Should there be limits to AI?

Whenever there have been innovations and advances, there have been rejections. And they have led to the birth of opposition movements, such as Luddism, revived today as neo-Luddism, a modern current critical of technology that highlights only the most negative aspects and its effects on health, work, or the environment. That it has effects is evident, and they depend on the use that can be made of it or is made of it, as with nuclear energy, which is used to generate electricity but also to make atomic bombs, like those of Hiroshima and Nagasaki. Taking all this into account, it is clear that AI cannot be a territory without any kind of limitation, as is the case with social networks. And even more so considering that it is in the hands of a small technological elite that should somehow be subjected to rules, as is the case in other economic sectors.

At the European Union (EU) level, the artificial intelligence law was approved, the first comprehensive regulation in the world to regulate this technology. But this is not enough. This technology cannot be left in the hands of a small group of multinationals from the United States, or from China. The EU must step on the accelerator to have its own champions, the only way to gain strategic autonomy and apply European values.

And governments must also have transition plans to address the consequences of AI in the labor market. The consequence of only considering the advantages from an economic perspective carries risks, as the founder of Microsoft, Bill Gates, recalled this week. "It will be the great equalizer or the worst source of injustice", stood out, while warning that "some jobs will disappear forever" if the right measures are not taken.

A few weeks ago The Economist was titled: "The Job Apocalypse". And a study by Brett Hemenway Falk and Gerry Tsoukalas, researchers linked to the Wharton School of the University of Pennsylvania and Boston University, also warned about it. These experts warned that if AI suppresses jobs to gain productivity faster than the economy can absorb them, there is a risk of eroding consumer demand and therefore penalizing the results of the companies themselves. These are all elements to consider.

As Garton Ash recently wrote in a metaphor that the director of the SoReDI chair at Esade, Liliana Arroyo, also uses: "It feels like we have ten Manhattan Projects – the military initiative that developed the atomic bomb – in our hands, but one AI Hiroshima it is not evident". The social risks are evident: from the substitution of human labor to the loss of cognitive abilities due to excessive use. Or the explosion of data centers and their impact on the environment, that has led the Spanish government to the public consultation to place conditions on its expansion.