haven't, hey, welcome to this. By the way, I cannot code. I cannot write a single line of code. I am a business guy, an idea guy, a strategy guy, a sales guy. I, I know how to do all these cool business things, and, you know, now I'm an AI guy. I've been doing AI for a long time now, and, and it's, and it's really, really fun, and I've taught thousands and thousands of people how to do these things. But, uh, but it's cool. I'm teaching people how to code and I'm not a coder. That's— I mean, we're in a whole new world here. Um, so what we're gonna do here is we're gonna say, all right, we're gonna use a language this morning called p5.js. This is, this is kind of just a fun kickoff exercise here. So using p5.js, uh, create me a mind-blowing visualization. All right, let's just see what this says. The request asks for a mind-blowing visualization. All right, so one thing that's kind of interesting is AI— while it's thinking for a second here, and you know what, I don't actually want it to use expert, we're going to switch to fast. Um, but one of the cool things it does now, and you know what, I actually want to Kind of forgot one little part of that request, which is no big deal. Format this so that I can just use the p5.js editor website and give me a link to that site. Okay. One of the things that AI can do now is it can— it understands what you mean. It used to be that prompting was super important and it just isn't anymore, right? It's mildly important. It used to be everything, but all of the AIs, right? The ChatGPTs, the Groks, the, you know, Manus and Replit and all these different systems that are out there, Anthropic and Perplexity, they really want to figure out what you mean, right? And so they figure out what you mean and then they structure their own prompt for you. So here's what I did. I basically said, okay, give me a mind-blowing visualization using p5.js. Show me where I can on it. It shows me a link here. I'm going to copy all this. I'm going to paste it. I'm going to delete all that. There we go. Okay, so here's my cool visualization. All right, I built something. I built code that works, that does something. Okay, I don't know if it actually is interactive. If I click— nope, no interactivity, nothing like that. Now, I mean, it's kind of cool. It's kind of like a screensaver, like, like that's a Windows 93 screensaver or something like that. This is actually what I would call a bad prompt, right? Why is this a bad prompt? It's not that it's structured bad, right? It's, it's clear. It knows what— it knew what to do. It did it. It accomplished the task. But here's why it's bad. It has almost no details. I, I left way too many gaps for the AI to fill in. So if all of you were to take this exact same prompt we would all get different, different answers. Right. And so because of that, you're going to get inconsistent results. And inconsistency doesn't work well. Right. If you've ever been to a golf course, you know what I mean? Okay. You hit it one time and it slices and you hit another time and you hook and you hit another time and you top it. Right. If you, if you slice every single time and you know you're going to slice every time, you can go win the PGA tournament because you're predictable. Right. You know exactly what to do. You can adjust, you can figure that out, right? And so predictability is critical. It's something that we want to be able to do. And so we need to learn how to talk to AI in a way that's going to be very predictable. So let's give it a little bit more details. Let's, let's have it not fill in quite so many gaps for us. So I'm going to say, okay, now create a, an interactive No, we're not going to do— yeah, well, interactive is fine. Interactive 2D animated model of the solar system. Yeah, we're just going to do that. Now, now there's two things that I want to point out here. One is I did not have to ask it again to do p5.js. It understands the context of my conversation. It already gets it right. It remembers what we're talking about. Oh, okay. Now he wants this. And so your employee, your little employee here is remembering, okay, yeah, I understand what you want. Now, the solar system, if I were to give this to a developer and just say, hey, I want you to create, you know, a model of the solar system, they'd have to go do a bunch of research, right? They have to figure out, okay, what is all this stuff, right? How, what's the orbital period and how big is it relative to all these other things, right? And I'd have to do a lot of research to figure that out. Luckily, AI knows everything about everything. When you ask it to do this, it actually has quite a bit of information. While this might seem like about the same length as the prompt I gave it above, it's actually pretty good. Now we're interactive. Now I didn't tell it how I wanted it to be interactive, so I left some blanks there. We'll see what it figured out. Animated, I also didn't tell it exactly how to be animated, but it's pretty obvious that it's probably going to just create some orbits around the sun. Let's see what happened. Okay, we'll go copy, we'll come over to p5.js here and we'll replace it. And hey, we have a working model of animated model of the solar system. Okay, pretty cool. Now it says scroll to zoom and press L to toggle labels, right? So I can, I can zoom in and out here. I can push L. Maybe. There we go. Yep. There we go. So I've now created You know, I have created fire, right? All I had to do was tell it what I, what I needed it to do. Okay. Pretty cool. So let's do one more little iteration here, and then you guys get to do, do your thing, and we'll have each of you do something and, and share your screens. Alright. So add a starry background with twinkling stars. Have the background rotating slowly so it brings the scene to life. Add a spaceship that I can fly around with my WASD keys, and if I hit the spacebar, it shoots. Yeah, it shoots. Have waves of Enemy spaceships, enemy spaceships flying in trying to get to the sun. If they hit the sun, I lose. Use R to restart. I just gave it some random instructions. I'm just streaming my consciousness here., with, with some, some ideas. But now I'm creating— I'm turning this, this model into a game, right? Um, it used to take years, you know, when video games were first coming out, right? When you were creating— and by the way, these types of— you're not using this, this graphics engine for, uh, you know, making Call of Duty, right? This is not Unreal Engine. This is more like Atari and, you know, Nintendo maybe, and even some things like that, right? You could say, hey, you know what, let's create a clone of Tetris or something like that. And it can do, it can do that kind of thing very, very easily. All right. So it's a little bit longer. Got a few more things. By the way, when you're copying, it's really easy to just use this copy button, right? Instead of going and trying to select it all. You do have to wait until it finishes thinking and finishes writing everything. Okay. All right. So I'm done now. I'm gonna push copy. Come back over to p5.js here, paste and play. All right, here we go. Oh, I died. It's kind of hard. First of all, it's not— it's not— oh, I have to use the mouse. Here we go. Here we go. Here we go. I'm going to die anyway. I got to keep working on that. So it's a little bit hard. But anyway, I created something that's pretty cool, right? You got the starry background. It's rotating. It kind of brings it to life. I can drive around. I use my mouse to aim. You know, pretty neat, pretty neat stuff. Okay, so I can obviously keep adjusting and talking to it, but now it's your turn. Okay, so go ahead, open Grok. You can do this in ChatGPT if you want as well, if you don't have Grok. But I want your first prompt to be basically copy. I'm going to give you guys these in the chat here. Okay. And it's going to— I'm also going to include this. Okay, so I want you to replace what I just sent you, create me a mind-blowing visualization, create me a something, right? And come up with an idea. It could be some fun game, it could be, you know, dinosaurs chasing someone, or, you know, a Flappy Bird clone, whatever you want to do, right? And let's see what you got. Okay, so take, take just a few minutes here, and as soon as you have something something up and working, then let me know and we'll— I want each of you to share your screens. I'm going to sit here and see if I can beat a level here. And by the way, guys, feel free to ask questions while you're working on this. This is an interactive course. If you don't know what to do, speak up.

I got my, I got, I got my first attempt. And it's actually pretty hilarious.

But all right, let's see it.

Okay, uh, it's in the action right now, so enjoy, folks. This is supposed to be a University of Utah player trying to tackle the BYU quarterback in front of thousands of crazy fans.

Uh, can you guys see it? Okay, yep, we're there now. All right, I love it.

Right here. So I didn't put any real descriptions. I mean, it picked out that BYU is blue, Utah is red, but yep. Pretty, pretty generic stuff. I guess the red dots are crazed fans.

But, uh, anyway, I love how it just used—

it's supposed to jump over that Utah player at some point.

Let's go Bear! Something's happening with the red player.

Yeah, there's, there's my attempt. I can try to fine-tune it while everyone else is working.

Yeah, you can obviously, yeah, you can obviously like say, okay, let's make it look a little better, make it faster.

It took, yeah, faster for sure, whatever, more realistic.

Hey Jon, when you get to the editor, is there anything, do you just Select all, delete, delete the stuff that was there, and then just paste, you know, what you get from, from your AI.

Okay, I think my one's working. Let me see if I can share my screen.

Okay, share away.

Uh, where are we? So I asked it to make a game where you have to scan an ear, um, and there's a loading bar with, uh, 10 seconds, and the— these asteroids are trying to hit you. So when you move away from the ear, it stops loading. You got to kind of stay in the middle, and then if you get hit by one, then it's game over. That's funny. It's kind of an interesting ear.

There you go, Martin, you've got— you've got a new addition.

I know, I'm gonna I'm going to show that to the developers tomorrow.

I love it. That's great. And you guys, these are perfect, right? I mean, look, they're not supposed to be pretty right now. That's not the purpose of this section. It's to be fun, to just be like, hey, I can make something that works and does something. And, you know, some of you have obviously had experience doing even more than this, and that's great. But this is actually super fun to do with your kids, right? Show them how to do this. Let them sit down, create games, tell them, hey, you know, just talk to it and it'll change it, it'll make it better.

And anyway, I'm getting a lot of these error messages, Jon, like this syntax error, cannot declare a constant variable twice.

That happens. Yeah, that happens. So, so what you, what you want to do when an error comes up is you want to, um, you want to copy the error, just copy exactly what it says and just give it to the AI and it'll be like, oh, okay, fix that. And it'll identify what it is, whatever. You know, the more information you can give it, the better, obviously. Got it.

Yeah.

Anyway.

Okay.

All right.

I got mine if you want to look at it.

Yeah, let's see it.

Share away. This is 10 people playing a game of telephone and laughing at the end.

They're not people. Let's see what we got. There you go, messages bouncing back and forth. That's awesome. Now guys, you, you might not be super surprised if a lot of video games actually start like this. Right? Today you can actually move forward a lot faster, but a lot of times the very first early versions of like Call of Duty or something like that, right? It's just stick figures or blocks running around, right? It's usually like a cube or a block or something that's just moving around. And then later they say, okay, now we're going to replace this block with this highly, you know, engineered, you know, model of a human or model of a whatever, right? And, um, and then they add the animations and they add all that and it just takes time to iterate, right? So a lot of times games really do start looking just like this. Minutes, right? And then you just go from there. But anyway, so that's cool. Um, but the fact that you can do it with just, you know, 4 minutes and no experience is, is what's crazy, right? To be able to model something, some sort of behavior, some mechanic that does something properly. Pretty cool. All right, all right, who hasn't gone yet? Jacob, how's yours coming?

It's going.

One sec.

I was like, I, I kept getting the error too, so I finally just got it to quit erroring, and I'm trying to—

I'll share my screen.

Okay. So I know I asked it to—

it'll let me—

well, it's not coming up now. One sec. Looks like almost like the interface was gone.

It's there, but it's not letting me share the screen.

One second. Interesting. This exercise is also a secret, um, kind of exercise to get you guys to make sure you know how to share your screen properly, because we'll do that, we'll do that later some more with other stuff. All right, what do we got here? Oh, you're herding cows.

Indoor cows into a cow pen. So basically it's just press, then you just move.

So just click, click on the green. Yeah, there you go.

Yep, little by little until—

oh, just as you get close, you just kind of have to— oh, that's cool.

And they move a little bit.

You got to add a dog that can—

yeah, yeah, I got to have a dog somewhere. Anyways, yeah, so it's just little—

that's actually kind of cool. So what happens— so when you get them in, it—

the counter goes up Yep, counter goes up, which it's got.

Yeah, I can see, I could see it there up in the top left. But yeah, um, that's pretty cool actually. Yeah, wait for it. One.

All right, what's up? Every time you get one in. So I don't know, I thought it'd be—

I think it's cool. I think it's— I think that's a fun one. That's a really clever one. I love how you guys all take all the things that you like and turn it into this. I think that's awesome. Thanks for sharing. Um, all right. Uh, Morten, last but not least, right?

Yeah, so I'm still, I'm still working on it. Um, it won't come up, but I can try and share my screen here. Um, we can work on it with you. Yeah, maybe, uh, that's, uh, usually I use Teams, but I think you, you can see my screen now, right? Yep, you can see it. Yeah, yeah. So I, um, I was basically trying to create like the sun that was sort of hitting the Earth while the Earth was rotating, and there was like planes and rockets on the Earth. Maybe it was too ambitious. But yeah, so this is what I see right now.

Yeah, so I got this. It looks like it might be trying to do some crazy stuff with images there.

Do you see that? Okay, yeah. But that's fine. Should I go back into Groq and maybe just—

No, that's cool.

Grab the—

Yeah, so it looks— Yeah, grab the error that it's showing there down there in the console. Grab all those errors. Oops.

There we go.

Yeah. So all of these, the whole thing. Yeah. Just grab all of that.

Okay.

Yep.

Like that. Yep. And then back to, back to Groq.

Okay.

Back to Groq.

Yeah.

And then just paste it in there. Just paste, enter. Okay. Okay. Yep. And then let's see what it says here. Okay. Fix copy and paste ready sketch error called here. Replace this.

All right.

Cool. Should I just copy it now? No, not yet. Wait a little bit.

Yes. No, scroll down. You kind of got to wait until it's done writing. Yeah. The way that AI works, guys, is it's, you know, at a very simple level, it's estimating which word it should write next. And it does that all the way until it's done writing all the words. Right. And so you just kind of have to wait until it finishes that full thing. Well, it looks like it just kind of froze there. Mm-hmm.

Stopped writing.

Yeah, it's weird. Yeah, so try just refreshing.

Just refresh the browser.

Okay, okay, see what that does.

Um, yeah, um, let me just go down, I guess.

Scroll all the way down. It's not—

it's not liking— almost like it's, uh, no.

It's just struggling a little bit.

I don't know why. No, now it's back. But I— obviously, it's a long one. So I need to go down and catch the last code, right? So let's see here. Okay.

Now we're—

let's see. This is the last one, I think.

This should be the last one. Let's see what happens. Scroll to the bottom, make sure it's finished. Yeah, yeah, yeah.

Um, so yeah, it looks like it, right?

It looks like it should be good.

So yeah, try it.

Okay.

Yeah, let's try and just click the copy button.

You don't, you don't have to put the top.

No.

Yeah, that's right.

Yep. You just push that little button. Yep. Head back over to p5.js editor.

Uh, see it there and I'll delete it.

Let's see. Yep. Select all, delete and paste. Um, let's see. So now we got a script error. Oh, so, so this is, this is fine. I'm actually okay that this happened. I want you to keep trying maybe a couple, a couple more times while we, while we keep talking here and then just tell us. You can stop sharing, but just show up, click on the script error down there, see if it gives you any more details than script error. It doesn't look like there are any more details. Just copy that, paste it in Groq, do it again, and just keep doing it a couple of times. This is very common when— I'm just going to have you stop sharing. I'll just click the button for you. So, um, it's, it's very, very common to have, uh, AI mess up, right? To do it wrong the first time, and then it goes in, you just kind of have to refine it and, you know, get it down. When I just did my solar system thing, I didn't get any errors. That's kind of rare, actually. I usually get one error when I'm— whatever I'm trying to do, like, I'm going to get some error, they have to fix it, whatever. But, uh, it's getting better too, right? I mean, AI is getting smarter and smarter and smarter, and, you know, give it a year, give it 5 years, whatever you want to call it, it's going to be near perfect, right? Even for very large-scale things. You know, one thing that a lot of people, a lot of detractors of AI will say right now is, ah, but it can't handle really complex, big, you know, enterprise-level things. Well, I mean, sure, maybe, maybe it can't handle 10 million lines of code right now, but that's because it has a context limit. It can't read 10 million lines of code all at the same time. Um, but guess what? Every single release from ChatGPT-3 to 4 to 5 to, you know, they all are going to have— they have massive increases. I think right now the current context limit in ChatGPT is 1 million, right? It's 1 million tokens. And then it's going to go to 5 million, and that's going to go to 10 million. And then eventually it's going to be at 1 billion, like it really will be. It's just a matter of compute increasing and AI compute increasing and, you know, needing the data centers to be able to handle it all. And that's another problem. But, but the reality is, guys, it will be able to do it perfectly., right? The fact that we can, as non-coders, sit here, tell it a little thing, and create a game that, that, you know, granted, is a game of the quality that you might see in the '80s, right? That's fine, right? Next year we'll be able to create a game that was created in the '90s, and the year after that we'll be able to create a game that was created in the 2000s. In 5 years from now, we'll be able to create Call of Duty with a prompt, right? Or Fortnite with a prompt, right? It's just gonna keep progressing, right? So anyway, I hope that was useful for everybody. The point of this exercise really is to be like, okay, I can do this. I can make something that works, right? I know, Martin, you're gonna get there. You see the concept. You're making it work. You're writing code. You're deploying it properly. It's just you tried something that was ambitious, and that's cool. That's fine. Totally fine. So I'm gonna share my screen again here, and we're gonna dive into Replit for a second, okay? And you're just gonna watch me first. We're not gonna have you guys jump into Replit for probably about 30 minutes.. But how many people here have not ever used Replit? All right. So only Cody. All right. That's perfect. Perfect. Well, very cool. Well, this is going to be really fun to watch and understand.

Jon, can I ask you a question real quick? So I originally had ChatGPT. That's what I paid for. And then I saw that you said use Cronk, right? So two observations I wanted to ask about.

Yeah.

I put that that prompt into both of them. The first one is what I demo— what I showed, and that was the Groq one.

Yeah.

And then I just did the ChatGPT one that wasn't working, and then I finally got it to work, which is completely different, and it looks much more advanced than the Groq one. So I guess, is that what you're saying from the consistency perspective? Because what I noticed is the Groq code was 81 line items and this is 270.

Yep.

So is that just how they interpret it?

