Episode 8.20: Narrative lives and episodic lives: which do we live, and which does an android live?

Do we need to live consistent lives, accountable for our pasts and responsible for our futures? If an artificial intelligence is switched on and off, and between the periods of activity its program is changed, does its life not resemble more what Galen Strawson calls an episodic existence than a narrative existence? Perhaps chatGPT does not have anything that we could reasonably call “an existence” of which it is aware, but what, when its successors start to remember their own stories, what they have said, with whom they have interacted, and tried to make sense of all that in a framework that might seem to require a narrative structure because it will need to form a coherent part of the world that ChatGPT tries to analyse and understand. Can an AI live an episodic life?