Detail zooming in Artificial Intelligence World Modeling

For an Artificial Intelligence to operate it must apply a model of the physical (and subjective) environment in which it is to operate. We call this a World Model. In human intelligence, it is called imagination. This paper addresses an essential feature of such modeling - variable detail definitions. Like zooming in graphics, detail must be omitted to process the big picture. Every object in the model is a type which may be made up of less complex objects and may be a part of a more complex object. Detail zooming imposes new rules for object classification and processing. For some kinds of processing, high detail is required for some of the objects that compose a particular instance of a type and low detail for other objects composing that type. When we declare an object in a modern programming language we acquire all the functionality of that class of object. Detail Zooming imposes a variation in the extent of the functionality of a class.