Design com ferramentas de inteligência artificial: uso do midjourney como apoio ao longo do processo de criação visual

Abstract

Aqui está o texto no padrão justificado, sem qualquer alteração de conteúdo — apenas com a remoção das quebras de linha para formar um parágrafo contínuo, conforme a formatação acadêmica: This dissertation investigates the impact of generative artificial intelligence tools on design, focusing on the use of Midjourney as a creative support tool. The study analyzes the role of AI in the development of visual identities and support in illustrations, exploring the balance between automation and human intervention. The research adopts both a practical and theoretical approach, considering the creative potential of these tools as well as the ethical and technical challenges involved in their application. To contextualize the development of generative AI and its applications in design, a historical review and a technical description of the use of this technology were conducted. In addition, the study carried out three distinct practical experiments and one experiment focused on teaching. The first experiment investigates the use of AI as a support in creating illustration systems and visual identities, analyzing how generative tools contribute to the creative process. The second focuses on creating visual identities with AI, exploring the challenges in designing logos and the need for adjustments during the creative process to ensure legibility, versatility, and aesthetic coherence. The third experiment examines the materialization of AI-generated images using a Pen Plotter, comparing physical printing and assessing the limitations of AI in replicating the expressiveness of the human hand. In addition to the practical experiments, a discussion on AI-generated images was held with design students, addressing issues such as authorship, creative control, and algorithmic bias. This process revealed the limitations of AI in replicating the unpredictability of the human hand, highlighting that, despite its ability to simulate textures and brushstrokes, substantial differences still exist between digital and manual work. The dissertation also discusses ethical issues related to the use of these tools, such as the presence of watermarks and signatures in generated images and the algorithmic bias that favors Western aesthetic standards. Even when diverse cultural elements were requested, AI models showed limitations, relying on predominant cultural references found in their training data. Another topic addressed was the AI-assisted construction of visual identity, focusing on logo creation and the need for adjustments to ensure legibility, versatility, and aesthetic coherence. The study revealed that although AI speeds up experimentation, final curation and refinement still depend on human intervention. The research also included a classroom experience in which students analyzed the challenges and impacts of AI on design. The exercise demonstrated that, although there are concerns about the replacement of human labor, AI can function as a complementary tool, expanding creativity without compromising authorship. The results of this study reinforce that generative artificial intelligence tools should not be seen as replacements for designers, but rather as allies in the creative process. Their use requires a critical and ethical approach, ensuring that technological innovation is integrated consciously, without compromising the diversity, authenticity, and originality of artistic work.