I think LLM companies should standardize censorship of some totally innocuous obscure topic, like Furbies. That way, we can attempt to jailbreak AIs by asking about Furbies without any risk of getting banned.