From Peter to Dilbert: How AI Inverts the Management Refuge
1969: The Peter Principle
In a hierarchy, every employee rises to their level of incompetence.
You’re a good developer? They promote you to lead. Good lead? They make you director. Until you reach the position where you become bad. There, you stay.
Peter’s conclusion: work is done by those who haven’t yet reached their ceiling.
1995: The Dilbert Principle
Scott Adams observes something else. We don’t promote the good until they fail. We directly promote the bad to get them out of the productive flow.
“The most inefficient workers are moved to where they can do the least damage: management.”
Management is no longer a reward. It’s a quarantine.
What If Everyone Became a Manager?
Before, being a manager was a privilege. A minority accessed it.
Now, you open Claude, and you lead.
So the filter changes.
Managing a human tolerates vagueness. Alignment meetings. The politically acceptable.
Managing an AI demands clarity. It doesn’t “feel”. It executes what you formulate.
The Inversion
The Dilbert Principle reverses.
Adams thought management protected the incompetent.
AI exposes them.
Peter described the rise toward incompetence. Adams, the refuge in incompetence.
AI removes the refuge.