The AI Compute Crisis, Why India is Lagging Behind DeepSeek and OpenAI: After DeepSeek, Indian social media has just one question: If China can, why not India? Why can’t we build a Facebook, or Google, or ChatGPT? Well, there isn’t one single factor. It’s a combination of many.
Factor 1: The R&D Funding Gap
Number one, of course, is money. And I’m not talking about bigger budgets; I’m talking about better budgets. We do not allocate enough money to research and development (R&D).
Just look at the numbers. The US spends 3.4% of its GDP on R&D. China spends 2.4% of its GDP. Israel spends 5.7%. What about India? Only 0.6% spending. This is simply not enough to create breakthroughs.
And it’s not just the government. India’s private spending on R&D is also less. Private companies make up just 36% of our total spend. In China and the US, it is higher than 70%. So, we need the government and the private sector to coordinate to put more money into research and development. That’s the first point.
Factor 2: An Unprotected Market
The second factor is the market. It was explained pretty well by Sridhar Vembu, the chief scientist at Zoho, an Indian multinational company. He says the Indian market is not protected, so startups are vulnerable.
I’ll give you a scenario. Imagine an Indian startup creating a homegrown Google. It’s not perfect, but it’s an alternative and it is homegrown. What do you think will happen to it? Do you think Indians will ditch Google? Do you think they’ll flock to this homegrown version? The answer is no. The startup will perish under competition from American tech giants. Their products are cheaper and well-established, so they will push out the Indian ones. In fact, this is what happened with India’s social media platforms too.
But something like this never happens in China. Their market is closed to American tech companies, so the homegrown ones have wiggle room to fail, to improve, and to finally deliver. So, it’s not always about the quality of the product; it’s also about the environment that they grow in.
Factor 3: Playing Catch-Up in Hardware
Which brings us to the final factor. India is already behind. AI is the cutting edge of technology. You had so many platforms leading up to it: you had the internet, semiconductors, mobile phones, smartphones, and laptops. India is yet to succeed in any of those sectors. And once you fall behind, you stay behind. That’s generally how technology works.
But what if you skip some of those steps? What if you focus directly on new technologies like AI? Then you will need help from others. For example, AI needs specialized semiconductors. Virtually all of them are made by one US company called Nvidia, and the US won’t sell them to you freely. Washington has capped advanced chip sales. You can’t buy as many as you like. Countries like India can only buy 50,000 chips by 2027. That is the upper limit; that is the cap imposed by the US government.
How much is India looking to buy right now? The plan is to source 10,000 chips. Which is not a bad number; OpenAI trained ChatGPT on a similar number of chips, so there is no reason India cannot do the same. The problem is the likes of OpenAI have unlimited access to chips. By the time we make a ChatGPT, they will be a few steps ahead of us.
How Do We Turn This Around?
So, these are the three main factors holding India back: lack of research and development, competition from American firms, and a massive gap with the leaders.
How do we turn this around? How did China do it? They anticipated the curb in chip supply, so China began stockpiling chips in advance. They’re also making their own. The likes of Huawei use China-made chips—apparently, they’re quite good too.
And all of this boils down to one thing: a national campaign. We keep hearing stories about US tech giants building out of basements. Well, that is just one part of the story. The US government was heavily involved in it. The CIA had its own venture capital firm in Silicon Valley. The Pentagon gave them contracts. That is how these tech giants flourished in the first place. In China, it’s the same story. Chinese firms get billions in state subsidies. They have a closed market to use as a training field.
We don’t have any of that. So the need of the hour is a streamlined national mission—one that brings together all stakeholders: the government, the startups, the private investors, and the ordinary people. If not, we will be left on the sidelines in this race.
