STEP 3 — PROMPT STRATEGY AND REFINEMENT
Improving AI Outputs Through Better Prompts
Embedded Files
Prompt refinement was an important stage in the workflow. At the beginning, a simple prompt produced a general answer. However, when more project information was added, the AI output became more specific and useful for the Silent Transition Student Clubhouse.
This stage helped me understand that prompting is part of architectural thinking. A prompt should not only ask AI to design something. It should include context, location, climate, users, concept, and design intention.
By refining the prompt, I was able to guide the AI toward a clearer architectural response that connected better with the project’s site condition, spatial sequence, and calm student experience.
Improved Prompt:
“Generate a concept for a student clubhouse in Malaysia that creates a transition from a noisy campus environment to a calm interior space. The design should include shaded circulation, arches, soft daylight, natural ventilation, tropical greenery, quiet study areas, and social student spaces.”
Refined Prompt:
“Develop a concept for the Silent Transition Student Clubhouse in a tropical Malaysian campus. The site is noisy due to road traffic, campus movement, and sports activities. The design should create a gradual journey from active public areas to calm interior spaces using spatial sequence, controlled daylight, shaded circulation, arches, passive cooling, and quiet student zones.”
Human Judgment
Human Judgment
A better prompt does not mean the AI makes the final decision. The designer still needs to review, edit, and select the most suitable ideas.
Figure 10: Prompt refinement improved the AI output by adding project context, climate, users, site condition, and design intention.
Page updated
Google Sites
Report abuse