And then one becomes better than the other? Yes. Yeah, the reality is that could be, that could go either way, right? Grok could be the one that does the 200 lines and then ChatGPT does the 47 lines or whatever it is, right? The reality is when you don't provide a ton of context, it has to guess, right? We're gonna talk about that, you know, here in a minute when we're looking at what Replit's doing. This is where most people go really wrong when trying to vibe code and build things and actually build a business around something that they're vibe coding. They just say, you know, build me a weather app. And they don't— you know, imagine if you hired a developer and you just gave him— you sent him an email and said, build me a weather app. That's it. That's all the instructions you gave him. He'd be like, all right. You know, you want to give me any details? Nope, just build it. Right. And the reality is you probably need to send them a 5 to 10 page PRD. Like, you got to send him something that says, here's really, you know, all the details. Here's why we're doing it. Here's the monetization. Here's what it should look like. Here's why it's going to be unique. Here's where it's going to stand out. Here's the audience we're going after, et cetera, et cetera, et cetera. Right. And if you get that, then he'd be like, okay, I got it. Right. And so that's what we're going to really learn how to do here with Replit. And, um, Anyway, so, um, all right, so let me jump into Replit now. Let's share my screen here and we'll pull this stuff up. We'll go to one of my many browsers. Oh, here we can close this out so we're not just looping a video here. Why is this stuck? There we go. All right, close that out. We'll come over here. I built something fun here yesterday and we're gonna build something new today. All right. I'm gonna build today. So I build something new every single time I've done this course, and I've done this course like 60 times. Like, I've done this one 20, 20— this is cohort 22, but I've taught it to tons of other people as well, and I always like to build something new. Um, all right, I have a client that, um that's a landscaping company, and they were asking me if I could build a, um, like a, a chatbot named Sunny. Their company's called Sunline, and they wanted me to build a chatbot that allows them to ask questions of their, you know, like, hey, what should we do here for this plan? Or like, and be able to like put all their company records and things like into it so they could have this kind of this a way to query all their systems and new employees or current employees or, you know, check on the status of a project or ask questions about how, you know, what's the best way to lay sod for this type of an environment, you know, whatever, right? So just kind of a landscaping expert. So I'm going to just kind of play with that today. I'm going to see if I can come up with something cool and I could show them later, but that'll be the little business I do here. So let's see here. All right. So build me a website for Sunline Landscapes. It's going to be a— going to be kind of a dashboard that helps them run their business. And the core feature will be an AI powered— whoops— chatbot that knows everything about landscaping as well as their business. The name of the chatbot will be Sunny. Okay, so this prompt is in my opinion, still bad. Okay, um, and the reason for that is because there's so many things that, you know, and it probably would take me, you know, 20 minutes to really figure out how to describe everything that I exactly need here, right? And so, you know, there's a couple things here. Add OpenAI to your outputs managed service, no API keys required, charges— well, that's kind of cool. Um, sure, do that for now. Integrating with OpenAI, it's actually— that's a new thing. I've never seen that before. Um, I— you can— I've installed OpenAI many times, but usually you have to do it through an API integration, and this is now just like directly set up. So that's pretty cool. And I wonder if this is an OpenAI integration that's like separate from my other one. I need to see what that— I need to look into that a little bit more. But anyway, bottom line is it's gonna— it's gonna build right? It's gonna build this really cool thing. So we will start with a design. Start design. Alright. So the reason why this is still a bad, a bad prompt is because I'm still leaving out a lot of details. It doesn't understand the why. It kind of understands the what, you know, but, but I've left so many details out. So let's, let's give a demonstration. Let's see here, Justin, we're gonna, we're gonna, we're gonna pick on you for this one, okay? I want you to, um, tell us how you get ready in the morning, right? You wake up and you're getting ready to go out, like, what's your, what's your routine?

Wake up, hit the john, brush teeth, take shower if I need to, get changed, and I'm either going to the gym or I'm going to do a first meeting.

All right, cool. Very simple and probably not too different from, from most of us. So, um, let's zoom in on one of those. Brushing your teeth. I want you to walk me through brushing your teeth. Like, tell me, like, pretend I don't know how to brush my teeth and you're gonna tell me how you do it and so that I can learn.

All right, grab toothbrush, grab toothpaste, apply toothpaste to toothbrush, turn on water, wet toothbrush, put toothbrush in mouth, Brush teeth, rinse, turn off water, put toothbrush back, done.

They swallow the toothpaste or— No, spit. Spit out. All right, cool. That's awesome. Thank you. All right, so very cool. So what you're doing there is what is actually happening right here. So AI has only been able to do what's called chain of thought reasoning. For maybe a year, maybe a little less than a year even. Chain of thought reasoning is you take— you have this chain of thoughts, right? And so you take maybe an overarching task like get ready. Okay. And, and then, yeah, you can get up and go to the bathroom. You're going to brush your teeth. You're going to, you know, have to have some breakfast. You're going to work out. You're going to shower. You can do all these things, right? And then if you zoom into each of those individual tasks like we just did with brush your teeth, Yeah, you've got all these little subtasks, right? And even within some of those subtasks, there's sub-subtasks, right? And so our brains do this automatically, right? We know, hey, I need to get ready. And really what that does is it kicks off a chain of events that we need to, we need to do all these things. And our body just kind of goes through the motions because we've trained ourselves to do that. Okay. AI is taking an idea of, hey, we need to build a website. Okay. A website probably is going to have header. It's probably going to have some buttons. It's going to have some messaging around this thing. This is a chatbot, so we're going to need an OpenAI integration. It's thinking through and it's building out all of these little tasks. Now, the reason why this is a bad prompt is because I'm leaving way too many gaps of things that it's going to have to fill in. How far off from my vision is this actually going to be, right? Probably pretty far, probably pretty far off. And so because it's going to be far off, it is— I might not be pleased with the result. I might be like, AI sucks, it can't do this very well, right? And so what we want to do is we want to give it a lot more information. Now, this is, this is cool. It did a bunch of stuff, right? AI chatbot Sunny powered by OpenAI interactive chat interface. A business dashboard overview, client management section, job project tracking, calendar view, professional landscaping theme design, mobile responsive layout for field access, whatever, right? So, um, cool. It actually thought through some things. I didn't ask it to make it mobile responsive for field access, but it's like, hey, this is a landscaping company, they're probably out on the job, they need to figure this out, let's make it easy to access on the phone. So again, that's part of its chain of thought reasoning. It's smart. Right? But it's like a smart employee, right? It's a smart, proactive employee that's going to just kind of say, all right, I'm going to figure out what this guy needs. I'm going to try to deliver the very best thing I can. But, but even though it did this kind of stuff, you know, like I didn't ask it for job and project tracking, like I might want that eventually, but right now I really just want to build a chatbot that can access company records and company stuff. Right? And so, you know, it's out being an overeager, you know, agent and doing that. And I think that's fine. Right? But what it is doing, which is really cool, is it's now doing something that's much, much more complex than, you know, the little video game we just made, right? Which, by the way, I think my video game might still be running somewhere. Yeah, here it is. Save the sun, right? Anyway, I slowed it down. I got it to where I could actually like keep up with the— keep up with the thing, and I can actually do that. I have a score of 10 now. Anyway, kind of fun. So, um, anyway, we're going to turn this off. I actually like this one. I kind of want to save it. I should save it, but it's okay. I can, I can come back to the Groq chat later anytime. So we'll leave that back for now. But now I'm not using some memory doing that. But what it's doing is it's writing way more code than just that 50 lines of code or 200 lines of code. Like, see, it's writing 90 lines of code here for the schema. This is kind of how the website's gonna work. And then here's the storage, almost 400 lines of code. Here's the OpenAI integration, 62 lines of code. Here's the routes and all the different websites and things that are gonna happen and user messages and errors and all this stuff. I'm over 1,000 lines of code and I've just been sitting here talking to you guys. This used to take lots of money, guys. This used to take so much money to do these types of things, to be able to put together fully functioning apps that can do this stuff. It's just mind-blowing to me. I've done it 1,000 times and I just am constantly amazed by how much better it keeps getting. So one of the cool things about Replit is version 1 of Replit with AI could do— it could think by itself for about 2 minutes, right? And so you tell it to do stuff, it thinks for about 2 minutes, and then it would ask you a question. All right, what should I do next? Okay, great, I'll do that. Okay, what should I do next? Right? Because it only had a certain context limit. Okay. Version 2 of Replit could think for 20 minutes, right? Which is amazing. Right? That now it can work for a long time. You tell it to do something and go grab some pizza, watch an episode of The Office, whatever you want to do, and then, you know, it's good. Version 3, which we're in right now, which came out just about a month ago, can think for 200 minutes autonomously. 200 minutes of AI work is easily 100,000 lines of code. Like, it is crazy. You can build a fully functioning CRM. In fact, I saw something yesterday, from the CEO of, of Replit, um, Amjad Massad. I'm going to pull it up because I thought it was a cool— it like, it's pretty cool. Let's see, he, he posts a lot. Uh, yeah, right here, look at this. Greenleaf replaces Salesforce with custom CRM built in 25 hours. Salesforce hates that, obviously, right? But you can build your whole new— anyway, amazing. 60 employees. They now have a custom CRM built with Replit. Unbelievable. Unbelievable stuff. That's what's possible these days, right, with AI. So unfortunately, we have to kind of sit and wait for this to happen. But we'll, you know, we'll, we'll do a couple other things while we're, while we're waiting here. So, but anyway, guys, the, the, the capabilities. In fact, one of the things I do want to do right now, and I'm sorry, I don't have this ready. Let me just, Pause really quick and then find this video. I want to play a video for you guys. Let me see if I can— let's see, what is it called here? Eric Schmidt.

Here we go.

All right, we're going to watch a quick little video that's just 2 minutes long while this is thinking and encoding for us. Okay, and then we'll talk about it.

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

Yeah, yeah, largely free is a big deal. Come on, let me click this. Um, It's, it's a huge deal to be able to, for a couple hundred bucks, like, boom, I just created a Sunline Landscape dashboard, right? Like, building something like this, guys, used to take $50,000 easily, and it would take months. And now, you know, while we're watching a little video and whatever, it's, it's building out, you know, this thing. Now, it's not fully functional yet, right? It's, it's right now just working. You know, it's kind of functional though. Right? Like, you know, it hasn't built the analytics page yet, it looks like. But yeah, or the settings page. It's, it's building these things one, one little bit at a time. So some of it works and you just kind of will go through and you'll start working on it and building it out. But, but look how crazy this is. The one thing it didn't do though, yeah, here's my chatbot. Yep, there's Sonny. I'm Sonny, your AI landscaping. Um, tell me about the sun. Typing, typing, typing. Wait for it. This is interesting. Storage services, OpenAI, landscaping. You are sunny. You should keep responses concise but informative. Yeah, it looks like it might not be working yet, and that's fine. Right. But because I told it actually just to create like a visual preview. So it's not, it's not all the way there yet. But it did way more than a visual preview, by the way. Like this is, this is crazy, right? It didn't build the settings out like we said, and that's totally fine. And it obviously didn't get this actually completely working. Wow, that's amazing that it saved that. I'm going to refresh this. Really interesting. Is there a storage? Anyway, we'll look at that later. But, but guys, isn't this crazy? What do you guys think about First of all, what you're seeing on the screen here and what Replit can do and what Eric Schmidt— Eric Schmidt, by the way, is the former chairman of Google. He's a billionaire. He's super smart, super connected. And by the way, he said this in April of this year, 6 months ago, right? And some of the things he said would take a year are already done. So what do you guys think? What are your thoughts just as you look at this?

What can't this do? Once it's like if you gave it the true BRD, like what is the gap between it being live at that point for you to be fully functional?

Can you do anything? It can be fully functional, fully live. I mean, like I said in the beginning, right, my, my GenAI API is fully functional. People are coming here purchasing things. You know, you can take, you can take assessments, you can do all these different things. It's a fully functioning website. That is earning money and earning millions. Actually, it's earning a lot of money. Right. And so, I mean, it can do anything. There really isn't a limit. Now, there's definitely an art to doing it right. Right. Like I said, there's a lot of people that fail, like they'll get 50% or 80% and they just don't know how to finish it. They don't quite get it all the way there. And, and usually that's not a matter of the system itself can't do it. It's a matter of the person in charge doesn't know how to push it through to the finish line. And I'm going to show you guys how to easily be able to push it through to the finish line. It's really not that hard. It's not rocket science, right? But if you understand a few things, you'll, you'll do better. Okay. So again, guys, I already said this, but I think this is a bad version. I think this is a bad version of, uh, or a bad prompt, right? I think, I think this Sunline Landscapes dashboard is cool. First of all, it gave me tons of other things that I just, I don't think I need yet. I don't know if I want to keep all of my projects and clients in here. I want to probably connect it over to their existing databases and things like that. But here's the thing, it has no idea that that's what I want. I didn't give it any of those details, right? So we're gonna back up. All right. Okay. We are gonna bounce into my favorite platform, ChatGPT. The reason I like ChatGPT, guys, now I don't know if it's my favorite platform, but it's my, it's the one I use the most, right? And the reason is because it's my project manager. Okay. I have tons of projects in here that I work on. I've got different GPTs and different projects. So many that it can't even actually load all the projects I want it to load. Um, it has— it's, it's loading a big cached thing right now. There's— I, I literally have like 70 projects, and they don't have a great interface for me to see all the projects without it taking— it takes like a minute for it to load the whole thing. So anyway, that's their problem. Um, so try to keep it, you know, kind of to where you want. But I have different clients in here. I've got my General AI Proficiency Institute. I'm, I'm talking to it about, you know, all kinds of different ideas. If you scroll back in this all the way back to February 27th when I started this company and the idea, this was my first chat, right? And I talked through it and I worked with it, right? So we'll, uh, we'll talk a little bit more about that in a second. But before we do that, I want to just talk a little bit about, um, about ChatGPT itself and some of the things that are behind the scenes that help it be very, very effective. Okay, first of all, um, Context matters more than ever in AI, right? Context, you know, when you're having a conversation, you know, if you just walk into a room and there's no context, you just walk into a room and walk up to a person and ask them, what should I do? They'd be like, I don't know who you are or why you're in this room with me or what I should, you know, there's no context, right? You gotta provide details, okay? And that is just as true of an AI. I really like to think of my AIs as my employees. I like to treat them like people. And the more I treat them like people with— and people, by the way, have deficiencies, but I calculate for that. I account for that. I understand it. I know what's going on. I know that sometimes maybe ChatGPT will make something up, right? I know that sometimes it's not going to remember that I asked it to do something. I know, you know, anyway, so But if you understand kind of the quirks of your employees, you can work with them really well. But the reality is I treat them just like they're a person. I don't expect them to know everything unless I actually share it. I got to be a good boss. I got to be a good boss. That's really what it comes down to. Okay, so the context inside of virtually any AI system is similar to this right now. In a given chat, right? So I can go back to a chat I was having on push-ups. Okay. I was talking to my sister about, you know, where, you know, how much of a difference is a push-up is, you know, if you go 90 degrees or if you go all the way down and touch your chest or your nose to the ground, right? What's the, what's the difference, right? So anyway, I can come right back to that conversation and I can say, what if I only go to 45 degrees on the push-up, right? I don't have to give it the context again. And you already saw this earlier in Groq, right? With the P5JS. It already knew we were using that language. I don't have to tell it what's going on. It remembers everything that I talked about in this conversation, and it can pick up right where we left off, and it's going to understand. Okay, at 45 degrees, your elbow's only bent slightly. Here's what you're going to do. 45, 90, full, whatever. Okay, so great. Okay. And so, so cool. So it's just going to give me all this stuff. It's just going to tell me lots of great things. Okay, so the context is remembered inside of a conversation. Okay, now what happens if your conversation— you have a conversation inside of one chat for 10 hours straight? Can it remember every single thing it said in that entire conversation? No, it actually can't because it has a context limit. It can only process so much data at once. Can it remember most of it? Yeah. Will it get it pretty good? Yeah, it will. But you have to understand that if you have one conversation, you just keep talking forever in that one conversation, it will become less efficient. So if I'm switching to a new topic, new chat. It's going to a new chat always. Okay. Short and concise is better, right? It's like talking to your wife for 10 hours straight and then expecting her like, remember what we were talking about 10 hours ago? Like that specific thing? Like maybe she remembers, maybe she's, you know, we're 10 hours in, she's forgotten everything. She's ready to go to sleep. Right? So anyway, treat it like a person. Okay, the second thing, the second place where you have context is in what's called— ChatGPT has actually maybe a little bit better than some of the other platforms that are out there. It has a memory. Okay. And the memory is inside of this personalization element here, where you have— you can, you can give it custom instructions. I actually don't use this specific part of it. I certainly could. I could tell it that I'm not a gastroenterologist, or whatever, I could give it some more interest. But it actually remembers things that I've talked to it about, right? And so, you know, this is actually something I don't even care, I can delete that, 'cause I don't, this was a random, you know, fake business I was creating about 60 posts per day. Anyway, but bottom line is there's all these different things. It remembered, it knows I'm an entrepreneur and a composer and a consultant. I'm a husband and father. I live on a 5-acre farm. It knows I'm fluent in Spanish and Chinese. I'm a kayaker. I'm a member of the Church of Jesus Christ of Latter-day Saints. Et cetera, et cetera, et cetera, et cetera. It knows who I am, and because it knows who I am, it is going to, within context, give me great answers, right? If I ask it a question about God and it knows I'm a member of The Church of Jesus Christ of Latter-day Saints, it's going to answer it differently than if it knew I was a Jew, or if I knew I was Buddhist, or if it knew I was whatever, right? The fact that it knows that, right, doesn't matter which one of those I am. It just matters that it knows that this is what I am. And so this is the kind of filter, the kind of prism I want it to look through when it's giving me an answer. Okay, ChatGPT is trained on over 20 million— the equivalent of 20 million books of data. Okay, it knows everything about everything. It does. So how does it filter and know how to present details to you or information to you unless you tell it? And so if you open up a brand new ChatGPT account, it is not going to be as effective or efficient as it is when it really understands what's going on here. So you've got a system-level context. It understands who you are. Then you have a conversation-level context. And then again, the thing that I love is the project-level context, right? Inside of a project, I can add files. I can upload company documents, things like that. So I can get it right up to speed on something I'm working on. Or I can just have it remember all the different conversations I've had with whatever project I'm working on. And so, anytime I'm like, hey, I need a new contract or I need an update to this or I need a proposal written for this thing, I just tell it and I give it some extra context about that client or whatever it is I need. And it knows what my business is. It knows what I care about. It knows my brand language. It knows my website. It just knows all this stuff. Okay? And because it knows all of that stuff, it's a very useful assistant. It's well trained on everything I need it to know about GenAI API. Okay, so we're going to start a new project here. And in fact, I wonder if I even have a project for— you know, it's starting to load a few more of those things. Looks like I don't have a project in here, so that's fine. But we're going to go new project and I'm going to call this Sunline. Okay. And we're going to say— and let's see here— investing, home writing, health, travel. Yeah, it's none of those things. Create project. Okay, super easy, right? Create a new project, and now here we are, we're ready to rock. Okay, I can add files, I can, I can, um, I'm gonna look at the settings here. Project only. Yeah, this is kind of cool. Actually, I wondered what the settings were. This is actually a new button. You can have it access memories from outside chats and vice versa, or you only want it to access its own memories. Its memories are hidden from outside chat. So that's kind of cool. Keep it all, you know, you don't want it to be part of everything else. It's actually super interesting. Um, all right, cool, good to know. I like clicking on new buttons. Um, there's— there were some new buttons in Manus as well. We'll play in Manus later, uh, probably tomorrow, probably tomorrow.

