Nitin Pai: India can shape how artificial intelligence diffuses through society

Today anyone in the world with above-average computer skills can hope to deploy open-source AI models that are close to the state of the art. This means that the shift in value creation from AI development to diffusion will take place far earlier than anticipated.

The winners in the age of AI will not only be those who’ve built the smartest machines, but also those who make use of those machines most effectively. AI, like fire, electricity and computing, is a general purpose technology in that it permeates and transforms all sectors of the economy.

Also Read: Nilesh Jasani: Snap out of the DeepSeek delusion and invest big in basic research

It also has extreme compounding effects and early adopters— economies, firms and individuals—could establish and maintain leadership for a long time. So strategists and policymakers must quickly direct their attention to how India adopts AI as much as how India builds it.

At the macro level, there are wildly varying numbers of what AI means in terms of additional productivity. A 2023 Goldman Sachs report estimates that AI will add 0.7% every year to India’s total factor productivity (compared to 1.5% for the US). 

That’s not a lot compared to almost 3% that India enjoyed in the decade of digitalization and mobile internet diffusion. But even that low figure is challenged by Daren Acemoglu’s conservative analysis. He predicts that AI will add only 0.07% annually to US productivity, which would imply a minuscule effect on the Indian economy. 

On the extravagant side, industry associations and consulting firms have, as usual, plucked trillions of dollars and tens of millions of jobs out of their hats as the long-term benefits of AI adoption.

The truth is that we know AI promises to have a massive impact on the global economy, including ours. We don’t know how big it will be, what path it will take and how fast the benefits and disruptions will arrive. Last year, an International Monetary Fund staff discussion note observed that around a third of jobs in the US and UK are at risk of being displaced by AI. Just over 10% of Indian jobs were similarly at risk. But non-exposure to AI is a double-edged sword, because such sectors and economies will be relatively less productive compared to their AI-infused counterparts, and therefore less prosperous. In general, advanced economies seem to be in the high-risk, high-return category, while emerging economies face lower risks and lower returns.

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

One way for India to benefit from this dynamic would be to refocus its education system on sectors that have the greatest potential for being augmented with AI. 

This could mean professional services requiring good judgement, industrial work employing robotics and high-tech agriculture. Interestingly, some modelling studies show that AI can benefit below-average skilled workers by lifting their productivity, at least in the short-term. This is an opportunity to ameliorate past weaknesses in the country’s education system.

A lot depends on the trajectory of diffusion. The two big dimensions of uncertainty are the following. Will AI be incremental, like a faster train, or transformative, like an aeroplane? Will it augment human labour or will it replace workers through automation?

My own analysis suggests that a path of transformative augmentation is most consistent with India’s interests. It will create a broad productivity surge in the global economy, enriching many of our important trading partners, creating opportunities for Indian firms and citizens. At home, it can help narrow the governance gap by improving the government’s capacity in many crucial areas. 

The technology and services sector will be an engine of growth and our younger, more adaptable population can take advantage of global opportunities. India can leapfrog into a different league. However, this will come with tremendous pressure on our education system, welfare frameworks and urban areas.

Also Read: AI chips: India must get its basics right to meet this great catch-up challenge

The worst-case outcome is one of transformative automation, where humans are rapidly replaced around the world in multiple sectors. This will cause global disruption and likely lead to a concentration of power and wealth in the hands of a few countries and firms. Even in this scenario, India has an opportunity to leapfrog its economy, but at the risk of severe unemployment crises, social stability and premature de-industrialization.

There are incrementalist scenarios between these two extremes, which favour the rich world and leave India with decent growth but without the opportunities to ‘pole vault’ into the prosperous world.

Contrary to the self-serving narrative that the US industry has been putting out, the pathways of diffusion are not inevitable or over-determined by the technology itself. Human agency and public policy can make a difference.

We have a degree of choice on where to take AI. That choice gets better if India can engender international cooperation towards trajectories of mutual interest.

The author is co-founder and director of The Takshashila Institution, an independent centre for research and education in public policy.