How to publish a children's book in a weekend using AI with Ammaar Reshi
Ammaar Reshi shares how he wrote, illustrated and published a children's book using Generative AI in weekend using ChatGPT and Midjourney
Hey Readers,
In today’s post, I want to share one of the first episodes of a new podcast I’m working on. In the podcast I will be exploring the world of AI to understand how it’s going to change the way we live, work and play, by interviewing creators, builders and researchers. In this episode, I interview Ammaar Reshi, a designer who recently wrote, illustrated and published a children’s book using AI! I highlighted Ammaar’s in my first post1 a few weeks ago as a great example of how AI is making creativity more accessible to everyone.
In the interview, Ammaar shares what inspired him to use AI to write a children’s book, the backlash he received from the online artist community and his perspective on how AI will impact art in the future. If you’re new to AI and haven’t yet tried using Generative AI tools like ChatGPT or MidJourney, this is a great video to watch because Ammaar also shows us step-by-step how he created his children’s book. This is a must-watch for parents, educators or budding authors who might want to make their own children’s book too!
To get the most out of this episode, I recommend you watch the video so you can see how all the AI tools we cover work > Youtube Video
I hope you enjoy this episode. I’ll be officially launching the podcast in a few weeks, so it will be available on your favorite podcast player soon. In the meantime, I’ll be sharing more episodes here as I record them and I would love your feedback in the comments!
Show Notes
Links from the episode
Ammaar’s Twitter post on how he created a children’s book in a weekend:
Ammaar’s book “Alice and Sparkles”: https://www.amazon.sg/Alice-Sparkle-exciting-childrens-technology/dp/B0BNV5KMD8
Ammaar’s Batman video:
- ChatGPT for story writing: http://chat.openai.com
- MidJourney for illustrations: Midjourney.com
- Discord for using MidJourney: https://discord.com
- PixelMator for upscaling your illustrations: https://www.pixelmator.com/pro/
- Apple Pages for laying out your book: https://www.apple.com/pages/
- Amazon Kindle Direct Publishing for publishing your book: https://kdp.amazon.com/en_US/
Episode Contents:
- (00:00) Introduction
- (01:55) Ammaar’s story
- (05:25) Backlash from artists
- (12:20) From AI books to AI videos
- (16:20) The steps to creating a book with AI
- (18:55) Using ChatGPT to write an children’s story
- (23:45) Describing illustrations with ChatGPT
- (26:00) Illustrating with MidJourney
- (35:30) Improving prompts in Midjourney
- (37:20) Midjourney Pricing
- (40:00) Downloading image from MidJourney
- (44:20) Upscaling with Pixelmator
- (49:25) Laying out book with Apple Pages
- (53:40) Publishing on Amazon KDP
- (55:35) Ammaar shows us his hardcover book
- (56:25) Wrap-up
Full Transcript
[00:00:00]
Introduction
ammaar: I think it has to start with your idea of a story, right? I think, you know, people might think, okay, you press a button, it spits out a book, but I think it has to start with your imagination. And then we will provide that to ChatGPT to kind of give us a base for our story. I think then we'll iterate with ChatGPT almost like a brainstorming partner. We're gonna go back and forth. We're gonna expand on characters and the arcs that we might want to, you know, go through. And I think once we have that, Then we go back to imagining again.
We have to think through how do you take that script and that story and you bring it to life, how do you visualize it? And that's where MidJourney comes in. And we're gonna generate art that fits that narrative and expresses that narrative in a really nice way. And then we can combine it all together with you know, pages to create that book format.
aj_asver: Hey everyone, and welcome to the Hitchhikers Guide to ai. I'm your tour guide AJ Aser, and in this podcast I explore the world of artificial intelligence to learn how AI will impact the way we live, work, [00:01:00] and play. Now, if you're a parent like me in the middle of reading a book to your kids, this thought may have crossed your mind.
Hey, I think I could write one of these, but the idea of writing and publishing a children's book to many. Is a distant fantasy that is until now, because generative AI is making it easier than ever for anyone to become an author or illustrator.
Just like today's guest, Ammaar Reshi who wrote illustrated and published a children's book for his friend's children in one weekend using the latest AI products including ChatGPT and MidJourney in this episode. Ammaar's going to show us exactly how he did it.
So Hitch A Ride with me is we explore the exciting world of generative AI in this episode of the Hitchhiker Guide to ai.
Ammaar's Story
aj_asver: Hi Ammaar it's great to have you on the podcast.
ammaar: Hi. Great to be here, AJ. How are you doing?
aj_asver: I'm great. I'm so excited for you [00:02:00] to join me on this podcast especially to be one of the first people I'm interviewing. It's gonna be a learning experience for me, and we'll work it out together.
But anyway, I'm so excited to have you here. I talked about you in my newsletter in the first post because I thought what you did by publishing your book, Alison Sparkles, was such a great example of how AI is gonna change the world around us and really make creativity and being a creator a much more achievable and approachable thing for most people.
Since your book was published on Amazon in December of last year, you have sold, what is it, over 900 copies. Is that right?
ammaar: Yeah. It's about 1,200 now, so yeah. Crazy
aj_asver: That is amazing!
ammaar: Yeah, it's been wild
aj_asver: That is so cool. And at the same time, you've found yourself at the center of a growing debate about AI and the future of art. So tell us how it all happened.
Rewind us back to the start. What made you a designer at a tech company decide to publish a children's book?
ammaar: I guess what kicked it off to go all the way back to two of [00:03:00] my best friends basically had their first kids. And I went to visit one of them. She had turned one years old. And I went over and it was around her bedtime where she grabbed my hand and took me upstairs. And I was like, what's going on? And they're like, she's picked you tonight to read her bedtime story . So I was like, wow. I was like, I am honored.
This is one step closer to being that cool uncle. She hands me this book I was something about getting all these animals delivered from the zoo. And so I was, I mean my friend was there and I was reading her this book and we were both laughing cuz we were like, this book makes no sense at all.
This story is so random. But she loved it. You know, she loved it. She loved the art, she loved everything about it. And it then kind of hit me in that moment. I was like, it'd be really fun to tell her story of my own, you know? I just had no idea how I was gonna go and do that yet.
I told my friend, I was like, I think the next time I come over, there's gonna be a book on her shelf. It's gonna be mine. And he was like how are you gonna do that? I was like, well, gimme a weekend. I'll figure it out, right..
I had already been playing with MidJourney like sometime in February [00:04:00] of uh, last year.
And so, You know, I, I knew generative AI and the artwork stuff was there and it was really cool. And Dall-E also had blown up around then. And so yeah I knew okay, if I wanted to illustrate this book, I could lean on MidJourney to help me with some of the creation, but I hadn't yet come across ChatGPT. And a friend of mine actually, just that moment, like that Friday message meeting, he's have you seen ChatGPT? I've been playing with it all weekend, all week. I've even created music and chords and like chord progressions and stuff with it. And I was like, that's crazy. Like I wonder if it could help me craft a story, like a children's story.
