How I Created a Cartoon Series Using Artificial Intelligence

How I Created a Cartoon Series Using Artificial Intelligence

What does it actually take to create an animated cartoon using artificial intelligence?

In this episode of Papa D's Corner, I pull back the curtain and show you my real-world process for creating an episode of Me and You and Kalamazoo. From the earliest story ideas to a finished animated scene, you'll see how I use AI tools to create characters, generate artwork, animate video, create voices, produce lip-syncing dialogue, and assemble everything into a finished production.

This isn't a technical tutorial for programmers or professional animators. It's a practical look at how an ordinary grandfather from Mississippi learned to use artificial intelligence to tell stories, create cartoons for his grandchildren, and bring ideas to life.

Along the way, I share some of the lessons I've learned about creativity, storytelling, and why AI works best as a tool rather than a replacement for human imagination.

If you've ever wondered whether artificial intelligence can help you write a book, create a podcast, build a website, produce a video, or even create your own cartoon, this episode is for you.

You may be surprised by what's possible.

In this episode:
• Story development and brainstorming
• AI image generation
• AI video creation
• AI voice generation
• Lip-sync animation
• Video editing and production
• Lessons learned from using AI creatively
• Behind-the-scenes look at Me and You and Kalamazoo

Thank you for listening to Papa D's Corner, where we explore stories, ideas, technology, and the things that make life interesting.

 For more episodes, visit Papa D's Corner and subscribe wherever you get your podcasts. 

