What cannot be imported, why desert architecture redraws the boundaries of AI in Jeddah

While engineering studios race to generate building façades with a single click, the rapid construction boom in the kingdom’s cities has run into a silent technical and cultural question: what computational models can truly design, and what they are completely unable to grasp? At the fifth International Conference on Architecture and Design hosted by Afif University in Jeddah under the theme “From Desert Heritage to Global Legacy,” the discussions paused at an explicit paradox suggesting that the future of urban planning may learn from ancient desert settlements as much as it may learn from artificial-intelligence algorithms.

It is not about nostalgic longing for the past, but about a precise engineering deconstruction of technology’s role in harsh environments. Dimitra Kanellopoulou, professor at the École Nationale Supérieure d'Architecture de Paris-Malaquais in France, explained that architecture cannot turn a blind eye to modern technological tools that ease life, but the difference lies in what she called vernacular architectural intelligence and the climate-adaptation expertise passed down through generations."What we cannot import, and what we must not forget, is vernacular architectural intelligence, and how humans learned to live in harsh desert conditions"Thus Kanellopoulou framed the strict limits of what ready-made mathematical models fail to transfer from humid continents to intensely scorching deserts.

This divide between automated calculation and human expertise was clearly outlined by Dr. Tarek Rakhha, co-founder and CEO of Lamarr.AI and associate professor at the Georgia Institute of Technology, when he emphasized the need to resist AI taking over the very practices of creative design. Rakhha argues that the greatest value of intelligent systems lies behind the scenes: automating the assimilation of engineering data, understanding construction mechanisms, and field monitoring to improve efficiency, while warning against the chase for speed and cost reduction that yields buildings lacking genuine quality and losing connection with human nature.

The practical impact of this discussion reaches directly to architecture studios and real-estate development firms in the Gulf and Egypt. Overreliance on Western visual-generation tools imposes costly operational and climatic risks, because global models have been trained on temperate environments that ignore the physics of natural ventilation, shadow distribution, and traditional mashrabiya techniques in cities such as Al-Wajh and historic Jeddah. The skill required today from a consulting engineer is no longer mastering automated generation commands, but possessing the critical ability to audit algorithmic outputs, adapt them to site-specific data, and limit AI’s role to supply-chain management and construction-site monitoring to reduce financial waste.

The shift reflected in the 122 papers presented at the conference from 25 countries confirms that the success metric is no longer the speed of plan delivery or tower height. The real gamble facing urban decision-makers lies in creating spaces that retain their cultural identity and withstand their climatic environment, rather than filling the urban landscape with glossy buildings run by algorithms that ignore the desert’s harshness.