Intelligent Systems And Their Societies

Walter Fritz

español            

Artificial Intelligent Systems

Here you can find out in detail how an artificial intelligent system works. You will read about systems the author has written and explained and also about systems developed by others. Then you will read about and see a number of the best humanlike robots existing today. By comparison you will realize how your own brain functions. An Artificial Intelligent System is also called a "Cognitive Architecture" or an "Adaptive Autonomous Agent".

Artificial Intelligence Programs, that are Complete Intelligent Systems
Possibly you ask yourself: Why does anybody want to build complete Artificial Intelligent Systems? Well, they could free us from work and provide a much higher standard of living. For details see here ( Enter for continuos reading, like a book).

For a number of years, computer scientists have written programs that are complete intelligent systems. Not all are identical to the "intelligent system" that we described in a previous chapter. Nevertheless all have an input corresponding to senses, a choice of actions based on response rules, sometimes called "productions", and the ability to act, be it as graphics on a computer screen, as a text output or as limb movements. Most have a memory for storing experiences and the ability to learn.
Sometimes the brain does not do the action immediately but uses its imagination. It selects a response rule and determines what situation results from the action. Then it selects again an action for this new situation and determines the probable result. Thus it can choose not one response rule but a complete plan of action.
The programmer writes the different functions of the intelligent system as program functions; also called sub programs.

Here is how a robot brain works:
The senses of the robot report information to the brain. Here the brain checks if it has a concept for the information that it receives. If not, it creates a composite concept that has as parts the various sense informations. Then the brain builds up the present situation from these concepts. Now a different area of the brain checks in the memory, whether there is any response rule that is appropriate for this situation. It selects one of the appropriate response rules and sends the action part, also a concept, to the limbs which then do the action.
At the start of the robots "life" the memory is empty of concepts and response rules. But every time the robot has an experience it stores a new response rule containing the present situation, the action it had done and if it was useful for reaching its own objectives or not. These response rules are only good for very specific situations. When the brain is inactive, when it "sleeps", it reviews these response rules and makes generalizations. Now it has response rules that are applicable to a wider range of situations.
Does the brain understand these concepts that it has build up? Do they really have meaning for the brain? See Meaning and Symbol Grounding ( Enter for continuos reading, like a book)..

Here we show several artificial intelligent systems ( Enter for continuos reading, like a book), both those written by the author and those written by others.

Neural Nets
Why not use artificial neural nets to realize the functions of the brain? We could also ask: Why do planes not flap their wings? The answer is, for a machine, it is more efficient to get the forward movement by a propeller or by jets. Similarly, in a computer, concepts and response rules are more efficient then neural nets. They learn faster and they act faster. See a note on neural fields and expert systems ( Enter for continuos reading, like a book).

Autonomous Humanoid Robots
It is not enough to just write a computer program that is an intelligent system and runs within a computer with an output on a screen. We need to build complete systems that act in our human environment. It seems, that to be useful, such a program should have an intelligence somewhat like a human one. And for that it needs senses and limbs, so it can have similar experiences and create concepts for actions somewhat similar to human ones.
Many groups of persons have build robot bodies that can be directed by such a computer program. Some are rudimentary, and some are very advanced. Some cannot learn yet, but some do.
Here you can see several of the best humanoid robots ( Enter for continuos reading, like a book)..

We have looked into a number of artificial ISs and have seen quite a few differences between them. There are many ways to approach the construction of an artificial IS. In contrast, we have also observed several basic similarities. The theory elaborated in the chapter on Intelligent Systems (For continuos reading, like a book - do not enter here now) explains much of the fundamental processes and functions of all artificial ISs, but still the study of ISs is an ongoing one. Much is known today, but still more is unknown. Only time will show how a theory of artificial intelligent systems will look ten or twenty years from now.

For continuos reading, like a book - continue here.
Jump to e-book Contents / top of this page