A STUDY OF SELECTED CHATGPT-GENERATED CONTENTS IN EDUCATIONAL DISCOURSE FROM NIGERIA AND GHANA: A FUNCTIONAL AND CRITICAL PERSPECTIVE

Abstract

Artificial Intelligence (henceforth, AI) has become a huge part of
academic culture used to generate contents for academic purposes.
However, while educational institutions take responsibility for the
content of AI technologies adopted for service delivery, no one is held
accountable for AI/ChatGPT’s content when students use ChatGPT
for academic purposes. In this study, ChatGPT educational discourse
was examined using Mood analysis and van Dijk’s triangulated
Critical Discourse Analysis (CDA) framework of discourse, cognition
and society. The objective was to determine how AI/ChatGPT
educational discourse was constructed. In a world driven by racial and
class ideologies of dominance and the psychological warfare, the study
specifically queried “whose opinion ChatGPT pushed in anonymity
when it responded to questions from students”? The data constitute
fifty (50) purposively selected ChatGPT-generated discourse, taken
from Nigerian and Ghanaian undergraduates, using Mood analysis
and Critical Discourse Analysis. It was discovered that ChatGPT is
highly opinionated, authoritative and speciously anonymous and
could be deployed as a hegemonic agent designed to perpetuate
subsisting narrative of western supremacy over others. The paper
concluded that such anonymity corrupts educational discourse and
threatens educational achievements of students, especially those from
Third World countries, who have gravitated towards AI technologies.
It is thus suggested that policy makers must be deliberate and
proactive about instituting a healthy control of ChatGPT’s influence
in educational discourses.