JOHN MAYNARD KEYNES always assumed that robots would take our jobs. According to the British economist, writing in 1930, it was all down to 鈥渙ur means of economising the use of labour outrunning the pace at which we can find new uses for labour鈥�. And that was no bad thing. Our working week would shrink to 15 hours by 2030, he reckoned, with the rest of our time spent trying to live 鈥渨isely, agreeably and well鈥�.
It hasn鈥檛 happened like that 鈥� indeed, if anything many of us are working more than we used to. Advanced economies that have seen large numbers of manual workers displaced by automation have generally found employment for them elsewhere, for example in service jobs. The question is whether that can continue, now that artificial intelligence is turning its hand to all manner of tasks beyond the mundane and repetitive.
Fear of machines taking jobs dates back at least as far as the Luddites, a group of British weavers who went on a mill-burning rampage in 1811 when power looms made them redundant. Two centuries on, many of us could face the same predicament. In 2013 Carl Frey and Michael Osborne of the Oxford Martin Programme on the Impacts of Future Technology at the University of Oxford looked at 702 types of work and . They found that just under half of all jobs in the US could feasibly be done by machines within two decades.
Advertisement
The list included jobs such as telemarketers and library technicians. Not far behind were less obviously susceptible jobs, including models, cooks and construction workers, threatened respectively by digital avatars, robochefs and prefabricated buildings made in robot factories. The least vulnerable included mental health workers, teachers of young children, clergy and choreographers. In general, jobs that fared better required strong social interaction, original thinking and creative ability, or very specific fine motor skills of the sort demonstrated by dentists and surgeons.
Life after work: What we鈥檒l do when robots can do anything
Prepare yourself for the future of work with our special report
Others find that list overblown. A recent working paper for the rich-world OECD club suggests that AI will not be able to do all the tasks associated with all these jobs 鈥� particularly the parts that require human interaction 鈥� and only about 9 per cent of jobs are fully automatable. What鈥檚 more, past experience shows that jobs tend to evolve around automation.
According to this more Keynesian view, technological progress will continue to improve our lives. The most successful innovations are those that complement rather than usurp us, says Ben Shneiderman, who founded the human-computer interaction lab at the University of Maryland. Witness for instance the prominence of 鈥渃obots鈥� at last month鈥檚 in Chicago. Such robots are designed to work alongside people, making their work safer and easier, not replacing them. 鈥淭echnologies are most effective when their designs amplify human abilities,鈥� says Shneiderman. They could help us solve problems, communicate widely, or create art, music and literature, he believes.
The weight of expert opinion is behind him. In 2014 the Pew Research Center, a US think tank, . The optimists outnumbered the pessimists.
That鈥檚 not to deny that AI is spreading into some surprising settings 鈥� whether it be organising nightly maintenance on Hong Kong鈥檚 subway system, or helping out with subtle legal research, as does ROSS, an AI assistant built on IBM鈥檚 Watson computer. This suggests that AI could still cause short-term turbulence in the labour market.
One unfolding example is the gig economy. Here AI systems serve up a platter of casual labour to a convenient app for consumers. Examples include the taxi firm Uber and outfits like TaskRabbit, which helps people find casual labourers to complete all sorts of chores. Although the gig economy is still small in absolute terms, a recent study of 1 million people who bank with JP Morgan Chase suggested that (see graph).
In such set-ups, workers are typically considered self-employed contractors, so the company has no obligation to keep supplying work or provide benefits like holiday pay or pensions. That has already led to strikes.
How can we adapt? The answer might simply be to update our social frameworks to reflect the new reality of work. Many countries are considering new regulatory frameworks for the gig economy. In the US Uber and Lyft, another taxi service, face ongoing lawsuits about the classification of drivers as contractors rather than employees. Drivers may vote with their wheels, too: Transunion Car Service, established in New Jersey last year, is an Uber-like taxi business owned by its drivers that promises health and retirement benefits.
Others are thinking more radically about how to reconfigure our whole relationship with work. That speaks to an important point: ultimately we, not AI, are in charge of our own destiny. Given the benefits of work for our health and well-being, maybe we鈥檒l opt not to abolish fulfilling, rewarding work. 鈥淭here will be inequities and disruptions, but that鈥檚 been going on for hundreds of years,鈥� says Shneiderman. 鈥淭he question is: is the future human-centred? I say it is.鈥�
This article appeared in print under the headline 鈥淣ice work if you can get it鈥�
