Career Progression in the Age of AI

The old ladder made sense when effort mapped more cleanly to value

Most people in tech are taught some version of the same progression model.

You begin by learning the fundamentals. You write more code. You fix smaller things. Over time, you earn trust. Then your scope expands. Eventually you are no longer judged just by what you can build, but by how well you think, how much you can own, and how many hard things become calmer because you are in the room.

That model still has truth in it.

But it was built in a world where implementation itself was a stronger filter.

If you wanted to know whether someone was good, you could often see it directly in how they debugged, how they structured systems, how they navigated ambiguity, or whether they could move from idea to working software without getting lost in the middle.

Now a lot of the visible surface area of competence is getting cheaper.

You can produce more code than before. You can scaffold faster. You can generate patterns you have not memorized. You can get unstuck without spending three hours trawling forums written by someone named HexBlade2007.

That is genuinely useful. I am not against it.

But it does create a new problem.

If more people can produce plausible output, then output alone becomes a weaker proxy for depth.

And once that happens, the old ladder starts to wobble.

The uncomfortable part is not that AI makes people faster

The uncomfortable part is that it blurs the old signals.

A junior engineer with good prompting can sometimes look more capable than they are.

A senior engineer who thinks carefully can look slower than the moment seems to reward.

A team can ship more while understanding less.

That last one worries me the most.

There is already a subtle shift happening in software. In some corners, engineering is moving from understanding systems deeply to coordinating enough output to keep momentum alive. Sometimes that is a healthy evolution. Sometimes it is just a nicer way of describing shallowness.

If the job becomes, "use tools to get the task done," then seniority starts to sound strangely cosmetic.

What does it mean to be senior if the thing you were told to master is being compressed into a prompt window?

What does it mean to be principal if a machine can generate the first draft of the architecture, the code, the tests, and the docs before your coffee cools?

I do not think the answer is that seniority disappears.

I think the answer is harsher and more interesting.

The ladder is not vanishing. It is being redrawn.

If AI keeps making execution cheaper, then the value of an engineer moves upward.

Not upward in title. Upward in abstraction.

The differentiator becomes less about who can produce the most code, and more about who can:

• define the right problem
• choose what should exist at all
• recognize when the generated answer is wrong in a dangerous way
• preserve product quality when speed is everywhere
• make good tradeoffs under uncertainty
• build systems people can actually trust

In other words, judgment starts mattering more than volume.

That sounds obvious, but I do not think most people have fully absorbed what it means.

Because judgment is harder to fake.

You can borrow syntax. You can borrow structure. You can even borrow momentum. But you cannot easily borrow taste. You cannot borrow responsibility. You cannot borrow the calm clarity required to decide what matters when ten plausible options are on the table.

That is where I think the next version of engineering seniority lives.

Less in raw production. More in authorship.

This is why early-career engineers feel uneasy

If I were entering the industry right now, I would probably be asking a slightly panicked version of the same question a lot of people are quietly holding:

That anxiety makes sense.

The old advice was simple. Learn deeply. Put in the reps. Become excellent at the craft. Climb over time.

But if the environment is changing fast enough, people start wondering whether the reps are still compounding in the same direction.

If it takes two or three years to become meaningfully stronger, what happens if the job itself changes shape before you arrive?

That is not laziness. That is a rational question.

Still, I do not think the answer is to stop growing.

I think the answer is to grow in a way that survives tooling shifts.

Learn how systems behave. Learn how products fail. Learn how people use things in ways the builder did not intend. Learn how to tell the difference between something that works and something that is trustworthy. Learn how to think.

The market may keep changing the interface of engineering. It will still reward people who can see clearly.

Art might be the better analogy now

For a while I kept thinking about engineering ladders like ropes. You climb. You gain height. You reach the next level because you have the strength and skill to hold yourself there.

Lately that analogy has felt less useful.

Art feels closer.

AI can help people make images, music, writing, software, and just about everything else with much lower friction than before. That changes the process. It lowers the barrier. It even expands who gets to participate.

But it does not remove the human part that matters most.

The difference between forgettable work and meaningful work is rarely just the raw material. It is taste, intention, emotional truth, restraint, and the invisible choices that make something feel alive instead of merely complete.

I think engineering is moving in that direction too.

Not because code is becoming art in some dramatic, hand-on-heart way.

I mean that the frontier is shifting from mechanical production toward discernment.

Toward seeing what should be built.

Toward deciding what should be left out.

Toward understanding what quality feels like before metrics catch up.

So, will there still be senior engineers?

Yes.

But I think the title will mean something different.

The strongest engineers will not just be the people who can build without AI. That is too narrow, and probably a little nostalgic.

They will be the people who can use new tools without becoming dependent on shallow thinking.

The people who can move fast without becoming careless.

The people who still understand systems, users, tradeoffs, and consequences when everyone else is busy celebrating output.

That kind of seniority might actually be rarer.

Not because the tools are making excellence impossible, but because they make imitation easier.

And when imitation gets cheap, substance matters more.