Training your contributors in AI: why learning to think with the tool is more important than knowing how to use it

Giving access to ChatGPT or Claude to an entire team is not enough: three weeks later, the tool is back at the bottom of the drawer. Training in AI means teaching your contributors to think with the tool rather than use it from time to time. We detail what this concretely changes, from verifying responses to the boundary between what you automate and what you keep in hand.

Lire la suite

Start with one person, grow on proof: how a collaboration with a dedicated team evolves

Many executives hesitate to entrust a function to a dedicated team, for fear of the big leap. In practice, you start with one person on a precise position, then grow on what has already proven itself, up to an entire department.

Lire la suite

An AI agent to write the first drafts of our content, we retain control over what matters

Writing a first version takes time. We prefer to dedicate this time to what matters: the message we want to convey and the accuracy of what we assert. We therefore entrusted this first step to an AI agent, while retaining control over what matters most: meaning, nuance and final validation, from idea to completion. A concrete look at this workflow: its advantages and its limits.

Lire la suite