By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
The Calum Johnson ShowThe Million Dollar Vibe Coder: The Fastest Way To Get Rich In The New Era of Al (No Code!)
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
Great conversations run long. Their best ideas don't have to.
uListen is free on iPhone, iPad, Android, Chrome, and Claude.
See howTranscript
60 min read · 12,471 words- 0:00 – 1:57
Intro (The Game Has Changed!)
- RBRiley Brown
I think that this whole movement of AI and vibe coding, you don't need to follow the rules because you have all of the knowledge from human history in an AI model. You can ask it any question, and you can just apply it. I would consider myself one of the best experts. I don't follow a playbook ever. And I've noticed the most successful people out there, one of them being this guy named LevelsIO, who's making around 300K a month, he has no structure.
- CJCalum Johnson
How do you actually scale this to the point that where he's making, like, $300,000 a month? Like, how do you get to that point where it's generating that sort of revenue?
- RBRiley Brown
Honestly, I think I should show the response to this instead of, uh, answering it. Can I just pull up my computer real quick?
- CJCalum Johnson
Okay, so today's guest is Riley Brown, and before we get into the episode, I have to share the inspiration behind this conversation. Riley is the definition of a vibe coder. He has no deep background or experience in coding. He's not a software engineer. However, he is an expert in using AI to build products and businesses super fast. He just raised $9 million for his AI app, Vibe Code. And so our mission, our intention with this episode was to show you the fastest way to build an AI business that you can launch and start getting feedback on in just a few hours. And so let me say this. In this conversation, Riley shows everything, and he's so passionate about this stuff that it really made for a great conversation. Nothing was held back. And one thing we actually did right at the end is that we set a challenge for you. And if you can do this challenge, there's gonna be a number of you that have successful AI companies in the next six
- 1:57 – 4:35
Step 1 – The $9M AI Sprint Challenge (Speed over structure)
- CJCalum Johnson
to 12 months. And so you're gonna see that at the end. Let's just get into the conversation. You are listening to The Calum Johnson Show. I just wanted to read to you what Calloway said about you. He said, "Riley's so deep in AI right now. He lives inside of ChatGPT. And what I love about Riley is that he's always experimenting. He'll spin up these business prototypes super fast. He'll have an idea and build the version one using AI in just a few hours. And it's not even about the tool, it's about how fast he gets to something that's real. That's what makes him dangerous." It's about how fast he gets to something that's real. That's what makes him dangerous. And so here's where I wanted to begin, because even as I was just reading it out, I start to think about that person at home that just has, like, a ton of business ideas. It's, like, saved in the Notes app on their phone, or, you know, when they go on YouTube, they just see, like, entrepreneurship videos. But at this moment in 2025 and us sitting here, they haven't yet launched something. What do you think is gonna be the value to that human being of listening to the conversation that you and I are about to have?
- RBRiley Brown
Yeah. I mean, first of all, love Calloway, and he experiments in his own way. He's a lot more meticulous than I am. I am much more of an explorer, I would say, in terms of I find cool technologies, and I, I obsess about them because I, I, I genuinely don't understand if you're not obsessed with AI right now. N- And not from, like, a guru sense. You know, like, a lot of people are selling snake oil in the AI space, and I think a lot of people get worried about it in that way. But, like, if you actually just look at what AI can do, you'll realize that anything that you do on the computer or could do on the computer is possible with just, like, a month of learning. And so hopefully by the end of this episode, you'll realize some, uh, f- good frameworks around finding problems to solve in your own life and then being able to take that problem and turn it into a solution through AI coding tools, which has, uh, become known as vibe coding, which is a movement that I started a company, uh, with the name. Like, we're literally building, uh, the Vibe Code app, and that's kind of what I'm gonna dedicate my next decade to. Um, but I wouldn't be in this position if I wasn't just, like, hyper-obsessed over this technology, and so hopefully, um, I can make some of this obsession rub off on, on the, the viewers.
- 4:35 – 6:20
What Is Vibe Coding? (One prompt, endless possibilities)
- CJCalum Johnson
Yeah. You know, we- we're gonna get into the, like, the fastest way that someone can build an AI business. But before, before we do that, you mentioned vibe coding, and I wanna, I wanna make sure that we give people the context because typically, when people get really excited about technology, they're, like, software engineers, or they spent, like, years coding. Vibe coding is something different. So to people that might not know, can you just explain, like, when you say vibe coding, what do you mean?
