Brooks's Law in the Age of AI Agents: Not Repealed, Just Mutated

Jan van den Berg's short essay began with an everyday scene. Four people were desperately trying to get a laptop to connect to a screen, while the author watched contentedly from the sidelines with no intention of joining in. When someone nudged him to take a look, he laughed it off: in his experience, adding more people would not solve the problem any faster.

That was, of course, a nod to Brooks's Law: "Adding manpower to a late software project makes it later." Frederick Brooks formulated it in his 1975 book The Mythical Man-Month, drawing on his experience managing IBM's OS/360 project.

First impressions deceive

Still, the author started to wonder: does the law hold up in the age of AI agents and large language models? At first glance, it seems not. The classic bottlenecks — lengthy onboarding, endless stand-ups and office politics — vanish overnight. An agent gets up to speed in milliseconds and never asks for a coffee break.

Look closer, though, and it turns out the law hasn't disappeared. It has mutated.

What hasn't changed

The famous "baby problem" still applies: nine agents cannot deliver a baby in one month. Sequential tasks remain sequential.

The overhead hasn't gone anywhere either; it has simply shifted. Instead of managing people, you now manage agents. Someone still has to define what "done" means, review the code, catch hallucinations and stitch the pieces together. And that someone is you.

Throw nine autonomous agents at a single codebase without crystal-clear orchestration, and they will cheerfully rewrite each other's work in a high-speed loop.

An amendment to the law

Van den Berg ends by proposing a 2026 update to Brooks's Law: "Adding agents to a project with unclear requirements gets you the wrong answer faster."

Source: Hacker News