Propaganda in Plain Sight
How Word Salad Masquerades as Expertise
We expect propaganda to look like propaganda—bold claims, obvious spin, clear self-interest. But the most effective kind doesn’t. I was just reading a book about AI, which was clearly intended to scare you, but the content is nothing but 500 pages of a word salad, with no logical argument that would stand a chance against reason. That’s a terrible, terrible of information to convince the world of untruths. Let me tell you more.
The most effective propaganda arrives dressed as serious analysis from credible sources, dense with information, yet built on a foundation of loose connections and missing logic. The result is a text that sounds authoritative while proving nothing. Engineers and domain experts can spot the gaps immediately, meanwhile everyone else walks away convinced they’ve just absorbed a profound piece of literature.
No, they don’t understand almost anything of course, but the title of the author, the references, the tone, it all makes it sound as if they did.
Consider the straightforward way a logical argument actually works.
If A leads to B, and B leads to C, and C leads to D, then D follows. The chain is visible, the conclusion is falsifiable, and anyone can check whether each step holds. When the chain is broken or absent, the honest response is to say so.
Propagandists do the opposite. They flood the page with loosely related facts, historical analogies, technical details, and speculative scenarios until the sheer volume creates an illusion of rigor. The conclusion appears at the end as if the preceding pages have earned it, when in fact they have only exhausted the reader.
This book I was reading was arguing that artificial intelligence is catastrophically dangerous and must be heavily regulated follows this pattern exactly.
It piles on information from multiple fields, mixes established facts with unproven assumptions, and arrives at its policy recommendation without ever demonstrating that the proposed dangers are probable rather than merely possible. Readers who work in AI see the missing logical steps and dismiss the argument. Readers who don’t see a thick volume written by a university professor and conclude that the author must understand something they themselves cannot follow. The book succeeds not because its reasoning is sound, but because its presentation flatters the reader’s sense of their own limitations.
The same technique works in reverse—using apparent expertise to make the possible seem impossible.
When Elon Musk argues that humans can establish a self-sustaining presence on Mars, the engineering case is straightforward: the laws of physics do not forbid it, the necessary technologies exist in principle, and incremental progress is already underway. The task is extraordinarily difficult, but difficulty is not impossibility. Yet critics respond with opinion pieces that highlight narrow technical challenges, funding uncertainties, or regulatory hurdles.
These points are real, but they are presented as decisive objections rather than engineering problems to be solved. Non-experts read the critiques and think, “The experts have identified reasons it cannot be done,” when what they’ve actually seen is a list of reasons it will be hard. The illusion of impossibility is manufactured not by disproving feasibility, but by burying the distinction between “difficult” and “impossible” under layers of authoritative-sounding detail.
In both cases the mechanism is the same. A lack of logical connection is disguised as depth of analysis. Volume substitutes for rigor. Credentials substitute for clarity. The reader who cannot immediately see the flaw assumes the flaw does not exist, or that only specialists can evaluate it. This is not accidental. It is a reliable way to move public opinion without ever having to win an argument on the merits.
The defense is simple but demanding: insist on seeing the chain. When someone claims that a conclusion follows from a body of information, ask exactly how each piece of information supports the next step. If the answer is hand-waving, speculation, or “trust me, I’m an expert,” the argument has not been made. Difficulty is not impossibility. Correlation is not causation. And a thick book full of words is still just a thick book full of words until the logic is laid bare.
The most critical piece of spotting propaganda? You must truly want to notice it.