“It is a great pleasure to be here. As indicated on the announcement, I have a double identity of politician and scientist, since I am a mathematician.

In January, during the visit of President Macron in China, I had the honour of moderating several Franco-Chinese panels on artificial intelligence.

One panel provided the opportunity for a varied audience to bring to the table all the different facets of artificial intelligence.

It was a discussion on science and technology, yes, but also a meeting point between the society and the development of experts. A meeting point between the interrogations and the great hopes on the progress and possible achievements. And then, in that instance, it was also the meeting point between France and China.

I also remember a visit to Shenzhen, in the big incubator, led by Cyril Ebersweiler, which was also among the examples of points of contact between France and China on the theme of technology and progress.

Artificial intelligence is a singular topic. Everyone remembers, I think, his first contact with the field. As far as I am concerned, it was when I was an adolescent and the topic was well featured in certain works I was reading with the passion of the young, somewhat geeky adolescent who wants to keep abreast of scientific progress. In this case, the works of Douglas Hofstadter.

At that time, artificial intelligence was a topic that demanded the understanding of cognitive phenomena, logic and reasoning. How surprised were the experts of the field some 30 years later, in the situation where we are currently, where the keywords that imply artificial intelligence have changed – be it as a science or as a social phenomenon.

The keywords which were the understanding and reasoning have moved to the background. Why were we talking about artificial intelligence? Because the plan was to reproduce the biological mechanism of reasoning. And then that objective was largely abandoned in favour of the objective of carrying out complex tasks.

And where, not so long ago, people were hoping that all-new algorithms would allow the obtainment of intelligent solutions, now, today, the debate is dominated by algorithms which have proved themselves and, which in certain cases, are more than thirty years old.

Not so long ago, artificial intelligence was a topic on which one distributed a few euros, to ascertain that researchers had something to keep themselves busy. Now it is the field in which all governments who intend to matter in the domain, ask themselves how many billions they will be able to mobilise to advance the subject.

What has happened in the meantime? The techniques that one thought to be futureless – like neural networks – have started to have the upper hand. The very big databases could be used in effective ways with statistical methods. The extraordinary augmentation of the power of computers and the means of calculation has allowed the processing of all these data.

And then, something happened, that no one has understood to this day, and which shows that the theory is not yet up to the level it needs to be: it is not known why the algorithms function more effectively than predicted. It is beyond theorists, and there is still a lot of speculation on this topic.

Let’s be clear. It is not because the algorithms have acquired their own intelligence. It is just because certain of the statistical conclusions in the structure of our world, of our data, had not been anticipated at all and are still not identified. There is a domain therefore which has built itself in an extremely experimental and pragmatic way, with a lot of research; and it will continue to be one of the keys in the subject.

In the course of this mission which I have led for the government and which has been the opportunity, during six months, to meet hundreds of actors and to travel almost everywhere in the world – around ten trips – I have appreciated to what extent artificial intelligence has developed in science and technology, but also with everything else. And, most importantly, how much it is a set of technologies that is changing the game – but without trampling upon what had been done previously.

There are revolutions such as the digital revolution, which have crushed certain niches. For example the email has replaced postal distribution, without postmen having any knowledge to transmit to email specialists. On the other hand, for artificial intelligence to bring about a revolution in industry, in health, industrialists and health specialists need to transmit a lot of their knowledge to artificial intelligence experts.

And the economic stories, or the stories in terms of development, which work well almost always say the same thing: we need experts in the development of sophisticated algorithms coming together to work hand in hand with experts on the topic which needs to be improved. The topic could be industry, health, transport or environmental questions, or defence, to mention some of the themes that are the most trending on the subject. This way of organising the processes, the cooperation, the work mode, will be one of the keys of the transition, of the upcoming revolution, which is already underway.

The government, who has followed our recommendations, predicts that there will be five [industries]. And on calls for tenders, they will be established with the responsibility to train many more students to be much more in sync with the various industries and to work in a completely interdisciplinary manner.

Beyond research, there are the other industrial questions I have spoken about, and there is a key aspect that immediately crops up as soon as we speak of artificial intelligence. It is the ethical question; it is the question of the goal of artificial intelligence. These questions are coming particularly in the European context, from a continent with needs to be reassured with regulations of protection of private life, with also a lot of questioning on the uses of artificial intelligence.

The strongest questionings have been coming from the United States with certain authors who have been very convincing in explaining how such and such practice has to be addressed with transparency and respect for values of reproduction. All of these questions will also be at the heart of the French and European strategy. With, in particular, in the French framework, the setting up of an independent ethics committee to judge – either under citizens’ direction or under the French government’s direction – the ethical application of such and such technology.

I will end by saying a few words on the French-China cooperation, which will be at the heart of the panel discussion in a few minutes.

Regarding international cooperation for France and wider European cooperation, it is essential to have a consequential economic market size as well as to have an important strike force in research, in addition to an environment that is as attractive as possible for our students.

Therein, France is destined to play a leading role, has the ambition and brings to the table, in particular, its assets: very developed research. [Well known is] the French mathematical competence, whose performances in terms of mathematical research are unanimously recognised as coming right behind that of the United States and, in terms of proportion, easily the first in the world.

And then, in terms of research in artificial intelligence, France is, here as well, among all of the first nations. One of our Chinese speakers in the panel discussion [I participated in earlier this year while] in Beijing pointed it out very strongly. When one looks at the superstars on the matter, one finds people like the French Yann Le Cun, the Quebecois francophone Yoshua Bengio who was born in Paris. And many other names which were in the French ecosystem.

On the Chinese side, as we know, the competencies are also considerable. And we have all been extremely impressed, at the global level, to see the rise of Chinese scientific performances in all the fields, and in this one in particular. And I refer not only to the quantity, but also the quality in constant growth, with more and more contribution in international seminars.

And then, of course, as we know, China succeeded in what nobody thought possible: the emergence of giants of size comparable to those of American giants. Leaning on the economic boom and the technophile enthusiasm of the population, among other pillars. Of course there will be great opportunities for collaboration between China and France; cooperation that will be all the more desirable as the world is in a phase where it is stretching – politically and in all fields.

On the question of artificial intelligence, a lot of media are setting the scene for a great economic and scientific war which would play out in the future between, on one side, the United States, on the other, China. That was the cover of The Economist not so long ago.

I think that no one in this room wishes to see the creation of a situation in which two camps would fight a merciless battle. I am in favour of the emergence of a world in which several actors hold the power and are capable of being in competition and in intelligent cooperation, with a shared vision to reproduce.

Thank you.”