2023-11-22 06:28:18
That artificial intelligence (AI) is not a threat to employment but an asset for social progress. This is the wish of the CGT which is organizing, this Wednesday, November 22, a first major conference on the issue at its premises in Montreuil, in Seine-Saint-Denis. Its title, « IA, menace your perspective? », sums up well the ambivalence of the subject, which the union world is only just beginning to grasp. A full house (300 participants expected) will attend the discussions, which will be concluded by Sophie Binet, general secretary of the CGT.
On December 19 and 20, it will be the turn of Solidaires to hold its “AI days”. Objective : “refine” a collective position in the face of “technical discourse” of employers. Activists will be made aware of these issues by researchers (sociologists, lawyers, etc.) but also by members of Solidaires who have already started to fight on the ground against an implementation considered too rapid of these technologies: postal workers, authors, journalists… and even tax inspectors.
Detect fraud risks
Benjamin Gandouin and Damien Robinet, national secretaries of Solidaires-public finances, will lead three workshops on December 19 and 20. The General Directorate of Public Finances acts as a pioneer within the State regarding the use of AI, they say, while deploring that staff are not “never associated with reflection”. “However, these changes have a huge impact on their work! »
An example: “data mining”, this technology which promises to predict the risks of tax fraud based on the history of past fraud. “Agents no longer have to identify the fraud risk criteria themselves, at the end of an investigation: the AI does it for them, and they check whether it is consistent or not. This enormously reduces their room for initiative, and the very scope of their action. The task becomes repetitive and time-consuming – because there are often errors. »
In front of “total opacity” of their management’s choices concerning the service providers requested, Benjamin Gandouin and Damien Robinet try to force their hierarchy to provide them with a certain number of documents: reports, impact studies, etc. At the same time, they launch into a communication that they want « offensive » to defuse the reassuring speeches according to which AI would not be “only a tool at the service of humans”.
“Let’s do some experiments!” »
These concerns are shared by certain private sector executives. “There is nothing trivial about the transformations brought about at work, and yet we never discuss them: there is no social dialogue on these issues,” regrets Franca Salis-Madinier. This Orange employee and member of CFDT Cadres has been interested in this subject since 2017, to which she devoted a book (1).
Like the other trade unionists interviewed by The crossshe refuses to make AI a ” demon ” but wants to anticipate the deleterious effects of this technology. “Under the pretext that the subject is perceived as technical, should it be hidden? Let’s do experiments, tests, get feedback on what works and doesn’t work! » Among Franca Salis-Madinier’s points of vigilance: inequalities between men and women“because women do jobs that are more easily automated.”
A first large-scale social plan
Will AI destroy many jobs? “For the moment, according to the International Labor Organization (ILO), there is no notable decrease,” advances Franca Salis-Madinier. At the CGT-numerique, Noël Lechat, one of the organizers of Wednesday’s conference, is cautious. “The studies are contradictory: we cannot decide. What we are sure of is that jobs will change. This is what must be supported. »
For the moment, in France, only one large-scale social plan has been justified by recent advances in AI. The French subsidiary of the Onclusive group, specializing in media monitoring, announced in September that it wanted to eliminate 217 positions out of 383 by June 2024. Faced with pressure from the unions, management, however, announced in mid-October a postponement of the plan. A new version should be presented to them in December. But internally, we are talking about a « sursis » more than one « victoire ».
———-
New AI professions
Specialists agree that artificial intelligence will create jobs – in proportions that are still difficult to quantify – but that it will be poorly paid and extremely fragmented work. Among these new AI professions:
– « prompt engineer » or « query engineer »: he writes the instructions for these systems to provide the desired response.
– developer specializing in AI : it administers and trains these tools to adapt them to a specific industry or company. Because ChatGPT is too general.
#unions #demand #real #social #dialogue