Main idea

Separating whoever designs from whoever executes had a quantifiable reason. Translating a need, documenting, drafting, formatting, keeping tickets up to date cost enough time to justify a dedicated role to absorb that plumbing. The division wasn't chosen for its virtues: it was paid for because it came cheaper than the alternative.

When that cost collapses, the justification falls away and the division holds only through habit, the org chart and the careers it created. Nothing about the product or the market has changed. In the product case, automation therefore doesn't create the problem of the separation: it removes the last practical excuse available — "in theory whoever designs should follow through to production, but in practice there's no time".

Hiring today on a job description built around the transmission of information amounts to optimizing an already obsolete circuit.

Why it matters

This gives a criterion for anticipating what happens to organizations: spot which divisions of labor rest on a cost that is falling, and which rest on something else.

And it defuses a false debate: the question isn't whether a tool can do a person's work, but whether what justified splitting that work still exists.

Nuances and limits

"Not long" is sometimes counted in decades. A division of labor is also held in place by power, careers, market conventions and collective agreements, long after its economic motive has gone.

And a collapsing cost isn't a vanished cost: delegated execution comes back as framing, review and supervision, which can be enough to justify keeping a role, but rarely the same one.

Open questions

• What clue lets you spot, before the reorganization, that a division of labor is held up by nothing but the org chart?