The question is if the prompts are deterministic and if the information you want to transport is exactly what the prompt will produce. I actually doubt this will work, as you never know what the text will be every single time the LLM produces it. Still I find it better than producing slop and spamming forums, emails and more with it.
IMO it is mandatory to read and control any AI generated output and always remember to:
THINK - PUSH - TALK
I understand the goal of justsendtheprompt to explicitly not involve a LLM in the first place, like a reverse lmgtfy, maybe we are in this mess because people with questions often get told âjust google it yourselfââŚ
Donât produce more noise, just ask the question directly and let the human you are communicating with decide what to do with it, thatâs exactly what I would want 
3 Likes
Exactly, thatâs why AI texts are always a pain to read. âTL;DRâ (âToo Long; Didnât Readâ) was invented for them⌠
Humans are (usually) trained for efficiency, in this case meaning transmit the most information with the least effort (words). AI is trained to âsound goodâ, i.e. creating beautifully constructed and formatted phrases regardless of how much useful information they might contain. Instead of summarizing, they expand. Form over function. 
2 Likes