AI Development for the Public Interest: From Abstraction Traps to Sociotechnical Risks

TL;DR

This paper examines the historical development of sociotechnical perspectives in AI subfields and proposes a unified pedagogical approach to integrate social context understanding into AI graduate education.

Contribution

It provides a comparative analysis of AI Safety, Fair ML, and HIL Autonomy, and offers a roadmap for sociotechnical graduate pedagogy in AI.

Findings

Different AI subfields have unique sociotechnical histories influencing their social integration.

Perceptions of Public Interest Technology vary across AI disciplines based on past risks.

A unified pedagogical framework can enhance sociotechnical understanding in AI education.

Abstract

Despite interest in communicating ethical problems and social contexts within the undergraduate curriculum to advance Public Interest Technology (PIT) goals, interventions at the graduate level remain largely unexplored. This may be due to the conflicting ways through which distinct Artificial Intelligence (AI) research tracks conceive of their interface with social contexts. In this paper we track the historical emergence of sociotechnical inquiry in three distinct subfields of AI research: AI Safety, Fair Machine Learning (Fair ML) and Human-in-the-Loop (HIL) Autonomy. We show that for each subfield, perceptions of PIT stem from the particular dangers faced by past integration of technical systems within a normative social order. We further interrogate how these histories dictate the response of each subfield to conceptual traps, as defined in the Science and Technology Studies literature. Finally, through a comparative analysis of these currently siloed fields, we present a roadmap for a unified approach to sociotechnical graduate pedagogy in AI.