Watch me Generate Wallpapers using Midjourney AI in Real Time!

In this live stream, the host walks viewers through the entire workflow of creating custom wallpapers using Midjourney, an AI image‑generation service accessed via Discord. He explains how to join Midjourney’s Discord server, set up a private server to avoid crowded queues, and craft effective prompts—including version flags (v4, v4.2), quality settings (Q2), resolution tags (8K), and aspect‑ratio specifications for phones and laptops. The stream covers the underlying diffusion process, how to upscale and create variations, how to manage the 25‑free‑prompt limit and paid subscriptions, and practical tips such as using ChatGPT to look up device resolutions. Throughout, the host engages the live chat, generates images on‑the‑fly from audience prompts (e.g., retro sports car, asteroid belt, Milky Way), and demonstrates upscaling, variation, and exporting. He also highlights the emerging role of prompt engineering as a professional skill. The session ends with encouragement for viewers to experiment, copy prompts, and continue exploring AI‑generated art.
Chapters
All right, cool. I'll get started here. Feel free to leave a comment, chat. I'll be able to see them and respond to them. Hopefully fast enough. But what I'm trying to do here today is, um, just gonna show the power of AI with generating art. And this is with, um, I'll be using, uh, Midjourney. It's one of the different flavors of AI that's out there right now. So a lot of people are using different ones. Um, there's DALL-E, OpenAI, there's other different ones for generating art. But for Midjourney, I did a, I did a live stream before, a couple few weeks ago, I want to say 4 weeks ago. And it was more of like just using this Midjourney— this is Midjourney's website, by the way— uh, using Midjourney to generate some art. And I was just, just showing how you can actually generate art using Midjourney, and it, and it looks really looks really good, um, the art itself. So I generated a few of those and I can show you. I have some on this site here. I— these are like recent ones that I generated. So from here you can see like it generates some really cool images and you can just go crazy with anything else that you want to generate. Um, you just give it a— you say what you want to get generated and it does generate that for you. So this here is an example of some of the ones that I generated a couple weeks ago. Um, so whatever just kind of comes to your mind, you generate. For example, this, I just put it as black surfer surfing through the galaxy and V4Q2. These are just like codes to let the bot or the AI itself know, I guess, what styles it should use on, on the image that generates. For example, this one just tells it like I use version 4 of this, uh, Midjourney AI to generate this art right here. And also this Q2 part is more of like an enhancement to your image. It takes a minute to generate, but that's just the, just the importance of it. So I'm just kind of like going around sharing how I generate these images, and so other people can know how this is done. But it looks really good. Like you can see, for example, this, this is I mean, human could probably— humans, we're really good at generating images and art and all that, but I think this one does also a good job of like enhancing our— yeah, it does look good, Asko. It does really look good. Um, and this is just like an example of how this API can— like, it does a really good job of generating art. And what I'm trying to do today in this live stream is I want to show you how you can create Um, wallpaper for your, for your, for your laptop, for your phone. I actually have one for my phone and it looks pretty sick. Um, let me just show you an example here. Let me do that. So I'll, I'll show here because I think I generated it somewhere in here. So we'll try to do the same thing today. Uh, and then I'll show you how it— like what the prompts I put in and how it showed up, because you have to also consider the aspect ratio of your phone and your laptop be different. But let me scroll up so you can see. It's like, uh, I put in like a retro car, uh, driving on a sand dune. And like all these are just kind of like different things I tried before. So let me go to that. Which one was it? The one I have on my phone right now, which I want to change eventually, is— where is it? Not that one. This one. Yeah. So for example, look at those ones. Like, this was— they look pretty good. And this is like a wallpaper for my phone right now. And this is what I'm using. Yeah, this is what I'm using to on my phone, and you can also use it on your laptop. For example, you're gonna see my laptop. You can see it's like this one was generated by that AI. I didn't really use it on— like, this is generated by it. I just use it as my background for my laptop. And you can literally generate any idea you have in your mind. Think of anything that you want to generate, type it in. It's going to generate something of this sort, right? So that's just what I'll be going through today. I'm just kind of like demonstrating how you can do this on your own by yourself. So first of all, to get even started, you have to go to their website, which is midjourney.com. You just type in midjourney.com. If you want to do