Perplexity AI CEO Aravind Srinivas publicly disagreed with Infosys co-founder Nandan Nilekani’s views on AI, advocating for India to focus on AI model development and real-world AI applications.
Mr Srinivas shared his views on X, praising Nandan Nilekani as “awesome” for his unparalleled contributions to India’s technological progress through initiatives like Infosys and UPI. However, he raised concerns over Nilekani’s advice to Indian AI startups, urging them to focus exclusively on developing practical AI applications rather than training large language models (LLMs).
Mr Srinivas wrote on X, “To be clear: Nandan Nilekhani is awesome, and he’s done far more for India than any of us can imagine through Infosys, UPI, etc. But he’s wrong on pushing Indians to ignore model training skills and just focus on building on top of existing models. Essential to do both.”
See the post here:
I am ready to invest a $1mm personally and 5 hours/week of my time into the most qualified group of people that can do this right now for making India great again in the context of AI. Consider this as a commitment that cannot be backtracked. The team has to be cracked and… https://t.co/g6ItsPL0uc
— Aravind Srinivas (@AravSrinivas) January 22, 2025
Mr Srinivas’s remarks were in response to Nilekani’s statements at the Meta AI Summit in October, where Nilekani advised Indian startups to steer clear of the expensive pursuit of building large AI models and instead prioritize developing practical, resource-efficient AI solutions.
“Our goal should not be to build one more LLM. Let the big boys in the (Silicon) Valley do it, spending billions of dollars. We will use it to create synthetic data, build small language models quickly, and train them using appropriate data,” Mr Nilekani had said.
He underscored the importance of scalable, cost-effective infrastructure and practical applications tailored specifically to India’s unique needs.
Mr Srinivas, however, offered a differing perspective, advocating for a more ambitious approach to AI development in the country. Drawing from his own experience, he argued that India risks missing out by avoiding the training of AI models due to the misconception that the financial costs are insurmountable.
“I feel like India is falling into the same trap I did while running Perplexity-assuming that training models will cost an exorbitant amount of money,” Srinivas said. He urged Indian startups to move beyond relying solely on open-source models and instead focus on building proprietary AI capabilities to achieve global competitiveness, particularly in areas like Indic languages.
“Elon Musk admired ISRO-not even Blue Origin-because he respects people who can achieve great things without excessive spending. That’s his operating philosophy,” Srinivas noted, encouraging India to adopt a similarly resourceful and bold mindset in AI development.
He concluded with a call for a paradigm shift in India’s AI strategy, emphasizing the need to develop expertise in training foundational AI models that can both address local needs and compete on a global scale. Offering his support, Srinivas expressed his willingness to assist anyone “obsessed enough” to take on this challenge, pledging to help open-source AI models for broader impact.