VIRTUAL: Applying Education

Abstract

Artificial Intelligence (AI) has become a powerful tool in various fields. For example, ChatGPT, a language model based, has been tested for its capacity to handle complex topics and has even been used by researchers to write paper abstracts. Nevertheless, architectural design requires careful consideration of utilizing AI for teaching, learning and practice. At the same time, Text-to-Image AI platforms can generate endless design alternatives by using keywords that may lead to strange images of the place. Therefore, this paper proposes using Repertory Grid Technique (RGT) instead to generate constructs that could serve as keywords during the architectural design process. First, five architectural engineering students were interviewed using RGT to elicit constructs about the meaning of architectural forms. Then, the authors used these elicited constructs as keywords in two AI platforms, Midjourney and Prodia. The resulting images were rated by seven practising architects, who found them reliable and reasonably relevant to the brand image of Al-Ain City in the United Arab Emirates. The study supports that using AI-generated images to construct visual scenarios can produce better results than relying on examples of existing buildings since the former can be tailored to individual preferences and needs. This method can help architects overcome the initial difficulties of using AI in architectural design and motivate students to create designs incorporating a local brand image of architecture. Furthermore, the results provide helpful feedback for AI and RGT developers to enhance and personalize the database’s input.