4
comments
This is one of the reasons I am comfortable with chatting with LLMs (I interact with real people too, don't worry)
Same, both points. The 3rd time it gave me a garbage answer, I could ask for what phrasing mistakes or assumptions I may be making leading to the bad output.
Doesn't make the faces my boss does when asked the same thing.
The LLM doesn't know what lead to a specific input either, it can only write a response guessing about what could cause it.
Ouch.