Another contentious aspect of the razor is that a theory can become more complex in terms of its structure (or syntax), while its ontology (or semantics) becomes simpler, or vice versa. Quine, in a discussion on definition, referred to these two perspectives as "economy of practical expression" and "economy in grammar and vocabulary", respectively.
Within AI, are there some examples that can help understand where the above occurs or where this may be of importance?
I have tried to form a possible example:
Suppose an AI software program has a high cyclomatic complexity. This would mean it has high syntactical complexity (is this correct?). Further, suppose the program operates at a high abstraction level then its semantic complexity could be low (again, is this correct?).