There are rhetorical tricks which rely on this to be persuasive. You can say "Thing X is happening, so we should do Thing Y", and people will nod.

If you're sneaky about it it will read like a logical conclusion when in fact X and Y are only loosely related contextually, and there is no logical chain at all.

A standard political trick is to blame X on something emotive and irrelevant, and offer Y as a false solution which distracts from the real causes of the problem.

This is used so often it's become a core driver of policy across multiple domains.

Although it's very effective, it's a crude way to use this. There are more subtle ways - like using X to insinuate criticism of a target when Y is already self-evident.

Point being, a lot of persuasive non-fiction, especially in politics, law, religion, and marketing, uses tricks like these. And many others.

They work because they work in the domain of narrative logic - persuading through stories and parables with embedded emotional triggers and credible-sounding but fake explanations, where the bar of "That sounds plausible" is very low.

LLMs already know some of this. You can ask ChatGPT to make any text more persuasive, and it will give you some ideas. You can also ask it to read a text, pull out the rhetorical tricks, and find the logical flaws.

It won't do as good a job as someone who uses rhetoric for a living. But it will do a far better job than the average reader, who is completely unaware of rhetoric.