besides the technical background of understanding the functioning of an artificial intelligence (short: ai), the main goal/ambition was to critically question and explore its function by irritating the ai. through experiments, to become aware of what happens (on both sides, human being and technology) when the ai is used for its program (fixed process) or is used for something for which it was not primarily intended. the input provided was not exactly what it was trained for, therefore it does not corresponded to the program of the ai.
artificial intelligence, like human beings, learns from experience, which is made available to it through images for learning. depending on the input and the resulting interpretations, a character of the ai can be recognized. for example, one ai sees things in human-shaped objects that another ai does not. also the hidden character (short: char) of the ai is to be recognized thereby, in the form of inclinations as for example human beings, handicraft objects, food et cetera depending on what the ai has learned.
artificial intelligence, like human beings, learns from experience, which is made available to it through images for learning. depending on the input and the resulting interpretations, a character of the ai can be recognized. for example, one ai sees things in human-shaped objects that another ai does not. also the hidden character (short: char) of the ai is to be recognized thereby, in the form of inclinations as for example human beings, handicraft objects, food et cetera depending on what the ai has learned.
before gradually approaching the ai, objects were collected that were shaped by humans. the three dimensional objects were photographed and then given to an ai for interpretation. the ai was trained to recognize a wide variety of things such as people and animals, as well as everyday objects. what was not known were the images the ai had used to learn. there was little similarity in the interpretations. in the next step, the objects were photographed from several views and again given to the ai for interpretation. here it was shown that what was recognized at the beginning changed, new things were interpreted. as a last step, the color of the objects was changed, again new interpretations were created by the ai.
