This content is locked. Please login or become a member.
One of the most interesting ways to use AI to make sure that you’re being radically respectful is to practice conversations that you are worried about. You’re going to have a conversation with someone. You’re worried that what you’re saying might be misrepresented or misunderstood. Practice this conversation with an AI.
Everybody hates role-plays with other people, but one of the things that I have found is that people don’t mind so much practicing conversations with AI. You can ask the AI, “I’m planning to say this.” You can describe that person, and the AI will explain to you where you might be going wrong.
However, I think that it can be pretty dangerous to use an off-the-shelf LLM to role-play a conversation, because often the AIs, the ChatGPTs are going to reflect and reinforce biases or they’re going to overcorrect. And we’ve seen both of these things happen. So I really recommend using an AI that’s been specially designed to recognize, to be aware of biases, and to help you avoid them.