A new study by PwC suggests that AI and related technologies could result in minimal net job losses. But isn’t technology supposed to eliminate the need for human labour? How can automation occur and yet not make a dent in overall job numbers?
The answer is that automation comes in many guises. While the media is fixated on machines that substitute for humans, less attention is paid to those that augment labour. And seldom is it recognised that machines might in effect create tasks, doing work that was never done by a human previously.
At least four types of automation can be distinguished:
#1 Substitution – The most conventional form of automation, substitution involves technology taking on a task that would usually be undertaken by a worker. Examples include self-driving cars, picking and packing machines used in warehouses, and software that can write simple news articles and company financial reports. Occasionally these technologies substitute for whole jobs, but more often they replicate individual tasks that in aggregate make up occupations.
#2 Augmentation – Augmentation expands the capability of workers, allowing them to achieve more and better-quality work in a shorter space of time. Examples range from CAD software used by designers to produce higher quality images, to robotic medical tools used by surgeons to make more precise incisions, through to bread-and-butter search engine technology that allows researchers to find more relevant information. In theory, these technologies take away tasks from workers, but the overall effect is to amplify their abilities (e.g. to complete successful operations).
#3 Generation – As well as mimic what workers already do, technologies can generate tasks that were never done by humans previously (or only by a very small number). ElliQ is a prime example. This ‘elder care assistant’ can remind people to take their medicine, set up video chats with family and friends, and recommend physical exercises depending on how sedentary a person has been. Given that none but the wealthiest of individuals have carers on hand 24/7, this device cannot be seen as encroaching on human turf. Technologies such as this one create work rather than capture it from others.
#4 Transference – A fourth form of automation is transference. This is where technology shifts responsibility for undertaking a task from workers to consumers. Self-service checkouts, for instance, have not done away with the job of processing items through tills. Instead they’ve merely passed on the responsibility to shoppers. Ditto with ticket machines in train stations. Someone is still entering the necessary journey details, except now it’s the passenger rather than a member of staff. This form of automation typically relies on basic software and sophisticated UX and UI Design.
These are not academic details. The reason for unpacking the different permutations of automation is to show that different technologies can bring about very different outcomes for workers.
Self-driving cars might well substitute for thousands of taxi drivers, but a new AI diagnostic kit could augment an equal number of health clinicians as they attempt to detect obscure diseases. Some technologies, whether it’s everyday domestic appliances in the home or sophisticated AI voice assistants, are more consequential for consumers than they are for workers or their employers.
Their impact on wages will also differ. Some may reduce the skillsets required for a role (think of those manning self-service checkouts vs. being a trained cashier) thereby lowering the barriers to entry and reducing the bargaining power of workers. On the other hand, augmenting technologies could improve the quality of a product or service and allow workers to charge more. Farmers might use AI technology to spot blight and boost harvest yields, while a personal trainer could use new monitoring software to create tailored fitness regimes for their clients.
Still, none of this explains how the introduction of technology might lead to a rise in job numbers (PwC suggest AI could spur the creation of 7m jobs over the next 20 years). Three of these forms of automation will lead to at least some job losses, while one – Generation – would at best have a neutral impact on the demand for labour.
Although some believe lots of new roles will emerge to create and maintain machines – think of machine learning engineers or cyber security specialists – the evidence to back this up is sparse. An investigation in 2013 by PwC found that just 6 percent of all UK jobs that year were of a kind that did not exist in 1990, while an OECD study found that only 0.5 percent of the US workforce is employed in digital industries that emerged during the 2000s.
So where will all the new jobs come from? The answer may lie in a little-known labour market dynamic that PwC describes as the ‘income effect’, and which the RSA has called ‘recycled demand’. This is where automation (mainly of the substitution kind) leads to greater productivity and falling prices, which in turn frees up money for consumers to spend in another part of the economy or indeed on the same product or service.
One of the best examples of recycled demand can be found in the transformation of the 19th century garment industry. It’s estimated that 98 percent of the labour required to weave a yard of cloth was automated as a result of new technologies, yet the number of textile weavers actually grew for a period because prices fell and demand was elastic. The same phenomenon is thought to explain the growth of legal clerk jobs in the early part of the 21st century, which happened despite the introduction of legal software that some feared would lead to job losses.
No one quite knows what the next wave of new technologies will do to the workplace. But as these examples illustrate, automation is rarely a straightforward process. Not every machine is a threat to workers, and even when they are, the dynamics of falling prices and increased savings can spur job creation elsewhere. As ever, what matters is making deliberate and informed choices – as investors, employers, educators and policymakers – to ensure the spoils of technology are shared as widely as possible.
With just 6% of workers feeling they have the most to gain from new technologies like AI, clearly we still have a long way to go.
Find out more about the RSA’s Future Work Centre here.
Alan Lockey Fabian Wallace-Stephens
Alongside the moral urgency of the pandemic, the challenges of growing economic insecurity and labour market transforming technologies require a new social contract for work.
What skills will workers need in the future? And how should educators, employers and policy makers respond to shifts in the labour market? We worked with Skills Development Scotland to explore these questions.