Transcript

SPEAKER_00 0:00

Today's show is for AI beginners and hobbyists who want to learn how to create a cartoon for their children without a lot of high-tech jargon. I'm going to show you how I do it so that you can too. Now I want to interject something here. This book was originally created for YouTube and I show, I demonstrate how I created these visually. And um so there may be a scene or two here that I refer to looking at the screen. Please forgive me for that. I'll try to not do that too much. But if you like, you ought to jump over to YouTube and watch this along with the audio. Now before we get started, you may be wondering why you'd want to watch me create a cartoon in the first place. Um over the past year, I've written books, produced a podcast, um, and created an animated series for my grandchildren called Me and You in Kalamazoo. What makes that unusual is that I'm not an animator and I don't have a team of artists working behind the scenes. For better or worse, I'm the writer, the producer, the editor, the voice director, the special effects department, and yes, sometimes even the janitor, and that's exactly why I wanted to make this video, because if an ordinary grandfather from Mississippi can learn how to use these tools to bring stories to life, maybe you can too. A year ago, if you told me I'd be creating animated cartoons for my grandchildren using artificial intelligence, I would have laughed. I'm not an animator, I'm not a programmer, and I certainly don't work for Disney. But today I'm going to show you exactly how I create my cartoons. Not the polished version, the real version. You're going to watch me go from an idea to a character to an animated video scene using the same AI tools that are available to ordinary people right now. If you're a parent, a grandparent, a teacher, a storyteller, or just somebody who's curious about artificial intelligence, I think you're going to find this fascinating. Now, this is not a detailed tutorial on every button, every setting, or every prompt. We can do that in future episodes. Instead, I'm going to show you my actual workflow, the process I use to create episodes of Me and You in Kalamazoo. By the end of this video, you'll see how I use artificial intelligence to create artwork, voices, animation, lip syncing, and finished cartoon scenes. And who knows, you may discover that creating your own stories is a lot more achievable than you thought. Let's get started. My assistants are ChatGPT, also known as Chatty Patty, Mila Groc and Claude. This show is for education and entertainment purposes. I'm here for the entertainment. They're here for the education. The very first thing that I do when I'm creating a new episode is to decide what the story is going to be, or at least come up with a general idea. Most of the time I start right here in Google Docs using the voice typing feature. The reason why I like doing this is because I can talk naturally, almost like I'm brainstorming with myself while taking notes at the same time. A lot of times I type, but this works better for me most of the time at this stage. Later I'll copy and paste these notes into Chat GPT and ask chat to organize my thoughts, create an outline, or rewrite something a little more clearly. But at this stage, talking to myself is just easier for typing. And Google Docs has the best voice uh typing tool that I know of, and it's free if you have Google anyway, and I think everybody does. Um, some of you may be wondering why I don't I don't just use Chat GPT or Grox voice mode to brainstorm directly with AI. Uh well the reason for that simple, right now I'm still exploring. I'm bouncing my ideas could deviate dramatically. I'm and I don't want to give my AI assistant a specific direction too early and then feel locked into it later. Uh you know, they have a tendency to remember stuff better than humans, uh, sometimes too good. And um, if I give them a piece of information and it sticks, they may keep inserting that idea that I've already gone past and changed my mind about. So I want to make sure that I formulate my ideas much better before I uh collaborate with AI too quickly. In this particular case, uh I already do know what my I want to name as my title, and that's not always the most of the time that's one of the last things you do, and AI can help you with that as well. One of the other things that AI can do for you after you've created your content, you've posted it on YouTube, is uh can help you uh study the analytics. YouTube provides data that shows how long people, average viewers watched, uh all sorts of valuable information. Too much for me to try to uh decipher. Well, what I learned is that two things. Number one, uh one of my most successful uh videos, which is very surprising, is uh the third or fourth one I made called When in Rome. The reason why it's surprising is my skills and technology uh were primitive back then, and so it's not the work that I would consider I'm the most proud of. And it's on a subject I didn't think kids would be interested in. The ancient Rome. I mean, uh kids don't know anything about that. But the reason why I picked that in the first place, my objective with this video, this cartoon series, is to not only entertain my child grandchildren and other children, but to uh educate them, not necessarily right now, but help them in the future. For example, grow up, get a little older and they're in school and they start learning about ancient Rome and Julius Caesar and Cleopatra and all that, they're far more likely to pay attention if they've heard of this before and have some type of interest and think, wait a minute, I remember having uh those characters appear in a cartoon series that I enjoyed as a child, and and they may pay more attention in school. The other thing I try to accomplish with this cartoon series is I try to teach subliminal, uh good moral values, family values, uh the importance of family, that type of thing. Um, so the other thing I learned with about this analytical study, too many of the children who are adults who would see the thumbnail on YouTube for my videos, they would see it and it would look interesting, so they would click on it. Too many of them were clicking away after 30 seconds or a minute, minute and a half, because uh that's where my introduction is. And they when they see a thumbnail and they see dinosaurs and volcanoes or monster trucks or race or spaceships. When they click on the video, they want to that's what they want to see. They don't expect to see uh Papa D reading a bedtime story to a couple of little children. However, that part of it is important to me and my grandchildren because that's this whole uh premise of this show is started with me reading bedtime