How to find meaning at work in the AI era. Or not

Last time I wrote about the grief that comes with being an engineer right now. One of the most common symptoms of that grief is losing your identity, and it usually starts with losing the meaning of what you do. I want to talk about that, partly for the peers going through it, partly because the topic fascinates me.

I’ll go through where meaning can come from, most obvious first. The last section is about not finding any.

Find meaning in the mission

Your company’s mission can carry it. Most companies exist to make money, but plenty of them now also have a mission about the environment, healthcare, education, social justice. If you’re in one of those, the meaning is sitting right there. What makes it real:

• Leadership says the goal out loud, on purpose, more than once a year.
• Somebody measures progress toward it, and those wins get celebrated like revenue wins do.

A mission nobody mentions after the all-hands is just marketing.

One warning: overselling meaning is worse than not selling any. The people who suffer most aren’t the ones who gave up on meaning at work. They’re the ones actively looking and not finding it. Tell a team their app is changing the world and you put them there. If the honest version is “we sell insurance and we’re good at it”, say that. People can work with honesty.

Find meaning in the way you work

Not everybody gets a job where the impact is the point. Someone has to sell the lemons. If that’s you, and you have no power to change the company, you probably took the job because you like some part of the work, or for the money.

So start small. There are moments in your day when you feel good. Write them down for two weeks, one line a day:

• What was I doing?
• Who was I with?
• What time was it?

Do you like giving the demo? Crafting an interface until it feels right? The wiring, the moment the whole system finally clicks? A pattern shows up faster than you’d think, and that pattern is your meaning. You don’t need anything bigger than that.

Then protect it. Psychologists call this job crafting and three axes:

• Task crafting: change what you do. Trade the report nobody reads for the migration you want to own.
• Relational crafting: change who you do it with. Pair with the person who makes the work fun. Get into the room where the decision happens.
• Cognitive crafting: change how you see what you already do. Nothing in your job description moves. You just stop treating the boring 40% as the job.

The third one is free, which is why I like it, and it works best in the rigid roles where the other two are impossible.

I’ll be honest about how well it works, though. The best trial I found on reframing didn’t make anybody happier. It kept them steady while the group that did nothing slid backwards. Still worth doing. Meaning leaks by default, and holding the line is a real result. Just don’t expect fireworks from a journaling habit.

We’ve all been told automation eats the boring work and leaves us the good stuff. A study presented at CHI this year mapped computer-based tasks across the whole US workforce and found the opposite. The tasks most exposed to AI are the ones workers describe as creative, autonomous and enjoyable. The safest ones involve reading a room, sitting with someone, building trust.

So the thing that makes your day is the first thing you’ll give away without noticing. Push the boilerplate to the model and keep the debugging you enjoy. Once a tool has owned a kind of problem for six months, you won’t want it back, and you might not be good at it anymore.

It can go the other way too. Developers interviewed about Copilot and ChatGPT said they spent the time they got back on the creative problems they’d been too busy for. Same tools, opposite result. The difference is whether you’re choosing or being chosen for. Ask for the choice. It’s usually available and almost nobody asks.

Maybe there’s none

This is what makes it fascinating to me. Looking for meaning at work is a privilege. Having the time to ask the question is a privilege too. It doesn’t feel like one, because work eats the day and we need it to pay the bills, but it is. You won’t always land a job with meaning in it. Sometimes there won’t be any for a while. That’s okay.

Buddhism and Taoism will tell you the chase itself is the mistake. Viktor Frankl, a psychiatrist who survived Auschwitz and wrote Man’s Search for Meaning, is the person most associated with the idea that meaning matters more to us than pleasure or power. He also said you can’t go and get it. It shows up as a side effect of living and doing what you care about. The harder you look, the less you find.

There’s a therapy called ACT, acceptance and commitment therapy, with a move I find useful here: accept that part of the job is meaningless instead of fighting it. Fighting is what hurts. Someone who can sit through a dull sprint without spiralling has energy left for the work that matters. That’s also why the approaches that help are about your values and your attention. The ones that fail try to convince you your job matters when you can see it doesn’t.

What to actually try

If you take one thing from this article, take something from here.

On your own, this week

• Reframe before you restructure. Cognitive crafting costs nothing and needs nobody’s approval. Take a task you hate and connect it to a person it helps.
• Write down your three work values, then open your calendar. Values you never act on are where the emptiness comes from.
• Keep the joy log. Two weeks, one line a day.

If you run a team

• Find the work that needs judgment: the ambiguous calls, the ethics, the conversations that need a human in them. Protect it from automation. It looks like inefficiency on a dashboard and it’s the part that keeps the job worth having.
• Make crafting possible. It needs psychological safety, a manager who says yes, and clear edges on what people are allowed to change. Without those, “craft your job” is a poster.
• Diagnose before you fix. Someone who feels nothing and isn’t looking needs something different from someone searching hard and coming up empty. There’s a ten-question survey for this if you want it on paper instead of in a 1:1 (WAMI).
• Ask everyone, once a quarter: which part of your job would you hate to lose to a tool? Then don’t automate that one first.

If you set the culture

• Design the job, don’t buy the perks. Real responsibility, craft and growth hold up for years. Snacks and a ping pong table don’t.
• Don’t inflate. Framing every role as world-changing manufactures the searching-and-not-finding state that hurts people most. Say what’s meaningful about a role and what isn’t.
• Don’t assume your framing travels. “Find your authentic self through your work” is a very Western idea. In plenty of cultures meaning comes from your role toward others, from duty, from getting slightly better every day. If your team spans continents, ask instead of exporting.

Not everything is meaning

Meaning is good, but your job doesn’t have to carry it. You can change careers. You can get laid off. Build meaning in your life outside work first: that’s the part no reorg can touch.

And meaning isn’t job satisfaction. You can be happy at work because the people are good company. You can enjoy a task done well. Hard work feels great when you’re getting better at it and going somewhere. There are a lot of ways to have a good job, and each of us has to find the ones that fit.

The shift happening in our daily work with AI is sometimes overwhelming. It’s on us, as people and as leaders, to build places we can stand to be in. Aristotle had a word for what’s left once the machine takes the execution: phronesis. Practical wisdom. Seeing what a situation asks for and doing it well. Taste, judgment, knowing which problem is worth solving at all.

A model can write the code. It can’t tell you which part of your day made you feel alive, and it can’t decide what’s worth doing. That’s still your job.

If you want the papers behind this: “Are We Automating the Joy Out of Work?” (CHI 2026) for the automation finding, “Code and Craft” for the developer interviews, and this trial for the reframing result.