this with me as well, you know, feel free to do that. Uh, jump on and get started. And if you run into any issues, let me know here in the chat. So we can troubleshoot together. I like troubleshooting together live. So from here you can see that you can join the beta. The beta is just like a pre-preview version of what's happening, and when you click on that, it takes you to the Discord channel. If you don't know what Discord is, just type in discord.com as well. You'll see it pops up. So Discord is basically like a chat for gamers. It's really good for gamers, but also like it's for everyone these days. Yeah. Just go on Discord. You can literally just do whatever you want to do on Discord. Chat, you can— there's so much to do. Uh, so let me just minimize the chat box so that we can prioritize the— but yeah, feel free to chat, leave comments on the, on the comment side right there. So this is what we'll go ahead and start. So you'll see. Okay, cool. Once you accept the invite, it will take you to somewhere like this. And this is a Discord, so you can create a channel. And this is the original— this is the Midjourney channel. It will take you here to their channel. And when you go to their channel again, you'll see a whole bunch of stuff going on here, right? So this is the home, and this is where everybody, I guess, in the world that's using Midjourney is at. So this is where everybody jumps in. You can even come to announcements and you'll hear like whatever new updates they're passing out or anything really interesting that they have, you'll see them generated here. And from there also, but I think the most important part that you want to jump into is you, as you're starting, you're gonna come into newbies here. And on the newbie section, this is for everybody that's pretty new, and you'll see what other newbies are generating, and you can see the text prompts that they're using here. And I guess just to preface all this, you have 25 prompts before you hit your limit because it's free until it's not free, right? So basically you have 25 of these prompts that you can use, which is a lot really if you prioritize what you— but feel free to go play with those 25. You have 25 for free, right? Why not go play around and see what you can, what you can come up with? And you can even use some of the prompts that I'm using if you want to experiment. But on the newbie section, you can see what people are generating. This is basically somebody generated this and The only words they used was people spend their time on screens and their phones. And that was kind of— that's what gave them that result. So feel free to just go there and experiment and see what you get from that. So yeah. And there's so many other channels within this server. So there's so many newbies. There's 1.5, 185, 5. And so many of them. And while you're generating this art in this, uh, channels, the only issue I have was probably that big issue, but, um, it is a big issue. Let me not minimize it. So a lot of people, when you generate your prompt, because there's so many of you in here trying to generate prompts and all this other stuff, you'll get kind of like lumped up between other people's stuff. So when you write a prompt, for example, let's just do here, imagine, so This is how you generate a prompt. You just write imagine and it will tell you imagine prompt, and then think of anything and type it in there. So I can say generate, or not even generate, just say 8K image. Let's see, imagine. Let me go back. I messed that up. Yeah, you have to type in within that black box right here and hopefully beyond that. So 8K. That's effective resolution. Of the image that I want, aka image of, um, milkshake maybe commercial, right? And let's use version 4 of this engine, which is the AI version 4 of this. And let's just use Q2 as well and type. So when you enter it, you You can see like it's waiting, right? And you already see like other people's prompts are getting flooded between what you, what you wrote, what you wrote, right? So you have to like keep scrolling up so you can see your stuff, right? Generate. Because other people's stuff are getting like— your stuff is getting buried underneath because everybody else is generating this stuff. So yours gets buried under here. So what I— solution that I figured for this was I would just go to— since you're already on Discord, create your own server here, random server, and invite this bot right here, the new Johnny Bot. Just click on it and then add to server. And when you click add to server, it will show you— add server, you pick. These are the servers that you have access to and you have like admin rights to. So then you're able to pick whichever one. For me, this is mine and I picked that one and then I invited it into my server. Um, in this case, let's just go back to what we Generated up here. After finished, find our our drink, our milkshake drink. So yeah, so here's our milkshake that we generated, right? This is our milkshake that we generated with that prompt. And after here, you can enhance whichever one that you like. If you like this one, you can enhance it and it'll look much better than that. So instead of like getting buried here and trying to find where we put our where our result is. Like I said, you come to your own server, and