Sociologists face the pressing question of how large language model-based technologies can contribute to social theorising. Several social scientists have called for a greater integration of generative AI tools in the social sciences and humanities to speed and scale up the production of social theory. In this article, we adopt a more cautious approach, reflecting on how the adoption of generative AI raises foundational questions about the construction, refinement and evaluation of critical social theory. To make this argument, we create and critically analyse a speculative piece of dystopian sociological fiction. This enables us to explore the values and practices of critical social theory authorship. We argue that authorship must be ‘recentred’, particularly with regard to four key values of authorial practice: (1) standpoint, (2) creativity, (3) pluralism and (4) accountability. In the present moment, the use of generative AI as substitution for authorial practices redefines and potentially threatens authorship in ways that problematises its particular use in generating critical social theory. With regards to standpoint, generative AI technologies run the risk of reproducing an ‘abstract masculinity’ that feminist scholars have long written against. The relational wonder at the heart of social theory creativity is undermined by generative AI use, as is the pluralism of creativity. Generative AI also flattens and overlooks pluralism of voice, as well as dilutes authorial accountability. We sociologists need to defend and rearticulate what we contribute to social theory and its concomitant politics, both in the theories themselves and in the project of social theorising. It is with this in mind that we call for the ‘rebirth’ - that is, the recentralisation - of the human author in critical social theory creation.
No takes yet. Share an insight, caveat, or question.
Higgins et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: