ChatGPT in Education: Risks to Fairness of Access

Abstract

In the rapidly changing ChatGPT world, human skills training is the need of the hour. With a large amount of data, ChatGPT is training itself day by day to understand human prompts and to provide appropriate results. A lot of prompt engineering courses educate learners in structuring the prompts based on how the ChatGPT engine processes any prompt and response. However, these courses do not target specific cognitive skills required by users (learners) to unfold, understand, and express what they really need and want to know from ChatGPT. In this paper, we look at ChatGPT as a tool to support knowledge acquisition and we discuss the questioning skill as an essential cognitive skill required to interact with and make optimal use of ChatGPT capabilities. We argue that risk to the fairness of access does not just stem from the monetary availability of such technologies, but it can also arise from the unpreparedness of the target users. The AI and ethics of AI literacy communities should also focus on training individuals on the cognitive skills needed to become optimal users of such technologies.