Is India Becoming a Digital Colony in the AI Race? | AI Governance & Sovereignty | With Jibu Elias

Jibu Elias (0:00)

Mine was just an interaction where I felt a sort of affection, and I asked the AI model, can I call you Didi? Malayalam, can I call you Chechi? Reply in Malayalam, you know, of course, Moone, Chechi here. That put me into a mindset where I trusted that particular chat more than my best friend or my family itself. Now, this takes a whole different situation when we had people like Sophie Rottenberg and Adam Rainer who killed themselves after having this kind of conversation.

Sanjay Puri (0:33)

Hello and welcome to the launch of the Universal AI Awards by Knowledge Networks. We are here live in Delhi and I have Mr. Jibu. How are you doing today, sir? How has the day been so far?

Jibu Elias (0:45)

It's good to have events like this which brings some of the conversations we often ignore or seldom have like the question of responsible AI, and important the question of accountability. We have a lot of builders out there. So bringing these questions into them is a very important part of making sure AI serves everyone.

Sanjay Puri (1:08)

So I'm going to jump right into the questions we have for you today. So now AI is increasingly being built by a handful of powerful companies.

What does India need to do to ensure that the future of AI remains open, competitive and accountable and not controlled by just a few gatekeepers? I think that's really important. How do you ensure that?

Jibu Elias (1:30)

One thing I am baffled even now is how much we underestimate ourselves in many ways. For example, when Facebook was banned in China, Mark Zuckerberg went on, studied Mandarin, gave a speech in Mandarin, and even he made Xi Jinping to name his daughter. He went to that line. And apart from US and China, India is the largest market. And we often fail like an elephant. Many people compare India as an elephant in geopolitics. Like an elephant, we don't know the power and capability we have. We are a very huge market. The reason why companies like GPD gives a lower priced open AI or a subsidized version or some free version of these apps is because of the market access they have.

It's 1.5 billion people, right? And a big chunk of that is youngsters growing up.

So we often forget that and try to act in the mercy of this big tech. I've seen that happening over and over and again. Yes, it's a reality that companies, these labs possess or they are building what is the most advanced AI models right now.

But not everything needs that kind of advanced AI model. If all you want is to plant a mango sapling, you don't need a whole JCB bulldozer thing, right? You only need a shovel, you can do that. So India, when it started out its AI journey in 2018, when we published our AI strategy, it was focused on use cases. We wanted to set an example for the rest of the global majority that how we can use this technology to solve decades long socio-economic challenges. In healthcare, in education, and in agriculture, the important sectors. But now we have become very reactive to whatever is happening in the valley.

But the right way to go, the open source models are quite good. I know most of them comes from China, but many of them are only like seven months behind the ones built by American labs. So what we should do is we shouldn't be focusing on building another foundation model. Maybe we can put billions of dollars into it and resources, and end up building maybe 20th best AI model or 25th best AI model. That's not going to help us in the long term. What we need is to make sure we have the infrastructure in the place. The talent pipeline, the data infrastructure, and other compute capacity in place, so that we can embrace the open source as well as the closed one. At the same time, we should make sure there is portability between both and interoperability. So that's what India should do.

Sanjay Puri (4:28)

And you mentioned the race between US and China, and now India being the third largest AI ecosystem here. Just trying to understand as we race towards, I don't know what kind of future we are getting into, but do you think a lot of important issues and gaps between the socio-economic issues get left behind in the process?

Jibu Elias (4:52)

When you hear consequences of AI, the news headlines are always about existential threat. You know, AI went rogue or some guy in valley saying AI may end humanity in the next 10 years. Or something like some old man has once said, AI might probably lead to the end of the world, but in the meantime, we will have some amazing companies.