​24th (RTAIM) Rebuilding Trust in AI Medicine Online Seminar

Maira de Chinque

#24 Maira de Chinque

[Microsoft Teams link | ID Meeting: 3421342035921| Password: yS3BU9rq]

21 Janaury 2026 | 5:00pm - 6:30pm (Lisbon Time Zone)

Linguistic Self-Correction in the Age of AI #Seminar 24: Human linguistic practices have always relied on implicit criteria of correctness governing what counts as acceptable or unacceptable expression. This presentation will examine contemporary, language-based AI systems specifically in their role as instruments of linguistic self-correction, arguing that their widespread adoption makes their normative influence on grammar and usage inevitable. As these systems are increasingly relied upon to correct, predict, and standardize language, they begin to shape not only grammatical practices but also broader judgments of right and wrong in both expression and meaning. Rather than focusing on artificial intelligence as a moral agent or autonomous decision-maker, this discussion approaches AI as a normative tool for grammatical self-correction and examines its influence on judgment practices that extrapolate mere linguistic ones. The analysis ranges from the nature of linguistic practices to the application of AI to it, the possibilities it opens, and the extent to which it somewhat reshapes human action.  The central claim is that language-based AI systems are increasingly positioned to function as criteria of judgment in linguistic practice. Through mechanisms such as autocorrection and predictive text, they generate and reinforce norms of correctness and acceptability in language. A concrete illustration is found in practices of self-diagnosis, where individuals privilege AI outputs over professional medical advice, resisting correction even when expert judgment is available. What is ultimately at stake is not merely the adoption of a new tool, but maybe a transformation of the grammatical and normative background against which meaning, judgment, and communication are formed. 

Short bio: Maira de Cinque is a Brazilian-Canadian philosopher, Assistant Professor at the University of Canada West and Part-time Studies Instructor at BCIT. Her work spans logic, moral philosophy, and philosophy of language, with research grounded in Wittgenstein and Rousseau. Her current interests focus on linguistic normativity and the role of language-based artificial intelligence.