FROM GPT 3 TO CHATGPT: POTENTIALITIES AND WARNINGS IN THE FOCUS OF BRAZILIAN ACADEMIC PRODUCTION

Main Article Content

Meire Nadja Meira de Souza
Paulo Vinícius Pereira de Lima
Karla Vanessa Gomes dos Santos
Carlos Lopes

Abstract

This exploratory study aims to analyze the approach of Artificial Intelligence (AI) in the construction of academic texts based on versions of the GPT - Generative Pre-trained Transformer language model, such as GPT 3 and ChatGPT, in the period 2021-2023. The analyses were carried out according to the following "a priori" categories: a) research approaches; b) research methodology; c) production centers and regions of these productions; and d) potentialities and warnings. A bibliographic review was adopted as the procedure and a systematic review was used as the technique, in accordance with the PRISMA method (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). The search was carried out on texts published in the Brazilian Library of Theses and Dissertations and articles on Google Scholar. A total of 93 articles, theses and dissertations were retrieved and 13 were included in the review. The results of the study indicate a predominance of research with a qualitative approach, varied methodologies and a greater concentration of publications in the Southeast of Brazil. Ideas associated with "warnings" are more prevalent than those linked to the potential of generative AI for algorithmic writing. We conclude that in the current Brazilian scenario (2023), studies on generative AI for writing will tend towards a progressive volume of interest in the subject due to experiences, conceptions and questions about its use in various areas of knowledge. Among these interests, in correlation with the results of this study, will be the ethical aspects and the concept of author and authorship associated with the topic of generative AI for writing. Clarity and critical reflection on the incorporation of AI are fundamental to guiding and outlining its role in research and knowledge production.

Article Details

How to Cite
SOUZA, M. N. M. de .; LIMA, P. V. P. de .; SANTOS, K. V. G. dos; LOPES, C. FROM GPT 3 TO CHATGPT: POTENTIALITIES AND WARNINGS IN THE FOCUS OF BRAZILIAN ACADEMIC PRODUCTION : . Conjuncture Bulletin (BOCA), Boa Vista, v. 16, n. 47, p. 599–620, 2023. DOI: 10.5281/zenodo.10222348. Disponível em: https://revista.ioles.com.br/boca/index.php/revista/article/view/2732. Acesso em: 18 nov. 2025.
Section
Articles
Author Biography

Carlos Lopes, University of Brasilia

 

References

ARAUJO, M. “O uso de inteligência artificial para a geração automatizada de textos acadêmicos: plágio ou meta-autoria?” Logeion: Filosofia da Informação, vol. 3, n. 1, 2016.

BARBOSA, L. M.; PORTES, L. A. F. “A inteligência artificial”. Revista Tecnologia Educacional, n. 236, 2019.

BOA SORTE, P.; et al. “Inteligência artificial e escrita acadêmica: o que nos reserva o algoritmo GPT-3?” Revista EntreLinguas, vol. 7, 2021.

BROCKMAN, G. et al. “OpenAI Gym”. Arxiv vol. 1, 2016.

CARMO, C. R. S.; CARMO, R. O. S.; MELO, G. D. “A inteligência artificial e os desafios na avaliação da escrita acadêmica”. Cadernos da Fucamp, vol. 21, n. 53, 2022.

CRESWELL, J. W. Projeto de pesquisa: métodos qualitativo, quantitativo e misto. Porto Alegre: Editora Artmed, 2007.

D'ALTE, P.; D'ALTE, L. “Para uma avaliação do ChatGPT como ferramenta auxiliar de escrita de textos académicos”. Revista Bibliomar, vol. 22, n. 1, 2023.

DANTAS, J. “Arte, profissão e tecnologia: as greves em Hollywood em 2023”. Boletim de Conjuntura (BOCA), vol. 15, n. 45, 2023.

DONATO, H.; DONATO, M. “Etapas na condução de uma revisão sistemática”. Acta Médica Portuguesa, vol. 32, n. 3, 2019.

DONEDA, D. C. M. et al. “Considerações iniciais sobre inteligência artificial, ética e autonomia pessoal”. Pensar Revista de Ciências Jurídicas, vol. 23, n. 4, 2018.

ENDERLE, J. D.; BRONZINO, J. D. Introduction to biomedical engineering. Burlington: Academic Press, 2012.

