Key points are not available for this paper at this time.
This article posits artificial intelligence (AI), as applied in instruction, is a more efficient means to old ends and that newer ends can be thought of. In particular, it is argued on the basis of some past research and on the basis of a Vygotskian view, that intelligent computer tools can not only simulate human cognition but, given specific conditions, humans can simulate computer's intelligence. That is, learners can internalize computers' intelligent tools and use them as cognitive ones. Internalization is discussed in terms of cognitive reconstruction of cultural artifacts of particular characteristics — tool-like nature, relative novelty of function, compatibility with learners' schemata, and most importantly — explicitness of operation. The article also discusses learning conditions that might facilitate the process of tool internalization, particularly “high road” learning whereby learners are intentionally mindful of a tool's mode of operation and logic. A preliminary study is briefly described showing that children are capable of internalizing the metacognitive guidance provided by a semi-intelligent Reading Aid and that they show evidence of internalizing the tool's “intelligence” thereby manifesting improved reading comprehension as well as improved essay writing. Four fundamental questions are raised for future research as well as questions of ethics.
Gavriel Salomon (Sun,) studied this question.