With extensive commercialisation of artificial intelligence (AI) and the advent of general-purpose generative AI such as ChatGPT, there has been intense debate on the effect of such automation on labour markets. According to a study by McKinsey, current generative AI and other technologies have the potential to automate 60-70 per cent of the employees’ time and as per a technology adoption scenario, half of the work activities could be automated by 2045.

The increasing ability of AI systems to automate tasks has spurred multiple studies that focus on AI’s labour market effects.

Professors Acemoglu and David Autor at the National Bureau of Economic Research (NBER) in the US have written a number of research papers delving into the details of such automation technologies on capital-labour replacement, labour demand, wages and productivity.

They have explained automation’s effect on labour demand as an interplay between the displacement effect and the productivity effect.

The displacement effect reduces labour demand by removing workers from automated tasks and squeezing them into a reducing range of manual tasks, thereby resulting in downward pressure on wages.

On the other hand, the productivity effect counteracts the displacement effect by increasing labour demand in the non-automated tasks, due to increased productivity in the automated tasks.

Some are of the view that with strong focus on automation enabled by technology, are we losing sight of possible new tasks where labour could be productively employed.

If so what could be the mobility paths for labour in tasks that are replaced by automation technologies? How can these mobility paths be enabled to minimise adverse impact on employment and wages?

Our research also shows that, as firms progressively automate the occupational tasks, the workers have to compete in an increasingly narrow range of manual tasks.

The overcrowding of workers in such occupations leads to downward pressure on wages. The examples of occupations in the IT services industry with easy to automate tasks are software testing and infrastructure management.

On the contrary, the workers associated with occupations having hard-to-automate tasks experience an upward growth of wages driven by increased demand. Examples of such tasks are IT consulting and software design. These sectors characterised by high labour productivity relative to capital.

Workforce mobility

These deteriorating wage inequality effects could be ameliorated if the displaced workers move across from the first category to the second through retraining. Such a mobility pathway structure can facilitate an efficient adjustment process.

According to a a recent IMF study, in an economy based on automation, investment in education produces the highest welfare gains such as wage growth and distribution of income in comparison to other policy responses such as taxation and infrastructure investments.

An important aspect of discovering mobility pathways for displaced workers is upskilling and re-skilling. Working professionals of today need to be constantly on the look out for continuing education opportunities to up-skill and re-skill.

First, the start-ups in the edtech market today enable such upskilling by offering short-term certificate programmes in emerging areas of technology and business. By forging strong alliances and partnerships with higher educational institutes (HEIs), the edtech platforms such as upGrad, Simplilearn and Great Learning have enabled such upskilling opportunities for working professionals to adopt suitable mobility pathways.

Second, the National Skill Development Corporation (NSDC) aiming to upgrade skill levels of the labour force in tune with international standards also plays a critical role in shaping the mobility pathways for individuals.

With the original objective being to upgrade the skill levels through vocational institutes, NSDC is also looking at avenues for furthering skill augmentation in emerging technology areas such as data science, AI and software development as India looks to become a $1 trillion digital economy by 2025.

Training for future

Third, is for the technology and business institutes in the country to train the workforce for the future. Though technologies such as AI and machine learning are evolving fast, they sometimes lack truthfulness and warrant human interventions. Since these technologies affect all humans and society at large, they need to be built with considerations such as minimising bias and discrimination in decision-making, proving counter-intuitive propositions, and assist but not take over human autonomy to name a few.

Initial steps towards this direction have been taken especially by the European Commission by enacting ‘Harmonized Rules on Artificial Intelligence (the ACI Act 2023)’. The Act strives to balance the socio-economic benefits of AI and the possible new risks and negative consequences for individuals or the society.

In light of the speed of technological change and the associated challenges, HEIs need to prepare engineers and business leaders of tomorrow with the required multi-disciplinary approach. The HEIs can lead the way in training such multi-skilled workforce that cuts across engineering, technology, business, social sciences, and management science.

Automation technologies using industrial Robots and AI, warrant businesses and governments to create occupation mobility pathways for displaced workforce through education and skilling so that it has the balancing effect on the labour market.

Sridhar is Professor, IIIT-Bangalore, and Upreti is Associate Professor, PES University. Views are personal

comment COMMENT NOW