Jon, I got a quick question for you. Um, I went through and kind of looked at my memories, um, that you— that you did.

Yep.

And it grabbed a whole bunch of just like really random stuff of maybe stuff that I was looking on, looking at. Yep. Uh, you know, from anything from my, like, I have a mini golden doodle to I have a barista espresso machine, all these different things. How do I, I mean, as, as you're trying to create an assistant of who I am, are you controlling that narrative or are you just allowing ChatGPT to create its own of like who you are as a person?

Yeah, that's a great question. And, you know, the, the reality is what they should do is allow you to add a memory, like, manually here, but they don't. They only allow you to delete them. Okay. Okay. And so if you want to do— if you want to add something to the memory, what you do is you just talk to ChatGPT and then say, here's who I am, I want you to learn all this and add it to your memory. Okay.

Because I just— I, I'm imagining As you're training, like, basically, it's just like a knowledge center of, like, who I am. And you're referencing that with all your projects of, like, if you know that I'm a 44-year-old financial services experience background, I mean, it keeps that perspective. So, it's just smarter. Again, that's kind of like what your message is so far is about perspective. It's about information that you provide, like, who are you? What's your perspective? What's the perspective of the people that you're speaking with? It just seems like that's a project in and of itself of as you're going and building AI platforms of like, this is who I am as a person.

That's why I like to use a centralized database like ChatGPT or a centralized system as my project manager. Because then no matter what I'm working on, maybe I'm going to go work in Replit or Manus or Cloud Code or VEO3 or whatever it is I'm going to be working on., but my project stays here. Then it knows who I am, what I'm about, what I'm trying to do. Then I take the prompts and code and organized information and I take it over to those other ones. And so I start with a place where it gets me and then I can quickly take it over because, because if I'm working in Replit, for example, Replit doesn't know who I am. It has no way to learn who I am. It doesn't have a memory. It doesn't know any of that stuff, right? It's just gonna do stuff. And so that's why I wanna bring it context that's created in depth here in ChatGPT when it understands exactly what's going on. So yeah, really, really good question. And this is, this is a pro AI tip, right? The more it knows you, the more it understands you, the better results you'll get from it because it will give you, you know, it's, it's, there's a lot of advice out there. There's a lot of good advice out there, but good advice for one person might be terrible advice for another.

Other question for you, you know, now that you have your whole business model in this, What's the security parameters to protect you and your business from someone like me walking in and saying, hey, I want to know everything about GenAI. Like, tell me, tell me the whole thing. Like, is there any protections for you that, that some of this is confidential?

Yeah. So I have right here improve the model for everyone turned off. Okay. So I don't, you know, Allow your content to be used? Nope. Now I can turn it on if I want. I actually don't really care that much. The reality is somebody could take my entire business plan, everything that I'm doing, and try to do it and likely fail. They will likely fail, right? Because ideas are a dime a dozen, man, right? I think actually they're probably worth less than a dime a dozen. It's all in the execution. It's all in, do you know how to take this idea and this concept and this model and then do you not? Actually push it out there, right? But the reality is ChatGPT can be absolutely 100% secure as long as you just have some of those settings properly, properly clicked and toggled. So, alright guys, so here's what we're gonna do now. One of the things that a lot of people forget to do with AI is, and maybe, maybe some of you have had this type of a conversation with ChatGPT. Have you ever, have you ever said, Hey, ChatGPT, I've got an idea. I want to talk about something and you tell it, you know, I want to, I want to. Yeah, here's an example, right? I live— there's a beautiful mountain outside my window here called Mount Timpanogos. It's right. Yeah, it's right there. Kind of hard to see. Anyway, Mount Timpanogos is out there. Really tall mountain. Second tallest mountain in Utah. And it is It's like a 6-hour hike to get to the top. Okay, so what if I— what if I— and I've actually done this, I've literally done this. In fact, I think it was— I think I did it recently. Um, I basically said, hey, I've got an idea. I wanna, um, I want to start a company where I sell ice cream at the top of Mount Timpanogos, right? People get at the top, they've been hiking for a long time, and You know, they'll probably want ice cream at the top. Okay, so first of all, this is a stupid idea because you're gonna have to hike ice cream up to the top of the mountain. It's ridiculous. Okay, it's not, it's not practical. Okay, but of course ChatGPT likes to glaze you and make you feel good, and so it says, you're a genius, this is awesome, right? Great work. Okay, and, and then, you know, it'll tell you all these different things about it, and then, and then they'll say, would you like me to create a marketing plan? And you say yes. Yes. Right. And then awesome marketing plan. Okay, great. Would you like me to put together some profit margin analysis? Yes. Right. And then, hey, would you like me to do this? Yes. And you know, you read everything and you just love it all. It's great. Okay. So now you get to the end of this conversation and you're like, I just built a business. Well, did you? Did you actually build a business? Because really all you did was you say, I want to build an ice cream shop on the top of Mount Timpanogos. And then yes, sure. Sounds good. Do that. Great. What have you actually contributed to the world at this point? Basically nothing. Okay. You've prompted ChatGPT. ChatGPT is very action-oriented and always wants to push towards the next action. Okay. It wants to do the next thing in the, in the process. And so we got to flip that. We got to flip that. We got to get control, right? You need to be the thought leader. And ChatGPT needs to be your thought partner. It needs to help you organize. It needs to help you kind of maybe even push back on some of your ideas, right? But the bottom line is you need to be the one in charge. Okay? So I'm gonna come back over here to wherever the heck my— this one is. Okay? And I'm gonna copy my little prompt over here that I put into Replit, and I'm gonna say, I want to build a website for Sunline Landscapes. SunlineLandscapes.com, www. Okay, um, put the period there. Sorry guys, I'm kind of like a grammar nut. Anyway, it's going to be kind of a dashboard that helps them run their business, and the core feature will be an AI-powered chatbot that knows everything about landscape as well as their business. Name of the chatbot will be Sunny. Um, they already have many other systems in place, so we'll want to tie into Aspire and monday.com and some other services to— in order to pull over some data, uh, via APIs. And other connectors. Um, okay, cool. So that's kind of the basic, like, here's what I want to do, here's my business idea. Okay, then here's where we're going to flip the script. I want you to interview me one question at a time about this idea. Um, I want you to ask me about You know, the, the USP, the unique selling proposition, what's going to be different here, the differentiators, the functionality— type that wrong— functionality, the branding, and anything else that— anything else you think or you need to know in order to help me make a solid business plan, um, for this idea. Okay. All right, so what this is going to do, guys, is it's gonna flip the script here. And I actually don't want ChatGPT-5 Instant. I kind of want it to actually— I'll switch it to Auto. That's fine. All right, so now what it's doing here is it's going to say, okay, great, John, Mr. AI user, you're going to have to actually think now. Okay? You got to just be careful when using AI to make sure that you're the one in charge. Okay? So what's the core purpose of the dashboard? In other words, when Sunline's team logs in, what's the main job they'll be trying to accomplish with it? They can't really do it in AspireMoney.com today. Those other platforms don't have AI-powered features. The chatbot is a key element that will allow them to you know, ask questions about their business, about landscaping, and about, you know, whatever else they need to find. I want it to be a more natural way to access data. Okay. And so then it will ask the second question right now. One of the things that's important here, guys, is asking— I want you to interview me one question at a time. If you say, I want you to ask me some questions, it will literally give you like 20 questions. And you know what our brains hate? Homework and tests and quizzes. And you know, when you see 20 questions, you're just like, I don't want to do this, right? But when you see one question, your brain can just focus in on it. It doesn't feel overwhelming. And, and you actually start to get asked some really, really cool questions. Okay. So, um, here's what we're going to do. We're going to take the next about 15 minutes. And you're all gonna do your own ideas. Okay? So I'm gonna copy this right here for you. I'm not gonna copy this part. Okay? You're gonna write, I want to build this, right? Whatever that thing is, give it as much detail as you want. Probably just maybe like 2 or 3 sentences tops. You'll get much more detailed later. Just kind of high level. I wanna build this. This is kind of the idea. And then I want you to copy this. I want you to interview me one question at a time. And so I'll paste this in here and You'll, you know, anyway, you'll, you'll be able to go in. It looks like I saw a question. How do you handle the self-affirming nature of AI and making sure you get unbiased answers? We'll talk about that. We'll talk about that later. It's a great question, Kalima. So, so anyway, guys, um, go through this interview process. You should be able to get through maybe 7, 8, 9, 10 questions or so in the next 15 minutes. Okay, so right now it's 10:33 on my, on my clock here. We'll go till about 10:48. I'm going to sit here and answer questions online. And we're going to start building our own projects here. But this is where you really get to be the boss and start to think through what is this going to be about. Kalima, you're going to be like, hey, you know, I've got 200,000 followers. This is what I'm doing. It's blowing up. I need to figure out my offering, right? Help me figure this out. Even if you don't know what your business idea is, help me figure out this business idea. Interview me, ask me about what I'm trying to do. Help me identify opportunities, right? And then just answer the questions, right? And keep going. Okay.

And obviously, guys, speak up if you have questions. Would you mind just scrolling up to your prompt just so I can leverage it?

Thank you. Yeah. Yeah, and I'll copy that whole prompt if you need, if you need. I'll throw it in the chat as well. Just the top part so you guys can see that.

All right. Hey, Jon. So does it have to be like website or app, or that doesn't matter at this point?

I would recommend you build, if it's more like an app, I would say web app. Have it going to be working towards a web app because you can always turn a web app into a traditional Android app. Thanks. We'll be building in Replit using React, and React can easily go to the web, iOS, Android, whatever. What's the USP again? Unique selling proposition. Why is your— why is your business special? Why is your idea worth doing? Hey guys, take another about 2 minutes. And by the way, for what it's worth, you know, I've spent 3 hours on this step before, right? This is the strategy part. This is where you're really thinking. And I hope, I hope you guys are like, your brains are like really activated right now. Just like, man, this is cool. I'm thinking about things I wasn't I wasn't considering, you know, it can open your ideas, right? It's not that you don't want AI to never come up with an idea, right? It certainly can help you do that, right? You just want to make sure you remain in charge and that it's taking your ideas, your thoughts, so that when the product starts to come to life, you, you're happy with it.

It's really delivering the vision that you had. I could see how this could definitely lead you down a rabbit hole of really fine-tuning, but which is great though.

And it's okay to go down those rabbit holes if you're going to build a business, like actually, you know, time in your life to something, you want to spend a few hours really thinking through it. In this, you know, in this course, right, kind of the goal is and the challenge is, can we build a business in a few hours, right? Which is, you know, just compressing that really, really tight. And so, we don't have time to really go deep. But anyway, long story short, guys, let's take what we have now. So, here's what we're going to do now. So, you've gone through a bunch of different questions. It's probably asked you about the feel and who's your customer and what's your company called and branding. Maybe haven't even covered all those things. Hopefully you've talked at least about, you know, what is it, why is it unique, what's going to be happening here, does it understand kind of what you're building? And, and now, um, we're going to, uh, we're going to build a business plan. Now one of the things I want to— I'm going to pick on Kalima again here because I really just want to kind of help him, you know, kind of guide a little bit. Um, and maybe you already did this, Kalima, but if I were you, I would be asking questions like, um, like I would be really giving it the details of, hey, I've got this Instagram page I've got 200,000 followers, I've got 50,000 DMs in the last 6 months, and I want to set up some automations with ManyChat and whatever. And I want people when they, when they comment and interact to be driven towards some sort of offer, right? And I need to figure out what that is. And let's help me build a business and a website where they can land and engage and pay me money and whatever, right? But that's where I'd be doing— I'd be, I'd be using AI to help me figure out what is this next thing. Okay, so assuming you guys all did that stuff, And I encourage you probably to go back and do this. By the way, I am recording this and I'll record tomorrow as well. You'll have the full recordings, right? And so when you, if you want to go back later and be like, hey, what was he saying in that spot? You can always go back. Okay. But the next thing we're going to do is this. I want you to take everything we've discussed and build out a 2,000-word plan. We can, yeah, 2,000, let's say 2,000-word business plan. Actually, use the, whoops, use the Canvas tool for this. Okay, so the Canvas tool, guys, is something inside of ChatGPT that's kind of like a document generator, right? Kind of like a Google Doc, right? But it's inside of this and you can see here now you can copy, edit, download. It's creating this little document inside of the chat, which is super cool. So I'll give you guys this thing. Go ahead and copy paste what I just gave you there. And by the way, it will ask you questions until you're blue in the face. It'll just never stop. So sometimes you just have to be like, all right, that's enough, right? We'll fill in the gaps later, okay? But when you ask it to build this business plan, now we've got an executive summary, we've got unique selling proposition, vision and mission, brand design direction, product overview, core functionalities, crew members, crew leaders, management, project managers, admin staff, public users, technology stack, all this kind of stuff. It's doing a lot of really, really neat things, um, and it'll, it'll figure out all this stuff. Ooh, JWT-based authentication. It knew, it knew that. That's pretty cool. That's pretty cool. Aspire Monday, QuickBooks APIs, OAuth 2, user experience, left sidebar, main panel, bottom right interaction model. Here's the business model. Here's where we'll be making money. This is pretty cool. It's pretty cool. Here's the roadmap. So yeah, obviously, you know, if you had time, we just, we just don't today, right? You'd read through every single word, right? I'm pretty good at scanning and seeing exactly what we're going to be doing here. But, but this is, this is amazing. I built a 2,000-word business plan now. And now I could take this directly over to Replit now and say, hey, get started. But I'm going to add one more step in here. That I think is valuable. Okay, so what I'm going to do is I'm going to go ahead and I'm going to copy my business plan. Okay, by the way, I found that 2,000 words is kind of the right length to ask it to do something like this, because if you tell it 1,000 words, it's just going to cut some details out. If you tell it 4,000 words, it just becomes a little bit too big. 2,000, really, like, that's a very specific number that I've really tested and played with. I've tried 1,000, I've tried 1,500, I've tried 3,000, I've tried it all, and 2,000 just seems to get the right, the right balance. Okay, so as once your business plan is created, you're going to click copy, right? And then you're going to come back to Grok, right? And now we're going to open a new chat and, um, we're going to ask it to do some market research for us. I want you to do some market research on this idea. Don't focus so much on the business plan structure itself, but more on the market and the product idea. Be sure to check X and what people are saying in this space, as well as many other sources you can find. Okay, and I'm actually not going to do fast. I'm going to do just auto, and we'll let it, we'll let it do what it does. Okay, so that's all, that's all I'm going to do. And I'm going to hold Shift while I press Enter, Enter so that I can then paste something else here. And I'm going to paste my business plan. Okay, I don't need the word count. We can delete the word count. Okay, so I'm ready to go. So I'll give you guys this little, uh, thing in your, in the, in the chat here. Okay, so you guys can grab this. So just copy paste that into Grok, Shift, Enter, Enter. Or return, return, and then paste your business plan and say go. Okay, and then it's gonna do a month's worth of research in, you know, 2 minutes or so. You know, landscaping's a good, good business to be in. It's always going to be new houses, there's going to be new buildings, new businesses, and someone's got to take care of landscaping. It's just like growth is built in, at least if you're in the right place. If you're in New York City, like, you know, kind of have to just go up there. Not as much landscaping. But if you live in Kansas, landscaping's not a bad business to be in. Very AI-proof as well, which I love. Okay, so anyway, bottom line is, um, I now have a pretty good market overview for, you know, landscaping and what kind of stuff is going on here. There isn't, there isn't a lot of stuff here. I mean, this is very, you know, this is a market, a niche solution in a market ripe for digital disruption. Like, this is a good idea. Like, I shouldn't just do this for my, you know, for, for my one client that's a landscaping client, right? I should, I should actually turn this into a real business that I could then sell to any landscaping company out there, right? That's what it's kind of thinking about and saying this is a really good opportunity. Key trends: sustainability, AI for efficiency, competitive landscape, Here's a couple little things that are happening here. Yeah, job or service autopilot. Here's some whatever user sentiments and discussions, growing enthusiasm for AI and landscaping, hardware, drones, opportunities and challenges, integration downtime, rate limits, require caching. Okay, whatever, no big deal there. So long story short, you'll read through your market research and that's cool. And then here's what I'm gonna do. It's very simple. I'm literally just going to copy this. I'm going to copy the, you know, the feedback and I'm going to come back to here and I'm going to say, please use the information below to refine the business plan. Be sure to keep my answers and thoughts prioritized but informed. Based on— let's see, actually, but informed based on this new information. Only adjust what you think actually improves business plan. Create a new canvas called improved business plan. Okay, Shift+Enter, Enter, drop all that stuff in there, and I'll copy this little thing for you. And so now what I've done is I've copied and I've copied the stuff from Grok, and now I'm gonna have this new little, little paragraph that I gave you here, paste the, the research and send. Okay, so guys, In order to do this without AI, this, this entire process would be many, many, many days, if not weeks, to try to get the level— like, the amount of research that Grok did alone— it searched 27 web pages, it found 15 posts talking about this, did research, found competitors, built out analyses, all these different things. Like, just that alone is at least a week of work. Right. And it takes you a minute, 2 minutes here. Right. It's just incredible the efficiencies that AI is bringing. So now we've got this improved business plan being built out. It's finding a few better opportunities here. Yeah. AI adoption remains below 20% in the landscaping industry, leaving a massive gap. Very cool.

