Keeping the skill that made you useful

The oldest finding in automation

Aviation learned this decades ago. Autopilots made flying safer and, over years, manual flying skills degraded — not because pilots became worse people, but because the practice that maintained the skill had been removed. Regulators eventually responded by requiring hand-flying practice, deliberately, on purpose, at some cost in efficiency. The lesson generalises: the skill you stop using is the skill you stop having, and you will not notice until the day the automation is unavailable or wrong.

That day arrives in office work as a specific moment. The tool is down, or the material is confidential, or the answer is subtly wrong and you are the only one who could have caught it.

The two losses

Yours. If you never write the first draft, your writing gets slower and flatter over a year or two. If you never work through the analysis, your feel for when a number is implausible fades. This is gradual and it is invisible from the inside, because the output on your screen is fine.

Your juniors'. This one is larger and it is not yours alone to manage. Professional judgement is built by doing bad work and being corrected. The trainee who drafts a poor contract clause, has it torn apart, and understands why, is acquiring something. The trainee who edits a competent AI draft is acquiring something much thinner — and looks more productive while doing it, which is what makes this hard to resist.

Two years of that and you have people who can operate the tool and cannot evaluate it. That is precisely the population you need on the day it is wrong.

What to do about it, concretely

Keep a practice ration. Some proportion of the work you do unaided, deliberately, because it is practice. Once a week, one document, from a blank page. It costs you thirty minutes and it is the cheapest insurance available.

Do it yourself first, then compare. Write your version, then ask for one, then read the two side by side. You get the benefit of the comparison, you keep the practice, and you learn considerably more than you would from either alone. This is the single best habit in this lesson.

Use it as a tutor, not a substitute. Explain why this clause is drafted this way. What am I missing in this analysis? What would a sceptical reader attack first? All of these leave the work with you and put the model where it belongs, which is beside you rather than in front of you.

Protect the reps for juniors. If you supervise: keep some tasks unassisted and say why. "Do this one yourself first, then we will compare" is a sentence that costs an hour and preserves a career. And be honest that it is slower — pretending otherwise makes it look like a ritual instead of training.

Notice the discomfort. If starting a document from nothing now feels harder than it did a year ago, that is the measurement. It is also reversible, and it reverses in weeks rather than years.

The distinction that decides everything

There are two ways to end up three years from now.

One is faster at producing documents and no better at the work — dependent on a tool, unable to tell when it is wrong, holding a skill set that has quietly thinned while the output looked fine.

The other is genuinely better: able to do more, having used the time saved to think harder, read more, ask the second question, do the version you could not previously afford to do.

The difference is not which tools you used. It is whether you spent the time you saved on the work, or simply produced more of it. The tools are indifferent to which you choose, and nobody at your organisation will make the choice for you.

Where this course ends

You now have a way to decide what to hand over, a way to brief it, a way to work from your own material, a set of lines you will not cross, and a way to prove whether any of it helped.

The last thing to keep is the one nothing here can supply. Your judgement about whether an answer is right is the whole of your professional value, and it was built by doing the work. Keep doing enough of it.