- This topic has 5 replies, 3 voices, and was last updated 2 years, 7 months ago by
.
-
-
Many artists have embraced AI and the ability to enhance images or create entirely new ones from just a textual description. The many available AI tools differ in their “creative” abilities. I find it amusing that often faces are very well detailed and accurate, but when it comes to hands, AI seems to have rather limited abilities. Often the number of fingers is unrealistic, and the shapes are deformed.
One of the tools I use often ignores parts of my text. Recently I asked for a boy on a sled as part of a longer input text. The generator in a half dozen tries never managed to put a boy on a sled. Is that because the learning set did not include such images? I settled for a boy on a pile of snow in my “Snowy Street” creation.
Like so many other tools and media, artists will try them. At first the resulting works are more like talking dogs – not praised for eloquence but admired for the modest accomplishments.
Attachments:
-
Well, I went back and tried again. This is what I asked for: “Boy on a sled being pulled by another boy. They are in the middle of the street. There is snow all around. The location is a modern urban area with tall buildings. There are shops with brightly light windows.” Here is what I got.
Attachments:
-
