Chatbots are an interesting topic. While I think they have their uses, they also have significant limitations, as well as ways they can be abused.
Useful things they can do include stuff like creating first drafts of email newsletters and even blog posts. But in such cases, the end result needs to be tidied up by a human before being used. I'm not sure what the text equivalent of the "uncanny valley" is, but posts written by chatbots somehow seem stilted and cold to me. I can't put my finger on it exactly, but they're just not quite "right".
I've seen far too many AI generated spam messages (particularly contact form spam), and sadly spam filters are losing that particular arms race, at least for now.
However, a post I saw elsewhere today was interesting. It seems Google can already identify bot-written product descriptions (mainly because the bots just steal and regurgitate whatever they can find online), and have started flagging and de-ranking them as plagiarised content. Personally, I think this is a good thing !
It's not really hard to spot them as the content created by such bots is so vague.