Integrating Kalama Sutta, Critical Thinking, and Learning by Reflective Thinking: A Conceptual Framework for AI Literacy in Education with Critical Judgment of LLMs

Abstract

This academic article presents a conceptual framework for critical literacy of Large Language Models (LLMs) in the context of education. A review of research, literature, and international reports indicates that LLMs play a significant role in learning, access to knowledge, and the creation of personalized learning effectively. However, their use still faces challenges such as bias, lack of transparency, and the risk of believing information without verification. To promote critical use of LLMs, this article integrates three essential concepts: the Kalama Sutta, which emphasizes not believing without reason; critical thinking, as a core skill in analyzing and evaluating information; and reflective Learning, which emphasizes reviewing and developing from direct experience. These have led to the proposed framework of “Critical Literacy of LLMs,” consisting of three steps: (1) Filtering Belief (2) Analyzing and Evaluating (3) Reflecting and Developing This new body of knowledge is not intended to limit the potential of LLMs but supports their use as “educational tools” aligned with social values, enhancing human capacity in a holistic manner, and contributing to the development of education to be responsive to the age of artificial intelligence.