From visual grammar to prompting: Lessons learned in creating differentiated messages with artificial intelligence

Abstract

This systematization of experience is based on a teaching strategy developed in the Graphic Design I course of the Graphic Design program, during the first semester of 2025, at the Western Campus (Sede de Occidente). The group consisted of 20 students and the strategy was implemented in the course's third project. The objective of this article is to analyze the design process and experience of Graphic Design university students in creating images with generative artificial intelligence, evaluating the effectiveness of the visual message when adapted to four generational groups (baby boomer, X, millennial, and centennial) for social media publications. A mixed qualitative and quantitative approach was applied. Among the results obtained, working with a real brand stood out for promoting research and visual adaptation to diverse audiences. The findings confirm time optimization and concept generation as strengths of AI, while its weaknesses lie in visual quality and bias. In the conclusions, the systematization demonstrated generative AI's capacity to differentiate visual messages according to the aesthetics of generational groups on social media; it also strengthened students' understanding of the link between brand and audience, fostering the development of critical prompting skills and ethical reflection on the limitations and opportunities of these emerging technologies in contemporary design.