ESP, Gender Biases and AI systems: Challenges in a present future
Author(s)
Gianluca Alfano
Date Issued
2024
Type
book
Volume
3
Abstract
The aim of this research paper is to highlight the phenomenon of bias and discrimination in AI systems, particularly in the field of machine translation, through a review of the current literature on the subject and by offering hints and solutions on how best to deal with the problem. The meanings of stereotype and gender bias are analyzed, followed by a description of the various weaknesses and biases found in artificial intelligences and where they may lurk along the production pipeline of an AI.
The dissertation is accompanied with some selected experiments and case studies, with a focus on four AIs adopted in the field of translation, such as ChatGPT, Google Gemini, Google Translate and DeepL.
This paper finally looks at some alternative and innovative proposals to mitigate the phenomenon of bias in AIs, emphasizing the importance of a multidisciplinary approach in this context.