The release of DeepSeek R1, a Chinese AI model, has sent shockwaves through the world. A significant milestone in the AI race, the release of R1 has proven that Chinese companies can now build foundation models at a level comparable to OpenAI’s GPT-4, Anthropic’s Claude, or Google DeepMind’s Gemini.
While Silicon Valley is still reverberating from shock, closer home, a different kind of question is on everyone’s mind: Why has India, with its plethora of software engineers, not been able to build AI the way China and the US have?
An Indian professional in the AI space has answered this complex question.
Why India lags behind
Indian-origin American author Sadanand Dhume took to social media to ask the pressing question of why India is lagging behind in the AI race. “For those who follow the Indian tech scene, what’s the best explanation of why China is able to birth DeepSeek and India is not?” he asked.
The question was answered by a man who goes by the name “GDP” on X. According to his profile, he works for Amazon in the artificial intelligence space.
The X user opined that India doesn’t have its own major AI foundation models because there’s no protected market to help local companies grow, unlike China, where the government supports AI development as a national priority.
He suggested that India needs a well-funded, government-backed AI initiative (like a $3B DARPA-style programme) to compete globally.
LLM is easy
The AI pro argued that building LLM or Large Language Models is actually quite easy.
“In AI, LLM science part is actually quite easy,” he said, explaining that the fundamental science behind these models—based on Transformer Decoder architecture—has been around since 2017.
While there have been improvements (like flash attention, Mixture of Experts (MoE), and fine-tuning techniques such as PPO/DPO/GRPO), these are relatively minor and openly available through research papers.
“LLM building is nothing that Indian engineers living in India cannot pull off. Don’t worry about Indians who have left. There are plenty in the country as of today,” the AI expert claimed.
Then why India does not have foundation models?
According to the AI expert, “it is for the same reason India does not have Google or Facebook of its own.”
“There is no protected market to practice your craft in early days. You will get replaced by American service providers as they are cheaper and better every single time. That is not the case with Chinese players. They have a protected market and leadership who treats this skillset as existential due to geopolitics,” he said.
In India, AI startups face intense competition from American service providers, who are often cheaper and better due to their head start. Without strong backing from the government or large Indian corporations, AI labs in India struggle to scale.
The X user also added that DeepSeek is just the tip of the iceberg when it comes to AI in China. "First some more shocks: You heard DeepSeek.
“Wait till you hear about Qwen (Alibaba), MiniMax, Kimi, DuoBao (ByteDance) all from China,” he said.
He estimated that China has more than 10 AI labs comparable to OpenAI, and another 50 tier-2 labs.
Source: Hindustan Times
Bd-pratidin English/Fariha Nowshin Chinika
