Jitesh's Domain

Game Designer. Producer. Gamer.


Leave a comment

My concepts with AI Art Generators

After a long break, here is a new post to give my readers an idea about new things I’m trying during my free time. My current obsession is making short stories, comics or other original concepts with the help of AI tools like Midjourney and Stable Diffusion.

Why AI Tools?

For a long time now, I’ve been wanting to work on my own concepts & short stories… to come up with something original and if possible, thought-provoking and creative. I’m not a proper artist. With the availability and access to these new massively engaging AI art generators, it’s quite helpful for me and others to generate concepts that align to our vision. Atleast, as close as one can get. It’s not without flaws, but the benefits they bring are enormous. That’s what I believe.

Currently, there is a rift among the art community on using these tools, even some of my friends and colleagues are a little split. In my opinion, as long as one can use the tech and the knowledge to get what they want to create, it’s all good and well. Few of my artist friends are already using these tools to save time and paint over, to further enhance the output. Truly, the possibilities are endless!

Some Concepts

I’m creating all of my story/comic concepts using Midjourney and Stable Diffusion. In the future, I’ll explore other AI art generators and perhaps mix and match certain things. All my work (for now) is mainly shared on my other instagram account: uns33n.comics. Do note, most of the images I shared on the website here are of medium quality.

Here is a sneak peak (already present on my instagram account):

Realistic Portraits

Check out some of my ultra photorealistic pictures made using Stable Diffusion:

Artistic & Creative

Midjourney gives me the best results when I want something to be more artistic, wacky and fantastical:

Old to Modern?

I also tried a sample to give a modern look or to make a character from a video game realistic. In this case, I went with Fran (from Final Fantasy 12), because I managed to find a good image to use as a reference. It is important to know that img2img also works, which can bring a lot of interesting and sometimes strange results. It depends on how we manage the parameters to achieve our intentions.

Someone completely different?

As I briefly covered img2img above, you can actually also shapeshift someone, create a doppleganger or a completely different person! Check my try:

These tools are also both supporting and competing with each other. All of them receive regular updates, improvements to their algorithm, more artistic possibilities, variations and so on.

I’ll continue making more concepts and stories. If you like the work, do support and give a like, follow and share: uns33n.comics.