If you read the actual article there's barely any evidence that any of what they are claiming is even remotely true. They talk about vague connections through certain words being used on YouTube that are, in their own words, inconclusive. And a bunch of anecdotal instances on reddit in which mods use "vibes" to detect AI slop comments and posts. And then finish with more anecdotes about some real world encounters that they think are written by AI.

I mean, no doubt that AI garbage is filtering into online discourse because let's face it, people are lazy assholes who want easy karma and updoots. But this is hardly evidence that actual conversational language is being altered by AI.