What's Jio’s '15-cent data advantage' in AI? Mukesh Ambani reveals to NVIDIA’s Jensen Huang скачать в хорошем качестве

What's Jio’s '15-cent data advantage' in AI? Mukesh Ambani reveals to NVIDIA’s Jensen Huang

Reliance Industries Limited (RIL) chairman Mukesh Ambani, in a fireside chat with NVIDIA CEO Jensen Huang at the NVIDIA AI Summit in Mumbai today, said apart from the U.S. and China, India has the best digital connectivity infrastructure of 4G, 5G and broadband in the world. "Jio took India from No. 158 in the world to No. 1 in the world in 8 years. We, as a single company, didn't know anything about this domain. But today, we are the largest data company in the world, with volumes equivalent to AT&T, T-Mobile and Verizon combined." Ambani says his telecom company, Jio Infocomm delivers data at 15 cents per GB compared to the U.S., which pays $5 per GB and the global average data cost of $3.5 per GB. "At Jio, we delivered about 16 Exabytes of data this year," he says. Huang, while speaking about Prime Minister Narendra Modi, says he "loved" it when he met the PM the first time and he asked him to address this cabinet about AI. "And I was so surprised. It was literally the first time any government leader, any national leader asked me to address his cabinet on this particular topic." Huang says he got inspired by the Indian PM when, around six years ago, he told Huang that AI could "elevate" the entire population of India. "There are so few people in the world who know how to program a computer. Programming is not easy. Here in India, it's the largest population of any programmers in the world." The CEO of the world's biggest chip-making company says AI could be put into the hands of every citizen, and it would elevate everyone. Mukesh Ambani also thanked Meta CEO Mark Zuckerberg, saying his name will be in history for making its AI model Llama 3 an open source. "This move of Mark will be written in the history when we look at it a hundred years from now. From an Indian point of view, we can use Llama as a base model. It allows all of us to develop on top of a state-of-the-art model. And surely then we can fine-tune, train, retrain, and do everything else. I am sure there is somebody in this audience who will have in the future an Indian model, which might be 10X of Llama."