some of this is honestly explicitly virtuous, engineering-wise, in that i'm much, much more likely to live at a wider local minima for longer before specializing if i don't have to look at it. like during early experimental phases, my work is _better_ for me looking at it less
TypeError"having to" i think maps to a misunderstanding. i don't hand off stuff i don't know how to do, i hand off stuff that i don't feel like doing because i can make a computer do it at 10x speed, meaning i can try more approaches for ~the same design effort