Customization Gap is India's Edge in AI

India is now ranked third in the annual Global AI Vibrancy Ranking, a score sheet issued by the Stanford University. In most metrics, India ranks comparably with the top two nations, the USA and China. Where India lags is expectedly, that of R&D. This is significant and while the gap of India with the countries behind it is also large, this is going to be the critical metric which will need to be improved upon, vastly. 

On a happier note, Indian public opinion in favour of AI, based on the same score sheet is higher than most nations, compared to say the USA. Where India also lags is in the infrastructure enabling AI. This is a deficiency which stops India to build upon the potential of AI, despite the strong public support. 

To address some of these critical issues for Indian context, Susan Athey, Stanford Professor of Economics of Technology, at the Stanford Graduate School of Business, in an interview with the Indian Express mentions, why she is so optimistic about a significant strategic opportunity for India to build its own AI ecosystem.

 The important factor she harps upon are:

1. AI is only useful if it solves a problem to the finish line. Global models lack the local customization needed for the domestic markets in terms of language, local services, and data context. This is where the local players come in to act as "implementing partners" to complete the job needed. 
2. Global AI companies provide solutions for general purpose things. AI for government, defence, and local businesses requires specialized providers, which can only be provided locally as these global players lack the customization needed for India. 
3. A crucial thing she also mentioned is that "a big part of the customer demand is government-led, which gives an opportunity to promote domestic companies."
4. However, Athey does express concern that developing economies could become "net importers" of AI, leading to a loss of leverage and high markups. For that, local control is absolutely necessary, so as not to be exploited by foreign entities.
5. Need for a proper Regulatory body in this sector is thus an important factor as AI can well be involved in all areas of strategic importance for the nation. 
6. She also advocates Open-source models that will allow India to download, customize, and own the technology forever, ensuring competition and sovereignty. India needs to own and control the AI to stay safe and independent.
7. What is highlighted in the statement of Athey is that India’s strength lies in the application layer. Global AI is built for everyone, but it doesn't understand India’s specific languages, local data, or unique services. This creates a gap that only Indian companies can fill - especially in specialized sectors like defence, infrastructure, or local businesses - but only with proper Regulator. 

Link https://indianexpress.com/article/business/global-ai-models-dont-customise-enough-india-has-opportunity-susan-athey-10432041/