The Surgical Team

The Problem

Small teams are productive but can't build large systems fast enough. Large teams have the manpower but drown in communication overhead. How do you scale without losing the efficiency of a small, sharp team?

The 10x programmer observation makes this worse. A few great programmers vastly outperform a large group of mediocre ones. You want the great ones doing the critical work — but they can't do everything.

Mills's Proposal

Harlan Mills proposed the surgical team model. One chief programmer does the creative work — design, coding, testing. The rest of the team supports the chief: a copilot as a sounding board, an administrator handling resources, an editor preparing documentation, a toolsmith building specialized tools.

This gives you the productivity of a single brilliant mind with the support structure of a full team. The chief programmer's time is never wasted on administrative or mechanical work.

How It Works

The key insight: not everyone on a team needs to be doing the same kind of work. Specialization lets you scale through coordination rather than through division of the creative work itself.

To build a truly large system, you coordinate multiple surgical teams, each responsible for a component. Communication happens between team leads, not between every individual. This keeps the communication channels manageable while applying talent where it matters most.