Do Content Warnings count towards generated response?

[shoopuff570]

I am using chat gpt for help writing a song. I get a few content warnings but I am using those to better mold my dark subject song into something that is more reasonable. My question is, does the AI bot read the content warnings and “remember” them? like ill push for a bad word, get a content warning, then ask it for a better word, but it seems to forget what the theme is. ill ask it for “Fk" in the context of "Oh Fk, Ive stubbed my toe” but it wants to “build a better future”. “I have stubbed my toe, ill build a better chair.”