▪ WHAT DID I LEARN ABOUT AI?

Through this exploration, I learned that artificial intelligence is a powerful digital tool that can support architects in multiple areas such as research, visualization, concept development, and problem-solving. Different AI platforms serve different purposes. Text-based AI tools help generate explanations, ideas, and structured content, while image-generation AI assists with visualization, atmosphere creation, and conceptual architectural rendering. Research AI tools help collect summarized information and references efficiently.

I also learned that the quality of AI-generated outputs depends heavily on the quality of the prompt provided. More detailed and specific prompts produced clearer, more organized, and more relevant results. This showed the importance of prompt engineering in improving AI performance within architectural workflows.

Most importantly, I realized that AI performs best when combined with human creativity, critical thinking, and professional judgment. AI can support the design process, but it cannot fully replace human imagination, emotional understanding, or cultural sensitivity in architecture.

▪ WHAT ARE THE LIMITATIONS?

Although AI offers many advantages, it still has several limitations within architecture and design.

• AI can sometimes generate inaccurate or misleading information.

• Certain AI-generated architectural visuals may appear unrealistic or structurally impractical.

• AI lacks emotional understanding, cultural awareness, and human experience.

• Excessive dependence on AI may reduce originality and independent creative thinking.

• Ethical concerns exist regarding copyright, authorship, and data privacy.

• Some AI tools provide generalized outputs that require professional review and correction.

Because of these limitations, AI should be considered a supporting design assistant rather than a complete replacement for architects and designers.

▪ WHAT SURPRISED ME?

The most surprising aspect of this exploration was the speed and efficiency of AI tools. Within seconds, AI could generate architectural concepts, visual renderings, organized research summaries, and structured written content that would normally require much more time to produce manually.

I was also surprised by how significantly small changes in prompts could affect the final results. Adjusting the wording, tone, or level of detail often produced completely different outputs. This demonstrated how important prompt writing is when working with AI systems.

Another surprising aspect was the realism and creativity of AI-generated architectural visuals. Some outputs appeared highly detailed and visually convincing even though they were created entirely from text descriptions. This showed the strong potential of AI in supporting future architectural visualization and conceptual development.