- RBRiley Brown
Traditionally, if you wanted to create an app, you would need to write all of the code by hand, and this involves writing a bunch of code files and then debugging all of those code files and then running it so that you have some sort of app working and running so that people could, you know, sign into your app and use it, and then maybe they could pay you for that. With AI, the same way that ChatGPT generates essays for college students or generates emails for people who work at Microsoft, AI now generates code. And in my opinion, it is the best skill that AI has. Because it's relatively deterministic and, um, these models have been trained on so much, uh, code, they can literally create full programs in a matter of minutes. And it allows people who are insanely creative and are really fast to act on their ideas to iterate a bunch of times. Um- And so you brought up that, uh, uh, that Kane said that I can create a solution or a, or a prototype of an app really quickly. Most of them didn't amount to anything.
- 6:20 – 10:08
Fail Fast, Learn Faster (Riley’s rapid-iteration secret)
- RBRiley Brown
But the fact that I do it over and over and over again has led me to a place where I created a successful app, uh, four or five months ago, and then I ended up putting it on hold a little bit to start a company, and now we're in S- I've just moved to San Francisco, um, and we're actually building a company around Vibe Coding. And I would not be in this situation if I wasn't just, like, constantly iterating and getting better at this skill. So it's still at the stage where it is a skill. And the leverage, if you have a good ... Uh, if you're a good critical thinker, which I think is the most important thing in the age of AI, like how good are you at recognizing problems and then, uh, solving those problems with prompting, those are gonna be the most dangerous people over the next few decades, uh, because it is still a skill and you still have to have the desire and the taste to actually create the thing. Um-
- CJCalum Johnson
Mm
- RBRiley Brown
... and so that's, that's kind of the philosophy and a little background on Vibe Coding.
- CJCalum Johnson
Yeah. You know what's interesting, and I, I even think about this in my story with, like, the first businesses that I start, that I tried to start. The thing that killed them is that it just took me too long to launch.
- RBRiley Brown
Yes.
- CJCalum Johnson
Like, I wanted everything to be so perfect before I let the world see it. And it's like when I was speaking with you, and even when we spoke before this, your superpower is your speed. Like, you're, you're the first one there. Can you talk about, like ... And just, I just wanna make it so real for people. Why-
- RBRiley Brown
Can I, can I say something? Can I say something on this? I, I have-
- CJCalum Johnson
Go for it
- RBRiley Brown
... this is actually something I'm insanely passionate about. I, I've had some friends in my life who were constantly deciding which path to take in life. And the way that I see it now, especially with the fact that even with just the internet, never mind AI, you can figure things out quickly if you want to, but you have to want to. And then I've had friends who are, you know, deciding, like, "I don't know if I wanna be this," or if, "I don't know if I wanna be this," or basically they don't know what, like, life to launch. And what people fail to realize is that just flip a coin, because if, especially if you're early off in life, I think you're more likely to do both if you just pick one, right? Not, you're not picking between the two. You're picking between whether or not you wanna be someone who makes things or someone who doesn't make things. And so just pick one, and then fail fast so you realize it's a stupid idea. Because if it's a stupid idea, you should want to know as quickly as possible. And so just do a bunch of stupid things over and over again. And I think that has really defined my, um, this decade for me, since 2020, is I just, I've just been doing and failing and, um, and then I created a, a reselling business, like, uh, in, like right after I graduated. And then, like, that was pretty successful. And then I realized I could make way more money as a content creator, and then I started making content. And then I g- I, I r- and then now I'm, I'm gonna make content for the rest of my life, but I also wanna start companies. And so now I'm just trying to take as many shots on goal until I find one that fits with my content. And so I've just ... If there's one thing I'm proud of myself for is it's just, like, my willingness to just fail really quickly. Um-
- CJCalum Johnson
Mm
- RBRiley Brown
... and I, I wish more people tried this because I think they're too attached to the identity. Like, "I can only be one thing," when really you can just take as many shots on goal as you, as you possibly can.
- CJCalum Johnson
Yeah. You know, I, I think you're so right. I think it's the identity. I also think that there's ... And I just think about myself, like back then. I think there's also, like, an embarrassment in failing.
- RBRiley Brown
Mm-hmm.
