AI and Art. Arguments for Practice

Abstract

Recent advances in the computer generation of pictures using methods and programs from artificial intelligence research, or, more precisely, machine learning, have once again raised the question of whether computers can make art. Based on A. Michael Noll’s early experiments with computer art from the 1960s, I argue by contrast that even the latest tools cannot do without human work and can only be part of an artistic practice thanks to this work. Rather than asking whether machines can make art, we should therefore develop creative practices in which it is possible to leverage the potential of new techniques for design and art.