The use of modelling and design software among architecture firms continues to grow, bringing added accuracy, flexibility and efficiency to the design process. But underlying all of the benefits that these tools have to offer, there has been a growing unease among some industry veterans.
It could be argued that advances in computer aided design cheapen the knowledge and experience of architecture professionals- making a significant part of their training irrelevant and even obsolete. Yet, I believe this is really just the tip of the iceberg. The real threat to architects and to the profession as a whole is Artificial Intelligence (AI).
Why AI is Such a Big Deal
AI refers to the “intelligence” of computers and software. Three decades ago, computers were hailed for their ability to access, manipulate, and analyse vast amounts of data that they could then use to reach probable solutions or assumptions. Today, when a computer or a collection of code is called “intelligent,” it means the system is adaptive, inquisitive, borderline intuitive and ultimately able learn from past experience. In other words, it has the ability to get smarter over time.
Of interest to architects in particular is the fact that AI is even making its way into uncharted territory: non-technical, creative tasks, such as writing and art. As computers get better at manipulating and analysing data, images, and 3-dimensional spaces, and as advances in speech recognition coupled with virtual and augmented reality blur the separation between man and machine, self-generating design solutions are no doubt on the horizon.
This leads to an intriguing question: Could an intelligent machine or piece of software navigate the complex needs, desires, ambitions and ideas surrounding a given project and come up with a viable, even likable, design solution?
We are definitely edging closer.
But, there is an even bigger picture over here. AI has the potential to influence the design and build process at every stage– from site research to the realisation and operation of the finished structure. This comes as the increasing complexity of design, engineering and construction meet up with economic pressures demanding that architects provide even more service for less money.
In this brave new world, architects could be replaced by algorithms and big data, and an architecture “technician” working in a construction or engineering firm could theoretically be put in charge of the kind of complex projects that require several human teams today.
Finding the Hidden Opportunities
The idea that a computer or a piece of software could be smart enough one day to do almost all the work of a trained architect may not be so comforting to those already in the profession. But for the industry to survive the worst thing professionals and the firms they work for can do is stick their heads in the sand.
AI isn’t going away… and while these technologies may not entirely replace the architect, it promises to dramatically transform the architect’s role in the scope of a project as well as how he or she is trained.
That said, here are three areas where AI may bring the biggest disruption to architecture. But at the same time these very areas offer the biggest opportunities. Discerning and courageous architecture studios can (and should) start experimenting with some of this technologytoday. By doing so, it gives them the best chances of maintaining their place in the design and build process:
- Automation. With automation, routine, transaction-based and analytical tasks, such as preparing schedules, measuring, calculating, and even evaluating safety concerns and zoning compliance, are performed by algorithms that can not only tap into big data sources, but both upgrade and improve itself over time. With the need for human-input into these areas eliminated, architects are then free to direct their attention to the overall quality and suitability of the design.
- Site and social research. The first stage of any project is collecting and analysing information about the project location as well as the people who will be interacting with the finished space and then matching this data up with the project goals. Today, much of that data is available online. Some forward-thinking BIM programs, such as Google SketchUp, are already starting to tap into these data pools in order to create a surprisingly accurate simulation of a site. This means architects can instantly conduct environmental and construction analysis without having to visit the location.
Moreover, for projects that will be used be a large group of people, AI can reference socioeconomic data to predict how various aspects of the space will be used and automatically draft plans that best represent and suit all those involved.
- Design autonomy. Perhaps one of the most threatening areas of disruption to architects is the loss of autonomy in the design process, since giving design autonomy to a machine would be synonymous with the loss of their artistic expression.
While this is certainly a valid point, creativity need not be lost; rather there would have to be a change to the way it is expressed. In this setup, the computer would become a partner, rather than a competitor, in the design process. The architect would “plug in” various project variables and goals, and then let the system suggest a range of plans which fulfill these parameters. This means architects would be less focused on the actual design process. In its place, attention would be given to deciding which variables, factors, and goals to emphasise over others in order to reach a certain outcome. The value (and talent) that future architects would bring would thus lie in their ability to arrive at a creative intention that stems from their very human understanding of the space and the people who will use it.
In short, when it comes to the issue of AI in architecture, it isn’t a matter of if; it’s a matter of when. While this technology may not entirely make the architect’s role obsolete, it will, for better or for worse, fundamentally transform it, and architects need to be ready when that happens.