The Debate: Why Hasn’t India Matched China’s AI Success?
In a recent social media discussion, Sridhar Vembu, Chief Scientist of Zoho Corporation, addressed a pressing question from Indian engineers: “Why can’t India achieve what China did?” While acknowledging India’s strengths in democracy and diversity, Vembu emphasized the need for grassroots innovation, asserting that “India’s strength lies in its ability to innovate and adapt to rapidly changing global dynamics.”
This conversation gained urgency following China’s latest AI breakthrough: DeepSeek. This Chinese AI model has shocked the global tech industry by outperforming OpenAI’s models, topping the iOS App Store, and challenging Meta in the open-source AI space—all at a fraction of the cost. DeepSeek’s success has sparked renewed concerns among Indian tech leaders: Why isn’t India leading in AI?
India’s AI Ambitions: Can It Compete?
India has recognized the urgency of building its own generative AI model. Union Minister Ashwini Vaishnaw recently announced plans for a large language model (LLM) powered by the India AI Compute Facility. This initiative has already secured 19,000 GPUs—including 15,000 high-end GPUs—far exceeding the government’s initial target of 10,000. The model, expected to roll out within 8-10 months, aims to enhance India’s AI capabilities at a highly competitive cost.
Highlighting the affordability of India’s AI mission, Vaishnaw stated that GPU computing costs will be under a dollar per hour—the lowest in the world—thanks to a 40% government incentive.
Why Is DeepSeek So Significant?
DeepSeek R1, China’s answer to OpenAI’s ChatGPT, reportedly cost just $6 million to develop, compared to OpenAI’s $540 million. Moreover, DeepSeek operates on fewer GPUs, drastically reducing operational costs. This frugal innovation—reminiscent of India’s Chandrayaan-3 mission—has sent shockwaves through global tech markets. Following DeepSeek R1’s release, American tech giants like Nvidia, Meta, and Alphabet saw billions wiped from their valuations.
The Key Challenges Holding India Back
1. Lack of a Long-Term Vision and Investment
Zerodha CEO Nithin Kamath pointed out that India’s technological struggles stem from systemic short-termism. He explained, “You can’t just buy GPUs and expect Indians to create groundbreaking AI applications. Without the right talent and an enabling ecosystem, all the GPUs in the world will be pointless.”
Kamath emphasized that India doesn’t provide an adequate research environment, which is why top Indian researchers leave for the US. “China spent over two decades systematically building its AI and scientific capabilities,” he noted, highlighting the need for sustained investment and research infrastructure.
2. The Hard Part of AI Innovation: Training Large-Scale Models
An anonymous AI researcher at Amazon, known as GDP, pointed out that the fundamental science behind LLMs isn’t the hardest part. “The real challenge lies in running large-scale AI training jobs across thousands of GPUs,” they explained. DeepSeek’s success stems from innovative parallel and distributed computing techniques, which drastically improve efficiency and reduce resource consumption.
3. Lack of a Protected Market
Unlike China, India lacks a shielded market where domestic tech firms can experiment and grow. GDP noted, “In China, AI startups are protected from foreign competition, allowing them to fail, iterate, and eventually succeed.” By contrast, Indian AI firms struggle to survive against American tech giants, which are often cheaper and better.
China also treats AI development as a national security priority, ensuring steady government funding and corporate backing—even when AI startups underperform. This long-term commitment to AI talent has led to a fiercely competitive ecosystem where only the best models survive.
The Solution: India Needs Its Own ‘DARPA’
To bridge the AI gap, experts are calling for India to establish a government-backed AI research fund. GDP proposed a DARPA-like national AI initiative, backed by $3 billion over three years, to support AI teams without immediate revenue pressure.
“We need a national fund that will support AI teams, ensuring they meet increasingly difficult benchmarks every six months. No revenue needed to survive for the first three years,” GDP argued.
Vembu echoed this sentiment, urging Indian engineers to embrace the challenge rather than blame systemic obstacles. “Chinese entrepreneurs never had it easy and still don’t. We can do this in India. We WILL do this in India,” he affirmed.
Can India Catch Up?
DeepSeek has proven that AI breakthroughs can be achieved with limited funding and computing power—provided the right ecosystem exists. For India, the key takeaways are:
- Long-term investment in research and AI talent.
- A protected domestic market for AI startups to grow.
- A supportive government-backed AI initiative (India’s own DARPA).
Without these structural changes, India risks falling further behind in the AI race. As Kamath warned, “With AI set to reshape the world, being mediocre won’t be an option.”
Leave a Reply