Not surprising, guys.

If you want to do a cool business idea, you're welcome to take my business idea. If you want, you can go sell this, go, go into— but go into blue-collar trades, right? Jacob, take this into, you know, cow management, you know, farm management, right? What kind of AI-powered dashboard could you have, right, where you add, you know, trackers to your cows and and have a dashboard that shows the location of them at any time and digitally tracks when the last time they got their vaccines was and all these different things. I don't know a ton about cattle stuff, but my wife has a couple of horses and she can go on cattle drives occasionally. So I know a little bit. I know a little bit. But bottom line is, man, there's so many opportunities in blue-collar spaces for people to go and bring AI to their industry.

Um, it's kind of funny that you mentioned that actually, because like almost all of that that you just said— so I just started, whenever I came into this, I'm— I just started my marketing business, like just a small blue-collar marketing business is all it is. And so all of this, like reading this right here, like landscaping is huge where we are. We're having like 8,000-home community that's about to come in, another 6,000-home community that's about to come in., and they're, I mean, it's just like, it's blowing up here. Plus cattle-wise, like I've been trying to develop this thing for like a couple years, but I just didn't know how to do it. Um, basically where you would take a drone and each cattle in their ear tag, they have like an RFID chip in their ear tag. And as you fly over with the drone, because here in Florida there's a lot of woods. And so if you fly over with the drone, as long as you're within— I think it depends on the RFID tracker, but as long as you're within 200, 250 feet of them, and that's pretty easy if you're in the air, you can track and see exactly where they are, set the drone on them, and then it literally follows them. So then you could find them with horses, you can find them with anything. So it's like everything you just said, there's so much, so much happened.

So it's pretty crazy. Yeah, see, it's cool, right? Like, yes, AI is going to replace jobs, guys. It is totally. But there's new jobs, there's new stuff we're going to create, new businesses, new better ways to improve the world, right? I just, I'm not afraid of, of progress. Progress is awesome. Does progress disrupt? Absolutely, right? But progress and disruption bring opportunity left and right. Okay. All right, guys. So now we've got a business plan. We've got an improved business plan. Okay. Now the second thing that I'm going to do here and the last thing we're going to do before we get into Replit and then we're going to take a break because I know you guys probably want to stand up. The next thing we're going to do here is we're just going to say, please create a complete feature list. Whoops. Feature lists for all front-end and back-end elements that need to be created. Don't include any specificity in regards to which technologies will be used. Just list out the features. Create a new canvas for this and call it feature lists. Okay. So one more canvas and then we're ready to rock here. Okay, so I'll give you guys that little thing. Go ahead and just drop that into your chat and we're ready to get into— we're ready to get into Replit and get started here. Okay, um, I'll give you guys just a second to get that feature list knocked out.

Jon, sorry, I want to make sure I'm doing this in the right order. Can you go back to the just like the process. So I did my— I got my business plan. It says write a 2,000-word thing. And then I went to Grok and I did the— I want you to do market research on this idea.

That's the step I got to. Yeah. So now you're going to take that research and if you just go into the, into the Zoom chat, right, you'll see the different prompts that I've dropped in there. Right. So just grab the one that says please use the information below. And that goes back to ChatGPT. Yep, you'll take that and you'll ask it to create an improved business plan, right? So grab that entire thing right there. So just copy that, paste the research that you got from Grok, and then click enter and it'll create a whole new thing called improved business plan. And then, and last one, the feature list. Yeah, I know I'm going kind of quick here, guys, and it, you know, I try to slow down a little bit. This is good. And, you know, the second half of this morning, you know, we're going to take a 15-minute break here. And so then we'll have about an hour and a half left. Um, after that, it'll be a little bit shorter second half of the morning.

I'm done. Well, I thought it was stalling the feature list, but maybe it's just thinking.

Yeah, it might just be thinking. Mine took a second to think. It has to review a lot of information there. Sure. Guys, once you've got a feature list, we're doing pretty good. Oh no, wait, looks like it froze my It thinks it froze. Stop. Maybe it did freeze.

Jon, do you have to do the canvas prompt portion in order to get anything exportable?

You don't have to. You don't have to, but the, uh, you, I mean, you can cop— I mean, obviously you can just copy paste text, right? But, but yeah, I, I, I use the canvas if I need, like if I'm making a proposal, if I'm making a, you know, a report or something, you know, I'm gonna, I'm gonna send over there. All right, I'm gonna keep waiting. For this. It's really struggling.

I don't want, I don't want it to refresh. I asked the question because I've, I've used ChatGPT to help build a deck for a presentation I'm gonna have, and then I also used it to write a book this past week, and it seems like the exportable feature is just terrible, and I don't know if I'm doing something wrong or it has to be listing that prompt as Canvas.

Yeah.

Part of it is ChatGPT is not very good at that. Okay. If you're gonna build a presentation, you're gonna use Gamma. Gamma? Mm-hmm. I'll show you guys Gamma. I'll show you guys that later. Gamma's awesome. Yeah. I mean, part of, part of being a good AI user is just knowing what to use where, right? Like, you know, you're on the farm and you're, you're out and you need to, you know, get some weeds. You know, there's the right tool for that depending on what, what kind of weed it is and where it is and what you're doing, right? And so, just like anything. All right. So we've got a feature list. We've got an improved business plan. Now it's time to go into Replit. So the last thing we're going to do before we go over to Replit, we're going to say, give me a 3-sentence description what we are building.

Okay.

So what you're trying to do here is just get, you know, just, hey, here's 3 sentences. And so we can tell Replit very succinctly, here's what we're building. Sunny. AI-powered dashboard. Okay, so I'm going to copy this and I am going to start a new project inside of Replit. Okay, describe the idea you want to build. I'm going to give it the 3 sentences. Okay, then I'm going to Shift+Return, Return. Okay, so that I can add some more stuff. So don't send yet. Don't send, don't send the request yet. Okay, I'm going to come in, I'm going to copy my feature list. I'm going to paste it, and it's going to paste as like— it will look like a little gray box, an attachment. You'll be like, wait, where'd it go, right? It's just a little gray box with an attachment. And you go up, you grab your improved business plan, copy and paste. Okay, please pay close attention to the attached Um, detailed. Okay, all right, so this is what your prompt should look like in Replit. You've got your 3-sentence description. You're going to use the thing I just dropped in the chat there, please pay close attention to the attached documents, and then you will have pasted your business plan and your feature list. Jon, what was the 3-second thing again? Yeah, the 3-sentence thing. If you look in the chat there, yeah, just sorry, maybe I didn't copy that one. I'll copy that for you. But it's just, yeah, give me a 3-sentence description of what we are building, right? Chat, paste, go. There you go.

Yes, you can grab that one right there. Yep. And you said we're doing the improved business plan and feature list.

Okay. Yeah, they should both be pasted. If you see— if you look at my screen, if you can see it, you'll see two kind of pasted documents. That kind of, they kind of show up as attachments because it can't deal with, you know, it's just a lot of text, right? Can't deal with 2,000 words. So you're attaching them as documents, and we're going to use them here in the initial thing, and then we're going to use them later too, um, in the project to give it kind of these guardrails, right? Yeah, this is going to be awesome. As soon as you're done with this, as soon as you've gotten to this point, 3 sentences, the one sentence says please please pay close attention and your two, you know, pasted documents.

Click Start Chat. And real quick, so to get it over as a document and not just copy and paste, you—

No, just copy and paste it. Just copy paste it. Okay. It will, yeah, it will kind of attach it and it'll kind of look like it's attaching it as a document just because it's a long text. You could download it and upload it. I mean, that's certainly, it would accomplish the same thing, but For now, just copy, go into Thing, paste it. All right, so your, your, uh, your little box should look just like mine. 3 sentences with a basic description of what's happening.

Please pay close attention to the attached documents. So Jon, you're putting in the feature list first.

Doesn't matter. Um, okay, I just did the feature list first because it was the first one I— when I was scrolling up. I was like, I'll grab this and then I'll grab the other one. Doesn't matter, doesn't matter the order you paste it in. So I'm gonna go ahead and click. Um, how many people have clicked Start Chat? Just one. Okay. All right, I want you guys to— I'm gonna just kind of wait for you guys to catch up to me.

Just sorry for the— so, so I basically, this is my first time in Replit. So I'm starting with please pay close attention to the attached documents and I put in the features and the improved business plan, right?

That's it. Yep.

And then the 3 sentence description.

Yeah.

Oh yeah.

Okay.

Got it.

Yep.

Does that come first, Jon, the 3-sentence description?

Yeah. Or does that not matter really? It probably doesn't matter, but yes, I would put it—

I would. Okay, okay, got it.

Okay.

It's interesting.

Mine's asking, it, it says it's a sophisticated B2B platform. What would you like to prioritize? Public marketing website, client portal, or both in a simplified form?

You're saying that's what Replit's saying? Yeah. Yeah. You share your screen. I'd love to see that. Yep. That sounds slightly different than other things I've seen. All right, cool. So what do you, what do you think?

I mean, what is your opinion? I mean, it's actually, I'm surprised that it picked up on that's exactly kind of what an end goal would be is, you know, what I'm trying to do is create just a website and then Essentially, I eventually need for the business to operate a client portal to engage with. So probably, you know, probably just for this activity, uh, marketing website. Honestly, I, I don't know.

It would take a lot more work, I think, for the client portal to actually integrate. Here's what I'm going to say. This is my opinion. Right. My opinion is say focus on the public marketing website as that's the most important, but build the, the infrastructure for the client portal. We're going to need that also. So the answer really is both. Well, let's put the emphasis on the marketing website today. Yeah. You know, so that's my, that's my opinion. Yeah. So yeah, go ahead and push start, guys.

Get your things. Yeah, go ahead. Jon, I got the 4, they gave me 4 questions. So yeah, it has just some questions. So I guess I should answer them before it can move on, right?

Yep. Yeah. Go when you— yeah, when you tell it what you're doing, just go ahead and click start the chat, guys. And then, and then chat, chat for a second. Yeah. Okay. And then, and then when it comes up, guys, it's going to give you an option to say build a visual preview or build the whole thing. I would say build the whole thing. Okay. And then we're going to take a 15-minute break while it builds and works. You know. All right, is everybody, has everybody started their, um, their chat now?

Yep.

Yeah. Hey guys, it's 11:14. Let's go. We'll, we'll give you an extra minute. Okay, so 11:30, we'll be, we'll be back. So I'm gonna, I'll turn off my video and stuff here and then, uh, but I'll just keep the meeting up and, and we'll, uh, we'll resume. Hopefully by that time everybody has some sort of website that kind of looks cool. We'll start checking them out. Thanks, guys. Thank you. All right, guys.

Welcome back.

How are your apps looking?

Still building for me. Yeah, still building here.

Yeah, when you tell it to do like the full thing, like, yeah, it does, it does take a second, but it looks super cool what, what it's, you know, built so far.

I mean, it's, it's unbelievable that I've been able to do that kind of thing. Yeah, for sure. So guys, while it's, while it's building here, I want to have, I want to have a little bit more of a discussion. So we watched Eric Schmidt earlier and talked about how, you know, programmers are all going to be replaced. But I want to, I want to broaden just a little bit. We can talk about that specifically, but what is— what kind of impact is this going to have on, on the world? I want to hear from the crowd. What do you guys think? What are your thoughts when you hear and you see some of what you're doing right now and you hear Eric Schmidt talk about it and just like where, where this is all going? What is— how is this going to change the world?

Or even just your own life? I think that's a good question. I, it's, I don't know if there's, I've been trying to find like really good ethical thought leaders on AI that lead with like a humanitarian approach and not just, you know, Altman and Musk and Meta just saying, hey, you know, we got 4 or 5 companies that are going to gobble up the, you know, language models and basically rule the, the world, right? And I haven't really been successful, like, what does this mean for everybody, right? Like, um, in my world where you have all of our customer support centers, so call centers, like level 1 and 2 support is essentially going to be eliminated in the next, or could be eliminated.

It's going to be human choice.

You just have to install it at this point. Yeah, yeah, yeah. Right. There's some, like, some technicalities and some accuracy details. And especially, like, in my world where there's a lot more compliance, like, sure, we have to be a little bit more— finance is going to lag behind everybody else for that, just for compliance reasons. But, you know, like, what happens when you lose all these different types of people? Like, there's a lot of good that comes from it. Like, for a customer, I don't want to go sit in another call tree. I just want to like get responses. And so I'm excited for that. But then the humanitarian side of like what happens to all these different people that are being replaced and can be replaced within the next few years on that wave. And I, I don't, I don't know anyone who's really talking about that potential solutions to that. So a lot, a lot of cool stuff for entrepreneurs, but a lot of scariness for just the general public.

I agree with that. Cause, and Justin, I want to go to you next, but I'm going to respond to Cody really quick. I think your last sentence there is spot on. Entrepreneurs, it's like a, it's fantasy land, right? Like you can do anything now. You just launch and you don't have to go hire people, right? Like, like I built GenAIPI writing over 100,000 lines of code in a weekend, launched a multimillion-dollar business. I didn't have to hire any developers. I used to have to hire developers. I used to have to, you know, there's this whole cycle of, okay, we need funding. And so you've got VCs and family offices and angel investors. And all this stuff, and that creates value, you know, or risk and, you know, whatever. But there's an economy there, right? That then is all of a sudden like, wait, I don't need that anymore. Now I could just go and I can bring that to life. I don't have to hire that developer, right? My CTO, I actually just hired my CTO last week, last Thursday, just hired him finally. I'm super excited. He's, I've worked with him since my last, he was my first CTO in 2016 and amazing. So I've worked with him for about a decade, but But I built this whole thing and I was like, Chris, I don't need you. And he's like, I get it, man. You're good. Like, it's all cool. And he's been doing something else for the last year, year and a half. And then finally I was like, okay, hey, I think I'm ready for you to come in and kind of take over the technical side. Like I can do it, but I need someone else to do it now. I need to focus on other things. And so there's still a role, but it's dramatically shifted. I just, you know, maybe I used to need to hire 17 developers. Now I only need 3. So, so yeah, it's It's very, very good for entrepreneurs. But, you know, all due respect to, you know, a lot of people out there, there's a lot of people out there that just, they're just really not capable of doing what we're sitting here doing right now and building a business and launching it. Like they, they need a call center job and, you know, where do they go? Really? Where do they go? Like I have a call center company that just hired me. I'm doing the kickoff call. It was at 1:30 after we finished this thing here. And they told me, they said, "Hey, we employ like 700 people and we want to get that down to 300." Like, okay, I understand why you as the business owner want to do that and you're going to do that because your competitors are doing that and you have to be competitive and you have to cut costs and you got to push profitability. And you literally have, as a CEO, you have a fiduciary responsibility to drive profitability for the shareholders. Like, legally, you can get sued. Sued if you don't do that, right? And, you know, I have been sued for that. I've never lost a lawsuit, but I have been sued for like, hey, did you make this decision like for the good of the company or was it for the good of you? You know, that type of thing. And so it's tough. It's a tough answer. I think there's a lot of discussion around that.

But anyway, let's go to Justin. Go ahead. No, first regarding the Replit, it says I need an API, ChatGPT API key.

So I guess I have to figure that out. Will, let's, let's do that. Let's actually pause our discussion for just a second and let's help you. So go ahead and share your screen and I'll walk you right through it.

It's very easy. It's actually quite easy. Yeah, I remember, I remember having to do that when I did my little app and I can't remember how to do it, but I figured it out.

