Enhancing Student Engagement and Lecturer Roles in Electrical Educational Programs using LLM Models

Today, LLM models have become an essential technology element, with applications in education constantly evolving. Its economic viability is underpinned by investments from large multinationals, with the needed vast amounts of data provided free of charge by users in their daily interactions with these systems, and the paid subscriptions that have recently emerged, with a tendency to become more expensive in the future. This scenario raises several critical issues in terms of governance, both at country level and within education and educational institutions.Europe is reacting to these changes. In France, for example, the development of Mistral AI stands out; in Spain, the ‘Alia’ project has emerged; and in Portugal, the ‘Amália’ model is starting to be available. This paper addresses this issue in an educational context, analysing the risks and opportunities associated with using different AI platforms, such as ChatGPT and Gemini, in teaching in electrical engineering. The changes implemented, the students' general feedback and their engagement throughout the process were carefully documented, as well as their perceptions of the use of AI in learning. During the semester, student-centred teaching methodologies were adopted, such as the use of explanatory videos and the flipped learning approach, providing extra support or pre-class preparation. In face-to-face sessions, students took part in tutorials guided by the main learning objectives, using artificial intelligence to reinforce their understanding of the content.The assessment, structured into various components spread over three curricular units, included quizzes, theoretical work, practical activities, sessions supervised by the educators in which the students clarified their doubts with each other, and a final exam. At the end of the semester, the students answered a specific survey on the use of artificial intelligence in the teaching-learning process. All these points and results are presented in the paper.