within your own server, once you invite the bot to the server, then you're able to come and generate your prompts here. Let's just generate the exact same thing. So imagine, um, create an image of milk— of a milkshake commercial. What milkshake? Commercial, right? And also we want to use the version 4. All right, let's send that in. And as you can see here, it's not being crowded by other people's, uh, generations, right? So you're able to see just your own stuff. But remember, others are able to see your stuff, so your stuff's sticking to populate in the other main channel. But this is only for you, so that way it's not like crowded within other people. You can pay extra money, more like a subscription service, to be able to have this private so other people can't see what you're generating at the end of the day, right? So you can see it's already like working on generating this. And I guess the tech behind how it generates these images, and as you can see here, they look pretty good, you know, for a first prompt, right? Um, what really happens here is that it It takes, like, it's been trained with all these images from Google, right? So for example, maybe you got all the images of like milkshakes on Google that me and you post there. Maybe you go buy a milkshake, you take a picture, you post it online, right? So it takes all of them and it's trained on all of them. So it kind of knows how milkshake looks like. And what it does is like it does what's called diffusion. So basically kind of like walks back on the image, adds noise to this image, the original images, It adds noise to it until the image is not decipherable anymore. You can think of it as like until it looks like a— remember those TVs back in the day when, uh, you know, the old Great Wall TVs, um, that would show like those like signal, like, um, noise on the screens and you'd see like the fuzzing and stuff? So it makes— it adds noise to the image until it looks like that. So while it's adding the noise, It's learning that, okay, this image, when I add this much percentage, it looks like this. So it learns that way. So when you tell it to generate the image, it starts from the noisy part and it brings it all— it knows how much I need to add until it looks really good. That's a very weird way of explaining it, but that kind of makes— if it makes any sense, that's like an idea of it. So once it generates the image and you like it, if you don't like it, you can hit this like recycle button. And basically what it does is kind of like regenerate the prompt again. So until you like, you get an image that you like. But in this case, maybe I like, maybe I like the second one. So it counts from 1, 2, 3, 4. So that's the order of it. So maybe I like, I don't know, maybe I like the second one and maybe the, I don't know. Yeah, maybe the second one and the fourth one, right? So U2, U3, U4. What does all this mean, right? This means upgrade. That means I want to upgrade that image to look better. And V down here means I need a— I want a variation of that. A variation of that could mean that instead of all these swirls coming this other way, maybe the swirls go the other way or somewhere. It just kind of gives you a variation of that to look something different but similar to that. And we'll test them out. So let's test out first making it look, you know, upscaling it. So let's upscale this second image, right? Let's upscale the second one and let's upscale the fourth one. Right. And for fun and giggles, let's make a variation of number 1. The background looks really nice. Awesome. And let's give it a moment. So now it's generating them and it's going to tell you fast, fast, fast. And this is basically allocating memory for your prompt. Right. So if you haven't used up all your prompts, it will give you the fast way of accomplishing this, or it gives you the fast— gives you allocated more memory to your image first generation. But if you hit your limit, it will kind of like stall your— it will prioritize other people first before it gets to you. So basically, you know, you get put behind the queue and it will take time to generate your prompts. And after you hit the 25 images generated limit, then you have to pay a monthly I think there's like a $10, $30, I forget, but they have it on their website for getting started on all that. I don't want to go into much of that. If you're interested, you can look at that on their website. Um, but let's look at the images. So this was the variation that we generated, right, for number 1. And you can see like it generated different variations. To me, it looks really weird. Like, I don't know, I mean, it thinks this looks nice because I mean, the straw is definitely out of place because like if you follow this straw, unless it's like bent here and goes in, this one looks more accurate going, but then it just doesn't look realistic, right? Because like this swirl is just gonna swirl different than that. Um, yeah, it doesn't make sense. I'll leave that alone. Now let's look at— it's finished upscaling the 2 that we had upscaled, right? And you can see like it looks pretty good so far for a commercial. So if you were like a designer, you're maybe you're making commercials, and you know, this is a good starting point to give you like maybe even ideas of how you want