AI: a new economic miracle or a bubble ready to burst?

IA nuovo miracolo economico o bolla pronta a esplodere
Yuri Brioschi
17/10/2025
Horizons

In less than three years, ChatGPT has grown from less than one million to around 800 million weekly active users, a figure that equates to 10% of the global online population.

This pace of adoption is unprecedented in technological history. For comparison, the expansion of the Internet took over a decade to reach a similar level of penetration. But the real revolution is not just in the product, but in the speed with which an innovation now generates adoption and economic value.

In the 1990s, the Internet had to build everything from scratch (cables, infrastructure, capital and digital culture). Today, thanks to that legacy (smartphones, social networks, global spread of the web), a service like ChatGPT can reach hundreds of millions of users in three years. The time to transform capital into value has compressed like never before.

The shadow of sustainability

However, beyond these extraordinary numbers lies a deeper complexity. Each new user and each request for artificial intelligence implies a significant increase in processing demands, a greater demand on server resources and ever-expanding energy consumption.

This raises a crucial question: will we be able to sustain this acceleration?

The adoption of artificial intelligence is proceeding at a dizzying pace, but it cannot be limited to the development of ever more advanced models. It is essential to build a resilient infrastructure, capable of meeting challenges on several fronts: technological, energy and environmental. The future of AI will not be defined solely by its ‘intelligence’, but by its ability to evolve in a sustainable manner.

Market fever: trillions of investments


This rapid adoption is reflected in a buzzing market with staggering numbers:

  • OpenAI is targeting an annual revenue of USD 13 billion. Already in the first half of 2025, it generated $4.3 billion in revenue, 16% more than in the whole of 2023. Its valuation has been estimated at around USD 500 billion.
  • Investments of $3 trillion are expected by 2028 in the AI sector.
  • Multi-billion dollar strategic agreements have already been made:

    • Oracle and OpenAI ($300 billion);Alibaba and Nvidia ($100 billion, with estimated total AI investments for Alibaba of $400 billion over the medium term)
    • Nvidia and OpenAI (100 billion, with Nvidia supplying chips for data centres);
    • Google will invest $15 billion in India to build the largest AI data centre hub outside the US;
    • Even in the more modest European market, ASML, as we have already written on these pages, has become the majority shareholder of Mistral (which in turn also has ties with Nvidia).

These investments, largely coming from the Tech giants rather than from the banking world, suggest an inherent soundness, while at the same time fuelling growth that, according to some estimates, will bring the contribution of the AI sector to about 40 per cent of the increase in US GDP.

The wind in our favour: The role of the weak dollar


The impressive growth in sales and profits of large American companies, including the ‘Magnificent Seven’ (Apple, Microsoft, Nvidia, etc.), in 2025 was significantly helped by an external factor: the record depreciation of the US dollar.

  • These multinationals generate 40-60% of their revenues abroad, pricing in local currencies such as the euro.
  • A weak dollar amplifies foreign earnings converted into dollars: if the USD/EUR goes from 1.10 to 1.20, a €100 product yields $120 instead of $110.
  • This effect acts as a ‘free’ catalyst for Big Tech, pushing earnings and contributing to market outperformance. Goldman Sachs, for example, forecasts the S&P 500 index to exceed 7000 points.

Stock market fears


Despite the euphoria, financial analysts are keeping a close eye on risks, particularly those from the US:

  1. Debt and Tariffs: The frighteningly high US debt ($37 trillion) and the impact of tariffs (wanted or threatened) by political figures such as Trump remain sources of volatility.
  2. Financial Alarms: Jamie Dimon, CEO of JP Morgan, said he was “much more worried than others” about a serious market correction in the next six months to two years.

As evidence that not everyone lives in the ‘gilded world of Tech’, some anomalies are noticeable: gold is at highs (driven by central bank purchases ‘dumping’ US Treasuries) and short-dated Microsoft Bonds are considered safer (yielding less) than US Treasuries of the same duration.

The explosive growth of AI offers a potential ‘explosive global growth’ scenario (some estimates assume +20% of GDP), but the question remains: if mass adoption slows down, would Big Tech be willing to sustain the monstrous investments (Meta alone forecasts costs of 150 billion in 2026) with revenues only in the medium term?

The future is played out on this balance between dizzying innovation and the sustainability of a market that is rewriting the rules of technological and economic progress.