The Automation Issue and the Presidential Race

While Donald Trump softens (or curdles, depending on the perspective) on immigration, while folks question whether Hillary Clinton is a particularly articulate zombie, automation looms like a tidal wave. Neither candidate has touched on the issue very much, which is a shame because it will likely begin striking during the next administration.

Automation is the process of making stuff happen without humans doing it. Self-driving cars (and trucks) will be one example of automation, but so will robotic cooks at fast food joints. Automated Teller Machines (ATMs) were an early victory for the robots, way back in the 1960s. But technology has come a long way since then, and I understand they will have ATMs for Bitcoin any day now.

There are some big concerns with automation (which isn’t AI, although AI is an important subject in its own right). One of the biggest issues is how you deal with job losses to machines. Some say this won’t be an issue, claiming that new jobs will be created to take up the slack. These people are both correct and wrong.

They’re right that humanity will figure out how to deal with the startling reality of no longer being needed to sweep floors, drive cars, weld, do taxes, cut hair. They may be right in the idea that we’ll find new employment doing creative tasks, but that remains to be seen. We may invent alternative modes of employment, instead. We may implement a universal income system, or pay people for activities, or shorten the workweek. Where they are wrong is the idea that we shouldn’t worry about it. Looking back at previous transitions, we should definitely have plans in place to protect people from the worst of automation.

The problem is, we’ve never done very well with transitions. We suck at integrating military service members back into society. We suck at integrating ex-convicts back into society. We fail to make adequate plans to rapidly retrain workers losing coal mining jobs or oil jobs when the price drops. And so on.

We fail with the people we like best, veterans. We fail to help women have children, despite the fact that everybody loves babies. There’s no reason to think goodwill has anything to do with it. We’re just bad at helping people transition.

Instead of facing automation, we’re trying to tackle issues that will be mooted by it. The outrageous cost of health care, a real and pressing problem, will likely evaporate as automation makes everything from basic tests to elaborate surgeries cheaper and safer. The main issue there will be regulatory: how to get a body like the FDA in shape to tackle changes to equipment coming at breakneck speeds.

A similar story will play out in education. While college is currently way too expensive, automation will probably lead to the elimination of all but the most historied campuses.

Now, many of these changes will take decades to fully develop, but the first wave of automation is likely to begin in the next four years. While it’s not an issue that will likely sway voters, it would be nice to see some discussion of it. Oh, global warming, too. Automation—figurative tidal waves; global warming—literal hurricanes.