a major technological revolution for the French but which needs to be regulated, according to a survey

According to the 2024 Barometer of critical thinking carried out by OpinionWay for Universcience, a majority of French people favor artificial intelligence which still raises ethical and legal questions.

Published


Update


Reading time: 3 min

For a majority of French people, artificial intelligence is a major technological revolution but one that needs to be supervised and regulated, according to the 2024 Critical Thinking Barometer.  (illustrative photo) (JEAN-MARC LOOS / MAXPPP)

Artificial intelligence (AI) applications are a major technological revolution for humanity, for six out of ten French people (61%), in the same way as the invention of printing was, for example. This is what emerges from the 2024 Critical Thinking Barometer produced by OpinionWay for Universcience published Thursday March 21.

But for more than eight out of ten French people (85%), it is necessary to supervise and regulate their use, because for three quarters of those questioned (77%), they raise new ethical and legal questions in terms of responsibility and rights.

Regarding the possibilities offered by these technologies, 68% of respondents believe that AI applications cannot make autonomous decisions, since they follow defined computer programs. 59% of French people believe that they are not creative and innovative, because they assemble already existing content. According to this survey, only 19% of respondents believe them to be perfectly reliable. As for whether AIs are more neutral than humans, the panel studied is more divided. 42% answer that yes, they are more neutral, while 38% believe that they are not.

Risks and benefits of AI depend on application areas

The survey also questions the vision that the French have on the advantages and risks linked to AI. And they remain quite divided, generally believing that there are as many risks as advantages to using these applications. In detail, it is the perception of risk that dominates the field of human relations. 37% of French people have this feeling, while 31% believe that there are as many risks as advantages to using AI in human relations. 10% see it mainly as advantages. The risk of AI is also predominant for information (33%) and employment (31%), although 34% of respondents also see as many risks as benefits for these two areas.

The French perception is also very nuanced on the contribution of AI to health, research, science and transport. Applications have as many risks as benefits in health (35%), science and scientific research (36%) and transportation (33%). But it is in these three areas that respondents see the most advantages. 29% perceive this advantage for health, 28% for science and scientific research, and 26% for transport.

The survey also reveals that the French mainly rely on artificial intelligence to detect errors in texts or computer programs, for 75% of respondents, analyze large quantities of data (73%), find information ( 70%), deepen their knowledge (65%), or even write a text (64%). Conversely, they are more suspicious when the question of accountability comes into play. Only a quarter of those surveyed (25%) trust AI applications to render legal decisions. 28% assure that AI can fly a plane with confidence, or drive a car (34%). And 33% have confidence in AI to prescribe medications, or to carry out a medical diagnosis (41%).

Scientists, museums and specialist journalists deemed reliable for evaluating AI

Finally, to assess the risks and benefits of AI, the French have above all confidence in scientists, museums and specialized journalists. The people interviewed for the barometer primarily favor scientists and researchers in general (68%), museums and centers of scientific culture (59%) and science journalists (56%). But respondents remain very suspicious of companies. Only 34% trust them and even 29% for digital companies like Gafam. Distrust of elected officials or political representatives is also strong. Only 22% of French people trust them to assess the risks and benefits of AI. The trust rating of influencers and public figures is also low, at 17%.

________________________________________

The 2024 Critical Thinking Barometer produced by OpinionWay for Universcience was carried out by an online questionnaire from January 31 to February 12, on a sample of 2,004 people, representative of the French population aged 18 and over.


source site-15