← All writing

Musing · · 1 min read

Be the human in the loop, not the human at the end

Full end-to-end automation has a hidden cost: catching up on what the agent did. Checkpoints along the way fixed that for me.

I first posted this on X on August 6, 2026. It's here as posted.

A lot of people in our profession are chasing full end-to-end automation. I get the appeal and even chased it myself for a while, but I’m not convinced it’s the way for us. And my reasoning is primarily because of the context loss and the catch-up cost.

When you’re only in the loop at the very end, you’re reviewing something you now have little to no context on. So, you end up spending a lot of time figuring out what the agent did and why, that by the time you’re caught up, you’d have been faster doing it yourself the old way.

What’s fixed this for me is breaking my work into its natural steps and putting myself in the loop between them instead of only at the end. More checkpoints, yes, but less time at each checkpoint, more context gained, and thus less catch-up cost at the end.

The work still moves fast. I just never have to reconstruct what happened, since I was in it the whole way. So aim to be the human in the loop and not the human at the end.

Where to next

Follow along: @cbarmorecpa on X ↗ · The newsletter ↗ · More writing →