AI in Architecture

Futures Rambling #96

By Laurie Aznavoorian

One of my favourite questions to ask at the onset of a workplace strategy is – What keeps you up at night? The intention is to gain an appreciation of disruptive changes in an industry that could significantly impact the type of workplace design they should have. Not voyeurism. As I always say, if we were debating the brief for a new film processing plant for Kodak and no one had the brains to ask about digital photography we’d be real buffoons.

The point of completing a workplace strategy is a quest for meaning. By considering a broader range of issues and imperatives when articulating the problem we are solving, we can get beyond the easy picking, rainbows and unicorns items that float on the surface when a brief is being compiled to leverage the physical environment to do more. Consequently, it came as no surprise when the sticky topic of automation rose to the surface when I asking architects this question. It appears they’re scared sleepless by automation.

It’s said technology has created more jobs and industries than it’s destroyed, but recent research from the US suggests mechanized robots, both humanoid or drone types, along with Artificial Intelligence may eliminate 6% of jobs in America in the next five years and it won’t just be low-wage work on the chopping board. Industries that rely heavily on data are at particular risk including: radiology, law and accounting. Some estimate 95% of accountants may lose their job in the next ten years.

According to some creative fields will be safer and therefore one could surmise the fear architects have of being replaced by robots anytime soon is unfounded. When you think about it one could attire a robot in black and give it a groovy haircut, but can a robot bring the passion and common sense architecture requires? I doubt it, but on the other hand, we too work in an industry that is heavily reliant on data which is what radiology, law and accounting have in common.

The challenge we breathers have is a computer outfitted with the right algorithms thinks faster and more accurately than a person, and that is why start-ups like The San Francisco’s Enlitic are doing so well applying deep learning to the analysis of X-rays and CT scans. They’re giving doctors a real run for their money in tests against human radiologists. The Enlitic system was 50% more accurate in detecting malignant tumours and had a 0% false negative rate, humans generally miss 7% of cancers.

They’re trialling another new technology here in Australia which will detect wrist fractures and sadly the early trials are not looking good for humans who are once again being outperformed by computers. I’m not sure why they’re testing that here, perhaps it has something to do with Mr Trump bullying the company into leveraging the skills of unemployed factory workers in the Midwest to read scans? Could be okay as long as they don’t blur their vision by drinking too much Wild Turkey, but then does it matter? Who’ll be able to afford to get a wrist set without health care?

Fellow architects might take comfort in the findings of a 2013 study that found the half of the workforce at high risk of losing their job due to automation were less likely to be in creative fields. The study highlighted architecture as being at a lower risk because it’s non-routine and highly paid, ha ha ha ha, this is of course in comparison to cleaners or burger flippers, not other fields that require a university degree, internship and nasty exam to get a license.

That study was done in 2013, clearly they hadn’t heard of Magenta. This is a project launched by the Google’s Brain team and inspired by DeepDream. Magenta uses machine learning to explore content creation and creativity. Yep, creativity. They’re currently using it to compile music and art. What makes Magenta possible is deep learning or deep neural networks which mimic how the human brain works. Prior to that machine translations were based on algorithms that used statistical methods to guess possible outcomes.

Go ahead, be smug, argue there’s no way a computer could possibly be as creative as a human. Tell that to Android Lloyd Webber the computer that wrote the musical Beyond the Fence, while the reviews weren’t rave: “this show is as bland, inoffensive, and pleasant as a warm milky drink”, it played in London’s West End which is more than many composers can claim. Similarly, Nick Montfort, a professor of digital media at MIT who wrote the novel “World Clock” using a computer and algorithms that outlined characters, locations and actions produced a smash hit.

I guess this means we architects should be scared and pay close attention to the words of Sebastian Thrun, an AI professor at Stanford, who says “we are just seeing the tip of the iceberg. No office job is safe.” Heck computers are already being used to create floor plans for housing projects and any robot worth its metal could probably draw a banquette detail faster than a junior and get the back angle and foam density correct to boot.

This is not great news for those of us trying to put shitty 2016 behind us and doing our darndest to diffuse negativity. By the way, experts say the first thing you need to do to diffuse negativity is to stop worrying and obsessing about things that have happened because it launches a cycle that is very difficult to extract yourself from – a slippery slope. They suggest acknowledging and accepting – that’s what I’m practicing when I repeat to myself ‘the new head of the EPA is a climate change denier – fantastic.’

Another method for forgetting is to be in a worse situation. We are so fortunate to live in a time where there are crackpot companies that do this! Of course it cost more than what the typical architectural practice is prepared to pay per person for an offsite, $950 US, but if you want you can go to Survival Systems and have a worse situation simulated. They’ll stick you under water in a mock plane crash with your co-workers. Imagining drowning with colleagues, that’s one way to forget about the nasty things that keep us up at night.



Aldermanjan, Leslie; “The Year of Conquering Negative Thinking”; The New York Times; January 3, 2017

Barrie, Joshua; “Computers Are Writing Novels: Read A Few Samples Here”; Business Insider Australia; November 28, 2014

Griffiths, Sarah; “Musicals Written by Computer is Heading for the West End… and Based on the Machine’s Calculations, it Should be a Guaranteed Hit” MailOnline; February 5, 2016

Grothaus, Michael; “Bet You Didn’t See This Coming: 10 Jobs That Will Be Replaced By Robots” Fast Company; January 19, 2017

Hyde, Rory; “Architecture in the Coming Age of Artificial Intelligence” Architecture AU

Kelleyjan, Tyler J; “Need Better Morale in the Workplace? Simulate a Plane Crash”; The New York Times; January 7, 2017

Morgenstern, Michael; “The Impact on Jobs – Automation and Anxiety”; The Economist; June 25, 2016

Shani, Or; “Is Artificial Intelligence Going To Take Your Job?” Forbes; August 29, 2016






Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: