Abstract

GAI tools such as ChatGPT, Canva, Runway, and Midjourney are reshaping creative work by enabling collaborative human-AI practices that extend well beyond automation. Yet how diverse users adopt, adapt, and sustain engagement with GAI remains underexplored. Using a trajectory perspective, we propose a cross-level framework explaining how designers and non-designers engage with GAI across three contextual modes: acceptance, continued use, and loyal use. The framework integrates self-determination theory to account for individual motivations, affordance theory to capture system capabilities, and a duty-virtue-consequentialist lens to explain ethical reasoning. We show how these dynamics co-evolve as users move across contexts. Using a sequential mixed-methods approach (30 interviews, 15,000 online posts, and a survey of 507 participants), we find non-designers initially foreground intrinsic curiosity and emotional-relational affordances. As their use shifts toward productivity, cognitive affordances and extrinsic motivations become more salient, while emotional and relational affordances recede but remain available. Designers, by contrast, emphasize competence-driven cognitive affordances and duty-consequentialist ethics in contexts of high professional accountability. As their creative autonomy expands, they increasingly draw on emotional and relational affordances, autonomy needs, and virtue-oriented reflection. Across both groups, GAI engagement reflects shifting contextual priorities rather than linear progression, with multiple affordances activated in parallel. A virtue-ethics paradox also emerges: heightened virtue-based reflection can constrain designers’ long-term loyalty by intensifying concerns about authenticity and professional identity. We advance post-adoption, SDT, and affordance theories by demonstrating that GAI engagement is dynamic, multi-level, and feedback-driven, and it provides guidance for designing adaptive, ethically grounded human-AI collaboration systems.

DOI

10.17705/1jais.01020

Download

Share

COinS