your commercials to look like or how you want to design a website, right, that involves chocolate or milkshake in this case. And this is the other one, um, as you can see, it looks, it looks super nice. Um, I'd say this looks pretty good. You just have to add a few words and you have your image. So that's a working sample you got. So that's a good starting point, but let's go back and get started with generating our wallpapers because that's what we're here for, right? And feel free to leave comments in the chat, or if you have any, also feel free to write your prompts, what you want me to generate for you. And I can send you the pictures too if you like them, the ones that we generate from here. So if you have a prompt that you want me to generate, let me know. I will generate it for you. So let's start by generating the car prompt, right? Let me start by generating that. And we want to generate maybe a wallpaper for your phone, right? And let me use the one I generated a while back, and then I'll show you how it looks like. So this is the one I generated a while back. So it says the 7— 750. By 1334, extremely detailed 8K image of a retro sports car driving not in the sand dunes but on sand dunes, right? And basically this, um, 750 by 1334, I think these are like, uh, dimensions for, uh, iPhone 8 Plus if I'm not wrong. You can look up dimensions and you can look that up, right? Matter of fact, let's just ask ChatGPT, right? While we are at it, I'm just kind of like combining all the AI stuff here today. Let's ask. Let's just ask ChatGPT. What are the phone dimensions? And save resolution. All right, let's see what it tells us. Here we go. Some phones resolutions include 720 by 1280, 1080 by iPhone XR, X2 Max XR. So I guess full HD. Quad 4K. Let's go by— let's go with 720 by, uh, 1280, right? For example. So let's come here and replace this 750 by that. Let's replace that by 720 by that. And let's run it and see what it gives us. So it's gonna run and give us, um, some images. Yes, I do have this recorded. This is from Asbo. Have you hope you have this recorded? I would love a copy of recording for best. Yes, this is recorded and it's going to be live on YouTube on my channel. You can see the YouTube channel down there. You go on YouTube and type in Ed Onikipimbo. It's going to show. It's going to show you my. And you'll be able to find these recordings after this. I'm also. It's also streaming live on. Facebook, Twitch, and YouTube. But you'll find these videos on YouTube. You can also find my previous recording on YouTube and you can watch that as well. I'm trying to go live as much as possible, so feel free. Also feel free to share your prompts that you want me to generate here. So this is a prompt I just generated for a retro car on a sand dune, right? And you can see how they look like. This is a retro car. It looks pretty good. But I think, I don't know which one would— let me just generate another one. I don't think I like any of them so far. This one looks okay, but the wheel looks a little weird. This one will look nice on— I guess I'm just trying to imagine what would look nice on a screen, right? On a screen. So let's generate some more. Maybe this one. Maybe number two. Let's let's upscale number two and see what we get. So it's generating some more. Yes, feel free to put your prompts on the on the chat and I can generate them live here. If you like them, I can send them to you. If you don't like them, and also please share your phone dimensions. Let me know what phone you have. Put it in the chat. Let me know what phone you have. Like what what type of phone you have. I'm gonna ask ChatGPT to tell us the resolution for your phone, and I can generate the image and send it to you if you like it, of course. And it generated more. I think I like this one and this one. This one looks a little too yellow for me, but, um, yeah, I mean, that's a phone resolution, right? You wanna— that pops. So that's upscale number 2, number 3. And number 4. All right, so also another tip for you to consider is that, um, when it's telling you job is queued, so it only takes, uh, I think 3 jobs at a time. So right now it's running 3 jobs, and if you have a lot of things running at the same time, so you have this one, uh, running, this one, the upscaling, and then you have this both The more we keep like adding more jobs to it, it's gonna keep it on a queue. So when that one ends, it's gonna keep running. When that one is finished, it picks up the next job and kind of runs it. All right, well, it's generating some stuff. Let's see. All right, just finished with this one. Zoom in so you can see it. Um, look, it looks okay. I can't say that it looks amazing or anything. Um, one of those— I don't know if that's a mirror or that was supposed to be like, um, part of the torch or the lights. Sorry, it just seems out of place, which is okay. Uh, if you I mean, if you're really good at something like, um, what's it called? Not Lightroom, the other one. Photoshop. If you're good with Photoshop, you can go ahead and like extra edit this image. But you can, you can tell like it's already giving you something that's really nice to start working with, right? And you can do like more with it. I think I like this more. This one looks much better. Um, I mean, it'll look really