Design manager Ammaar Reshi used AIs ChatGPT and Midjourney to create the text and images for a children’s book, which he titled Alice and Sparkle.

It looks like a book, but is it really a book? Any writer, including myself, is probably likely to say no: Books take months and usually years to write, and illustrations for children’s books are usually painstakingly detailed. The word choice, the aesthetic of the book, and, of course, the inspiration–the preciousness of it all–makes us feel like book-writing is a different process than simply iterating over and over using an AI.

Reshi spent hours discarding hundreds of images that were too spooky for a children’s book.

“Now, let me tell you, some of those results were absolutely wack. It would have become a horror book if I put those early illustrations in,” he said.

Yes, to me, a lot of AI art still has that Coraline’s-mom-with-button-eyes quality, but the finished images Reshi selected do not look horrifying. However, it’s strange that when using AI to make art, we have to do this kind of crawling out of the uncanny valley…teach a machine what is and is not creepy.

So what is this, if not a book? Reshi created this piece of work in about 72 hours, and put it on Amazon. He originally created the book for a friend who had a child.

It seems that Reshi is genuinely interested in computers and what they can do — his passion for computers is real. His passion to create a book with AI is real. And the work he spent selecting good images and iterating on them…those are real work hours, too.

It’s not the same kind of work I do when I go for a walk, an idea pops into my head, and then I make up playlists for characters in my book (this is the anthem of my newest protagonist, Daria Muñoz). I have read a lot of books, but I’m not exactly compiling content from other books to churn words out.

(The other thing is: I’m always going to write. I don’t want AI writing to take over, but if it does and I’m out of a job, I’m still going to write books and stories and posts in my spare time. I have to; it is who I am.)

At the same time, the language we use seems to suggest that we do think a person is doing something. “I made this using Lensa.” “I made this using Midjourney.” We don’t say these things are creating on their own. We do acknowledge the human component: The human has to give a command to set everything in motion.

Yesterday, a friend of mine used an AI to generate some news headlines. As someone who writes for news, I was curious…but then, I actually felt kind of relieved. The AI was generating news for November 20, but didn’t mention the holidays, shopping, or Thanksgiving. It was just general headlines.

Human news is in part centered around our reality–what is going on, what we anticipate. Our artwork is the same…it’s us reacting to what is going on around us. Even our obsession with AI art is just more of humanity responding to what is happening in real life.

Alice and Sparkle is a book, but it’s a different kind of book. We don’t have a genre for it yet; maybe we will soon. This creation is what happens when a human being who usually doesn’t write children’s books tries to get a machine to write a children’s book. It’s an exercise in what computers can do by someone who is fascinated by computers.

And honestly, I’m not threatened by that. I think I can coexist with that.

I can respect his passion as I’m sure he can respect mine.

We will always need human artists.