Why automation is good for job growth

Think Progress Team

Tuesday 1 August 2017

The pace of artificial intelligence (AI) progress is quickening, and with it the number of studies into the effects it may have on society. As robots become more capable, experts have predicted everything from regional depression caused by zero employment to a utopian future in which everybody enjoys a paid-up, work-free life.

The debate about the positive and negative effects of automation will run and run. The truth is probably somewhere in between, but what is overlooked is the very real possibility that autonomous machines won’t just make some jobs obsolete.

They will create them too.

A looming dilemma

Industry 4.0 is well under way. There’s always been hand-wringing at the prospect of a machine making someone’s job simpler. If the job becomes simpler, the person doing it becomes more disposable.

At the outset of the Industrial Revolution, a group of weavers and textile workers became unhappy at the number of mechanised looms appearing on factory floors. They worried that their skills and craftsmanship would become redundant. Instead, one person could oversee 10, 20, 50 power looms working far faster than an equivalent number of human labourers.

Those desperate workers sparked a destructive uprising throughout England, and came to be known as the Luddites. Today the term is attached to anyone opposed to new technology. And while it has come to signify small-mindedness, the original Luddites did recognise one thing. While machines may not have wiped out the need for textile workers in the early 1800s, they did fundamentally transform the nature of the work those people did.

James Bessen is an economist at Boston University School of Law. In a feature in The Economist, he talks about the advent of the ATM and its implications. In America in particular, bank staff worried about redundancy as the holes-in-the-wall took over the mundane task of counting and giving customers their money. Indeed, the average number of US bank clerks fell from 20 per branch in 1988 to 13 in 2004.

But those numbers don’t tell the full story. Over the same 16-year period, the number of high-street branches rose by 43 per cent. ATMs meant banks were saving money and could open new branches according to customer demand, creating more jobs.

Automated number-crunching

To give a more recent example, professional services firm PwC has developed a machine-learning system that predicts the outcome of the UK’s gross domestic product (GDP). During testing, the AI construct was tasked with ‘forecasting’ the last five years’ worth of GDP results, as if the results were unknown. When compared with the actual GDP data from the Office of National Statistics, the AI was 92 per cent accurate.

But as PwC pointed out, it would still need “input and judgement” from human economists to interpret the information and identify any problems.

Amazon has installed over 40,000 robots in its depots over the last three years, yet the rate at which it hires staff has barely changed, according to data on Quartz. This suggests that automation is helping to keep prices down, increasing both sales and the need for people to work in its warehouses.

We’re seeing job fields that didn’t exist 20 years ago, directly because of the explosion of digital technology. If cute, robot carers for the elderly mean there are fewer human carers in care homes, it might also mean more care homes can be opened, following the US bank branch trend.

To prevent a Luddite mindset forming, we need to be sure we’re looking at the whole picture as we embrace new technology. Just as importantly, we need to make sure those who might feel threatened can see what we see too.


Building the next-gen data centre

Where traditional and web-scale apps co-exist