The fact that a robotic hand can delicately manipulate an apple is impressive, of course, but the probability that this is a universal specialization is very low, say p(specialization_universal | apple) < 0.15. We should see this more as a functional adaptation to a specific use case. For example, if we wanted to pick damaged fruits without tearing them, the robot's capacity would be judged on its adaptability flexibility to each fruit, not on a supposed general hierarchy.