Sine hominum? Artificial intelligence and the concept of authorship

Abstract

In the recent years, we have been witnessing dynamic development of artificial intelligence. More and more technologically advanced machines, which thanks to the implemented software imitate human cognitive processes, have already become able to generate works undistinguishable to the audience from human-made ones. The unprecedented degree of autonomy characteristic for the software based on artificial intelligence, which makes it able, to improve its functioning on its own, by using deep learning, seems to marginalize the role of human as a creator. This paper attempts to answer the question whether the advancement of the aforementioned intelligent systems endangers the concept of authorship, generally recognized as lying in human domain only. First part of this article focuses on the attempt to explain the term „author” in the context of copyright. Secondly, I present the basic operational principles of the systems based on artificial intelligence, highlighting the differences between them and the previously existing systems. In the next part of this paper, I examine AI’s generative activity in terms of its similarities and differences to the human creative activity. Basing on the results of said examination, I consider whether the emergence of AI’s works can induce a change in the current perception of an author as a human being.