The robots are coming... but what will they do when they get here?

In this blog post, CROWC's Dr Fergus Bolger considers the possible consequences of increased automation in the workplace.

Growing pains

The rapid recent developments in AI have raised questions about the consequences for future employment (see, for example, reports from McKinsey, Forbes magazine, Deloitte and others). These reports suggest both winners and losers but, in contrast to most previous technological innovations, most disruption is expected amongst relatively high-paid knowledge workers such as data analysts and content creators.

However, there is a second technological revolution in the wings – advances in robotics could also have a profound impact on labour, and in this case the brunt of change is likely to fall on more traditional physical work, particularly in manufacturing.

Symbiosis

Of course, the ongoing generative AI revolution is linked with the putative robotics one – robots can do more things if they are smarter (e.g., autonomous drones on the battlefield) – so advances in AI, and computing more generally (e.g., edge computing), are drivers of advances in robotics.

Other drivers are technological developments in materials and engineering that could potentially transform robotics (e.g., battery innovations, artificial muscles, and new additive manufacturing techniques). Despite these drivers, most of the technology applied to robotics remains in the lab, with the majority of commercial robots showing only small advances in functionality from robots of the 1950s.

The Advance Research and Invention Agency (ARIA), established in 2022 by the UK government to fund “high-risk, high-reward” research, is hoping – through one of its first major programmes – to kickstart the robotics revolution in the UK by putting millions into moving new robot-relevant tech out of the lab. But to what end?

In the funding call for the current programme, the stated goal is to “release the bottlenecks in robotic dexterity and create vastly more capable and useful machines” that will lead to increased productivity at the same time as the number of able-bodied workers declines.

The stated belief underlying this goal is: “A world where robots free humanity from physical labour is not only possible... it is imperative if we wish to boost longevity and prosperity”. I think this belief needs examining in more detail.

Is this what we really want?

First, do we really want to “free humanity from physical labour”? There are well-established benefits to exercise for healthy minds and bodies. Also, many people enjoy working with their hands (e.g., the growth of “Maker Culture”).

So, perhaps we don't we want to eliminate physical labour altogether, just take the unpleasant parts out – this would imply robotics that are integrated with, and enhance, human activity (e.g., wearables, prosthetics, exoskeletons) rather than autonomous machines as is commonly envisaged. Thus, we need to consider what kind of physical labour is harmful, and which is not, and also, most of all, what people enjoy doing and what they do not.

Secondly, I question whether “prosperity” is the right goal here – yes, we want to boost productivity (i.e., improve efficiency of production) and, through this, quality of life, but do we really need more material wealth? Both productivity and well-being may be related to prosperity, but not necessarily. For example, increased productivity could mean we have the same stuff in less time, while increasing the amount of stuff we have has a marginally decreasing effect on well-being.

So, we need, as a society, to seriously consider what we want to maximise with robotics and other technological advances. AI and robotics seem to have great potential to add to human capabilities as well as to replace them, and I propose that we should concentrate our resources on the former rather than the latter.

Accepting ARIA’s premise that we need more robots, in addition to needing a clearer plan of what we want robots to do for us than ARIA seems to have, we also need to think about how we want to manage the robot revolution. For example, ARIA argues for ‘ubiquitous’ robots and, while recognising that to be ubiquitous, robots also need to be affordable, there is no mention of sustainability.

Ubiquitous robots implies using lots of resources to make them – where will these resources come from, and how will making robots impact other important goals such as reducing poverty and negative effects on climate and biodiversity? For example, will next-generation robots be competing with electric cars for minerals that may be unsustainably and/or exploitatively sourced?

Stopping to think

A robot revolution in the near-to-medium term seems probable, particularly given the efforts of agencies such as ARIA – already influential in the US through various advanced research project activities (ARPAs), and central to Mario Draghi’s recent blueprint for a more competitive EU.

Such a technological revolution, like others before it, will likely bring both costs and benefits to society. It seems to me that we have the opportunity to actively manage this particular revolution to maximise the societal benefits and minimise the costs.

As it currently stands, although not completely unprincipled, the strategy is incoherent. Pumping lots of public money into developments of technologies potentially useful for making ‘better’ robots will most likely lead to a transfer from the public coffers to the hands of tech companies and their bosses – and billionaires such as Bezos and Musk already have their fingers in the robot pie.

Dr Fergus Bolger, Senior Lecturer, School of Management