Ai Art

Before the "Mid Journey" period it used to take a long time to produce a fusion of two characters, as it involved a long sketch phase, a creative characteristic combination phase where the most prominent characteristics of the characters were combined, and of course, the execution phase which could take several days.

‍

My first experience creating a fusion of two characters was while studying visual communication design at “H.I.T.” In an illustration course during my second year, we were asked to create a fusion of two different characters. The challenge was to see if we could identify the most prominent features of the characters and create a new character that combined those features so that a neutral observer could immediately recognize both characters upon looking at the illustration.
‍

‍

‍

‍

Is it the same in Mid Journey?
‍

The answer is both yes and no.
Just like I had to identify essential features of each character to create a new character that was recognizable, it was similar in Mid Journey.

I learned that if I relied too much on the AI to do the combinations, it wouldn't work. As evidence, in my early attempts, the result was completely disconnected from the final product.

Throughout the creative process, I had to identify and define the prominent characteristics I wanted to be present in the work and ensure they were there, whether it was the color of the suit, an iconic item the character must have, a hairstyle, a cosmic portal in the background and more, while i'm using prompts and reference images to create more accurate result.
‍

‍

Why its important to present the process?

‍

As a designer, I understand the importance of presenting a case study of the creative process, so the viewer can understand the struggles and creative decisions made along the way.

That's why I presented the creation process of several characters I created to emphasize that this isn't "hocus pocus." Some people, especially those who haven't experimented with this technology, think all we do is type a few words and voila, everything is perfect. But that's not the case.

It's true that the Mid Journey can produce anything that's registered in a way that is perceived as  "beautiful" according to the conventional definition of beauty. However, note that all the images that are not a result of careful consideration, but instead are the result of overly simple prompts, look alike. For example:
‍

"a pretty woman portrait."

‍

Mid Journey has a default look and feel that it uses when you don't define a style, and people are getting tired of this generic style because it's everywhere. That's why there's a push for uniqueness and the creation of more complex and accurate prompts for our imagination. We should be directing and training the robot to suit our desires, not the other way around.
‍

‍

‍

As seen in my processes, I always started from the most basic place and checked if the Mid Journey knew the characters and knew which elements to combine.

‍
In most cases, it didn't know or didn't know how to do it well, so I had to make several attempts to improve the prompts through trial and error while using reference images of the characters I was trying to create. With each attempt, I got closer to the result I was aiming for.

‍

‍

As a Pokemon fan, I used Pikachu as the basis for a series of combinations I wanted to try with Mid Journey. The goal was to test its limits and see how well it could perform character combinations. The results surprised me, and I invite you to see for yourself the rest of the series and artworks I created with Mid Journey.

‍

More artwork processes

‍

‍

‍

‍

‍

‍

Single artworks

‍

‍

‍

‍

‍

‍

‍

‍

‍