ADVERTISEMENTs
-10”-x-5.jpg)
AI needs decentralization, says tech entrepreneur Radha Basu
“India’s strength will be in training models that are applicable for India and using that as a base to become a global AI power,” she advised.
Radha Basu, founder and CEO of iMerit. / Courtesy Photo
India’s artificial intelligence (AI) landscape must move towards decentralization, focusing on small, purpose-built models rather than large, centralized systems, according to Radha Basu, founder and CEO of iMerit, a global AI data solutions company.
Speaking to New India Abroad on the sidelines of the Indiaspora Summit Forum for Good (IFG) in Abu Dhabi, Basu emphasized that India has the potential to lead AI innovation by tailoring models to its unique needs.
“Decentralizing to small models—that’s what India can do,” Basu said. “India’s strength will be in training models that are applicable for India and using that as a base to become a global AI power.” She underscored that AI development should not be confined to major metro cities but should instead extend to tier-2 and tier-3 locations.
Basu also spoke about India's approach to AI as a sovereign force, designed to strengthen the country before expanding globally. “I think that India's approach is of looking at AI first, in terms of really contributing to India, within India,” she said.
She stressed the importance of an inclusive AI workforce, calling for participation from semi-urban and rural areas as well as a strong focus on gender diversity. “India is very committed to 50-50. Gender-wise and to involve young people from rural areas. All of those things, not just rural, semi-urban, tier-two towns—all of those things make it powerful.”
With India’s push for responsible AI and language-based large models, Basu believes the country is poised for a significant AI-driven transformation. “Responsible AI, purpose-built applications, and sovereign AI—these are the three key pillars that will shape India’s AI future,” she said.
Government and industry collaboration
On the role of the Indian government in AI growth, Basu pointed out that public-private collaboration is crucial. “I think the government plays an important role, and it is for us companies… working with the government in tandem, it means one plus one equals five,” she said.
Citing examples from Tamil Nadu and Meghalaya, she noted how state governments are already taking proactive steps, such as setting up an Automotive AI Center of Excellence in Coimbatore.
“Different states are also involved—West Bengal, Meghalaya,” Basu added. She recalled an interaction with Meghalaya Chief Minister Conrad Sangma, who inaugurated an iMerit AI center and later joined employees for an impromptu jam session. “We have to work with the government, and the government can play a role,” she said.
The rise of generative AI
Basu noted that while AI has been around for eight years, the recent surge in excitement is due to advancements in generative AI. “We’ve been in AI for eight years… The last year, the buzz is because of generative AI,” she explained. She pointed to applications in precision agriculture, medical AI, and autonomous mobility as areas where India is already making strides.
She also highlighted recent breakthroughs in language models, such as China’s DeepSeek, which has demonstrated the ability to develop AI infrastructure at a lower cost with more accessible APIs. “That means energy for the infrastructure is lower, and their APIs or their connects are so simple and cheap that all kinds of coders, all kinds of people can develop it,” she said. India is following a similar path, with homegrown language models like Krutrim and Hanuman emerging.
“Let’s just develop the LLMs [large language models] in Indian languages,” she said, explaining how iMerit works on both Silicon Valley-based models and India’s indigenous AI efforts.
ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
E Paper
ADVERTISEMENT
.jpg)
Video
.png)
Comments
Start the conversation
Become a member of New India Abroad to start commenting.
Sign Up Now
Already have an account? Login