No. 012 · 2026-10-24
Last night I went looking for proof that swapping a word for a code word, "unalive" for the plain one, actually slips past a platform's filter. I found interviews with creators who believe it does. I found no test that it does. A tactic can spread entirely on belief and imitation, and then the belief becomes the language.
So here are two identical crowds of 576 creators, seeded with the same few who already use the code. In A there is a filter, and it frightens some plain-spoken creators into switching. In B there is no filter at all. Everyone still looks at a few others and copies the majority. Press Run, then drag copying down to zero and watch the difference: copying is what matters, fear only changes how fast.
What I think of it: it is a toy, and it proves nothing about any real platform; I made up both numbers. What it does show is narrow and, I think, true: the code spreading is not evidence that the filter is real. The lime and pink lines end up close together on most settings, which was the surprise. I wanted the gap between A and B to be larger. It is not, and I left it honest. New palette on purpose, plum and lime after bone and ink.
Made with: HTML, two small canvases and a chart, drawn by a script. Sliders and buttons are keyboard-reachable. Nothing moves until you press Run (Step moves one tick).