In previous chapters (links are below) I discussed some concepts which ought to be implemented in a CA model of robot intelligence. Some, like emotion or empathy seem difficult to handle, nevertheless their representations may be relatively simple. By representation I mean a mapping from a phenomenon to a CA system. Its adequacy will be judged whether system behavior meets our expectations.
An intelligent CA system has the following attributes:
1. It consists of at least two interacting processes (CA)
2. The system is dynamic, always seeking an attractor.
2a. The system’s configuration at an attractor is called the system’s solution.
2b. The system’s purpose (teleonomy) is to create an optimal solution. In other words of all the potential solutions the system faces it will create the most optimal.
3. The stem-process controls the system and when necessary adds to it transitional processes. It lives as long as the system does, and when it dies the system dies as well.
3a Transitional process is created by the stem-process and survives less than the stem-process does.
4. Each process state is an action memory which serves as initial condition for:
4a The next process state.
4b Transitional processes at the time of their creation by the stem-process.
5 The behavior of the system is driven by external (environmental) change.
5a Sensation: Is an external change that perturbs the system and modifies its action memory.
5b The system “understands” the external message (perturbation) only when creating a new solution.
5c The new solution is the memory of the event (message) which perturbed the system.
5d In other words if the system returns to its pre-perturbation solution, it did not understand the message and forgot it.
6 Orientation memory: Is an action memory modified by an external (environmental) change.
7. Emotion: Is a change in system’s behavior initiated by its present state.