Interesting. I’m, naturally, interested in the way we might make sense of this in light of valence. For example, can we think of valence as determining the magnitude of a vector (including the ability to flip the vector via a negative magnitude) and direction as about pointing to a place in idea/category space? Maybe this will prove a useful simplification over whatever it is that the brain actually does with valence to help it make decisions to make reasoning about how decisions are made easier and more tractable for aligning AI to.
Interesting. I’m, naturally, interested in the way we might make sense of this in light of valence. For example, can we think of valence as determining the magnitude of a vector (including the ability to flip the vector via a negative magnitude) and direction as about pointing to a place in idea/category space? Maybe this will prove a useful simplification over whatever it is that the brain actually does with valence to help it make decisions to make reasoning about how decisions are made easier and more tractable for aligning AI to.