Dots Study
Generative dot waves and grids designed to destabilize perception. Sine waves modulate dot size diagonally across the composition while subtle position drift prevents visual anchoring. The high-contrast patterns trigger optical artifacts—ghosting, flicker, apparent motion—where static geometry exploits visual processing failures. Each generation explores a different point in the algorithmic space.
Click to enlarge ↓
Inspiration
Inspired by ideas in Vertigo: Op Art and a History of Deception 1520 to 1970, and particularly the work of Bridget Riley and her Static series of paintings (1966). Riley is a foundational figure in Op Art who believes "perception is an active process—a dynamic relationship between the viewer and the artwork." Her dot paintings employ hundreds of carefully placed circles that create illusory movement in static compositions.
This project was an early experiment in AI-assisted creative coding using p5.js. Where Riley hand-painted each dot, I wanted to see what an algorithmic approach might reveal. Starting with the simplest possible initial state (dots in a grid), I altered parameters to push the boundaries of the original idea while trying to maintain the illusory perceptual effects.
If the original painting represents an equilibrium within a larger parameter space, I wanted to see what the more extreme or surprising states would look like. Would they create the same illusions, or different ones? Would another variation prove equally satisfying? Would the original ever emerge?
I found that the model (GPT-4o) didn't have a good sense of what would be interesting aesthetically but was a great partner for iterating quickly based on my visual feedback and guidance.
I left the algorithm intentionally unpolished so viewers could explore the space themselves and decide what constitutes a "good" or "bad" output.
Live Version
Chris De Giere
Generative algorithm and digital image using p5.js
400×400 pixels
2025