A year from now, these apps are going to be absolutely mind-blowing. You'll probably be able to prompt a whole

SAS company into existence. Right. Yeah. And the kind of stuff that you'd have to be a developer to do before,

you'll be able to do with these tools. Right. Yeah. So if you're in business, I would be thinking about what that means in terms of opportunities;

like, what are you going to be able to do with this new technology once it gets that good? The way I've been using these

tools is to create really basic, simple mini-apps that sort of clear bottlenecks in my business. Hey, welcome to The Next Wave

podcast. I'm Matt Wolf, and I'm here with Nathan L., and we're entering into a world

right now where the gap between having an idea in your brain and actually seeing a working app, a working piece of

software on your computer screen—that gap is closing quickly. And we're entering a really exciting time

right now where anybody can write code to do pretty much anything they need to do. And there's a whole new crop of

up-and-coming AI coding tools that we've been hearing so much about, and we wanted to put them all through their

paces. Tools like Bolt, tools like V0, WebSim, Replit AI—these are some of those AI coding tools that we're

starting to see everywhere, and people are just raving about them. So we wanted to find out what they're good at, what

they're not so good at, and really put them through their paces and see what we can build with them. So let's go ahead

and just jump right in and play with some of these up-and-coming AI coding tools. We recently did an episode with Riley

Brown where we walked through building out an entire app step-by-step with Riley, using mostly Cursor for that.

project, but there's been like this new crop of AI tools popping up, like Bolt, new and V0 from BCEL, and what was the

other one? Replit Agent. Yeah, right. Yeah, so we've got all of these new AI tools that are

popping up that are just making it so much easier to code. And I mean, like, even easier than what we saw with Cursor.

Right, like with Cursor, uh, even when Riley Brown showed it to us, we kind of started with a template, and then we had

like a starting point, and then we used Cursor to build on top of that template that Riley provided us. Right? Well, with

these other tools like Bolt and V0, you don't even need that template. Right? You just go in there and you say, "I want

an app that does X," and it just builds it and it gives you a working app. So that's

what we want to do on this episode is play around with some of those tools, put them through their paces a little bit,

talk about the implications and, uh, see what we can get them to build for us on this episode. Yeah, I guess let's just

jump into Bolt. Yeah, so Bolt is like really simple. Right? This is this is the screen you get when you first, uh, log

into Bolt. You can come down here and choose what you want to use for your code. Right? You've got Next.js and Remix

and Svelte and React. I use Astro. Most people don't even know Astro. It's like, I don't know what Astro is.

Yeah, yeah, yeah. My website's all I coded in Astro, which is like—Astro is a relatively new framework which, for like static websites,

is probably the best, like, 'cause it loads the least JavaScript, and so it's the fastest to load. This kind of stuff, but

it's it's a bit intimidating to learn, so it's it's awesome that they have something like this here. Yeah, well,

This one, you don't even need to pick which sort of, uh, framework you want to build on top of, but you can just

you know, give it a concept and it'll build it out. I like to ask it to build a

game first, because I always feel like games are a lot more complex than what most businesses are going to use it for,

and it's sort of like a really good demonstration of something more complicated than what you're probably

even going to use it for. Last time I did it, I had it build Tetris. We see everybody on Earth use the Snake example,

so I don't want to do Tetris or Snake. I was thinking maybe like Space Invaders. Have it build Space Invaders real

quick, so I can just give it the prompt: build a Space Invaders game. And then just like that's

it. That's the whole prompt; four or five words. And then you can see on

the screen, it's just coding. I'm not—I'm not touching anything. It's going and writing all the code. You can see over on

the left, it's already, you know, got all the pages and everything's already built and ready to go. It's just—it's—it's

just creating it. I'm not doing anything. And, uh, let's see. Let's see if just one prompt is enough to make this work. We

can see here, over on the left, it says, "I'll create a modern, beautiful Space Invaders game with React and TypeScript."

It will feature smooth animations, particle effects, and responsive controls." So here he says, "I've created a modern Space Invaders game with the

following features: smooth canvas, particle effects, uh, score tracking, modern UI, responsive controls with keyboard input, beautiful start and game

over screen, performance optimized with proper frame timing." So you can see over here on the right of my screen, we've got

Space Invaders. I can click 'Start game'. I don't know what the controls are. Okay, so I've got a square that moves.

