Biological living systems in general exhibit complex and diverse dynamics. The latter, in particular, is essential, since diversification increases the odds of survival of an organism while reducing the risk of extinction of the population. Primarily, diversification is a consequence of the randomness in the replication process of a biological cell, which eventually manifests into a motley set of macroscopic features of an individual. These heterogeneous features of individuals constitutes for diversity in population. Cancer is a prime example of such a complex system where the transformed cells exhibit plethora of disparate features, which in turn makes modeling their dynamics quite challenging. In this paper we consider cancer as a prototype of a complex living system and provide two contrasting perspective for studying and modeling it. Based on this we illicit a deeper role of diversification in the evolution of cancer. Following this, we ask ourselves how can we model these diverse dynamics in a multiscale setting. We address the shortcoming of the existing multiscale modeling technique by providing an abstract but mathematically rigorous framework for deducing stochastic evolution equations at the macroscopic level starting from a microscopic description of the involved dynamics. We achieve this by making use of the connection between stochastic process and the semigroup operator generated by them. In particular, we look at the semigroups generated by Levy processes and their connection with the characteristic functions and Levy symbols. The latter turns out to represent pseudo-differential operators using which we eventually provide a mechanism for constructing stochastic evolution equations. Altogether, this provides the framework for modeling diverse dynamics at the macroscale starting from the microscale.