A machine, not an intelligence: how Chinese programmers imagine generative artificial intelligence

Imaginaries have become an essential analytic lens for understanding how emerging technologies like generative artificial intelligence (AI) are envisioned. While existing studies have begun to explore these collective visions, they remain predominantly focused on a macro-level perspective, emphasizing national and public imaginaries, while overlooking those shaped by key professional groups. This study addresses this gap by examining how programmers, who occupy an intermediary role between generative AI and society, imagine this transformative technology. Drawing on insights from the Social Construction of Technology (SCOT) and the concept of imaginaries, we adopt a mixed-methods approach combining computational text analysis of online forum posts with a national survey of 14,445 Chinese programmers. Online discussions reveal two imaginaries: “AI-as-Machine,” which frames AI as a powerful yet controllable tool, and “AI-as-Intelligence,” which casts it as an autonomous and potentially disruptive agent. Survey evidence further confirms the dominance of the “AI-as-Machine” imaginaries, particularly among technically proficient programmers. By centering programmers’ pivotal position, this study not only grounds imaginaries in concrete metaphors, but offers a micro-foundation for understanding China’s rapid, tool-driven adoption of generative AI.