Artificial Intelligence

For a long time, economic research relied on statistical tools, mathematical models, and certain assumptions of stability. That foundation enabled a valuable understanding of the economic world—though often general and detached from specific realities. But today, the context has changed.

The Era of Digitalization

Digitalization surrounds us and leaves constant traces: searches, interactions, satellite images, real-time sensors. These are richer, more diverse data, with a speed and depth that traditional methods alone cannot fully capture.

The Role of Artificial Intelligence

This is where artificial intelligence comes in. It allows us to see invisible patterns, process overwhelming volumes of information, and detect signals we would otherwise miss. It goes beyond our human capabilities and expands our possibilities.

Challenges and Responsibility

But for it to be a true ally, we must acknowledge and address its challenges:

We don’t always know where it gets its data.
Many AI methods operate as "black boxes," generating opacity and making verification difficult. Unlike traditional models, it is nearly impossible for developers or users to ask AI to explain in human terms how or what it has learned. This unintelligibility undermines transparency and trust in its decisions, making them difficult to audit and understand.
Sometimes it repeats mistakes with an authoritative voice.
AI models, trained on large volumes of public and unstructured data, can reproduce and amplify the biases and prejudices inherent in that data. This can lead to the perpetuation of conventional or even mistaken ideas, presented with false confidence. Moreover, social media algorithms can create personalized "echo chambers," fostering discord and amplifying biased information.
And it can impose false consensuses, leaving no room for dissent.
Historically, human information networks have often prioritized social order, even at the expense of truth. AI, if not managed with strong self-correction mechanisms, could amplify this tendency. In an unregulated information market, AI may prioritize sensationalism and falsehoods over truth, thereby imposing consensuses that lack a real foundation and silence dissenting voices.

Overcoming these limitations means building more transparent systems, promoting diversity in data sources, and contrasting results with human judgment. Because AI is powerful—but it will only be transformative if we guide it responsibly.