Executive Briefing: AI in human resources—Use cases and ethical issues

Description

AI and its related subtechnologies are being introduced into operational decision making throughout the enterprise. The most promising and risky experiments involve the way people are selected and utilized, but the use of AI in HR raises the specter of software product liability. John Sumser offers an overview of the available use case solutions and the accompanying ethical issues.

Use cases: There are currently 22 use cases for which solutions are being proffered, ranging from flight risk predictors to organizational learning acceleration in call centers. John walks you through each of them.

Latency: Generally, AI techniques excel at solving problems where goals are consistent, rules are fixed, or both. In human organizations, goals vary, rules change, and consistency is always a slight improvement away from reality. Our organizations are complex adaptive dynamic systems. When machine intelligence learns the current status quo, it is just a question of time before it is interrupted and the machine has to relearn. This means that machine recommendations have a hard-to-measure variable quality that confounds decision making.

Ethical issues: There is a range of potent and disturbing ethical issues. The quantification of work guarantees that currently configured systems will never find the Einstein in your company. Numerous social psychology experiments have shown the human tendency to follow instructions from an authority figure. Employees are likely to do that. A key question is how to teach the workforce to argue with machine recommendations.