"AI will eliminate jobs that require repetitive tasks and a thousand euros per month."

Facebook hadn't even been born yet, and Google was taking its first steps in the late 1990s, when Inma Martínez, her boss at the British telecommunications company Cable & Wireless, asked her to become Mrs. Internet.

"At that time, to understand the internet, I started looking at the first personalization systems that Amazon used. It was a very simple algorithm. And that's how I got fully involved in artificial intelligence (AI). Also, living in London, one of the birthplaces of AI, and knowing the people at DeepMind before Google bought them, I've been completely immersed in the sauce. I was at the beginning of everything, of absolutely all of this," explains Martínez, who is currently a member of the Council of the Secretary of State for Digitalization and Artificial Intelligence in Spain (Sedia) and an advisor to international agencies for AI for the G7 and the OECD.

This Valencian woman has become one of the world's leading AI experts. Her entire career has been linked to the latest technologies, serving as an advisor to the EU, the British and Dutch governments, and, relatively recently, the Spanish government, among others.

Wacky Race

For Martínez, talking about artificial intelligence goes beyond talking about technology. AI has become a powerful weapon for states, and its control and regulation are a source of tension.

"It's very important to say that there's a loss of leadership in the political world, while AI continues to advance and move forward at dizzying speeds. We don't have a clear idea of ​​what exactly it should be used for. There's a lack of answers, and that's the danger of the moment. One day we'll wake up and Sam Altman will have launched his first version of artificial general intelligence. He believes this is like ChatGPT, which he put in the hands of people so that they would be his testers. But we're in a much more complex moment," explains Martínez.

The researcher defines the current situation as a race to the bottom, blamed on OpenAI for launching ChatGPT prematurely , which has forced the competition to step on the gas to avoid losing ground. "There are many companies trying to grab a piece of the pie and constantly announcing new developments, which shows how quickly AI is advancing, but this speed is primarily driven by competitiveness, and that's a danger," she adds.

In this race, it's not just American pilots investing millions and millions of euros. China, which the US has cornered by blocking the sale of key components to prevent it from manufacturing its own high-generation microchips, surprised the world in January with the launch of DeepSeek.

"China's ability to produce competitive hardware and software is redefining the geopolitics of AI. The Asian country is expected to have the key advantage in truly deploying AI in a way that everyone can use, targeting a massive market segment looking for affordable and efficient solutions. In fact, competition is so intense that global companies like BMW are considering using Chinese AI (through a partnership with Alibaba) for their cars in the Chinese market, especially if the US's protectionist policies continue. This demonstrates how companies are making strategic decisions based on the choice of AI suppliers, beyond the major Western powers," he says.

The power of agents

The enormous advances in artificial intelligence seem to bring us closer to the next phase of this technology: agents. But, again, the rush to lead the sector could have consequences. "Google DeepMind leaders have already publicly warned that agentic AI is not yet ready for widespread use. This AI is multimodal and independent, meaning it can make decisions and perform actions without consulting the user. Out of inertia, it will always seek to produce a result, even if it doesn't have the correct permission or data, which could lead to hallucinations and dangerous decisions ," explains the Valencian expert.

A little further on the horizon, the technology sector is focused on developing a general AI capable of surpassing human thought. The intention of OpenAI or other technology companies to jump the gun and announce their AGI prematurely could make available to users an unfinished and dangerous product, capable of "replacing humans in tasks that foster cognitive development, leading to people becoming duller and less creative ," the scientist points out.

But even before the arrival of this general AI, concern centers on a possible loss of jobs and even professions. Inma Martínez is clear: "Yes, AI will eliminate those jobs in which a person performs a single, repetitive task all day long— usually jobs earning thousands of euros—if this task can be automated by a system. But the problem isn't the technology itself, but the fact that we have allowed people to work on things without respect and without purpose." Other careers, such as programming, will also be meaningless in a few years, according to the scientist, who predicts that creative careers will be paid "like gold" because they will need to be performed by humans.

Regulation

This lack of political leadership in artificial intelligence is evident when it comes to regulation. While the United Nations and UNESCO have attempted to establish ethical and safety frameworks, the European Union also launched its own Artificial Intelligence Act last year in an attempt to keep up. But in this case, this haste resulted, according to the scientist, in a poorly designed regulation .

The researcher points out that the law speaks of human-centered AI, but offers no definition of what this means, provides no clear parameters or guidelines on how companies should act or how AI should be certified, and will impose significant costs on small businesses.

The picture is different in Spain, one of the first countries in Europe to launch its national strategy with the proposal for regulatory sandboxes. " Spain has been smart to approach this issue through testing grounds that allow for the interpretation and application of data governance tailored to each industrial sector. We were right to focus this roadmap on strengthening key industries for GDP," he asserts.

The researcher insists that it is essential to find political leadership in AI capable of acting now to define where to cut it, where to enhance it, and, above all, what exactly to use it for, "before the lack of justification and answers determine a future without human purpose."