upvote
Thanks, I find it very interesting as well. I think very many people would assume they must be interacting with another person, and I don't think there's really a way to _prove_ it's not that, just through conversation. But we do have a lot of mechanisms for understanding how others think through conversation only, and so I think the approach of having a clinical psychiatrist interact with the model make sense.
reply