And then I wanted the story to be something a little meta, you know, something about. But also a little personal. I remember as a kid, like my dad let me play with his computer when I was like four or five years old, right? And I, that kind of led me down the path of going into tech and like all of that and that curiosity.
And so I basically wanted to mash those two. It was [00:05:00] this young girl who's curious about technology, and specifically about ai and then ends up making her own. And that's essentially the prompt that I gave ChatGPT and that's what set off, the path into making this book.
aj_asver: That is such a cool story. I think the bit you talked about where you're like reading this book with a child and you're lucky because as an uncle you don't have to do it a hundred times.
ammaar: That's what he said. Yeah.
Backlack from artists
aj_asver: Yeah. It can get pretty tiresome and you're like, and you do wonder often wow. This book doesn't make a lot of sense, but like hundreds or thousands of copies have been sold.
It's really cool that you took the initiative to do that. Now, you were in the Washington Post recently in an article titled he made a children's book using AI, then came the Rage. Talk us about that. Why did you make so many people angry with the children's book?
ammaar: Yeah, that was also dramatically unexpected. So when I first created the book on that weekend, my goal was like, I need a paperback in hand as soon as possible. And Amazon, KDP, which [00:06:00] I think is the most underrated part of this whole discussion, where it's like there's a platform out there that can get you a paperback within a week , just upload a PDF.
No one talks about that. I used KDP, got the book and initially, just gave it to my friends. That's, that was the goal. And it was out there. And then, I was like, this would actually be really fun to share with other friends as well. And so I put it on my Instagram and have a good amount of friends who are not in tech, and so they replied and I put in my story. If you want a copy, I'm gonna gift it to you. Let me know. And so a lot of folks were like, oh, this is so cool. I want a copy. I want a copy. And everyone's reacting in a super positive way. And so this is forming my initial um, You know, opinion of how the story is going to be perceived.
It's going to be perceived in a way where everyone thinks this is very cool concept, right? Like the same way I thought it was. But boy was I wrong.
Then after the Instagram story, and I think this was the most Instagram messages I've received in a while in one go. A friend of mine was like, you gotta put this on Twitter. I think everyone should see this. And I was like, okay, yeah, [00:07:00] why not? So I quickly drafted up a tweet and put it out there first 24 hours. Tons of love and praise. You know, I had parents reaching out saying, I wanna make a book with my child over the weekend. Can you walk me through it?
Then I vividly remember it was like 4:00 AM the next day. And my phone is just buzzing like it is going. And I'm getting all these messages and it's you're scum. Like we hate you, like you're stole from us. I was just like, whoa. What is happening? And at the time, I hadn't read up on what had happened with Lensa or any of these other AI tools that were allegedly being trained on copyright material and and art without consent.
And I had not, I was just not aware of any of this. I was digging into it and I was like, why am I getting all this hate all of a sudden? And I realized that I'd been retweeted by some artists who had very large followings. We're talking about 40, 50,000 people. And I think then Tim Ferriss liked the tweet and I just blew it up to another yeah, I think he has like a million followers who, another, huge set of folks.
And then it was spreading across like illustrators, [00:08:00] writers, artists, and. and I know why they have these concerns, right? It's like you're seeing a piece of technology that is doing what you do on a day-to-day basis without you involved at any single part of that creation of that cycle, right?
I think AI, we've been talking about it for forever, right? Especially if you've been in the tech field for a long time. , but it's felt abstract. It's not tangible. And now you finally have this physical object that blends in with everything else that's out there, a children's book. And if I didn't say it was created by ai, I think most people would've just said, oh, it's just another random children's book. And so I think that provoked the discussion and created that fear that, wow, this is not. AI is good. It's so damn good that you could publish a book and it could be out there and no one would know, if you didn't say so.
And I think that struck a chord and that led to the reaction that I got. Death threats, everything in between. It was quite, quite the storm. I found myself caught in right after that.
aj_asver: Going [00:09:00] through that experience you saw this other side of ai. , as people in tech that have been using ai, probably with this optimism and excitement, you don't really realize how folks that aren't as familiar with it or folks that really find it threatening to their livelihood see how it might impact them.
One thing I'm curious about is did you experience or hear from any artists that thought about it in a positive way or excited about AI and how it could help them?
ammaar: I did get a few dms and there were DMs. They literally said, I don't wanna put this out there. They started the DM with that. So I think there was already this fear of backlash based on how the broader community was reacting to this. But they had said that They find this really exciting because of the way it could speed up their workflows.
They saw it as an opportunity to brainstorm faster, come to concepts much faster than they could, and then use, their skill to hand draw and expand on that. And I think that's really promising too. It's that view that this is supplemental and not a replacement and actually can fit into your existing workflows.
I think that's really exciting. . And I think the problem is like the rhetoric [00:10:00] around all of this is stuck between it's being conflated, right? It's one on the one hand, this is a very exciting progression in technology because you're enabling a new set of creators. People who, could imagine all of these things, but couldn't dream of creating them because they were limited by their skills.
Maybe they couldn't pay for the courses or anything like that, right? And now your imagination's limit and the technology is just empowering you to create, which I think is amazing, right? We're gonna get a whole new set of people. I'm sure there are young Spielbergs out there that now can create things they couldn't be for.
Very awesome. But it's being conflated with the other, A aspect of this, which is it is being trained on copyright material, it is being trained on work that's out there without people's consent. And I. , you do have to address that. But at the same time, you can still be excited about where this is going.
And unfortunately, it's become this blob of an argument where it's just bad period. And I think that's what people need to unpack because there are a whole set of artists out there who also think this is very cool and would just like it to be trained in a more ethical way. Which yeah, I would be So for that, I'm not [00:11:00] anti artist in any way.
It's Let's enable the artist to create, cuz again, their imaginations allowed this tool to exist and create some things that we couldn't even imagine. So imagine what they could do combined with this tool. I'm sure we'd see stuff we haven't even imagined yet which is even more awesome.
aj_asver: Yeah. A lot of people make that argument in the same way that you know, that I think this is like a classic business argument of the early part of the 20th century where the factories came and people thought it was gonna replace jobs, but it actually created more opportunity and more growth in the economy, but it's hard to see that viscerally when you see it impacting your work.
