These last years, we have talked a lot about the adoption of artificial intelligence in businesses. Tester Claude, deploy Copilot, train the teams, identify use cases… a necessary first step, but now insufficient. From now on, the issue is less the adoption of AI (although many organizations have not yet deployed tools or trained their employees) what it concretely changes in our way of working.
It is entirely possible to integrate new tools while maintaining exactly the same habits, the same validation circuits, the same meetings, or the same repetitive tasks. In other words, we can adopt AI without changing much. It can even be used to speed up processes that would first be worth questioning.. Automating a useless process doesn't make it smarter. This is when we really begin to transform the ways of working : when we stop looking only at the tool and become interested in the work itself. Which tasks still have value ? Which ones could disappear ? Which could be reduced ? Which could be improved using AI ? And which ones do we want to continue to entrust to humans ? The DRIP matrix allows us to pose this reflection in a simple way. It is based on four verbs : Delete, Reduce, Improve, Preserve.
Delete
It’s about removing what no longer needs to be. Before trying to automate a task, you have to ask yourself why it exists. Some re-entries, manual consolidations, repetitive searches or systematic production of documents are sometimes habits more than real needs. So the first question shouldn't always be "How can I automate this task?" ? » but rather “Is it still necessary ? »
Reduce
It means reducing the time spent on what remains useful. Prepare for a meeting, synthesize several documents, structure an initial analysis, comparing information or producing a first version of a support are tasks which retain their interest but which no longer necessarily require the same human time as before. Saving time is only useful if it allows you to devote that time to something else. : analyser, to exchange, decide, to accompany.
Improve
Improve what you already do : this is probably where AI is most interesting, when it no longer only serves to go faster. It can be used to test several scenarios, to confront an idea, to look for counter-arguments, to identify blind spots, or to further personalize a communication. Provided, of course, not to delegate our critical thinking to him.
Preserve
Finally it is a matter of deciding what must remain human. The “Preserve” is perhaps the most important part. Not everything that can be entrusted to an AI must be entrusted to an AI. A difficult conversation with a colleague, feedback, a complex arbitration, understanding a human situation or responsibility for a decision does not depend on the quality of a response produced by a tool. And this is where the role of the manager really changes.
Humain is I : who does what ?
As AI takes over some of the execution, the role of the manager is refocused on something else : decide, arbitrate, control. It is not his vocation to become the one who masters the greatest number of tools or who writes the best prompts.. Its role is increasingly to determine what can be entrusted to AI, what needs to be achieved with its help, and what must remain profoundly human. This also requires setting clear rules, to check the results produced, to develop critical thinking among teams, and to avoid two opposite excesses : refuse AI on principle, or want to introduce it everywhere.
We looked at the discovery phase. We must now move on to that of transformation. For leaders and managers alike, the right question is therefore no longer “How to use AI more ? » but rather “What does AI really need to change in the way we work? ? »
