Seeing the Matrix, Switching Abstractions, and Missing Moods — LessWrong