How long until robots rule the job market?

Next time you buy a soccer boot, or any other Adidas footwear, it’s very likely that not only will it have been manufactured in Germany, reversing a decades-long policy of outsourced manual production to South East Asia, but it will have been made by robots.

The company has build a 4,600 square meter prototype factory in Ansbach, deep in the heart of Bavaria called Speedfactory and designed for robot production. By reducing the impact of labor costs, the company has returned home. It makes sense: Adidas’ main rival, Nike, has been reducing the size of its workforce at its plants by investing in robotization, and has announced that it too is developing a fully automated plant, although its plans are less advanced.

But Germany is not the only country headed down the road of automation. Last week, Foxconn, the biggest manufacturer of electronic goods and China’s single largest exporter, said it was shedding 60,000 jobs from one factory, replacing them with robots. The factory, which will now have a workforce of 50,000, is just one of many the company owns, and the rest are due to follow the same path following a deal with Google and that in conjunction with other Chinese companies pursuing the same policies, will see millions of workers laid off.

Adidas and Foxconn insist they do not intend to move over to total automation, and the Chinese government insists that robotization is an opportunity to train more workers, and we can also argue that robots can’t replace us because they lack empathy, but it’s clear that we humans are going to have to find our place in completely redefined value chains, and that applies to all kinds of work, not just assembly line production.

Look around: people are being replaced by machines in Starbucks, inAmazon’s warehouses, to answer students’ questions at Georgia Tech, tomanage investment funds, and make and serve hamburgers in McDonalds, and even to work as lawyers. The very few cases where the opposite is happening, such as Mercedes Benz’ decision to replace robots with humans on some of its assembly lines, are exceptions to do with the difficulty of reprogramming robots for shorter manufacturing runs, a problem that will tend to disappear as machine learning improves.

Machine learning will change automation forever. Robots are no longer about the mechanization of repetitive tasks, but can now take decisions in real time and are better informed and more precise than humans: we see this with self-driving cars, which will potentially reduce accident rates drastically. From the moment a robot can develop certain skills, there is no going back to humans.

The industrial revolution eliminated a great many jobs, but the end result was the biggest generation of productivity and wealth in human history. That’s right, it’s more than possible that robotization will end up creating more jobs than it threatens to destroy. But that transition and skill development will be a major challenge.

(En espa?ol, aquí)