IA on the job: should an urgent request come first?
As requests accumulate, asking AI to prepare a priority list seems reasonable. The management question is more specific: how can an AI-assisted ranking avoid favouring a visible deadline over a production issue that is less immediately apparent?
What the literature establishes
Meng Zhu, Yang Yang and Christopher K. Hsee (2018, Journal of Consumer Research) showed in their experiments that people may favour a task with a short deadline even when another task offers a greater benefit. Urgency can therefore influence choices beyond what the value of the outcome would justify. These studies concern human decisions, not rankings generated by AI.
The mechanism that protects a priority
James Y. Shah, Ron Friedman and Arie W. Kruglanski (2002, Journal of Personality and Social Psychology) describe goal shielding: activating a goal to which a person is committed can inhibit the accessibility of competing goals. This helps explain why an explicitly maintained priority can withstand competing demands. Applied to AI-assisted work, the hypothesis to test would be that a clearly stated production objective protects prioritisation better than a mere collection of deadlines.
The premature conclusion
One might conclude that asking AI to rank tasks by importance rather than urgency is sufficient. Yet importance is not a property that the tool can always infer from a message: it depends, among other things, on the consequences of postponement and the dependencies between activities. An approaching deadline may genuinely be critical; the distinction lies in making the reason explicit, not dismissing it. (our executive and employee training programmes)
What the evidence does not establish
These studies do not demonstrate that an AI-generated list worsens priorities on a shop floor. Nor do they establish that a better-written instruction corrects the problem. In practice, comparing two rankings reveals little if nobody has documented the operational consequences that should guide the decision before examining them.
A practical check in Neuchâtel
Within SHR's « IA on the job » programme, an exercise could use anonymised requests from microtechnology, watchmaking or precision microelectronics: inspecting a batch, ensuring equipment availability or answering an administrative query. Before consulting AI, production and quality managers would document the consequences of postponement and known dependencies for each request, then assign a reference priority category. A ranking based on messages alone would then be compared with one also receiving this information, without changing the actual production schedule. The measure would be the proportion of requests considered critical in the reference assessment but classified as deferrable by the tool, retaining the inputs, outputs and reasons for the decisions. To go further: explore the AI on the Job training in Neuchâtel in the canton of Neuchâtel, or browse our executive and employee training programmes in Switzerland.
In pictures: AI on the Job in Neuchâtel
- ai on the job
- Neuchâtel
- research
- training neuchâtel