The Case Study of the Galaxy Nexus Café project showed the support and enhancement of AI to architectural design process for the research, concept development, prompt development and visualization. Various AI tools were employed throughout the project, such as Perplexity AI for researching architecture, ChatGPT for brainstorming concepts and prompts, and Bing Image Creator for creating visual content. These tools enabled the rapid design process, and provided the ability to rapidly investigate design ideas and design possibilities.
The most successful part of the workflow was that it was able to merge the Neo-Futurist style and Tropical Modernism together in a manner that is visually futuristic yet still reacting to the tropical environment of Malaysia. The AI tools were particularly good at coming up with atmospheric ideas, futuristic shapes and visual inspiration. The quality of the final outputs was also enhanced using iterative prompt refinement. The refined images were more realistic, balanced and appropriate to a student café, when compared to the first set of images created. The lighting, vegetation, sense of space and finishing of the materials gave the design more realism and coherence with the planned architectural concept.
But the process also identified a few drawbacks of the use of AI for architecture. AI-generated outputs were sometimes overly technologically advanced with no regard to functionality, realism, or climate responsiveness. Some images presented appealing aesthetics but lacked functionality with regard to shading, ventilation and user comfort. This proved that AI can sometimes go out of its way for architecture to be creative rather than feasible. For this reason several prompt adjustments and evaluations were needed in order to get a satisfactory result.
The human element was very important at every stage of the process. Creating a realistic and functional design that is also comfortable and environmentally appropriate could not be entirely determined through AI outputs. Different results, weaknesses, prompt improvements and the best-looking visual direction can only be determined through human evaluation. Personal judgment also played an important role in the design being appropriate for the students and the hot-humid weather conditions in Malaysia. The process proved that AI is a great creative asset but architects must still make decisions on the final design.
The workflow increased the efficiency dramatically, requiring less time for brainstorming, visualization and concept exploration. AI tools were able to quickly generate various architectural concepts and moods within a short time, rather than manually producing numerous conceptual sketches. This facilitated experimentation and fostered greater creativity in the process of design.
The project also pointed to the dangers of over-reliance on AI. Excessive use of AI might lead to the creation of less original, critical, or technically evaluated work if designers blindly copy-paste content without going into analysis. It is also important to note that the generated results from AI might not always be realistic or environmentally friendly, so they should be thoroughly checked. Overall, the project has highlighted the need for a harmonious relationship between AI assistance and human creativity, architectural expertise, and critical thinking to ensure that the generated ideas lead to tangible and impactful architectural solutions.