Reactively and Anticipatory Behaving Agents for Artificial Life Simulations.

Reactive behavior is still considered and the exact opposite for the anticipatory one. Despite the advances on the field of anticipation there are little thoughts on relation with the reactive behavior, the similarities and where the boundary is. In this article we will present our viewpoint and we will try to show that reactive and anticipatory behavior can be combined. This is the basic ground of our unified theory for anticipatory behavior architecture. We still miss such compact theory, which would integrate multiple aspects of anticipation. My multi-level anticipatory behavior approach is based on the current understanding of anticipation from both the artificial intelligence and biology point of view. As part of the explanation we will also elaborate on the topic of weak and strong artificial life. Anticipation is not matter of a single mechanism in a living organism. It was noted already that it happens on many different levels even in the very simple creatures. What we consider to be important for our work and what is our original though is that it happens even without voluntary control. We believe that this is novelty though for the anticipation theory. Naturally research of anticipation was in the beginning of this decade focused on the anticipatory principles bringing advances on the field itself. This allowed us to build on those, look at them from higher perspective, and use not one but multiple levels of anticipation in a creature design. This presents second original though and that is composition of the agent architecture that has anticipation built in almost every function. In this article we will focus only on first two levels within the 8-factor anticipation framework. We will introduce them as defined categories of anticipation and describe them from theory and implementation algorithm point of view. We will also present an experiment conducted, however this experiment serves more as explanatory example. These first two levels may seem trivial, but growing interest in implicit anticipation and reactive anticipation in last two years suggest that this is worthwhile topic. The simplicity of them is intentional to demonstrate our way of thinking and understanding of the topic. The comparison with other results is possible only when these levels are integrated into the complex architecture, as will be shown below that can also act as standalone. [ABSTRACT FROM AUTHOR]