Abstract

The provocation offered by this position paper is that a new perspective on the design of tools based on generative AI should be adopted, one that tackles at once the problems of the social impact of a new revolution in work and of the "mechanization of cognitive tasks". This perspective should be based on two tenets: (i) design should be human-centred (Shneiderman, 2020), going beyond the human-centric perspective endorsed by the AI Act; and (ii) design should be values-based (Friedman & Hendry, 2019), going beyond the typical tension between utility and usability.

While human-centric suggests that technology should ultimately serve human goals, human-centred is a process for designing tools that places human practices and goals at the center of conceptualization and implementation. In this process, alongside understanding users and their contexts in terms of (tasks’) needs and impacts, ethical and eudaimonic values have to be fully considered.

Share

COinS