- CJCalum Johnson
And so, you know what, I just, I just wanna debunk this because even though, like, you're right now building a startup and raising money for a startup, like millions of dollars for a startup, uh, based on the fact that you just experiment really fast and, like,
- 10:08 – 14:13
Step 2 – Live Demo: See AI conjure a fully branded expense tracker in seconds
- CJCalum Johnson
test things, like you didn't have this super technical coding background. Can you talk about-
- RBRiley Brown
Yeah, I mean, I mean, to be fair, to be fair, I am assembling the team of the smartest engineers in the world, and we're going to ... And, and they all know consciously, like we're trying to build the app that replaces them, right? Their, their path to t- try- we're trying to get rich. Like, and I, I I'm not afraid to say that. Like, we are taking the big shot, right? I can create a content business in my 30s and 40s and 50s and, and build wealth slowly. Like, this is my one true shot, and, um, I've always wanted to do this. And it's going to require the smartest engineers in the world. Like, we are trying to build the tool that allows anyone to code, and to do that it will require a ton of engineers. Some people have this idea that I vibe coded the Vibe Go tool. That's not what happened. But it was the fact that I made so much content on Vibe Coding that led me to the ability to network with the best talent, uh, engineers in the world. By the way, the engineers on the team, which are, like, world-class, don't like coding. They like building, and they're really good at coding, but they would r- much rather have AI code and just, like, bring their ideas to life. And so that's, like, what we're hiring for at this company is, like, people have a bias towards action and just, like, making things real, and we're gonna ... That philosophy that I have in life, like, that's gonna be the center of the company. Sorry if that's like a, a ramble on the side, but I wanted-
- CJCalum Johnson
No, no
- RBRiley Brown
... to just throw that out there.
- CJCalum Johnson
It, it's perfect. And I, and I wanna stay there just for a second, uh, before we get into the practical stuff. Can you do two things? One is tell us a little bit more about what it is that you're building, 'cause it's fascinating. I think it's gonna be valuable to the people that are listening. But before that, tell us also about, like, your first experiment with, like, building an app or a business with AI, 'cause I'm sure it wasn't as successful as what you're doing now. But give us those two point of views. The first time that you built something, uh, experimenting with AI and building products, and then what you're building now. Okay. So we're gonna get back to Vibe Coding with Riley in 60 seconds. But before we do, today's sponsor is gonna make sure that your finances stay on track Introducing our friends at Rocket Money. And here's what I wanna know. Let me know if you can relate to this. One of the most annoying things is paying for a subscription you didn't even realize is still active. It's happened to me numerous times, and so heading into 2025, I made an adjustment, and I started using Rocket Money to clean house. Rocket Money is a personal finance app that lets you manage subscriptions, lower your bills, and build a custom budget and grow your savings all in one place. And so here's what happened when I started using Rocket Money. They surfaced every subscription that I'm paying for. I was shocked to learn that I'm paying for some things twice. And with just a few taps, I was able to cancel all of the dead weight. And it's not just that, it also scans your bills for savings. It pings you before charges even hit your account, and its Goals feature automates deposits towards whatever you're funding. And all of the members who use their premium features have saved up to $740 a year, and collectively over $500 million in canceled subscriptions. And so if you wanna get your spending on track and join the five million other members using Rocket Money today, go to the link in the description, and when you use that link, you can get started with Rocket Money for free. So thanks again to Rocket Money for sponsoring today's episode, and let's get back into it. Give us those two point of views. The first time that you built something, uh, experimenting with AI and building products, and then what you're building now.
- RBRiley Brown
Yes. Okay, so I'm gonna answer the second part first. So first thing that I ever built was in something called Claude Artifacts. Honestly, I think I should show the response to this instead of, uh, answering it. Can I just pull up my computer real quick?
- CJCalum Johnson
Do it.
- RBRiley Brown
I could explain this way better on, uh, screen share.
- CJCalum Johnson
Let's roll. That's what we're doing today, man. Let's go
- 14:13 – 18:53
One Prompt → Building a mobile app
- CJCalum Johnson
into it.
- RBRiley Brown
Hell yeah. Great. Okay. So the first tool that I ever used where, uh-- And, uh, the first tool that a lot of people used where they realized that they could actually do things with AI and code was Claude Artifacts. And so Claude is j- a ChatGPT competitor, and they created this Claude Artifacts feature. And so, for example, I could say create, um, a, create a simple calculator, uh, app in Artifacts. And so Artifacts is this, like, side window that pops up in Claude. And by the way, I think this is a great way to just learn about front-end design. And so you can see here the code is generating here on the side, which was different than anything that came before it. Previously on ChatGPT, you could generate code, but if you're a non-coder, you don't even know where to put the code. It just generates, and you're like, "What's the big deal of code? Like, what does it even do?" But you see here, could you-- Uh, you can see my s- uh, you can see how it rendered as a calculator, right?
- CJCalum Johnson
Yeah.
- RBRiley Brown
And so now we have a calculator that should, in theory, work, right? Five times five. And so you can create these simple applications. This is a very simple, um, example. But, like, we could say, like, uh, make it bigger and, um, make it bigger and call it Riley's calculator. And this is when I realized we're entering this era of, like, personalized software about-- And this is kind of the essence of vibe coding, right? You generate something, and then you edit it, right? And now you have this-- And you see it's just editing the code. Claude's getting a lot faster. It's getting a lot better. This is the new Claude 3.7 model, which is their state-of-the-art coding model. And you can kind of watch it generate and edit the code, and then it re-renders, and now it's making another edit, and you'll see it render again.
