REFLECTION

4. Comprehensive Reflection

What worked well?

The multi-stage AI workflow significantly accelerated the conceptual development of the Wave Pavilion by integrating environmental analysis, concept generation, and architectural visualization into a single design process. Research AI provided valuable climate data for Kajang, including solar exposure, humidity levels, and prevailing wind directions, ensuring that the project remained grounded in environmental performance.

The Text AI effectively transformed environmental requirements into architectural strategies, helping develop concepts such as the wave-shaped roof, central wind corridor, open courtyard, and passive shading systems. The Image AI then translated these ideas into visual representations, allowing rapid exploration of form, materiality, and atmosphere.

What failed?

The primary limitation was the Image AI's lack of understanding of structural feasibility and construction logic. Early renderings produced unrealistic roof spans, unsupported cantilevers, and exaggerated glazing systems that would be difficult to construct in reality.

Some visualizations focused heavily on dramatic aesthetics while overlooking practical considerations such as structural support systems, material limitations, and environmental performance. As a result, several design iterations required significant modification before they could be considered realistic architectural proposals.

Which AI tool was most useful?

The Research AI (Perplexity) proved to be the most valuable tool within the workflow. While Image AI generated compelling visualizations, the environmental data supplied by Research AI provided the foundation for the entire design process.

The solar analysis, wind patterns, humidity data, and climate-responsive recommendations directly influenced the development of the wave roof, the courtyard, the wind corridor, and the building envelope strategy. Without this information, the project would have remained a purely aesthetic exercise rather than a climate-responsive architectural proposal.

Where was human judgment required?

Human judgment was required throughout every stage of the workflow.

Phase

AI Limitation

Human Intervention

Research Phase

Produced generalized climate recommendations.

Selected strategies appropriate for a tropical student clubhouse.

Concept Development

Generated expressive forms without considering construction practicality.

Simplified geometries and improved structural feasibility.

Spatial Planning

Prioritized visual impact over functional requirements.

Reorganized spaces according to circulation and program needs.

Visualization Phase

Produced unrealistic roof spans and excessive glazing.

Refined structural systems, roof proportions, and façade design.

How did the workflow improve efficiency?

The AI workflow significantly reduced the time required for environmental research, concept development, and visualization. Tasks that would traditionally require multiple software platforms and extensive manual testing were completed within a much shorter timeframe.

The workflow enabled rapid iteration of roof forms, spatial arrangements, daylighting strategies, and environmental responses. Multiple alternatives could be explored quickly, allowing stronger design decisions to emerge earlier in the design process.

What are the risks of depending too much on AI?

The greatest risk is prioritizing visual appearance over architectural reality. AI-generated images often appear highly convincing despite lacking structural logic, construction feasibility, or compliance with building regulations.

Overreliance on AI can lead to designs that ignore important considerations such as load transfer, material performance, fire safety requirements, maintenance needs, construction costs, and user experience. While AI is highly effective as a design assistant, it cannot replace architectural judgment and technical expertise.

Conclusion

The Wave Pavilion demonstrates how AI can enhance architectural design by accelerating research, concept generation, and visualization. However, the success of the project ultimately depended on the combination of AI capabilities and human critical thinking. The most effective workflow emerged not from AI alone, but from the collaboration between artificial intelligence and architectural judgment, resulting in a climate-responsive, functional, and buildable student clubhouse.