Okay, so today I wanted to mess around with creating some AI art, specifically focusing on a character named “Ji-Hyun Suh”. I’ve seen some cool stuff online, and figured, why not give it a shot myself? Here’s how it went down.
Getting Started
First things first, I needed an AI art generator. I’ve dabbled with a few before, so I just picked one I was somewhat familiar with. Nothing fancy, just something to get the job done.
The First Few Tries…Yikes
My initial prompts were pretty basic. Stuff like, “Ji-Hyun Suh, portrait,” or “Ji-Hyun Suh, standing.” Let me tell you, the results were…not great. We’re talking distorted faces, weird body proportions, the whole nine yards. It was clear I needed to be way more specific.

Adding Some Detail
I started thinking about what I actually wanted to see. I figured a specific art style might help, so I added “in the style of Artgerm” to my prompt. I also threw in some descriptive words like “beautiful,” “detailed lighting,” and “intricate details”.
This definitely helped! The images started looking more like actual art, and less like a digital nightmare. Still, Ji-Hyun Suh was looking a little generic. Time to refine further.
Getting Specific with the Character
I remembered some key traits. Short hair. So, into the prompt those went “Short Hair”.
The “Aha!” Moment
I played around with the settings on the AI generator. There’s usually a bunch of sliders and options, and honestly, I mostly just mess with them until something cool happens. I think I increased the “detail” setting and maybe tweaked the “style influence” a bit. And boom! Suddenly, I was getting images that actually looked pretty good. They had that soft, detailed look I was going for, and Ji-Hyun Suh was finally starting to resemble, well, Ji-Hyun Suh!.

Iterating and Refining
From there, it was just a matter of making small adjustments. I tried different variations of the prompt, like adding “looking at viewer” or changing the background slightly. I probably generated dozens of images, picking and choosing the ones I liked best.
Final Thoughts
Overall, it was a fun little experiment. It definitely took some trial and error, but I learned a lot about how to coax the AI into generating what I wanted. It’s all about being specific with your prompts, experimenting with the settings, and being patient enough to keep iterating until you get something you’re happy with.
I’m by no means an expert, but it was cool to see how much you can achieve with just a few keywords and some button-mashing.