Artificial Life or Alife is the study of natural processes by the creation of artificial systems designed by humans according to the principles observed in nature.
The purpose of experimenting with simple models of naturally occurring systems depends on your perspective...
Natural systems display capabilities which are not only intriguing and highly complex but also eminently useful, (especially for the organisms concerned). Where a human-designed system requires similar capabilities, it is wise to look at nature for inspiration.
If simple enough models of the natural phenomenon can be made to exhibit similar properties, then we can transfer the principles of nature into our own designs to achieve system behaviours which are hard to design by brute force. In essence, the evolutionary process represents billions of years of parallel experimentation into the efficiency of different system designs. By seeking inspiration from nature, we can raid this massive store of pre-optimised designs without having to repeat the process of trial and error ourselves. Fields which have already benefited from this approach include decentralised computing, pattern recognition and automated design.
Some people hope to derive results in the natural sciences, e.g. biology, sociology, game theory. In other words they hope to discover things about the naturally occurring systems in their own right. Chris Langton, to whom the term 'Artificial Life' is attributed, was largely focused on this approach in his original exhortation for people to adopt Alife experimentation.
An example of artificial experimentation leading to results in the natural sciences can be found in Paleontology; the phenomenon of punctuated equilibria found in evolutionary development:
The dating of the fossil record indicates that the transitions of organisms from one form to another does not take place through continuous and gradual modification. Instead, after long periods of little or no change to an organism's form, novel forms emerge suddenly with no apparent warning. This phenomenon is also observed in processes of artificial evolution, in which candidate designs or solutions to a problem are retained or rejected according to a selective regime similar to that found in nature - the fittest survive. Consequently the phenomenon of punctuated equilibria can be examined in an entirely artificial environment, and various hypotheses about the phenomenon can be explored through experimentation examined without even having to touch a living thing.
The natural systems which are explored in artifical life approaches are very diverse.
In some cases, capabilities of individual organisms are examined, such as the developmental process of multicellular organisms. In other cases, dynamics of cooperative organisms are explored, such as the self-organising and self-regulating processes observed in colonies of social insects. Equally the competitive dynamics such as evolution can be exploited for useful properties. The technological exploitation of the Artificial life approach is not restricted to the direct transfer of the designs found in nature. Higher level results from the natural sciences, for example in information theory, graph/network theory or economics, can also be exploited to identify approaches to problems in technology, and establish novel approaches for self-organising, self-regulating and adaptive systems.