Managing in the Age of Artificial Intelligence training in Lausanne · canton of VaudSupervising AI-Assisted Work: What Automation Research Teaches Us

Supervising AI-Assisted Work: What Automation Research Teaches Us — SHR, Lausanne
Supervising AI-Assisted Work: What Automation Research Teaches Us — SHR, Lausanne

The manager reviewing an output produced with AI does not face an unprecedented problem. Automation research has long described what happens to humans placed in supervision of a high-performing system.

The supervision trap

Raja Parasuraman and Victor Riley (1997, Human Factors) formalized the notions of use, misuse and disuse of automation: a system that is reliable most of the time encourages the cessation of verification, and error occurs precisely on the day the system fails. Parasuraman & Riley, 1997

Automation surprises

Nadine Sarter, David Woods and Charles Billings (1997) described how operators lose track of what the automated system is doing and discover late an unexpected mode of operation. The problem stems less from competence than from the design of the human-machine interface.

Supervising AI-Assisted Work: What Automation Research Teaches Us — Lausanne, Vaud
Supervising AI-Assisted Work: What Automation Research Teaches Us — Lausanne, Vaud

What this changes for a manager

This work concerns aviation and industrial systems, not language models. But the mechanism—increasing trust, decreasing vigilance—is documented with sufficient stability to be taken seriously in the tertiary sector. (our executive and employee training programmes)

The limits of general guidelines

Asking everyone to 'remain vigilant' amounts to entrusting the burden of verification to individual goodwill. Under time pressure, it is the first expense to be sacrificed.

A practical implication, verifiable in Geneva

In a team spanning higher education institutions, university hospitals and international headquarters, define for each type of AI-assisted deliverable an explicit control point—a figure, a source, a sensitive formulation—verified by a person other than the author. Counting errors intercepted over a quarter indicates whether supervision truly exists. AI Training in Geneva helps build this supervision framework. To go further: explore the Managing in the Age of Artificial Intelligence training in Lausanne in the canton of Vaud, or browse our executive and employee training programmes in Switzerland. AI training in the canton of Vaud.

In pictures: Managing in the Age of Artificial Intelligence in Lausanne

Managing in the Age of Artificial Intelligence training in Lausanne — in practice
Managing in the Age of Artificial Intelligence training in Lausanne — in practice
Managing in the Age of Artificial Intelligence training in Lausanne — hands-on workshop
Managing in the Age of Artificial Intelligence training in Lausanne — hands-on workshop
Managing in the Age of Artificial Intelligence training in Lausanne — on the ground
Managing in the Age of Artificial Intelligence training in Lausanne — on the ground