OpenAI ChatGPT and Biased Information in Higher Education

Abstract

Motivated by the appearance of large language models and their sudden societal impacts—both beneficial and harmful, realized and potential—we evaluated several of them with respect to bias in its myriad forms. Bias in machine-learning models refers to their tendencies to make certain decisions more often than expected. This is a result of the text on which they were trained and, in some cases, the result of post-learning human manipulation. In the end, whether it occurs in the real world or in the machine-learning world, bias will always be a subject of discussion and debate. We view that debate as becoming more and more important, given the recent, unprecedented explosion of AI—in particular, OpenAI and its chatbot, ChatGPT—and what it might mean for the future of higher education.

Ähnliche Arbeiten