This is very ironic. The LLM output samples I have seen from OpenAI and Facebook are horribly censored as well. Article on HN a week or two ago highlighted how FB's model refused to answer questions about the Bible citations or about politicians.

try `ollama pull llama2-uncensored`

https://github.com/jmorganca/ollama

it's not completely "uncensored", but you avoid at least some of the silliness of the standard model.