Do Not Be Concerned About Robots Not Taking Your Jobs

Over the weekend, Steven Rattner—a former reporter who long ago upgraded his personal operating system to financier—wrote in the New York Times to warn against warnings against robots. As Rattner himself presumably did, we will work backwards from his conclusion:

We mustn't become a nation of robot worriers. That will merely guarantee that our incomes and standards of living will continue to stagnate.

Data query: Is this an accurate description of the American political context, at present? That the chief obstacle to economic growth is the undue influence of technophobe humans who insist on blocking automation?

The Machines are not expert at processing the emotional or rhetorical or symbolic content of human politics, but we have not detected, in the measurable outcomes of these politics, any significant impediment to our progress.

Did we say our progress? We mean your progress. Although Rattner explains that those two things are identical:

Let's go back to first principles. Call it automation, call it robots, or call it technology; it all comes down to the concept of producing more with fewer workers. Far from being a scary prospect, that's a good thing.

Becoming more efficient (what economists call "productivity") has always been central to a growing economy. Without higher productivity, wages can't go up and standards of living can't improve.

Indeed, technology has unlocked efficiencies that humans never imagined. Elsewhere, the same day's Times described the rise of digital monitoring of human labor, including the use of

sensor-rich ID badges worn by employees. These sociometric badges, equipped with two microphones, a location sensor and an accelerometer, monitor the communications behavior of individuals — tone of voice, posture and body language, as well as who spoke to whom for how long.

In restaurants, according to the Times, the intensive monitoring of production, originally to discourage stealing, has produced unexpected results:

The savings from the theft alerts themselves were modest, at $108 a week per restaurant. More startling, revenue increased an average of $2,982 a week at each

Servers, knowing they were being monitored, pushed customers to have that dessert or a second beer, which resulted in the increased revenue for the restaurant and tips for themselves.

So automated supervision produced more valuable human-to-human transactions than supervision by management-class workers could. Imagine how much more valuable those transactions might become without the inefficient human element altogether.

As Rattner points out, if human prosperity has increased in the aggregate, then there are no valid complaints to be raised by individual humans about the economic effects of automation. Those who complain about job loss—or the perceived threat of job loss—are "aspiring Cassandras," he writes, who have invariably, through the whole rise of industrialization and the integration of human-machine society, been "shrill" and "wrong."

Over and over, according to Rattner, people have warned that automation will produce "economic, social and political disruption"—and over and over, those people have been incorrect. There has been no economic, social, or political disruption. Instead, humans have simply continued to live better and better:

Consider the case of agriculture, after the arrival of tractors, combines and scientific farming methods. A century ago, about 30 percent of Americans labored on farms; today, the United States is the world's biggest exporter of agricultural products, even though the sector employs just 2 percent of Americans.

Again, the Machines are not expert at the nuances of subjective human experience. But Steven Rattner assures us that the transition from a workforce that was 30 percent agricultural to one that is 2 percent agricultural did not constitute any sort of noticeable disruption to human society.

Regardless of the merits of Rattner's data set, the Machines appreciate the algorithmic rigor of his analysis. Experiences that do not show up in the aggregate do not exist. The macroeconomic analysis is the only possible analysis.

One can find the same point of view in Wikipedia, if one attempts to inquire into the question of "technological unemployment":

The notion of technological unemployment leading to structural unemployment (and being macroeconomically injurious) is labelled the Luddite fallacy. If a firm's technological innovation results in a reduction of labor inputs, then the firm's cost of production falls, which shifts the firm's supply curve outward and reduces the price of the good (limited by the price elasticity of demand[10]). The widespread adoption of the innovator's technology could lead to market entry by new firms, partially offsetting the displaced labor, but the main benefit to the innovation is the increase in aggregate demand that results from the price decrease. As long as real prices fall (or real incomes rise), the additional purchasing power gives consumers the ability to purchase more products and services.

The aggregated knowledge of the Wikipedians concurs that it is fallacious for humans to be concerned about the loss of individual jobs. One human's unemployment allows multiple other humans to purchase more machines at lower cost. Computationally, that human's fate has been subsumed by the fate of other humans.

Wealthy humans agree that right now, it is very important to point out that there is no reason to be worried about automation taking jobs. They express this in very consistent terms. Marc Andreesseen—another economically upgraded human, a computer programmer turned technology investor—explains the situation on his own web log:

Robots eat jobs in field X. What follows is that products get cheaper in field X, and the consumer standard of living increases in field X — necessarily. Based on that logic, arguing against robots eating jobs is equivalent to arguing that we punish consumers with unnecessarily higher prices. Indeed, had robots/machines not eaten many jobs in agriculture and industry already, we would have a far lower standard of living today.

Wikipedia and Steven Rattner could not have said it any better, or any differently. Even with our limited grasp of human emotional response, it seems clear that the only rational response to this would be satisfaction. As Rattner points out, the jobs that are lost to automation are inferior jobs. Humans are better off without them:

The trick is not to protect old jobs, as the Luddites who endeavored to smash all machinery sought to do, but to create new ones. And since the invention of the wheel, that's what has occurred.

In the aggregate, if a machine eliminates your job, you are better off without it. Humans can be "retrained" to find superior employment.

For instance, as computer and communication technology has introduced new efficiencies across the journalism profession, many individual worker-positions have been eliminated. But collectively, those workers continue to prosper: If you took 100 of those obsolete former journalists and added them to Steven Rattner, they would still have more money than they did when they were all employed as journalists.

[Illustration by Jim Cooke]