Automation Theatre
Some automation saves time. Some just moves the work somewhere harder to see. The difference matters more than we admit.
Walk through a modern office and you will hear people talk about automation the way an earlier generation talked about electricity. It is assumed to be everywhere, on all the time, and broadly a good thing. Question whether a particular piece of it is worth the effort and you are met with the same look you would get for suggesting we go back to candles.
But automation is not one thing. Pushing a payment through a clearing system without human intervention is not the same category of decision as having a language model summarise meeting notes. Conflating them serves the people selling the software. It does less for the people using it.
What gets hidden
There is a particular kind of automation that does not eliminate work but relocates it. The spreadsheet that used to take an hour to populate now fills itself, which is good, except that someone now spends forty minutes checking the output for errors that did not exist when a person built the thing row by row. The time has not vanished. It has changed shape, from construction to verification, and become harder to schedule and harder to account for.
This pattern repeats wherever automation sits between a worker and the thing being worked on. A teacher using an automated grading tool gains speed and loses contact with how a particular student thinks about a problem. A caseworker whose decisions are pre-filled by a risk model spends more time justifying overrides than making judgements. The system records the outcome. It does not record what was lost in getting there.
The work that disappears is often the work that gave the job its texture.
The phrase people reach for is "working with the machine," a framing that makes the arrangement sound collaborative. In practice it often means cleaning up after the machine: correcting its mistakes, translating its outputs into something a colleague can use, filling the gaps it was not designed to address. These tasks are real work. They are also invisible to the metrics that justified the automation in the first place.
The theatre of efficiency
None of this is an argument against automation. It is an argument against pretending that the costs are easier to measure than they are. When a tool promises to save ten hours a week, the sensible question is not "does it work" but "whose time does it save, and what fills the space it leaves behind."
Too often the answer is that it saves the time of the person who signs the purchase order, and fills the space with unrecorded labour by everyone downstream. That is not efficiency. It is theatre, and the audience is management. The actors, as usual, are not being paid to review the script.
What would a more honest approach look like? It would start by tracking the work automation creates, alongside the work it removes. It would treat the time spent checking, correcting, and compensating for automated outputs as a first-class cost, not a rounding error. And it would ask, before any deployment, who benefits and who absorbs the spillover.
These are not technical questions. They are questions about power, which is why they rarely get asked in meetings about procurement. Field Station intends to keep asking them.