A CHAT WITH DEEPSEEK ABOUT 'SOMETHING ELSE'

CHINA'S DeepSeek distinguishes itself from other AI models like ChatGPT, Bard and Claude by focusing on cost-efficiency, optimised computing methods and scalability.

- C P SURENDRAN

CHINA'S DeepSeek distinguishes itself from other AI models like ChatGPT, Bard and Claude by focusing on cost-efficiency, optimised computing methods and scalability. While models such as ChatGPT (OpenAI) and Bard (Google) rely on thousands of graphic and tensor processing units to support their architecture—resulting in high computational and operational costs—DeepSeek uses fewer chips and simpler coding to achieve efficient performance, reducing both cost and resource requirements.

In 1923, Sam Altman of ChatGPT met with fawning Indian business tycoons, in which he wished good luck to his competitors, making it clear that he was far too ahead in the game. "They can try," he smirked. This was pretty much what Nvidia chief Jensen Huang, too, implied last year in yet another Indian VIP meeting, where one of the chief interlocutors was the actor Akshay Kumar.

Last week, I was gratified that American technocrats lost out. I am the crowd: a hero bites the dust, and my day is made. So, on Sunday, determined to contribute to the further destruction of the status quo and to add to the traffic of DeepSeek, I asked the Chinese AI, who its founder was. Was it Liang Wenfeng? I had just read an article and did know a little about the 40-year-old entrepreneur, a coding genius who repatriated from the US to his homeland.