The artificial intelligence boom has captured the imagination of markets, with billions of dollars pouring into the sector across every corner of the global economy.
History offers sobering lessons about transformative technologies that reshaped civilisations while simultaneously destroying the fortunes of early investors who backed them.
The railways transformed 19th-century Britain and America, yet the speculative frenzy that accompanied their construction wiped out enormous sums of private capital along the way.
The internet similarly revolutionised commerce, communication, and daily life, but the dot-com crash of the early 2000s left countless investors nursing catastrophic losses despite the technology itself proving world-changing.
The core paradox at the heart of AI investment is that a technology can be genuinely transformative and still fail to reward those who funded its rise.
Competition is one of the most powerful forces working against outsized investor returns, as rival firms race to deploy similar capabilities and drive down the profit margins of any single player.
When a technology becomes widely available and commoditised, the economic gains tend to flow to the businesses and consumers who use it rather than to the companies that build it.
The enormous capital expenditure required to develop and maintain AI infrastructure means that even strong revenue growth can struggle to translate into the kind of profits shareholders are expecting.
Nvidia (NVDA), which manufactures the chips that power AI systems, has so far been a rare example of a company capturing substantial financial value from the current wave of AI investment and enthusiasm.
Yet even dominant positions in fast-moving technology markets are rarely permanent, and new competitors, alternative chip architectures, or shifts in demand could erode that advantage over time.
The broader question facing markets is whether the current valuations placed on AI-linked companies accurately reflect realistic future earnings or represent a speculative premium that may never be justified.
Investors piling into the sector today are, in effect, making a bet not just that AI will succeed as a technology, but that specific companies will successfully monetise it before competition erodes their edge.
The most likely outcome, if past technology cycles are any guide, is that AI will prove every bit as powerful as its proponents claim while the financial rewards remain stubbornly concentrated or disappointingly thin.