Disruptive Education: Integrating ChatGPT into an Active Methodology for Teaching Sciences and Mathematics

Resumo

Context: The integration of Artificial Intelligence (AI), marked by the advent of ChatGPT and other Large Language Models in December 2022, has demonstrated transformative potential in Science and Mathematics educational practices. This technology offers prospects to revolutionize conventional pedagogical methods and reduce socioeconomic disparities. A teaching model is proposed that employs Generative AI tools in flipped learning environments, allowing students to prepare at home through simulations that act as Sophotechnic Mediation. Objectives: An active methodology is presented where the use of ChatGPT at home precedes and prepares classroom learning, applying the Cognitive Networks Mediation Theory. Design: Hypercultural Mediation was utilized with simulations developed via ChatGPT, alongside mediations arising from students' previous experiences or integrated into classroom activities. Environment and Participants: Conducted with high school students, the research focused on the Physics curriculum, exploring concepts of Light and General Relativity Theory. Data Collection and Analysis: Following the ideas of Bachelard and Mortimer, the External Mediation Level Profile was created to visualize the most effective mediations for creating mental images. Results: A significant correlation was observed between the mental images formed and the students’ gestures, suggesting a deepened understanding of the discussed concepts. Conclusions: Integrating AI with the Cognitive Networks Mediation Theory can help address traditional teaching limitations by considering individual student characteristics and fostering the development of modeling and problem-solving skills.