Midjourney’s AI-generated images of potatoes seeing French fries for the first time and looking shaken. I needed to do lots of refreshes to get the results I wanted.
Basically, an AI digital art image of potatoes getting surprised and shocked at seeing what happened to their fallow potato friend. Getting slaughtered and becoming French fries is a funny-looking image.
It’s rendered to look realistically using the “realistic octane render 8k” phrase. I used --ar the aspect ratio property in Midjourney of 10:16 to create a vertical image. I explicitly told Midjourney to render the fries “in front of him”. The composition was perfect in many of the AI-generated images.
I could get more variations with different facial expressions, add a kitchen knife, or even change the scene background. But I think it looks well defined for the two main subjects in the image, and it leads to a quick realization of the witticism behind it.
When generating AI art using Midjourney, try to consider at least the composition and choose an appropriate aspect ratio using the –ar property. Then you can carefully describe where you want the subject’s images to appear. It’s kind of like carefully directing the scene. Otherwise, it can lead to results that don’t look great.
On another side note, something when you go to an experiment, you might get interesting compositions. But generally speaking, after a lot of experiments with the Midjourney Discord bot, I found out that it’s better for me to choose the aspect ratio ahead of time, but still, I give myself room for experimentation of course.