Business still tend to associate an employee’s use of a tool with an increase in their performance or even a transformation in their working methods. They are delighted to see employees using all the technologies they make available to them and conclude that a collective dynamic is underway. AI even accentuates this phenomenon by offering very quick individual results, especially since these tools are already widely adopted in employees’ personal lives, which can give the impression of a transformation underway.
However, when we examine what is happening within teams, we quickly come back down to earth. Individual use does not transform anything until the team reorganizes the way it works. This observation is not new and has been seen in all past technological waves involving so-called productivity tools. AI does not pose a new problem but, once again, reminds us of the lessons we did not want to learn from the past.
In short:
- Local improvement is insufficient without questioning a team’s structures, objectives, and operating methods.
- Collective benefits only appear when organizational adjustments are fully implemented.
- Collective ownership is a concrete process through which a team transforms isolated practices into new shared practices.
- The value of AI depends on the team’s ability to integrate it consistently into its operations.
- Real impact always emerges from collective effort, not from isolated individual initiatives.
First level: I know how to use
Knowing how to use a tool says nothing about its impact on work and, more generally, on production activity. An employee can obtain content in a matter of moments, simplify a task, or speed up a step, but that is not enough to transform a team, given that in business, you are never alone but rather a link in a chain, a member of a team that produces something collectively with a large number of interdependencies. As long as use remains isolated, nothing will change.
We believe we are making progress because an individual is working faster, while the workflow remains the same, dependencies remain, and the team has not changed the way it works. Individual use gives the impression of progress, but it is only an impression. It is a prerequisite for the next steps, but a prerequisite that, although necessary, is not sufficient for the business to derive any benefit from it.
Second level: the team must be able to organize itself around the new practice
The impact begins to appear when the team agrees to review its operating procedures. For individual practice to transform into collective progress, the entire workflow must be reviewed, each task reevaluated, certain steps redefined, and decisions made about what needs to be streamlined or redistributed. A task performed more quickly does not create any progress unless the entire sequence is rethought. In other words, saving time on one task is meaningless if the entire workflow is not brought into line (Local optimum vs. global optimum and the theory of constraints: why your productivity gains are sometimes useless). A simplified step does not free up anything unless another step immediately absorbs what is freed up or can keep up with the pace.
This transition from the individual to the collective level requires more than superficial coordination. It is necessary to understand what each person contributes, what needs to change, what needs to remain, and what needs to evolve. The team must give itself the opportunity to stabilize what works and discard what does not produce results. This adjustment work does not depend on technology but on a team’s ability to observe itself at work and choose a new way of organizing how it operates and rethinking each person’s tasks. AI does not transform work but forces the team to decide how it wants to organize it (Second level: the team must be able to organize itself around the new practice
The impact begins to appear when the team agrees to review its operating procedures. For individual practice to transform into collective progress, the entire workflow must be reviewed, each task reevaluated, certain steps redefined, and decisions made about what needs to be streamlined or redistributed. A task performed more quickly does not create any progress unless the entire sequence is rethought. In other words, saving time on one task is meaningless if the entire workflow is not brought into line (Local optimum vs. global optimum and the theory of constraints: why your productivity gains are sometimes useless). A simplified step does not free up anything unless another step immediately absorbs what is freed up or can keep up with the pace.
This transition from the individual to the collective level requires more than superficial coordination. It is necessary to understand what each person contributes, what needs to change, what needs to remain, and what needs to evolve. The team must give itself the opportunity to stabilize what works and discard what does not produce results. This adjustment work does not depend on technology but on a team’s ability to observe itself at work and choose a new way of organizing how it operates and rethinking each person’s tasks. AI does not transform work but forces the team to decide how it wants to organize it (Business design before architecture: putting the business back on track).
Third level: the activity must be reorganized
After the team, we reach a third level, where the activity itself is reorganized. Here, it is no longer local tasks and operating procedures that are adjusted, but the workflow as a whole, from input to the final usable and valuable result, most often involving several cross-functional departments. We identify what is slowing down progress, review the sequence of steps, redistribute responsibilities, and redefine each person’s contribution. If a team saves time but the overall flow does not speed up, that time saving does not exist. If one step moves forward but another continues to slow down the whole process, the impact is zero.
