The main thing to remember:
The key figures:
- ROI multiplied by 3.7 for every euro invested in generative AI
- Doubling the number of users Co-pilot every quarter
- 90% of employees Save time on a daily basis
- 83% of employees report increased job satisfaction
The major trends
- Change management is more of an issue than the technology itself
- The emergence of a single entry point for all HR and productivity tools is essential to the adoption of this tool
- An inclusive approach to skills development on prompting is gaining momentum in HR development plans
- Concrete benefits are already being observed: diversification of profiles and early detection of psychosocial risks
Introduction:
Generative AI is profoundly transforming the way people work and the employee experience. To understand the challenges and opportunities of this revolution, BFM TV and Frédéric Simontel brought together three recognized experts: Jean-Christophe Dupuy (Modern Work Director, Microsoft France), Eugenia Fornieri-Macchi (Senior VP Global Talent & Learning, Danone) and Paul Courtaud (CEO, Neobrain). Their feedback sheds light on best practices and the prospects for this major transformation.
The interview:
State of play and evolutions of generative AI
Continuously growing adoption
Frédéric Simottel: Two years after the arrival of ChatGPT and one year after the launch of Copilot, where are we with the adoption of generative AI in business?
Jean-Christophe Dupuy:
The numbers are very encouraging. An IDC study last November shows that we have gone from 55% to 75% of use of generative AI at work between 2023 and 2024. Two major objectives are emerging: first Productivity gains individual employee, but more and more, we are seeing a desire to transform business processes to increase turnover.
The return on investment is particularly significant: for one euro invested in generative AI, the return is 3.7 times, or even ten times for the most advanced organizations.
Our World Trade Index study out of 31,000 employees in 31 countries shows that 90% save time, 70% are happier at work, and 79% see opportunities for growth thanks to the new skills acquired in generative AI.
Technological innovations and data analysis
Frédéric Simottel: What are the major developments at Copilot after one year of existence?
Jean-Christophe Dupuy:
The evolution is considerable. The integration of GPT-4 made it possible to Multiply the relevance of the answers by three and to significantly improve the processing speed. A major advance concerns data analysis: while previously we could only process numerical structured data in Excel, we can now analyze unstructured and non-numerical data.
Concretely, imagine a company that receives hundreds or thousands of verbatims in its satisfaction surveys or its pulse surveys. Copilot can now analyze this qualitative feedback, identify trends, main concerns, and even detecting weak signals that could escape traditional human analysis.
Feedback: deployment at Danone
Frédéric Simottel: Eugenia, how did this adoption happen at Danone?
Eugenia Fornieri-Macchi :
We see AI as a real opportunity for the company and our 100,000 employees. Our approach has been structured around three main axes:
- First, we created An AI committee with representatives from each function and geographic area to anchor the AI strategy in our business strategy.
- Then we launched the Danone Microsoft AI Academy to train all employees. We started with leaders and their teams with learning expeditions to help them seize strategic opportunities.
- Then we organized an event 'AI for All' simultaneously on all our sites, with podcasts in 15 languages and Microsoft experts on site. The result was impressive: 10,000 licenses distributed in 48 hours.
Maintaining momentum and continuing education
Frédéric Simottel: How do you maintain this dynamic on a daily basis?
Eugenia Fornieri-Macchi:
Success is a daily occurrence. We have set up AI champions in each function who train their colleagues with Micro-learnings, like our 'Copilot Coffee'of 30 minutes. Our conviction is that it is necessary to anchor the use in daily life so that employees can see it the concrete benefits and make their jobs evolve thanks to AI.
A lever for adapting the transformations experienced by companies
The Neobrain vision: transformation of jobs and mobility
Frédéric Simottel: Paul, what is Neobrain's vision on this transformation?
Paul Courtaud (Neobrain):
I co-founded Neobrain based on an observation: jobs are changing more and more quickly, and companies often respond to these changes by cutting on the one hand and recruiting on the other. Our approach to AI Aims to create bridges between jobs that are evolving, using AI to identify transition opportunities.
Our collaboration with PWC in the United States and Mexico are a perfect example of this approach. Initially, when our AI proposed candidates for missions, managers selected only 2% of atypical profiles.
By combining our matching AI with Microsoft's generative AI to explain the recommendations, we went to 15% of selected atypical profiles. It's a concrete example of how AI can promote diversity and inclusion.
Impacts on the employee experience
Frédéric Simottel: How does generative AI concretely improve the employee experience?
Paul Courtaud:
The impact is multiple.
- First, it renders The most active employee in his career. Faced with an HR, he can now say “I have imagined three possible paths, here are my skills, what do you think?”
- Second, it allowsexploiting data that was previously underused, such as annual interviews. In a company with 100,000 employees, we can now detect weak signals, identify abnormal managerial situations, and act preventively.
- Finally, a crucial point is toavoid the multiplication of tools. That's why we integrated Neobrain directly into Microsoft Copilot. When an employee questions his evolution, the system first offers him internal opportunities before looking outside.
2025 Perspectives of Generative AI
The next challenges ahead
Frédéric Simottel: What are the major challenges for a successful adoption?
Jean-Christophe Dupuy:
The main challenge is not technological but organizational. We see that 53% of our customers understand what they can do with these solutions, but only 43% think they have the right skills internally to take full advantage of it. The question of prompting is central: it is necessary to develop a true science of the relevant prompt to obtain the best results.
To facilitate adoption, we let's develop functionalities that automate certain results without requiring a complex prompt. For example, automatic email prioritization or instant data analysis. The arrival of autonomous agents will also allow you to automate certain tasks according to predefined triggers.
The next steps in behavior change
Frédéric Simottel: What are the prospects for 2025?
Eugenia Fornieri-Macchi:
Our priority is twofold: make our systems evolve in line with the evolution of technologies and personalize support by profession. A marketing director will not have the same needs as a purchasing manager or an HR manager. The aim is to encourage a culture of continuous learning.
Paul Courtaud:
The challenge is to continue democratizing access to AI while maintaining theThe transparency of algorithmic decisions. Integration with existing tools and the ability to explain The recommendations of the AI will be essential to maintain the trust we have earned from users.
Jean-Christophe Dupuy:
We continue to enrich the solution with new features. The focus will be on intelligent automation and the personalization of experiences. Autonomous agents represent a new generation of applications in the world of AI that will profoundly transform the way we work.
The Generative AI FAQ
What is the average ROI of generative AI?
For each euro invested, the return is on average 3.7, which can reach 10 for the most mature organizations.
How to effectively train employees?
The key lies in short and regular training courses, customized by profession, with a network of internal champions.
What are the first use cases of Generative AI to focus on?
The most impacting generative AI use cases concern:
- Data analysis and synthesis
- The preparation and follow-up of meetings
- The identification of internal mobility opportunities
