Nobody is hiring for the steno pool

Not so long ago, every large office had a room full of people whose job was to type. They were hired on words per minute and accuracy, and they were good at it.

I don't see many of those job postings anymore.

I don't say that to sound uncaring. The work didn't vanish. The bottleneck did. Once everyone had a keyboard and a word processor, typing stopped being a profession and became something we all just do. "Computer" went the same way. It used to be a job title, held by a person.

That pattern is the best guide I know to where AI is taking us.

Every wave removes a bottleneck

The tech writer Rene Ritchie laid this out in a video in March 2023 that crystallized it for me.

His arc goes like this. Before the internet, you spent days in a library just finding the raw information before any real work could begin. The web made finding things trivial. Mobile made it trivial to work from anywhere. Generative AI is now making the first draft trivial.

Each time, a chore that used to eat most of the effort shrank to almost nothing, and the valuable part of the work moved up a level.

The last 20 percent

Here is where it has moved now. AI can produce the outline, the first draft, the prototype, and the summary in minutes. It's often confidently wrong, so it needs a human mind guiding it.

What's left for people is the part that was always worth the most: choosing, judging, polishing, and directing. Call it the creative 20 percent.

That's augmentation, not replacement. But notice what the 20 percent includes: checking the other 80.

Yes, some jobs go away

I won't pretend otherwise. The steno pool is gone, and some of today's roles will follow it.

History offers some comfort, though. The typists didn't become unemployable. The work changed shape, and the people who learned the new tools did fine.

For leaders, that's an obligation. Redesign the roles and coach your people through the change. For everyone else, it's the lesson I learned the hard way: keep learning, or get left behind.

The trap

Ritchie makes a point I keep returning to. Every earlier wave of better tools could have let us do the same work in less time. Instead we did more work in more time, and ended up busier.

AI is about to hand organizations a great many freed-up hours. What happens to those hours is a leadership decision. You can fill them with more of the same, or you can spend them on better work.

Continuous transformation, for people

The march of technology doesn't stop, so there's no point at which the learning is finished.

The answer is the one we apply to systems: iterate. Small, steady upgrades to your own skills beat one panicked retraining every decade.

Nobody is hiring for the steno pool. Plenty of people are hiring for what the typists learned to do next.

This article grew out of a post that first appeared on LinkedIn. Join the discussion there.