Artificial Intelligence and the production of academic papers at university

Abstract

The popularization of artificial intelligence (AI) brings benefits to science and improves the productivity of everyday tasks. However, replacing a manual task does not always bring benefits, as in school activities. Here, we reflect on the use of AI for the production of academic work by university students, raising the main questions: What is the impact of the use of AIs that generate texts on the academic writing of undergraduate students? How can we identify texts produced by AI in their academic work? What strategies can be used to create activities in undergraduate courses that require active student participation, avoiding dependence on AI tools? To answer these questions, we use information obtained from activities carried out in three classes of a course on reading and producing academic texts at the Federal University of Minas Gerais. In the analysis of these activities, common characteristics of AI-generated texts are identified, and ways to propose challenging activities that make it difficult to replace written production with AI requests are also pointed out. We also reflect on ethical issues and on what may lead students to use AI to do academic work. It is hoped that these results will help teachers deal with advances in AI in the university context.