DeepSeek AI model is a sobering idea for India

The Chinese artificial intelligence breakthrough reveals how far behind India is in the race for technological prowess

- LATHA JISHNU

CHINA'S TECHNOLOGICAL breakthroughs tend to create headlines; and as they burst onto the scene, a befuddled world is usually asking how it was done. But nothing has been as spectacular as the emergence of the DeepSeek artificial intelligence (AI) model that was developed by a young hedge fund manager on a slender budget in Hangzhou, and is just two years old. We are familiar with the wide swathe of destruction it unleashed in the US, shaving off nearly US $1 trillion from the market value of the big guns like AI Open, Nvidia and a host of other players. The shock and awe worldwide was unparalleled, prompting the head of a stellar Big Tech company to hail the debut of the Chinese startup's R1 model—R stands for reasoning—as "impressive". Another termed it "the most amazing and impressive breakthroughs" he had ever seen.

But what of India? I was keen to know what the country made of DeepSeek, particularly because it was a puny David that had taken on the titans of the tech industry like Open AI. As a rule, the Indian government tends to ignore any news demonstrating China's technological prowess but makes what are patently delusional claims about its own progress. If pressed for answers, ministers usually claim that India would soon be a leader in some technology or the other with vague claims that are seldom questioned by the media. In 2022, for instance, the then Union telecom minister Ashwini Vaishnaw claimed at a very late launch of 5G in the country that Indian developers had many of the technologies required for development of 6G, and that the country would soon be a leader in the next-generation network since they held many patents in this field. This was stated when India was still playing catch-up on 5G with the rest of the world (see "India patently way behind on 5G", Down To Earth, 16-31 October, 2022).