Adam Smith opened The Wealth of Nations with a pin factory. One untrained worker makes a handful of pins a day. Ten workers, each doing one of eighteen steps, make tens of thousands. Specialisation won, and for almost 250 years it kept winning, anywhere the work was repetitive enough to split.
The part people skip is Smith’s warning. A person confined to a few simple operations, he wrote, becomes “as stupid and ignorant as it is possible for a human creature to become.” In software we have a gentler word for it. We call it a hand-off.
Because in software the unit of work is not the pin. It is the understanding. Divide the labour and you divide the understanding, and the gaps between the specialists fill up with rework, brittle interfaces, and defects nobody owns.
AI changes the arithmetic. When one person can carry a feature from discovery to build to test, with a model doing the mechanical parts, the case for splitting a small team by job title gets weaker every month. So we did the unfashionable thing and let the boundary drift back toward the generalist.
Like the novel this borrows its title from, it is really a story about patience. You specialise late, once the system is stable and the work is genuinely worth dividing. Not on day one because an org chart told you to.
Eleven years in, we are still resisting the pin factory. Some things are worth the wait.