One of the interesting things about AI is there's been this inability to predict what will be easy and what will be hard for ai. So for example, you and I can learn to drive a car maybe in about 10 hours, and that actually is really hard for AI. But for me to become an illustrator or a master chess player as a different example, would probably take many years, right? I couldn't imagine myself becoming an illustrator. If I pick up a pen or, or a, a paint and try painting, it would take [00:12:00] me a really long time. But AI has learned to unexpectedly quickly, which is really interesting. And you went through this experience now after you went through this and experienced that, that kind of wave of first people being really excited and then people really being mad at you how did you feel after that? Did it stop you from wanting to go out there and make more stuff like this? Or did you keep going?
From books to videos
ammaar: it Was interesting because in the beginning it kind of forced some introspection, some reflection on like, why did it strike such a chord? What could I have done differently? I think realizing and seeing that, okay, this is copyright material uh, in these companies as well of really kind of dodging this question, right?
I think I watched an NBC nightly short where David Holz from MidJourney is asked Hey, you're caught in this copyright arguing. He's like, I don't wanna talk about that basically. I think they, these things aren't being addressed and. I was like, okay, I still really enjoy creating with this. I think it's allowed me, like I said, I felt empowered when I was using these tools and I'm sure a lot of other folks do as well. But I was, I was not going to continue to monetize with the art I think. [00:13:00] The book itself was just an experiment. And then, and again the only way to get that paper back really quickly was to put it on Amazon.
I didn't take the book down, which is another counter-argument you could give to me. Okay, if you don't wanna monetize, why is the book still up there? well, I, we left the book up there because I think it, it sparked an interesting discussion. And that discussion is now at a point where, we are having this, we are having this conversation. The Washington Post has elevated this conversation and I think the pessimistic point of view is oh, this is a quick cash grab or whatever. But I think the somewhat optimistic point of view the realist in you says okay, this discussion has now gotten so much light that we might make more progress on our concerns.
I think if you look at it that way the pragmatic point of view I think. . I think that's why the book is still up there. It's, we are now talking about this all over the place and it'll might actually lead to progress with, with getting these companies to maybe do things in a more responsible way.
So that's where I like, did not continue on the monetization path, but that didn't stop me on the creation path because I think it's still really cool to see how these tools can [00:14:00] bring your ideas to life. And a book was just one of the ideas that struck me that. But then as I was browsing Twitter one other weekend , I saw this like very short animated clip where someone had just animated this person, walk into a tavern, seen this mysterious and I've got sucked into this very short clip.
I was like, this is so intriguing. And. and that was before I realized that it was actually made using immuno generative ai. And I was like, what? Like how did they do this? There was no source to who made this video. And so I had to just spend like a few hours just de decomp and thinking about how they did this in reverse engineering the process in my own head.
But it got me so excited again because film is one of my. Favorite like things, it's like a hobby of mine to watch and log all the movies I watch. If I didn't study computer science, it definitely would've been film. And so to then see an avenue where I could tell my own story and animate it and create it was super exciting.
And yeah. And then one weekend I was like, you know [00:15:00] what? I'm gonna do this . I'm going to try to create my own short. And I do it with one of my favorite superheroes, Batman, which I think has like such a great audience as well. Like lots of people love the superhero. So it. To tell my own Batman story, and that's when I decided to make a little Anma animated short.
And again, tweeted about it this time, racing for impact a little bit given previous reactions. And within a few days it got 7 million views. So it was insane to see how it, resonated with folks again, but also of course struck a chord again.
aj_asver: 7 million views.
ammaar: Seven
That's amazing.
Okay. Ammar. You wrote illustrated and published a children's book and put it online for the world to buy in a weekend.
Yeah.
aj_asver: many people may not believe that's even possible. And in fact, many people that are listening to this podcast may not have tried Chat G P T or MidJourney or any of the other cool generative AI tools that you and I have tried.
And [00:16:00] so I thought it would be a really fun exercise for you to teach me exactly how you did it and for us to make a children's book.
Are you gay?
ammaar: let's go. Let's do it
How to write a book using AI
aj_asver: Let's start at a high level. Tell me the three to five steps that you went through to publish Alison Sparkles a weekend.
ammaar: Yeah. I think it has to start with your idea of a story, right? I think, you know, people might think, okay, you press a button, it spits out a book, but I think it has to start with your imagination. So I'm sure AJ, you've got a fun story we can craft together, but you know, it'll start with what you think that is.
And then we will provide that to ChatGPT to kind of give us a base for our story. I think then we'll iterate with ChatGPT, almost like a brainstorming partner. We're gonna go back and forth. We're gonna expand on characters and the arcs that we might want to, you know, go through. And I think once we have that, Then we go back to [00:17:00] imagining again.
We have to think through how do you take that script and that story and you bring it to life, how do you visualize it? And that's where MidJourney comes in. And we're gonna generate art that fits that narrative and expresses that narrative in a really nice way. And then we can combine it all together with you know, pages to create that book format.
aj_asver: Okay, so to recap, we've gotta come up with a story using our imagination. That's probably the one area where I might be able to help. Then we're gonna
ammaar: let's go. Yes.
aj_asver: we're gonna ask ChatGPT to help us turn it into a story, for a book. Then we're gonna. Iterate on it a little bit and develop the story. Then we're gonna take that and work out how to turn it into illustrations. Also think about the character arc and the different characters involved and how we describe them.
And then we're gonna try and put it all together into a book which we're gonna publish. Is that correct?
ammaar: Sounds good. Yeah.
aj_asver: All right. So we need a story now.
I've got this idea for a book that I wanna publish. It may have been inspired by a book I read [00:18:00] recently. It's a great fairytale. It has a prince, he meets a princess, they fall in love. There's some tragedy, but it has a happy ending. so the character in the book is Prince Harold. he's young, he loses his mother
ammaar: Oh, man.
aj_asver: And the only way for his heart to be mended is for him to meet someone special. And he
this actress named Megs and they fall in love and their love eventually men's his heart, he live happily ever after.
ammaar: That's a great story. Yeah. I th I think we can make something off of this for sure. Let's do it.
aj_asver: The first step you said was, we're gonna take this story and we're gonna put it in ChatGPT. So let's walk through that and I'd love you to show me how you use chat GB t
The URL is chat.openai.com will include the link in the notes for the podcast too. And you need to sign up, I think you can use a Google account on Microsoft account or use a, your email address. But ChatGPT is completely free so anyone can try this just like Ammaar's about to show us.
Using ChatGTP to write story
ammaar: All right let's work off of that original story [00:19:00] And you know, I like to start with something that just gives us a little bit of a base to work off of. You know, we could do, write a short story about a prince, whoops, can't spell, prince, called Harold, who loses his mother at a young age.
And the reason I'm giving it that is because I want it to introduce some tragedy into the short story, and let's see if it can kind of work off of that. He eventually meets the love of his life, an actress named Max.
