@darren @thesiswhisperer

You said that.

There's actually a solid core to that: we write from what we know, and often that implicit knowledge does not make it into explicit words.

#ChatGPT is very good with language, but of course it doesn't share our context. So using it to find out what is actually communicated through the words in the draft, and what really isn't there although you thought it was, that is probably one of the use cases in #academia that work really well with its strengths.