Yeah. Yeah, it's not hard. Go and share your screen, Justin. We'll walk you, we'll walk you through it and everybody can learn. So an API, guys, an API, for those of you that don't know, Maybe you do, but it's very simple. It's a connector. It just connects one service to another, one program to another, one piece of data to somewhere else. Okay, so in this case, we need to connect into, uh, to how to do this. So go back to Replit where it's asking you for the key. Yeah, so you asked ChatGPT, but what you could have done is just scroll up a little bit. It tells you exactly how to do it. If you haven't created one yet, you can get an API key from here. So yeah, yeah, Replit is very good. You shouldn't ever have to talk to ChatGPT about what's happening in Replit because Replit is a very— now, um, you can't log in here. Oh yeah, you can continue with Google. You don't have an OpenAI account yet. You have a ChatGPT account but not an OpenAI account. Right? And so they're a little bit different, but go ahead. This will work, right? Go ahead and click in here and it'll take you through the signup process. And okay, cool. So now we're gonna just create a new secret key. So go ahead and click that button. Now in a best practice world, you're not gonna be sharing your screen while you're creating a secret key. Luckily it's gonna be super long. We won't be able to copy it and we don't need, we don't care. So go ahead and click secret key. Name is optional. You don't need to do it, but just go ahead and click it. And then it's just going to give you this cool thing. So click copy and head over into Replit and paste it. Then click continue. And then it's going to say, do you want to save this? Add to account secrets? Yes. Go ahead and click add to account secrets. Perfect. Okay. And you're done. Now you're not actually quite all the way done. There's one more thing you got to do. Go ahead back to ChatGPT to the, or sorry, to the OpenAI tab. And click Done. Okay, now what you have to do is you have to click Start Building up there in the top right.

Yeah, good.

Yeah, you can, you can call it whatever you want to call it here. Go ahead and fill this out just so that Perfect. And then, yeah, you don't even need to fill those things out. I think just click Create Organization and then click I'll Invite My Team Later down at the bottom. And you can generate an API key if you want. Do I need it for anything? You— well, just click I'll Do This Later, actually. See, just click I'll Do This Later. But what you do need to do is this. Okay? This is where we'll have you stop sharing your screen. You do need to add $5 of credits. If you don't, it won't work. So got it. Okay. Yeah, you'll be like, why isn't it working, right? So, so guys, APIs, most APIs cost money. Um, many are free, right? They have like a free tier up to like a certain number of credits, right? Um, but even though he's adding a $5, you know, whatever in there, a budget, it will take him a month or 2 months or 3 months maybe to use it. Like, it's, it's like fractions of a penny per request. It's very, very, very, very, very small costs. APIs are built for scale, right? They want you to be able to have thousands of people using it. And if you have thousands of people using it, you probably have customers as well, right? And so, um, and again, some, some APIs are free.

The ESPN API, for example, totally free, and it can do cool stuff, really cool stuff. So, anyway. What about the Google APIs, Jon? Are they free too? Because I actually had that thing come up with a Google API and I just said just ignore for now and continue building MVP. But is Google free or do you have to go into Google?

So, yeah, you'll need to go in. You'll need to make sure you set up an account with their API and their— what's called their, I think, Google console. And no, it's not free, but they do give you a bunch of credits to start with. Yeah. Okay. So the testing phase is usually going to be free, right? And guys, I love, you know, there's a lot of really great APIs out there. Like, I'll show you another one, weatherapi.com. So sometimes weather is a great thing to throw into an app for whatever reason. You might need the weather or something like that. But if you click on their pricing, you'll see, you know, you get a million calls per month for free, right? And then after a million, you'd have to start paying, and then, you know, you're at $7 a month, right? Again, if you're using weather for something and you're getting more than a million calls per month, you're making money. You can afford your $7, or you can even afford your $35, right? You can afford that kind of stuff. But if you're using an AI, Anthropic API, Groq API, OpenAI API, any of these systems out there, they're going to require that you put like a $5 budget or something just to like get it started. But then your usage is again very, very, very minimal. So not that difficult to do. But APIs, guys, are really awesome. They're really important. APIs really, in my opinion, are what start to make your app unique, right? Because you create your interface that does what it does, But then you start having these unique combinations where you connect to Google Docs and then you connect to Google Maps, or you connect to the weather, you connect to the ESPN API, or you connect to whatever API it is you're wanting to do— HubSpot, Salesforce, um, Instagram, you know, whatever it is you're trying to do. And now you're saying, okay, now when I do this, it automatically does this and this and this and this, and you're starting to create this automated kind of value chain that creates unique value. So that's what APIs are all about. All right, let's see if mine— yeah, mine's still, uh, mine's still going. That's okay, that's okay. Um, all right, um, let's continue this conversation for a minute while we continue to wait. I, I really, I really want to hear some more thoughts. Thank you, Cody, for yours. Um, how, how else is this going to change the world? What are we going to do? What do we do about all those call center people?

Where are they going to move to? I had some thoughts just about, um, yeah, in general. I remember, what did they call it, it was like the WWW wave, like once the internet kind of became huge. For me, I was born in '92, I was just a kid when that was going through, but I remember me and a lot of friends, like a lot of millionaires were made in that first wave of the internet and I always felt like I got gypped and left out, but this feels like that is happening now. And, uh, the only barrier of entry is figuring out how that changes things and getting ahead of it. And, uh, that's the thing that's hard. I mean, the first thing I think of is, uh, data centers, but like, I'm probably $50 million short of being a stakeholder in a data center, but like when, when there's no barrier to services that used to cost money, like how do you provide value and what does that look like? And I just need— I probably just need a few hours to sit down and think about what that actually means and how you get ahead of that wave because it seems like it's cresting. But yeah, you're probably a better person to ask about that, Jon, just like where you see it going.

I've thought about it a lot, but I'd love to hear your thoughts. Right. And I certainly have my opinions, which they're, you know, that's all they are.

I'll give you my view because my presentation next month is right around this topic. I'm presenting on, on what transformation means to the financial services industry aligned with regular regulations as well. And in a lot of ways, due to the nature of how fast these changes are occurring and how slow, or traditionally slow, the adoption is in banks. Uh, there may be, you know, the shiny object, and I just went through this of trying to wave the shiny object to draw attention to it, but to actually get that type of momentum and have the proper governance around it, I think that's going to be slow-moving for certain industries like healthcare, like financial services and some others, right? But then there's ones that you mentioned where there's really true low-hanging fruit. You can go out there and replace people like that. And it just becomes a matter of how quickly you get individuals like yourself, what I'm striving for, where you can help kind of curate the pathway for these companies to have a softer landing with their employees, to use these tools as best as possible. I think before I mentioned to make them the Tony Stark that could become Iron Man because they'll have the capabilities at their fingertips. So their productivity in theory should go up. However, to all, I think you made your major point is you won't need as many of them in order to do that because the growth from the revenue or the productivity, it'll probably hit a limit because of the people, not because of the technology. Uh, secondarily, I'm also making a business out of this. My, my I'll show— it looks like that worked with the API, but I created a framework called Make It Make Sense, MIMS. And essentially it's how to take this and think about it in a more simplistic way, because one of the biggest breakdowns is from a cultural perspective for people to understand that they should not necessarily fear the tool. They should embrace the tool and understand how to use it. So more people should be going to your classes and we should propagate this more broadly so folks understand how to use it..

And I think right now they just don't besides Googling or using ChatGPT like Google. Yep. Yeah. A lot of people use ChatGPT right now just as like, oh, this is the next search engine.

Right?

And it's, it's just not. Right? When you start to get into, and it's because, you know, ChatGPT can't do what we're doing in Replit. Right? And I'm gonna show you some other platforms over the next day or 2 that, that are even, you know, even more impressive. Right? It's just like, Holy cow, I just don't have to spend time on that anymore, right? The standards are being raised. One cool stat says 60% of the jobs that we have today did not exist in 1940, right? And now, that's a long time period. That's 85 years. But the bottom line is the world changes, right? I mean, think if you talk to somebody back in the year 1850 and you said, you know what, there's going to be a time when only 1 out of 100 people are farmers. And people would be like, no, everyone's got to eat. Like, everyone's got to have their own. Like, that's impossible, right? And so, you know, obviously I'm sure there were people that were against the tractor, right? And there were definitely people against electricity and against the Industrial Revolution. And oh man, we don't want machines to take our jobs, right? But obviously we've made it, and our, you know, We can't even imagine having to do certain things these days now because they're just done for us, right? When the internet came out, like you were talking about, a lot of people are like, there's no way anyone's ever going to bank online. No one wants to just move some digits around. They want to feel the cash. They want to talk to a banker. It's relationship business. It's this, it's that, it's whatever, right? And so, people push back on it. And so, you are right too, Justin, right? It's going to be the banking and financial health. Healthcare. There are some industries that will lag partially from a regulatory standpoint and also because there's huge lobbies and huge things that are going to hold it back artificially and do that kind of stuff. And then there's others where San Francisco and Silicon Valley, Silicon Slopes here in Utah, the tech world, the tech world is not going to slow down and they are going to just change. This year so far, on average, if you just kind of average it out across the whole thing, jobs a day have been lost to AI. Okay. And, but I saw a really, really interesting, somebody sent me the article yesterday. I'm gonna find it actually, cuz I think it's worth, it's worth pointing out and, and I need to, I need to really read it. I actually only read the, the title here basically, but check this out. IBM laid off 8,000 employees to replace them with AI. But what they didn't expect was having to rehire as many due to AI. Right? And so this was interesting. It's not that they're like, oh crap, we need those people back. They actually went through unexpected AI-driven growth. So their company did. It was a positive move for their revenue growth. But then it brought so much more business that they had to then rehire those people to do higher-level things, right? So they didn't rehire them to do the things that AI was, you know, had replaced those workers for, right? That worked actually. But what happened was the growth of the job just changed, right? Think about it. I like to think about McDonald's, okay? There is right now, I can tell you McDonald's is working on the one-man McDonald's, right? The whole thing can be run by one person, right? Why? Well, because flipping burgers is a pretty robotic task. Why not just have some robots to do that, right? And hey, we need to drop some fries in, and then we need to pick this up and move it over and drop it into things. It's very like— it, it, it's like a vending machine, right? It just— it knows what it needs to do, and it's very common, right? The Big Mac is created the same way every time, at least you hope it is, right? When it's done by a robot, it's probably even going to be better than a human, right? Because it's going to be perfect. It's like, perfect, we got this right. AI is going to take your order You know, maybe if you go in, you'll use a kiosk or something like that, but then your order will just pop out. It'll probably be right, right? It's, it's not going to be messed up. And so, um, what does that leave humans to do? Well, it leaves humans to interact more with customers, maybe. It leaves humans maybe to focus more on expanding, on marketing, on whatever it is, right? Um, there's certain industries or certain types of industries that are going to be highly automated away, and because of that they will become cheaper. AI, the end result 15, 20 years from now when AI is really, really embedded into our society is there will be less poverty and more people in homes. Just that's how it's going to be. Are we going to have to have UBI? I don't know. Maybe. That's one solution. Universal basic income is definitely something that people are thinking about. What happens if all those call center people just can't be reskilled and retrained fast enough to be able to put somewhere and we're just going to go to 20%, 30% unemployment? That doesn't seem like a great outcome. And that's part of what society is about. Yes, I do believe huge in personal responsibility. We are responsible for feeding our families. But society also can act as a safety net. For those that can't for whatever reason, right? Maybe they have mental issues, maybe they have physical issues, and we all go through, we all go through, even though it's those of us that are healthy, we sometimes have mental issues and physical issues, right? I've got, I've got scars all over my body. I've had 16 surgeries over the last 10 years, and sometimes I'm just down, right? Um, because, you know, I'm an athlete and maybe I'm not very good at being an athlete because I just hurt myself all the time. Um, no, I just, my, my knees wear out and stuff. But anyway, bottom line is, guys, We got to figure out as a society how to help those that can't quite figure it out. I do predict pain. I predict a painful transition. 10 to 15, 20 years from now, I can picture the, the better life for everybody. But right now, more jobs are being lost. Now, IBM's a pretty interesting example, but not all of them, you know, that call center that, you know, has called me up and said, hey, I want you to do this. That call center is, is something that, that is going to be a net loss.

Right.

And someone's going to have to figure out where those people go. Those people, again, it's their responsibility. I really do believe that it's not up to the government. It's not up to your neighbor. It's not up to whoever. Like, you ultimately have the responsibility. But again, we as a society, as our, as our social safety nets, I think we have the obligation to go and, and see what we can figure out. So, um, all right, other thoughts, guys? Any, any other thoughts on this before we kind of move on? And obviously keep bringing them up. You're just like, hey, you know what, I want to share something. So, um, feel free to do that. Um, I'm going to go ahead and, um, share my screen.

Can I ask you one question, Jon? When you present, you know, your proposal to these companies that are coming to you Is there any point in which you're providing an insight to that? You're doing it from a savings or an ROI perspective, but is that discussed?

Absolutely. Big time. Totally. By the way, guys, if any of you want to do what I do, right, whether you're teaching classes like this or working with, you know, I do much higher-level stuff than this. Consulting, and I do fractional chief AI and leadership stuff, all kinds of fun stuff. Um, take my business model, copy my website, go do it, go build it, right? There's about 100 million businesses that need to learn this, and I can't teach them all. This is the biggest opportunity ever for any of our lifetimes, right? And it's bigger than the internet. It's 10 times bigger than the internet. Um, it's bigger than electricity. It is bigger than anything. And the speed at which it is getting bigger is unprecedented. And so ask me anything. I am an open book. I will tell you exactly how I make every dollar that I do. If you're like, hey, I love Jon's business, I'm going to go do that. That's great. Become an AI expert. Okay. All right. So I've got now a first version of my Sunny here, and I want to see your projects too. So what we do here, guys, in Replit, this is where it starts to get fun. Now, the rest of what we've done, I think, is actually fun too. But this is where we start to say, okay, Okay, let's bring this to life. So now I'm gonna click sign in to get started and use Replit auth. What I'm gonna do, you know what, I'm just gonna say login. It's okay, I'll fix that later. I actually hate Replit's auth system. I think it's really, really dumb, really bad. And but whatever, that's okay. Here we are. And now I'm into my dashboard. Okay, so immediately, anybody see something weird? Yeah, header. Good morning, John. Good morning, John. All right, cool. So we've got, we've got kind of some double stuff here. So I don't know what is actually going on. Sometimes I can be like, okay, this button is off or this specific thing is off. You know, whatever. But what I'm, but here I'm just going to say we're going to do a screenshot. So I'm just going to go, you know, Shift Command 4 and do a screenshot here. I'm going to say why Is there a double? Good morning, Jon. Let's make sure we get this fixed. All right. I can also chat with Sunny. Um, tell me about the sun. I'm experiencing technical difficulties right now. Cool. Awesome. So yeah, guys, you, you really just kind of start pushing through. Yeah, it looks like that's in the header. And on the dashboard. Because when I went to chat with Sunny, we kept this— some of the spacing's off here. We've got a, a 3 up here. I don't know what the 3 is, so that's, that's fun, whatever that means. Yeah, projects isn't created yet, and that's fine, partially because we're going to be connecting, I think, into, um, like QuickBooks and things like that, and just hasn't connected any of that stuff yet. All right, I can see the duplicate greeting. Let me fix this. It's interesting. New dashboard. Yep, none of that works yet. That's okay. So when you get to here, guys, here's what's, here's what's going to happen. You're going to just start acting like a user, okay? You, you all know, like, how many of you guys have more than one app on your phone? Yeah, everybody's gonna be raising their hands here, right? How many of you guys have used multiple websites? You guys have used dozens or hundreds or thousands of websites and apps throughout your life, and you generally kind of know what to expect, right? A chat bubble is usually in the bottom right, so that kind of feels good, right? If the chat bubble was up here in the top left, you might be like, it kind of feels weird, can we move the chat bubble to the bottom right, right? Or, you know, maybe the sign out bottom left here, that kind of feels right, feels good. I can also click right here, guys, I can look in the screen size instead of being full size desktop, I can switch it to a phone, right? And so now I can look at it and examine it. Oh, hey, look, I've got the menu in the way of this, and so I'll have to tell it that. But you just kind of act like a user and go through and say, okay, does this feel weird? Do I like this? Does this— when I click the button, does it do something? Like if I click View Recommendations, it doesn't do anything. When I click Ask Sunny a Question, it doesn't do anything. Okay, so I need to— yeah, I need to fix that. Okay. Here's— oh, now we've lost our head of the chat. It's interesting. New conversation, new conversation. It's creating tons of new conversations here. So obviously there's some problems that we're going to need to go through. And it doesn't mean that it did it wrong necessarily, guys. It— what it really means is that it probably hasn't done that thing yet. It went so far and then it stopped and said, okay, what should I— what should I do next? Okay, now this little 3 right here, I have no idea what it is, so we're going to get rid of it. I'm going to click this little select element button right here, and I click that, and then I can come up here to this little 3, and I can say, what the heck is this? Delete it. Okay, and, and again, I'm not being technical here. This is not a, um, this is not something that only technical people can do. The cool thing about AI is the language of AI is whatever language you speak.

Right?

