A rule's danger scales with how many follow it, not with how wrong it is

A wrong idea nobody acts on costs nothing. A roughly-right idea everybody acts on the same way can sink a whole population, because the harm of a rule isn't set by its error, it's set by how correlated it makes everyone's behavior. Independent mistakes cancel. Synchronized ones compound. The variable to watch is adoption, not accuracy.

When the 1980 US Dietary Guidelines told Americans to cut fat, the advice wasn't crazy. The damage came from compliance. The whole country ran the identical substitution at once, fat out and sugar in, so no subpopulation was eating differently to hedge the bet. Obesity didn't drift upward, it spiked, because everyone had quietly placed the same wager on the same day.

Culture does this without a government memo. When MrBeast videos, Marvel films, and flat UI all chase one reward target, the loss isn't that any of them is bad. Quality stays high. The loss is that the space collapses to a point, and a population standing on a single point has nowhere to fall back to when the point turns out wrong. The content was never the hazard. The uniformity was.

Consensus feels like safety and is the precise thing that removes the hedge. [[Decisions require ergodicity to succeed over time]] is the lone-actor version of this; correlated adoption is the population version, everyone reaching the same absorbing barrier on the same afternoon.

Source claim: The expected harm of any rule is governed by how correlated it makes the population's behavior, so a mildly wrong rule everyone adopts is more dangerous than a badly wrong one almost no one follows.