Dynamic Reconfiguring of GPT-4 Based Tutors to Become GPT-4 Based Teachers in Underserved Areas in Africa and the Environs

Abstract

The popular GPT-4 API boasts excellent academic records of passing a number of standardized tests. As such it is used in a number of tutoring systems providing students and pupils with access to artificially intelligent tutors when help is not easily available. Tutoring and teaching, however, are different. In many underserved areas in Africa and the environs, students and pupils may attend formal school but then after hours no tutors are nearby to assist them with their studies and homework. Previous research by the authors has described mobile tutoring artifacts linking the popular Whatsapp mobile chat app with GPT-4 to provide that tutoring. In other underserved areas (especially in war torn areas), however, there may not be even teachers to provide the initial classroom instruction or Zoom based instruction. In such case these mobile tutoring artifacts need to be dynamically reconfigured to act as full teachers to these students. This paper describes the research to dynamically reconfigure GPT-4 based mobile tutors to become GPT-4 based mobile teachers wheGPT-4, chatGPT, Whatsapp, Tutoring, Java Programmingn necessary.