This paper demonstrates a novel, literature-based approach to the consideration and regulation of advanced computational and data systems, in particular artificial intelligence (AI). The paper contributes uniquely to the discourse of AI through its proposal that although we may think of AI as presenting a distinctly contemporary challenge, in fact, its roots can only be explored by starting with the discourses, narratives and social imaginaries of the long nineteenth century. To illustrate this, the paper looks at one of the most widely recognised cultural threats associated with big data and AI systems, namely the threat to privacy. It explores how the concept of privacy might be viewed, and ultimately managed, differently through the application of an historical and literary frame of reference, deploying a mixed-methods close and distant reading of novels in English and German. For this, we focus on two specific literary exempla, each of which illustrates the function of privacy in society. This close reading is complemented with exploration of experimental data generated by using a modern privacy dictionary to query a large corpus of nineteenth-century novels and a modern test corpus comprised of policy discussions targeting privacy issues related to the rise of big data and AI.
Read full abstract