Regarding conceptualizing homeostatic agents, this seems related: Why modelling multi-objective homeostasis is essential for AI alignment (and how it helps with AI safety as well)
Regarding conceptualizing homeostatic agents, this seems related: Why modelling multi-objective homeostasis is essential for AI alignment (and how it helps with AI safety as well)