Last year, the worldwide management consulting firm McKinsey & Company issued preliminary findings on automation across all industries and most human endeavors. They’re still collecting and collating data from around the world, with a full report expected sometime in early 2017, but last week they released additional findings detailing what workers—as well as owners, policy-makers, and our future robot overlords—can expect in the next decade.
The McKinsey report looked at 800 different occupations. The researchers also determined the technical feasibility of automating more than 2,000 individual tasks across those occupations, developing a more accurate and nuanced picture of what proportion of each job could be automated. It then grouped occupations into sectors and give us a rough idea of everything from how an overall industry might be impacted by automation to how an individual worker can expect their job to change over the next ten years (if they still have a job by then).
The takeaway? It’s not all doom and gloom, but if you want to keep your job—or get a better one—you might want to be a better human.
What Drives Automation Adoption
Occupational changes abound, but job losses might be less significant than many imagine. Some jobs generally seen as low-skill might be far safer than they seem. The good news is that the current findings suggest that automation will leave many occupations at least partially intact. You will, however, need to be flexible to compete with reprogrammable computers.
Most jobs won’t disappear altogether, the report predicts, but many workers in both blue and white-collar positions are likely to find many of their tasks being automated, either freeing up their time for more “human” tasks or reducing the number of workers needed for a given occupation. A version of Murphy’s law applies: anything that can be automated will be, so it’s on workers to make themselves indispensable. Companies can then choose to use their workforce to increase the service level to consumers or reduce their workforce to give consumers a more competitive price.
So what are the human-dependent tasks that are difficult to automate? As you may have guessed, things like interacting with customers and creative thinking. These workers will fair better than their counterparts who perform rote tasks that don’t require a great deal of flexibility, emotional intelligence, or personality.
Yet unskilled laborers shouldn’t despair. Not entirely, anyway. Technical feasibility isn’t the only force behind the decision to automate. McKinsey & Company identified four other factors that drive the adoption trend:
- Cost of automating
- Cost and supply of labor
- Benefits of automation (others than direct cost savings)
- Regulatory and social barriers to automation adoption.
Plenty of people still aren’t happy with the idea of sharing an assembly line or cubicle with a program or mechanical claws.
The general rubric for determining how well an occupation is protected from automation is summed up in one line from the report discussing Education, the sector least likely to face significant automation disruption:
“The essence of teaching is deep expertise and complex interactions with other people.”
Even the most sophisticated piece of software doesn’t possess true expertise. Deep Blue wasn’t an expert chess player; it’s programmers just knew enough about chess to give it rules and goals. Deep Blue used its processing power to map out moves, determine probabilities, and act according to its algorithms. It couldn’t use any sort of psychological advantage (nor was it susceptible to any). True expertise in the game depends on human interaction and creative thought. Deep Blue was good at winning, but it still has blind spots that can’t be programmed.
Automation From Front To Back Of House
Automated cooking robots exist, and can do most basic cooking tasks faster and more accurately than relatively unskilled human cooks, but none possess the creative flair of an executive chef. Automated carts could easily replace front-of-house servers, but they are unlikely to create the atmosphere that most restaurant patrons are looking for from their waitstaff.
A closer look at McKinsey’s findings in the food prep and service world yields several other key insights. The study found that a vast majority of kitchen tasks and occupations are highly susceptible to automation when considered purely in terms of technical feasibility. However, the labor costs for kitchen workers are so low (around $10/hr on average) that the move to automation doesn’t make sense for the thin margins within which most restaurants are operating. For institutions like hospitals, though, where large-scale food preparation is an auxiliary service and creativity in the kitchen isn’t a sought-after goal, automation is a far more attractive and likely proposition.
At the same time, a fine-dining restaurant that has higher margins could see substantial profit increases if it automated most of the kitchen work. A single chef with perhaps 1-2 human assistants can develop new dishes and menus and ensure the final preparation, while the vast majority of the prep work and the cleaning could be machine-driven. The quality of the consumer experience wouldn’t be significantly diminished and the human interaction and culinary expertise behind the meal would still be present.
Fast food restaurants, on the other hand, could go in the opposite direction. Data-crunching software would develop meals based on supply trends, replacing the chef, while kitchen workers would keep low-paying jobs that are cheaper than the costs of automating. The service level at most fast food restaurants being what it is, switching to automated ordering and payment might not be a bad thing either.
Retailing & Retooling
A similar divergence is likely to happen in retail, another industry that McKinsey reports is highly susceptible to automation from a technical standpoint, while noting that human interactions between salespeople and customers are still in demand. Setting up merchandising displays and ringing up customers could easily be automated, but the level of emotional intelligence needed to help someone select a pair of shoes is best left to the people who have walked in a some themselves.
So where does that leave you? We all spend at least some time on tasks that can be automated. If your job is largely made up of tasks that require little expertise, virtually no creative problem solving, and limited interaction, it might be time for some human conditioning, Mr. Roboto.