If you speak Spanish, it speaks Spanish. If you speak English, it speaks English. If you speak Southern, it speaks Southern right back to you, okay? Um, if you speak Morse code, it will understand that too. It can understand whatever you want to talk to it in, okay? And so because of that, you don't have to be afraid. You don't have to be afraid of it. Just talk to it in the same way you would talk to a developer. If your developer brought you this dashboard, you'd be like, hey, Dude, this looks great. Like, that's awesome. What the heck is this little 3? Oh, I don't know how that got there. Let me delete it, you know? And so, you know, we'll delete it. We'll go through and doing that. But you would fix this just in the exact same way that you would anything else. Now, one of the things that it's doing right here right now is it's actually actively looking at and it's starting up a browser to begin testing. It's going to test it and then you can look at my thoughts will show up here. You can see what it actually thinks as it's working through things. You can turn that off if you want. Okay, when I click Agent Tools, you've got App Testing on autonomy level: low, medium, high, or max. Okay, low is going to generate and execute on task lists. It can actually do quite a bit. Okay, at medium, it adds this reviews latest code changes and fixes issues found. High, it's going to expand the review scope to entire app, so it's going to take time to look at everything holistically. And then max, it's going to do a lot of stuff. Now with max, you're going to get, um, you're going to get pretty significant, uh, not delays, but you're gonna have to wait. You're gonna be waiting a long time while it works on tasks. Max is where it can work 200 minutes on its own. Okay. Now it won't use, it won't work 200 minutes on its own for everything, but if you gave it like a really complex thing, Right? Lots of tasks, lots of things. Build the entire CRM, make sure every single functionality works, build the privacy policies, build the connections, connect the APIs, do everything. Like, you just tell it to do everything. Then yeah, it might take you a while. Okay. I like to hang out in medium and I like to turn off app testing when I'm moving faster. Okay. Once I'm getting better and better and I'm like, okay, I know what I need to do here. I need to move this. Okay. I want you to do that now. Now I want you to do that. I'm gonna, I'm gonna turn it down. But back when I built GenAIPI, none of this stuff existed. It was not there. I will also say this, this is, this is important. These options may result in higher costs. Yeah, you put it on max. Yeah, you can expect to knock, you know, you'll blow through $100 of credits in a day if you work for like 8 hours at max. It's going to cost you some money. Okay. So anyway, user beware, user beware. There's some other things you can do here. High power model. High power model is something I would only do 5x the cost, right? I would only do higher power model if you're running into a very complex problem and it can't figure it out. And I'm only going to use higher power model with the plan. Okay, so over here you've got these, you've got this function, you've got this little function here. Let me get rid of this annotate button. Okay, build means it's going to make test iterate, it's going to do stuff, it's going to be very action-oriented. Awesome. Okay, then you've got this plan, you know, thing where it's just ask questions, plan your work. I use this so much. Okay, I actually almost always use it when I want it to do something. I'm going to say, okay, I want to get this functionality working now. Tell me your plan to make this work. And then it's going to go through, it's going to think, it's going to look at all the code, it's going to figure everything out. And I'm only going to use the high-powered model. If I know it's very complicated or if I've already tried a couple of times and it's failed, then I would say, okay, I want you to make a plan, figure it out. What is going on? Oh, I've discovered this now. Now that I've used the higher-powered model, then I'll come back and I'll turn that off. I'll turn off high-powered model. I'll come back to like medium or high or low or whatever I want to do. I'll switch it back to build and I'll say, great, do the plan. Right. Because the AI is capable of building at a very high, high level and following a plan., right? But the higher model can devise a better plan. I hope that makes sense and I'm not confusing you guys. And then I actually don't know what edit means, like how that is different than select element. I'm going to try this. Ah, that's cool. That's where that button went. Okay. I wonder, they actually added this back in. They took it away for a minute. Now it's back. This is cool. So edit allows you to just click something and then it brings up a thing here. And so instead of saying view recommendations, You know, you can say view recommendations again, right? And so it can do that. You can change something live. You have to then click save changes and then it will, it will change the code to do that. You can change the font and say, you know what, I want this extra light or, you know, I want to make this really big and I want the margin to be, you know, higher or whatever it is, right? Much higher, zero, whatever it is. Okay, pretty cool. I like to be able to do some of that visual, visual stuff there. Now I'm not going to keep those changes because I don't need to do that. So edit's cool. That's awesome. They added that back in. That used to be tied into the element selector. Man, you guys are so spoiled. Replit is so awesome now. This is really cool. But what you're doing here, guys, is you're trying to get to the point where your app has no, what I like to call, paper cuts. Let's talk about GenAI API. When I launched this, I built it out, I built a website.. But if I come here and I land here and I'm a first-time user, okay, if I try to click on one of these questions, I might quickly figure out, okay, this doesn't, this doesn't actually do anything. It's just kind of whatever, right? But I need to be able to click here. This needs to work. I need to be able to click not now. I need to click here again. I need to click X. Okay, you gotta test everything. Start the test. Okay, that works. I need to click through the test, see what happens, right? Um, says, you know, roughly 3 minutes remaining. Okay, great, whatever. What happens when I get to the end of the test? Does that work? Does it break? You have to click through every single button and you have to make sure that there are no paper cuts. Here's a paper cut that will cause somebody to say, I don't know if I trust this company, right? If you come to the bottom and you click on privacy policy and nothing shows up, okay, or it goes to a 404 or something like that, right? Or maybe you scroll to the footer and you click contact and that doesn't work. Or, you know, obviously you want contact to work. That's how you're going to make money, right? But you click on Instagram and there's not a page there, or there's no posts, right? There's no posts on the page, right? You want there to be posts. They want— I haven't— I actually never ever ever posted here. I haven't posted here in a long time. Um, but anyway, actually I think I posted yesterday. Did I post yesterday? I don't know. Anyway, I was on the news yesterday, so that was fun. But anyway, bottom line is, guys, um You got to go through every button and, and you know how, how it should work, right? The login should work. You need to click, hey, don't have an account. You need to test the forgot password functionality. You need to build all this stuff out. And there's a lot more about forgot password we need to talk about in a second here. But that is the goal here. That is the stage. Once you're to this point in Replit, now it's a matter of does every button work? Let me also caution you and say Sometimes Replit puts a bunch of things in there. Maybe I don't need crew management. Just because the button's here doesn't mean I need to make it work. I could say, you know what, I don't think I need that functionality. Let's remove this button. Gone. Hey, you just simplified. You just removed a bunch of work, probably 2 hours of trying to, you know, perfect a system that you could have spent other time. For reference sake, GenAIPI took me roughly 35 hours to build. Okay, and I had to build a lot of stuff. I had to build out the assessments. I had to build out live courses— or not, sorry, not live courses. I had to build these courses, right? I had to build out all the functionality in them. I had to be able to get all the payment working for it, right? What happens when you try to buy something, right? Processing, processing. Oh, you got to be logged in. Okay, let's log in. Now I'm, I'm an admin, so I already have access, right? But I had to test all this different stuff. Right? If I wasn't an admin, what happens? Okay, then it goes and it forwards to this and all this stuff happens. I then had to build an admin for myself. Here's my users management. Here's schools. I've got a BYU thing that's going now, which is really cool. Here's course content, modules, and lessons. Here's live courses, right? Here's the different schedules. Here's the participants in those schedules. Here's the course content. I actually used to have this. I turned that off. Here's reviews, right? If you send me a review, I can, I can grab this, you know, and I can put it in here and I can say, hey, Cody K., right? Job title, whatever. And I can add those reviews and then I can check the course page and see what those reviews look like. They'll show up right here, right? It's super cool. And I can click right here on reviews and look at all this cool stuff, right? But it takes time to just build all these things out and test them out. And a lot of this stuff didn't exist when I first launched it. Test results, for example, did not exist. But I had somebody say, hey, I did your certification, right? And, and I, but, but I, I didn't get my like certificate and I was like, yeah, I didn't actually build that out yet. I kind of waited for somebody to buy it until I built that functionality out, right? And they were able to go in and take the test and get a score, but they didn't get like the certificate. And so I had to create the whole certificate thing and then I had to create a way to be able to look up that, that certificate, right? And so I could come in here, I can see my test results, I can see their test results. I can click View Certificate and boom, then I can eat. And I had to create this download PDF functionality. I had to make sure that actually downloaded, right? And so it's just a matter of, of going through and saying, does the experience actually work right? There's a lot of technical things behind the scenes that might not be perfect, but my first test is, does it work and does it feel weird when I do anything? Okay, the paper cuts, right? Avoid the paper cuts. If you can get through the whole experience and say, hey, I met, you know, Kalima, right? Whatever your offer is. Okay. Hey, I came from Instagram. I landed on a site. It showed me an offer. It said it was $17. I clicked buy. That gave me access. I got an email confirmation. I got into this thing. I was able to find some information here. I got my value. Hey, I'm not calling Kalima saying, dude, give me my money back. Back, right? I feel like I was able to get through the experience without feeling like I got gypped, right? So, so that's, that's the idea, is you're trying to avoid people saying, uh, I don't trust this, right? No one trusts anything these days, right? Like, people's kind of default, you know, when they land on a website is, is this real, right? And, and then especially if it sounds too good to be true, right? So you get— there's a lot of things you gotta, you gotta work through on that part. And you can work with ChatGPT. ChatGPT can help you work about great marketing, great language, great figuring things out. By the way, for what it's worth, I put every word that, that is on this page— well, why is that flickering like that? It's like there's a fast refresh or something happening. Um, no, I don't want Settings. I did not want that. I wanted Developer Tools. Tools, Developer Tools, Errors. Got some errors here.

That's interesting.

It's just continually fetching the auth, which it shouldn't. It should only do that once. I don't know why it's flickering. Kind of weird. But anyway, guys, we'll teach you how to do some of these things too. Um, what is the 12 issues? What are the issues? 6 errors, 30 info. Anyway, you can go through and you can look at that stuff later. I don't know why it's flickering. I've never seen it do that before.

Maybe it's just the browser right now. So anyway, so, Jon— yeah, yeah, go ahead, Jon. Um, so when you're kind of going through all these paper cuts and maybe, you know, adding a page, adding a section, all that stuff, um, do you, do you just use that, uh, like, do you just chat with the, uh, with the agent?

Do you use the planning tool, or how do you, how do you do it, uh, basically? It's quick, right? So, so here's what I'll do, right? Yeah. I'm going to click here, ask Sonny a question.

Okay.

Here's a paper cut.

Yeah.

This doesn't work. Okay. Yeah. I just, I'm going to go like this. I'm going to say this button doesn't do anything. Got it. Okay. I don't even have to tell it what to do. I'm not sure exactly what it, what it should do.

Right.

I can obviously think through that a little bit more, but I could just be like, hey, this doesn't do anything. Ask Sonny a question. Obviously it needs to be connected to Sonny. We need to make sure Sonny works. And so, yeah, also when I click on this, I, I can't access it. When I click on the chat with Sunny now, it's kind of messed up. So we gotta obviously fix that. Let's try one other button and see if that's a header, a header issue. If I click on projects, well, yeah, that's not gonna work because it's— all right. So cool. Now I've got another problem. Here's a cool thing though. I can click on, you know, whatever. When I click on chat with Sunny. It goes to the page but does not allow me to do anything because most of the content isn't displaying. It's just cut off. And then here I'm actually going to turn on— oh yeah, app testing is on. I'm actually going to let it do that. Now here's a cool thing. I can say go and so it finished that. I'm going to do one more thing here. I'm going to say Integrations. Okay. So the integrations page. So I can just do this. I can say, oh, doesn't work. Make it work. Okay. And I can give it another thing. Right? And then what it does is just adds a queue here. Right? So it's gonna finish its other task first and it's gonna come. So you can be working, testing, planning, doing whatever. And, and, and obviously, you need to be careful if you fill up a queue with a bunch of things that might interact with each other. Because if it changes one thing, then that might change the next project or the next element. So you want to do that. So I like to kind of bounce between things, right? And so maybe I told it to work on the Chat with Sonny. Now I told it to work on the integrations page. Now I'm going to click on notifications. Hey, notifications doesn't work either, right? The other thing I could say is this. This is another thing I could do. And I'm not going to do this because it would take forever and we wouldn't be able to do the class anymore today.

But I could just say, make sure every button, page, and element on the entire site works.

That's all you need to do to get it to, to do full regression testing. Yep. That's incredible. Yep. Crazy. And so then it will go through, it will build the finance page, it'll build the crew management page, it'll build the integrations page, it'll build the notifications page, it'll make sure the sign out So make sure the knowledge base is there. We'll do all of that stuff. Now you're still gonna have to go through and be like, okay, does this work in the way I want it to? Yeah. Maybe, maybe not. Whatever. Let's change it. Hey, let's redesign this. Right? And by the way, for design and stuff like that, guys, it's super easy. I usually hold design till the end. Right? Because you're gonna change a bunch of things. Whatever you design in the beginning, you're gonna end up changing anyway. So just make sure the functionality's there, then get the design. I actually kind of super love this design that it gave me. It's kind of really good with Sunline Landscape. Like this is Sunline Landscape, just so that you guys know, landscapes.com, right? They kind of have a dark theme, but this kind of yellow-orange color, and it was able to identify that, right? But it's kind of dark, and I'm like, I want my dashboard to be bright and pretty. It's named Sunny, right? The logo is terrible. It tried to make a sun with lines and circles, and it did not do that well. We'll replace the logo later. That's again another thing that we can do. Another thing you can do is, say, hey, I want this to feel more like Salesforce design, or I want this to feel more like apple.com and they're like really futuristic design, right? And it can be like, great, I'll look at apple.com and see what they're doing and I'll apply it to your site. So cool that you can just rip off other people's designs and take what you like, change it, make it your own. But yeah, super cool. So, all right. Wow. They're going, it's going fast here. All right. Integrations page added a dedicated page for managing integration. There's sync status. Okay, so here's what I'm gonna say. It just, it just told me it did that. This is what I see when I go to the integrations page. Okay, I know what happened. It didn't connect the route. It probably built the page and it didn't connect the route, and that's, you know, its own fault. Uh-oh, I accidentally gave that on max autonomy, so that might take a second. We'll come back to medium here. Um, it shouldn't be that bad. Again, max autonomy just means it's going to do everything it needs to do and, and it won't ask any questions. It's just going to try to fill in all the gaps. And so you don't want to use max autonomy, um, later on. I actually really love using it early on when I'm like, I just need everything to work and then I'm going to go back and do it. It's kind of like, how many of you guys have ever built a PowerPoint presentation and the hardest part of that presentation is just starting, right? Like once you have like 8 slide decks or pages, whatever, like it's actually pretty easy, um, to just go in. I mean, I'm going to tweak this, I'm going to do this and whatever. It's great to have an assistant or someone give it to you and be like, hey, let's change this, let's change this. Starting is just so hard. And so I'm just like, okay, just, just build what you think it could be. And then I'm going to come back in and, and figure this out. So anyway, Yeah. Can you guys see that? All right, so the idea that I'm building out on, and luckily, I mean, I've been talking with ChatGPT for months about this product offering that I've been thinking about. It's called Origin Genealogy. So I'm offering reports to be sent to people. Essentially, they will submit some of their ancestral information, and Right. And so you'll get better with this. Like any tool, you're going to get better with it with, with time. All right. So let's let this start up here and we're going to show a couple other things that are really, really important. Okay. One of those things is already open here. I'm going to close it so you can see how I open it. But up here on the right, you've got tools and files. Okay. And one of the things that we want is a database. Okay. You might click on that in your app. And it will say there's no database. Would you like a database? Yes, you need a database. Okay. But the database is where all of your data lives. Okay. Your users, for example, are going to live in here. All right. We've got Jon at FlowState Leadership. This is one of my things. We've got a test at Sunline Landscapes, whatever. Here's some names. It's created a couple, you know, basic users for me. But one of the things that you want to do when you go into the database, guys, the database is where all of your information lives and it's just a bunch of tables, tables and, you know, rows and columns. Of information, but these rows and columns are the things that populate everything that happens out in your app, right? So let me give a more complete example of that using GenAI API, right? So here's my, here's my, you know, product, okay? If my database is very significant, right? There's tons of things here, tons of users, tons and tons and tons of data in here. Luckily, all the passwords are all hashed, hashed and stuff. I can't see that stuff. You want to make sure that you don't know other people's passwords. You don't want that liability. You want to make sure it's hashed and using good stuff. It will automatically do a lot of that stuff for you. But I have so many different things in here. Okay, so let's look at one example of something that is database-driven. Okay, so a lot of people will do something, uh, when they— especially when you're creating with Replit and some of these live coding tools, sometimes it will say, okay, The user wants the site to look like this and it wants the price of this product to be this. Then what it will do is it will do something called hard coding. Hard coding means it's going to put it into the code and if you want to change that thing, like this word should is part of the code. I can't just go in and change it in the database or anything like that. It's hard coded into the code of this website. Now, some things that's okay. That's actually okay to do that. We can change the code later. But what if I scroll down here and look at something like, like this? Why get AI proficiency certified? It's $99 per certificate. Okay. So what if I want to change the price of that to say $59? Okay. And see if I get more purchases at that price. Okay. Well, it's $99 here. I can click on certification. It says $99 right here. And $99 right here. Okay, and there might be another spot or two, but there's, there's 3 spots already. So if it's hard-coded in, if I decide I want to change the price to $59, I'm gonna have to go in, I'm gonna say, all right, please change the price here, here, and here to $59. Well, that's not very efficient. Okay, the better way to do that is to be able to be database-driven. And man, this used to be a list and now these columns. Let's see here, uh, certification tests. I think that's where it is. So I only have one certification test and I've got, you know, some different things here. Here's the price, but because it's database driven, I can change the price to $59. I can save that change and now I come back to my preview and I'll refresh it. And now it's going to load that database driven element and now it's $59. Amazing. Okay, $59. I can come back to the homepage where I had it down here. $59. Super cool, right? So that is good design, good database usage, right? Your database is where all of your information is going to store that is then going to tell— it's going to inform what is displayed here, right? If you hardcode data here, that's useless. What you want this to be doing— because right now it is hardcoded, it's hardcoded fake data, mock data right now. Because they just wanted to show how it works. At some point, we're going to switch this to database data, and it's actually probably going to pull through my QuickBooks or Aspire or Monday or whatever APIs and data. It's going to pull that in. It's going to sync it, right? Record synced. Record synced means, yeah, there's data in QuickBooks, but there's also going to be data in the database that's going to show some of those types of things. And so it takes time for your system to kind of understand what's going on here. For example, if you look at the database right now, I don't see any information in here related to finance. There's no information about like— so there, well, there might be some. Here's project, here's ID, name, description, status, budget amount, actual costs, right? Those two numbers are something that are probably going to show up somewhere, but how are those calculated? Where is it pulling the data from? What's the math? What's the equation behind it, right? I don't actually have to figure that out. But I do need to be able to tell it what I want. So I need to be able to say, okay, in QuickBooks, under this section, this is where you're going to find projects. Here's where you're going to find the budget amount. Here's where you're going to find our actual costs at the end of the day. Right? Or maybe one's pulling from QuickBooks or one's pulling from Aspire, which is their project management tool, and one's pulling from QuickBooks. And so bottom line is you're just going to have to understand it's like, it's like putting together a puzzle. Right? And it's not a hard puzzle because you don't actually have to put the piece where it goes. You just have to know that the piece exists and then it needs to go somewhere. Okay, so this can feel a little bit overwhelming, guys. I know it can, especially if you've never dealt with a database before. But the thing I love about Replit is that Replit makes it so easy to do all of the things you need to do inside of your app. I'm Kalim, I'll get you in just 2 seconds here. If you try to use other vibe coding tools like Cloud Code, which you've used, Kalim, right? You have to try to set up Supabase and/or use Vercel and use all these other things to be able to deploy and have databases and things. You have to use all these systems to tie them together. Replit has all of this stuff here. You can publish here, which means you're hosting it right here in Replit. Your Secrets Manager, this is your APIs, all your API keys. You've got, what else is important? Git, you can push this to GitHub and save version control. You can do all these crazy things. You don't have to know how to do all this stuff. You don't have to know how to do all this stuff at once. Right now, all you need to know is I've got my app, I've got data. Data is either hard-coded or it's being pulled from somewhere, either the database or somewhere external like an API. That's it.

