Big Tech companies remain publicly confident that artificial intelligence (AI) will eventually justify the huge sums of capital pouring into the sector. Yet, concerns have grown over less-than-commensurate returns and high market volatility amid reports that OpenAI may defer its long-awaited public offer of equity.
All this suggests that the breakneck pace of today’s AI boom may not be sustainable. Goldman Sachs projects capital spending of $5.3 trillion by hyperscalers between 2025 and 2030, much of it in AI chips and data centres.
Should today’s AI narrative falter or investment plans go awry, though, it may prompt investors to look anew at AI latecomers like India.
While Microsoft, Amazon, Alphabet and Meta invest heavily in AI infrastructure, OpenAI has a web of alliances across the ecosystem and hardware firms like Nvidia have been recycling their AI windfalls into this space through equity and other deals that exchange chips for future computing capacity.
In its March Quarterly Review, the Bank for International Settlements highlighted that a growing share of AI infra investment is being financed through devices that resemble debt but stay largely off corporate balance sheets. The result is a circular capital structure that exposes AI firms, infra providers and investors to one another’s fortunes.
An abrupt market correction could thus prove more disruptive than the railway, power and web analogies often invoked to reassure investors.
While even overbuilt rail, electricity and fibre-optic networks proved useful for decades, the AI build-up may pan out differently as tech hardware turns obsolete sooner. Advances in model architecture, algorithmic efficiency and specialized chips are already reducing the need for brute-force computing.
Moreover, if investors begin to foresee an AI future that lies less in ever-larger frontier models and more in local models of practical utility, capital may migrate.
Here’s where India could enter the picture. The country is on the global map already for some parts of the action. Over the past year, players like Amazon, Microsoft and Google have lined up investments of over $55 billion to expand their AI and cloud footprint in India; others could follow suit.
Sure, the US and China are way ahead on development, while our chip ambitions remain nascent, sovereign AI infra is still being set up and local language models are only beginning to mature.
In capital deployment, India cannot match global leaders. But a slower AI race can change the rules of rivalry.
History offers precedents. Japan did not pioneer the industrial revolution. South Korea didn’t invent semiconductors. Yet, both exploited tech transitions to create globally competitive industries.
A correction that shifts AI’s centre of gravity from frontier models to local applications and diffusion would favour many of India’s domestic capabilities, such as a deep software talent pool, world-class IT service firms, robust digital public infrastructure and a huge domestic market hungry for tech solutions.
On the flip side, a sharp drop in global AI investment could also reduce local venture funding and weaken our startup scenario. So the opportunity is not guaranteed.
However, in a country that’s still laying the foundations of its AI ecosystem, a repricing of expectations that shakes up the global market for AI assets but keeps AI’s promise intact may offer us just the strategic opening we need. On an untrodden path, as Bob Dylan sang, the swift don’t always win what the worthy do.