Main image of article Maybe Automation Won't Kill as Many Jobs as We Think
Will the rise of artificial intelligence (A.I.) bring about the employment apocalypse? That’s been the dire prediction of a number of pundits and economists. The reality, however, may turn out altogether different; a new report from the Organization for Economic Cooperation and Development (OECD) suggests that some 10 percent of jobs in the United States are ultimately vulnerable to machine-driven automation—a significant percentage (especially if it’s your job on the line) but much lower than some of the more dour estimates. The report calculates an occupation’s risk of automation by examining the “bottlenecks,” or the aspects of the job that would prevent the machines from taking over a job from a human being. These bottlenecks include social intelligence (i.e., the aptitude to “negotiate complex social relationships”), creativity and complex reasoning, and the ability to perceive and manipulate an uncertain work environment. In line with other papers on the topic, the OECD fears that jobs in agriculture and manufacturing are highly automatable, whereas several classes of professionals are somewhat more protected. “The occupations with the highest estimated automatability [sic] typically only require basic to low level of education,” read the report. “At the other end of the spectrum, the least automatable occupations almost all require professional training and/or tertiary education.” Another worker category in automation’s cross-hairs: younger workers, who are often assigned the kinds of tasks and jobs easily taken over by a machine capable of executing routines. On a macro level, that could make things a little rough for youth employment over the next few decades. Even if this report puts the risk of job loss due to automation as somewhat lower than other sources, machine learning and A.I. really will have an impact on the employment market over the next few decades. The best way to prepare is to improve on those “bottleneck” abilities—those soft skills that are difficult to routinize or replicate. In any case, the report itself is well worth a skim.