AI and the Transformation of Scientific Knowledge

Abstract

This article proposes a rethinking of the role of artificial intelligence in contemporary science through an analysis of Harry Collins’s concept of expertise. The point of departure is the observed tension between normative accounts of expertise – grounded in ideas of socialisation, tacit knowledge and membership in a form of life – and the empirical practices in which AI systems increasingly perform epistemically significant functions. The article contends that traditional deficit-oriented critiques of AI, which focus on the absence of human properties, are losing their explanatory power in the context of the widespread adoption of large language models. The article further argues that with the integration of AI, expertise ceases to be primarily an individual property and increasingly emerges as a feature of hybrid socio-technical systems, whose epistemic capacities arise from the interaction between human and non-human agents. In conclusion, the paper demonstrates the need to reconfigure the anthropocentric conceptual apparatus of science studies to enhance its explanatory adequacy. It proposes a shift in research focus toward the empirical analysis of AI’s actual use in scientific practice and the reconstruction of the emergent normative orders within these systems.