Capabilities provided by artificial intelligence (AI) systems and related technologies such as robots and driverless vehicles could provide a 10% boost to the UK economy by 2030, but there are concerns about the large number of jobs that could be lost in the process.
The conclusion arose in a report published by the Department for Business, Energy & Industrial Strategy (BEIS) on the potential impact of AI on UK employment and the demand for skills, based on research by consulting firm PwC.
Focusing on two main questions relating to the impact of AI in the economy, the researchers used a mix of expert workshops and publicly available government data on employment and job displacement to understand whether AI and related technologies would continue to trigger significant changes in labour. It also discussed the scale and form of that AI-led disruption going forward.
Regarding the effects of AI on employment, the researchers noted that AI and related innovations should be very positive in terms of increased productivity and income levels. On the other hand, the study predicted a “broadly neutral long-term” impact alongside “significant structural shifts” in labour patterns due to AI and related tech in the next 20 years just as in other waves of technology advance throughout the course of history.
The report predicted that senior level and professional roles would be positively affected by AI technologies that build up steadily over time, while net employment effects would tend to be more positive for higher-paid occupations and graduates than for lower educated groups.
Age and gender variations were not significant in terms of net employment effects predicted in the research, but it noted that younger workers in entry-level jobs involving routine tasks may be more vulnerable to automation.
Moreover, the report pointed out that clerical and routine manual work might see negative impacts of advances coming through in the next five to 10 years as robotic and automation processes evolve across all sectors of the economy.
The researchers said not all individuals performing manual tasks would be affected in the same way, however, with roles such as truck and taxi drivers likely to see negative effects of AI over the longer term – approximately 20 years – once driverless vehicles were adopted at scale.
The greatest gains in net employment brought by AI over the next two decades would be seen in the health and social care sectors, according to the report, which suggested this area could also see more severe skills shortages as the technology is more widely adopted. The information and communications sector might also see net gains, it added.
Of the sectors predicted to see net employment losses as AI adoption advances, wholesale, retail, finance, public administration and transport were listed as the most impacted areas.
In geographic terms, the report said there would be job losses and gains from AI nationally, but positive impacts were more likely in London and the South East. Conversely, industrial areas in the Midlands and the North of England were likely to see a less positive effect.
The findings reinforced the idea that increasing AI adoption would widen existing education and skills-related income and geographical inequalities over time.
The need for a policy agenda that maximises the economic and social benefits of AI was mentioned in the report, as well as actions to help people adjust to these changes by the public and private sectors, such as digital education and upskilling focused on human skills that are complementary to AI and investment in areas of the UK that could see negative employment effects from AI.
Separately, BEIS also launched a report on the implications of robotics and autonomous systems (RAS) across UK sectors.
According to that report, RAS could be a solution to key challenges to continued economic growth in the UK, such as labour productivity. Reducing manual labour and increasing higher value output is seen as a priority area to be addressed if growth in living standards is to be sustained since the UK has an ageing population and a low birth rate.
On the other hand, the report noted that this was not straightforward for the UK, since the country had failed to capitalise on previous waves of robotic development and was now lagging behind other nations.
The total economic impact of RAS uptake across the sectors of agriculture construction, energy, food and drink, health and social care, infrastructure and logistics is estimated to be around £6.4bn by 2035, according to the report. In most UK sectors, however, it found there was a gap between adoption and the potential rate of automation that could be achieved.
Barriers identified in the report range from addressing drone regulation in agriculture and the cultural hurdles and lack of skills in relation to human-robot interaction in construction and logistics, to technical complexities in energy and infrastructure and the ethical concerns in the use of RAS in health and care.
While the study accepted there was room for policy to mitigate some of these barriers, it also noted that the success of these interventions would depend on various factors and the feasibility of policy changes should be further investigated in separate research.