My intuition is right with you. They will actively avoid anything that the end user thinks is a camera while packing it full of sensors that allow interactions

My intuition is right with you. They will actively avoid anything that the end user thinks is a camera while packing it full of sensors that allow interactions with (mapping/object recognition of/etc) the world around them.