Does the Emergence of AI in Architecture Herald a New Era of Design and Practice?

Authored by: Dean Hodcroft, Head of Europe
Updated on: April 12, 2024

AI has the potential to reshape the architectural landscape the world over. This presents a dual-edged sword - offering unprecedented design capabilities while challenging the very essence of architectural practice.

The La Sagrada Familia is perhaps one of the most famous landmarks globally, famous both for the striking pose it cuts through the Barcelona skyline and the fact that it is still unfinished, after more than 140 years. Its architect, likewise, is one of the most famous the world over.
What would happen if we could all be an Antoni Gaudí at the press of a button? Can the power of Artificial Intelligence (AI) enable anyone to construct an architectural marvel like La Sagrada Familia? Essentially, does technological advancement bring us closer to emulating the genius of Antoni Gaudí? Perhaps.

For now, we know the integration of AI in architecture and the built environment has started what will be seen in the future as a transformative era in design and practice, offering a suite of innovative tools and raising profound questions about the future of the profession.

The integration of AI in architecture marks the beginning of a transformative period in design and practice. This shift is not just theoretical; buildings are already being designed and built with the help of AI. Venture capital firm A/O Proptech highlights the significance of this trend, noting that AI startups focusing on the built environment have garnered $12.3 billion in funding over the past three years.

Xkool was one of the first AI tools for architects, where structures like the La Sagrada Familia are potentially just a few clicks away, but other technologies like Midjourney and Stable Diffusion facilitate text-to-image generation capabilities that are routinely mistaken for photographs. Other niche AI tools exist, like 3DGuru, Hypar and Delve, tools which can do everything from speeding up early concept design to interiors and predicting how much sunlight each property in a development would receive. Xkool stands at the forefront of AI in architecture, experiencing rapid growth with over 50,000 users in China and the recent introduction of its image-to-image AI tool, LookX, the first AI tool trained on data for and created specifically for architects.

The company was established in 2016 by Wanyu He and some former colleagues, driven by a vision to innovate beyond traditional architectural practices. Xkool's platform harnesses AI to offer comprehensive solutions, from creating masterplan layouts that adhere to specific requirements like daylight exposure, spatial standards and local regulations, to detailing interiors and construction. Additionally, Xkool has developed technology capable of converting 2D building images into 3D models and transforming lists of room dimensions into practical floor plans.

There is conversation around how AI's most prolific impact on architecture will be to eradicate the profession - but that seems to be a very long way off from happening. Nicolás Valencia from ArchDaily suggests AI's role in architecture is not to supplant human creativity but to empower professionals with efficiencies. For example, had Gaudí had access to an AI when he was designing the La Sagrada Familia, would it be the same structure today - and be finished already? Between his genius and AI as a technological marvel, it's hard to think that it couldn't be done.

A significant issue with AI in architecture is its inherent cultural biases, but also well documented technological biases. Andrew Kudless, an Associate Professor of Architecture at the California College of the Arts in San Francisco, points out AI's reliance on existing data pools can lead to overrepresentation of certain styles and underrepresentation of others. This bias is not necessarily intentional but reflects the predominance of certain data in the AI's training. Put in other words, if architects all use the same AI platforms, what happens to true originality - and our built environments?

One of the most promising aspects of AI in architecture is its potential to enhance sustainability. AI can optimise energy efficiency, thermal comfort and material selection, significantly reducing environmental impacts.

However, this transformative potential is accompanied by its own set of challenges. As AI integrates into architecture, the industry faces the challenge of balancing traditional values with new demands in data literacy and interdisciplinarity.

My ultimate hope is that architecture is no exception to the overriding premise of AI, whereby the human brain is freed up from excessive menial tasks to be more imaginative and expressive. That could be pretty exciting for the future of architecture in particular. Who knows what 22nd century designs will look like?

Published on: April 12, 2024