Navigating platformized generative AI: Examining early adopters’ experiences through the lens of data reflectivity

Abstract

Following the widely noticed launch of ChatGPT 3.5 in November 2022, an unprecedented number of users have started to experiment with generative AI for communication purposes. Prior studies have shown how users are commodified by platforms, and the unprecedented development of generative AI raises hence once again questions of platform dynamics vs user agency. In this study, we argue that platformized generative AI (GenAI) actively ‘talks back’ to their users, prompting them to act accordingly. Theoretically, we develop the concept of data reflectivity as a critical lens, showing that users exhibit reflective practices. These allow them to reflect upon their own role in relation to platformized GenAI and thus alter their patterns of action. The empirical case study, carried out in Spring 2023, draws on a survey with 60 early adopters and 14 subsequent semi-structured interviews, collected in an NGO operating in Southeast Asia. The thematic analysis shows that users relate to platformized GenAI in three distinct but related ways: (1) as a happy helper to organize and systematize knowledge and information; (2) as a creative tool to generate ideas; and (3) as a conversation partner for personal and life-related matters. In conclusion, we discuss the findings critically in relation to overall platform dynamics and the notion of systems ‘speaking back’ and further suggest that future research should aim to bring the two research fields of datafication and user studies even closer together.