AI MUSIC
Interviews
AI-generated music is a fascinating, never-ending subject. Coeckelbergh asks a pertinent question: “if a computer is said to compose music, is the machine really ‘creative’? And, is the product really art?” (Coeckelbergh, 2017, 286). If the machines themselves are created by humans, and are programmed by humans with different algorithms, is there any room for machine creativity? Coeckelbergh justly notes that the algorithm, not the human, is the artistic agent: the human creates the code that programs the machines, but the end product is not made by a human, which is what terrifies everyone. Indeed “it seems that the creativity is no longer […] in the programmer but has migrated to the technology.” And this is especially the case for Artificial Intelligence, as when the machine has the capacity to learn, “the process cannot simply be reduced to the execution of a code written by humans” and machines enter a sphere that had always been reserved for humans. (Coeckelbergh, 2017, 286)
In The Turing test and artistic creativity (2010, 409), Boden argues that for an program to pass the Turing Test, it would need to be able to produce an artwork that would be (1) indistinguishable from one generated by a human (2) seen as having as much “aesthetic value as one produced by a human being”. Coeckelbergh summarizes the two opposing views on what it means to artistically create. The first one is an expressivist view, implying that when one creates, one express something from his ‘inner self’, his ‘authentic self’, as if creativity was related “to something ‘in’ the person.” (Coeckelbergh, 2017, 290). According to this view, machines can’t produce any artistic creation as they lack consciousness, a “true origin” and imagination. “Humans, by contrast, can be original, authentic and so on, and in works of art, they can express this originality and authenticity.” (290) The second view is the pre-modern one, that assumes that one ought to imitate (mimesis) as well as one can one's environment and nature. In this second view, machines can be included in the domain of artistic work, and could in fact surpass many humans.
For Coeckelbergh, the problem of machine creativity is not much different than the problem of human creativity: according to him, “the machine faces the same challenge as the human artist: the challenge of creating a product that qualifies as art. […] Machines can in principle create art, this does not always mean that (1) it actually happens that art is created and (2) that it is good art. Like human artists, some machine artists may be better than others and some works of art they create are better than others.” (293)
Of course, we always come back to the question of whether or not creative machines should mimic human creativity or not (as a process or/and as a finished product). I believe, as does Coeckelbergh, that there are other, non-human forms of creativity. The acceptance of this concept will be, in my view, the complete post-anthropocentric turn: the human won’t be the measure of all things and the judge of all things. Coeckelbergh summarizes this very well: “the humanness, by itself does not make it art.” (298) Furthermore both art/craft and technology can be seen as “forms of poiesis, as ways of revealing and bringing into being.” (299)
One reason humans are reluctant to associate machines and art is because I believe that there has been an aura created around the concept of “creativity”. But as Boden notes, “creativity is not magical. It is an aspect of normal human intelligence, not a special faculty granted to a tiny elite. There are three forms: combinational, exploratory, and transformational. All three can be modeled by AI—in some cases, with impressive results.” (Quote by Cope 2015, 309). For data-driven AI, the process works like this. Here is a proof of the need to de-deify the aura surrounding creativity. This allow us to further complete the post-anthropocentric turn, as humans’ capacities are de-mystified and “de-exceptionalized.”
