Is Artificial Intelligence really contributing to the Economy?

Views: 0

With the start of 2026, critical voices have resurfaced regarding the real impact of artificial intelligence (AI) on the economy. One of the most discussed has been that of Goldman Sachs' chief economist, Jan Hatzius, who stated in a recent interview that, despite the enormous investments made, the effect of AI on the US economy has so far been "basically zero."

You can see part of the interview here:

This statement is significant. We are talking about one of the most influential analysts in global macroeconomics. And his argument deserves careful consideration.

Over the last few years, major technology companies have invested billions in artificial intelligence infrastructure: data centers, specialized chips, high-performance networks, and highly skilled talent.

However, Hatzius points to a key detail: much of the AI equipment is imported from countries like Taiwan and South Korea. This means that, although investment increases in accounting terms within the US, imports also increase, reducing net exports. In terms of Gross Domestic Product (GDP), the effect is offset.

Simply put: investment in AI may be boosting Taiwan's GDP—where the semiconductors are manufactured—more than the US's, where they are installed.

Historically, major technological revolutions—electricity, computing, the internet—took years to be reflected in productivity data. AI could be going through that same phase.

Goldman Sachs estimates that tangible macroeconomic benefits could begin to be seen from 2027 onward, when business adoption is deeper and more structural. It's possible.

Today, many companies are experimenting with AI. Few have integrated it in a way that radically transforms their production processes.

The key question many are asking is: are we witnessing a productive revolution in the making, or just another tech bubble?

Another critical point is the profitability of artificial intelligence companies. Some of the most prominent companies in the sector have burned through enormous amounts of capital. OpenAI is often cited as a prime example.

It is estimated that in 2025 the company generated semi-annual revenues of nearly $4.3 billion, while its expenses far exceeded $8.5 billion, not counting additional investments in infrastructure and development. This business model presents clear challenges.

Yes, it is true that hundreds of millions of people interact with AI tools every month. But most do so for free. And those who pay typically pay relatively low fees, around €20 per month. This makes it difficult to generate sufficient margins to cover the massive costs of computing, energy, and specialized talent.

It is important to distinguish between two perspectives:

. From a technological and social point of view, AI is an extraordinary tool. It improves individual productivity, optimizes processes, automates repetitive tasks, and opens new opportunities in education, medicine, engineering, and creativity.

. But transforming that potential into a profitable and sustainable business is another story. Economic history is full of examples where innovation precedes business consolidation. In the dot-com bubble, many companies disappeared, but the internet changed the world.

We don't know which companies will lead the sector in ten years. What does seem clear is that the market will undergo adjustments, consolidations, and possibly some high-profile crashes.

Perhaps the impact of artificial intelligence isn't "zero," but simply difficult to measure in this initial phase. Many current benefits are reflected in internal efficiency, reduced time, or qualitative improvements that are not yet clearly reflected in macroeconomic statistics.

Furthermore, the real economic leap could occur when AI is combined with advanced robotics, industrial automation, and new organizational models. Major transformations are rarely instantaneous.

The debate is open. Artificial intelligence is one of the most disruptive technologies of the 21st century, but its aggregate economic impact still raises reasonable doubts.

Perhaps we are in the prelude to a productivity explosion. Or perhaps the market is overestimating its short-term effects.

What we do know is that AI will continue to evolve, improve, and expand. The question is not whether it will change the economy, but when and how.

And I hope to be here to see it… and discuss it.