O uso da inteligência artificial no ensino de química : uma análise das ferramentas para uma aprendizagem significativa dos conceitos de propriedades coligativas

Abstract

This undergraduate thesis analyzes the role of artificial intelligence (AI) as a support tool in Chemistry teaching and learning, focusing on the concepts of Colligative Properties, which are notoriously challenging for students. The research compared the responses of generative AIs—ChatGPT, Copilot, and Gemini—to a set of questions about the topic and applied a questionnaire to 19 students to investigate their perceptions of these technologies. The analysis of the AIs demonstrated that, while all were capable of providing correct definitions and examples, they showed divergences in clarity, depth, and accuracy on complex topics such as the Van't Hoff factor. The student survey, in turn, revealed that they are "digital natives" who see AI as a valuable resource for obtaining detailed explanations and visualizing concepts. However, they reject the idea that AI can replace the teacher, reinforcing the educator's essential role as a mediator and knowledge curator. The study concludes that AI is a powerful pedagogical ally, but its effectiveness depends on the teacher's critical mediation to ensure conceptual correctness and foster critical thinking, culminating in the development of a practical guide to optimize this interaction.