<![CDATA[This article aims to clarify and clarify the dual interpretations of the term' deep learning' in the context of mathematics education in the era of artificial intelligence. The term is often reduced to merely an AI-based technology that relies on the internet. In contrast, in the pedagogical domain, deep learning refers to a learning approach that emphasizes deep conceptual understanding, connections between ideas, and the transfer of knowledge to new situations. This study adopts a conceptual review approach based on literature analysis, using secondary sources such as journal articles, books, and policy reports published between 2000 and 2024. The findings show that deep learning technology holds potential to support mathematics learning through features such as handwriting recognition, automated evaluation systems, intelligent tutoring, and adaptive learning. However, the implementation of this technology also faces serious challenges, including limitations in contextual data availability, uneven digital infrastructure, the opaque nature of model interpretation, and issues of ethics and data privacy. On the other hand, the pedagogical approach to deep learning places the teacher as the main actor in designing meaningful learning experiences. Therefore, the integration of technology and pedagogy must be carried out critically and contextually. Educational innovations in the AI era must remain grounded in humanistic principles and an awareness of students' sociocultural realities—especially in the diverse context of Indonesia.]]>
Copyrights © 2025
