AI adoption: which problems remain out of sight in Vevey?
AI adoption often begins with tasks whose data is accessible and whose results are easy to demonstrate. The management question is specific: does this selection leave out important problems that are less immediately suited to the tool?
What the literature establishes
James March (1991, Organization Science) distinguishes the exploitation of existing knowledge from the exploration of new possibilities. His model shows how learning can favour nearby, predictable improvements at the expense of more uncertain search. Applied to AI adoption, this framework invites a distinction between improving familiar tasks and investigating problems that are still poorly defined.
The mechanism: learning can narrow the search
Daniel Levinthal and James March (1993, Strategic Management Journal) analyse the myopias of learning, including the tendency to favour effects that are close in time and within the organisation. A team can therefore become more effective within its usual remit without gaining a better understanding of what happens elsewhere. In an AI project, this suggests a risk worth examining: cases that are easy to document could absorb attention at the expense of difficulties spanning several departments.
The premature conclusion
One might conclude that straightforward applications should be abandoned in favour of exploratory projects. These studies do not justify that reversal: exploitation remains necessary, while exploration involves costs and uncertain outcomes. The issue is instead to check that the ease of an initial application does not become the sole criterion for selecting subsequent ones. (our executive and employee training programmes)
What these studies do not demonstrate
These articles do not concern generative AI: the first uses modelling, while the second develops a theoretical analysis of organisational learning. They therefore establish neither that AI reduces exploration nor an optimal proportion of exploratory projects. A list of AI projects is also insufficient to identify blind spots: it must be compared with an inventory of problems compiled independently of the tool.
A practical check in Vevey
In Vevey, with its context of global food businesses, group headquarters and Riviera SMEs, a manager can ask business functions to list their priority problems before presenting what AI can offer. For each problem, the team then records whether it is being addressed with AI, addressed otherwise or left without action, together with the reason. The proposed measure is the proportion of priority problems left without action primarily because accessible data or an obvious AI application is lacking, which can be verified in this register at the next review. Within SHR’s programme « Conduire l’adoption de l’IA dans son équipe », this exercise would support a concrete management decision: choosing the tool to fit the problem, rather than the problem to fit the tool. To go further: explore the Leading AI Adoption in Your Team training in Vevey in the canton of Vaud, or browse our executive and employee training programmes in Switzerland.
In pictures: Leading AI Adoption in Your Team in Vevey
- leading ai adoption in your team
- Vevey
- research
- training vevey