Artificial intelligence has recently benefited from decisive progress in the power of processors and servers and in the expertise of growing masses of data, with new algorithms and increasing investments in research.

The government has entrusted Cédric Villani with a mission on artificial intelligence. The mission's objective is to study the actions needed to enable France and Europe to be at the forefront of the AI economy, to describe international best practices in the application of these technologies to transform and improve public policies, to identify priority applications to be deployed within the public sphere, and to open the field to a national reflection on the impacts of AI, considering its effects on work and the ethical issues it raises.

At the same time, the Minister of Labor and the Secretary of State for Digital asked France Stratégie toanticipate the impacts of artificial intelligence on the labor market.

These two missions come after the publication of the France IA report of March 2017 and the report of the Parliamentary Office for the Evaluation of Scientific and Technical Choices (OPECST)

In a report published last March, the OPECTS was concerned about the catastrophic representations of AI in the media and popular culture and their impact on public opinion. Listing a series of polls that had highlighted public concerns about AI, the OPECST report deemed it essential to "demystify AI". "In conclusion, the report proposed to "train and raise awareness of the general public to artificial intelligence through communication campaigns, the organization of an international exhibition on artificial intelligence and robotics, and the broadcasting of educational television programs" (proposal n° 13), to " train and raise awareness of the general public to the practical consequences of artificial intelligence and robotization " (proposal n° 13).

Polarization of opinions in the face of the development of artificial intelligence

Since the same question was asked by Odoxa in identical terms in 2016, March 2017 and June 2017, there is some stability in the responses.

However, when we examine the socio-demographic or socio-professional profile of the respondents, we get the impression that the answer to this question is largely determined by the relationship that people have with digital technologies, the intensity of their digital practices or the perception that people have of their "digital skills", which vary according to age, income level, professional situation or level of education.

The feeling of fear increases with age

While the feeling of fear increases almost mechanically with age, there is a relatively moderate gap (13 points) between the responses of the youngest (18-24 years) and the oldest (65 years and over).

 Confidence in Artificial Intelligence grows with education and income levels

The level of fear or confidence also appears to be indexed to the various forms of social stratification (CSP, income and qualifications).

The gap in responses is 19 points between CSP + and CSP - and 18 points between the lowest and highest incomes,

It is 17% between the least and the most qualified (Note that Odoxa uses a classification of diplomas in 3 categories (diploma below or above the baccalaureate: the differences would probably be more marked if Odoxa had used a more refined categorization).

AI worries women more than men

The feeling of fear towards AI is more pronounced among women than among men. The gap in responses between men and women has even widened in one year, from 12% to 17%.

This gap may be due to the fact that women perceive themselves (rightly or wrongly) as less competent in their relationship with digital technology than men. Logistic regressions would be valuable here in order to neutralize age effects: as women live longer than men, the differences observed are probably attributable in part to this age factor or to other factors.

Reticence towards AI is fairly widespread in Europe

The question asked was: " In general, do you have a very positive, rather positive, rather negative or very negative image of robots and artificial intelligence? The gradation of the proposed answers in four levels (Very positive, Rather positive, Rather negative, Very negative) leads to less clear-cut judgments.

The average of positive judgments (very positive and rather positive) is 61% for the 28 EU member countries: France is slightly behind (55% positive judgments).

There are significant differences among the 28 member countries between those who welcome AI most positively and those who are most reserved. The 28 member countries can be classified into four groups:

• The most positive towards AI (less than 20% of negative judgments): we find the Scandinavian countries.
• The moderately positive (between 20% and 30% of negative judgments): this includes Poland, the United Kingdom and Italy.
• The reserved: (between 30% and 40% of negative judgments): we find, in particular, Germany, Austria, Spain, Belgium, Hungary and France.
• The most reserved: (more than 40% and more negative judgments): Greece and Luxembourg are among them.

88% of Europeans believe that " robots and artificial intelligence are technologies that need to be managed with caution": 93% in France.

44% of Europeans think that "their current job could be done by a robot or an artificial intelligence in the future": they are only 37% in France, 38% in Germany, 39% in the Netherlands.

Références : 1 2 3 4 5

Launch of a public consultation on artificial intelligence by Cédric Villani

• Build a data policy adapted to the challenges of AI
• An ecosystem for agile and disseminating research
• Establishing an ethical and trustworthy framework for AI development
• Anticipating and controlling the impacts of AI on work and employment
• AI for a sustainable and ecological economy
• Stimulate a strong dynamic and catalyze opportunities in key sectors
• Establish a general breeding ground for the development of AI