That's really how all apps work.

Okay.

Kalima, go ahead. Yeah, you kind of answered part of that. My question was gonna be two-part, uh, whether or not Replit is where the website is hosted. And also I had a question about, um, like for e-commerce purposes, if Replit is just the end-all be-all for, uh, adjusting that, or in your experience, if you've seen it better to like add in some sort of integration with existing e-commerce stuff, or is it all just one-stop shop?

It's a great question, and it's actually part of our discussion we're going to have tomorrow called Surf, Don't Swim. Um, but the quick answer is if you're selling a widget or piano music, you know, or, you know, whatever it is, you might do better to just go to Shopify. Shopify has taken care of all this stuff for you. Do you want to have to maintain an entire, you know, system, e-commerce system? Do you like— for example, here's a very, very simple example. What happens if I forget my password here right now? How am I going to get my password? Well, I'm going to need to be able to push a forgot password button. What happens with a forgot password button? Usually it says, what's your email address? Great, then it's going to send an email. Well, is Replit capable of just sending an email? Like, is this website connected to email? No, we're going to have to use an API. We're going to have to connect to some sort of email system, right? So I could say that right now. I say, hey, um, I don't want to do that because I have to do that, but basically I would say, um, let's build a, you know, forgot a, a forgot password a working forgot password functionality. I also want to change the entire auth system to JWT instead of using Replit's auth. I'm going to go into plan mode for a second here. We're just going to talk to it because I think this is a great thing to look at. This is actually the next thing we need to work on anyway, is how to really think about that full experience, and then how do I actually communicate and start to bring it to life? I personally, Kalima, for you would be building in Replit. You want the flexibility of being able to control exactly what you're building, exactly what you're selling. You're not just selling a widget, you're selling access to maybe FamilySearch or maybe Ancestry or some APIs through there and some guides and different things like that, right? It will be very, very difficult to do that in Shopify. Shopify is cool if you fit into like this bucket, right? Right. You want to go outside of the bucket and do something custom, now you have to know how to code. Like they don't have AI over there to help you out in the same way. So this allows you to just do exactly what you want. Okay. So here's the, here's the plan, right? It says, hey, I understand you want to migrate from RapidAuth to JWT. All right, cool. Significant architectural change will affect multiple parts of the application. Before I create a plan, let me quickly understand your requirements. Do you want users to sign up with email password instead of OAuth? Yes. For forgot password, should we send password reset emails? We would need an email service like SendGrid. So it knows, it knows that you need this. And not only that, just like when I think it was just, yeah, Justin was asking about how do I set up the OpenAI API? It will tell you, click here, grab the API key, paste it here. It'll tell you exactly how to do it. Okay. And then you want to keep the OAuth options, Google, GitHub, etc., go to pure email password. Okay. So one email password is what we want, but we may add OAuth back later, although we won't use Replit's version. We will need something like resend for forgot password emails. And yeah, okay. Anyway. Go, please. Please wait for the agent to finish working or turn off plan mode. What's JWT? JWT is JSON Web Token, right? But it's basically, um, so right now the, the sign out, sign in process here with this— by the way, you also have to check your sign out, right? A lot of people don't, don't do that. Sign in to get started. Um, it brings you to this Replit auth testing thing, and it's just weird. Replit has its own auth system that will allow you to log in with like Google and things like that. And, and that's kind of fine, but it, it, I just really don't like it. I'm not a fan of their auth system. Okay. And so what I want is I just want an email and a password, or sorry, yeah, or sorry, a username and a password, right? So this would be— whoops, I didn't mean to do admin. I meant to go, um, GenAI API and click on the admin. Actually, no, we'll just go like this. If I just go log in here, I just have a username and a password. That's all I want. Right? I just want something very simple like that. I don't want to do OAuth and log in with Google, log in with Facebook. Like, I just want JWT. That's what this is.

So JSON Web Token, it's just what it's called when you have just a plain username and password.

Okay. Does it cost you money to have the OAuth? No. Okay. No, it doesn't. Yeah, you know, some of, some of what you have to understand also with, um, Wow, we just did a lot of stuff here. Questions for you, new system, breaking change impact, new backend features. Great. Build the system. Okay, plan, build. We'll make sure it's on and we'll give it high. Okay, just let it go. Okay, so, um, all right, cool. I want to see your projects and I want you guys to ask questions as we do this as well. Okay, so who wants to, who wants to show their screen? Tell us what you're building. Even if you don't have questions, I just still want to see, like, show us, show us your cool stuff. What has Replit built for you? I could show mine real quick. Yeah, let's see it. Let's see. And walk us through. I'm interested what you talked to ChatGPT about and what your plan is now to, to monetize your 200,000 followers. Yeah, can you guys see that all right?

Um, so the idea that I'm building out on, and luckily, I mean, I've been talking with ChatGPT for months about this product offering that I've been thinking about. It's called Origin Genealogy. So I'm offering reports to be sent to people. Essentially, they will submit some of their ancestral information and between AI and some VAs that I'm going to bring on, um, we'll create a 4 to 6 page, like, preliminary report just, uh, showing people what we can find online. Ideally, a lot of this can be outsourced to AI once I get access to some API keys, but, um, it did a pretty good job for what I told it to do. Um, when I had input all the information into Grok, it gave like a pretty substantial bio on me and like what I've been doing, which is kind of cool. So it created that for me. Frequently asked questions. I was digging through these and they're actually like, I'm surprised at how specific they got. Um, and accurate. Uh, a good contact page. I think what I like the most about it is that it's simple. Like I'm not trying to be this huge service, at least at the beginning. But yeah, people go to start your journey and then I'm still playing with the price, but this looks good. There are a few things I need to change. One of the big issues with Ancestry is depending on your country of origin or where you're from, a lot of different countries have different like levels of access to information online. And so I want to create a preliminary, um, like survey where people can say where their ancestry comes from. And if it's from a country that's too difficult for us to look into, I'll give like a warning or caveat and then it'll take them to check out.

Um, I connected. That's super easy. All you have to do is say, where do you, what country do you live in? Right. And, and if it clicks one of the countries that, that you have that, then have, say, Replit, make sure this happens with these countries.

Yeah. Yeah. Like a warning message. But yeah, I mean, it turned out great.

I'm using dark mode because I prefer that. But yeah, that looks pretty good. Love it. Honestly, your site looks really awesome. I'm just blown away, guys, like that we can in just a couple hours, you know, whip up something that looks this good, right? Does it need some work?

Sure.

But is this almost ready for prime time? Yeah, you got to still connect some payment APIs, things like that. I would use Stripe I already connected it, that one. Then you're— yeah, you're, you're rolling. So obviously you want to make sure you test it and play with it and figure out what your delivery is, right? So obviously collecting payment is one thing, delivery and figuring out that automation is your next step. This is cool. Um, good.

All right, who else wants to show what they've got? Morten, you're muted. Yeah, yeah, I can show mine. Uh, it's not as, uh, I'm impressed with yours, Kalima, but I'll show it for sure. Um, so you guys, can you see? So basically what I'm, what I'm building is, um, it's like a restaurant concierge app for business diners, or, you know, for businesses. I think a huge pain point that I've experienced in my career is you go to sales expos or if you take clients out or if you do like an internal company event, you go out for dinner in any city in the US or Europe, it's always kind of a hassle to, you know, especially if your party is above 8 or 10 people, it's always like, oh, where should we go? Who can fit us in? What should we do? So what I'm trying to create is like this, this agent or this app where you can say, okay, we're in New York, we're at this expo, we're 10 people, we're in this area, give me 3 options and I'll select option 1, whatever. And then the agent actually will go through the whole booking and make the reservation for you and confirm it back to you. So all you gotta do is basically say where you are, what you want to eat, how many people you are, you know, maybe the price level, and then you want the app to give you like 3 options. The reason why I'm saying 3 is because I think in a busy world, you know, you don't want to be presented with 18 options to go through. You just want like your tire, you just want 3 options, and then you, you, you pick one of those and the agent does the rest for you.

That's pretty cool. That's kind of the mission. Yeah. I love it. But what, what works so far?

Like, can you log in? Can you do any of that stuff?

Yeah. So I think I'm having the same problem like you are. Yeah, you're in the Replit Auth.

That's not a big deal. Yeah. Yeah. So, so I think, um, so, so right now I'm not able to do a whole lot.

Um, I think that probably requires me to kind of, yeah, you know, I guess I can, I can click around a little bit, but, uh, a couple of random things. Yeah. I, what I would do is go ahead and give it the, the thing that just says, um, replace Replit Auth with JWT Auth. Auth. Replace Replit auth. Yep. Auth with JWT auth. Like that? Yep. Send. Yep. Yep. And it'll, it'll just get to work on that. It might take it 10 minutes to do it. It's kind of a big change. It has to build some new systems. Okay. But it's, yeah. Yeah. Okay. It's good. I usually, and I do it, I usually put that instruction in the initial thing.

I say use JWT auth. Don't do that.

So, okay.

Yeah, but yeah, otherwise, you know, I think there's a little bit of tweaking and, you know, I can make it prettier, put some pictures in there, all that stuff. But yeah, obviously that's why— that's why it took you 35 hours, John, right, to create your—

oh yeah, your site. So yeah, guys, like, yeah, the fact that you can even see anything and buttons to click on, yeah, get an idea and all this stuff, like, it's just still mind-blowing to me. It's so awesome, right? Like, you're going to be able to do so many things here. I mean, once you, once you get past this, this awe thing and you start to get in and start playing with it, and you could, you could tell it, hey, let's change the design to this, let's make it more modern. And you can— and AI is actually kind of funny. If you insult it a little bit, it will actually work a little harder for you. If you're like, this looks like, you know, something a crappy designer built in the '90s, right? Then it'll be like, oh, I'm sorry, and then just like build you something amazing. It's just like shiny. Lights and things anyway. So very cool. I love, I love seeing it more. All right, well, it says something. So by the way, guys, you don't have to show, um, I'm just giving an opportunity to do so, right? I love seeing what you guys do.

I'll go. Yeah, go for it, Cody. The mine's still building, so I don't know what I did, but it's like still like, uh, uh, still working on some stuff, as you can see right here. Probably because it's doing a client portal But can you guys see my screen?

Just make sure. Yeah.

Yeah, we can see. So one of the things, I actually have a small business that I haven't, I, I made it 2 years ago and I haven't really done anything. Essentially it's just piggybacking on, um, promotional swag items for businesses. Um, and one of the, I, I've been in employee recognition space for years and when I've called most of the swag companies I've dealt with, they're like, well, what do you want to do? Like, tell us what you want. And I'm like, well, what's. What's available, you know, because they have, you know, send you to a catalog with like a million items. And I'm like, well, what's the human psychology? Like, what, what can I do here? And I've had a really difficult time. I end up spending all the time picking what I want without a good, like, a consultation. Yeah. Um, and so I've, you know, I've spent probably $400,000 at least over the last, you know, handful of years, and I've had to do all the work. So I'm like, what the heck is this company doing for me? So what I would like to do is just kind of, um, create a SWAG recognition profile. So I come from the employee experience space and then tie in with the recognition programs that are either developed by HR or a lot of companies have it just in different spots within the organization and create a comprehensive structure that's based upon the milestones of an organization. Those milestones typically are like just anniversary dates or whatever, like those easy ones, but there's also great milestones for training, onboarding, day-to-day tasks that, like, let's say in financial space, this is my 1,000th account that I've opened successfully. And with 100% accuracy. So like each company has their own kind of like really key milestones that they're all tracking and they're like, people, we've got to do this better, we got to do this better. And so you have all these like mid-managers or smaller managers trying to figure out how to recognize people for this, but put it all in one system where it's just like it works, right? And then my job is to— here's items that we have, uh, talk to me about your employee experience journey and coordinate that whole, um, situation in a fast, friendly, easy manner.

That's—

so that's why I'm building a front end and then a client portal where they can kind of experience that. So that's kind of what I'm trying to like build. Um, there's a lot of local companies, or like there's like AwardCo, there's Motivasi, there's a lot of these companies, but they're switching more towards a high-end enterprise level SaaS, super heavy SaaS model. And I don't think I, like right now I'm not trying to get there. I'm just trying to go to Jon, as you're growing out your business, right? Like, you know, you're going from 10 employees to 20 to 30, 40. Like how do we get integrated with you? And so the system takes care of all your recognition needs with the consultation instead of you having to hire someone in HR to go spend a whole bunch of time doing that. So anyways, this is kind of the website that it, created. I think the big key is like a consultation. So that's what brought out here. It's got my value prop like listed. There needs to be some type of clickable options. So whether you wanna do, if you want a consultation or if you wanna do self-serve, build that out here. But I mean, the buttons seem to work pretty well. I've still just kind of Every time I click on something, it kind of refreshes. So I would like to have the whole thing refreshed. But anyway, I think overall it's pretty amazing. I don't like the color scheme. That's easy to fix. Yeah, yeah, that'll be— it's more just like the content, like do I like the wording, things like that, the flow.

Um, but yeah, this is kind of what my— so what my page looks like. Super cool. I love the business idea, by the way. Um, there are a couple of other companies out there that do some stuff like this, but, but not exactly kind of, kind of what you're doing.

Um, yeah, like, I don't know if you've heard of Awardco. Yeah, so like Awardco, actually, I use, I use Awardco where I work, and it's the, it's still very— despite us using Awardco, the way it was implemented was very much especially with the larger organizations. There's, you know, 7 departments within the company. They all have their own recognition specialists and no one's coordinating with Awardco. So money is going everywhere. Branding is going everywhere. It's just, it's just an actual mess. And there's no coordination across like, what are the key milestones? So there are things like that. And I'm not, you know, Awardco, you know, is worth like $5 billion right now. I'm really just trying to go after this smaller market right now to see what I can get before you go into a full-on like SaaS model sales thing where you're competing with, you know, some of the, some of the bigger boys. I think there's an underserved market where they're not looking at smaller businesses. Um, and maybe I can come in there and I don't need to be the first to move. I just need to be, uh, a mover that provides value.

So that's kind of what my business plan is. Very cool. Very cool. I love it. I think it's fantastic. Yeah, I guess let it keep, keep cranking. Yeah. Building stuff over there. Just let it, let it go. All right. Awesome.

Anybody else want to show what they've got? Mine's still building.

Like early stages, it looks like it literally says this is a very complex app. That's awesome. Did any of yours like pop up like a cost?

Mine popped up, you're at $5.36. Yeah, I got $3.36.

That's because it defaulted to the turbo mode or whatever. It would be interesting to have it like kind of like a calculator in the corner of like where you're seeing your dollar amount just so you can kind of— what are we talking about when I click on this button or that button? Are we talking you got a burn rate of $0.30 per half an hour?

That's not incorporated into your monthly fee. That's in addition to it. No. So what it is, is you basically are prepaying for $25 of credits every month. So you have a minimum of $25 you get, and then you use up those $25. Once you hit $25, it'll say, hey, you've hit your credit credits for the month. Do you want to keep going? Yes. And then, you know, so again, for reference sake, I built GenAIPI with about $140, okay? Which is crazy. It's crazy, right? I mean, the fact that you're looking at like a $3 bill and you're like, you know, that was on Turbo, right? Like, you're— you pay $3, guys, for a real developer to look at your project for 30 seconds, literally. Right. This is building your whole project, right? That would take $10,000 for your first preview and then another $40,000 to get all the buttons working and then another $100,000 to get it actually ready for customers. And now you're at $150 grand and okay, yeah, now you're ready for customers, but now you have, you don't even know if customers are willing to pay for it. Right. Whereas now you can be like, I'm going to spend $15. I'm going to see if I can get some customers to pay for it. Then I'll spend another $20 or another $40 or whatever it is. Right. Again, I think it's probably about $100. $100 is what you should expect to really get a fairly complex thing, you know, put together. Um, uh, Morten, I bet you can do yours for $50. I think yours is actually complex, right? It's cool, it's very powerful, but it's not like crazy dashboards and all this stuff. It's like, I need to connect to a couple APIs, I need a good experience, results, it's ready to go, right? And then just work on making it like a really slick, smooth experience. Super easy experience. So anyway, very cool.

