Algorithm vs. Architect: Will Machines Design the Cities of the Future?

Published on May 3, 2016 | by Lidija Grozdanic
Coded City

 

Discourse concerning automation and artificial intelligence has really come into the limelight in the last few decades, and we have seen both optimistic predictions and dystopian visions of the future. Are machines taking our jobs away, or are they helping us move away from menial tasks and pursue more creative endeavors? Do architects have any reason to panic?

Technology-driven development is undeniably changing job markets across the world. Taxi drivers, clerks and bookkeepers, among other professions, are likely to completely disappear in a matter of decades. Apps and robots are replacing repetitive activities: drones can survey crops and deliver packages, and self-driving cars are expected to dominate city streets within our lifetime.

Computers are reaching a level of sophistication that allows them not only to perform mechanical, repetitive tasks but also operate on a higher cognitive level. While optimists see this change as an opportunity for economic growth and innovation, others are voicing their concerns over its social implications, such as economic inequality and unemployment.

According to a research conducted by the World Economic Forum, over 5 million jobs will be lost to automation by 2020, affecting mostly white-collar workers in administrative and office jobs.

Carl Benedikt Frey and Michael A. Osborne from Oxford University, authors of The future of Employment, have created a table that ranks professions by probability of computerization, according to which people working in administrative support, factories, retail and service industries are most likely to be replaced by computers.

In the AEC industry, model makers, technicians, drafters and urban planners are at a much greater risk of disappearing compared to architects, interior designers and civil engineers. Frey and Osborne give architects a 1.8% chance of being automated, compared to a 93.5% chance for accountants, a 96.3% chance for restaurant cooks, and 86.4% for real estate agents.

A new McKinsey report suggests a somewhat different impact of automation on job prospects. Instead of thinking of automation in terms of it eliminating entire occupations, the report suggests that many jobs will be redefined rather than eliminated—at least in the short term.

During the 1950s, before architecture and engineering firms shifted from having shared computers to individual machines, companies transitioned from mainframes to ‘minicomputers’. Due to steep prices, these machines had to be operated by specially trained workers. Designers would bring their work to the CAD Department and wait for hours to receive plotted outputs, which would then be turned in for revision before being returned to the designers.

This back-and-forth seems cumbersome by today’s standards but back in the day this new system produced work that had previously taken 10 or 12 people to complete. It also eliminated and introduced job profiles through a process that still continues to evolve.

In the last 15 years, employment for architects has increased by 25%, with significant differences between specific job profiles. Computerization has generated some new occupations within the field of architecture, in particular the mainstream profession.

While employment opportunities have diminished for architectural drafters, other profiles such as BIM specialists, digital making technologists and communication managers are experiencing growth. The chances of the architectural profession disappearing any time soon are slim, but long-term predictions imply that the role will most likely be redefined.

Another important question is: to what extent can machines substitute higher cognitive processes required in designing a building?

California-based company Aditazz uses methods derived from the semiconductor industry to create algorithms that simulate hundreds of viable design options for state-of-the-art healthcare facilities, completely eliminating the need for countless drafts and revisions. They developed an automated tool that offers the possibility of exploring a number of designs and operating options in a fraction of the time it takes traditional methods, allowing for intelligent trade-offs based on real data.

Furthermore, these can be translated into instructions for a robotic system that casts construction components. Aditazz conducts operational simulations that show performance metrics, bottlenecks of efficiency, utilization of key resources, patient wait time, etc.

Amazon and Google are also working on creating automated design solutions. Engineers at Google developed a platform with online planning applications that standardize and automate design and construction processes, promising to save up to 50% in construction costs, and cut project development time by up to 60%. The project, initiated in the Google X laboratories, was later rebranded as Flux.

Are these developments a threat to architects? Not likely. Both Aditazz’s and Flux’s automation tools are solutions based on machine-man combinations. Technology is nowhere near supplanting architects. It can address most of the quantitative aspects of architecture, even some of its qualitative characteristics, but when it comes to dealing with context, taste, aesthetics and negotiation, computers lack the higher levels of adaptability and superior cognitive skills needed.

***

Architects don’t need to worry about losing their jobs or becoming obsolete, at least not in the short term. Even in the long term, it seems that technology actually redistributes labor from areas susceptible to automation into other sectors, eliminating certain types of jobs and boosting others.

Economists at the consultancy Deloitte released a study showing that, in the last 140 years, technology has actually created more jobs than it has destroyed in England and Wales. This cycle of eliminating and creating jobs can be painful, but it will hopefully occur at a slow enough pace that will allow workers to adapt and grow.

Comments