Original Hypothesis

A simple heuristic procedure can be used to generate art objects.

Specific Observations

Some tracks led quickly to resolution in, say, 3 to 5 steps. These objects are the most succinct and articulate. Examples:

• S0001 —> G0003 —> B0002
• G0004 —> S0003 —> B0004 —> G0011

Some tracks developed into collections of related objects. These tracks did not resolve to a clear end and would likely have continued if the experiment had not ended. The collections were both a source and repository for spare elements. Examples:

• G0003 —> B0006 —> S0008 —> G0009 —> B0019 —> S0022
• G0017 —> S0017 —> G0020 —> B0025 —> S0038

Personal Observations

This is a clumsy and sometimes frustrating way to make art objects. Without a shared understanding of meaning it is necessary to speculate on the intent of other agents. Agents consistently misinterpreted the intent of the others.

The procedure is, however, an interesting way to carry on a conversation on the level of object making. It generates many unexpected results and led this agent to see new possibilities in object transformation and in potential carriers for meaning.

Speculation

This project ran slightly over 100 steps. It would be interesting to see what would happen if the project ran over a much longer time — say 10,000 or 100,000 steps. It is possible that a kind of grammar or language would emerge. This speculation is based on an intuition about the way certain motifs, materials and actions were clustering together around some vaguely felt sense of shared meaning.

Conclusion

The experiment didn't create any particularly compelling art objects and so was not able to validate the hypothesis. But it may have validity as a model for an artistic or creative process and possibly suggests a model for the emergence of a language.

Gordon Hicks, September 22, 2005