And since AI is often compared to the invention of electricity, I like to point out that electricity alone did not create productivity gains in factories because it initially only changed the energy source, not the organization of work or the design of machines. As long as factories were content to replace steam with centralized electric motors, they kept the same workshops, the same flows, the same waiting times, and the same waste, so productivity remained virtually unchanged. Spectacular gains only came when factories were redesigned around small decentralized electric motors, new production lines, improved parts circulation, and more precise work coordination, combining technology, organization, and management rather than electricity alone.
What if you get stuck at the first few levels?
When you remain at the individual or even team level without progressing further, you virtually check in no gains in terms of activity or workflow. Those who have been driving forces in adopting the technology or who have simply made the effort to try to use it may backtrack simply out of mimicry or because they don’t feel any real impact. Worse still, not knowing what to do with the time saved, it may be misused (Without governance, the gains from AI are virtual).
Watch the team before watching the tool
Ultimately, the impact of AI depends less on individual mastery than on a team’s ability to change the way it works or on the business’s ability to redesign its work and processes. Use only creates potential, and the organization only progresses when that potential is exploited.
Transformation never comes from a single action, but from a collective effort that adapts. A tangible and lasting impact always comes from the collective.
Bottom Line
What we are seeing with AI confirms what we have seen in the past and even what industry learned long before knowledge workers: local improvement is never enough. A team only progresses when it agrees to review the structure of its work, redefine certain objectives, review responsibilities, and reorganize its operating methods. As long as these adjustments remain pending, local gains do not contribute to collective gains and add nothing to the organization, and the initial promises fade away.
Ultimately, collective ownership is not an abstract goal but describes how a team gives coherent form to what it discovers and experiences. It explains how an isolated use case becomes a new way of working and reminds us that AI only adds value if the team gives itself the means to integrate this new development. Impact always comes from the collective, never from the individual alone.
To answer your questions…
Local improvements remain limited until the team reviews its organization. Individual use of AI can produce one-off gains, but these do not translate into collective progress unless objectives, responsibilities, and methods are adjusted. Without this harmonization, efforts remain scattered and initial promises fade away. The real impact only becomes apparent when the team structures and shares its learning, transforming isolated practices into a collective dynamic that benefits the whole.
Integrating AI requires giving a common form to discoveries made through individual experimentation. The team must clarify what it is seeking to improve, reassess certain responsibilities, and adapt its operating procedures to prevent the tool from remaining confined to a few personal uses. This collective effort transforms AI into a real lever for progress rather than a mere gadget. By aligning practices around a shared framework, the team creates a solid foundation for integrating AI into its operations on a lasting basis.
Collective appropriation makes it possible to transform scattered attempts into a new way of working. It is not an abstract concept, but a prerequisite for developing team cohesion around AI. Without it, uses remain isolated and the impact remains marginal. By sharing what they learn and adjusting their practices together, members build a common framework that multiplies the effects of AI. This collective dynamic enables the organization to extract lasting value from the new developments introduced.
Relying solely on individual initiatives exposes the organization to a lack of concrete benefits. Gains remain scattered, do not add up, and do not change existing processes. Teams then progress at different rates, creating misalignments and weakening overall efficiency. In this context, the promises of AI fade away in the absence of structured integration. Ultimately, the organization risks missing out on a useful transformation, having failed to bring together the necessary collective conditions.
To transform an isolated practice into a collective standard, the team must work together to analyze what works, what needs to change, and how to integrate these lessons into their routines. This involves regularly sharing feedback and a shared commitment to adjusting objectives and methods. By creating a shared framework, the team gives consistency and depth to emerging practices. This pooling of knowledge makes it possible to move beyond individual experimentation and establish a truly renewed way of working that has a lasting impact.
Image credit: Image generated by artificial intelligence via ChatGPT (OpenAI)
Related articles:
Who is handling your artificial intelligence projects? Probably not the right people.
Adoption and impact of AI: lessons (and limitations) from the latest McKinsey and BCG studies
AI adoption: methods put to the Test
AI adoption does not replace productive appropriation
How can you prepare your organization for AI?
AI in the workplace: going beyond augmentation to actually transform
