Simulation of Intelligence Evolution in Object-Oriented Systems

In classical artificial intelligence and machine learning fields, the aim is to teach a certain program to find the most convenient and efficient way of solving a particular problem. However, these approaches are not suitable for simulating the evolution of human intelligence, since intelligence is a dynamically changing, volatile behavior, which greatly depends on the environment an agent is exposed to. In this paper, we present several models of what should be considered, when trying to simulate the evolution of intelligence of agents within a given environment. We explain several types of entropies, and introduce a dominant function model. By unifying these models, we explain how and why our ideas can be formally detailed and implemented using object-oriented technologies. The difference between our approach and that described in other papers also - approaching evolution from the point of view of entropies - is that our approach focuses on a general system, modern implementation solutions, and extended models for each component in the system.