I’m not sure I buy the analogy. In a forum, which presumably is moderated not censored in your example, it’s an exchange between people with the intent in keepi

I’m not sure I buy the analogy. In a forum, which presumably is moderated not censored in your example, it’s an exchange between people with the intent in keeping it civil and on topic. You have many humans attempting to collaborate and the moderators job is to ensure the humans don’t wander off into human inanity and to protect both the discourse and the participants. An AI model is talking with myself, about subjects I want to discuss in private with a second brain. If I want to talk to myself about erotica, why can’t I? If I want to yell obscenities at myself, why can’t I? I can understand why Microsoft or Google don’t want to give a chat bot that does these things, or offer dangerous medical advice, or whatever, but if it’s my own device and a model I’m running myself, why on earth would I want a censored model? I.e., I see why the service provider wants to give me a censored model, but do I want a censored model? Why would I?