Listen to this story
The conversation around building AI in India is finally taking shape and it’s time to take stock of our standing in a world engaged in a blind race of technological advancements. When OpenAI CEO Sam Altman visited India, he said that no one would be able to replicate what OpenAI has done with ChatGPT. To this, Tech Mahindra CEO CP Gurnani had famously said, “Challenge accepted”. And in just about two months, the company launched Project Indus, an indigenous LLM that could speak in many Indic languages.
Most recently, at Reliance’s 46th Annual General Meeting, Mukesh Ambani announced his plans to build India-specific AI models. The Jio chief is probably one of those bigwigs from India who can actually challenge OpenAI. “India has scale. India has data. India has talent. But we also need AI-ready digital infrastructure that can handle AI’s immense computational demands,” said Ambani.
Through Jio, Ambani was able to give broadband connectivity to everyone and definitely delivered on the promise he made — can he do that with AI?
A little too late?
Last year, Reliance acquired a 25% stake in Silicon Valley-based Two Platforms, a company focused on advanced technological projects aimed at creating interactive and immersive experiences through AI interactions, for $15 million. The company is also investing in developing up to 2,000 MW of AI-ready computing capacity.
When it comes to capital and investment, nothing is going to stop Ambani from making a comparatively greater model. As for the talent, Reliance can hire people from across the globe, to some extent even existing leading companies if they wish to. Alternatively, Ambani can simply outsource it to some other company and build an AI model for India (and maybe call it AI-JIO).
Arguably, it might be a little late to come up in the AI space even with Tech Mahindra’s Project Indus still in infancy. Moreover, for the project to be successful, Gurnani has said that it requires contributions from every Indian to build the dataset. An Indic language dataset would just be the first step towards building an AI model to compete with ChatGPT. Still the bid is on and the way from here is only upwards.
Meanwhile, India’s ambition of becoming a semiconductor superpower is still in the making. The government has been investing thousands of crores for the projects, but the delays and inefficiencies in the process have resulted in the loss of one-and-a-half years. To make up for the loss, Semicon India 2023 gave a little hope with all the investments, and partnerships that Vedanta has promised with a ‘world-class technology partner’. Even AMD and Micron are planning to invest in India.
The recent success of Chandrayaan-3 mission has also brought India into the spotlight. Becoming the fourth country to land on the Moon, and the first to touchdown on the lunar south-pole has been a huge feat for the nation. This has also pushed countries like Japan to announce partnerships for the future missions with ISRO.
Interestingly, Elon Musk’s Starlink is expected to offer satellite services in India. The company has already been in talks with the Indian government to expedite the process. Musk also finds it “impressive” that some of the biggest companies in the world are helmed by people of Indian-origin.
Meanwhile, no one can deny the mega success of UPI. The whole world has been trying to replicate the success story in as many ways as possible. Countries like France, Australia, Singapore, Saudi Arabia, Bhutan, Nepal, Oman, Sri Lanka, and Pakistan have already received the UPI technology from India.
So, even though building AI systems may seem a little distant on the horizon for us, the country has already proven itself as one of the leaders when it comes to technological advancements, and is striving for more. As of now, he country can lead the way for quantum computing if IBM comes back to India, or be a semiconductor hub or even crack AI. The possibilities are endless.