Maine’s Workforce Challenges in an Age of Artificial Intelligence

Abstract

Artificial intelligence will improve productivity, expand the economy, and significantly alter many jobs. To accommodate these changes, Maine will have to upgrade workforce skills in a rapidly changing economy. This article recommends policy proposals in response to the rise of artificial intelligence, including (1) training programs for current and displaced workers; (2) revamped postsecondary education programs to provide a wider group of students with the skills necessary in a postindustrial society; and (3) a much closer relationship between government, employers, and educational institutions to develop the future workforce for Maine. The paper also looks at the deliberations about workforce development in the early twentieth century as the United States transitioned from a largely agricultural economy to an industrial one for insights from the past in arriving at educational programs suitable for a postindustrial society. I her inaugural address, Governor Janet Mills pledged to develop a first-class workforce in Maine to address the frustration of employers who cannot find workers and the dissatisfaction of workers stuck in dead-end jobs without the skills for advancement. The governor recognized the complexity of this challenge by noting that technological innovation will radically alter the way Maine people live, learn, and work, which is why she announced the formation of an Office of Innovation and the Future (Mills 2019). Over the next decade, the acceleration of the adoption of automation, including artificial intelligence and robotics, will likely exacerbate the mismatch the governor identified between workers and employers. While new technologies will increase productivity and economic growth, they will inevitably result in many jobs requiring new skills and some workers requiring new occupations. This article explores the policy implications for Maine’s workforce from the rise of artificial intelligence and discusses (1) effective transition programs that train employed and displaced workers in new job skills; (2) postsecondary education programs that provide a wider group of students with skills necessary to work productively in a postindustrial society; and (3) a much closer relationship between government, employers, and educational institutions to develop Maine’ future workforce. The article also examines the deliberations over workforce development in the early twentieth century when US society transitioned from a largely agricultural economy to an industrial one, as a way to highlight what we can learn from that experience as we transition to a postindustrial society. THE SHIFTING NATURE OF WORK T globalization, digitization, and automation of the US economy has increased productivity, held down inflation, opened up new markets, created moreefficient supply chains, expanded the economy, and given consumers quality products at lower cost. But these advantages came at the expense of thousands of American jobs, particularly good manufacturing jobs. In the 1980s, manufacturing represented Maine’s largest employment sector, but the number of manufacturing jobs in Maine has been cut in half since that time (Maine DOL 2013). Over the last half century, the United States has shifted from an industrial manufacturing economy to a skilled service economy with an emphasis on finance, education, health care, and information technology. These service fields require higher skills than the jobs in the old industrial economy. Now, two out of three jobs require training beyond high school, often a twoor four-year degree (Carnevale 2016). Most of the job losses from trade took place in the first decade of this century; more-recent losses are coming from automation. Much of the political rhetoric during the 2016 presidential campaign focused on trade and immigration as the culprits upending industries and jobs. But job losses from trade have been eclipsed by the far MAINE’S WORKFORCE CHALLENGES MAINE POLICY REVIEW • Vol. 28, No. 1 • 2019 12 greater job losses from automation and artificial intelligence—impersonal forces that are harder to scapegoat, negotiate with, or turn back. A McKinsey study found that 45 percent of work activities could be automated with current technologies (Chui, Manyiku, and Mirenadi 2015). Increasingly, many routine jobs are becoming casualties of automation: robots have replaced factory workers, ATMs have eliminated bank tellers, and kiosks and scanners have jettisoned clerks. In the future, high-skilled jobs will no longer be immune to automation as computer algorithms augment or even replace some professionals in law, medicine, and finance. Ironically, today’s workforce shortages may well accelerate the adoption of automation. Automation drives down labor costs and has been adopted in factories not only in high-cost regions like the United States, but also in lower-cost regions like China and Mexico, ensuring that the combination of automation and low-cost labor will make those regions competitive with any region with high labor costs. In many areas of the economy, employees of the future will be those who manage automated technology. Artificial intelligence has the power to change the nature of work for many people, but the pace of adoption and the extent of the disruption are still the subject of debate. A rapid adoption of autonomous self-driving vehicles, for instance, could dramatically displace millions of workers, but a more gradual and partial adoption, especially in a growing economy, will have far less impact on drivers. The Case for Significant Disruption As an extreme case, the historian Yuval Noah Harari offers a dystopian vision that imagines artificial intelligence doing to workers what the automobile did to the horse and buggy. In his analogy, many of today’s workers will likely be the horse that lost its job and never found another useful function in a motorized economy (Harari 2016). Harari reasons that the power of artificial intelligence is unlike past work-saving innovations that took over manual tasks and pushed human work up the conceptual ladder. Harari makes the case that innovations from artificial intelligence are radically different from past innovations. He argues that since humans can perform only manual and conceptual tasks, there is nowhere for us to go when artificial intelligence outperforms humans on both manual and conceptual tasks. The algorithms in artificial intelligence already dominate the stock market, social media, and search engines. Their capacity to take into account millions of variables when making split-second decisions will likely extend to the rest of the economy. Harari sees the possibility of artificial intelligence eventually taking over the work of millions of people, creating what he calls a “useless” class. What will societies such as Bangladesh do when we can produce products cheaper through automation than they can produce with their low-cost labor? People in Silicon Valley are already envisioning the need for basic universal income because many people will be unable to find employment in an economy dominated by artificial intelligence. Harari makes clear that he is not making a prediction or setting a timetable, rather he is dramatizing the power of artificial intelligence. The Case for More-Moderate Disruption More-moderate positions on the impact of artificial intelligence also predict dramatic change although without massive displacement of workers. A report by McKinsey Global Institute forecasts that 60 percent of current occupations could automate at least 30 percent of their work and as much as 33 percent of work activities could be displaced by 2030 (Manyika et al. 2017). The report predicts that up to 14 percent of workers will have to transition to new occupations. The timetable for these disruptions, however, will depend on the pace of adoption and whether multiple sectors in the economy adopt the technology simultaneously. But even with the adoption of artificial intelligence, the report paints a possible positive scenario of jobs growth if economies expand, incomes rise, use of health care increases with an aging society, and other economic stimuli (such as spending on infrastructure, energy, and climate change mitigation) take place. In contrast with Harari’s position, the argument for more-moderate disruption adopts the traditional view that while innovation disrupts the economy as some jobs and industries become obsolete, new technology ushers in new opportunities and jobs. Both cases, however, share the view that technology will disrupt the economy and that employers, employees, governments, and educational institutions will have to respond to the change. The McKinsey Global Institute projects continuing decline in jobs associated with physical and manual skills and basic cognitive functions and a shift toward higher cognitive and technological skills, as well as MAINE’S WORKFORCE CHALLENGES MAINE POLICY REVIEW • Vol. 28, No. 1 • 2019 13 increased importance in social and emotional intelligence (Manyika et al. 2017). Artificial intelligence will increase the importance of entrepreneurship and problem solving in technologically rich environments (Bughin et al. 2018). The Brookings report on Automation and Artificial Intelligence forecasts 25 percent of US employment will face high exposure to automation over the next few decades and more than 70 percent of current tasks at risk of substitution. Another 36 percent of the workforce will experience moderate exposure, which means that more than six out of ten jobs will see some—or even all—tasks associated with that job taken over by automation. The lowest-skilled workers will be most vulnerable, but they will not be the only ones affected. New technologies will create an urgent need for educational programs that continually upgrade workers skills, or we risk having large portions of the population ill equipped for the work of the future (Muro, Maxim, and Whito