- CJCalum Johnson
It's pretty-
- RBRiley Brown
And so this was kind-
- CJCalum Johnson
It's pretty crazy just watching it happen in live time. Like, it's figuring out the code in live time.
- RBRiley Brown
Yes. And they just released a new model where instead of having to regenerate the entire code files, it can directly just edit the code files. Um, I don't even really know how it works, but, like, the point is, is you don't have to know how it works. The same way that you're using a camera right now to film the podcast, um, and maybe you're a camera nerd, but I have no idea how it's working. I just know cameras work, and I can record a podcast and connect with people on the internet. Software is moving to the same, um, in the same direction, which is the reason behind our whole thesis here, uh, behind Vibe Code. And so Vibe Code, this is Vibe Code. I have it open on my phone, but I'm streaming it to the computer. Um, and you asked me what it is, and all it is is an app that lets us build mobile apps. So what I wanna do first is, like, we can just record our voice and, um, please create an app that lets me record my voice and it converts it into a YouTube video outline with a hook, a, um, a, uh, introduction, and then outline the different body paragraphs. But, like, make it YouTube specific. So use the Whisper model to, uh, transcribe what I say and then convert it using, um, AI and make it, uh, red and black.
- CJCalum Johnson
You know what? Just, just as it's doing this, I think one of the things that I think is just difficult when you're, like, first building something is, like, identifying a pain point, and it's actually kind of step one of the process, right? Which is identify a painful workflow and find a problem worth solving. And so in this example, it's like, okay, the problem worth solving is, like, how do I produce great YouTube videos faster?
- RBRiley Brown
Yes. Right.
- CJCalum Johnson
And with what you've just spoken into this app, it's now generating an actual product, an app, an idea-
- RBRiley Brown
Yes
- CJCalum Johnson
... like a phys- something real that could solve that problem.
- RBRiley Brown
Totally. And by the way, this will actually take a while because this has two AI features, so it, it's probably generating, like, 13 to 14 different code files, and the AI agent has to go through and generate everything. So this will take time. What I wanted to do really quickly is let's go to one of my favorite tools. And so this is w- um, one thing that I have just been doing more since ChatGPT came out is I've just become a really visual-- I just rely on visuals all the time. Um, how do I go home here on...
- 18:53 – 22:50
Screenshot to Website (How to Design your landing page in seconds!)
- RBRiley Brown
Um, oh yeah, Create new. So what is-
- CJCalum Johnson
So what is this tool you have open right now?
- RBRiley Brown
This is Whimsical diagrams. There's other ones you could use, Excalidraw, TL Draw, um, uh, Miro. It doesn't matter. But I really like to diagram everything out, because if you ever diagram something out, you could just give it to AI and it will understand what you're saying. And so I think diagrams are like-- that's just how I think, right? And so, like, let's try and come up with an idea for an app. Um, one thing that, um, I think would actually be kind of fun is just, like, converting a-- Do you ever watch a YouTube video and it has, like-- like, it's an educational YouTube video, but-- and you wish that, like, you could take whatever framework they're speaking about and turn it into an app? Maybe this is just me. Maybe this is not an actual real problem. But, like, you know, you know who Paddy Galloway is?
- CJCalum Johnson
Yeah.
- RBRiley Brown
Paddy Galloway is, like, a genius YouTube scripter or strategist.
- CJCalum Johnson
Like strategist, like growth expert on YouTube.
- RBRiley Brown
Yeah, he's-- Yes, he's, like, a YouTube strategist, and I wanted to make an app that, like, um-- Actually, you know what? That might be a good thing. Like, why don't we just take his philosophy and turn it into a mobile app?
- CJCalum Johnson
Uh-huh.
- RBRiley Brown
Because that was one thing that I wanted to do a while ago, or turn it into an app. Because the people's problem, right? When you get started on YouTube, the problem that people have-- And so I always go like this. Like, I, I go problem, and then we will drag this down here, and then we have, like, ideal state, right? The problem with people getting started with YouTube is they don't know, like, there's so many things you don't know. Like, you don't know how to organize all of your thoughts. And I think the central problem is, like, you don't know what questions to ask yourself. And so what if we created an app that asked you a question, and then it used AI, um, to ma- uh, to ask you questions, and all you had to do was answer five questions, and then it wrote a script in the format of, like, Paddy Galloway, you know?
- CJCalum Johnson
So, so j- just quickly to, to jump in and add context.
