The robot revolution
Huge technological advancements mean that automation could threaten jobs worldwide.
When the Googledesigned computer AlphaGo beat the world number two player at strategy board game Go by four games to one in March1, it was hailed as a milestone for artificial intelligence (AI). This was not simply because a computer had beaten a human; it was also because Go is so complicated, with more possibilities than there are atoms in the universe1, that it was assumed a computer would not be able to see through the millions of potential but pointless moves that human intuition is able to navigate.
AlphaGo’s victory was a tangible sign that machine learning is coming of age. Dramatic improvements in processing power – Moore’s Law holds that the processing power of computers doubles every two years2 – combined with exponential increases in the amount of information available through technology (known as big data), are transforming the types of tasks that can be undertaken by AI and the speed with which it undertakes them. Driverless cars could supplant the trucks and delivery drivers who currently shift goods around the globe; complex legal documents might be checked by intelligent machines rather than trainee lawyers; hedge funds have been using automated trading systems for years, and automation is spreading through financial services – known as ‘robo-advice’.
But robots are also penetrating other, less obvious, areas of industry. In agriculture, they are being tested for tasks such as automated weed control, sensing when fruits and vegetables are ripe, and checking the moisture and nutrient levels in soil3. Merrill Lynch’s report on the robot revolution4 highlights how automation has already had a significant impact on the industry. For example, in 1960 a single farmer in a country using machines in farming could produce enough food to feed 25 people. Today, that figure is 155, and by 2020 it predicts that will have risen to 200. Further automation will be a crucial factor in helping to feed the world’s growing population. Merrill Lynch also says that food production will have to increase by 70% by 2050 to meet the world’s needs.
Others, however, point to the potential costs of the robot revolution. Predictions of the impact this will have on jobs are increasingly gloomy. In January, consultants from Deloitte warned that more than 11 million of the UK’s 30 million jobs are at high risk of automation5, with the wholesale and retail sector at greatest risk. In the US, President Obama’s economic report earlier this year pointed to research indicating that more than 80% of low-paid jobs could face automation.6
Of course, experts have been warning for years that robots will take our jobs: in 1930, economist John Maynard Keynes talked of the emergence of a new ‘leisure class’ and predicted that, by 2030, the working week would have shrunk to just 15 hours as routine tasks were automated.7 But as Andrew Haldane, Chief Economist at the Bank of England, said in a speech last year: ‘Debates on the relationship between jobs and technology stirred again during the 1960s in the US, during the 1970s in advanced economies, and again in the 1980s and 1990s in the UK and parts of Europe. In each case, the prompt was rising rates of unemployment. And in each case, this debate subsided as unemployment rates fell.’
However, he did say there were signs that things could be different this time.
‘Moving into the 21st century, this debate has once again been re-kindled. The prompt this time has not so much been rising rates of unemployment. Rather, it has been the emergence of smart machines, jet-propelled by modern computing. These machines are different. Unlike in the past, they have the potential to substitute for human brains as well as hands.’
He points out that unskilled workers have been the main losers from technological advances. ‘Some displaced workers have not been on the up escalator. They appear instead to have “skilled-down”, typically by taking a job for which they are over-qualified. They have become not unemployed but “under-employed”. Across the EU, rates of under-employment average around 15%. As in the early 19th century, by adding to the unskilled pool of labour, this will have dampened unskilled wage growth.’
But Diane Coyle, Professor of Economics at the University of Manchester, says there is no need to panic. ‘How worried should we be by the speed of change? Not as worried as the most alarmist predictions suggest.’ She points out that electronics company Foxconn recently announced plans to install 30,000 robots.‘ It currently has just 10,000,’ she says, ‘and has meanwhile taken on an additional 100,000 human workers. ‘
Jobs will only disappear when businesses invest in computerised equipment. History provides many examples of technologies spreading slowly because the old ones remain profitable, because additional investment is needed, or because there are other barriers to achieving the necessary level of investment.’
That is not to underestimate the need to prepare workforces for automation, however. Professor Coyle believes that governments must focus on education. ‘An obvious requirement is for more and better education. This may sound banal but little has been accomplished, despite the fact that the policy world has been talking about the need to improve education for more than a decade. Most European countries have systems that deliver an elite with high cognitive skills.
‘The need now is to focus education and training on non-routine – cognitive and non-cognitive – skills that will ensure people are able to complement, rather than be substituted by, new machines.’
1 https://deepmind.com, March 2016
2 www.mooreslaw.org, April 2016
3 www.intorobotics.com, July 2015
4 www.bofaml.com, April 2015
5 www2.deloitte.com, January 2016
6 www.whitehouse.gov, February 2016
7 www.bankofengland.co.uk, November 2015