Concept · part of What the filter rewards

Picture a status update that people copy and paste, some making small edits along the way. A study of more than 460 million instances of Facebook memes, Information Evolution in Social Networks, found that the resulting spread of versions mostly looks like chance. Most variants spread no better or worse than their parent. They pile up by random copying plus a rich-get-richer effect, where versions with many copies get copied more. Statisticians call this a Yule process.

The neutral drift baseline is that picture: what spread looks like with no preference at work. It is a null model, a simple map to hold real data against, and what you can see depends on the model you bring. Selection only becomes visible as a departure from it, as when the same study found phrases (“paste if you agree”) that reliably beat chance.

In the book’s terms, drift is what lossy retelling produces: a directionless cloud of variants for selection to pick from. The baseline also limits the book’s own argument. Most of what travels is neutral copying, so engagement-tuned platforms shape the meaningful slice of what spreads while the long tail just drifts. The book’s verdict: “structural, yes; total, no.”

Read the full argument

How it connects


← Meme ecosystem competition · What the filter rewards (4 of 10) · Replicator ontology →

Explore all the ideas · Read the book in order · Open the interactive map