Terminal D, Miami International. An autonomous wheelchair rolled past me. A passenger sat in the chair. Nobody was pushing it.

Two guys at the coffee kiosk were watching. I overheard one say, “Someone just lost a job.”

That remark sent my thoughts upstream: to the people who designed the chair, built it, and would keep it running.

The wheelchair did not appear by magic. Someone wrote the navigation software. Someone built the sensors and manufactured the chassis. Someone has to maintain the fleet and handle failures.

The visible task had changed. Much of the work behind it was out of sight.

The visibility problem

The wheelchair pusher is someone you can see. The engineer debugging the software and the technician replacing a worn motor are harder to picture.

I think that shapes how we talk about automation. The task being replaced caught their attention; the work behind the chair caught mine.

Looking upstream gives me another part of the picture. It does not tell me whether more jobs were created than lost. I would need evidence for both.

Acemoglu and Restrepo’s research on US industrial robots found employment and wage losses in places more exposed to robots. That study covers industrial robots, not airport wheelchairs.

It gives me a reason to take displacement seriously, rather than assume that new work cancels it out.

The transition still has a cost

If you are the person who just lost work, new jobs somewhere else do not pay the rent. They may need different skills or be in a place you cannot move to.

For me, retraining, financial support, and a practical route into new work belong in the discussion.

I still see opportunity in what that chair represents. I also want to know who can take part in it.

Follow the work

Watching the chair, I could see one task being automated. I could not see the full employment picture.

To understand that, I would need to look at the jobs removed, the work created, and whether the people affected can move between them.

The chair made me look upstream. It did not give me an employment total.