Automation has a massive footprint in many industries, but as it makes its descent on architecture, it raises many eyebrows, as well as questions. What is the scope of something mechanical and limited in a field that has its limitations bound by the imagination of the designer? Does AI replace the architect, does it spell doomsday for the designer? Well, the answer is a yes and no, for it is more complex- the extent of A.I in design. Let’s probe into what A.I means for architecture- and its journey through the years.

Artificial Intelligence is the result of years of progress in automation and development in design. First seen in Corbusier’s modular designs, a modifiable template became a popular tool for designers to use, giving the user a sense of control. By the late ’80s, computer-aided design (CAD) would explode all over the design world, making drafting easy, quick and shareable. CAD marked a new era of design development, people relying on computers to realize design in 2D.

Though CAD had 3D possibilities, it still wasn’t compatible or advent for A.I yet, which has introduced a plethora of possibilities, from the floor plan to façade design. The prospects of A.I were initially noticed with the origin of parametric design, which has now picked up to become a popular form of design- facades as well as volumes. The repetition of elements in progression creating interesting forms was the beginning, the volumes could be tested for temperature and context with the click of a program. Here, the doors of A.I opened.

AI is incomplete without Generative Adversarial Neural Networks (GANs). These programs are a keystone in programmable design, GAN’s can learn from the statistical data given to them to find an anomaly or a “solution”. Very simply put, a ‘Generator’ creates images from a data set, and the ‘Discriminator’ can give feedback about the output.

How is AI changing construction in architecture- Sheet1
GAN’s Architecture Process ©Stansilas Chaillou

Organizational data in a framework is a standard in architecture, something that a GAN can fabricate in a short period. As these programs and generational software are incrementally progressive, it can create an array of options to choose from, with the set of data entered. As these GAN’s pick up pace in the design world, they might be able to produce reprised designs instantaneously, when they reach one plus ultra.

How is AI changing construction in architecture- Sheet2
Style Transition using AI ©Stanislas Chaillou

The applications of such programs are multitudinous. Stanislas Chaillou, from the Harvard School of Design (HSD), clearly lists the possibilities of these systems now- including style transformation- where a space changes geometry and its form to suit the art movement it has been associated with. A layout assistant- divides interior spaces by footprint and emulates tested floor plans, each one as viable as the last. Furnishing is also possible- with thousands of layout possibilities generated in a matter of seconds for a single house.

How is AI changing construction in architecture- Sheet3
Furniture Layout Iterations ©Stanislas Chaillou

A.I’s plus point is that it can learn, in moments- from the previous set of data entered. Hence, like a brain, it is learning, and as this article is being written and read by you, an A.I program has already generated plans in a matter of seconds. Artificial Intelligence, no matter the field, will always continue to improve. From generating orientations of buildings, the walls and even the sunshades of windows, A.I already can do it all. The question remains- will A.I replace architects?

Though the capabilities are infinite- artificial intelligence is limited by the initial data or any data that has been fed to it. The programs can generate thousands of designs, at least more iterations than any architect in a short period, but it misses some important elements. Among others- the passion for design.

How is AI changing construction in architecture- Sheet4
Circulation Diagrams ©Stanislas Chaillou

Architecture is the holy matrimony of art and science- through science is logical and a numbers game, art is the birth of the need and a product of imagination and creativity. These elements are unbound, they are intangible elements that machines can never have, at least for the next few decades. Humans have the upper hand- a hand that feels and soaks in the essence of every project or design attempted. Correspondence of architects and artists, designers and fields of a multi-disciplinary style challenge the smallest aspects of design- things that are spur of the moment and created by the creative complex that the brain houses.

Till machines perfect the human brain, the answer is no, A.I will not replace architects. It may be able to generate designs in a blink, even aid designers to a level never even imagined or thought of before, but the duty of the designer remains to breathe life into design. For, in the end, that is what separates a generated design and a crafted design- the pulse of passion.


Nishant Verma is a designated college nerd and has been writing ever since you could define the term “bullying”, first to vent out feelings and eventually to an amateur writer. Pastimes include productive activities- reading, writing, movies, the history of music and architecture, with whom he enjoys a love-hate relationship.