Generational Shift
Using A.I. to make art (or something)
I’ve been making A.I. art. But maybe it’s not art? Even if it’s sort of beautiful.
Not just a few pretty pictures: I’ve been generating many thousands of A.I. images. But is it me making the images, or the A.I.-powered, image making, prediction machine? (Namely, Stable Diffusion.)
To be more exact, I’ve been making thousands of A.I. video frames, 800 images at a time, and upscaling them into 4K videos (also using Stable Diffusion).
These pictures are all frames from those videos.
800 frames at 24 frames per second gets you around 30 seconds of video.
Here’s an example (without sound):
I then make music scores (also 30 seconds) and add them to the videos. (Link below.) It helps the weirdness go down a little easier.
The music is probably the most satisfying and just fun part of the process.
But to zoom out again — what am I actually doing here? I am the maker of these videos — managing prompts and programming motion. But I’m mostly using trial & error to direct a system I still can’t predict and don’t understand.
So, again: is it art? Or is it merely computers, vomiting mindless hallucinations?
(Is the music just human Lipstick on the robot Pig?)
You tell me! The first few videos are going up today:
