We are standing on a concrete floor in Toyota City, and a man named Taiichi Ohno is bending down with a piece of chalk. The year is somewhere near 1950 — the story is vague about this, and we will come back to why. He draws a circle beside a machining line, straightens up, and waves over a young manager. Stand there, he says. Watch. Then he walks away.
The manager stands. An hour passes. The line does what lines do: parts arrive, parts wait, a worker walks somewhere and comes back, a machine idles for reasons nobody has written down. When Ohno returns and asks what he saw, the first answer is not good enough. Neither is the second. He walks away again.
Somewhere in the third hour, the floor changes. Not the floor itself — the seeing. The walking resolves into a pattern. The waiting acquires a shape. The pile of parts beside the second press stops being scenery and becomes a question. The manager has stopped looking at a factory and started seeing work.
What the circle teaches has little to do with patience. It is a claim about where knowledge lives: what is actually happening exists in exactly one place, the place where it happens, and it does not survive the trip into a report. The Japanese word for that place is gemba — “the actual place.” You cannot delegate the standing. You can only do it.
This walk asks where the circle should be drawn in 2026, when the work is a stream of tokens passing between a developer and a machine. The street runs from this floor to that one, and it only works in order.
You carry three questions the whole way, and they never change. Purpose — what problem is this work solving, and for whom? Process — does each step create value, or does it wait, wander, and get done twice? People — are the people engaged in improving the work, or merely enduring it? Every station comes back to these three. Pick them up. We are walking.