Arun Maira: The impact of AI on Indian jobs is a distraction we can do without

On one hand, there is optimism that AI will improve the world in ways that human intelligence has not been able to. On the other is the fear that we don’t know how this artificial (or non-human) intelligence will change the world.

Also Read: Jaspreet Bindra: The ethics of AI will matter more than the technology

There are two responses to uncertainty over how it will turn out. 

One is to predict future outcomes for a world with more AI and design policies accordingly. 

The other is to regulate AI beforehand, even before we agree on its impact, so that we can prevent its bad side from destroying us and get more of the good. 

But then, “The only thing you cannot predict is the future," someone once said. I asked ChatGPT who did. It said many people have said it through history and one cannot attribute this eternal wisdom to anyone. 

But why is the future unpredictable? There are two reasons.

One is that multiple forces interacting with each other in non-linear ways change the state of complex systems and produce outcomes that will not be visible until the system has changed. Social, economic, environmental and political forces will be impacted by any transformative technology, and they will, in turn, react to the technology and change its course.

Ten years ago, the hype was ‘Industry 4.0,’ a theme pushed by the World Economic Forum and other lobbies. Consultants made predictions of how jobs will be affected and what new skills will be required. There were wide differences in their predictions of skills in demand and how employment will be affected. 

A 2016 Confederation of Indian Industry report that I guided, titled ‘Future of Jobs in India: Enterprises and Livelihoods,’ noted the wide variation of their estimates. For example, forecasts of jobs in India’s automotive sector in 2025 varied from 3.4 million to 19 million. 

There were methodological reasons for the variations. First, it was non-systemic. The models excluded social and political forces that are not easy to quantify. Second, some estimators were biased. Those vested in new technologies tend to make positive projections. 

The other reason why transformations led by new technologies cannot be predicted is that their trajectory depends on how they will be regulated, and by whom. This is a political process.

Early developers of transformative technologies always want control over their proliferation to retain their own power. With AI, the world is at a ‘nuclear’ moment again, 80 years after the power of nuclear energy was unleashed in Hiroshima. Regulation of nuclear energy has been controlled since then by its first large producer.

AI evangelists promote its benefits in healthcare, education, commerce, etc. Others are concerned about AI’s dark side: loss of privacy and the oligarchy of private companies controlling tech platforms and AI applications, who seem to be controlling our lives and even our governments. AI is giving the most powerful even more power. 

Thoughtful people, not yet brain-washed by AI evangelists and tech corporations, do not want to leave AI regulation to the conscience of rich companies and individuals.

Also Read: Sam Altman’s strategic ambiguity on AI sounds a lot like doublespeak

Let’s take forecasts of AI’s benefits by its promoters with a pinch of salt and look to the International Monetary Fund (IMF), which takes a more systemic view and recognizes that the impacts of AI will differ across countries, depending on their economic, social and political structures. 

The IMF says that AI will impact only 26% of employment in India. And that it will benefit 14% of job categories and have a negative impact on 12%. So the overall impact of AI on jobs and incomes in India will be insignificant in the bigger scheme of things. 

From a big-picture perspective of the Indian economy, AI is a distraction. Instead, we must look at the structural impediments that prevent the growth of decent jobs and incomes.

A fundamental problem is the way that policies for structural reforms are developed. The IMF has examined why many governments have diluted IMF-backed structural reforms over the past decade (and even backtracked). 

The IMF’s World Economic Outlook report of October 2024 concludes that stakeholders are not consulted sufficiently before policies are designed because experts think that common people do not understand their own problems as well as they do. 

Consultations, if any, with common people are pro forma and insincere. Unsurprisingly, says the IMF, reforms are resisted when they are imposed on people.

In conclusion, AI is marginal to India’s challenge of jobs and incomes. Better solutions will be found by improving the process of reforming labour, land and agriculture policies. Solutions cannot be found by data analysis, nor by think-tanks and experts alone. 

Also Read: Indian agriculture needs support. Reform farm prices at the very least.

To understand how the economy should be reformed and technology applied judiciously, our policymakers must listen more intently to common people. This could guide how we go about increasing employment and incomes faster in the lower half of the pyramid, which would give businesses an impetus to invest in growth.

The intended beneficiaries of structural reforms to increase jobs and incomes are farmers, workers and small enterprises, principally those in the informal sector, not providers of finance and technology that fund research and publish reports. As the IMF report suggests, their inputs are vital to making policy progress on their behalf.

The author is former member of Planning Commission and author of ‘Transforming Systems: Why The World Needs A New Ethical Toolkit’.