So it didn't really work on the first try, uh, which is often the case with AI. So I'm just going to give it the prompt,

All I see is a square that I can move around on the screen. I'll just give it that

little amount of info and let's improve the game. Now that's going to go on and code some more random side story, but

this is reminding me of back in the day when I launched my startup, gy.com. It had this waiting list, and when you

would join the waitlist, after you gave your email, you would be prompted to fill out this puzzle, which

was like using our logo, and you had to put the puzzle pieces in correctly. And it timed how fast you could do it,

and then we had a leaderboard, and everyone was like sharing this. It was like this crazy viral thing. It'd be

kind of cool to replicate something like that where you have a little game and use that as a marketing

tool of some kind. We'll have to see if, uh, V0 can do that when we switch over

to V0 to test that out. Right now, it's just going through, writing all the new code. You know, there's a whole bunch

of different files that it's already generated, and it's just jumping around between the various files and fixing the code. You can already see here: it's one,

two, three, four, five, six, seven different files it's already worked on since I gave it that last prompt. Yeah, it's so

cool that it gives you all the code and like does all the folder structure and everything for you. Like I

remember some of the early AI website generator apps or whatever. A lot of them didn't give you code; you just saw

like the final product and it's like, okay, now what? Yeah, yeah, for sure. So here it says left and right to move, space to

shoot. So hopefully this time we have a working game. A little bit closer. Apparently, I died already. I think this may be

complicated for it, something. Maybe Space Invaders is a little bit too complicated. I know I can get there with like two

more prompts probably, but what I really wanted to talk about on this episode was the concept of like just creating little

mini-apps for your business. I think a lot of people see these tools like Cursor and Bolt and uh, V0 and Replit and

all of these kinds of things and they think, well, I'm not trying to build a software company. I'm not going to go and

launch a SaaS; that's that's just, you know, I've already got my business. I don't want to go do that, right? But the way

I've been using these tools is to create just like really basic, simple mini-apps that sort of clear bottlenecks in my

business, right? Like one thing I was running into is a lot of times I'm pulling images off the internet and they

get pulled as webP files. Right? Like if you generate an image in DALL-E inside of OpenAI and you download it to your

computer, it downloads it as a webP file. I really hate webP files; they're kind of annoying to work with. I want it to be

converted to a JPEG file. I can't use webP files in DaVinci Resolve when I'm editing my videos to overlay them onto

videos. I need JPEG files, right? So I created a simple little app with a single prompt where I can drag and drop

as many webP files into that into that little app, and it would just convert them all into JPEGs like instantly. In

fact, here, let me let me actually show you the little app that it developed. It just created this little Python script.

called webpconverter.pdf right there. I made it so you can, like, change the folder; pick wherever you want.

them to automatically be saved. And then I've got—let's see—I've got these three WebP files that I just dragged

over here. One's a picture of a tongue that I downloaded for some reason; here's a picture I made—that's not your tongue, okay?

Good. I think this is like a doll. Yeah, this is a dolly image, and this one's also a dolly image, but they're all WebP

files. And so if I grab all three of these, just drag them and drop them into this little app that I built here, and it

says, "Conversion complete. Successful: 3, Failed: 0." And then if I look at my saved folder here, it just saved all

three of those as JPEGs.  Here's a really, really simple app. I know there are other apps on the internet that already

can do that kind of stuff, but I wanted to be able to just do it in bulk really quickly without, you know, ads all over

the screen and all that kind of crap that you get when you go to those online file converters. And so I just—I

had uh, Bolt build me this little simple app really, really quickly. Downloaded the files on my computer, and now, like, if I'm

working in DaVinci and I'm trying to find, you know, images to overlay into my videos, I just have this little thing

open on my screen. Whenever I download an image, I throw it in there; it gives me the JPEG back. It just saves me a few

extra seconds on converting those files because the way I used to do it was I would open it up in, like, the

Windows image viewer app, and then I would resave it as a JPEG file. And this is just like way quicker. So, right, that's

the way I've been seeing these apps is just like little ways to sort of solve bottlenecks in my workflow, like

What's something that I just... I want to solve this little problem. I don't really need to give this app away or sell it.

It's just for me, and I found that really, really helpful. So, you know, I think anybody can look at their business and

go, 'Where are the bottlenecks? Where are the things that slow me down just a little bit? Can I create a little quick

script that's going to make that thing just slightly easier, speed up the process just a little bit?' And that's

what I'm loving about these tools is just like finding those little things that just make my life a little bit

easier. Yeah, and the kind of stuff that you would have to be a developer before to do any of this, right? Or, you know,

now if you're... you could be a CEO and have an idea of something that would make your business a little bit better, a

simple tool that maybe before you'd have to hire an entire team of engineers to build it. And now, if it's something very

simple, then you could just use something like Bolt. Look, if you're curious about custom GPTs or are a pro that's looking to up

your game, listen up! I've actually built custom GPTs that help me do the research and planning for my YouTube videos, and

Testing 4 New Ai Coding Tools: Bolt vs V0 vs Replit vs Websim.AI

Scripts