Expectations And Expertise: Can STS And AI Benefit from each other?

• In Event: Metaexpertise? STS, Metascience, and Data Science - I: The Technologies of Meta-expertise

Abstract

During my doctoral research, investigating empirically and historically the interplay of expectations and expertise in the hype cycles of artificial intelligence (AI), I encountered the following research obstacle: conducting early literature review on the aforementioned topics, instead of STS articles investigating the social role of promise and legitimacy in shaping AI, scholarly search engines I used (ironically, AI-powered) returned technical articles about predictive software systems or knowledge-based expert systems. Often, articles about expert systems involved the question of prediction (as in medical diagnosis), which allowed me to reflect on the interplay between expertise and expectation, and how the problem of legitimacy influences not only decision making but also foresight strategies. Therefore, I came to observe patterns of similarity between the discipline I investigated (AI) and the discipline through which I investigated (STS). In this first public presentation of this observation, by focusing on the cognitive objects of “expertise” and “expectation,” I suggest that the contemporary maturity of AI and STS, reveals a common aim at apprehension and coding of knowledge structures; the former by the use of symbolic representations and statistical reasoning, and the latter by devising metaphors and narratives, inviting a novel alliance between the technical and the social.

Author

• Vassilis Galanos, University of Edinburgh