What happens when people have to jump in to clean up AI’s mess?

Imagine an artificial intelligence system that has performed reliably for months. Its recommendations are almost always right, and the employee reviewing them has learned that close scrutiny rarely changes the outcome. Then something unusual happens.