AI assistants: when should direct dialogue be preserved in Vevey?
In corporate headquarters and SMEs across the Riviera, a no-code assistant can act as an intermediary between departments that do not always attach the same meaning to the same words. The management question is specific: when does this intermediary facilitate coordination, and when should direct dialogue be preserved?
What the literature establishes
Richard L. Daft and Robert H. Lengel (1986, Management Science) distinguish uncertainty, arising from insufficient information, from equivocality, arising from the coexistence of different interpretations. Their theoretical framework explains why more information is not necessarily enough to coordinate an organisation. An assistant that completes a file may address a gap; this does not mean it resolves disagreement about what a request means.
Conveying information is not reaching agreement
Alan R. Dennis, Robert M. Fuller and Joseph S. Valacich (2008, MIS Quarterly) distinguish the conveyance of information from convergence towards a shared understanding. Their theory links these processes to media capabilities, including the ability to revise a message, revisit it or exchange messages rapidly. Applied to AI assistants, this distinction suggests separating the preparation of case information from the dialogue needed to agree on its interpretation.
The premature conclusion
One might conclude that assistants should handle only simple requests, while every complex issue requires a meeting. The distinction lies elsewhere: a technically complex case can be handled asynchronously if its criteria are shared, while a short request can conceal disagreement between departments. The useful criterion is therefore not simply the difficulty of the content, but the need to build a common understanding. (our executive and employee training programmes)
What these studies do not demonstrate
These two articles offer theoretical frameworks for organisational communication, rather than experimental evaluations of generative assistants. They do not establish that direct dialogue is always superior, or that an assistant necessarily fails to facilitate agreement. In practice, however, a fluent response does not prove that the departments concerned attach the same meaning to a priority, an exception or a commitment.
A practical check in Vevey
Within SHR’s programme « Créer ses assistants et agents IA sans coder », a relevant exercise in Vevey would be to configure an assistant that prepares requests between purchasing and quality teams, at a global food-industry headquarters or a Riviera SME. During a predefined pilot period, record the proposed acceptance criterion for each request and ask both departments separately whether they approve it; where they differ, arrange direct dialogue rather than another automatic reformulation. Measure the proportion of closed requests subsequently reopened because of differences in interpretation, comparing it with a baseline period involving comparable cases. This local measure would not establish causality on its own, but would make it possible to check whether the arrangement supports agreement rather than merely producing messages. To go further: explore the Building AI Assistants and Agents Without Coding training in Vevey in the canton of Vaud, or browse our executive and employee training programmes in Switzerland.
In pictures: Building AI Assistants and Agents Without Coding in Vevey
- building ai assistants and agents without coding
- Vevey
- research
- training vevey