Okay. So let's see what it's creating for us. So once upon a time, there was a young prince named Harold. He lived a life of luxury. Wow. it was a life that was missing something very important, the love and care of his mother. Wow. This is this is really building up the way you were describing it.
She passed away when he was just a boy, leaving him with a broken heart and a longing for someone to fill the void. Wow. [00:20:00] As he grew up, Harold became known for his adventurous spirit and charming personality. you say that lines up with how you imagined Harold? AJ
aj_asver: just like the Harold I was
imagining. It's like ChatGPT knew exactly what I was thinking about.
ammaar: Yeah. Amazing. Let's continue. He traveled the world, met new people and tried new things, but he was still searching for that. Someone special. One day he went to a play and met actress named Megs. She was the star of the show, and as soon as Harold laid eyes on her, he knew she was the one he had been searching for. Damn You can already uh, sparks.
Yeah. Megs was suggested intended by Harold, and soon they were inseparable. They went on adventures together, laughing and falling more in love every day. Harold finally found the love and comfort he had been missing for so long, and he knew that Megs was the missing piece to complete his life. This is quite the fairytale.
Wow.
The two of them got married in a grand ceremony with friends and family from [00:21:00] far and wide coming to celebrate their. Harold and Megs lived happily ever after. And their love continued to grow stronger with each passing day. the end, Harold realized that even though he had lost his mother, he had gained so much more in Megs. She was his rock, his partner in life, and the love of his heart. And together they built a life full of happiness, venture and endless love.
aj_asver: Wow. Chachi b t essentially wrote this story based off a few prompts that you gave it.
So you said write a
ammaar: one prompt.
aj_asver: right? One prompt, which is write a short story about a prince called Harold, who loses his mother at a young age. He eventually meets the love of his life, an actress named Max, ChatGPT, has this whole story. And in it, you know, Harold becomes known for his adventures.
He has a broken heart, but then he meets this actress at a play, which is really cool. And the two of them get married in a grand ceremony and they live happily ever after. And all of that ChatGPT came up with. So we've kind of already got parts of the story, so what's the next step? How do we turn this into a book?
ammaar: Yeah. [00:22:00] So the next step is gonna be taking that story and putting that I just like to work in Apple Notes, but putting that story in there and then thinking about what scenes we're imagining for that part of the story. So let's brainstorm a little bit, AJ, and figure that out, and then we can go to MidJourney and start to illustrate that stuff.
What do you think?
aj_asver: That sounds good. Let's try it.
ammaar: All right, so I've just copied the story. I'm just gonna paste it in.
Okay, so we've got a beginning. A once upon a Time there was a young prince named Harold.
What do you think we could do here? I think it could be maybe a young boy in a kingdom, you know, to show that Prince growing up
aj_asver: think that could be cool. Let's think about what it might look like. I'm thinking like maybe a redhead be kind of cool
ammaar: yeah. You're really original with this. Yeah.
aj_asver: I'm thinking the kingdom would have a lot of castles,
From the medieval times. That might be kind of cool.
of rolling green hills.
ammaar: I love that. very scenic. Yeah.
aj_asver: we could add some soldiers wearing red [00:23:00] tunics and really tall black hats made out of fur. I'm
you know,
ammaar: I've just never seen anything like that. That sounds so incredible.
aj_asver: We're kind of coming together with a story right now and you are writing it out and you are doing it in Apple, in your Apple notebook, but
this like in a Google Doc or
you could write out really anywhere. And what you're doing is you're essentially taking a story that ChatGPT created and just kind of describing the different scenes.
ammaar: where, right?
aj_asver: exercise to do and we're using a lot of our imagination. But I was curious, could chat g p d do this too? Like it, could it describe some of the scenes?
ammaar: Sure. Yeah. Why not? Let's just switch back to ChatGPT and pick a paragraph and then tell it to describe what that scene might look like and let's see what it gives us.
aj_asver: That would be really cool.
Describing illustrations with ChatGPT
ammaar: All right. So we're back in ChatGPT, we're still in the middle of our conversation with it. And let's, you know, let's just take, let's take that first paragraph and say, describe how you would illustrate [00:24:00] this scene. And I'm just gonna paste in the thing and let's see what it does
Now there we go. This is a really great description of what it described.
So we started with the first paragraph, which just to remind everyone was, once upon a time, there was a young prince named Harold.
He lived a life of luxury, but it was a life that was missing something very important, the love and care of his mother. She passed away when he was just a boy, leaving him with a broken heart and the longing for someone to fill the void. We asked basically to describe how it would illustrate the scene.
So here's what we got in response. To illustrate the scene of Prince Harold's life, I would depict a young prince standing in a grand castle surrounded by riches and luxury, but with a lonely and sad expression on his face. The room would be spacious and well appointed, but with a feeling of emptiness.
Wow. In the background, there would be a portrait of Harold's mother with a veil of sadness covering her face, symbolizing the [00:25:00] loss of her love and care for her son. The window of the room would show a view of a gray and gloomy sky reflecting the prince's heartache and longing for someone to fill the void left by his mother's passing.
This is very dramatic and emotional. To emphasize the Prince's loneliness, I would have him standing alone with no one in sight and his head bowed and sadness. The overall mood of the illustration would be somber with muted colors and soft light, but with a hint of hope in the prince's eyes hinting at the bright future that lies ahead for him.
aj_asver: So that was really cool because when we were talking about it, we kind of had a few words or
ammaar: Yeah,
aj_asver: it. But when we went back to ChatGPT and actually just said, Hey, ChatGPT I want you to illustrate, how you would describe this part of the story. It actually went into a lot of detail to do that.
And what it means is for someone that isn't necessarily good at imagining this and really describing it to illustrate it, you can actually shortcut that and let ChatGPT do some of that work for you, which is really cool. So we've got an idea, a scene now for the [00:26:00] first part of the book. How do we turn this into an illustration?
Illustrating with MidJourney
ammaar: So this is where MidJourney comes into play.
aj_asver: so what is MidJourney Ammaar?
ammaar: Yeah. So MidJourney is the generative AI tool we're gonna use where chat G B T is kind of giving us text descriptions of things.
MidJourney allows us to visualize that. So with a prompt, the same way we gave it to ChatGPT, it's gonna create images for us that we could start using and, you know, a book. So let's dive in.
the first step to using MidJourney is heading over to MidJourney.com. And you know, you come across this very trippy homepage very eighties hacker esque. And honestly that's probably reflective of what it feels like using MidJourney.
It's not the most approachable but it's not that hard to get started. So what you wanna do is click Join the Beta, and once you click that you're going to be guided to Discord. And so this is where you need a Discord account to join essentially the MidJourney channel. It's like joining a Slack channel, basically where you can then talk to the MidJourney [00:27:00]bot and provide at the prompts.
