Hermeneutic rationality as a critique of algorithmic clinical practice

Abstract

This article problematizes the contemporary concept of explainability in Generative Artificial Intelligence from a hermeneutical reading of medical rationality. Based on Husserl's diagnosis of the crisis in European sciences, it argues that modernity has established an apodictic and mathematizing reason, whose influence extends to medicine through the biomedical paradigm and, more recently, algorithmic clinical practice. Explainability, by claiming transparency and predictability as epistemic ideals, reiterates the modern project of a technical rationality that reduces meaning to formal causality. In contrast, it proposes hermeneutical rationality as an alternative model of explanation, founded on understanding, language, and intersubjective justification. Inspired by Gadamer, Cassirer, and Aristotle, the article maintains that medical practice is essentially interpretive and prudential (phronesis), requiring a form of intelligibility that is not confused with technical traceability, but with discursive responsibility. It can be concluded that hermeneutic explainability constitutes, simultaneously, an epistemology and an ethics of care: to understand is to respond to the other, and to explain is to become responsible for the intelligibility of the clinical encounter within an authentically human rationality.