So
the robots are coming for our jobs, are they? Yawn. That’s such an old
story. Goes back to Elizabeth I and the stocking frame, if my memory
serves me right. Machines have been taking our jobs forever. But
economists, despite their reputation as practitioners of the “dismal
science”, have always been upbeat about that. Sure, machines destroy
jobs, they say. But hey, the new industries that new technology enables
create even more new jobs. Granted, there may be a bit of “disruption”
between destruction and creation, but that’s just capitalist business as
usual. Besides, it’s progress, innit?
We have now lived through what one might call Automation 1.0. The paradigmatic example is car manufacturing. Henry Ford’s production line metamorphosed into Toyota’s “lean machine” and thence to the point where few humans, if any, are visible on an assembly line. Once upon a time, the car industry employed hundreds of thousands of people. We called them blue-collar workers. Now it employs far fewer. The robots did indeed take their jobs. In some cases, those made redundant found other employment, but many didn’t. And sometimes their communities were devastated as a result. But GDP went up, nevertheless, so economists were happy.
Now we’re embarking on Automation 2.0. This is largely driven by technologies employing machine learning (ML) and big data, what we misleadingly call “artificial intelligence”. The types of job it targets are different from those addressed by Automation 1.0: they have some cognitive content but also a lot of routine. We call them white-collar jobs. And the new machines can often do them adequately or well.
"One of the things we are learning about digital technology is that it has become an amplifier of inequality"
Which may explain why people are beginning to be more agitated about
the widespread deployment of the technology than they ever were about
Automation 1.0. Early studies of the likely impact were pretty alarmist.
For example, in 2013 Carl Frey and Mike Osborne in Oxford predicted that nearly half of the 700+ job categories used by the US Bureau of Labor were vulnerable. In their book The Future of the Professions,
Richard and Daniel Susskind foresaw a radical impact on professional
experts such as accountants, lawyers and management consultants. The
word got out that maybe a lot of high-status employment might be
vulnerable to automation and people began to fret about the
hollowing-out of the middle class. After all, there isn’t a functioning
democracy without one. (Strangely, the same liberal democracy has
apparently been able to survive the unemployment of millions of
blue-collar workers. But we will let that pass.)We have now lived through what one might call Automation 1.0. The paradigmatic example is car manufacturing. Henry Ford’s production line metamorphosed into Toyota’s “lean machine” and thence to the point where few humans, if any, are visible on an assembly line. Once upon a time, the car industry employed hundreds of thousands of people. We called them blue-collar workers. Now it employs far fewer. The robots did indeed take their jobs. In some cases, those made redundant found other employment, but many didn’t. And sometimes their communities were devastated as a result. But GDP went up, nevertheless, so economists were happy.
Now we’re embarking on Automation 2.0. This is largely driven by technologies employing machine learning (ML) and big data, what we misleadingly call “artificial intelligence”. The types of job it targets are different from those addressed by Automation 1.0: they have some cognitive content but also a lot of routine. We call them white-collar jobs. And the new machines can often do them adequately or well.
"One of the things we are learning about digital technology is that it has become an amplifier of inequality"
Last week, Oxford Economics, a high-end consultancy, unveiled the findings of its latest peek into the future in a report entitled How Robots Change the World. Like most of these inquiries, it foresees a “great displacement” of employment by Automation 2.0. But this displacement, the report says, will not be evenly distributed around the world or within countries.
“Our research shows,” it says, “that the negative effects of robotisation are disproportionately felt in the lower-income regions of the globe’s major economies – on average, a new robot displaces nearly twice as many jobs in lower-income regions compared with higher-income regions of the same country. At a time of worldwide concern about growing levels of economic inequality and political polarisation, this finding has important social and political implications.”
The report makes for pretty sobering reading. It claims that, on average, each additional robot wipes out 1.6 jobs. In lower-income regions of the world, each machine displaces 2.2 jobs, but only 1.3 in higher-income areas. The researchers compiled a “vulnerability score” for different regions in five countries – the UK, USA, France, Germany and Japan. The resulting maps confirm that employment in poorer areas will be hit harder by automation. “The regional inequalities that exist within countries,” it concludes, “such as England’s north-south divide, could be exacerbated by the rise of the robots.” The report notes that this trend “has important implications for policy design in advanced economies pursuing international competitiveness through automation”.
You bet it has. One of the things we are learning about digital technology is that in almost every area of its deployment it has become an amplifier of inequality. The tech companies that control it employ almost no one in comparison either to their profits or to non-tech companies of comparable size and scale. Volkswagen, for example, employs nearly 656,000 people worldwide. As of December 2018, Facebook employed only 35,587.
Likewise, tech companies pay derisory amounts in tax in the territories where they make colossal profits. In 2018, for example, Amazon paid nothing in US federal income tax on more than $11bn in profits before taxes. It also received a $129m tax rebate from the federal government. Automation 2.0 is likely to be very profitable for the companies that deploy it, but it’ll be governments that will be left to pick up the pieces. Some progress.
No comments:
Post a Comment