And so for those who are not familiar with what Discord is, it's basically it's basically Slack for communities. So if you've ever used Microsoft Teams or Slack or any of these tools that allow you to communicate with people, discord is aimed at those niche communities of people who just are hobbyists in different fields.
Could be movies, you know, cartoons, games really started with video games actually. And and now you know, it's got AI enthusiasts. So this just a channel a home for anyone who's really interested in generating art to join this channel and have conversations and share their creations with each other.
Okay, so in discord. So this is Discord. This is what it looks like. You know, you've got essentially. seems like a chat app, like an instant messaging app.
If you've ever played with MSN Messenger back in the day, or Yahoo Messenger, this is just like that. Or AIM I think in the States. But what we're going to do, I said that, you know, MidJourney had its own channel where you can see the community and here's just a preview of that. It, you know, you can see someone's creations, looks like they've made some [00:28:00] incredible art about, you know, with Spider-Man and The Hulk.
Beautiful style. But what we wanna do is we want to communicate with the MidJourney Bot. And the MidJourney bot is essentially this automated, you know, chatbot just like ChatGPT that lives inside of Discord. And we're going to tell it what we want to see and what we want to imagine, and it's going to end up creating images for us.
So let's jump over to the MidJourney bot and start generating some images.
All right, so we are in MidJourney and we are about to bring Prince Harold to life, . Let's see what we had. So AJ, your description was a young boy in the kingdom redhead with freckles has a lot of castles, scenic soldiers. All right, so I'm just gonna copy that over. And let's work with, let's work with this.
To use MidJourney, the first thing you wanna do is you type slash Imagine and so slash imagine allows you to start entering your prompt and creating images and artwork. [00:29:00] And so I copied that description that we had written in Apple Notes, and I'm just pasting it in MidJourney, and I'm saying young boy in a kingdom redheaded with freckles.
Kingdom has a lot of castles, scenic soldiers with large black hats. And I'm actually just gonna add a few more adjectives to this. Maybe I'm gonna say a prince actually in a kingdom, a young prince. I wanted to create and generate some of that royalty and that feeling of royalty in this image, right?
And I'm also gonna say a children's book illustration to make it, you know, approachable and approachable art style, children's book illustration.
aj_asver: That bit is really interesting that you just talked about the art style, because I've seen so many different styles of art come out in MidJourney and it seems like a big part of successfully using a product like MidJourney is really being able to describe the kind of style of art you want. You know, it does everything from like photorealistic art that looks like a photograph all the way through to, you know, comic style art that might look like, you know, a graphic novel [00:30:00]or a comic. And then, you know, in between. Like I'm sure we could get it to generate an illustration that looks like something from the Pixar universe, for example. That's really cool. And then when you put it in and you say Children's book illustration, is that enough or do you try to add more?
What's kind of some of the tips you have for making a really good prompt in MidJourney?
ammaar: Yeah it's really interesting you asked that because I think the more specific you can get, the better your results are. And sometimes you really need to, you know, I think this prompt, if I was gonna critique it, I would actually say it's probably got too much going. You know, it's going to try to get that children's book style.
It's going to try to bring a young prince in. We're giving descriptions to the prince. We're describing the kingdom around them. So there's a lot going on here, and I think we're gonna, you know, we'll generate and we'll see where the results are. But then the really the way to get better at this is to whittle down that prompt to get to just the parts that are yielding exactly what you want.
And you know, our prompt here, children's book illustration, A young prince in a kingdom redhead with freckles. And what's gonna be interesting there is will it know [00:31:00] to associate the redhead with freckles with the prince. We don't know that yet.
aj_asver: Yep.
ammaar: Kingdom has a lot of castles. Now where's it gonna place this print?
Because now we're describing kind of the background and the setting and the, you know, it's a scenic background, so that's also gonna be interesting. And then soldiers with large black hats, so another character has been introduced as well in this prompt. Usually I like to stick to one character in a prompt or maybe two max.
And, you know, the setting behind them. But, you know, I think this is just gonna be a really great way to see where we end up and, you know, work with that.
aj_asver: That was a really good point, by the way on the number of different characters. Oftentimes when I've seen general AI has had one character in it, and there's a few times where I've tested out with multiple characters that it kind of finds it a little bit harder to discern between the two. Let's see how this one works.
And then if we need to, we can always remove the the second characters.
ammaar: Yeah. That sounds great. Let's let's do it. So I'm just gonna hit enter and we are gonna start generating some art. So the thing I really enjoy about using MidJourney is you start to get a [00:32:00] preview of how that image is being composed live.
aj_asver: And for folks that don't know how MidJourney works, it actually uses AI approach called Stable Diffusion. And what that does is it actually takes images and destroys the images into noise. Like the noise you would see on an old television you know, before we had cable just with like white and black dots, and then it learns how to take that noise and rebuild it back into an image again. And so what MidJourney and other products like it stable diffusion and like Dali have done is they've taken millions of images and done the process of destroying it to noise and building it back up into image. And over time as you do that, you train the model to learn exactly how an image is made in all the different parts of an. and how to take text that's related to that image using a separate model, which is usually a language model and associate them together. And the process that we just saw that am that Ammar showed us is actually that live process of the idea that Ammar fed into MidJourney, the prompt being turned from noise into an [00:33:00] image bit by bit. And cool. What do we have?
ammaar: All right, so the first thing it does is it gives us this grid of four images, and it just likes to give us a few options to work with. And so looking at this grid, we can see that it has indeed placed castles behind this prince. And really it's a closeup shot of the prince. We didn't even specify that, but it decided to kind of give us that perspective, which is pretty cool.
It looks though that while it nailed our scenic background and that magical kingdom feel our young prince is wearing the outfit that we had planned for the soldiers. So he is wearing that black hat. He's dressed in military uniform, I have to say. He looks pretty sharp, it's really cool to see that.
While in some of the images it kind of dropped those freckles. It looks like it's not a fan, but in others it seems to be present. So we see four images, 1, 2, 3, and four, and. . You can see there's a slight variation in the art style. You know, the first one is a little bit it's got a little bit of that painted, like touch to it.
The second [00:34:00] one's a bit more of a, you know, pencil drawing kind of feel on the third. The fourth one is like almost very hyper-realistic to some extent. So now the cool thing is MidJourney then presents eight buttons to you. Nine actually. U1, U2, U3, U4 and V1, V2, V3, V4.
So what does that mean? U1, U2, U3, U4
the image on the grid. So image number one, image number two, you can choose to upscale that. So if you like any of these images and you actually want to use it somewhere, you can click u2. And we're gonna get image number two on that grid in a high resolution format that we can start using elsewhere.
Now, if we like what's happening in image number two, but maybe we wanted to see something a little different, we'd click v2 and that just means variations on image number two, variations on image number one or three or four. And so that's just a fun way to remix and kind of still get to a style that you know you really like before you settle on something.