stories to my children, and and these stories are basically bedtime stories and dreams and imagination based on that. So I I want to include that, but I need to have action at the very beginning, right away, long enough to basically get the children committed to sticking around for another minute to suffer through my introduction. We're gonna have action right from the start, and um, so based on that, the name of this episode is going to be Return to Rome: Mystery of the Missing Eagle. Why mystery? Because I used uh when I was a kid, Scooby-Doo was one of the most popular um television cartoon shows, and the reason why it was so popular, it was because there was a mystery. We always wanted to stick around and wait and see what the mystery was solved. Well, that's what we're gonna do here. We're gonna have a mystery that's going to be solved at the very end. And um, so based on that, I start with Google Docs bouncing ideas off this, and then the next thing I do is uh I want to um determine who my characters are. I have a little advantage because it's gonna be a remake. So I already know, for example, Princess Anna will be Cleopatra, and but we've got to get them into some uniforms. So that's the first thing I'm gonna do is I'm gonna show you how I create the characters uh with their clothing. But first, I'm gonna show you how I organize my workspace, and before that, right now I'm gonna show you what I do with all of this um dialog that I just typed into Google Docs. I will copy that and paste it into ChatGPT and ask Chat GPT to summarize and clear it up for a script. So what you've been reading or listening to is the result of that. So the first thing I do is I'll have my workspace set up this way. I'll have my two Google Chrome uh pages separated where I have Chat GPT on the left. That just works better for me, and I'll usually make that uh a little bit thinner because I don't need as much workspace with chat GPT as I will with some of the programs that I'll use over here, and on uh on the right side of the screen, I'll start with with Grok as my first tab, and it's usually on the imagine uh tool because this is where I'm gonna make my videos, and uh these will be the two that I use the most, and so I have it set up this way, but my second tab will be 11 labs, and I'm mostly doing uh text to speech here, and then I have Hajien that will I'll have set up, and this is where I have my avatars, and we'll get into that a lot more later. Uh, then I have open art set up as my other tab. Open art it uh is a uh is not really an AI by itself, it's a it's a door if you want to call it that. For example, if I wanted to do lip syncing, um it might default to cling, but it gives you all of these other choices. So that's what open art is. C dance, I could go over. There's a whole lot of different AI tools. Well, open art uses the the what they think is the best one for these uh variety of of situations, okay? So those are my main tools. A new tool that I haven't really uh not ready to endorse yet because I just discovered it last night and just started studying it. It's called sudo. But one thing I want you to remember is we're not going to, I'm not going to give you product demonstrations on any of these products. I'm literally literally going to illustrate to you the process that I go through and creating. We'll have another video or videos probably on each one of these individual uh products. So the first thing I do is I have to create a prompt. A prompt is really the term we use for giving a set of instructions. We're prompting AI to do something. I want AI to create an image. And so this is a prompt. If you'll notice the everything in bold, that's kind of a standard language that I use on all prompts. Uh now, down at the bottom where it's bold, it says no one else is in the scene, make the background transparent, make no other change. Those are interchangeable with others, but the reason why I put them in bold is the uh area in the middle is going to be unique to this particular image. And this is just my technique. The reason why I do it this way is I've learned that if you include the specific uh language that I have, for example, I tell them what aspect ratio, I tell them what style I like the Pixar style lighting, and and I put all that information in there, I get better results. Uh, then what I do is um I've got a uh an image of Princess Anna, she's my recurring character, and I want her to look like um Cleopatra in this one. So I've given the instructions and I always start out with a transparent background and have for that for each character to begin with. That way, if I need to create different scenes and so forth, I can have that one reference image that I can use to make sure my characters are consistent without confusing AI with the background, and so I'll do that for each one of my characters, but I'll show you, let's take a look at how the process works. So I start by going over to Chat GPT and I tell ChatGPT that I want to create an image, then I want to give it the reference images. These are the images that I want it to refer to to in order to create my new image. So I start by uploading an image of Princess Anna because that's my standard character, and I want to make sure that uh I'm consistent episode after episode. I want any you know, Princess Anna and the voice that goes with her, the at least the face to always look the same. And so I want Princess Anna to be dressed like and have the same hairstyle and jewelry as you see in this Cleopatra, but I want her uh face, eyes, skin tone, all that to look the same as Princess Anna. So I paste those instructions uh over to and then I hit the send button and chat GPT goes to work. Now, while that is creating, I'm gonna do something I don't normally do uh at this stage, but I'm doing this for you. I'm gonna go over to 11 Labs and I'm gonna give um uh uh a line for Princess or new Cleopatra to say to you in Princess Anna's voice, and then I'm gonna create a little short video, and um that'll show you the entire process from creating an image, giving it a voice, lip syncing, and creating a video all at one time. So after her voice has been generated in Leaven Labs, I download that. The image has been rendered in Chat GPT, I'll download that. Next, I'm going to write a prompt to give to Chat GPT to create a prompt. That may be unnecessary, that might actually be um an extra step, but I have found it works really well for me. ChatGPT writes better prompts than I do, and I know how to write prompts to Chat GPT that's successful. So I've got a standard prompt that I use, and you can see it here on the screen that I'm gonna give a chat GPT to tell Grok to create a video of Cleopatra standing in ancient Rome. You can read the