This week I was tasked with doing Middlebury Art’s demystifying Ai- digital detox assignment. I used two AI photo generators to create art based on a prompt.
My prompt was J Cole on Mount Everest.
Image 1:
While both weren’t what I was expecting this image was the worst. The character has no facial features, no distinctive characteristics, and looks out of place and uncomfortable. I know that technically I would be uncomfortable in the cold but still. This character could be anything and I feel as though the perspective is wonky I am seeing certain characteristics of the mountains that I shouldn’t be able to see or wouldn’t see in something man made.
Image 2:
https://deepai.org/machine-learning-model/text2img
This art was the best out of the two options. The AI produced more distinguishing characteristics and I could tell that it TRIED to recreate Jcoles characteristics. I am also not sure why AI has a problem with creating hands. In both of these, the hands are in the pockets which is okay but it would be nice to see some differentiation in poses. The mountain here, however, is blurry.
I don’t see AI taking any artists’ jobs any time soon. AI will need a lot of advancement to replicate the complexity of human art. Also, Ai creates art that feels tone-deaf to me. As of now, it is not possible for AI to eventually take an artist’s job but it could happen in the future.