While existing research indicates the utility of generative artificial intelligence (GenAI) in assisting language learners, how human–AI relationships may condition learners' being and becoming remains underexplored. In this article, we draw on Barad's agential realism to discuss findings of a posthumanist phenomenological study on 110 Chinese students' experiences with GenAI in a L2 writing workshop. Five forms of human–AI relationships are identified from the metaphors that the student participants employed to describe their experiences: user–tool, king–minister, friend–partner, student–teacher, and slave–tyrant. Across these entangled relationships, human and GenAI agents are differentiated in ways that foreground the figure of a ghostly GenAI native speaker, experienced as competent, efficient, and self(‐)less. By contrast, learners' weaknesses are brought forth as deficient, slow, and unreliable. As our analysis reveals, the ghostly GenAI native speaker is rendered possible and intelligible through a range of existing apparatuses, whose enactment has summoned a dark energy, reconfiguring the way in which native‐speakerism exerts influence upon language learners—it reinforces a technical view of L2 writing and excludes learners not from native‐speaker communities but potentially from the sociality of L2 writing as a whole, with foreclosed possibilities for imagining diverse ways of relating to the target language.
No takes yet. Share an insight, caveat, or question.
Bi et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: