MidJourney Comicbook Experiment

I posted here about this comicbook experiment last year, but I’m revisiting it now because there will be a follow-up soon connected to it. Nothing of major importance, but I wanted to explain where and how these few pages of comicbook came to be.

Fifteen years ago, I wrote a short sci-fi story about an agoraphobic woman living in an old Victorian house at the edge of a village near an ancient forest. She lives with her two kids, her older sister, and a small childlike robot adored by the children. One day, during a blizzard, the kids and the robot get lost in the forest. The mother, trapped by her own fears, is unable to search for them. When the robot returns with unsettling news, unimaginable events unfold for the two sisters.

Over time, this simple tale expanded into a vast and intricate universe, populated by various characters. The story now spans from ancient kingdoms and their epic wars with elves to a futuristic space opera set in a collapsing Solar system, where humanity is battling evil corporations along with a monstrous alien force.

When MidJourney launched over two years ago, one of my first prompts was to create characters from this universe. I was thrilled to finally have a medium to visually explore and enhance the world I had built only through words.

Although comicbooks were never my primary focus, I joined some MidJourney Facebook groups, and one in particular caught my attention: the AI Comic Book Creators group. That’s where my journey into the world of comicbook experimentation truly began.

During late spring 2023, while MidJourney was still in version 5.1, I decided to experiment with creating a comicbook using MidJourney. I picked the ending of the original story for this comicbook experiment because I could better control that scene visually and it was easier at that moment. Additional details about this comicbook experiment are available below in this short presentation.

While traditional comic creation involves a team of illustrators and layout artists, AI allows you to handle it solo—yet the workload remains immense. Key comicbook principles like visual storytelling, panel layout, and speech balloon placement were central to the experiment. My first attempt was using realistic AI-generated photos, but I quickly realized that illustrations were the true language of comicbooks. After many trials, I managed to refine the design of my main character, Dr. Clark, maintaining her consistency across different angles, moods, and poses.

Choosing supporting characters and backgrounds followed a similar process, experimenting until finding the right fit. A key lesson I learned was about balancing text and visuals—comic dialogue needs to complement the images without overwhelming the reader. I learned about ChatGPT's Fountain-Style Format during the process, but it was too late to use it for this project. However, it was a valuable lesson for future projects.

This project has been an ongoing experiment, filled with trial and error, but it's a rewarding process of learning how to create something new with AI.