Views: 0
Every day, we read headlines about artificial intelligence, and every day they seem to contradict one another. Some say AI will surpass us in a few years; others call it an exaggeration. Some argue for slowing down its development, while others push to accelerate it. Amidst all this noise, it is natural to feel lost.
And yet, one issue deserves our attention above all others: the safety of these systems and the services that are already part of our daily lives.
For most of us, AI remains a "black box." We type a question and, within seconds, receive a surprisingly accurate answer. What we don't see is everything happening behind the scenes: the training, the data, the model's internal decisions... and the enormous energy consumption behind every query.
The data centers that train and run these systems consume electricity at a rate that is hard to imagine, and a significant portion of that energy still comes from polluting sources. This is the other, less visible side of every instant answer we receive.
We have now reached the point where AI is beginning to improve itself. One of the most striking recent developments is that some companies are already employing thousands of AI agents working in parallel to design and train the next version of their own models. In other words, AI is collaborating to build an AI more advanced than itself.
It sounds like science fiction, but it is the direction the entire sector is taking. All the major companies are competing to launch the most powerful model, and there seems to be no stopping this race.
The problem is clear: if we already struggle to understand how current models reason, what will happen when these systems accelerate their own development with increasing autonomy? The faster they advance, the harder it becomes for humans to understand—and, above all, control—them.
We hear rhetoric that does not always align with the facts. From time to time, a tech leader publicly calls for a pause in AI development to allow for the establishment of control mechanisms. Such statements raise reasonable doubts: is this a genuine concern, or merely a way to look good in the public eye while internal development continues unabated?
At the same time, a contrary message is coming from certain governments (such as the US): we must accelerate without regulatory brakes to avoid losing the technological race to other powers. These are two opposing narratives—both highly influential—yet neither is primarily focused on citizen safety.
I believe AI is a fantastic technology that requires rules. It is a tool with enormous potential for medicine, education, science, and daily life. The problem lies not in the technology itself, but in the lack of control over who decides how it is developed and who benefits from it.
Without oversight, there is a risk that these advances will end up concentrating power and wealth in very few hands, rather than distributing the benefits across society as a whole. And this is already happening.
However, the responsibility also lies partly with us. In democracies, a significant part of this problem stems from a very human source: we elect representatives who subsequently fail to prioritize our safety or our interests. Furthermore, AI itself can become a highly effective tool for manipulating opinions and decisions.
Yet, we retain something that no machine can take away: the ability to think for ourselves, to stay informed, to demand transparency, and to make choices based on our own judgment.
Ultimately, it is partly up to us.
EN 
ES