You're not wrong, but the suspicion is that the person used AI (ChatGPT or similar LLM) to write answers to the reporters' questions. No evidence presented for

You're not wrong, but the suspicion is that the person used AI (ChatGPT or similar LLM) to write answers to the reporters' questions. No evidence presented for that, though, so while I could well believe it I think your criticism is valid. I do think that the BS-as-a-service aspect of LLM's makes it harder for many people to tell if they're talking to an expert. The optimistic scenario is that this will eventually cause people to be more suspicious of trusting any online source they don't otherwise know anything about.