You used to be able to detect bots by their short, lazy, often poor english messages.
Now itβs the opposite.
If anyone on the internet types anything longer than the length of a tweet, and it has flawless grammar, you automatically assume it was written by chatgpt.