When most folks think about the impact of robots on the economy, the knee-jerk reaction is anxiety. We worry that robots, and by extension automation and AI, will take over jobs and leave many of us unemployed.
As automation experts, we wanted to take this opportunity to get the facts straight, look at what's happened in the past and what’s likely to happen in the near-future.
Yes, robotics and other technologies are handling more and more low-skilled tasks. But they’re also boosting productivity, creating new, high-skilled jobs – and generating more jobs than they replace.
What everyone is afraid of: Disappearing jobs and redundant workers
This is, at its core, the biggest argument against robotics, AI, and automation: Someday, technology will do everything better than humans, and companies will use it to get rid of their workforce.
We've already tackled this topic in one of our previous blog posts, and discussed why certain terrifying numbers—nearly 40% of U.S. jobs lost to automation, roughly one-third in the U.K. and in Germany by the 2030s—need to be taken with a grain of salt, and why automation may instead end up being a boon for workers and company alike, assuming companies and governments can navigate through the initial growing pains.
In fact, there are a number of studies that point in the opposite direction, and demonstrate that technology adds jobs. Researchers at Deloitte reviewed 140 years of census data from England and Wales, starting at the height of the first Industrial Revolution, and came to a clear conclusion. “The dominant trend is of contracting employment in agriculture and manufacturing being more than offset by rapid growth in the caring, creative, technology and business services sectors,” the authors wrote. “Machines will take on more repetitive and laborious tasks, but seem no closer to eliminating the need for human labor than at any time in the last 150 years.”
When we envision a massive robot-takeover, what we often forget is that creative intelligence is what’s needed to oversee whatever task or project is being tackled. And this is where we, the people, have a huge lead.
Robots are single-minded specialists, not jacks-of-all trades
An easy way to figure out whether the robots are taking over, or not, is to look at the jobs they are most likely to impact. Until now, these have generally been low-skilled, physically-demanding jobs. Large segments of the mining and manufacturing industries, for example, have already been significantly impacted by automation and robotics in many countries. The next wave of jobs likely to be affected involve driving—deliveries, taxi, and long-haul trucking. There's a very clear reason for that. These tasks are all considered predictable physical work. Unlike more complex areas, the labor involved in those sectors—generally low-skilled, exhausting, and/or repetitive in its nature—can increasingly be replaced by machines.
It is one thing to build and program a robot to perform a series of limited tasks on the factory floor or as part of an assembly line. Assembling, finishing, moving, painting, welding, and drilling are all areas where robots excel. But on-the-spot, reactive problem-solving is another matter entirely. What happens when something goes wrong? What happens when an order needs to be stopped or changed immediately? What happens when a task needs to be paused or modified on-the-fly to meet whatever need that suddenly arises? Unpredictable changes demand intervention, the capability to juggle multiple puzzle-pieces and make them fit. That's something modern-day robotics simply can't handle.
Those interruptions, the unexpected variables that constantly pop up during an average day of work, might someday be solved by AI, of course. But we are far, far from it. Yes, AI did beat a professional player at DOTA 2 this year in a simplified one-on-one fight. The popular video game typically involves two teams of five players and another 100 playable characters. The AI could not handle all those variables, however. And yes, the game of Go has now been dominated by AI, but Go is a game of complex pattern-finding in a controlled space. Difficult for humans to master—but the perfect battlefield for computers powered by AI, which can quickly process billions of possible configurations.
These computers are not the fully sentient brains of science-fiction dystopias, however. They're deep-learning, predictive, number-crunching geniuses. Like robots, they are currently great at specialized, focused tasks, but they can't match the reactivity, creativity, empathy or the sheer breadth of the human mind.
The more robots, the more productivity… and the more manufacturing jobs
An interesting article in Harvard Business Review reports on a paper published by London’s Center for Economic Research that concludes that robots are “substantial drivers of labor productivity and economic growth”. The paper’s authors, George Gaetz and Guy Michaels, demonstrate that robotics have boosted productivity growth on a scale similar to the steam engine in the 1800s.
They also point out that countries like Germany and Sweden, which use industrial robots far more than the U.S., lost far fewer manufacturing jobs than the U.S. between 1996 and 2012. As the authors conclude, data suggests that robots are not causing a net loss of jobs. They do, however, appear to create increased demand for higher-skilled, higher-payed jobs – and fewer low-skilled, low-paid jobs.
Even China, which lagged far behind industrialized countries just a few decades ago, is experiencing its own "robotic revolution" – and has been for years. What's happening there is not that different from what is happening in the Western world. Established industrial and fast-developing economies alike are quickly integrating robots in more and more ways. And while low-skill jobs are indeed suffering, skill-based jobs are growing at an incredibly fast pace—fast enough that China desperately needs more skilled workers, rather than fewer workers.
Robots aren't taking over—they're working for us and with us
Increasingly, robots are part of our everyday lives and we’re working right alongside of them without giving them a second thought. There is an ever-growing symbiosis between people and tech that often goes unnoticed. One person in the supermarket oversees five or six self-checkout machines. Marketers and analysists use software to automatically crunch numbers and publish reports. Robotic shelves—such as the ones found in Amazon's warehouses—help humans complete their tasks more easily. People are rarely completely removed from the process; in most situations, they're using robotics the same way you might use Siri on your phone, as assistants that make a variety of tasks easier and more efficient.
People will stick around—and continue to be employed—for two major reasons. The first is that there are many area where people are used to, and prefer to deal with other people directly. The service industry is an obvious one, but that's not all. AI might spot a medical condition better than a medical professional can, but when it comes time to communicate the diagnosis, patients want to talk to another person, not a robot. The medical professional needs empathy, not raw processing power, and must also be ready to answer a variety of unexpected questions.
The second reason? Our ability to create and be aware of the fact that we create. Sure, algorithms have been tasked to try and make art, but the results are mixed, to say the least. The so-called "hard problem" of consciousness is a long way from being solved, and we know that it plays strongly into our creative impulses. Humans are biological decision-makers, but also rule-breakers. We're willing to ignore empirical evidence and established patterns, or react to the unexpected in seemingly illogical yet intuitive ways—all of which are the perfect breeding ground for creative moves. It's why academic papers and books focused on philosophy of the mind are interested in jazz music and the relationships between cognition and theatrical improvisation. Leaps-of-faith and unexpected risk-taking are things machines simply don't do well. In most cases, they don't do it at all.
If AI ever comes close to matching us in those inherently human areas, then maybe they'll also be able to guide robots and progressively replace us. But that is a gigantic if.
Until then, we humans are not going anywhere. Robots will be used the same way we've used other machines and devices for thousands of years, only in more effective ways. This is true of artificial intelligence, too. Deep Blue may have originally beaten Garry Kasparov, but chess AIs have been repeatedly humbled by the centaur model—consisting of humans and machines working together to defeat their opponent, making up for each other's blind spots, and enhancing their respective strengths. Algorithms may help generate content for musicians, but only a human will use that content in a way he or she approves of. Machines can worry about architectural function and the building process, leaving us to focus on form and designs other people respond to. And in our own field, automation, we know that we'll continue to need people who emphasize with and can respond to our customers' needs, and devise new, creative, and often unexpected ways to solve the problems they face.