And then you've got that little spiny arrows, and that just means restart this whole prompt again. I didn't like anything. And so that's [00:35:00] if you just want a little reset, but you don't really wanna change the description. I actually really what we have here and I almost want to double down on, you know, number three.
What do you think AJ?
aj_asver: Awesome. Yeah I mean, it's really cool to see how it generates those four different images and how they're kind of different. And one thing I've noticed about that is, You a little bit of choice of, you know, which image you want to go with. All four of them look pretty great. The only thing I'm thinking about is that black hat kind of makes them look a little bit more like a soldier than a prince.
So I'm wondering if we need to change it up a little bit to remove the hat. What do you think?
ammaar: Yeah, we could try that. So why don't we pick an image we like.
I kind of like number three and I'm just gonna click v3. And what's cool here is MidJourney then brings up that prompt again and allows me to change any part of the description so that I could tweak that prompt and maybe change specific detail about it.
So in this case, you know, let's remove that Soldiers with Black Hats part from our prompt. And so I'm just gonna delete this bit. and I'm gonna add a young prince [00:36:00] in a kingdom redheaded with freckles wearing a crown. And so our final prompt is children's book illustration. A young prince in a kingdom redheaded with freckles wearing a crown.
The kingdom has a lot of castles and scenic background, and we removed the part where it was black soldiers with black hats. That's gone. So with that, we can hit submit again. Let's see what we get.
aj_asver: So how long does it usually take for MidJourney to give you a response? And does it depend on how many people are using it, or is it fairly consistent?
ammaar: It it can be pretty fast. You usually get images generated within 30 seconds. But that's if you're on the paid plan. So if you're on the paid plan, you have what they call fast image generations and fast image generations allow you to basically prioritize your prompt amongst all the other people that might be generating at the same time.
So you get priority in the queue and you get your images faster. For most people though, you'd be on the relaxed queue, and in that it can take up to five to 10 [00:37:00] minutes, sometimes even to get some of your images back. So it can be a bit of an involved process.
aj_asver: Some of these products are starting to use speed and access to getting back the thing you're asking for as a way to kind of differentiate their priced versus free version. You know, the free version, you can get a response, but it might take a bit longer.
And then the fast version, no matter how busy it is or how many people are using MidJourney you'll get a response much faster, but you have to pay for it. How much does it cost?
ammaar: Yeah, so I think the basic plan is $10 per user per month, and then $30, and then just introduced a pro plan, which is $60 per user per month. That just means you get more, more fast image generations which they've capped per plan which is pretty interesting.
aj_asver: Awesome. So if you are finding that you use this tool a lot, or it's useful to you or you want to kind of come back and make lots of different images, it could be worth spending that extra 10 bucks a month to, to get faster feedback from MidJourney.
ammaar: Totally.
aj_asver: Cool. So MidJourney is working away in the background, and for folks that don't know how this works, the MidJourney team have [00:38:00] thousands of servers or hundreds of servers running in the background that are essentially taking all these prompts from all these different people that are using the Discord channel and churning out these images, sending them back to us so we can take a look at them.
ammaar: Exactly. And so it looks like we've got new output and the one thing you're gonna notice is. Our prince looks like an entirely new person. And that's because every fresh prompt just starts from scratch. Again, you're not maintaining your previous history. It's not chat g b t where it understands the context that it was in prior.
This is just starting fresh again. And so we're working with a new set of images and a completely fresh looking print . What I really like though about the images we just got is this crown looks fantastic. You have another great closeup shot of the prince in all four images and you're seeing that somber personality of his come through and the way he's looking, you know, in the distance, but composed calm for a young child.
aj_asver: Yeah. For folks that are listening to this, by the way [00:39:00] and I encourage you to watch the video in the show notes. The princes already do have this kind of somber look to them. And the fourth one in particular
it kind of looks like some a prince that's you know, experienced some hardship but has some hopeful hopefulness.
He's looking directly at the camera, but a little bit off center, kind of with that kind of vulnerable look in his eyes. And it looks like the perfect prince for us, I think. So that's the one I would go with. Let's go with number four.
ammaar: I completely agree. The first one was looking much more like a soldier, I think, in this path. But this one has that innocence in his eyes, which is beautiful. So let's get an upskilled image of that one. So I'm gonna click U4 and that's going to pick the fourth image on the grid and upscale it.
And the fourth image is the one we really enjoyed. And so now what's gonna happen is MidJourney is going to actually add a few more details to this image. Might even change the way the prince looks a little bit because it's trying to, you know, create a high resolution version of what was otherwise a very small square grid.
And we're gonna see that come [00:40:00] to life in.
aj_asver: And that's a really cool feature that MidJourney has actually, is that it takes that small image and kind of fills it out. And it's not just enlarging the photo so we can see it bigger. It's actually adding extra detail to make it a higher resolution image. And as you mentioned, when it does that, it's not just adding extra detail in pixels, but it might actually add extra features into the image itself, which is really cool.
You know, now we get another reveal of seeing what the upscaled image is gonna look like, which I'm so excited to see.
ammaar: Exactly. I think that's one of the most fun parts of the process. You just, you know, you're imagining something, it's coming to life and then it's going in a direction maybe you didn't expect, but you enjoy and want to keep. And I think that process of iteration and back and forth that, you know, we saw with ChatGPT when it was kind of expressing the story and we were able to expand on it, MidJourney kind of does that visually.
And that's really fun as.
aj_asver: Yeah, that's one thing I actually really love about generative ai, and especially if you're not someone with any experience actually using tools like this. And, you know, if you haven't ex had experience using something like Photoshop to come up with an illustration is it's much more of an iter iterative [00:41:00] and conversational experience where you're like going back and forth with this AI to come up with the answer. And it just feels a lot more approachable, and I would dare to say a lot more human in the interface than, you know, pointing and clicking on a bunch of different buttons and a bunch of different you know, squares and circles and stuff to try and make this image. And so to me that makes it way more approachable for anyone to really try and have a go at this.
And, you know, in the space of about 10 minutes of iterating, we've now come up with this awesome very vulnerable but hopeful prince standing in front of his castle that we can use in our in our story.
ammaar: Absolutely. So you can see it's made him a little more stern in this final image that we've gotten but still see that childlike innocence. He's still kind of looking in the distance. Slightly off camera. One other thing we can do real quick is we can just hit light, upscale, redo, and that is just another way to maybe say the MidJourney.
Hey. Give it another shot at that upscale process that you just did, and again, maybe try to tweak some [00:42:00] of those details because I still really liked that original prints had a bit more innocence to him and less that let stern looks. I'm really hoping that light upscale redo button gets us that.
