Menneske og drømmemaskine. Posthumane myteskabelser om AI hos Amalie Smith

Abstract

Summary In the present article, I analyse the process of making the artwork Machine Learning I II III (2018) by Danish artist Amalie Smith to examine how it conceptualizes a posthuman myth about computer vision and AI. Smith’s artwork investigates the phenomenon of computer vision through using a convolutional neural network to represent what this network is thought to “see”. The artwork is an aesthetic manifestation of the invisible operations of the machine learning algorithm in a form that is visible to humans, and thereby it engages aesthetic speculations about how computer vision works. The artist provides an insight into how machine learning algorithms interpret images in a way that is radically different from humans, but at the same time greatly affects human reality because of the algorithmic culture that permeates contemporary societies. I read Machine Learning I II III with posthuman thinkers Rosi Braidotti, Donna Haraway and N. Katherine Hayles to show how Smith imagines a co-creative relationship between human and technology that neglects a myth about “machine” and “human” as distinct and isolated categories; a myth where symbols, human and algorithmic intelligence, weaving, and written discourse intertwine to make the artwork. In this way, Machine Learning I II III moves towards a posthuman myth of computer vision and AI where it is impossible to unentangle human and technological forces.