good on the phone and you can go and like increase the light or increase the coloring a little bit more if you want to. And but this is already good to go, just go to your phone straight up if you want. And I think that's a really good thing about, um, playing around with this, uh, AI. And it's pretty much free, and for the first 25 prompts, remember that. So it's free for the first 25 prompts, so feel free to kind of like play around with this. Um, I'm waiting for some of you all to send me some prompts. Send me some prompts. And basically a prompt— don't be scared, like, to share anything. Like, a prompt is basically words. It's just like this. You can tell me, like, you want an image of whatever. Yeah, the main thing here is I told it to give me a retro sports car driving on a sand dune, and that's what it produced. So feel free to share what you, what you want me to pull, If you want me to ask it. And like I said, if you like the images that I generate with your prompt, let me know and I will be able to send it to you after this. So feel free to share what you want me to generate. It could be anything. It could be anything really. You can see this one. All right, cool. What else? What else? Tell me what I need to generate. I want to be like generating by myself. So please type on the chat. Send me what you want me to generate. I could generate it for you, and then I still have a lot of free credits to use on this account, so please feel free. I'm gonna, I'm gonna use most of them, so you can help me use most of them. Um, let's go with the retro sports car again. On a racetrack, let's use the version 4.2, and let's use a style of— I'm just telling it to give me a style variation of 500. And then let's also use Unreal Engine. Unreal Engine. Um, all right, no QA, thank you. I'm gonna switch that next after this one, so let's just queue it behind this one. So imagine family. I'm just gonna add a few stuff in front of your, uh, prompt here as well, so it Extremely detailed 8K image of the Milky Way galaxy. And just use version 4 for better detailed image and v2. And let's give it a style. It's going to go with the styles of 500 and then give us from there variations, different images variations from that point. It's going to take a minute to generate that, but it's basically prioritizing quality over speed, which is why we're getting that. So it will take a minute to generate. You can see these images. I mean, they look all right. I think the Unreal Engine gave it a more cartoonish look here, which is not bad, but it's definitely something that I personally don't like as much. But this could be You know, good starting point for some stuff. Um, let me upgrade upscale number 4 and number 2. All right, Asbel. Oh, it generated— look at that, like it's doing a freaking good job at this. Um, which one would you like me to upscale? You'd like me to upscale all of them? Let me just show you. So this one is number 1, number 2, number 3, and number 4. Uh, which one would you like me to upscale? You want me to upscale all of them or individual ones that you— is there one that you like most? I think we should definitely upscale number 1. Number 2, why not? Let's just upscale all of them, right? Um, like, it doesn't hurt. We have a lot of credits to use here, so I'll just upscale all of them. And boom, let's just give it a moment. So all of them are queued up. We'll give them a moment. It's gonna take a minute to run, but this is just basically, um, Just number one is fine. All right, cool. Um, I guess I just upgraded all of them. We'll see, we'll see what comes off of them. But I like this, uh, and this is for me— the reason I'm doing all this is just for exposure purposes again, um, for some people that are here to watch or some people that want to watch this later, um, just for me to show you what's out there, what's happening. Even if you're not like super into this AI tech stuff, just being Being aware of them and being able to play around with them, I think, is very important. To be aware and just to know that these things exist, I think that's enough to get you going. Let's see, so we have this Flexor Sports Car. I mean, it's all right. I'm not too crazy about it. Let's see. Let's continue. Not too bad. You can use it. I don't know. I wouldn't use it for my phone. But I don't know, uh, you could probably go to a wall somewhere, maybe. Um, but yes, anybody else has any other prompts while we wait for this? If you have any prompts, please put it in the chat. I'll be able to generate one. Um, some like the one, uh, Albo just shared here for me to generate. Okay, yeah, I mean, yeah, they don't look that bad. I'm just bad. I'm just being really picky. risky here. I think it depends on who. But here's the one you, you asked to, uh, as well. Let me open it in a different browser so we can take a look at it. See, it's— it looks, it looks pretty good. I like it. Like, it's— you can zoom in and see how it looks like. Like, this would look, this would look great somewhere, right? Like, this is what it imagined. It took all— it basically just went and found all the images of the Milky Way galaxy Taxi impressions on Google that and and you have to think of it like what this AI is doing is not like it's not like inventing something overly new that humans can do