But let's see what happens.
aj_asver: Yeah. And as you do this, one thing to bear in mind is that there's a lot of randomness involved in using generative ai. You know, it isn't something where every time you put the same prompt in, you'll get the same result. And so we're actually gonna see what a different look might look like and exactly how you just talked about that.
We now see a prince that does look a bit more innocent and is a bit closer to our original version. The other cool thing to bear in mind is you're doing this in discord in a chat channel. So every single step of the process, you get to see that as an image, and you can always go back to the step and download that image if you don't like what you get when you know, do the redo.
ammaar: Exactly nothing is lost. You can just keep expanding on your iterations, which is really fun and looks like it's giving us that young, innocent prince that we wanted in a high resolution version, which is awesome.
aj_asver: Very cool. So we have our Prince Harold. He's taking [00:43:00] shape. We're seeing him in front of his castle. We have an upscaled image of Prince Harold that you created using generative AI using MidJourney, and we did it. By first having the story that we made in chat G p T. And then from the story we created some scenes, which we described.
Now we described it ourselves, but we also showed you how you could describe the scene in chat G P T and use some of that language from describing in chat G P T to actually then create the prompt that you put into MidJourney as well. And you know, there's many different ways you can approach this.
You can use your imagination, you can ask chat g p t to help you. It's a really fun process and one thing I've really enjoyed about using generative AI is that kind of back and forth fun of trying different ways and experimenting. And I really encourage everyone that's listening to this and wanting to try it out for yourself to just experiment a lot and try different processes, try different ways, click the different buttons and try the different variations to see what you get. And don't be afraid to, you know, experiment.
ammaar: Absolutely. It is just as much of a process, of just [00:44:00] iteration and back and forth and, you know, brainstorming with this bot, you know, almost to come and bring your ideas to life.
aj_asver: All right, so we've got this image of Prince Harold. He's staring into the distance. He's hopeful, he's looking a little bit sad, but this is kind of the beginning of our story. How do we get this onto a page and then turn it into a book
Upscaling with PixelMator
ammaar: Yeah. So the first thing you're gonna notice is even though we've gotten, what I would say is a high resolution image of what was on the grid, it's still not high resolution enough for print. And so we're going to use another tool called Pixel Made for Mac which you can get on the Mac App Store.
And it's got this amazing feature called Super Resolution, which uses machine learning to try and expand the image that you have to a high resolution version and maintain the details that were there. And I'm gonna show you how you can do that in just a couple of clicks and it's just magical.
aj_asver: So we've got this image, but we wanna make it even bigger. And I assume that's because in order to. this as a book, you need a super high [00:45:00] resolution image. And so you're gonna use a different tool for that. Now remind me, that tool is called Pixel Meter
It's a Mac app, that you can install off the app store.
We'll include a link to that app in the show notes for anyone that wants to try it out. And there are many apps like this that do upscaling that you can Google for as well. But we're gonna try this one.
ammaar: Exactly. Awesome. So I'm going to save this image from MidJourney to my desktop and then I'm gonna fire a pixel made and we're just gonna blow this image up, so that's ready for prime time.
We're in PixelMator and
all I've done is dragged that image that I saved from a journey onto the PixelMator icon on my dock. But if you just fire open the PixelMator app, you can just open image, select the image you downloaded and it opens up. So now the main thing I want you to focus on is look at the resolution of this image.
It's 1536 by 1536 pixels. So it's not that big. It's a bit of a, it's a square, right? But I want this to be really big. I want it to be good enough for a book cover. I want it to be good [00:46:00] enough for pages that are gonna come in high quality print. And all I'm gonna do, there's just two clicks and we're gonna make that way bigger.
So I just click the three dots in the corner and hit super re. and it works like magic. And so now it gives me this before and after where you can see the details that are actually preserved, there's no blurriness. So notice that we've got this 1536 by 1536 pixel image. You know, it's a good enough size square, want something a lot bigger, something that could be printed and high quality on pages, on book covers, posters, whatever. So to do that, I just have to click the three dots in the corner here.
And then you wanna scroll down and look at this thing here in the side, which says super resolution. And so when I click that, it's going to blow up this image and take our image from the small square 15 by 1500 to something a lot bigger. And there you go. So in just that one or two seconds, you [00:47:00] know, you're seeing that there's no level of detail that's really lost between the two.
But more importantly, that image is now a 4,000 by 4,000 pixel image. and you know, I can just keep going. I can just go again and click super resolution and it's a little slower this time because it's making it huge, right? And you probably don't have to go this big, but you can, and this is helpful if you're making an animated short and you want it to be a 4K resolution shot.
But now you can see it's a 13,000 by 13,000 pixel image to really high quality. And, you know, you can zoom in and you can see that there's no blurred lines or anything like that. This is gonna work great for our book.
aj_asver: This is really cool and it's a little bit harder to see it in the video, but I've tried this myself and seen how impressive it is to see the amount of pixels that it fills. And so it's something worth definitely trying yourself. We're not just making the image bigger. We're actually adding more pixels to it, and it's done using artificial intelligence to actually [00:48:00] work out where the image should add pixels and what those pixels should look like. And now we get an image that's big enough for us to put on a book.
ammaar: All right, so now that we have our high resolution image, let's just put this on a page of a book with the description we got from chat G b t to bring it to life.
So the first thing I'm gonna do is I'm going to export the image that I just created with PixelMator and, you know, pick the JPEG format. It's gonna be a five megabyte image. Gonna click export here, and I'm gonna just choose my desktop and save it. I think because I'm sharing my screen a certain way, you can't see the save dialogue, but just imagine that classic Max Save dialogue is there and I'm just choosing my desktop.
And now we're gonna move over to Pages, which allows us, comes free with all Most Max today. And that just allows us to compose our book by giving us the right templates to start doing that. And it's really easy.
[00:49:00]
aj_asver: Okay, so we've got a high resolution image. We've got some texts that we generated with ChatGPT. We took that image from MidJourney. We scaled it up using PixelMator, and now we're gonna put it into pages,
Laying out book with Pages
ammaar: absolutely. It just comes with your Mac. And if it's not on your Mac pre-installed, it's free on the app store so you can easily grab it.
So now I'm in pages and you know, when you click new document, you're asked what kind of document, choose a template. And on the left hand side you can see basic reports, books, letters, where after books here, right? So you can click books and you can see that it's got a bunch of, different templates for us to work with.
A basic photo book, contemporary novel you've got a storybook. There's some really nice ones for us to work with here. I like the storybook one. It looks like a children's book. It kind of fits the vibe we're going for. So let's go ahead and create with that [00:50:00] one.
So I clicked that and now I have essentially a template to work off of. It's given me a few different layouts, right? We've gotten the cover page a page here, which tells a little bit of a story with a big image as like the focus. A page that's just dedicated to an image.
