No References: the ChatGPT from a latourian perspective and the Double Click trap

Authors

DOI:

https://doi.org/10.48075/rfc.v25i41.30761

Abstract

Generative Artificial Intelligence has sparked considerable interest among the population, with ChatGPT and its wide media coverage being a notable example of this phenomenon. However, the development of Artificial Intelligence models does not occur without facing challenges and limitations. One of the main hurdles encountered in ChatGPT is the scarcity of reliable references, which can result in inaccurate or even false information. The rapid growth in the use of generative AI can exacerbate such a problem, which, from the latourian perspective, can lead to the Double Click trap. This paper focuses on the problem of sources and references, which ChatGPT insists on not providing or, in some cases, even falsifying. As the use of generative AI becomes more and more widespread, the issue of lack of sources and references may become even more serious, given that users may consider ChatGPT an infallible oracle whose reliability may grow by force of habit.

Author Biographies

José Victor Rodrigues Catalano, Universidade Federal de São Carlos

Mestrando no programa de Ciência, Tecnologia e Sociedade da Universidade Federal de São Carlos (UFSCar). Possui graduação em Administração (2018) pela Universidade Paulista (UNIP).

Bruno Rossi Lorenzi, Universidade Federal de São Carlos

Pós-doutorando no programa de Ciência, Tecnologia e Sociedade da Universidade Federal de São Carlos (UFSCar). Possui graduação em Ciências Sociais (2009), mestrado em Ciência, Tecnologia e Sociedade (2012) e doutorado em Ciência Politica (2018) pela Universidade Federal de São Carlos (UFSCar).

Published

26-05-2023

How to Cite

RODRIGUES CATALANO, J. V.; ROSSI LORENZI, B. No References: the ChatGPT from a latourian perspective and the Double Click trap. Revista Faz Ciência, [S. l.], v. 25, n. 41, 2023. DOI: 10.48075/rfc.v25i41.30761. Disponível em: https://e-revista.unioeste.br/index.php/fazciencia/article/view/30761. Acesso em: 10 jun. 2025.

Issue

Section

Artigos