The Ghost in the Machine: Why We’re All Becoming AI Babysitters

Why the future of work is less about creating from scratch and more about curating meaning.

Hey founder 👋

I read a piece recently about how live-coding engineers are slowly morphing into AI babysitters. They don’t mind it because instead of hand-writing every line, they’re corralling, reviewing, and nudging AI output into something that works.

But the more I thought about it, the more I realized: this isn’t just engineering. It’s law, finance, writing, design — pick your field. We’re all slowly becoming overseers.

Editing AI: The Missing Connectome

I’ve been editing a book lately that was heavily AI-generated. And honestly? It’s one of the most fascinating editing challenges I’ve ever had.

AI can spin up a structure, 10 chapters on how to build an app, each expanded with neat subpoints. But the continuity? The flow? The connectome… the human web of connections and meaning is missing.

In code, that doesn’t hurt as much. If it runs, it runs. In writing, though, the gaps are glaring: repetition, clunky similes, the same “it’s not X, it’s Y” rhythm over and over. Filling in those connective tissues requires human experience, memory, and taste.

Overseers, Not Originators

That’s what most of us are really doing now: not generating from scratch, but shaping, pruning, editing. Saying:

• “This matters, this doesn’t.”

• “That law is real precedent, that one isn’t.”

• “This character arc sings, that one falls flat.”

The AI will keep trying to please us. But what gives the work value is not its output, it’s our discernment.

The Exhaustion Factor