In an attempt to create a theory for describing such a human phenomenon as a possibility to self-learning, we need to create an instrument for dynamic modeling of sense-to-sense (S2S) [3] associations betweenheterogeneous objects. This instrument would help understand the nature of the cause-to-effect relationship [4] and the creation of new knowledge.In this article, we describe one of the instruments, sense derivative, that sheds light on the nature of forming new knowledge in the field of Artificial Intelligence.