Research on the visibility of social Bias in ChatGPT’s responses and interactions
Xin Li
University of Sydney
DOI: https://doi.org/10.59429/esta.v11i4.8470
Keywords: ChatGPT; Social bias; Gender bias; Bias detection; Data privacy
Abstract
The article examines the phenomenon of social biases manifested by ChatGPT in its responses and assesses the impact of these biases on user experience and social perceptions. The study identifies potential sources of social bias in ChatGPT’s interactions with users, with a particular focus on gender, racial, cultural, and political biases. The study indicates that the biases exhibited by ChatGPT may originate from constraints in its training data and inherent biases in user input, which can influence the objectivity and reliability of the generated content. Furthermore, the article examines the manner in which ChatGPT’s biases are perceived and discussed by users on social media, and proposes potential solutions, including the reduction of bias through strict regulation and enhanced model training.
References
[1] Bartoletti, I. (2023, 9:32 PM · Mar 19, 2023). Retrieved
[2] Gault, M. (18 January 2023, 1:00am). Conservatives Are Panicking About AI Bias, Think ChatGPT Has Gone ‘Woke’. Retrieved from https://www.vice.com/en/article/93a4qe/conservatives-panicking-about-ai_xfffe_bias-years-too-late-think-chatgpt-has-gone-woke
[3] Li, L., Fan, L., Atreja, S., & Hemphill, L. (2023). “ HOT” ChatGPT: The promise of ChatGPT in detecting and discriminating hateful, offensive, and toxic comments on social media. arXiv preprint arXiv:2304.10619.
[4] Motoki, F., Pinho Neto, V., & Rodrigues, V. (2023). More human than human: measuring ChatGPT political bias. Public choice. doi:10.1007/s11127-023-01097-2
[5] Patton, D. U., Landau, A. Y., & Mathiyazhagan, S. (2023). ChatGPT for Social Work Science: Ethical Challenges and Opportunities. Journal of the Society for Social Work and Research, 14(3).
[6] Paul, Wu, C., Morency, L.-P., & Salakhutdinov, R. (2021). Towards Understanding and Mitigating Social Biases in Language Models. arXiv pre-print server. doi:None
[7] arxiv:2106.13219
[8] Prates, M. O. R., Avelar, P. H., & Lamb, L. C. (2020). Assessing gender bias in machine translation: a case study with Google Translate. Neural Computing and Applications, 32(10), 6363-6381. doi:10.1007/s00521-019-04144-6
[9] Ray, P. P. (2023). ChatGPT: A comprehensive review on background, applications, key challenges, bias, ethics, limitations and future scope. Internet of Things and Cyber-Physical Systems.
[10] Rozado, D. (2023). The political biases of chatgpt. Social Sciences, 12(3), 148.
[11] Sebastian, G. (2023). Exploring Ethical Implications of ChatGPT and Other AI Chatbots and Regulation of Disinformation Propagation. Available at SSRN 4461801.
[12] Singh, S. (2023). Is ChatGPT Biased? A Review.
[13] Urchs, S., Thurner, V., Aßenmacher, M., Heumann, C., & Thiemichen, S. (2023). How Prevalent is Gender Bias in ChatGPT?--Exploring German and English ChatGPT Responses. arXiv preprint arXiv:2310.03031.
[14] Wahbeh, A., Al-Ramahi, M. A., El-Gayar, O., El Noshokaty, A., & Nasralah, T. (2023). Perception of Bias in ChatGPT: Analysis of Social Media Data.
[15] Wen, Z., & Younes, R. ChatGPT vs Media Bias: A Comparative Study of GPT-3.5 and Fine-tuned Language Models.
[16] Zhou, J., Müller, H., Holzinger, A., & Chen, F. (2023). Ethical ChatGPT: Concerns, Challenges, and Commandments. In. Ithaca: Cornell University Library, arXiv.org.