basically it's it learned from us definitely it took all these images from us we gave it inputs it learned and what it does it it it learns and then it it does better it kind of like expands our imagination it combines all the images that we've ever put on the websites or on the internet, basically. And it tries to imagine something totally different from those. It just uses your words to try to mix and match what you've learned to create something novel from that. I'm sure you've seen some of them that people are like generating new faces that don't exist, for example. And those are like really good, I guess, use cases that are out there. Um, yeah, but there's so many of them. Okay, here's another one. They look really good. Let me know if you want me to send them to you after this, um, because I mean, they look pretty good. I'll be able to send them to you. Also, if you want, um, you can tell me your phone, uh, like what phone you want to have those on, for example, and we can generate the one that matches the resolution for your phone so that way you're able to match, or it's able to match your resolution of your phone if you want me to. All right, Oriana, welcome! Generate asteroid belt, let's go! I like these vibes with space today, I like them. Let's keep them coming. All right, let's generate the asteroid belt. All right, extremely Let's see, keep typing over this. Once you delete the prompt, you have to have it within that box. So let's start with extremely detailed image of the asteroid belt. And let's use version 4. Let's use Q2. And variation style, variations up to 500 different variations. Just gonna like sample them and send us which one's the best. Um, give it a moment as well. Here's your Milky Way. It looks really amazing. It looks full of life. Um, yeah, I like what you— which I like, I like that all of you are, um, are sending in. Please send in some more. Feel free, feel free to share some more. All right, look at that. That is— whoa! Oriana, maybe you can tell us more about asteroid belt. Are you, um, are you an astrophysicist? Are you into astrophysics, um, or anything of that sort? Because you can help us interpret some of this because this one looks really amazing. I can imagine like the asteroid belt is, um— Yeah, it is, it is amazing. I can imagine the asteroid belt being being like very, um, very cold, a lot of rocks, because that's like the expanding, um, part of the, uh, universe, of our universe, at least of our Milky Way. Um, yeah, that's a lot cooler, right? Yeah. Um, yeah, I think all of them, like, this one's really cool, and we haven't even upscaled them yet. That's the crazy part. So, um, Oriana, let me know which ones you want me to upscale. Upscale? You want me to upscale all of them? Because I think I'll upscale all of them. I'll just ask you a lot of them. Why not, right? Let me just ask you a lot of them. And I also found something really cool. Um, if you— I'll show you guys after this, but, um, guys and girls, um, but one thing that I found like really interesting is, um, you can have this and like there's something, I don't know, they call it like the Nolan case. Like, um, let me find that image and I'll be able to to share. Like, you can have— like, basically you can have those— like, we can put any image that you generate. For example, you can generate this and put them in like a glass container, and it's going to look like freaking awesome. Um, at least that's how I think. Uh, I'm not sure if everybody thinks that's cool, but yeah, please send me some more. Send me some more prompts. I wanna— I like this. I like where all this is going. Generate some more. Um, I have free credits, so we can use them as much as we need to. I don't have the limit. I think I passed the limit of 25 a long time ago. So I paid some money to have this extra playtime to play with them. So this is taking a little longer to generate, but let's give them some time. In the meantime, let me find the ones I generated a while back that are in that glassy mold and whatever. Uh, let's see. All right, I think I found it. All right, here it is. Oh man, this is— this looks sick. Seriously, thank you for the prompt, Oriana. This one looks like— look at, look at this, it's like rocks. It's like molten rock or something, and we have like—I don't know—is that—is that like lava inside of it? Not even lava. It's like heating up, and it's like moving so fast too. You can see like the dust coming off of it. But I don't know, like it's vacuum of space, so I don't know if you would like really see. Oh yeah, I guess you would see it because I think I remember when the comet came by a long time, like not a long time ago, but a few few years ago, we were able to see. It was at the Oumuamua. if I'm not wrong. You were able to see like the stardust behind it, like the dust from it, and that was pretty interesting. So man, I like this. Oof, look at that. That looks— I like the depth and just kind of like the clarity of it. You can see like— but it's very strange because that looks like a surface right there. Would that be like a surface of some bigger rock something, I'm guessing. But that looks really cool, I like it. Let's see what else. Let me know if there's anything else you want me to generate, leave it in the comments. Let's get this— most of this