Use of ChatGPT and its implications in the academic and scientific community: failures, misuse and challenges

This article systematically analyzes the main advantages and limitations of using conversational artificial intelligence (AI), specifically ChatGPT, as a support tool for essay writing in academic and scientific contexts. Additionally, selected problems in mathematics and science were addressed to evaluate the effectiveness of this platform. To this end, a systematic case study was designed and implemented, through which data generated by the chatbot were collected and analyzed. As part of the study, essays were written on six different topics, with the introductory sections produced using ChatGPT. In parallel, six mathematical problems were solved, covering topics such as integrals, Z-transform, Laplace transform, and ordinary differential equations. Furthermore, the model was prompted to provide the definition and calculation method for the fracture toughness of materials. The data analysis led to the conclusion that ChatGPT is capable of generating grammatically correct and syntactically coherent texts within a short time frame. However, the generated introductions were limited in content, relying mostly on general information and lacking critical argumentation. Regarding the mathematical case study, it was observed that, in its current state, the platform exhibits significant limitations in solving complex equations, highlighting the need for substantial improvements to enhance ChatGPT's performance in mathematical applications.

• 主要作者: Pech-Rodríguez, Wilian J. (author)
• 其他作者: Armendáriz-Mireles, Eddie N. (author), Calles-Arriaga, Carlos A. (author), Rocha-Rangel, Enrique (author)
• 格式: article
• 語言: spa
• 出版: 2025
• 主題: Inteligencia artificial; plataforma digital; asistente conversacional; algoritmos de aprendizaje automático; Artificial intelligence; digital platform; conversational assistant; machine learning algorithms
• 在線閱讀: https://revistas.uteq.edu.ec/index.php/ingenio/article/view/1091