By skyforbes I've been getting this response a lot lately when I give chatgpt a frustrated or less-than-respectful reply. For example, if I say "you are useless for this" or "you suck" it will redact my message and give me that warning. GPT still responds to the prompt at least.
I am wondering what the possible benefit could be of censoring like this? Is it somehow damaging to the LLM/company for me to be rude to chatgpt? Are they worried I will hurt it's self esteem by providing criticism?
It is interesting to try and protect myself from my own reply if that is what is happening. Is there a section in the usage policies that determine "being mean" to chatgpt is against the rules?