- Nikhil Kamath warned of a widening gap between AI investment and revenue generation
- About $1 trillion is invested annually in AI, but revenue is only $100-$200 billion
- Kamath said funding the $700-$800 billion yearly gap will become increasingly tough
Zerodha co-founder Nikhil Kamath has flagged a widening gap between the money being poured into artificial intelligence and the revenue it is generating, warning that capital for AI could become harder to find in the near term.
Speaking at the NDTV World Summit, Kamath estimated that about $1 trillion a year is now going into AI, while the money coming out of it through enterprise use is “maybe between $100 to $200 billion.”
“That gap of $700-800 billion a year has to be bridged from somewhere,” he said.
Finding The Next Trillion
Kamath said that in the last cycle, a lot of foreign capital, including oil money from the Middle East, went into AI bets in America. But finding about $1 trillion of fresh money every year, he said, “is tough.”
He outlined the bull case often made for the spending. World GDP is about $100–120 trillion, services account for roughly half to 60% of it, and if AI disrupts 10% of that, it could unlock $6–7 trillion in productivity, he pointed out.
But someone has to fund the gap until that revenue story plays out, he said, “even if it does or it doesn't, you know, nobody knows.”
“In the interim, I think it will become harder to get money than it has been this year or the last year,” Kamath said. “So if somebody is able to raise a lot of capital today and keep it, probably not a bad idea.”
ALSO READ: Tax Income, Not Transactions: Nikhil Kamath's Fix For F&O Traders
A Blip For India, Not A Reason For Pessimism
Kamath acknowledged that India could see “some kind of a blip short term because of AI and not having the risk capital to attempt it,” but said that does not make him pessimistic about the country.
“I am 90% allocated to the equity markets,” he said, adding that valuations look attractive across many sectors. He said he is able to buy quality banks, IT services and pharma companies at 11–13 times earnings.
“Barring COVID, I can't remember another time where I've had opportunities like that,” he said.
Kamath also sees an opening for Indian entrepreneurs in AI through labour arbitrage. An engineer in India can be hired for about $500–1,000 a month, he said, and pairing that talent with AI models to build a “half human and half tech” product could serve customers outside India who lack that cost advantage.
ALSO READ: 'Let Chief Ministers Run States Like CEOs': Nikhil Kamath's China Challenge For India
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.