Generating AI illustrations for my Smartle word game
Creating art with AI is a fascinating and exciting process. However, it can also be unpredictable and sometimes disappointing. Recently, I attempted to generate a picture using AI where the English alphabet letters would be scattered, and a woman would be sitting among them, trying to form a word. The purpose of this picture was to illustrate Smartle, my daily word game where players swap letters to form five-letter words.
Some women, but do you see any letters? I don’t.
As I began the process, I was excited to see what the AI could produce. However, it quickly became apparent that the AI had no knowledge of the English alphabet. I tried multiple times, using different AI picture generators, but the results were always disappointing. Instead of the desired image, the AI generated pictures with random and seemingly unrelated elements.
This one doesn’t seem related at all! No letters, no woman..
Although the AI-generated images were not what I had envisioned, the experience was still valuable. It was a reminder that AI still has limitations and that it is essential to understand these limitations when using AI for creative purposes. Perhaps in the future, AI will develop to the point where it can generate the specific type of image I was envisioning, but for now, it seems that some things are still better left to human creativity.
These are probably the closest that I got to my idea. There are woman surrounded by some elements which resemble letters and seemingly trying to organise them.
