The Indian government is pushing full speed ahead to make the country an attractive technology base, especially in the hot fields of AI and semiconductors, but ambition and caution have the country at a crossroads.
The government last week cleared the path to establish a comprehensive AI policy under the banner of the “IndiaAI mission.” The $1.2 billion initiative provides the hooks to create a comprehensive national AI infrastructure, which includes creating AI models and providing computing infrastructure.
The IndiaAI mission will be a public-private partnership to catalyze AI innovation, said India’s Ministry of Electronics & IT in a press statement. India will also establish an AI computing base with over 10,000 graphics processing units (GPUs).
“Further, an AI marketplace will be designed to offer AI as a service and pre-trained models to AI innovators. It will act as a one-stop solution for resources critical for AI innovation,” the ministry said.
CEOs of major companies have visited India to get a piece of the AI opportunity in India. The country could become the third-largest economy in the world by 2030, surpassing Japan and Germany, according to S&P Global.
“India’s nominal GDP measured in USD terms is forecast to rise from $3.5 trillion in 2022 to $7.3 trillion by 2030,” said S&P’s Rajiv Biswas, Asia-Pacific chief economist at S&P Global Market Intelligence, in a research advisory published in December.
India’s business environment is also becoming friendlier, which was not the case a decade ago.
A business environment readiness survey by the Economist Intelligence Unit in 2023 saw a market opportunity, technology readiness, talent pool, and foreign trade and export controls as India’s strengths. The political environment is rated as India’s weakest link.
There’s a lot of money to be made in India. AI is still a greenfield opportunity in the country, with very limited use so far. The country has a small GPU install base compared to the U.S., which is expected to grow over the next two years.
Almost all major tech CEOs — Elon Musk, Andy Jassy, Sundar Pichai, Pat Gelsinger, and others — have met with Prime Minister Modi in the last two years.
Nvidia CEO Jensen Huang visited India last year and announced a partnership to create an AI infrastructure with Reliance Industries, a diversified company with technology and telecom divisions. Reliance runs Jio, one of India’s largest mobile phone service providers.
Reliance is establishing a data center infrastructure that will deploy Nvidia’s DGX Cloud boxes CPUs and GPUs. Reliance is creating a foundational AI model to host AI applications for 450 million mobile customers.
Nvidia said in a statement, “The AI infrastructure will be hosted in AI-ready computing data centers that will eventually expand to 2,000 MW.”
Reliance’s CEO is billionaire Mukesh Ambani, whose family members are majority shareholders. Ambani shelled out a cool $120 million on his son’s wedding earlier this month.
An Indian data center operator, Yotta, is working with Nvidia to expand AI use. The company expected 4,096 Nvidia H100 GPUs to be operational by January this year and said that 16,384 GPUs will be operational by June 2024. That number will scale to 32,768 by the end of 2025.
India’s technological transformation took a sharp turn during the pandemic when the large cash economy pivoted quickly to mobile payments. Street vendors were quick to adopt services like PayTM for touch-free transactions, and mobile devices are now a mainstay for customer-facing services provided by private organizations and the government. Reliance wants to use Nvidia to put AI at the top of the infrastructure.
Microsoft’s CEO Satya Nadella visited India last month and talked mostly about AI. The company announced a program to upskill 2 million people with AI skills by 2025.
Microsoft and Nvidia are also working with Indian organizations that can adapt to India’s major languages. India has 22 official languages, according to India’s Department of Official Language.
Most AI opportunities go indirectly through the Indian government, which is more of a watchdog of technology implementations. However, potential issues have popped up that could serve as a cautionary tale for Silicon Valley.
A recent advisory from India’s tech ministry stated that tech companies required explicit permission to use AI models, LLMs, or algorithms deemed undertested and unreliable. That was on top of concerns within the Indian government on deep-fake videos, with the Indian election happening later this year.
Silicon Valley companies did not react kindly to the Indian government’s heavy-handedness, which resulted in a backlash.
Indian minister Rajeev Chandrasekhar clarified the government’s stance with further guidance, saying the “advisory is aimed at untested AI platforms from deploying on Indian Internet” and “advisory is aimed at the significant platforms and permission seeking from Meity is only for large platforms and will not apply to startups.”
Chandrasekhar’s comments focused on the unreliability of LLM responses, which has been a topic of controversy even in the U.S. India’s advisory provides guidelines on the penalties for misinformation or illegal use of platforms that violate India’s laws.
Chandrasekhar last month called out Google when its Gemini model labeled PM Modi a “fascist” in response to a prompt query.
“These are direct violations of … of the IT act and violations of several provisions of the Criminal code,” Chandrasekhar said on X.
He went on to say “safety & trust is platforms’ legal obligation” and that”‘ sorry unreliable’ does not exempt from the law.”
Chandrasekhar, minister of state for electronics and IT, has a distinguished tech background. He is a chip designer and was formerly an Intel executive. He went on to lead BPL Mobile, one of India’s first mobile providers.
But the sour episode has raised questions about whether the advisory is an advisory in name or whether it is legally binding. Executives in Silicon Valley are aware that India is not the easiest place to do business, but the Indian government is trying to change that perception, but not at the cost of safety and ethics.
Chandrasekhar isn’t against Google. Late last year, he met with a Google executive to see how the company could work with the government and its partners. However, the limits posed by experimental uses of AI could stop the latest technology from reaching India’s shores.
The $1.2 billion AI overlay is still much smaller compared to India’s ambition to become a semiconductor giant. India is offering $10 billion in incentives to attract chip makers to the country and recently approved plans to set up three chip factories with $15 billion in public-private investments.
