84% of developers use AI tools. 46% actively distrust what they produce.

Both numbers are rising at the same time.

That is not hypocrisy. It is exhaustion.

The Wrong Diagnosis

Most of the conversation treats this as a job-market story. Will AI replace developers? Will salaries fall?

Wrong question.

The experienced developer who feels worst is often still employed. Still shipping. Sometimes shipping faster than before.

What breaks first is not the paycheck. It is the answer to a question nobody had to ask before:

Who am I as a developer?

Two Camps, One Fault Line

Some developers call it the Coding Depresh — a grief not for employment, but for identity.

A fifteen-year veteran ships a pipeline in an afternoon that would have taken three days by hand. He expected pride. He got grief. Not for a person. For an identity.

The other camp feels the opposite. More capable. More output. Closer to the problem, further from the syntax.

The split is not seniority or skill.

It is whether your identity lived in the typing or in the building.

If “who you are” was “the person who writes excellent code,” AI hits at the identity layer. If “who you are” was “the person who makes the thing exist,” the same tools feel like leverage.

The Question That Sorts You

Do you love writing code, or what you get out of writing code?

Most of us never had to separate those two. They came bundled.

AI unbundles them.

That is not a verdict on your worth. It is a map of where you actually live.

Why the Paradox Makes Sense

You use a tool you do not fully trust because the alternative is slower.

You verify output you did not write because “almost right” is worse than wrong — it looks finished until it breaks in production.

66% of developers say they spend more time fixing almost-right AI code.

That is not adoption failure. That is vigilance without resolution.

Working with something you cannot delegate your judgment to is tiring even when it works.

Where I Land

I am in the second camp.

I never tied my identity to typing every character. I tied it to deciding what should exist and making it real.

AI did not take that from me. It removed the part I was willing to let go of.

That does not make the grief unreal for people who loved the craft itself. Fifteen years of identity does not evaporate because a tool got faster.

Grieve if you need to. The loss is real.

Then ask the question.

Do you love writing code, or what writing code gets you?

What Survives Either Way

There is a difference between not caring about the code and caring deeply but not typing most of it.

The taste that took years to build is still doing the real work. Knowing when architecture is quietly wrong. Knowing when output is almost right but not quite. Knowing when to reject what the model produced.

That instinct is not a bug. It is the sharpest instrument you own.

The job was never typing. It was judgment.

AI made that visible by automating the part that looked like the job.

Key Takeaways

The Depresh is identity disconfirmation, not job-loss fear.

84% use AI. 46% distrust it. Both can be true.

The fault line: craft identity vs outcome identity.

Grieve the loss if the craft was the point. Then ask what you actually loved.

Judgment, taste, and systems thinking survive either way.

The typing was never the whole job. It just looked like it.

Data from the 2025 Stack Overflow Developer Survey.