Epistemologi Artificial Intelligence

Kebenaran, Validitas, dan Otoritas Algoritmik

Abstract

The development of artificial intelligence (AI) has brought fundamental changes in the way knowledge is produced, validated, and accepted in various sectors of life. Algorithmic models, especially deep learning, generate predictions and recommendations that are often treated as operational truths even though the inference process is not fully explainable. This study analyzes how AI changes the understanding of truth, validity, and epistemic authority from the perspective of the philosophy of science, and links it to the ontological and axiological dimensions in modern knowledge production. A qualitative approach based on philosophical analysis is used to integrate the thoughts of Popper, Kuhn, Lakatos, Van Fraassen, and Floridi. The results show that AI shifts knowledge from rational justification to performative and statistical validity, and challenges the position of humans as the primary epistemic agents. This study asserts that the epistemic transformation triggered by AI requires ontological and axiological reflection so that the development of knowledge remains in line with humanitarian principles and ethical responsibility