A PwC study in 2017 estimated the world would gain $15.7 trillion by 2030 if artificial intelligence (AI) was adopted across nations. The study said that AI would first lead to productivity enhancement, and a major portion of gains would accrue from consumer-side effects. China, it had said, could see its GDP rising by around a fourth as it was using AI more aggressively. Although the study did not estimate how much India would gain from using AI, new research by Icrier along with Nasscom and Google shows that even a marginal increase in artificial intelligence adoption may add 2.5% to GDP in the immediate term. Moreover, it highlights that if the government spends the Rs 7,000 crore it had envisaged for the national AI programme, GDP could get boosted by as much as $86 billion. The way Icrier sees it, as AI becomes what it calls a general purpose technology—like the internet—its impact rises; essentially, then, the pace of India’s digitisation drive will determine how fast AI is adopted.
To understand how fast the adoption of AI can take place and its impact on total factor productivity, Icrier studied 1,553 firms that have some software investment. What it found was that there was a huge gap in the use of AI, suggesting a large untapped potential. AI-intensity was defined as the ratio of software investment to total sales, and the study found that, for instance, in the case of agriculture, while the average AI intensity is 0.001, the maximum intensity was 20 times as much. For electrical and optical equipment manufacturing, the difference between the average and the top in the industry was 145-times; it was 742 in the case of trade and in the case of services, the average intensity was 0.159, while the maximum intensity was 110.
The report, however, argues that businesses alone won’t be able to push AI, the government will have to play a bigger role, by setting up a nodal AI agency to push for AI-adoption and also drive business, government and academia partnerships. Another suggestion is to initiate large-scale skill development programmes to get the workforce ready for AI-adoption. What is worrying, however, is the slow pace of digital adoption so far, though the pandemic has helped speed up things a big; both the education and health sector, for instance, are likely to see faster adoption of AI techniques. A related problem is that of cybersecurity where India needs both a national strategy and a governance structure that is more well-defined.