<紀要論文>
ChatGPT and Anti-White Bias

作成者
本文言語
出版者
発行日
収録物名
開始ページ
終了ページ
出版タイプ
アクセス権
JaLC DOI
概要 ChatGPT is a free, extremely popular artificial intelligence (AI) chatbot capable of answering questions, providing explanations, and creating texts such as computer code – or even entire university e...ssays – based on simple queries from the user. While portraying itself as neutral, it also includes disclaimers that it may occasionally provide biased content. Since its launch, the English-speaking media and Internet have been awash with reports that ChatGPT is “racist,” and these accounts are, almost without exception, focused on how non-Whites are misrepresented and maligned by the AI. However, at present, there has been little media attention given to – and no academic studies examining – ChatGPT’s treatment of White people, despite anecdotal evidence of racial double standards and anti-White bias in the Anglophone media and Internet in general. In the present qualitative study, a series of single-question “Q&A sessions” on racial identity and racial self-advocacy were administered to ChatGPT in parallel, with the only difference being the substitution of four different racial groups: Blacks, Asians, Latinos, and Whites. Comparative analysis of the responses to five salient questions seemed to clearly demonstrate biases against Whites and racial double standards, at least for the current context. This report provides a cautionary tale for university educators likely to encounter students considering the use of ChatGPT for essay and research paper assignments.続きを見る

本文ファイル

pdf 51_p053 pdf 3.23 MB 2,518  

詳細

PISSN
NCID
レコードID
査読有無
タイプ
登録日 2023.11.30
更新日 2024.12.24

この資料を見た人はこんな資料も見ています