Generative AI: getting started with confidence

For generative artificial intelligence, 2023 was the year of recognition. With ChatGPT leading the way, individuals and businesses have become aware of AI tools' vast potential. While some rightly express reservations, this technology is set to become indispensable in business transformation and development.

The importance of taking the lead

A study by Gartner highlighted a glaring deficiency: only 4% of the surveyed decision-makers consider themselves sufficiently mature in generative AI. They struggle to move beyond the "zero stage" of the ChatGPT test, which involves allowing users to test the tool without a defined objective or use case. However, 70% of them believe that this technology offers more advantages than risks. And rightly so: from improving operational efficiency and productivity to personalizing services, stimulating innovation and achieving economic gains, the benefits of generative AI are numerous and serve as a powerful catalyst for development and competitive advantage. It is essential to seize its potential today.

AI, AIs

Artificial intelligence: any tool used by a machine to replicate human behaviors, such as reasoning, planning and creativity (Source: European Parliament)

Generative artificial intelligence: a category of AI that aims to produce new data resembling that created by humans, whether in the form of text, images or music (Source: Bpifrance)

"Trustworthy" artificial intelligence: must exhibit the following three characteristics, which should be upheld throughout the system's lifecycle:

  • "Lawful," ensuring compliance with applicable laws and regulations
  • "Ethical," adhering to ethical principles and values
  • "Robust," both technically and socially, as, even with good intentions, AI systems can unintentionally cause harm

While each of these characteristics is necessary in themselves, they are not sufficient to achieve trustworthy AI. Ideally, these three characteristics should work in harmony and overlap. If tensions arise between these characteristics in practice, society should strive to address them. (Source: European Commission)

Getting on board: Yes, but how?

Implementing generative AI within a company requires a methodical approach. Firstly, it is important to familiarize employees with this technology to demystify what may seem complex and enigmatic. Training sessions, conferences and interactive workshops are a good starting point for teams. The second step is ideation, through workshops that identify and prioritize use cases. Only after these two essential phases can we move into "project mode" to establish the most suitable methodology for delivering tools quickly and deploying them through a change management plan to support employees, including training programs. By following this process, we can avoid pitfalls such as collecting inadequate data or introducing cognitive biases, which could turn a virtuous instrument into a counterproductive tool.

The need for a single entry point

Managing such a process is complex and poses a significant challenge for any company. This is where we come in, by offering our new one-stop shop that brings together all of the Orange Business expertise in artificial intelligence. With 4,000 experts dedicated to providing end-to-end support, we assist businesses of all sizes with trusted technologies, both in France and beyond. Supported by Orange Cyberdefense, our solutions ensure secure use of generative AI and guide towards ethical use cases that aim for a positive impact on companies, their employees, customers and society.

Michaël Deheneffe
Michaël Deheneffe

As VP Data & AI and Director of Marketing, Innovation, and Strategy for Digital Services at Orange Business, Michaël leads development projects within the group's ESN activities and is an evangelist for data, AI and generative AI.