- RBRiley Brown
Yes.
- CJCalum Johnson
So Paddy, um, Paddy Galloway, he helped, like, MrBeast grow his following. Obviously, MrBeast is, like, the biggest YouTuber on the planet. And so the problem that you correctly, like, identified, Riley, is that people just lack experience. So you don't actually-- You're, you're, like, creating videos and posting them because you don't have the experience because you just haven't been on YouTube that long. You don't know what the problem is. Like, you don't know what you're missing. And so essentially what you're saying is, what if we actually created, like, an AI, like, YouTube growth coach that basically took on the, like, experience, personality of Paddy Galloway, and now it's asking you questions and it's helping you through, like, your, your YouTube journey.
- RBRiley Brown
Yeah.
- CJCalum Johnson
Like, that's a product that people would pay for, which is basically-
- RBRiley Brown
Absolutely
- CJCalum Johnson
...what you just identified in a few seconds.
- RBRiley Brown
Hundred percent. Hundred percent. And this was just something I thought of last night, like, uh, when I was, like, trying to come up with, uh, some ideas for this episode. Um, w- I think it could be really interesting to create something like this. And I think this is an important theme for people. If, if you're watching this video and you work in consulting, if you work in, um, like if you're, if you're a contractor or you help anyone achieve a specific goal, I think it's a great place to start because you can now take whatever's in your brain and you can convert it into software. And if you think of, like, what a ghostwriter does or, um, or anyone-- Like, if you were to hire Paddy Galloway or hire a, uh, someone who writes content, um, what they're gonna do is they're gonna ask you a bunch of questions, right? Because they wanna know what you're an expert on or what you have, um, you know, uh, you, you can establish some sort of credibility on something. Uh, and they're gonna ask you a bunch of questions, and then they're gonna take tho- those questions and convert them into a script that will work based on their own
- 22:50 – 31:00
Deploy in 15 Seconds (Instant feedback loop)
- RBRiley Brown
expertise. Well, AI is actually-- can, first of all, it can ask you questions, and then you can respond, and it can dynamically ask you questions 'cause it's AI-generated based on what you've previously said. It's not deterministic. Like, it's not like a questionnaire that you would fill out, right? It can ask you questions dynamically, and AI will actually know what information it needs to gather from you, and then it can generate a script in that format. Does that, does that make sense?
- CJCalum Johnson
Yeah. And is, is, is the-- When you say dynamically, uh, what you're saying, Riley, right, is that it's like, it's personalized to you.
- RBRiley Brown
Right.
- CJCalum Johnson
So it's not just a generic questionnaire that anyone-- It's, like, made for you.
- RBRiley Brown
Right. And in here, we can do it through just prompting on ChatGPT. Look at this. So this is actually the best place to start. By the way, prompting is programming. There's this connotation around prompting in AI that it's like, oh, you're a prompt engineer. You're this, like-- It's pompous. It's like prompting is so easy. Don't-- That is a psyop. That's fake. Prompting is so important. And not-- There are a lot of people on the internet who sell prompts or, like, or, like, you learn use these prompts, but, like, the-- And that i- that might be a scam, but the, the underlying ability and skill of prompting is probably the number one thing because it, it is the f- it is the fundamental, um-- It's just being able to communicate your ideas effectively and then get an answer from a PhD-level system, right? If you can Formulate questions and formulate workflows through prompting, you are going to be a great vibe coder or engineer because you can take any prompting workflow and turn it into software for your company. Let me show you an example.
- CJCalum Johnson
All right, quick break. I wanna put you guys onto something brand new that's completely changed how I handle email, and it's called Notion Mail. They're the sponsor of this episode, and it's just launched. And so I've been using them for a few weeks now, and it's seriously helped me up-level how I use email. Less chaos, more clarity, and I'm just not wasting as much time in my email inbox. And so how it works is that Notion Mail connects directly to your Gmail, and then it just goes to work, and it uses AI. And so it sorts and organizes your inbox based on actually what matters to you. No more digging through clutter trying to find emails over here. It puts the emails you care about front and center. Honestly, to me, it's been like having a powerful AI engine directly in my inbox. It's made me way more efficient. It's helped me draft better replies, adjust tone. It even suggests responses. And so if you're tired of typing the same email over and over, or you just feel like you're wasting too much time in your inbox, then Notion Mail is for you. And so here's how you get started. You can connect your Gmail to Notion Mail for free using the link in my description. So go to the description, use that link, and not only are you helping yourself when you use that link, you're also directly supporting the show. So thank you, and let's get back to the episode.
