The Shift: the missing rung

How organizations used to grow expertise

Most organizations grew expertise by accident. A wide base of newer people did a large volume of routine work: first drafts, data gathering, reconciliations, standard cases, document review. Someone more experienced corrected it. Over hundreds of repetitions, people learned what good looks like, which exceptions matter and how to handle a client.

Very little of this was designed. It happened because the work existed and someone had to do it.

What AI changes

AI is very good at exactly that base layer of work. The productivity gain is real. But the work was also the training, and when it goes, the training goes with it unless someone puts it back on purpose.

Early evidence points the same way. Payroll research from the Stanford Digital Economy Lab finds the effect landing on young workers in AI-exposed jobs first, mainly through fewer hires rather than layoffs. The authors call these early, descriptive findings, and I treat them that way.

Three compressions

The shift is easier to manage when you break it into three forces:

1. Pipeline compression. Fewer entry points, and less senior time for each person learning.

2. Learning-curve compression. Judgment (the ability to make good calls) is needed sooner, with less low-stakes practice behind it.

3. Ladder compression. Fewer, wider levels, so growth is harder to see and to reward. This one is a common organizational response to the first two, not a direct effect of AI: the pressure is real, but the flattening is a choice.

Why it matters later

The early symptoms are visible now: overwhelmed new hires, overloaded managers, people leaving from the middle. The serious costs arrive three to five years on:

• a succession gap nobody can hire their way out of,

• oversight of AI that has quietly become a rubber stamp,

• people who used AI to do the work without learning the work.

A consequence, not a goal

When the base narrows and the middle is redeployed to oversee AI, the classic pyramid can turn into a diamond. I use the word to describe what tends to happen, not to recommend it. Ready Sooner doesn't move anyone toward a target shape. It helps you see the pressure and respond deliberately, close to the work, in hierarchies and in networks of teams alike.

Read the full argument

The Missing Rung is a discussion paper on this shift: the evidence, the three compressions, and a set of practical tools, all offered as proposals to test. It is in review now.

Ask for the discussion draft 

Source: Brynjolfsson, E., Chandar, B. and Chen, R. "Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence." Stanford Digital Economy Lab, revised August 2026. https://digitaleconomy.stanford.edu/publications/canaries-in-the-coal-mine