In this chapter, the author explores the impact of generative AI on academic writing from the perspective of equity and diversity. The current focus of AI development is explored, and its premises are analyzed. The author highlights the significantly skewed nature of the training corpus used to develop AI models and argues that a biased corpus would only generate biased answers. Moreover, the author argues that although generative AI has the potential to reduce the time and cost associated with language learning as well as the conventions of academic writing, using AI to overcome gatekeeping, such as peer reviews, may encourage its users to conform to the norm and standards, resulting in further biased research output. Additionally, the current neoliberal climate may put competitive pressure on students and researchers to increase productivity by using generative AI. The resulting normativity centering on the WEIRD (Western, Educated, Industrialized, Rich, and Democratic) cultures and voices would threaten the diversity of human knowledge production. Teachers and researchers should revisit the purpose of writing and its role in education, as the voice of AI could be extremely biased against L2 English writers, non-WEIRD cultures, and the marginalized, ultimately generating responses which lack benefit for humanity.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Equity and Diversity in Writing: What Does AI Generated Normativity Mean for Second Language Writers?

  • Kei Kawanishi

摘要

In this chapter, the author explores the impact of generative AI on academic writing from the perspective of equity and diversity. The current focus of AI development is explored, and its premises are analyzed. The author highlights the significantly skewed nature of the training corpus used to develop AI models and argues that a biased corpus would only generate biased answers. Moreover, the author argues that although generative AI has the potential to reduce the time and cost associated with language learning as well as the conventions of academic writing, using AI to overcome gatekeeping, such as peer reviews, may encourage its users to conform to the norm and standards, resulting in further biased research output. Additionally, the current neoliberal climate may put competitive pressure on students and researchers to increase productivity by using generative AI. The resulting normativity centering on the WEIRD (Western, Educated, Industrialized, Rich, and Democratic) cultures and voices would threaten the diversity of human knowledge production. Teachers and researchers should revisit the purpose of writing and its role in education, as the voice of AI could be extremely biased against L2 English writers, non-WEIRD cultures, and the marginalized, ultimately generating responses which lack benefit for humanity.