And this is great because we've got enough here to work with to kind of lay out our book and give it that style. So let's just use that image that we had of the Prince and, you know, start our story the way ChatGPT kind of helped us describe it. So what I'm gonna do is I'm going to grab that image from my desktop, and I am just going to drag it here and bring it to life by putting it in there.
So as you can see, I just dragged my image over to Pages and it gave us the Prince as the focus of this first page, which looks great. We can obviously also just double click and slightly move it. If you wanna maybe emphasize this crown a little [00:51:00] bit. You could also resize a little bit, so maybe, it's not exactly that whole width, but you're getting a scene of that castle as well in it.
Again, it was cropped out before, slightly above the chin. Let's give him some of that room. And there you go. You've got the prince looking there at the reader, which is pretty fun. And now I'm gonna go grab that description we got from ChatGPT so I've just copied the text that we already had saved, and I'm gonna paste it in here.
So one thing you'll notice is messed up the formatting, but that's okay. What we're gonna do is go to text and just choose the body template text that it's already given us. And that'll fix it. But now you'll notice that it's a little out of the box, right? It's not fitting. So we can just slightly resize that and it fits into place but it's also given us these nice headers that we can work with.
So what I'm gonna do is actually, I'm just gonna cut out this once upon a time from this part and paste that in here. And that [00:52:00] just gives it that nice dramatic storybook feel as you're reading it. And I don't think we really need this one, so I'm just gonna go delete that. And now what we've actually done is we've created a little bit of room for ourselves, so we can actually increase the size of the text again and drag the box down just a little bit.
And there you go. There's our first page of the storybook for us to go ahead and roll with.
aj_asver: Wow, that was amazing. So we went from an idea for a storybook to then generating the actual story using ChatGPT to then creating some descriptions for what the first scene might look like. Putting that in MidJourney, going through a few iterations to come up with an illustration, and then using PixelMator to make that image even bigger so that we can use it in print.
And then we brought it into a Mac app called, pages, which is available on your MacBook. And we've used a template to turn it into the first page of a story. And this is amazing. So we've actually started writing a story now, and obviously you can keep doing this and you can [00:53:00] continue the rest of the story, which is what exactly what I'm gonna do now that Ammar has shown me how to do it.
But there's one piece that I still don't understand how you did Ammar. So we, let's say we made whole book.
How did you get it? Into the Amazon bookstore, how do you make it into an actual book that someone can have in their hands?
ammaar: Yeah, for sure. So once you've kind of composed all of your pages, what you want to do is export this book and that export functionality is available on pages. You just click on the MAC toolbar, click export, and export to pdf. And so once you have that pdf, you're going to go to Amazon KDP and publish your title.
aj_asver: what's Amazon KDP?
Publishing on Amazon KDP
ammaar: Yeah. So for those who don't know,
KDP is the Kindle Direct Publishing platform for Amazon, and that essentially allows anyone to publish their own Kindle books. They just make you sign up, provide tax information if you're selling that book. And then you can go and fill out the information about your book.
So the title, who the authors are illustrators, description, and then you [00:54:00] just upload your pdf. Then once you're on KDP, it's just gonna ask you if the formatting looks good. You can also enable if you want it to be a paperback and click next, next, next, and you're done. And then it's live on the Amazon store. Within 72 hours if approved.
aj_asver: That is amazing. So within 72 hours of having this book ready in pages, you can actually have it on the Amazon bookstore. And now it's not just a Kindle book. They actually give you a paperback version or a hardback version too, that people can then buy. And did you have to spend any money to publish this book?
ammaar: I didn't have to spend any money apart from the MidJourney subscription. That was the only thing that cost money here.
aj_asver: That's so cool. So you really didn't have to pay for publishing the book. Amazon is basically kind of putting the book there and then if people want it, I guess they print it and then send it to 'em. So it's kind of on demand.
ammaar: It's on demand. Print on demand is this new phenomenon actually, that they they've somewhat pioneered, which is, yeah, they don't have to hold all this inventory. They can just print on demand with all these printers across the world. [00:55:00] And the way they make money from it is if there's a sale, they take a cut from your sales.
And so you don't have to put up any upfront capital. So if you're someone who wants to put your book out there, you don't have to worry about buying a load of books and hoping that someone's going to buy them and hold that inventory. Amazon's taking care of that for you.
aj_asver: Wow. So really thanks to Amazon, anyone can be their own publisher, which is so cool. And I'm gonna go do this. I'm so excited to make this into a real book and publish it and then get a copy of it delivered to me so I can show everyone and all my friends just like you did. Now, do you have a copy of your book?
I really wanna see what it looked like.
ammaar: Here's the book, it's the hardcover version. And, you know, it's it's in my hands which is kind of crazy, right? Yeah, I love some how some of these images came out.
And yeah, the print quality as well was also really impressive. This one's my favorite image in the book kind of shows Alice all grown up with her robot friend. So yeah, there you go.
aj_asver: That is so cool. And one of the things to also think about is if you don't wanna publish this book to everyone and you just want to kind of make a [00:56:00] gift, let's say it's for Mother's Day or for your friend, or maybe you want to gift it to a niece or a nephew, you can just publish it and then buy one copy.
And I guess after that you can stop publishing that book. So it's, that's kind of a cool hack to just make a custom book for yourself.
ammaar: Absolutely. Yeah. I think this is a great way to just gift personal stories to people, which, you know, is just, it's amazing that we can now go ahead and do that and realize that.
aj_asver: thank you so much Ammaar for being down to show us exactly how you illustrated and published a book in the space of a weekend. And anyone can try this. Just follow along with the steps in the video. It's very straightforward. We'll include all the links in the show notes as well, so you can give it a try to.
Ammar, so glad that you were able to make it. I really appreciate you joining me for one of the first podcast episodes . Do you have any future plans to create any more cool stuff? One area you haven't ventured into is music. Should we expect an Ammaar Reshi billboard top 10 anytime soon? Yeah.
ammaar: I [00:57:00] think I'm waiting for Google to release that model so that we can start working with some music. I've also been seeing a lot of these 3D tools out there now where you can text to 3D models and so might be fun to make a little video game with generated AI assets. That could be really fun.
I'm just gonna let inspiration take the wheel, so I'll let you know when that happens.
aj_asver: Amazing. Thank you so much for listening. I really appreciate it. If you enjoyed this podcast and wanna learn more about generative ai, about the world of ai, and follow me as I explore, artificial intelligence and the way it changes how we live, work, and play, just subscribe to this podcast and your favorite podcast player and we'll have more episodes coming up soon too. Thank you, Amara, and have a great day.
ammaar: Thanks for having me. Thanks.