Cognitive labour is decoupling from human population growth. That is the premise behind Adani Group CFO Jugeshinder 'Robbie' Singh's argument that frontier AI companies should be valued differently from traditional businesses.

"Cognitive labour is decoupling from human population growth," Adani Group CFO Jugeshinder 'Robbie' Singh said, arguing that artificial intelligence could allow businesses to expand the amount of work they perform without having to increase their human workforce at the same pace.
Singh was explaining why he believes AI companies such as Anthropic should be valued differently from India's largest listed businesses. As AI agents take on tasks ranging from writing code and analysing information to handling business processes, their deployment could expand without requiring a proportional increase in human employee.
Singh's remarks came in response to Uday Kotak, founder and director of Kotak Mahindra Bank, who questioned India's position in the global technology race after Anthropic's valuation reached $1.5 trillion ahead of its Nasdaq debut on Thursday.
Anthropic, the maker of the Claude family of AI models, priced its initial public offering (IPO) at $83 per share, above its earlier indicated range of $70–78. The offering of 980 million shares is expected to raise around $81.3 billion, valuing the company at approximately $1.5 trillion on a fully diluted basis. Founded in 2021, Anthropic is set to begin trading under the ticker symbol ANT.
In a post on X, Kotak pointed out that the valuation of the five-year-old company exceeded the combined market capitalisation of India's 10 largest listed companies. "Are we undervalued, or do we need to catch up in the new world? Or both?" he asked.
Singh described the comparison as a "textbook category error in asset pricing", arguing that traditional companies and frontier AI developers operate in fundamentally different economic categories.
Banks, energy companies and consumer businesses, he said, generate returns from an economy driven by human population, consumption, credit demand and physical activity. Their growth is therefore linked to factors such as demographics, household spending and working hours.
Frontier AI models, in contrast, could power what Singh calls a parallel "synthetic workforce" made up of autonomous AI agents. These systems can perform tasks such as analysing documents, writing code and handling business processes, potentially operating around the clock without requiring a proportional increase in human employees.
Singh describes this emerging source of economic output as "synthetic GDP". His framework visualises total output as a combination of human-generated economic activity and output from AI-driven systems. The term is a conceptual description rather than an officially recognised measure of GDP.
He argues that investors are not valuing Anthropic only on its current earnings, but also on the possibility that its AI models could become the underlying infrastructure for this new digital workforce.
"Frontier models aren't SaaS tools; they are the 'infrastructure of intelligence' for a parallel synthetic workforce," Singh said, describing Anthropic as a "call option on synthetic GDP". In financial terms, he is comparing the valuation to a bet on the possibility of a much larger future market, rather than a valuation based solely on present cash flows.
Singh also argued that the economic opportunity from AI would extend beyond companies developing the models. Large-scale AI deployment requires computing infrastructure, semiconductor processors, data centres, cooling systems and reliable electricity supply.
"Intelligence is written in code, but it runs on electrons and sovereign infrastructure," he said, pointing to the role of energy, power transmission, ports, metals and computing capacity in supporting the AI economy.
The argument is also relevant to Adani Group's own investment plans. In February 2026, the group announced a direct investment commitment of $100 billion by 2035 to develop renewable-energy-powered, hyperscale AI-ready data centres. It said the plan would build on its existing 2 GW data-centre platform and expand towards a 5 GW target. The group also expects the investment to catalyse an additional $150 billion across server manufacturing, electrical infrastructure and sovereign cloud services.
The broader question, therefore, is not just which companies will develop the most powerful AI models, but which businesses will capture the economic value generated by their adoption. That could include model developers, semiconductor makers, cloud providers, data-centre operators and companies using AI to improve productivity.
Singh also invoked the dotcom boom, arguing that the collapse of internet stocks in 2000 did not invalidate the long-term transformation brought about by the internet. He said AI could represent an even larger shift, while acknowledging that market corrections could occur along the way.
His response reframes Kotak's question: rather than comparing Anthropic directly with India's largest traditional companies, Singh argues that investors should consider the economic value of the new AI infrastructure and workforce taking shape.