- RBRiley Brown
Let me show you an example. So we can actually just turn on search the web here. Um, and we could say, "I want you to research"... I don't know why we're using-
- CJCalum Johnson
This is in, this is in ChatGPT. This is in Chat-
- RBRiley Brown
This is in ChatGPT. Sorry, I move kind of fast. I switched from ChatGPT to Claude because I know that more people watching this, I would say more-- way more people have ChatGPT than Claude. So this is-- And I'm not even gonna use the o3 model, so even if you're on the plus tier, which is twenty dollars a month, you will have access to this. And I don't know why we're choosing Paddy Galloway as our guinea pig today, but that's fine. Um, so I want you to research Paddy Galloway and, um, and since we're searching the internet, we can just do that, and I want you to come up with questions to ask a user, right? To ask a user, um, that would be useful for writing a script. Um, so just list those questions here. And actually, I think it'll be cool. Let's throw, uh, let's throw in the o3 model, and we'll, we'll search. And AI is now beginning to think, and I think a lot-- not enough people truly understand the fact that AI is thinking now. And you can actually see the thinking steps. And so now he's saying like, "I think the user wants me to gather information about Paddy Galloway, a YouTuber known for analyzing YouTube growth strategy." So p- one of the tools that o3 has is the ability to go out on the internet, find information, bring it back to the chat, right? And so now it's reasoning, and then it's going to literally think. Now it's searching the web. So it reasoned. Um, by the way, our YouTube app is done, but we can come back to that in a second. Um, and so the point is, is this, is like you can create these workflows in ChatGPT and use them every day. And once they're super useful for you, that act of honing the craft and making it useful to you means it might be useful to someone else. At that poi-- at that moment, then turn it into an interface, right? And, and make it into an app and try and sell it. That, that I think is one of the most important things in the modern world, is just to use ChatGPT as a way to solve your own problems.
- CJCalum Johnson
Just quickly before we move on, I wanna double-click on something for people, which is actually the way that you constructed that prompt. 'Cause one of the things that I've learned, uh, with AI, and I remember someone-- I had a friend, and he told me this a year ago when I didn't really know anything about ChatGPT. He said the biggest mistake that people do, especially beginners, is they treat ChatGPT like Google. So they ask it like a very direct, short question, and they get-- they-- Like, Google gives a transactional response. ChatGPT, the way that you're use-- you're meant to use it is actually like a conversation. And so, like, even in the way that you, uh, structured the prompt, you gave so much context on what you wanted. And so I just-
- RBRiley Brown
Yeah
- CJCalum Johnson
... wanted you to, to elaborate on that, Riley.
- RBRiley Brown
I think that's only partly true. I think that's only partly true because, uh, first of all, I didn't give it a lot of context. Instead, what I did is I turned on search, and it-- what it did is, since these AI agents are thinking, it went out and found relevant context for me, which, by the way-
- CJCalum Johnson
Mm-hmm
- RBRiley Brown
... that's happening. And al-- because I, I think that there's two s-- I think that, yes, you shouldn't treat it like search in the sense that, like, it, it's not just gonna give you a list of details that's ranked by how much money these, these companies are spending on Google Ads, right? It's inherently more useful because it gives you, like, an AI-generated response. Um, but, but in the same way that, like, you said you didn't fi-- uh, you didn't create your businesses, uh, because you, like, took way too long to launch, you don't wanna take too long to, like, find the perfect prompt because you just wanna iterate really quickly, right? I said, "I want you to research Paddy Galloway," right? And then it did all of this research. So it went out for me and found context, which, by the way, is gonna get a lot better. And ChatGPT just added memory, meaning it can actually go through, if you give it permission, go through all of your previous chats to look for relevant context to make sure that it, it-- So it'll basically start guessing why you want that information. But after it went through all of this process, it came up with all of these questions. And then so w- now imagine we're cr- imagine we're Paddy Galloway, and we're trying to turn this into our own app, and we're gonna sell it. And, and so now what we could do is we could read these, and we could say, "Okay, I like one, uh, um, I like one,
- 31:00 – 35:22
Monetize & Scale (Adding payments, security, and growth tactics)
- RBRiley Brown
four, seven, uh, nine, and ten. List those only." Um-
- CJCalum Johnson
And this is, like, the conversation piece of it, right? Like, it's collaborative.
