How to tell if AI will take your job

How to successfully steer your career into the AI Age

Perfecting Equilibrium Volume Four, Issue 5

Well, we're living here in Allentown
And they're closing all the factories down
Out in Bethlehem, they're killing time
Filling out forms, standing in line

Well, we're waiting here in Allentown
For the Pennsylvania we never found
For the promises our teachers gave
If we worked hard, if we behaved

So the graduations hang on the wall
But they never really helped us at all
No they never taught us what was real
Iron and coke, chromium steel

And we're waiting here in Allentown
But they've taken all the coal from the ground
And the union people crawled away

The Sunday Reader, May 18, 2025

I ran out of room behind the cocktail table and he was finally able to corner me: What did you think of my talk?

Feola Rule Number 22: Never ask me a question unless you really, really want to know the answer.

It was fine, I told him, but it was instantly obvious that wasn’t going to fly.

Fine, I said to him. Can you name ANYONE who is making bank in digital photography?

He had given a presentation the previous week to the Society of Information Management, a worldwide group of IT senior executives. He’d written a book about technology strategy, and his talk covered the cautionary Tale of Kodak’s Failure To Strategize.

Kodak, he’d said – and written in the book – had invented the digital camera, then failed to strategize how that would change the business, and consequently collapsed.

Every complex question has an answer that is logical and simple.

And wrong.

Kodak was in the camera business the way Gillette is in the razer business. Gillette gives away razers so it can sell you the blades.

Kodak was in the film business. Digital cameras don’t need film. No film; no business.

The reason for this diatribe is not to have fun dunking on this guru. (OK, maybe a little bit.) The point is that every wave of technology has requirements and friction points, and businesses spring up to service both.

When the next tech wave hits, it not only washes away the now-outmoded older tech, it washes away those businesses. The destruction of the old order creates space and opportunity for the new world to take shape and be born.

But the old world of the dying Industrial Age had structure. The new world of the Information Age will have structure. Some day. After the struggles of birth are finished.

Here we are in-between, in the Interregnum between the companies and jobs of the Industrial Age, and the still nascent companies and jobs of the Information Age

Here be monsters, and uncertainty.

But we have no choice but to live in the age in which we are born, so what do we do now that we are here?

We understand, so we can plan.

If your job is part of an Industrial Age technology like film, then your job is going away. Here’s reader Lynn Gardner with another example: Highways of America are littered with the remains of careers that don’t exist anymore. I and thousands of others as Manufacturers Representatives used to call on clients and keep them apprised of what was selling and what was not. We were all replaced by Point Of Sale cash registers that provide clients instant information on what is selling as opposed to a visit from me. There are many more examples, but the moral of the story is don’t be the last one holding on to a career or idea after its time has passed.

Lynn is right, and that’s our discussion today: How to tell in advance whether your career’s time has passed. For blue-collar manufacturing jobs, the key question was whether your job could be automated. If your job was tightening bolts on an assembly line, then yes: you were going to be replaced by an assembly-line robot.

Today white collar jobs are threatened by AIs. And the problem with these threats – assembly-line robots, AIs – is that they not only put individuals out of jobs, they destroy the surrounding communities, surrounding these newly unemployed.

We’re all familiar with what happened to the Rust Belt manufacturing cities. Economists David Dorn, Gordon Hanson, and David Autor did a study on the effects of factory closings. Adjustment in local labor markets is remarkably slow, with wages and labor-force participation rates remaining depressed and unemployment rates remaining elevated for at least a full decade…

This is why Learn To Code has always been not just bad advice, but stupid, and frankly abusive. Teaching coding to West Virginia coal miners and Pennsylvania steel workers didn’t help when all the entry-level coding jobs were in Mumbai. And it won’t help office workers today when all the entry level coding is being done by Curser and GitHub Copilot.

The correct strategy is to move to a new field before your career is replaced, and the supporting community dies. The question, of course, is how to know if your job is in the line of fire.

That answer, fortunately, can be boiled down to one word:

Judgment.

Does your job require judgment? Or does it require accurately following procedures?

Here’s a career that’s doomed: Junior consultants. Big consulting firms like Accenture have platoons of junior consultants who turn research reports written by senior staff into PowerPoint decks for client presentations.

Let’s say a company hires the consultants to see if outmoded tech is the reason sales are off in the EU. The senior folks do the analysis and write the report. Junior consultants then turn those endless pages of reports into colorful PowerPoint slides.

But these slides have exacting standards, down to stuff like the exact fonts and character sizing for footnotes and slide page numbers. I was once hired by one of the big consulting firms to do such a report. They paid five figures. I enjoyed the work; it was definitely worth it.

Then they wanted a PowerPoint. After I agreed, they sent me those guidelines.

If I had known, I would have charged them six figures.

But now Pirate Wires reports panic among junior consultants over an AI PowerPoint creation tool called Deckster launched at the Boston Consulting Group.

They are right to panic. This is a perfect job for an AI to replace: turn information in one extremely complex format into a different format that is equally complex. Note that there’s no judgment involved here, just moving from one format to another.

So-called AIs are no kind of intelligence, artificial or otherwise. They are Large Language Models that are excellent at assembling patterns and mapping one pattern to another.

They are lost when judgment is required.

Let’s say you work in a law office. Formatting a legal brief into the required submission form for a particular court? No judgment needed, and an AI can do that.

Choosing which legal brief best makes your case AND will persuade Judge Lawrence “Get to the point or I’m not reading the rest of your brief” Law? Judgment needed, and subtle judgment at that.

Let’s say you work in interior design. Choosing Modern furniture that fits your client’s room? An AI can do that.

Furnishing a room for your client who claims they only like Modern furniture, but actually doesn’t? Knowing how to design and create a satisfying space for a client that knows what they want but just doesn’t have the words to explain?

That’s a job for humans. Judgment needed, and subtle judgment at that.

So here’s the task: To survive and thrive, each of us must look closely at our careers. Does our job require following procedures like trains down a track? No matter how complex those procedures are, Large Language Models – AIs – will replace those jobs.

Or do you have a career with no tracks, and no train? Is steering through your day like one of those crazed Mission Impossible car chases, making snap judgments through attacking enemies, innocent parents pushing baby carriages, ice cream carts and the like?

You’re good to go for decades.

And after that?

Well, we’ll let the denizens of the 22^nd Century worry about that.