AI and robotics: Without a rapid ramp-up, there’s no salvation for Europe!

The first humanoid robot in history is called Talos. According to this myth, which is over twenty-eight centuries old, Hephaestus, god of fire, forging and metals, and a genius craftsman, created this bronze giant to serve King Minos as the guardian of Crete. Three times a day, Talos circles the island, hurling rocks at foreign ships to prevent them from docking. If any sailors manage to land, he burns them by hugging them to his fire-reddened chest. But Talos is more than just a killer robot: he also dispenses justice in the countryside, using bronze tables of the law that he carries with him. So he also has intelligence.

This myth, long considered a minor part of Greek theogony, has taken on a whole new dimension in our time. Although it lacks any great scientific credibility, the story of Talos takes us back a long way in time to the idea of creatures more or less in the image of Man, but servile, even indefatigable, used to do all the tasks that humans want to get rid of. The etymology of robot, derived from the Czech ‘robota’ meaning ‘chore’ and ‘slave’ in Old Slavonic, is unequivocal.

However, with the help of artificial intelligence and machine learning, robotics is opening up new prospects for humanity, both promising and frightening.

The use of AI models can enable robots to perform their tasks more efficiently, learn from experience and interact more intuitively with the real world, with all the expected benefits in sectors such as healthcare, manufacturing and logistics. This progress is dreamt of as collaborative and inclusive, the idealised fruit of synergy between humans and robots.

But there is a darker way of looking at the future of a world dominated by robotics and artificial intelligence. Isaac Asimov, both science fiction writer and scientist, foresaw the potential dangers of robotics and proposed three laws to ensure that this technological advance would not ultimately harm humanity[1].

Whereas the human brain is limited in its development by the cranium, artificial intelligence can develop ad infinitum. Everything depends on the intentions and choices of those who build it, the powers at their disposal and the goals they set themselves. All these prospects are a source of anxiety for all those who wonder what will become of them in a world where man could lose control of his own destiny.

These exciting prospects, as well as these legitimate concerns, restore the role of politics, which has a heavy responsibility to anticipate and prevent the risks that these technologies could present – in a word, to regulate. Civilisational issues must take precedence over economic considerations, which must be conditional on them. To do this, more than ever, we need strategists who can plan for the short, medium and long term.

It has to be said that in France we have a lot of progress to make when it comes to anticipation. You only have to look at how little control we had over the Internet epic to conclude that it’s high time we made progress in this area. A summit on action for artificial intelligence will be held in Paris next February, and this is excellent news. But that is not enough.

We must do more than just talk, analyse and theorise. We urgently need to anticipate and assess future developments, in order to determine their benefits and risks. So we need to be very clear about what we want and what we don’t want. Then we need to act, and fast.

We need to plan to step up investment in innovation, particularly in robotics and AI, so that – if it’s still possible – we don’t fall behind the leading nations. Such an ambition calls for a massive recovery plan.

In these times of abysmal public debt, the very word ‘stimulus’ may seem taboo. In other countries, however, it is not. The Chinese have not hesitated to inject 150 billion dollars to boost their economy, particularly innovation.

On the other side of the Atlantic, during the American election campaign, the debates barely touched on the problem of debt and deficits, because in the eyes of the candidates they were not essential in view of the technological challenges.

In Japan, the debt, which is expected to exceed 260% of GDP in the coming years, does not worry observers all that much, because over 90% of it is held by the Bank of Japan and Japanese citizens, and is massively devoted to innovation. In our country, 53.2%[2] of the debt is held by foreign investors, so we don’t have full control over it. More than its extent, it is its nature that poses a problem, and what we do with all the money we borrow.

For a recovery plan to work, we need first and foremost to return to genuine forward thinking and planning, as we have known them in the past, and to restore the lustre of the Haut Commissariat au Plan. ‘Countries that think about their future are ahead of the game. A nation has the right to know where it is going’, François Mitterrand rightly said.

We must then choose the sectors on which the recovery will focus, and not run the risk of giving money to companies that will relocate and practise tax optimisation, as we saw with the tax credit for competitiveness and employment (CICE), a large part of which went to shareholders, or with the reduction in wealth tax (ISF).

We also need to be able to draw on the right human skills, given that, according to a study published by Dell and the Institute for the Future, 85% of the jobs of 2030 do not yet exist. As public and private sector organisations are all called upon to adapt to the age of AI, they must rapidly map out their short- and medium-term needs and recruit the essential skills.

This means recruiting prompt engineers, AI UX designers and ethical AI managers, all of whom were non-existent two or three years ago. There is an urgent need here too, given the lack of interest among young French people in scientific disciplines.

All these efforts and all this promotion of excellence will be in vain if we are not vigilant about the cybersecurity of AI. In pointing out that AI tools remain components of the information system, the CEO of Thalès, Philippe Caine[3], rightly draws our attention to the risks of extracting confidential information or manipulating models. Here too, there is no room for error.

Will France be able to make the journey to excellence on its own, when so much ground has already been lost? It is doubtful. Together, European businesses and governments need to equip themselves with considerable computing resources to train and operate large-scale AI models. As things stand, however, the continent’s most powerful supercomputer, Jupiter, is finding it very hard to be the master of the clocks, in the face of rivals such as xIA.

Only the promotion of excellence in the field of AI can strengthen Europe’s potential, and that of its members, and enable it to face up to global competition. Because it is a major (r)evolution in the history of humanity, AI (and its mastery) plays an essential role at the heart of geopolitics.

In the Minoan myth, Talos, like Achilles in the Iliad, collapses for a simple weakness in his heel that escaped his creator. It would be a shame if the Old Continent, having underestimated the stakes of AI, were to suffer a similar fate.

[1] Law number 1: a robot may not harm a human being or, by remaining passive, allow a human being to be exposed to danger;

Law number 2: a robot must obey orders given to it by a human being, unless such orders conflict with the first law;

Law number 3: a robot must protect its existence as long as this protection does not conflict with the first or second law.

[2] https://www.vie-publique.fr/en-bref/294571-dette-publique-qui-detient-la-dette-francaise

[3] Pas d’intelligence artificielle sans cybersécurité, Les Échos du 12 décembre 2024