Addressing automation pt. 1: Separating fact from fiction
By Ehud Paz and Kelsa Trom
As artificial intelligence and other emerging technologies rapidly progress, there’s been intense, sometimes fearful, discussion around a simple, honest question: Will workers be replaced? In this three-part series, we will address the realities of automation from the perspective of an impacted worker, and shine a design lens on how we might shift the conversation to be more inclusive and productive. (And none of these articles were written by a robot, we promise.)
—
A few years ago, I spent some time in a processing center at a credit card company. I was working with a team that processes credit applications and customer service requests to help improve their customer experience. I met a lovely employee there. Let’s call her Amy.* Amy was the expert in her department. She knew every system, every template, every acronym. While she whipped new requests through a labyrinth of various checks and balances on papers, Post-its, command line terminals, and gray interfaces without missing a beat, I couldn’t help but wonder how much of our economy is still reliant on managing these dated and complex systems. We started thinking about how to make her job easier and create better experiences for her customers:
- Software systems that could better connect all the different data sets and applications she uses …
- Natural Language Processing technologies to capture and make sense of faxed paperwork and forms she receives …
- Most of the remaining paper forms she used could be digitized …
- A set of technologies called Robotics Process Automation (RPA for short) which could write rules to automate many of her tasks with a greater degree of consistency and fewer errors…
…and so on.
Then we got uneasy. If her employer acted on all of these things — what work would Amy have left to do?
These are not quite the technologies central to the plot of an episode of “Black Mirror” or a feature story on the nightly news. These are the technologies very quickly and quietly automating workplaces in back offices all over America. Amy had no idea these technologies were a potential risk to her economic security — in fact, they felt benign. And while the usual spin we apply to automation — that robots will take over repetitive or boring tasks, freeing people up to do new, interesting work — is true (in theory), how realistic is it for workers like Amy?
Movies like “Her” and an assortment of take-your-pick critically acclaimed sci-fi series set very high expectations for what AI could be. Some of the more current and less fantastical examples — footage of robotics-powered car factories, Amazon’s hyper-efficient distribution centers and even our trusty Roomba — heavily influence our perceptions of automation technologies. Every week, our inboxes and newsfeeds fill up with headlines forecasting varying predictions of doom and gloom and quizzes to see the likelihood of whether a robot will steal our job.
While many of the marvelous technological advances we hear and read about are very real and very much on the horizon, the onslaught of hype is making it harder for workers to understand who is actually at risk and, more importantly, what we can do about it. A 2016 study found that two-thirds of Americans expect that robots and computers will do much of the work currently done by humans within 50 years, but 80% of them don’t believe such automation will happen to their job.
And therein lies the problem: how do we help people not only understand these new realities, but offer them the tools to shape their futures?
Late last year, I joined a research program studying the impact of automation on office workers. One of our teammates spoke with a few office employees in a training program while their company in Pennsylvania automated many of their tasks. They told us stories of executives showing the company presentations of self-driving cars and sophisticated AI assistants. This only resonated with a few; for most, it felt abstract and intangible. It was almost like watching Steven Spielberg’s “Minority Report” and asking workers to place themselves in a fictional future.
They didn’t think some all-powerful robot was going to be able to do their job, but at the same time they felt scared of something they did not fully understand and something they could not see. Their company was also struggling to figure things out — executives could not predict how the new technologies they were implementing would impact headcount. In many ways, for everyone, it was like trying to diagnose a complex health issue: uncertainty felt worse than concrete bad news.
The automatable activities associated with office occupations are spread out across many jobs, titles, industries and regions. This makes it really challenging for someone to understand if their work is at risk. In the U.S., there are more than 19 million office and administration jobs. Let’s look at bookkeepers and accounting clerks, as an example. Some of the day-to-day tasks in these roles include processing payroll, customer invoicing, processing credit and checks, testing for compliance and fraud, and running financial reports. All of these tasks can currently be automated, and some of the projections associated with these jobs are not promising. The McKinsey automation study projects an 86% automation potential; and the Frey and Osborn study projects a 98% automation potential. There are about 1.7 million people in accounting clerk and bookkeeping positions spread out across many industries. In theory, if many of these jobs were to be eliminated, the impact would be spread out far and wide. The next “rust belt” would be all around us.
Here is where we might fall into the fallacy of painting with a broad brush.
Many automatable clerical tasks are not neatly tied to specific occupations, and many back office roles can vary greatly in responsibilities, especially outside of the United States. In many environments, these types of jobs rely more on human-to-human interaction than the routine processing we assume. For example, in the bookkeeping, accounting and auditing clerk occupations, an OECD study found that nearly 75% of all employees perform their jobs with some form of group work or face-to-face interactions.
If I think back to my interactions with Amy, she shined in her ability to manage relationships with clients and vendors — a critical business function for her group.
So, how do “back office” workers know if they are in jeopardy and know how to get ahead of change?
First, we need to deconstruct the traditional notion of a job and focus on its components — skills and outcomes — in new and inventive ways. Programs such as Skillful by the Markle foundation are doing inspiring work to shift us toward a “skills-based economy.” This begs the question — if we have been cataloguing human skills, and machines are learning to catalogue their own skills, can we start bringing together a common framework for human and machine skills? This could be very powerful to help policy makers, business leaders and potentially everyday people identify: (a) where skills are becoming less or more relevant, and (b) where there are opportunities to connect human and machines to create new value and solve different problems. At Fjord, we believe the collaboration between machine and human is the most effective scenario. A report by Accenture indicates the same, that the combination of human skills and AI may be more promising together, projecting that investment in AI and human-machine collaboration could boost company revenues by 38% by 2022.
Second, organizations and workers need to think about how we diversify our employee experiences. Gone are the days of sticking to a job, climbing the ladder and retiring with your gold watch. Just as we found ways to diversify our finances to navigate uncertainty, we need to think about our work and life experiences as a portfolio that needs continual rebalancing. We may need to routinely ask ourselves the question: am I professionally overexposed in a certain area? From a business perspective, in what has been dubbed the “talent war” for the past two decades, it’s been harder and harder to attract, train and retain people with the right skills. This could be an opportunity for businesses to find something new in their people versus relying on finding new people alone. Businesses can take the lead on being more creative with how we think about diversifying employment experiences and enabling continuous learning.
Third, those of us in the design and technology industries, in particular, need to start taking more responsibility for addressing the potential downstream consequences of our innovations. There is already some major investment by the likes of Google, Microsoft, Amazon and Accenture in skills training, which is a great start, but it’s not enough. We need to think about how we can better empathize with people outside of our industry bubble so we can speak in a language that is relatable, personal and productive. Our industry can sometimes be disconnected from the people we impact, and we can be more inclusive in designing a way forward. Less paternalistic talking down, and more respectful talking with. (More on that later in the series.)
Sometimes I replay that day with Amy and her team and I wonder if we both had access to the same facts how things might be different. I look around my office and wonder what we are taking for granted and where we may have blind spots. When will our work need to drastically change and will I personally be in a place in my life where I can handle the change?
How this fourth industrial revolution will impact people like Amy and everyone else is unclear. One thing, though, that is certain: we can’t just sit and wait to see if all of the predictions come true. We need to work together to shape a future where we have better transparency to identify change and more flexibility to adapt to it.
Technologies will, and should, continue to become obsolete. Human beings should not.
*Name has been changed.
Ehud Paz (his friends call him “Eudi”) is a strategist at Fjord New York, focused on designing for the future of work and education. Kelsa Trom is a content designer with a background in learning and development and a passion for interdisciplinary growth. She is based in New York.
Illustration by Annie Lin, a visual designer at Fjord Austin.