- RBRiley Brown
Yeah. And you're, you're, you're going to-- 'Cause, like, our goal here is to come up with what questions do we wanna ask each user when they use our app that we're creating. We're pretending we're Paddy Galloway for a minute, and we're coming up with the questions that are most important. And again, we do not need to have the o3 model on. This is overkill. I just wanted to show the fact that it can pull in additional reasoning. It does take longer. We could easily have used the GPT-4.0 model. In fact, I think OpenAI should change it so it automatically selects it, and it's also searching the web, which we didn't need to have on. And those are little things that you'll learn as you use it more, um, because basically all it should do is list these right here. So it's doing something-- It's taking way too long of a process. Um, but yes, it is a conversation, as you're saying, and, and you're, you're just trying to reach some sort of goal, and you don't need to do it all in the first prompt. Um, so let's imagine for a second that, um, that we want these questions, right? Like, according to Paddy Galloway, rule of clear niche, what's your channel's single sentence? So if we change it to GPT-4.0 and turn off search, we could say, "Don't include, uh, the part about, um, Paddy Galloway." Like, just ask the questions concisely, the same five. Um, okay, let's bring it back to the app. So earlier, you know, just a little bit ago, we, uh, wanted to create an app, and this is the Vibe Code app, the mobile app, and this is what it created. This is on my phone, right? I can use this on my phone. No idea if it's gonna work, but let's see. I want to make a video on, um, u-using AI effectively to create content and, um, help people use AI to build applications. All right, let's see if this works. Always-- It doesn't always work on the first try.
- CJCalum Johnson
But it's also like-- It's like how long have we been at it? We've been at this for, what, 15 minutes, and we already have something-
- RBRiley Brown
No, this is one prompt. This is one prompt, and it worked, right? It's not perfect formatting, but it literally used AI. We created this app. This is a mo- working mobile app that compiled on our phone. This isn't a web app. This is a mobile app. And what we can do here is-- So notice here that the bold formatting, right? It's trying to bold this. Um, and wait, what does this d- button do here? Let me press this button. Oh, I think that copied it.
- CJCalum Johnson
Riley, maybe w-- Just explain, like, what is it that people are seeing.
- RBRiley Brown
Yeah, here. Let's make it bigger. Um, make all of the text on the screen that shows the transcript when it's done significantly bigger and make sure the bold formatting is working properly and make the title of it bigger and all of the icons bigger. I wanted this to be big. [chime] So we can just ask AI to make these changes. This is vibe coding, right? We're building a mobile app, and so this is one of the things all these different companies-- Bolt. You can use Bolt, Lovable, Replit, Cursor, Windsurf. All of these tools are b- are building tools that most people are making web apps, right? These are not apps that would go in the app store. We're building a mobile app that builds mobile apps to start, at least, because no one else is doing it. And so this is kind of our wedge in the market, and I think this is one of the reasons why we were able to raise a pretty good amount, uh, because no one's done this yet. And I just think so much of the next generation is just on their phones that they're gonna really, um, just wanna be able to create apps from their phone, like in line for coffee. Anytime you get an idea, pull out your phone, just toss it in. That's our ultimate thesis. I'm wondering, like, do you, do you think you would-- do you think this is something that you would wanna do just, like, when you're out and about walking, chilling, hanging out?
- CJCalum Johnson
Well, I think, I think about the, the person that's watching at home and, and we spoke about it right in the beginning, right? Of, like, you have
- 35:22 – 1:06:12
Step 3 – Saturday Challenge (Ship your own AI app by Sunday!)
- CJCalum Johnson
so many ideas, but they're just that. Like, they're just dreams until you make it real, right? And so I love the fact that you can go from the idea to, like, something tangible really fast.
- RBRiley Brown
Yes.
- CJCalum Johnson
And, and, and that's how you start building momentum with something, right? And I think, I think building momentum is what makes a business work. It's what keeps you doing the thing-
- RBRiley Brown
Right
- CJCalum Johnson
... is actually having something tangible.
- RBRiley Brown
And what I like about vibe coding-- And, like, we could, we could literally do the same thing in a different tool. So, like, we could build a web app, right? And we could go into something like v0. v0 is actually my favorite tool to build web apps. Uh, we could go to v0.dev, which is another vibe coding tool. And I really liked what you said there about, um, the fact that you could just create it really fast and, like, if you don't like it, just change it. That's, that's where I think-- That's why a lot of stubborn programmers hate the idea of vibe coding because, because engineering is so expensive. Like, engineers cost a lot of money. Being able to coordinate between a ton of engineers is hard. It's, it's, it's expensive from a time perspective. Th-they are very-- They plan. They plan a lot, so most of their time is spent planning. And the, the essence of vibe coding, um, which, like, all of these tools like, like Cur-- uh, Replit or Bolt is another company right now. They're in San Francisco. They're having-- It's a million-dollar prize. Like, people are coming here for a hackathon, and it's all vibe coders. Like, no manual coding is allowed, and, like, all of these people are just gonna rapidly experiment, and I guarantee you someone's gonna create a multimillion-dollar company at the end of this event because-
- CJCalum Johnson
Mm-hmm
- RBRiley Brown
... that is where we are in the, in the cycle of softwares. You can just make stuff, and then you can iterate on it, and then you can sell it if you have distribution, um, which is something that you could think about, building a, a super valuable app for your audience.
- CJCalum Johnson
Yeah. It's pretty, it's pretty cool. Okay. So we're, so we're in V, we're in V0 now.
- RBRiley Brown
Yes.
- CJCalum Johnson
We put in the five questions. Just, just for-- so that people can understand, what is it that V0 does? And I guess you can just show it, but...
- RBRiley Brown
Yes. So V0, let's start off with something really simple, and then let's work our way up. So, um, this uses Next.js, which in my opinion, all you need to know is that I think it is, like, the highest quality for landing pages and simple applications. Like, it's very high-quality design. It just looks professional. Um, and so we could say create, um, just create a simple interface, uh, for an app called -- let's call it Patty, uh, [chuckles] GPT, um, for YouTube script writing. And these web apps look really cool because it, it's gonna open up a side panel, and it's gonna show you all the coding, just like Claude Artifacts earlier. And it is going to create a web app, and it's really easy to get these on the internet. I think V0 does a really great job at allowing you to deploy landing pages and simple web apps to the internet. And their tool's getting a lot more powerful. Um, and we'll, we'll see it pop up right here, and I think it's the best at design.
- CJCalum Johnson
Yeah. Y- correct me if I'm wrong, Riley, but I remember we did an episode a few weeks ago with Greg Isenberg, and one of the things that he did, uh, when we were using V0 is he literally went and said -- Like, he got screenshots of, like, Apple's, like, landing page and, like, Warby Parker, like, and Nike and, like, these big companies, and it was basically able to create a design that actually kind of took design inspiration from just pages that I liked on the internet.
- RBRiley Brown
Uh, y-
- CJCalum Johnson
Oof
- RBRiley Brown
... 100%. Look at this. So we could go to -- Let's go to, uh, one of their competitors, uh, lovable.dev, and what we can do -- I actually really do kind of dig this, their page. We could screenshot this right here. We can copy this to our clipboard. We'll go back to V0, and we can say -- So this is what V0 created, right? Maybe it's not the best design in the world, but it has, like, what's your video idea about, um, script length. It gave you a little toggle that you can edit and casual and friendly, and then you can type in your idea and press Enter. Now, no AI features were added. This is dummy text. Like, this is just, like, f- um... It uses, like, dummy text to mimic what the app will look like-
- CJCalum Johnson
Mm-hmm
- RBRiley Brown
... which is why I think it was so important for us to just add AI features natively because in order to use AI features, you need something called an API key, and most developers don't wanna do that. But I'm gonna digress on that, because what we can do here is, let's say we're just focused on the appearance. We can paste that screenshot that we just took on Lovable. Let me see. We'll screenshot this, and we'll drag this in. So we just uploaded that screenshot and say -- I'm gonna say this. And also, one thing about vibe coding is I use my voice for almost everything. I want you to please make the app stay exactly the same in terms of functionality, but from a design perspective, make it look like the screenshot. And so now we can run this, and the tool that I just used to use my voice is -- I think it is the most underrated tool on the Mac, which it's called WhisperFlow. Um, I actually do know the founders a little bit, um, but it is an app that just sits in the background of your computer and lets you, like, talk to your computer. You never have to type, and it, like, it has a way better -- It's way better than the one that's built into the Mac, and that's part of the fun with vibe coding is you don't even need to touch your keyboard.
- CJCalum Johnson
So you, you talk into your computer, and then what? It transcribes it, or it just means that-
- RBRiley Brown
Yes.
- CJCalum Johnson
Okay.
- RBRiley Brown
It immediately transcribes it.
- CJCalum Johnson
Very cool. And what was that called again?
- RBRiley Brown
And then, like-
- CJCalum Johnson
WhisperFlow?
- RBRiley Brown
WhisperFlow, but it's spelled weird. So it's W-I-S-P-R-F-L-O.
- CJCalum Johnson
Okay.
- RBRiley Brown
Um, for those of you who want to try it out. Um, I think it's really good. Other people use another tool called Super Whisper. I don't think it matters that much, but it's makes coding a lot more fun 'cause you don't have to, like, type it out every time. I can speak a lot faster than I can type. And yeah, it's generating.
- CJCalum Johnson
So it's literally building, like, the front-end design of how this page is gonna look based on-
- RBRiley Brown
Yeah. I mean, it's editing it
Episode duration: 1:06:13
In the uListen app, the transcript lights up word by word as you listen. Get uListen
Transcript of episode J_GCvfto07c