Prime Minister Narendra Modi is attempting to pave the way for India in a fierce competition to create frontier models as the country begins one of the biggest artificial intelligence summits in the world on Monday.

The India AI Impact Summit, which could be the biggest assembly of AI luminaries to date, is anticipated to bring together world leaders, tech titans, AI startups, and investors in New Delhi. Researchers Yann LeCun and Arthur Mensch are among the guests, along with Sundar Pichai of Alphabet Inc., Sam Altman of OpenAI Inc., Dario Amodei of Anthropic PBC, and Alexandr Wang of Meta Platforms Inc.

French President Emmanuel Macron will give the keynote address on the final two days of the summit, February 19 and 20, and PM Modi will then make his remarks.PM Modi sees the summit as an opportunity to highlight India's engineering prowess and large tech-savvy populace as factors that could help the country win the next stage of the global AI race. Data from more than a billion people, identified by the biometric ID system Aadhaar, powers the nation's digital infrastructure. Despite its late beginnings, it has a history of rapidly scaling technology; it missed the boom in personal computers but went on to become a dominant force in software services, going from a small number of landlines to about a billion smartphones in less than 20 years.

According to Abhishek Singh, additional secretary at the Ministry of Electronics and IT, "India is trying to compress decades of development into years by overlaying AI over existing digital identity, payment rails as well as health care, education, and governance stacks." "And what gets built for India won't stay only in India."
The nation's digital identity and payment system is already being exported. Aadhaar's architecture served as the inspiration for MOSIP, an open-source platform that is currently assisting nations including the Philippines, Morocco, and Uganda in creating national identification systems. Some countries are creating digital payment platforms atop the same scaffolding.
According to Stanford University's Institute for Human-Centered AI, India is the third most competitive country in the world in terms of AI, after the United States and China.

Global tech firms are taking notice. In India, OpenAI and Anthropic are establishing their businesses and pursuing government organizations, developers, and business clients. In order to cater to one of the fastest-growing markets for models like ChatGPT, Gemini, and Claude, Google and Meta are extending their data centers. Nvidia Corp. views India as a counterbalance to US export restrictions on high-end processors made in China, but its CEO abruptly withdrew from the summit, citing "unforeseen circumstances."

However, industry observers warn that India's AI growth may be hampered by years of underinvestment in technology research and development. The nation's true success, according to Aakrit Vaish, head of the AI-focused fund Activate, will come from bolstering its research environment so "we aren't just a testing lab for Silicon Valley's algorithms."

Models that are locally attuned are already being developed. This week, voice-first systems for dozens of Indian languages will be introduced by researchers, reflecting the country's linguistic variety.

The government-backed BharatGen, which was created by integrating the research capabilities of India's leading engineering schools, will introduce Param2, a 17-billion parameter model that supports 22 Indian languages, during the summit. With support from Lightspeed Venture Partners and Khosla Ventures, Sarvam AI will introduce a larger model that prioritizes speech. In order to assist revolutionize industries like classrooms, clinics, and agricultural fields, both projects seek to collect more data and make low-cost AI available to a large public.

A dilemma for the Chinese ecosystem, growing competition from such local models may further postpone the profitability of AI businesses in India for US companies.

The intentional emphasis on cost has the potential to revolutionize the market. "In India and much of the developing world, cost is not an afterthought," stated Rishi Bal, CEO of BharatGen. "Our model is designed to accelerate adoption in critical areas across governance, education, health care, and farming."

According to Himanshu Tyagi, co-founder of Sentient AI, a San Francisco-based company sponsored by Peter Thiel, India could catch up if it concentrates on fields like robotics and sophisticated reasoning for science, as "the next wave of intelligence will use data not on the internet."