All right, Jacob, you want to show off?

One sec.

Mine is not exactly like I wanted it. I got a lot of tweaking to do on it, but, uh, it kind of took some things from my Facebook and from my, my overall Facebook page and redid some stuff, so it did it a little bit differently, but Um, basically it's just trying to get people to come to the website to try to get me to help them market in any way. So I usually do social media, uh, analytics. I do like Google Ads, SEO, different things like that. And so just trying to get somebody to come in, it's basically a subscription. So first month is like a trial they pay, but they still it's a trial and there's no commitment.

And then after that, they go into a contract of however long they want to do it for. Got it. Very cool.

How many hours a month do you work for them for $1,000 a month? I don't know yet.

That's one of the things I've been thinking about trying to figure out. And maybe let me ask another question.

Why $1,000 a month? Is that just kind of what you—

I just threw that on there. I don't That's not like a—

all right, something that I was thinking about right then. I just threw it on there real quick.

But cool.

And just kind of went over services. It's a little bit plainer, a little bit more plain than I would like.

Oh, it's definitely— but, but yeah, so, so here's what I would say right now, um, on just on your design. Um, first of all, it's got the margins wrong. It's not centering it. It's not doing a couple things right. And so I would just, I would just tell it and say, hey, this design looks like it was made by a monkey. I need you to like really go through and do this right. Make sure it's centered.

Make sure the margins are right.

You know, whatever. All right. I'm doing it right now. Yeah. Like, literally, you can say that and it'll, it'll, it'll get offended a little bit. Right.

And yeah.

Anyway, I can stop. You want me to stop sharing then? No, no, no. You're good. That's good. I want you to. I want you to. Yeah. Yes. Yeah.

Center it. Fix the margins. So Jon, can I ask you a question? Like, so when I was building my first app all by myself, I got to this point and I was like, is this really what this process is supposed to be like where I'm going through every little thing and nitpicking? And so I, I stopped cause I'm like, I'm doing something like I'm doing something wrong. This doesn't work. It's not what you were saying. People give up on AI before they're just like, I'm just, I do, I have to really do this. So now it's taking a ton of my time cuz you don't realize, well, if, hired this out, you know, this is going to be a month and a half project and cost me $30,000, whatever it is, you know. So is there, I mean, to get to your level, were you just screwing around in this so much to find out like how everything works? Or is there like a learning course of like, I just want to focus on Replit and I'm going to explain what the difference between build and whatever, all these different things so that You know, Replit seems to be like the core of all these, like this is where the magic happens, right? This is where you actually take the cool stuff you did in Groq and ChatGPT and then move it over to an actionable product. Like the beautification of what your backend is, is it just trial and error over and over and over again? And, and it's, AI is just so new that no one's really come out and just said, hey, we're gonna do like a week-long course about out exactly how to get the best out of Replit.

Yeah, well, a couple things. Obviously you've got access, you know, to me now. You also have access to, um, one of the things— so we're kind of— we got 10 minutes left today, guys, and so this is a good time to start talking about some of this stuff. Um, this is great, Jacob. We can, we, um, let it, let it just kind of redesign it and be— you know, the AI doesn't care, right? If it sucks, tell it sucks. And, and, you know, hey, let's move this and take a screenshot of it. Be like, look, you've got everything pushed against the left side here. That's wrong. Let's get it centered on the screen and get the margins, get the padding right. And if it can't do it, if you ask it a couple of times and it can't do it, say, I need you to completely rip out your formatting and start over. Like you're doing something wrong anyway. But a couple of things, a couple of resources, guys, you're going to have beyond just the course, right? Obviously, we're going to do this today. Tomorrow, we'll do a little bit more in Replit. We'll start to learn how to bring in maybe some AI voice assistants, how to build your own your own GPTs, your own AIs. We're going to do image generation and video generation. We'll talk about sales and marketing, how to automate some of that stuff, how to automate social stuff, like lots of automations and more advanced agentic work tomorrow. Okay, so that's going to be, that's going to be really fun. And it's really just to expose you. And then you, you have to, yeah, you gotta, you know, if you want to learn how to kayak, you can watch as many YouTube videos as you want, but you got to go sit in a boat at some point. That's how you're going to learn how to understand what the water feels like underneath your boat as it comes through or as you get thrown over into a rapid. You learn when you do that. You learn from doing, and I did learn from doing coding. But that being said, part of what I'm trying to do here is help you guys, is help you avoid having to do all of those different things. Let's jump in here. Let's go drive right here. Let's log in. I think this is— I think it's in this FlowState email address. All right, here we go. Build a Business in a Weekend with AI. So this is a folder I've prepared, and there's all these advanced workflows about how you, how you go through and, and use the Replit agent. Here's, you know, breakdowns and different things, lots of tips and tricks in here. There's another, there's another really great one in here. Software and Tools and you go into Mastering Replit right here, this is awesome. If you take time and you actually read through these, in fact, we're going to do this before we finish up. In fact, I might do it right this second. But one of the things that we've done is we've built a, we built a business plan, right? And then we built a feature list.

Okay.

And so what we want to do though, right now, those, those plans are sitting here, right here. There's a little thing that says Open Files right next to your Publish button. And here's all your files, all the code that it's created. Okay. And so in here you've got in attached assets, you've got your improved business plan, right? This is the plan that we made, whatever. This is cool. Here's all this stuff here, right? This is great. And so, but they're not really usable in attached assets. We need to pull them into the project. Okay. And so we're going to find Connect Resend. Go ahead, do that. Nope, we're not going to do that right now. We're going to do that in a minute. So right now what I'm going to do is I'm just going to say this. I'm just going to paste that, that thing. I will drop it into the chat here so you guys have it as well. Okay. But what this is going to do is it's going to create a guideline. So every single time it is, yeah, every single time it is building something out, it is going to refer to your business plan. It's going to refer to your feature list, and it's going to turn it into a useful kind of guide and what I like to call bumper lanes, right? It's going to constantly be like, okay, remember why we're doing this. Remember why we're doing this. Remember why we're doing this. Here's the context, right? Because it's not going to remember everything, and you're going to start a new chat, and you're going to do this, and you're going to move it, and you're going to leave, and you're going to come back, and you're going to be working on it for 30 hours or 15 hours, whatever it is, and it's, it can't remember every single thing you've typed to it and every single, you know, whatever, right? There's too much information there. Okay. And so it will do this, you know, it's a queue, it's, you know, it's got 15 tasks to do right now. So it's got to, it's got to work on a few things to get my forgot password functionality and switch to a JWT, the JSON Web Token, right? Login stuff here. But that little thing right there is super, super useful. You got that, that's one thing that's really good. Here's how you optimize some of the things here. Here's how you debug and move faster. Here's some of the problems you might run into. Here's how you handle APIs. Trust me, these are like little— they're not like comprehensive guides. They're more like, here's a tip that no one's talking about. Use this tip and it will work for you. That's the stuff that's in there. Then there's a bunch of other things like that. I'm also right now, guys, working on a, um, something that you will all get access to because you've paid for my premium product of, you know, this How to Build a Business in a Weekend with AI course. You will get— I'm, I'm building out right now, uh, pre-recorded modules that are going to be, you know, tons of videos. They're gonna be okay, we're gonna, we're gonna focus on this task for a minute. Let's talk about these buttons. Let's talk about this interface. Let's talk about this thing. This little extra, you know, what, what, what is the security scanner do and what do you— are you going to use the extension store and how are you going to use the playground and all these little things in here so that you really can be like, okay, I know what's going on. And so I'm actually a partner with Replit in a big way. Um, the CEO of Replit— it actually kind of happened because I built this GenAI API thing out, right? And then Replit's, um, I just go to replit.com. I don't think I can because I'm logged in now. Yeah, let's go to an incognito tab. If you go to replit.com and you just come down here and you'd be like, hey, you know what's going on, check this out.

Boom, right on replit.com, Jenny API.

Awesome.

I know that, I know that you've had some— I, I've watched you, you've had some interactions with either the Joe Rogan Show or the Replit guy.

Yeah, I know that your name's been circulating. It's been pretty cool to follow that. Yep. So it's been fun. I, I, anyway, I work really close with them and, and I'm, I'm now working on a deal with them where I'm going to become their first, uh, official training partner, right? Wow. So if you're, you know, using Repwood, you're trying to figure out how to do it, they'll be like, hey, go to Jon, he's got courses, right? And so, you know, my Repwood training courses, things like that, that'll be something that you could buy on the website. Like I said, you guys are going to get access to it because you're, you you're in the, in the crew now, in the, in the, in the inner circle. And I'm not, you know, a lot of people that do like all the course gurus and stuff, there's always like some upsell to the next thing. And I'm really not here to do that. That's not why I did this. When I started GenAIPI, I had zero intention of doing this course. This was not something I wanted to do. And in fact, this is the last course I'm doing of the year because I just don't have time. I have to run my actual business, which is My business really is this now, AI leadership. We do fractional chief AI officer stuff. And so I work with companies, they pay me anywhere from $10,000, $15,000, $20,000, or $25,000 a month per client. And I have a team of 7 people now and we help deliver all kinds of cool stuff, strategy and transformation and implementation and training, all kinds of stuff. But this is taking my time and it's way more profitable than sitting here and teaching you guys, right? Not that I don't— I actually love teaching these, and so I'm going to keep doing them, but I'm not going to do any during November and December. It's just too busy with holidays, and I'd rather spend time with my kids than teach some more people how to do Replit. That's just how it is. Um, so anyway, guys, you're gonna have access to this. You're also going to have access to— and I'll give you this link right now, we'll just open it up. It's on Facebook. Um, there is a Build a Business with AI. Okay, this is an invite only, and you guys are welcome to join it. You don't have to, but there's a good number of people in here, 113 people. So it's a growing community, but it's, it's really awesome. It, it, you know, it has a bunch— any— in order to be in this course, you have to have taken— or sorry, in order to be in this group, you have to have taken this course. And some of these people like Brantley Brumley, I ended up hiring him. He's amazing. He took the course back in April. He's one of the first people that took it and he just took off. He built two businesses that were cash flowing over $10,000 a month. He just built stuff that just like started going really fast. And I was like, dude, you're amazing. And so I hired him. I was like, I need you to help me do cool stuff. So he's actually the head of my fractional chief AI officers at GenAIPI. And so it's really awesome. So anyway, lots of really cool people. They're sharp. Ask questions in here like, hey, I'm stuck in this, I don't know how to do this, how do I get through this thing? People will answer always. Brantley's always the first one to jump in and answer, but other people will jump in and communicate. So go ahead and, you know, click on the, you know, request to join or whatever, and I'll go through these. Yeah, it looks like Kalima's already in, so approve. So as you guys show up here, I'll make sure that happens, but you'll have some ongoing, you know, support there. Um, you guys also have my cell phone number, right? I send it to you in the, in the, you know, pre-whatever. And if you're like, hey Jon, how do I do this? Just ask me, right? I mean, tomorrow we'll be done with the course, 4 hours, you know, later, and, and then you'll be like, well, what do I do? What do I do now? And how do I do this? And hey, how do I really focus on this? I'm happy to just, you know, help you out, right? Um, I can't, I can't do a ton. Obviously I'm, I'm very, very busy, but if it's like a quick answer, like how do I do this, or it's a 5-minute question Hey Mitch, I'm gonna call for 5 minutes. Totally, let's do it, right? If you need like really advanced help, well, maybe you can hire me, right? And I can, you know, our team, my fractional chief officer team or whatever can, can do it. But it starts at $10K a month, right? It's, it's a little bit more expensive. So, um, anyway, another cool thing that I'm doing, guys, that I'll tell you guys right now— I'm announcing this tomorrow morning at 6:45 AM, so before we jump on the call, but you guys get early access to news. So I was hired a few months ago and I haven't really been able to talk about it until right now, but I was hired by Tony Robbins and Dean Graziosi to build out this new thing called AI Advantage. And so Tony and Dean, this is definitely an AI picture. I did not make this picture. That's AI. I actually built all the content for this website. I built the entire first version of this website. In Replit, right? And built it back in June. But, but they— what they do is they take, you know, whether— yeah, here we go. There's their, their page. They've done it for wealth, they've done it for health, they've done it for real estate, they've done it for all this stuff. Now they're doing it for AI, right? And so they want to take AI to the world and really help people figure it out. And so courses and webinars and, and events and, and all these things, although it's very low-level stuff. Like, we go more advanced in this course than you will ever get to in the stuff that AI Advantage is doing. It's really like, how do we take people that know nothing and they're scared and they want to do it? And let's talk about mindset and let's talk about all this stuff. But they basically said, look, Jon, we're going to talk to Sam Altman and Steve Wozniak and Amjad Masad, the CEO of— we're going to talk to all the experts. We need you to be our chief AI officer and we need you to make sure that we're doing this right, that we're saying the right things, that we're teaching the right things, that we're focusing on the right things, that we're not going too deep, that we're not going too shallow, right? How do we find the right thing, right? And so the next thing, I do recommend joining this event.

It's free, right?

You can click reserve AI Advantage, whatever. It's a 3-day event. This is live now. You can sign up now. It's November 6th, 7th, and 8th, 100% virtual, 3 hours a day. Okay. I do recommend doing it. I think you will get a lot of value out of it. There'll be a lot of great people. On there. And then while my picture is not here, I'm the, I'm the bonus guest that they're going to bring in. And so I will be, I will be on stage with these guys and it'll be super fun. The last bonus guest was Alex Hormozi. I don't know if I, you know, measure up to him, but, uh, but anyway, um, super fun. Yeah, I totally built all this stuff. This is so cool. And I brought Amjad Massad, the founder of Replit. Um, I, I met him through all this other stuff and I was like, hey, Dean, do you want to, do you want to bring Amjad into this? He's super cool and Replicat's amazing. And he's like, yeah. And so anyway, they, they hit it off, right? But yeah, Mark Baniyev will be there. Rachel Woods is super sharp. Sabrina, I love Rachel and Sabrina. They're so cool. They're so cool. But anyway, GenAIPI is really cool. After, after this, this will be kind of a future announcement in about another maybe month or so. But GenAIPI is going to be a part partially owned by Tony and Dean and they will, you know, just a couple percent, but they'll be on the cap table. They'll have that. But then GenAIPI will be one of the only companies like external companies that they actually endorse and recommend for businesses that want to do this because they're helping again this kind of lower end, low level, kind of more automated courses, evergreen content. Whereas GenAIPI is very hands-on, like we're going to be leadership in your company. High-end stuff. Super cool. So we're going to be— they're not going to try to compete. They're not going to try to do that. They're just going to push businesses to GenAI. So I'm going to turn this thing into a $50 million company by the end of next year because of this. So anyway, guys, super fun. This was a fun first day. Lots of cool stuff. The purpose of the first day really is to get us to the point where we've got something built in Replit. A lot of people spend the rest of their day working on their app and just building it out. By tomorrow morning, you'll be like, I have so many questions now because I've run into all these problems anyway. So definitely bring those questions, or, or, you know, take the day off, you know, go, go have a good, you know, rest of the day, whatever you want. But, uh, join the Facebook group. Um, I will send a follow-up email tomorrow with the recordings, with that Google Drive folder, with the, uh you know, the Facebook group. Let me do this right now. I know some of you are just, you know, some people are a little more ambitious and that's fine. So I'm going to drop the, the Google Drive link into the chat here. So those of you that want to go and click on that or copy that, again, I'll send it in the, in the follow-up email tomorrow. So you'll have access to all of these things. But yeah, guys, good first 4-hour. We're 50% done. Nice work. Cool.

Awesome.

Tomorrow we'll talk about a lot more advanced stuff. We're gonna— tomorrow, guys, will be a little less like waiting and, and, uh, you know, watching for Replit. And, you know, we'll do some stuff in Replit and we'll have some Q&A and like come with questions if you have them. But I'm gonna, I'm gonna show you guys 15 platforms tomorrow. They're gonna be fun. Some blow your mind. It'll be, they'll be really, really fun to do. So awesome. Thanks, guys. Real quick, thanks a lot, Jon.

Hey, you guys can drop off, but Justin, yeah, what's up?

Does any of your offering have anything around the enterprise stuff? If you remember, that was one of the, the original reasons why I reached out and then you recommended this course.

I was just curious. Yeah.

What do you mean enterprise stuff? The, the, your bread and butter, your $10,000 to $25,000, like that, like, like frameworks around that stuff, or is that some, something completely separate for people that have purchased this course?

Yeah, that's, that's really separate. You know, The way that I work with enterprises is it's relatively custom, right? There is some order to the madness, right? But we really come in and we analyze and we work with you and we figure out how does the business, what does the business need? Where are the holes? Where are the gaps? Where are the opportunities? And then we prioritize, make plans, train, do all kinds of stuff and figure it out. But it really comes down to the frameworks that you'll see on my website on the AI leadership page, right? We're coming in and providing strategic guidance. We're getting our hands dirty and doing stuff and implementing things and then leveling up all staff members to be more proficient with AI. So anyway. Okay.

All right.

All right. Thank you.

Awesome.

Okay.

Thanks, guys.

Thank you. Appreciate it. Thanks, Jon. Bye. See you.