Every major technological transformation brings with it a dual dynamic. On one side, it generates concrete, measurable change — changes destined to reshape the way we produce, work, and make decisions. On the other, it builds a narrative: often a powerful one, capable of fuelling expectations, enthusiasm, and — in financial markets — valuations that anticipate the future before it has materialised.
Artificial intelligence today inhabits precisely this unstable equilibrium. It is, without doubt, one of the most significant innovations of recent decades. But it has also become one of the most compelling narratives in finance. Understanding whether and when AI represents an investment — and when it is, above all, an emotional lever — is one of the central challenges for anyone managing their wealth with genuine awareness.
AI is not a sector, it is a process
The first mistake — understandable, yet dangerous — is to think of artificial intelligence as a homogeneous sector. In reality, AI is not a “segment” in the way that energy or banking are. It is, rather, a process that cuts across multiple layers of the economy.
There is an infrastructural dimension, comprising data centres, semiconductors, networks, cloud computing, and energy. This is a very concrete world: capital-intensive, subject to industrial cycles, requiring substantial investment, and delivering returns that are far from linear. There is then the dimension of models and platforms, where development costs are enormous and global competition is rapid and often relentless. Finally, there is applied usage: AI embedded within services, software, and the business processes of traditional sectors such as healthcare, finance, logistics, industry, and professional services.
It is precisely here that a crucial distinction lies. The economic value of AI does not arise automatically in those who “do AI”, but frequently in those who use it most effectively. In many cases, the real benefits manifest not as new growth narratives, but as greater efficiency, cost reduction, and improved margins — less visible effects, but far more robust.
Where real value is created
If one attempts to strip the subject of its hype, the economic value of artificial intelligence concentrates in a few fundamental mechanisms.
The first is productivity. AI enables the compression of time and cost in repetitive, support-oriented, or data-intensive activities. This can translate into higher margins, or into the capacity to grow without a proportional increase in costs. The key point, however, is straightforward: the benefit must be measurable. If AI adoption remains a communications exercise without affecting the numbers, the value stays potential, not actual.
The second mechanism is scale. Certain models perform better as the number of users and available data increases. Yet growth alone is insufficient: it must also be possible to defend pricing, retain clients, and make the service genuinely integrated within decision-making processes. When AI becomes easily replaceable, the risk is that the technology swiftly turns into a commodity.
The third element concerns capital structure. AI demands substantial and continuous investment. This favours those with access to financial resources, but also exposes them to cyclical risks. Paying perpetual-growth valuations for businesses that are, in practice, industrial in nature is one of the most common errors during periods of enthusiasm.
When AI becomes principally a narrative
The boundary between investment and narrative is often crossed without one’s noticing. It happens when the price of a share or fund incorporates a near-perfect scenario: sustained growth, elevated margins, an absence of obstacles. In such cases, AI need not “fail” to generate disappointment — it need only grow somewhat less than expected, or at higher cost.
Another typical signal is the confusion between adoption and monetisation. A technology can be adopted rapidly without generating adequate profits. Indeed, the more ubiquitous it becomes, the more it tends to compress prices. This is a counterintuitive paradox, yet it is central to understanding why not every technological revolution has delivered strong stock market returns.
There is also the communications dimension. When artificial intelligence becomes the keyword in investor presentations, yet fails to translate into a structural improvement in results, it is reasonable to ask searching questions. Narrative finance is not necessarily deceptive, but it is intrinsically fragile.
An underappreciated risk: concentration
One frequently overlooked aspect concerns concentration. Many investors are exposed to AI without having consciously chosen to be, simply because a small number of large groups carry ever-greater weight in global indices.
This means that portfolios that appear well-diversified may, in reality, depend substantially on a restricted number of companies and a single dominant narrative. This is not an error in itself, but it is a risk that must be recognised. The real vulnerability arises when concentration is unintentional.
Investing without being swept along by the story
Approaching AI in a balanced manner means relinquishing the idea of “playing the future” and focusing instead on portfolio coherence. An exposure may well make sense, but it should rarely become the cornerstone of one’s strategy.
The most underrated element remains discipline: defining a maximum allocation, accepting volatility, rebalancing when enthusiasm drives everything too high. These are unglamorous decisions, but they are essential to preventing a sound idea from becoming a damaging experience.
Conclusion: less myth, more method
Artificial intelligence is set to have a profound impact on the economy. Precisely for this reason, it has no need of mythologising. Like every great innovation, it will bring benefits — but also imbalances, competition, margin compression, and phases of adjustment.
The genuinely useful question for the investor is not whether AI will change the world. It is understanding who will succeed in converting this transformation into sustainable cash flows, at what price, and with what impact on the overall risk profile of the portfolio.
When those questions find answers, AI can be an investment.
When they are ignored, it remains principally a narrative. And in financial markets, narrative is often the most elegant form of emotional leverage.