What Are You Actually Working For?

The rules kept changing, and at some point we stopped being surprised. What we didn't expect was the question underneath.

For most of my adult life, the path through a career felt like it had some shape to it. Learn a skill. Become useful. Get better. That particular sequence might not lead anywhere glamorous, but it felt stable in a way that mattered. The ground stayed where you put your feet.

Then, sometime in the last few years, I started noticing a different kind of conversation. Not about job titles or promotions. Something quieter and harder to pin down — people asking, in different ways, the same question: what is any of this for?

The question isn't just tactical. It isn't "what skill should I learn next?" It's more unsettling than that. It's "what does it mean to be good at something, to build something, to offer something — when the thing I built last year might be obsolete, and the thing I'd build next year might not need me in it?"

I don't think that's an anxious question. I think it's an honest one.

Why the Question Gets Existential

There's a reason AI uncertainty doesn't just produce tactical anxiety — "should I learn Python?" — but a deeper unsettledness about identity and purpose. It's because earlier waves of disruption mostly eliminated categories of work. That's painful, but it has a shape: this job is gone; here is a different job; go learn it.

What feels different now is that the uncertainty is harder to locate. It isn't one kind of work disappearing; it's the nature of contribution itself that's being renegotiated. If I write well, and a model can write well, then my writing is still good — but it may no longer be necessary. Those two things — quality and necessity — used to travel together. Now they don't always.

That gap is what produces existential questions. Not "I don't know what job to have" but "I don't know what it means to be useful in the way I thought I was."

The Stoics had a useful framework for situations like this: the dichotomy of control. Epictetus put it plainly: some things are in our power, and some things are not. Our opinions, desires, and the choices we make about how to act — these are in our power. The market for our skills, what technologies get built, which companies rise — these are not.

That distinction doesn't make the uncertainty less real. But it changes where you put your effort.

Meaning as the Durable Asset

Here's an observation I keep returning to: the things I've found most worth doing in my work have almost always been the things I'd do even if they weren't going to be "mine" in the end.

Not because I'm selfless — I'm not particularly. But because the experience of doing them was its own reward in a way I couldn't fully fake. The clarity that comes from explaining something difficult until it's actually clear. The feeling of getting a system right, not just functional. The moment in a conversation when something lands and you can see it in the other person's expression.

These are experiences of meaning. And they're not particularly amenable to automation — not because they're technically hard, but because they're not separable from the act of a human being doing them. A model can produce a clear explanation. It can't experience the process of becoming clearer. A model can optimize a system. It cannot have the satisfaction of getting it right.

Viktor Frankl's central insight from Man's Search for Meaning is more useful here than any career advice: meaning isn't found in outcomes but in the act of engaging. That sounds like a platitude until you're in a situation where your outcomes are uncertain and your engagement is the only thing you actually have.

This is not a counsel of resignation. It's not "don't care about outcomes." It's something more like: build a practice that you can inhabit regardless of how the landscape shifts. The practice is yours. The landscape is not.

What You Actually Control

When I try to sort through the work I do and ask "what of this is genuinely mine?" I find the answer is always in the verbs, not the outputs.

Not "I write" but "I notice what's being left out of a conversation." Not "I build software" but "I hold complexity in mind long enough to simplify it for someone else." Not "I manage people" but "I ask the question that breaks through the story someone's been telling themselves."

Those verbs are harder to outsource than the nouns. They're also harder to describe on a resume, which is probably not a coincidence.

The reframe I find most clarifying: a career is not a product. It's a posture. The product might become less valuable as AI improves. The posture — curiosity, care, the willingness to engage with what's actually hard — doesn't have a depreciation schedule.

Seneca wrote about how the wise person treats circumstances as material to work with, not as the source of their identity. "What does it matter how much is entrusted to me, if I work well?" The work can change. The practice of working well is yours to keep.

Finding What No Model Can Imitate

There's a practical exercise I've found worth doing. Not to future-proof your job description — that's mostly noise — but to find the parts of your work that feel most like you, in the sense that removing you would remove something real.

It works like this. Take a week's worth of work — not the polished parts, the actual messy middle. The conversations, the decisions, the small adjustments you made that you can barely remember. Ask, for each piece: what was the human act here? Not the output. The act.

Was it holding two contradictory ideas in suspension long enough to find a third? Was it reading someone's hesitation and adjusting? Was it caring more than was strictly required? Was it bringing a particular perspective that changed the shape of a problem?

The things that keep showing up in that inventory — those are the signal. That's what you're actually bringing. That's worth developing and protecting and doubling down on, regardless of what models can or can't do next year.

This isn't about outsmarting AI. It's about understanding yourself well enough to do the work that only you can do — and getting better at it.

A Career Is Not All of a Life

One more thing worth saying: the question "what am I working for?" sometimes gets answered not by changing your work but by widening the frame.

In a yogic sense, the work of a life isn't primarily economic. The outer career is one expression of something larger: a disposition to contribute, to be useful, to give what you have. That disposition can find expression in many forms, some of which don't have a job title.

I find I need that reminder on the weeks when the economic uncertainty feels loudest. The part of the work that matters most might not be the part that's most at risk.

The question "what are you actually working for?" is worth taking seriously. Not because it has a clean answer, but because sitting with it honestly — without deflecting into tactics — tends to surface something true about what you value and what you'd protect if the circumstances changed.

That's worth knowing, regardless of what happens next.

FAQ

Is it worth investing in skill development if AI might make those skills obsolete?

The skills most worth developing are tied to judgment, relationship, and meaning-making — things that are hard to automate not because they require technical difficulty but because they require a human doing them. Narrowly technical skills are more volatile; broad human capacities compound over time. Invest in both, but don't conflate them.

How do I find meaning in work that feels purely mechanical?

Start by asking what the mechanical work enables — not what it produces, but who it allows you to be or serve. Often meaning isn't in the task itself but in what the task is connected to. If that connection is genuinely absent, that's real information about whether the work is worth continuing.

What is the Stoic dichotomy of control and how does it apply here?

Epictetus identified a hard line between what is "up to us" — our judgments, desires, and actions — and what is not — market conditions, other people's choices, technological change. The Stoic move is to put serious effort only into the first category while accepting the second with equanimity. Applied to AI: you can't control what models get built, but you can control how deeply you develop your distinctly human capacities.

Can meaning actually function as a career strategy?

Not in the narrow sense of "meaning will protect your job." In the wider sense: people who have a strong internal reason for their work tend to adapt better to disruption because they know what they're trying to preserve. That coherence is both more satisfying and, practically, more resilient than chasing whichever skill is highest value in any given year.

What if I genuinely don't know what I value in my work?

That's a real answer, not a failure. The exercise described here — reviewing a week's work for the human acts, not the outputs — is a starting point. Most people find the thread if they look with enough honesty. If the thread is absent, that too is worth knowing.