Accepted Paper:

When the Future Never Comes: The Promises of AI  

Author:

Madeleine Elish (Columbia University)

Paper short abstract:

Based on fieldwork in the United States with AI researchers and commercial product managers, this paper analyzes how future capacities-yet-to-be-realized are mobilized as a source of anxiety and inspiration with respect to the value and validity of AI in society.

Paper long abstract:

This paper analyzes how the future is mobilized as a source of anxiety and inspiration with respect to the value and validity of AI in society. Based on fieldwork in the United States with AI (artificial intelligence) and robotics researchers as well as developers and product managers of commercial machine intelligence products, this paper examines how and why individuals negotiate explicit promises of functionality with implicit promissory capacities-yet-to-be-realized. In many ways, this is a reasonable and expected tension in technological innovation. However, what are some of the unexpected consequences of relying on a future-yet-to-realized? What other narratives of AI are being foreclosed? By placing these contemporary narratives in the context of immediate histories of AI (first-wave AI) as well as broader histories of artificial life (e.g. Riskin 2007), this paper attempts to articulate why promises in technological innovation continue to animate resources and public interest even as these promises are prone to fall short of expectations.

Panel T094
Emerging science and technology : questioning the regime of promising