Field Note: Human in the Lead

“Keep a human in the loop.” It’s the phrase everyone reaches for when they talk about using AI responsibly — a person who checks what the system does before it goes ahead. But lately I’ve been thinking that it’s a potentially dangerous approach, and I want to say why.

The first problem is what it does to you. Being kept in the loop makes you passive — you sit at the end and approve whatever the system hands you. It’s comfortable, but it isn’t especially empowering, which is one step toward losing out to automation altogether.

The second problem is what it suggests: that the agents know what they’re doing, and that you, the operator, know and trust that they do. Neither is safe to assume yet. I’ve written before about the Context Gap — my own inbox agent once processed a scam correctly by its own rules and queued me to ship an exercise bike to a fraudster, marked HIGH, due today. My agent did its job by its own rules; it just didn’t know enough about me or about actual thinking to make an accurate judgement call. Trusting that it knew what it was doing is exactly the assumption that would have printed the label - and that is dangerous thinking.

So I’ve started using a different phrase, and I’m going to lean on it going forward: human in the lead.

The difference is who’s in control. In the lead, you’re running the process — you decide what it’s for, and you answer for what it does. I build on these tools every day, but I don’t let them run without me.

We need leaders in new fields, and we need humans in the lead.

— Jim

→ Related, on where this is already being built into the infrastructure: Cloudflare built the agent cloud in a week — and kept the human at the controls​

The system I keep the final say over: most of my task setup live in Notion. If you’re a startup or solo founder, you can get 3 months of Notion Business free — unlimited AI, no credit card through my link. Start your 3 free months →​

Notion affiliate link — I get a small credit if you start a trial