Comment by amypetrik8

> should be held as if the companies have vetted the content and it is theirs.

It's not "as if" it "is". There is a scant difference between an LLM and a recommendation algorithm picking what to "say" out of what, 100 billion or more messages? Because the pool to choose from is so enormous, speech becomes not what one person typed but what one algorithm plucked out of the haystack to show, to influence, and to manipulate for financial gain.