DeepSeek's Rapid Advancement Reignites Debate on Indian Foundation Model Development

Sam Altman, CEO of OpenAI, met with Indian IT Minister Ashwini Vaishnaw on February 5, 2025, to discuss the country's plans for artificial intelligence development. Altman was previously sceptical about creating Indian AI foundation models, particularly regarding cost-effectiveness, but in his recent statements, he highlighted India's leading role in the global AI revolution.

The OpenAI CEO acknowledged that they have learned much about developing smaller models, and while their training still requires significant resources, their implementation is becoming increasingly accessible. According to him, this development could bring tremendous creative momentum, in which India must also play a key role. The change is partly driven by the success of Chinese DeepSeek, which, with an investment of just six million dollars, developed an AI model that outperformed OpenAI's solutions in certain areas. The DeepSeek-R1 model is open-source and delivers outstanding performance despite its extremely low development costs, which has given new momentum to debates surrounding Indian AI developments.

India, however, still faces serious challenges in developing foundation models. According to Nithin Kamath, CEO of Zerodha, one of the country's biggest hindrances is short-term thinking and the "jugaad mentality" characteristic of problem-solving. This approach prioritises quick and temporary solutions over long-term, well-founded developments, impacting business life, regulatory strategies, and technological innovations. While the Indian tech sector focuses primarily on adapting and enhancing existing AI models, an increasing number of experts warn that developing foundation models is inevitable. According to Aravind Srinivas, CEO of Perplexity AI, India cannot afford to build upon others' technology and must prove that it can achieve outstanding results in AI, similar to its successes in space research.

Altman's recent visit and OpenAI's growing presence in India suggest that the country may become a more significant player in global AI developments. Despite the challenges ahead, more and more people in the technology sector believe that in the long term, India can not only catch up but potentially play a leading role in the AI revolution.

Sources:

India poised to lead in AI space, says OpenAI CEO Sam Altman

India emerging as OpenAI’s second-largest market, CEO Sam Altman urges country to focus on AI development.

BusinessLineS Ronendra Singh

Global AI race: India to build own foundational models; 18,693 GPUs to power compute facility - ET CIO

Put simply, foundation models in generative AI are large, pre-trained models that form the base for a variety of AI applications.

PTIPTI

Why is India behind China in AI race? Zerodha’s Nithin Kamath blames ‘Jugaad’ mentality for it amid DeepSeek buzz - The Economic Times

Zerodha CEO Nithin Kamath has criticized Indias approach to AI innovation, highlighting the country’s reliance on short-term solutions and the “Jugaad” mentality. He pointed to Chinas recent AI breakthrough, DeepSeek, as an example of how long-term research investment has propelled their tech progress. Kamath argued that India lacks the right ecosystem for talent and research, causing many top researchers to move abroad. Why India is falling behind China in AI race? Zerodha’s Nithin Kamath blames ‘Jugaad’ mentality for it

Economic TimesET Online

Deepseek’s rapid progress renews debate on Indian foundational AI model

Deepseek AI model leads global race, with Indian firms leveraging open source advancements to drive AI excellence.

The HinduAroon Deep