I suspect a new class of spam is emerging that is strategically placed positively or negatively sentimented text that is designed to be ingested by other LLMs about a certain topic. This spam doesn't need to be convincing to humans, just convincing to a mindless LLM that's giving an answer on the topic at some later date.

It is more likely a new type of joke, whereby the poster pretends to be a really bad chat robot, perhaps implying that the article itself could as well have been produced by an (equally mindless) AI system.

It certainly rings like one, you're not alone in thinking that