Abstract
The article was created for the Cognition and Artificial Life XX conference. It is dedicated to the reflection of the development of computer, or generative art, which from the beginning tried to release the original strict determinism of computer programs into framework-determined, but highly unpredictable results.
This raises the question of whether and since when one can talk about intelligence in these cases, or about intelligent artistic decision-making. Despite all the randomness, the results are limited in advance, even to unfathomably large files. Even the procedures that claim to be artificial intelligence – be it traditional symbolic modeling (Harold Cohen's AARON) or the use of artificial neural networks (Obvious group projects) – do not fulfill the imperative of openness of art as a system.
It seems, however, that projects using the principles of artificial life can by their very nature work with increasingly complex rules and increasingly complicated structure, while maintaining internal consistency despite all the openness.