This Algorithm Writes Fake Artist Statements That Sound Real
Entertainment By Elena Boaghi | November 19, 2017
You walk into a gallery and see a video installation on the wall. It’s called Movement, but you have no idea why. You look to the small plaque on the wall for an explanation, and read the following: “That ontology can in turn only be interpreted adequately under the guidance of the meaning of movement.”
There’s a good reason this artist statement is so inscrutable. It was generated by a machine learning algorithm to sound just like the barely decipherable art-speak that can usually be found on the walls of museums and in artist statements.
In fact, this computer-generated statement is itself an artwork called Variable by the Instanbul-based artist Selcuk Artut. At the press of a button, Variable generates a new artist’s statement for one of eight different screens. Each screen chooses a title for the art piece, like “movement,” “presence,” or “weakness,” and then each statement is concocted with a machine learning algorithm that was trained on the book Being and Time by the philosopher Martin Heidegger. Artut writes that he was inspired by the book’s complexity; his algorithm remixes Heidegger’s text and transforms it into new meditations on the nature of being, an appropriate topic for contemporary art, which so often ponders the same questions. All the statements sound appropriately vague and baffling–just like actual artist statements.
While the work is supposedly a critique on the limits of algorithms and a chance to “reflect on poststructuralism’s ontological questions,” it actually serves far better as a critique of how obscure and opaque the art world can be. After all, while the statements it generates are slightly off grammatically, they still convey the incomprehensible jargon artists tend to use when describing their work.
But if you let your imagination wander enough, perhaps you can glimpse the meaning of time and being through the algorithmic haze.KS