China has dropped a bomb on the global markets with ‘DeepSeek AI’, an artificial intelligence (AI) assistant that has leapfrogged other AI models including ChatGPT to become the biggest threat to US hegemony in AI sector. Now taking a cue, India, the next door neighbors of China, has claimed that it is building its Large Language Model (LLM) to develop an indigenous AI assistant, a la, Chandrayaan (Moon Mission) style.
The entry of DeepSeek AI resulted in collapse of markets including crypto as the makers have claimed to develop the chatbot at a marginal cost compared to U.S. based OpenAI, the makers of ChatGPT. Now, as per a report in The Indian Express, India’s Union IT Minister Ashwini Vaishnaw has claimed that the world’s most populous nation is building its own LLM model under an umbrella project ‘IndiaAI Mission’ with a whooping Rs 10,370 crore ($1196 Million) budget.
According to Vaishnaw, India has already selected 10 companies to supply over 18,000 graphics processing units (GPUs) for the project.
“Over the last one and a half years, our teams have been working closely with startups, researchers, professors etc. Today, we are calling for proposals to develop our own foundational model. The model will take care of the Indian context, languages, culture and devoid of biases,” said Vaishnaw to reporters.
Ahead of PM Modi’s scheduled visit to the AI Summit in France, many are contemplating whether India would initiate her own AI project model that is much more cost efficient than both OpenAI and DeepSeek. Now without wasting more time, let’s delve deep into this topic and find out whether India is capable of producing an AI model like DeepSeek and what’s the government plans are regarding this.
According to a report in The Indian Express, several scientists and academics in India believe that the country can build an AI assistant at the same cost or even less than $5 million that the DeepSeek team allegedly invested to create their AI chatbot.
This raised questions about the Indian capabilities in the AI sector and also sparked discussions and debates across all social media platforms. On the other hand, several Indian technology experts believe India can build a better AI model with a similar financial framework.
India Already Took The First Step
India has already taken the initial steps to make their foundation strong in the AI sector. According to the latest reports, Union IT minister Ashwini Vaishnaw has announced ambitious plans for Indian AI infrastructure and their recent achievements.
According to Vaishnaw, the Indian Government is accelerating its computing power by enrolling double the initially planned Graphics Processing Units (GPUs) for AI development. Initially, they planned to enroll 10,000 GPUs and now they empaneled 18,693 GPUs including 12,896 Nvidia H100s and 1,480 Nvidia H200s.
He said “Overall, the net high-end GPUs we have is 15,000 and noting that even a model like DeepSeek was trained on just 2,000 GPUs.”
The government is not only focused on creating a robust AI infrastructure instead they want to create AI models that will reflect India’s unique cultural and linguistic diversity. Vaishnaw stated that they launched a framework to develop an AI model that includes the Indian context, the Indian languages, and the culture of the country and avoids biases.
Strategic Focus Areas of India AI Mission
According to media reports, initially, the government will focus on some key sectors where AI can have a significant impact, these sectors include Healthcare, Education, Agriculture, weather forecasting, and some more. These areas are critical for addressing pressing national issues, and the government has already initiated funding for 18 projects aimed at leveraging AI in these sectors.
Expected Timeline and Budgeting
Minister Ashwini Vaishnaw gave an estimated time for developing these AI models. According to him, India will develop AI models in the next six to eight months which is the maximum limit and if things go well the rosy outlook will be four to six months.
Moreover, to make the IndiaAI Mission more successful, the government will provide a subsidy on GPU access. the government will provide a 40% subsidy to researchers and academia which will cut down the cost to Rs 100 per GPU hour which is lower than the global rate of approximately Rs 200-240 per GPU hour.
Final Thoughts
With the emergence of DeepSeek AI countries across the globe now joined the race of AI development so India. The government of India has a strategic plan to develop AI models that not only compete with models such as DeepSeeK AI, OpenAI, and more but also reflect the culture, languages, and context of the country.
Moreover, India needs an AI model that understands the cultural phenomenon of the country and also the diversity that India poses. Picture this, a local lady from Tamil Nadu starts her day by speaking to her AI assistant in Tamil. “Vanakkam,” she says to her basic smartphone, “En nilathil enna nadalam?” (Hello, what’s happening in my field?). The AI assistant also helps her in communicating in other languages of India.
India’s AI should reflect the unity and diversity of this nation and only time will tell, how quick will India roll out its indigenous AI model.
Also Read: DeepSeek Skips Questions on Arunachal Pradesh and Kashmir being part of India
The Crypto Times – Read More