AI: The Technological Revolution That Powered the Markets in 2025 (2/3)

No single force shaped the trajectory of financial markets in 2025 more decisively than artificial intelligence. AI became the dominant narrative through which investors justified elevated valuations, extreme concentration, and rising risk-taking. It offered a forward-looking story powerful enough to neutralise concerns about slowing growth, mounting debt, and geopolitical fragmentation.

At its core, AI is a genuine technological revolution — a productivity leap comparable to electrification or the internet — and it is already reshaping entire industries. But financial markets do not price long-term societal benefits; they price microeconomic realities: corporate cash flows, capital intensity, and sustainability. It is at this level that a growing disconnect emerged.

In 2025, the AI ecosystem was defined far more by capital expenditure than by profits. Training large-scale models, building data centres, securing advanced chips, and powering energy-intensive infrastructure required investment on a scale rarely seen outside heavy industry or national infrastructure programmes. The result was an unprecedented surge in capital spending across hyperscalers and AI platform providers.

The explosive rise of NVIDIA — the undisputed hardware backbone of the AI boom — to a valuation approaching $5 trillion reflected not only technological leadership, but also a global investment race. Companies and governments rushed to secure computing capacity, often committing capital well ahead of proven end-user demand. For many, these investments were less about immediate returns than about strategic positioning: buying today to avoid being left behind tomorrow.

As the year progressed, markets began to question the sustainability of this investment cycle — and, more importantly, the circular nature of its financing. What remained largely overlooked, however, was the issue of terminal profitability. Valuations increasingly discounted a future of sustained hyper-growth with little regard for return on invested capital or saturation risk. Early data now suggest that only a small fraction of AI ventures will ever reach meaningful profitability, raising the prospect that a significant portion of the roughly $1.5 trillion in planned investments may ultimately be written down.

The financial profile of the sector underscored this imbalance. OpenAI, widely viewed as the technological vanguard of the AI revolution, is expected to generate roughly $19 billion in revenue while losing close to $13 billion — a stark illustration of how far monetisation lagged behind investment.

In this sense, AI in 2025 functioned less as a conventional growth engine than as a belief system. It allowed investors to rationalise extreme valuations and historic concentration. By year-end, just ten highly overvalued stocks accounted for roughly 45% of the Nasdaq-100’s capitalisation and 34% of the S&P 500 — a precarious configuration should growth expectations fail to materialise.