Will it be five per cent of jobs as McKinsey suggests? Or are we looking at 14 per cent, as predicted by the OECD? Or could AI actually lead to more jobs, as PwC suggests?
The media has an insatiable appetite for sucking up and scatter gunning predictions of the number of jobs likely to be lost to automation. And who can blame them? It is a fascinating subject and one that terrifies and excites the public in equal measure.
But a new Populus survey commissioned by the RSA suggests workers should be less alarmed about the prospect of losing their job to a machine, and more with how it will affect the way they go about their work. This includes how they are recruited, the degree to which they are monitored, and whether their jobs are orchestrated from above.
Automation is a threat to some jobs, but not to work writ large. For all the talk of self-driving cars, checkout-less supermarkets and fully automated warehouses, machines are still constrained in what they can do. Moreover, they tend to eliminate tasks rather than whole jobs, and often expand the capabilities of workers.
In our research we have come across instances of robotics being used by care workers to lift and carry patients, factory machines removing demanding tasks that cause repetitive strain injuries, and chatbots used in airline call centres to partially automate responses to customer queries, allowing staff to concentrate on more complex tasks.
The hysteria that surrounds mass automation risks distracting us from the subtler ways that technology is changing the world of work. And many UK workers agree.
While 32 per cent are concerned about losing their job to a machine, 50 per cent are worried that new technology could lead to them being excessively monitored at work. Meanwhile 39 per cent fear they will have less freedom to work as they would like if technology is used to plan working patterns.
These concerns are not unsubstantiated. Several companies now sell software to aid what has been called ‘algorithmic management’. Hirevue produces AI-powered video technology for interviews – a tool that can detect anomalies in intonation and facial expressions. Another company, Veriato, sells software to log worker behaviour on office computers, including browsing history and even keystroke patterns.
A recent ethnographic study of long distance truck drivers found that electronic monitoring led to them feeling pressured not to take mandated breaks. Elsewhere, a Californian worker took her employer to court after she was allegedly dismissed for uninstalling a phone app that tracked her whereabouts 24 hours a day.
These technologies are not destined to suppress workers. They may even tackle the bias that is inherent in human decision-making, or instil healthier behaviours. What’s more, they could boost flagging levels of productivity, without which we can forget about a return to real wage rises. The point is that we need to make deliberate and informed choices in how we design, develop and deploy this technology. Nothing can be left to chance.
The RSA’s new Future Work Centre aims to lead this conversation. Whether it is calling for limits on the use of intrusive technologies in the workplace, or partnering with tech companies to devise new ethical standards for workplace tools, we will continue to push for these machines to be designed and deployed for benevolent ends.
What we won’t do is get lost in the thickets of automation predictions. The fully automated workplace is a long way off, but the algorithmic workplace is just around the corner.
Ben Dellot is head of the RSA Future Work Centre and associate director, economy, enterprise and manufacturing