AI and the future of work

Artificial intelligence dominated discussions at the World Economic Forum in Davos in January.  An IMF report, released prior to the start of the conference, catalysed debate with its prediction that AI will reshape the world of work, impacting the majority of jobs worldwide and ultimately leading to greater inequality.

AI will lead to job displacement

Almost 40% of global employment is exposed to AI, according to the report.  However, there is a high degree of variability between countries in the extent of exposure.  60% of jobs are expected to be impacted in advanced economies, with the figure much lower for emerging markets (at 40%) and low-income countries (at 26%).  The UK is an outlier.  According to the report, 70% of UK employment is highly exposed to AI.  This, it attributes, to the high proportion of workers in professional and managerial roles.

Exposure is high among skilled professional roles

One surprising feature of the report is how significant the impact will be on high-skilled roles.  While there is a certain logic in the use of AI in routine and automated tasks, such as the use of robots in warehouses, AI is also expected to transform occupations which involve a high degree of cognitive function.  Judges were cited as one occupation that is highly exposed to AI because of the degree of reasoning and textual analysis involved in the work, skills at which AI is particularly effective.

But certain roles are well-placed to harness the benefits of AI

Yet the picture for many roles is not as bleak as this may suggest.  The report identified that certain highly exposed roles also had high AI complementarity, meaning that AI had the potential to enhance productivity and efficiency in the role without subsuming it.  Surgeons were cited as an example because the high degree of responsibility in the role means there is limited scope for unsupervised use of AI.  These roles are well positioned to leverage the benefits of AI in performing tasks and taking decisions. 

By contrast, roles in which AI could autonomously complete tasks, such as telemarketing, were found to have much lower AI complementarity.  These roles were at much greater risk of being replaced by AI.

Inequalities will widen

The analysis of the impact on the global economy indicates that AI is expected to worsen overall inequality.  The differing impact between countries arises from different workforce compositions in terms of broad occupational groups and readiness to adopt new technology.  While advanced economies faced a greater risk of labour displacement, particularly in the short term, they were also much better placed to take advantage of the emerging AI growth opportunities.  The net effect for each country will depend on its ability to innovate, adopt and adapt to AI.

Within countries, it was expected that AI integration would disadvantage certain groups.  In particular, the report concluded that older workers may be less adaptable and struggle to embrace technologies, while younger, well-educated workers were best placed to reap the benefits of AI.

Trust in AI will make a difference

The overall impact of AI will be strongly influenced by the extent to which society trusts the technology, with higher levels of trust leading to a greater degree of social acceptability. The report identifies that robust regulatory frameworks are essential to govern and foster AI advances and will be critical in broadening societal confidence in AI tools.  By way of example, the report cites the fact that judges are shielded from job displacement because society is currently unlikely to delegate judicial rulings to unsupervised AI.

Preparation for AI is key

The main takeaway from the report is the importance of being AI-ready.  The report states that preparation for AI adoption is essential to harness its potential and mitigate its inherent risks.  For employers, this means embracing the new technology, rolling out training to upskill employees and being open to the need to adapt.