Original article: ¿Será la IA la nueva Babel del siglo XXI?

Is AI the New Tower of Babel for the 21st Century?

Just as the builders of the Tower of Babel believed their technology would overcome their limitations, there is a contemporary risk of overestimating AI’s ability to unify human knowledge. In reality, the accumulation of biased data and technological hallucinations often exacerbates confusion rather than clarifying truth, potentially leading to a massive linguistic and cultural disconnection.

By Germán Gorraiz López

Parallelism with the Tower of Babel

The myth of the Tower of Babel tells how humanity, in trying to build a tower to reach the heavens, was punished with the confusion of languages, preventing cooperation and scattering peoples. In the current development of AI, the risk is both structural and human. As different nations and corporations create their own artificial intelligences in isolation, we edge closer to a scenario of global misunderstanding.

It is likely that a phenomenon similar to the Tower of Babel could occur with AI due to geopolitical fragmentation, the absence of universal standards, and disconnection amid closed technological ecosystems.

The Illusion of Omniscience (Hubris): Hubris is understood as excessive pride or arrogance of a mortal who dares to compare themselves to the gods or challenge the cosmic order. When a human falls into hubris, they inevitably provoke the wrath of the gods, who send Nemesis (divine justice or punishment) to restore balance through their downfall or tragedy. Just as the Tower’s builders thought their technology would resolve all limitations, today there is a risk of overestimating AI’s capacity to unify human knowledge.

Factors driving an «AI Babel»

Technological and geopolitical sovereignty: Countries such as the USA, China, and EU member states are striving to create their own sovereign AI systems, subject to differing laws, ethical values, and censorship regimes. Consequently, an AI trained according to Western norms will possess an ideological and operational «language» distinct from that of an AI developed in Asia or the Global South.

Model incompatibility (multi-agent systems): Technology companies are creating closed, proprietary models—so-called «walled gardens.» When one company’s systems attempt to communicate or negotiate with another’s, the absence of universal communication protocols leads to errors, distortions, or the loss of critical information.

Cultural and linguistic biases: Most large language models are trained predominantly in English or other dominant languages. Although local models exist, the loss of cultural nuances associated with minority languages ​​creates profound gaps in interpretation, effectively isolating entire communities from the advanced AI ecosystem.

The AI ​​Winter

An AI winter could be triggered by a lack of short-term profitability. If current massive investments fail to yield expected returns, major tech companies will likely cease work on open standards; in a bid to maximize every euro of profit, they would further lock their technologies within «walled gardens.» Instead of a globally interconnected AI landscape, we would face isolated systems that function only with a subscription to a specific brand, leading to a fragmentation of knowledge.

Massive linguistic and cultural decoupling

Should funding dry up, only a handful of wealthy nations (such as the USA or China) or massive corporations would remain capable of investing in the necessary infrastructure. Consequently, minority languages ​​and developing nations would be excluded from technological progress, as training models for these contexts would no longer be profitable. This would lead to a massive linguistic and cultural decoupling—the very essence of the Babel myth. In summary, the AI ​​financial bubble would not bring the technology to a complete halt, but it would shatter the dream of a universal, open, and collaborative AI, leaving us with expensive, fragmented tools incapable of communicating with one another: the new Babel of the 21st century.

By Germán Gorraiz López