the uh prompt right here, and then after that, I'll load that into Movavy and I will play that, and then I will uh screenshot or uh the the end image, the the last scene of that, then I'll use that image into uh open art to create a lip syncing video, and then I'll load that back into Mov IV and play it for you so you can see the whole process. I've copied and pasted the prompt that chat GPT ChatGPT took my instructions, my prompt, turned it into a better prompt for me to create a video. I can do an image or a video using Grot. I can uh make it a six-second video, or I can make it a 10-second video. I'm gonna do a 10-second video. So we're gonna hit go button and we're gonna let that happen. And when this finished after this is finished generating, I'll download that and put it into Movivi. And for the sake of time, I'm gonna speed this up a little bit. When Grok finishes, I'll download the video and it goes into my computer on download, and then I'll go into Movav and I'll reach down into my downloads on my hard drive and pick up that video and pull it into Movavi for editing. And once it's in there, I'll insert it and I'll play it. I'm gonna snip off the very front end. Movavi will always begin the video with whatever your reference video was, and my video had a green screen background, so I wanted to clip that off. That doesn't look good. So I'm gonna play it, and when I get to the last frame, I'm going to do what we call a snapshot. So I'm gonna pause it right here. I'm gonna go over to my Movavi and do a snapshot. That's basically a screenshot, I guess you could say, of whatever you see on the screen. And now I've just created another image of Cleopatra standing by the fountain. Now, this is a steel image, so I'm going to upload that steel image into uh open art. Now I've already created the the uh line in 11 labs, but I wanted to show you that step in case you forgot. So now I'm gonna upload the the image into open art, and I'm going to go to downloads and get my audio that I created from 11 Labs. This is using Princess Anna's voice, which was already pre-created. And then I'm gonna hit create, and that'll take a minute. And while that's creating, oh by the way, I'm gonna add a little prompt. This is something you can do in open art that you can't do in HeyGen, which reminds me of why I'm using OpenArt. Hey Gen is a better option for a lot of lip syncing if you're doing a podcast or if if you're gonna reuse the same exact image over and over and over again, same style, and there's no movement. But if you're going to have just a one-off like this, open art's a lot faster and a lot better. So let's go see what the end result is. Now that open art is finished rendering the lip syncing video, I always expand open art, it's easier to see. So I'm gonna download that lip syncing video from open art. Then I'm gonna go back to Mov IV and I'm going to upload that lip syncing video that open art just created, and then I'm going to uh add that to my timeline, and of course, I'm gonna put the original video in here as well and in front of it, and then I'm going to play this and show you what the end result was. I could add some music to this first part, but I don't think I'm gonna do that. I think you'll get the idea. So let's see what the total final result is. Hello everyone, thanks for bringing me Cleopatra back to life. Next, let's talk about video editing. This is where everything finally comes together. Up until now, we've been creating individual pieces. We've written the story, generated images, created video clips, recorded voices, and produced lip sync animations. But none of those things are a cartoon by themselves. The magic happens in the editing room. For me, that's usually Movave video editor. This is where I arrange all of the video clips into the proper order, trim scenes that are too long, edit narration, insert sound effects, create titles, and make sure the story flows the way I intended. In many ways, editing is where the real storytelling happens. By the time I reach this stage, I will have dozens of separate AI-generated files, images, videos, voices, sound effects, music, and graphics all have to be assembled into a single production. The editing process is where all of those individual pieces stop being separate assets and become a finished story. And once that's done, the cartoon is finally ready for its audience. But don't let that scare you. If you have patience, you can do this. If you don't have patience, don't even begin to try. And once that's done, the cartoon is finally ready for its audience. Before we wrap up, I want to share a few lessons I've learned along the way. One of the biggest misconceptions people have about artificial intelligence is that you simply push a button and it does all the work for you. At least in my experience, that's not how it works. AI is an incredibly powerful tool, but it rewards specificity. The more clearly I can explain what I want, the better the results usually are. In fact, one of the most valuable skills I've developed over the past year has very little to do with technology. It's learning how to communicate my ideas clearly. The AI doesn't know what's in my head. It doesn't know what kind of story I want to tell. It doesn't know what my characters look like, what message I'm trying to convey, or what emotion I'm trying to create. That's still my job. The human being still provides the idea, the creativity, the judgment, the storytelling. AI can help me create images, voices, video clips, and special effects faster than ever before, but it can't replace imagination. It can't replace life experience. And it certainly can't replace the desire to create something meaningful. I like to think of AI as a toolbox, a very powerful toolbox, but just like a hammer doesn't build a house by itself. AI doesn't create great stories by itself. The tools are important. But it's still the person using those tools that makes the difference. Five years ago, I never would have imagined that I'd be creating cartoons for my grandchildren, producing documentaries, writing books, recording podcasts, building websites, and creating educational content with the help of artificial intelligence. Yet here we are, the technology is changing rapidly and the opportunities seem almost endless. And if you're willing to learn these tools, experiment, make mistakes, and keep improving, there's no telling what you might be able to create. Thank you for spending some time with me today. If you enjoy this episode, I'd appreciate it. If you'd subscribe to the channel, leave a comment, and let me know what you'd like me to cover in future episodes of Papa Dee's Corner. Until next time, keep learning, keep creating, and keep sharing stories worth telling.