AI and the scientist: On the fracture of epistemic authority

The figure of the scientist has long occupied a distinctive place in the epistemic order. From the Enlightenment onward, the scientist was not merely a skilled technician but a disciplined inquirer – a producer and guardian of knowledge – whose legitimacy rested on epistemic virtues such as curiosity, integrity, and humility, as well as on empirical rigour. Today, however, this position appears less secure. The systems we now refer to as ‘artificial intelligence’, and more concretely large language models (LLMs), produce outputs which, in certain domains, rival human work.