The identity crisis nobody is naming

I have been wrestling with something fairly personal, and I have come to think a lot of people in the engineering trenches are feeling a version of it.

It is a quiet identity crisis, running underneath all the excitement about agentic coding.

Where the joy actually came from

For years, the joy of programming came from a particular kind of problem-solving. The granular crafting of logic. The satisfaction of a function that finally comes out clean.

That was not incidental to the job. For a lot of us it was the job — the reason we kept doing it after the novelty wore off, the thing that made a hard week feel worth it.

As agentic systems move into the workflow, that dopamine regulation mechanism shifts. The source of professional meaning moves somewhere else, and it moves without anyone announcing it.

It hits differently when the how changes this fundamentally. You can be genuinely excited about the capability and still feel something go quiet.

Why this deserves saying out loud

I think this goes unspoken for two reasons, and both are bad ones.

The first is that saying it sounds like resistance. In a conversation where every second post is about velocity gains, "I miss the part that is being automated" reads as failure to adapt. So people do not say it, and then each of them assumes they are the only one.

The second is that it does not fit the frame. This is being discussed almost entirely as a technical and economic shift — tools, costs, throughput. Those are real. But this is a profoundly human transition, and treating it as purely technical means the human part goes unmanaged.

It shows up anyway. It shows up as the senior engineer who is quietly disengaged despite shipping more than ever. It shows up as the resistance that looks like a security objection but is not really one. It shows up in the twenty percent holding back — some of them for good technical reasons, and some for this one, which nobody has given them the language for.

What I think this actually asks of us

Not nostalgia. I am not arguing for less automation, and I do not think the craft was sacred.

But it does demand a deliberate disconnect from old patterns and a fairly courageous reconnection with what this new world asks. Those are two separate acts and most people only attempt the second. You cannot reconnect to something new while still measuring yourself against what used to make the work feel good.

Concretely, the questions worth sitting with:

• What mental models do we need to build psychological safety in a landscape where the ground keeps moving?
• How are those models different from the ones we inherited? Most of what we were taught about mastery assumed the skill ladder stayed still.
• What does the new craft actually consist of? Context engineering, specifying correctness, designing verification — these are real skills with real depth. They are not lesser. They are just unfamiliar, and unfamiliar work does not reward you the way practised work does.

That last point is the one I keep returning to. The satisfaction has not been removed. It has been relocated to a place most of us are not yet good enough at to enjoy. Being bad at something new feels identical to having lost something, and it is not the same thing at all.

If you lead a team

You are probably managing this whether or not you have named it.

The engineers who look like they are resisting the tooling are not all resisting the tooling. Some of them are grieving a version of the job, and there is no forum where that is an acceptable thing to say. Creating one costs you nothing and it is the difference between people moving through this and people quietly deciding they have aged out.

The other half of the job is making sure the new craft is visible and legitimate. If the only thing your organisation celebrates is throughput, you have accidentally told everyone that the part they used to be proud of no longer counts, and that the new thing is not a craft at all — just faster typing.

It is a craft. It is worth saying so.

Where I have landed, provisionally

I am not through this. I am still working out which parts of the old satisfaction I should let go of and which ones were load-bearing.

What I am fairly sure of is that the transition goes better when it is discussed than when it is not, and that the industry is currently discussing every dimension of this except the one that determines whether people stay.

The goal is preserving the human ingenuity and the passion that drove this industry, while the tools underneath it change completely. That is not a soft objective. It is the one that decides whether any of the rest of it works.

---

This is the part of agentic transformation that gets left out of the rollout plan, and it is usually the part that decides the outcome. Tell me where your team sits.