Two University of Oxford professors have caused quite a stir in the employment world with their September 2013 research paper, “The Future of Employment”. Michael Osborne and Carl Frey predict 47% of US jobs are at risk of disappearing over the next 20 years due to continued improvements in computerized applications of artificial intelligence and robotics.
We’ve seen the massive impact of automation on employment before, such as what happened to individual farming field hands in the agriculture industry or to manufacturing workers during the rise of automated equipment in factories. But this time, it could be even more significant because technology isn’t simply able to replace low-skill workers, but also many of what we label as mid-level workers.
In their recent interview on WGBH/Boston radio’s Innovation Hub, the authors explain that recent advancements in computer technology, particularly in machine learning and automation, show that computers are able to perform “subtle judgments” which previously were thought to be the exclusive domain of the human mind. Combined with the incredible power of data storage and retrieval, this foretells the end of a wide range of jobs. Predicted examples include sports referees and umpires (today’s computers are increasingly adept at analyzing at large datasets without bias), truck and taxi drivers (see Google’s fleet of fully automated cars and major automakers’ predictions of mass-produced driverless vehicles in 12 years), and aspects of other jobs, such as paralegals, whose time spent on finding the appropriate passage in case law to support a legal argument will no longer be necessary.
However, those workers whose jobs depend on originality, human perceptiveness and social interaction, such as doctors, lawyers and art designers, as well as consultants and team managers, are expected to continue to be safe. In addition, the numbers of those in gardening and other artisan jobs featuring unique, hand-crafted quality should increase.
Conversely, some jobs that one might think as low-skilled will remain in demand, such as child care and housecleaning. The latter is due in part to the wide range of variability from item to item and home to home (e.g., how tightly should a vase be grasped by a robot in order to clean it without breaking it?). But for work environments that are highly structured with consistent layouts, temperatures, object types, and relative locations — such as hospitals, factories, and warehouses — they lend themselves to automation.
You may be thinking how much this will impact sourcers and recruiters. Those who generate candidates purely through online sourcing are at risk, unless they have mastered the automation tools by which their company gathers data, and thus are seen as internal subject matter experts. Deep researchers whose work transcends recruiting into the realm of competitive intelligence, who need to use a significant amount of creativity and judgment, are probably safer, as are the recruiters who use perceptiveness and human interaction in their candidate engagement process via phone and social media.
The authors would not be surprised if the computerization trend leads to social unrest among many lower-skill workers (the “rage against the machine” scenario), but feel their ability to revolt may already be too compromised with the decline in labor unions and the seemingly irreversible trend towards offshoring (not just in recruiting) for cost reasons. It also may not come to that, as they say, because we have historically found ways to move people into new occupations through education, or evolve enough aspects of their current work to add value beyond technology itself, so jobs “at risk” doesn’t necessarily mean complete elimination. More good news: with the increased availability of free or low-cost education through massive open online courses (MOOCs), the rate of learning should accelerate.
But as the paper’s authors advise American employees in their interview, it’s time to learn to work with computers, not like computers. Sourcers and recruiters should take heed!
