I've been using #inktober as an excuse to play a game of man v. machine. I ask Night Cafe for an AI rendition of the prompt which I put on SteemIT.
I then take a pathetic stab at drawing an image which I put on VY Brainium.
It takes me a really long time to draw a picture. I often fail to complete the image.
I am interested in using AI to enhance images. When I complete a drawing, I try enhancing the image with Night Cafe which I post here.
This means I have original posts for three different platforms.
The Guidebook Prompt
I wanted Night Cafe to generate a cartoon guidebook that was guiding people. I could not get AI to create the image I wanted. So, I asked for an image of a person walking through Florence holding a guidebook.
The AI image is absurd. The dome on the Duomo should be more pronounced. Here is the AI image.
For my drawn image I decided to draw a picture of a person being misled by a guidebook.
Imagine a guide book that tells a spelunker to turn at the yellow column, pass the green crystal and cross the delicate bridge.
The guidebook is 40 years out of date. Delicate bridge collapses a long time ago. So, my hapless spelunker is being misled and stepping into an abyss.
I drew the underlayer of the image with Prismacolor. I then outlined items in the picture with a fountain pen to produce this image
I should note that tie actual image is much darker and saturated than the photograph. The pictures I've taken of late appear washed out.
I asked NightCafe to enhance the image.
The image was too bright. The image is in a cave. Caves are dark; so I adjusted the brightness to make the image darker to produce the image at the top of the page: