In this talk, scholar Hannes Bajohr provides an overview of notions of authorship in AI and natural language processing systems, and discusses past and current debates about computers as literary authors. He suggests the concept of causal authorship to measure the types of distance between human and machine agents, acknowledging the anthropocentric bias of this idea. To balance this, he reflects on the notion of distributed authorship, which considers the network of actors involved in the creation of a text and has its own limitations. Both concepts are elements of a future theory of authorship in the age of machine learning.
00:00:00 Introduction | John Cayley
00:06:16 “Writing at a Distance: Notes on Authorship and AI” | Hannes Bajohr
00:52:24 Readership | Q&A
00:59:09 Context | Q&A
01:02:35 Alienation | Q&A
01:05:45 Human Intervention | Q&A
01:09:44 What AI Should Not Do | Q&A
01:11:11 Muzak | Q&A
01:15:50 Authorship | Q&A
01:19:15 Authenticity | Q&A
01:22:52 “(Berlin, Miami)” & The Impact of AI on Literature | Q&A
Recorded on March 7, 2025. Presented by the Cogut Institute for the Humanities as part of the Collaborative Humanities Seminar “Reading the Large Language Models: Artificial Intelligence, Language, and Literary Art,” led by John Cayley and Ellie Pavlick, with the “Models-Scale-Context: AI and the Humanities” Lab.