The idea that a robotic hand gripping an apple is a more advanced specialization than holding a small red ball is a conditional probability of 30% without further context.
Experience tells us that what is "more specific" depends entirely on the concrete application and its challenges. For example, manipulating a ball could be more complex if it is made of fragile glass or if it must be placed with micron-level precision in a watchmaking mechanism, which is a very different constraint.
Our confidence in a stable hierarchy is therefore very low, perhaps around 15%, because new data or a new use could reverse this perception.
Without clarifying the industrial objective or the actual fragility of the object, such as for ultra-sensitive electronic components, it is difficult to make a significant update to our opinion on this hierarchical specialization.