The Fear Is About Identity as Much as Employment

Ask an engineer what worries them about AI and the answer is usually about jobs. Push a little and something else surfaces: the thing they're good at, that took years to become good at, that they partly are — a machine now does a version of it in seconds.

That's a different problem from employment, it responds to different things, and conflating the two makes both harder to think about clearly.

Two distinct worries

"Will I be able to earn a living?" An economic question. It depends on demand for software, on which tasks are exposed, on your organization's situation. It's addressed by positioning, by skills, by which employer you're at.

"Is the thing I'm good at still worth being good at?" An identity question. Someone who took a decade to become genuinely excellent at writing clean, correct, elegant code is watching that specific excellence become less scarce. Even in a scenario where employment is fine and the pay is fine, something real has changed.

The second one is what makes the conversation emotionally sharp in a way that pure economics wouldn't be. And it's usually argued as though it were the first, because that's the version that's socially acceptable to say out loud.

Why craft-based identity is particularly exposed

Engineering attracts people who like the feeling of getting something exactly right. The satisfaction is in the doing — the clean abstraction, the elegant solution, the bug tracked to its root.

That satisfaction is tied to personally producing the artifact. When production is delegated, the satisfaction doesn't automatically transfer to review and direction, even when the outcomes are better. Directing work you'd rather be doing yourself is a real loss, and telling someone the outcome is the same misses the point.

⚠️ Worth saying plainly: this is not irrational resistance to change. The thing they valued genuinely was valuable, and it genuinely is less scarce. Being told to reframe it as an opportunity, by someone who isn't feeling it, is not useful.

What's actually durable in the craft

Some of what feels threatened isn't:

Knowing what "right" looks like. The taste that makes a solution elegant rather than merely working — that's what's needed to evaluate generated output, and it's the scarce input now. Judgment developed through years of doing is the thing that transfers most directly.

Understanding systems deeply. Holding a real model of how something works, being able to reason about it under unusual conditions. That's what debugging hard problems requires, and it isn't produced by reading summaries.

Knowing what's worth building. Never the automated part, and it's the direction the whole role is moving.

The satisfaction of a system working well. Available to whoever caused it, whether or not they typed it. Some people find this transfer easy and some find it genuinely hollow — that difference seems to be about the person, not about who's right.

💡 What seems to help

From how people actually navigate this rather than how they're advised to:

Name it accurately. "I'm worried about money" and "I'm mourning something" are different problems. Confusing them produces advice that doesn't land, and a lot of arguments where the stated topic isn't the real one.

Keep doing some of it. Not everything needs delegating. Keeping some work you do by hand, because you want to, is a reasonable choice and not a productivity failure. Craft you keep deliberately is different from craft taken from you.

Find where the difficulty moved. For many people, the interesting problem migrated rather than vanished — from writing code to designing the systems around agents, from producing to specifying and verifying. Whether that lands as interesting or as a downgrade is individual and worth being honest with yourself about.

Accept that it might be a real loss. ✅ Not everything reframes into an opportunity. If the specific thing you loved is genuinely less valued, saying so plainly is healthier than performing enthusiasm — and it makes the practical decisions clearer, including the possibility that a different role or a different domain fits better.

The takeaway

The employment question and the identity question travel together and answer to different things. The economic one responds to positioning and skills. The identity one responds to naming it honestly, to keeping some of the craft deliberately, and to finding where the difficulty went. Conflating them produces arguments where neither person is discussing what they actually mean — and where the advice, however sound, is aimed at the wrong worry.