In movie versions of the robot apocalypse, machines become self-aware, decide that humans pose a threat, and then move to exterminate us. A more realistic and immediate threat is that automation in general, and AI in particular, will pose a threat to existing jobs on a scale not seen before. Imagine, for example, the true impact of self-driving cars and trucks on people who earn their living as drivers.

Technology, finance, and manufacturing sectors are enthusiastic about the ability of automation and artificial intelligence to increase productivity. Workers who may lose their jobs due to advances in technology may not be as hopeful.

Unclear Predictions
Some of the fear surrounding automation can be traced back to an especially bleak prediction in an academic study published in 2013, “The Future of Employment: How Susceptible are Jobs to Computerisation?” Its authors, from the Department of Engineering Science at Oxford University, estimated that up to 47 percent of American jobs are at risk of automation. That figure grabbed headlines and garnered media attention.

But as a 2017 Guardian article points out, the study “doesn’t make any explicit predictions such as ‘47 percent of US jobs will disappear.’ It simply says that these jobs are exposed to automation.” For example, the Organisation for Economic Co-operation and Development (OECD), an inter-government agency of developed nations, estimates that job loss in its member countries will be closer to 14 percent.

Clear Implications
Whether it’s 47 percent or 14, many studies done to date suggest that automation will lead to the loss of millions of jobs. A PricewaterhouseCoopers report estimates that men with lower levels of education are likely to be hit hardest, as they make up the vast majority of the workforce in “sectors that are the at the highest estimated risk” for automation. These include transportation, storage, and manufacturing. The report also notes that job growth in other industries isn’t likely to solve this problem without intervention in the form of retraining and education incentives.

Unchecked, higher unemployment in these sectors will decrease social mobility, increase income inequality, and lead to greater degrees of wealth concentration. In addition to the consequences for individual workers, these combined effects will likely reduce economic growth overall. A report from Citi warns: “The [economic] drag reflects wasted potential and a skills mismatch within labor forces and also that more unequal societies are less successful at investing productively for the long term.”

Help Wanted
Despite the clear implications for jobs, there are few discussions, especially policy discussions at higher levels of government, about ways to address the impact of new technology on employment. A relatively positive spin is that large-scale AI platforms will match customers with services, potentially creating new jobs. But many of these jobs are inherently less stable, tend to offer lower wages, and often don’t provide health care and other benefits.

To combat large losses of jobs from innovation, broader thinking will be required. Aggressive, long-term job retraining, led by both government and industry, will be essential to keep workers working. In the United States, “62 percent of executives at companies with more than $100 million in revenue believe retraining and reskilling must be at least half the answer to addressing their skills gap.” However, many of these same executives prefer to hire new employees instead of making corporate investments to training an existing talent pool. Tax incentives and credits for retraining workers could help change that approach.

To maintain social cohesion, keep workers engaged in the economy, and avoid a possible recession, it will be crucial to address the wealth and income inequality caused by the rise of automation and AI. Hard policy discussions regarding changes in education, new workforce training and social programs, large-scale reinvestment in infrastructure, and potential tax increases on high earners are sorely needed — and are not taking place.

Hopefully, the political climate will become more favorable to serious policy discussions about this issue and others. In particular, young voters are more open to thinking about new issues and new approaches. As it becomes clear that technology advancements will contribute to job losses and income inequality, there is more potential for a coalition of citizens and non-government organizations to highlight the unintended consequences of AI.