Weekend Reflections #4 | The Molting

[Views are my own]

I got my first computer when I was seven. A Commodore VIC-20. I have been in front of a screen for around forty-five years.

That means I have watched several waves arrive up close. The internet. The dotcom collapse. The BlackBerry era and its quiet death. The smartphone.

Each time, the same question surfaced: will the jobs survive?

The better answer, each time, was not yes or no. It was: some jobs disappeared, but many roles moved.

I have been thinking about why that is, and whether the pattern holds now.

The question I keep returning to is a simple one: when a technology makes part of your job redundant, what actually happens to the job?

The obvious answer is that it disappears. That is the fear most people carry right now with AI. But history suggests a more complicated answer.

I started tracing the history of a few familiar roles. Not in a formal way, just following the thread backward through time. And what I found was not simple disappearance. It was molting.

Take sales.

Before industrial manufacturing, the person who sold things was often the same person who carried them. The merchant moved across trade routes with physical goods. Her value was access: she was the only one who had the silk, the spice, the thing you could not get locally.

Then infrastructure changed. Products became abundant. Information became cheaper. The merchant became the catalog salesman, then the relationship manager, then the CRM-powered solution seller of the last decade.

Each time, the execution layer automated. Each time, the judgment layer survived.

Or take the person who protects systems.

In the pre-industrial world, protection meant physical presence. The watchman stood at the gate. Then locks got better, factories arrived, and the threat became sabotage, not bandits. Then mainframes introduced the sysadmin, managing passwords and physical server rooms. Then the internet made perimeters obsolete, and monitoring replaced patrolling.

In each case, the role did not die. It shed what the technology could now do, and kept what it could not.

Or take the person who finds the truth in numbers.

The royal scribe recorded harvests because literacy was rare. The actuary calculated risk because computation was rare. The dashboard builder used SQL and Tableau because visualization was rare.

Each time, the commodity layer automated. Each time, the role survived by moving one step further into interpretation.

None of these roles disappeared. They kept the judgment core and released the execution layer.

Some roles did disappear. The lamplighter. The elevator operator. The typing pool. They are not coming back.

But look at what they had in common: once the execution layer automated, there was nothing left underneath. The task was the role. Strip it, and you have an empty title.

The roles that survived were the ones with something beneath the task: responsibility for an outcome, judgment under uncertainty, accountability for what happens next.

That pattern has repeated often enough to matter. Not a law of history, but a useful lens.

Your role keeps going as long as there is something beneath the task: judgment, accountability, taste, context, or responsibility for an outcome.

AI may challenge this pattern more deeply than previous technologies because it does not only touch execution. It also touches drafting, analysis, synthesis, and parts of judgment. That is why the question is not whether judgment disappears. It is whether our judgment moves higher, faster than before.

And yet I understand why this moment feels different.

Because it is.

The collapse is not in the pattern. The pattern is the same. The collapse is in the timeline.

In the 1800s, you had a generation to adapt. Your grandfather learned the old way. You inherited the transition as a natural part of growing up in a changing world. The pace of change fit inside the pace of human life.

By the 1990s, that window had compressed to roughly a decade. Enough time to notice the shift, retrain, and arrive on the other side of it with a career still intact.

I lived that version.

When the internet arrived, I had years to move from coding for its own sake to understanding what it meant for products, teams, and markets. When the smartphone era came, the window was shorter, but still measurable in years.

Now the transition from one era to the next is happening in months.

Not in the abstract. Not as a forecast.

In job postings today. In the performance reviews being written this year. In the tools being retired inside companies that were still buying them last quarter.

The historical pattern says: your role will not disappear all at once.

It will be stripped.

The repeatable parts will move into tools. The coordination layer will move into systems. The visible output will become cheaper.

What remains is judgment: what to ask, what to trust, what to ignore, what to do next.

That is not new.

What is new is the speed.

Previous generations had time to become different workers. The window between waves used to span a lifetime. Then a career. Now it is measured in quarters, and the quarters are moving faster than the retraining programs.

So the question is no longer whether your role will change.

It is which layer of your work you are still defending, and which layer you need to shed before the market sheds it for you.

I have done this a few times already. Each time, the forgetting felt like loss. Each time, it turned out to be the move.

What part of your work are you still protecting because it used to define your value?