Learn how I work with my visual coding agent, Nadia, to rapidly create a piece using a coding language that I’ve never learned.
Generative animation always seemed interesting, but felt out of reach. Learning a visual coding language has been a "someday maybe" thing for ages, but I never had time for it. Once AI came around as a coding assistant, I figured... why not have it write THAT kind of code?
Processing / P5.js is an older language that I've always been curious about.
Some other work people have made with it:
For IMCAS, a conference I make visuals for on short timelines, I wanted to see if AI could help me work in a coding language I'd never learned. I downloaded Processing, showed my AI visual assistant Nadia some references, and art-directed her in plain English: how it should move, what to speed up, by how much. She wrote the code, and we iterated fast.
I showed her some references, and she created this scanning projector.
Then I gave her the company's logo, and told her to swing the projection side-to-side.
I liked it. but needed denser points for a clearer logo.
I liked the clarity, but missed how mysterious the sparse version looked. So I had her to oscillate the density slowly back and forth. The result, made in about a day, would've been impossible in After Effects in that time:
Is this cheating? I don't know. I didn't draw the lines myself. But it looks exactly like what I had in mind when I started, and I kept giving commands until I saw what was already in my head. Is that any different than a user interface? Isn't AI just acting as a front end for these tools that never had one before?