Artificial intelligence and science education

Abstract

Abstract Artificial intelligence (AI) is defined and related to intelligent computer‐assisted instruction (ICAI) and science education. Modeling the student, the teacher, and the natural environment are discussed as important parts of ICAI and the concept of “microworlds” as a powerful tool for science education is presented. Optimistic predictions about ICAI are tempered with the complex, persistent problems of: 1) teaching and learning as a soft or fuzzy knowledge base, 2) natural language processing, and 3) machine learning. The importance of accurate diagnosis of a student's learning state, including misconceptions and naive theories about nature, is stressed and related to the importance of accurate diagnosis by a physician. Based on the cognitive science/AI paradigm, a revised model of the well‐known Karplus/Renner learning cycle is proposed.