'One-tenth the cost, 10 times faster': Ex-HCL CEO warns cheaper AI models threaten Big AI

Former HCL CEO Vineet Nayar on Monday pointed to the rise of younger developers building large language models at a fraction of the cost and much faster. He raised questions over how AI companies will generate returns on their massive investments.

"I can tell you, because I work in artificial intelligence quite extensively, that the innovation that is coming behind is at a much lower cost and much faster," he said in an interview with India Today.

The new models may not solve every problem, Nayar said. But they do not need to. "It will not be able to solve 100% of the problems, but the very fact that it can solve 80 or 70% of the problem at one-tenth the cost and 10 times faster is a market disruptor. The business model is getting disrupted," he said.

Don't Miss: ‘No amount of US pressure should justify recklessness’: Sam Altman backs 'slow' frontier AI pace

Nayar's comments came amid a debate over the future of AI after researchers and top executives at leading AI companies, including Anthropic CEO  Dario Amodei and OpenAI's Sam Altman, warned about the risks of developing the technology too quickly.

"There are lots of young people who are developing LLMs at half the cost and 10 times faster. And therefore, when you bring in this kind of 'sher aaya' concept, that kind of innovation really dies down. So I would take that with a pinch of salt," he added.

The former HCL CEO has dismissed the more extreme warnings as part of an industry hype cycle, while arguing that AI will eventually transform industries and education.

But he suggested that an immediate question for the industry was revenue to justify the enormous investment. "There are trillions of dollars of investment not finding a return on investment," Nayar said. "And, all these young people who are coming from behind are able to do 70% of what they (top AI firms) are doing. So where will the revenue come from?"

In Case You Missed It: 'Sher aaya, sher aaya': Ex-HCL CEO says AI doomsday warnings are hype, won't kill humans

Nayar also questioned whether AI had yet produced a business case comparable to the internet. "And there is no business case like the internet, like a search, or like an Amazon which is going to result in billions of dollars," he said.

Asked whether the concentration of AI power among two or three large companies could be deeply unsettling and unhealthy for society, the former tech CEO said he was less worried, as such concentration has long been a feature of capitalism.

"If you look at the capitalist society which we have built for ourselves, operating system, there is a table of two. Search, there is a table of two. MRP and ERPs, there is a table of two. Pharmaceutical, there is a table of five. Financial services, there is a table of five," Nayar said.

"In a capitalist society, the amount of money you can throw into innovation creates that kind of monopoly and duopoly - and that is inevitable in our scheme of things. But so far we have survived that."

He cited Microsoft's dominance in operating systems and the subsequent rise of Linux as an example of how technological concentration can eventually be challenged.

"Microsoft came in, and it was the exclusive operating system used across the world. Then Linux came in. So we have survived that," the former CEO said. "And therefore I would be less worried about the fact that there are three people or four people talking about AI."