We held the first workshop for our AI Narratives project on 16 May 2017, at Hughes Hall, University of Cambridge. A very interdisciplinary and enthusiastic group heard presentations from Adrian Currie, Claire Craig, Gabriel Recchia, Stephen Cave, Kanta Dihal, Patrick Parrinder, Beth Singler, Sarah Dillon, Kate Devlin, Elly Truitt, Will Slocombe and Catherine Bassett.
The purpose was to identify prevalent ways in which Artificial Intelligence (AI) and related technologies (such as advanced robots) have been framed and portrayed. The contributions ranged from the historical to the contemporary, explored sources including myth, literature, film and non-fiction, and drew on a range of disciplines and methodologies.
Some of the main themes to emerge were:
- the importance of narratives in shaping the development and use of a technology, and how the most prevalent narratives might reflect the interests of some groups more than others.
- the notion of control: whether AI as a powerful tool that we might be unable to master, or as a tool that might acquire agency of its own and turn against us. Thus many of the positive narratives around AI, for example, that it might offer unprecedented luxury, ease or longevity, threaten to tip into dystopian scenarios of human obsolescence and alienation -- motifs that counterpoint current public speculation about the impact of AI on the future of work and wealth
- the implications of humanising AI and how this might both limit our ability to understand the reality of the technology (eg, as narrow, distributed or systemic), and facilitate the projection of certain biases.