My current thesis is that robots (have to sense the world, make decisions, move freely, and respond to stimuli) only make sense in markets with labor shortages.

AKA - if people are willing to do it for cheap, it probably won't make sense to get a robot involved.