stuff generated. Oh, yeah. Oriana, let me know as well which one you want me to send it to you. I can send you any of this, um, if you'd like. Um, down here and see which ones we have, if we have some more generated. Oof, look at that one! That one is pretty sick. I like that one too. Oh yeah, you can see like the, the entire expanse over there of the universe, at least of, of the, of the galaxy. I think the main thing here for me that I've always been realizing is like we always kind of like limited with your imagination really on what you can. I think prompts, writing a writing a really good prompt is something that it's very undervalued, but it's very important to get like what you want. So I think there's there's even a job now that's come up that's called prompt engineering basically. You get employed by somebody to write prompts and generate art based on that, and it's, it's a job now. That's a new job that's popped up from all this new, um, AI development advancement that we have going on. So it's pretty cool to see some of this stuff coming up. So yeah, I think there's so many other jobs that we don't know yet that's still going to pop up. Cool. All right, which other ones do you guys want me to, or Want me to generate? Let me know. Oh, I was gonna do the the one that I mentioned earlier about imagine. I put a prompt from a while back when I did this. So a man on the moon, and this is basically the dimensions of what I want. So just basically the entire prompt is a man on the moon, and then I reuse this other part. Which is the dimensions of what I want. And display case of knolling item teardown, transparent data visualization infographic, high resolution OLED graphical user interface display, micro details, Octane render, photorealism, photorealistic AR. This is just extra prompts that I got from the main channel, uh, from, from Midjourney Discord channel, and it made it look really nice and We'll see, we'll see how it, how it pops up. But it should show like a person, a man on the moon basically. And these other ones is like the aesthetics of it that you're gonna see on the side of it. Um, so what I've been doing as well is just kind of like copying like from here to, to right there and then copy. And then I just changed the front part here and it's going to generate something of what I want, but then all these other parts stay same. So basically it's going to generate a different person or a different thing, but then we'll keep the case almost similar. So see here what it generates. So it's still generating. I think it's done now. Yes, it's done. And you can see like it looks super cool because you can think of it like, like for example, this— I wouldn't even imagine like building something like that. All right, like it looks pretty cool. Um, so I'll go with upscaling number 1 and number 4. And let's see what comes off of it. Number 1 and number 4. Let's see what comes off of this one. Is the music loud enough? Just let me know. Or I put like some chill music in the background. I can change it to lo-fi, whatever. This app comes with like some Not so great music selection. If I was able to like do this by myself, I would probably find some other music. All right, let me change the music. Let me put some lo-fi music. Let's see what— how that sounds. All right, cool. So this one is still upscaling. It's taking a minute, but you can see like it, it does imagine and puts it. The case of knolling thing is, is basically the framing that you're seeing around here. That's the extra part of the prompt, but then whatever you you tell it to generate, it puts it within that. It embeds it in there. And for example, this looks really cool. I like it because it's like somebody walking out of the screen. Um, my other idea was thinking of like, if you can get something like this in real life and put it in your office desk or something, or in your house somewhere, like this would really— it would look really cool to have like an actual— probably look a little heavy. Um, but I think that would look really, really good. Oof, look at that one. That one looks pretty sick. Um, let me zoom in a little bit. Uh, and one thing I also realized with, um, if you tell it to write something or if it needs to write something like a Nike shoe, right? Maybe you tell it to generate a Nike shoe, for example. It will generate it and it will write some random words down there that kind of includes— sometimes it includes the words that you, that you mentioned a little bit. Maybe, but it would try to obfuscate that. And you can see here it does create some, some images. This one is walking through, somebody walking through that. Yes. I get my water bottle. All right, so that would be it. But, um, I think I'm gonna, I'm gonna stop the live stream from here today, and, um, we'll continue from here maybe some other time. But you can see, like, this was a simple example of what you could do You can generate way more than this. So you can feel free to copy my prompts and experiment with, uh, with on your part as well and see what you generate. All right guys, see y'all another time.
More from Tony

Summarize Videos and Research Companies with CrewAI + Gemini Models [Quickstart Guide]

How to Make Your AI Agents More Reliable with CrewAI Task Guardrails (Step-by-Step Tutorial)








