CONCEPT
Color as Channel
The information-theoretic observation that color in a visual artifact carries distinct semantic channels — black for structure, red for alarm, blue for link, orange for emphasis — each with its own cultural history and cognitive load, and each transformed when AI begins producing output in every channel at once.
Color in visual communication is not decoration; it is channel. Each named color in a document — black text, red warning, blue link, orange highlight — carries accumulated semantic weight from centuries of convention. The reader who sees red in a paragraph processes it differently from black before she has consciously registered the word. This is the pre-attentive layer of visual cognition, and it is the layer at which AI-generated imagery most unpredictably intervenes. A language model that produces inline SVG with four distinct colors is not making four aesthetic choices; it is selecting four positions in a shared cultural signaling system, often without understanding what the positions mean. The test of whether this matters is whether the resulting artifact communicates or merely performs communication — the fluency trap extended into the visual domain.
In The You On AI Field Guide
The color channels