The Impact of AI on the Future of the Job Market

James Golding

When one hears the question, “How will artificial intelligence affect the future of jobs and wage inequality?”, many people assume the negatives. However, artificial intelligence will enhance productivity, accuracy and speed of workers, and in many cases it is predicted that AI will compliment, rather than replace, labour. This article will discuss the positives and negatives of AI when considering the future labour market, and income inequality. 

In the case of high-skill workers, such as data scientists, software engineers, bankers, they often perform more cognitively demanding tasks. These can be enhanced by AI in the idea that artificial intelligence can support the workers by performing research, or skim-reading and highlighting the main points of a piece of writing. Many large companies in the areas of accountancy, banking and law, such as PwC, one of the 'Big Four' accountancy firms in the UK, highlight that a major skill they look for in applicants is an awareness of artificial intelligence, and to have skills which complement AI, such as digital skills, and adaptability. This demonstrates that AI can help assist high skills workers, without reshaping the job market. On the other hand, AI can automate tasks that would otherwise take humans hours to perform, such as digital analysis and law research. This could cause firms to increase productivity by increasing their capital (AI), and decrease their labour costs by laying off workers. This therefore is a reason why AI could cause high-skilled workers to be laid-off, and increase unemployment in this sector, via displacement effects. Overall, high-skill workers may benefit disproportionately, but this benefit won’t be across all areas of high-skill work, as some may be more replaceable by AI than others. 

For middle-skill workers, such as receptionists, administrative jobs, machine management, these jobs are particularly susceptible to AI automation. AI is particularly good at basic, repetitive tasks such as data entry and basic organizational skills. This change in the labour market is already being shown by machinery in itself, with robots performing manual organisational tasks previously performed by humans, as seen in Amazon and Ocado warehouses. AI will only extrapolate this decrease in middle-skill workers as it displaces the labour factor of production. Despite the opening of jobs for managers to lead AI-including projects in these areas of work, middle-skilled workers face the biggest displacement threat, especially if their tasks are more repetitive and simple. 

For low-skill workers, such as cleaners, factory workers etc., these jobs tend to be manual, repetitive and often involve dexterity skills. As many of these jobs require physical and interpersonal interactions, AI will not be able to perform these role for now. Furthermore, AI is unlikely to enhance productivity in these jobs either, as these jobs tend to be simple and it is not economically effective to pay for AI capital to support these jobs. Overall, AI will have little impact on the market for low-skilled workers, although the economic growth caused by AI may lead to an increase in demand for low-skilled workers, as labour is in derived demand. 

One policy to tackle the widening inequality gap for worker’s wage, is the implementation of a change of the school curriculum to include soft skills, and training in the use of AI. Many modern schools ban the use of AI, or tell students not to use it. This policy would modernise the school system and change the skills of future workers to be more adaptable and trained with AI. This program would both tackle the problem of displaced workers who can’t find new jobs due to structural unemployment, but also tackle the main problem of low-skill workers not being able to use AI as an assisting tool. A limitation of this is that the students have to be motivated to learn, and there is an opportunity cost for this scheme as the government could spend the money on defence, for example, or paying back debt to narrow the budget deficit. Also, there may be a time lag between the implementation of this policy, and the benefits it brings. This policy would help to successfully fix the problem of income inequality due to labour market polarisation, but it is not sufficient. It would need to be combined with policies such as subsidised reskilling programs, or/and increased unemployment benefits. 

In conclusion, the implementation of AI in future firms will reshape the job market in multiple ways: it will lead to an increase in the demand for high-skilled labour, particularly those workers with adaptable skills, it will lead to the wage stagnation of middle-skill labour, as the market shrinks due to the implementation of AI capital to replace labour. Some research suggests that AI could reduce inequality under certain conditions. However, in practice, the main expectation of economists is that the wage inequality gap will increase.