Here's the existentialism I was looking for :) I'm not sure how we can verify that empirically. Current LLM AI will hallucinate about the feelings it "has" and

Here's the existentialism I was looking for :) I'm not sure how we can verify that empirically. Current LLM AI will hallucinate about the feelings it "has" and argue why it is a person, but only because it was trained on human language. The more human we make something seem, the more we blur the line until we might have a day where people upload their consciousness permanently without knowing if "they" will wake up inside the computer or just some copy of themselves that answers all the "don't kill me, I'm a real person" questions will wake up in the computer. Human software is still intrinsically tied to human hardware, so this seems like it would be far easier to create zombies than true humans, even if the checksum matches after an interstellar transmission.