FERREIRA, R. C. V. et al. “O surgimento do Chat GPT e a insegurança sobre o futuro dos trabalhos acadêmicos”. Cadernos de Dereito Actual, n. 21, 2023.

FREIRE, P. A importância do ato de ler: em três artigos que se complementam. São Paulo: Editora Autores Associados, 1988.

FREIRE, P. Educação e Mudança. Rio de Janeiro: Editora Paz e Terra, 1976.

FREIRE, P. Pedagogia dos sonhos possíveis. São Paulo: Editora da UNESP, 2001.

FREIRE, P. Professora sim, tia não: cartas a quem ousa ensinar. São Paulo: Editora Olho d' Água, 1993.

FUJITA, J. S.; PEDROSA, J. M. B. F. “Inteligência artificial, algoritmos e o impacto das novas tecnologias nos processos judiciais da sociedade da informação”. Scientia Iuris, vol. 26, n. 1, 2022.

G1. “‘Inteligência artificial’ é a palavra do ano do dicionário Collins”. G1 [2023]. Disponível em: . Acesso em: 20/11/2023.

GALLO, S. M. L. “Chat GPT: hiperautor ou não autor?” Traços de Linguagem, vol. 7, n. 1, 2023.

HELMUS, T. C. Artificial Intelligence, Deepfakes, and Disinformation: A Primer. Santa Monica: Rand Corporation, 2022.

LANDIM, W. “Chat GPT: o que é, como funciona e como usar”. Mundo Conectado [2023]. Disponível em: . Acesso em: 26/09/2023.

LIBERATI, A. et al. “The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies that Evaluate Health Care Interventions: explanation and elaboration”. Annals of Internal Medicine, vol. 151, 2009.

LOBO, L. C. “Inteligência artificial, o futuro da medicina e a educação médica”. Revista Brasileira de Educação Médica, vol. 42, n. 3, 2018.

LOPES, C. C. “Ética e pesquisa em Educação: subsídios”. Anais da Reunião Nacional da Associação Nacional de Pós-graduação e Pesquisa em Educação. Rio de Janeiro: ANPEd, 2023.

MARANHÃO, J. “A pesquisa em inteligência artificial e direito no Brasil”. Consultor Jurídico [2017]. Disponível em: . Acesso em: 01/09/2023.

MCCARTHY, J. et al. “A proposal for the Dartmouth summer research project on Artificial Intelligence”. AI Magazine, vol. 27, 1956.

MOHER, D. et al. “Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement”. PLoS Medicine, vol. 6, n. 7, 2009.

NICOLELIS, M. O verdadeiro criador de tudo: como o cérebro humano esculpiu o universo como nós o conhecemos. São Paulo: Editora Planeta, 2020.

NOVAIS, P.; FREITAS, P. M. Inteligência Artificial e regulação de algoritmos. Brasília: Ministério da Ciência, Tecnologia, Inovações e Comunicações, 2018.

OPENAI, L. L. C. “ChatGPT: optimizing language models for dialogue”. OpenAI [2022]. Disponível em: . Acesso em: 29/09/2023.

PERES, F. “Alfabetização em saúde no ChatGPT: explorando o potencial do uso de inteligência artificial na escrita acadêmica”. SciELO [2023]. Disponível em: . Acesso em 11/09/2023.

PRICE, C. C. ChatGPT’s work lacks transparency and that is a problem: Objeticve analysis. California: Rand Corporation, 2023.

RUSSELL, S. J. Artificial Intelligence: a modern approach. New York: Prentice Hal, 2009.

RUSSELL, S. J.; NORVIG, P. Artificial intelligence: a modern approach. Upper Saddle River: Pearson Education Limited, 2010.

SICHMAN, J. S. “Inteligência Artificial e sociedade: avanços e riscos”. Estudos Avançados, vol. 35, n. 101, 2021.

SILVA, A. P. P. N. “O Estado da Arte ou o Estado do Conhecimento”. Educação, vol. 43, n. 3, 2020.

TECMUNDO. “A origem do ChatGPT: conheça a história da OpenAI”. TecMundo [2023]. Disponível em: . Acesso em: 26/09/2023.

WEINER, N. “Algumas consequências morais e técnicas da automação: uma refutação”. Revista Science, vol. 132, n. 3429. 1960.

WIEZENBAUM, J. Computer power and human reason. San Francisco: Freeman, 1076.

Most read articles by the same author(s)