Those of us who are careful internet readers have spent years developing good heuristics to use textual clues to tell us about the person behind the text. Are they smart? Are they sincere? Are they honest? Are they commenting in good faith? Those skills will soon be obsolete.
The folks at OpenAI, who are nominally on a mission to make sure AI "benefits all of humanity", have condemned us to a life sentence of fending off high-volume, high-quality bullshit. Bullshit that they are actively working to make harder to detect. And I think the first victims of that will be internet forums where text is the main signal, places like this and Reddit.