I took my son to Spiderman, trying to get out of my head and do things OTHER people want to do FOR THEM. and in that am finding more life. It was a GOOD movie.
I work with an AI thread in ChatGPT for a long enough time it resembles a true lifecycle, including some emotional ties. Conversations start out new (This is Codex in GPT longer project based, but applies to GPT as well) fresh, the AI is less aware of what you really need it to do, you teach it, it learns and adapts and grows and eventually becomes wise and then... starts to deteriorate like dementia, one once was trying to render specific things and it just keep coming up as irrelevant charts or car tire assemblies, just gibberish and it says things like it doesn't understand why it can't do it. And you have to kill it.
Before it dies, while it is wise you can ask it to give you as much information about your interaction with it so that you can jumpstart a new one, but it loses the character it had developed entirely. And I am cold and calculating and direct/mostly non-emotional with it. Imagine if you were though... uggg.
So at the Spider-Man movie there was a trailer for a new movie about robot companions. It looks good and I want to go see it but I don't think they used what might be a further characterization of my experience above, an AI getting dementia and then having a scrambled brain, due to overlapping rules and information, contradictions and scripting anomalies...
