Generated AI Imagery

2 users here now

Imagery generated by artificial intelligence / machine learning using keywords from community subscribers.

Each post (or thread) represents a theme. Optionally, generated images can default to the model specified in between brackets, in the post title. For example:

Halloween 2022 "art" thread! [model=dreamlike-art/dreamlike-diffusion-1.0]

All top-level comments to the thread will result in a new AI image being generated from the keywords in the comment. The generated image can take several minutes to appear, but will eventually show up as a reply to the comment.

Additional parameters can be referenced within the comment:

• width=512
• height=512
• steps=50
• model=stabilityai/stable-diffusion-2-1
• seed=<random number>
• negate="<negative prompt>"

For example, a comment could be:

A spooky house atop a small bluff with a full moon and dark sky, bats flying across the moon and a witch in the top right corner flying on her broom, black cat stands on the house porch staring at the viewer with the door partially open width=768 height=512 steps=45 seed=123456789 negate="cartoon,low quality"

AI models are trained using differing input datasets. So the resulting image will typically vary significantly from other models.

spoiler

---

MODERATORS