This is really interesting. At this point you seem to be modelling real human memory

In my opinion, this should happen inside the LLM dorectly. Trying to scaffold it on top of the next token predictor isnt going to be fruitful enough. It wont get us the robot butlers we need.

But obviously thays really hard. That needs proper ML research, not primpt engineering

But do they not need a warrant to track someones movements?

Alot of these tech vendors have been a way to launder data gathering to avoid neeidng to get warrants