Acculturation to major language models has several challenges. A large number of employees must be massively training while taking into account the company’s AI culture.
The enthusiasm around the generative rebates the job market cards. While Chatgpt d’Openai, Gemini de Google or Claude Anthropic make it possible to increase employees and to unload them from the rebellious and time -consuming tasks, the mastery of these large models of language (LLM) becomes an issue of employability. Becoming an expert in the art of prompt makes it possible to accelerate a career or, failing that, to remain in the race in terms of skills expected by the market.
According to a study by the Robert Half recruitment firm, a majority (53%) of employees say they are ready to undergo training “to evolve towards a new role if their work became partially automated or that their employer asked them to change their role and acquire new skills”. Otherwise, 17% of them say they are ready to change their employer. Thus, proposing training to Gen AI tools not only allows companies to accelerate their adoption but also to make it a lever for attractiveness and retention of talents.
A multitude of free educational resources
However, in another study published by LinkedIn, a year ago, only 38% of French companies set up a support program for their employees to help them familiarize themselves with AI. A figure that questions. If the training was sorely lacking during the thunderous launch of Chatgpt, at the end of 2022, there is today a plethora of generic training online and mostly free. These educational resources, of the MOOCs type, describe the general principles of generative AI and demonstrate some use cases.
Linkedin Learning, the eponymous professional social network training platform, has posted some 250 free courses. Institutions France France Digital University and Bpifrance University or Digital Learning specialists such as Openclassrooms and Coursera offer complete training courses. The American digital giants are not to be outdone that it is AWS (Skills to Jobs Tech Alliance), Google (digital workshops) or IBM (AI Academy). Meta has associated with Simplon to organize face -to -face workshops.
Easily accessible, these educational content has the advantage of acculturing all employees on the scale. Based essentially non -technical populations, it will be a question of selecting training which popularizes sometimes complex concepts but in clear language, by multiplying concrete examples. We will favor short and modular formats, of a micro-learning type, which are easier in overloaded agendas.
Adopt good practices, develop critical thinking
These generic training courses only describe the general principles of generative AI. They do not take into account the maturity of an organization in this area. The CEGOS Institute offers a self-diagnostic tool, called AI Profiler, which classifies organizations into four categories (explore, Learner, adopt and transform) then, according to the reached level, suggests priority training actions.
“Standard contents do not take into account the specific IA culture of the company, nor its regulation and ethical policy”, adds Grégory Gallic, head of offer and expertise at CEGOS and AI training expert. “However, the responsible integration of AI is based on a clear framework and support adapted to the challenges of each organization.”
To reduce the risks of bias, hallucinations, leakage of confidential data and damage to intellectual property, the training course must integrate authorized uses and good practices in accordance with the principles of the code of conduct and the ethical charter specific to each company.
Developing critical mind among employees is also crucial so that they do not take the results of generative AI models for cash. Another Bottom-up type movement this time consists, according to Grégory Gallic, “to enhance the field initiatives by inciting teams to test and explore AI in a supervised manner, in order to promote controlled and sustainable adoption”.
Tailor -made training and workshops
Beyond the training courses such as “3h chrono to raise awareness of artificial intelligence”, Cegos thus offers tailor-made sessions which integrate the risk aspect as well as workshops allowing the employee to get their hands dirty. He is experimenting in pairs with a CEGOS trainer or an advanced user, who takes the name, according to organizations, of IA champion or IA ambassador.
“This mentoring plays a key role in the progressive dissemination of AI, based on internal relays capable of supporting their peers in the appropriation of tools”, judges Grégory Gallic. “Experimenting with a colleague is more powerful than with a trainer”.
After acculturation, we enter a second step in that of professionalization. It is a question of training in the contributions of AI on the specific profession of the employee that the latter evolves in the field of HRDs, marketing, communication or project management. “It is an essential transition to move from a simple awareness of a concrete and effective integration of AI in business processes,” continues Grégory Gallic.
Cegos also offers a one -day Bootcamp AI for members of the management committees. After being made aware via an inventory of the generative AI, managers participate in thematic workshops on its contributions to marketing, sales or HR. At the end of the day, they pit on what they have retained and evoke the field of possibilities that the generative AI allows, according to them. “Management must be an engine of change, because its commitment facilitates the adoption of AI at all levels of the organization,” insists Grégory Gallic.
Form “just in time”
More than for other technologies, AI requires, moreover, a “just in time” training, that is to say provided when the employee must take advantage of the newly acquired knowledge, whether it is to write a blog post, create or retouch a visual, personalize a marketing campaign or prepare a commercial interview.
Mendo is positioned on this niche. After proposing an ADD-in for Excel, this French start-up designed an extension that natively integrated into Chatgpt or Copilot from Microsoft. This learning tool will teach prompting art in real situations through small training capsules and practical exercises.
“Personalized, the learning path differs according to the profession exercised”, indicates Quentin Amaudry, CEO of Mendo. “It is progressive starting with a case of basic use, as summarizing a text, to gain power and go as far as the data analysis. The program aims to discover the potential of the generative AI but also its limits by developing a critical mind in the face of the results obtained.”
Mendo, which has major accounts such as CNP Assurances, Bred Banque Populaire, PwC, EY, GRDF or Rocher Group among its references, intends to accelerate the adoption of generative AI and therefore associated productivity gains. “As with Excel, master Chatgpt or Copilot requires following a learning curve,” observes Quentin Amaudry. The statistical data reassembled by Mendo highlight the departments most in advance and, by gamification mechanisms, promote emulation between services. An interesting element to promote a progressive rise.
