AI : Bubble or Boom?

 

A financial bubble is, at its core, a collective illusion—an economic cycle driven more by emotion than by fundamentals. It begins with genuine innovation, accelerates through exaggerated expectations, and eventually collapses when reality supplants belief. Every bubble follows a recognisable life cycle: stealth phase, when early investors quietly accumulate positions; awareness phase, when institutional players join in; mania phase, when retail investors flood the market in fear of missing out; and finally, the blow-off phase, when prices crash under the weight of their own excesses. By every measurable indicator, the global artificial intelligence boom has entered the later stages of this cycle. The symptoms are textbook—soaring valuations, concentrated gains, speculative capital chasing the same narrative, and a feedback loop between hype and investment. The parallels with the dot-com era are too striking to ignore.

Let’s start with the numbers. In 2024 alone, AI-related stocks contributed nearly 80% of the S&P 500’s total gains, with Nvidia, Microsoft, and Alphabet at the centre of the surge. Nvidia’s market capitalisation crossed USD 3.5 trillion, briefly making it the world’s most valuable company—worth more than the entire GDP of the United Kingdom (approximately USD 3.2 trillion). Meanwhile, price-to-earnings multiples across the AI sector have expanded to levels reminiscent of 1999, despite most companies having limited or no immediate profitability. This is not to deny the legitimacy of the technology itself. Artificial intelligence is transformative—it is reshaping industries from finance to pharmaceuticals. But bubbles are not built on falsehoods; they are built on exaggerated truths. The internet was real, yet dot-com valuations still collapsed. Crypto was innovative, yet it too succumbed to speculation. AI today sits on the same precipice—a revolutionary technology inflated by irrational expectations.

The structure of the current boom is even more concerning because of its circular financing loop. Major technology companies are investing in their own customers to sustain momentum. For instance, AI startups raise venture funding to buy Nvidia chips, while Big Tech firms like Microsoft invest in AI labs such as OpenAI, whose products, in turn, fuel Microsoft’s valuation. This self-reinforcing system creates the illusion of unstoppable growth—but it is, in essence, growth built on leverage and sentiment, not on organic demand.

Macro-level dependency is adding further fragility. Nearly half of the recent U.S. GDP growth can be attributed directly or indirectly to AI-related investments. Strip away that component, and America’s growth rate would drop sharply. What we are witnessing is not a balanced economic expansion, but a concentrated capital bubble, where trillions of dollars are funnelled into one sector while others—like renewable energy, manufacturing, and healthcare—are comparatively starved of investment. The psychological foundation of this mania lies in the misconception of AI’s capabilities. Despite being labelled as “intelligent,” these systems do not think—they predict. They generate statistically probable outputs, not original reasoning. Yet the market has priced AI as if it possesses limitless cognitive potential. When that collective perception corrects, valuations will follow swiftly.

No one can time the exact collapse of a bubble, but the triggers are almost always the same: slowing growth, capital tightening, or a loss of belief among retail investors. Once the marginal buyer steps back, liquidity evaporates and prices tumble. The crypto winter of 2022 and the dot-com crash of 2000 both unfolded this way—sudden, sharp, and psychologically brutal. Can this AI bubble deflate gradually? History suggests otherwise. Market psychology rarely allows for soft landings. The rational path would be to moderate valuations, redirect capital into sustainable applications, and temper short-term expectations. Yet greed tends to persist until fear takes over. Artificial intelligence will undoubtedly endure beyond the current frenzy—just as the internet did after 2001. The infrastructure being built today—massive data centres, high-performance chips, and advanced models—will define the next era of technological progress. But between here and there lies an inevitable correction. When it comes, it will remind us of a timeless truth: technological revolutions create enduring value, but financial bubbles never do.