Work Without Meaning

AI is not just changing how we work. It is quietly taking away the part of work that made us feel like we mattered.

Nobody is going to tell you this in a company all-hands meeting. Nobody is going to put it in a press release or a LinkedIn thought leadership post. But if you have been sitting at your desk lately, doing your job with an AI tool running in the background, finishing in two hours what used to take you six, and still feeling vaguely hollow about it, then you already know what this piece is about.

The conversation around AI and work has been almost entirely about one thing: jobs. Will AI take your job? Is your role safe? Which industries are most at risk? These are real questions, and they deserve real answers. But they are also the wrong questions, or at least not the only ones. Because what researchers are starting to notice, and what a lot of workers are starting to feel, is that AI is doing something quieter and stranger than taking jobs. It is draining the meaning out of the ones that remain.

And that might be the harder problem.

What the Numbers Say, and What They Don't

Let's start with what everyone keeps talking about, because it matters.

2026 is the year that something shifted. Not dramatically, not in a single moment, but in the way a room slowly gets colder when someone leaves a window open. For years, AI was described as a tool that would augment workers, make them more productive, free them up for the creative and strategic work that machines couldn't do. And for a while, that was largely true.

Then came agentic AI. Not just AI that answers questions, but AI that takes actions, completes tasks, makes decisions, and reports back. AI that doesn't wait to be asked. Researchers at IMD describe 2026 as the year when agentic AI begins displacing jobs rather than merely augmenting them, and the displacement, they note, is not going to be uniform. The roles most at risk are white collar, junior to mid level, the exact roles that millions of people have spent years studying and training for.

The World Economic Forum projects that by 2030, 92 million jobs will be displaced and 170 million new ones created, a net positive of 78 million. On a spreadsheet, that sounds fine. In real life, the 92 million displaced roles are happening now and the 170 million new ones are not fully here yet. The gap between those two moments is where a lot of people are currently living.

Workers with strong AI skills are commanding wages up to 56% higher than their peers, according to PwC. That is a striking number. It tells you that the market has already decided there is a new premium skill, and everyone who doesn't have it is already falling behind, whether they know it or not. Only 25% of workers receive any formal AI training from their employers. The other 75% are figuring it out alone, or not figuring it out at all.

The Part Nobody Is Talking About

Here is what I find more interesting than all of that. Here is the part that the workforce reports tend to bury in footnotes.

Harvard Business School researchers, in a piece published at the end of 2025, made an observation that has stuck with me. They described 2026 as the year we would begin to see the second-order effects of AI on work, not just whether people have jobs, but what those jobs feel like. And their example was simple, almost mundane, and somehow devastating.

Customer service. In the past, when you had a problem with a product or service, you called someone. That person helped you. And the act of helping, of being genuinely useful to another human being who needed something, gave that worker meaning. It was a small transaction but a real one. A person helped a person, and something passed between them.

Now that conversation is handled by a chatbot. The employee, if they still have a job at all, does something adjacent, something upstream or downstream, something that involves less direct human contact and more screen management. And what researchers are starting to document is that the meaning that used to come from that direct contact, that simple sense of mattering to someone's day, is harder to find in the new version of the role.

This is the second-order effect. And we have barely begun to reckon with it.

What Meaning Actually Comes From

To understand why this matters so much, it helps to be honest about where the meaning of work actually comes from.

It is not, for most people, the salary. The salary is necessary, but it is not what makes you feel good about your job on a Thursday afternoon. It is not the company mission statement either, despite what a lot of onboarding presentations would like you to believe.

For most people, meaning at work comes from a handful of very specific things. It comes from feeling competent, from doing something hard and doing it well. It comes from connection, from the relationships built with colleagues and the people you serve. It comes from autonomy, from the sense that your judgment and decisions matter. And it comes from contribution, from the visible, tangible sense that your effort made something better for someone else.

AI, deployed badly or thoughtlessly, threatens all four. If the hard parts of your job are now done by a machine, the competence hits are fewer and farther between. If your customer interactions are mediated by a chatbot, the connection is diluted. If your decisions are increasingly suggested or made by an algorithm, the autonomy narrows. And if you can no longer see the direct line between your work and its impact on another person, the contribution becomes abstract and hard to feel.

You still have a job. You just feel less sure why you're there.

The People Who Are Actually Winning Right Now

There is an irony at the center of this story that nobody at a tech conference seems willing to say out loud.

Nvidia CEO Jensen Huang said it, though, and it landed like a small bomb in certain circles: the next millionaires, he suggested, will be plumbers and electricians rather than techies. Not because plumbing and electrical work are glamorous, but because they cannot be automated. They require a physical presence, a human body in a specific place doing a specific thing that a software agent cannot do. They require judgment in chaotic real-world environments that no language model can fully replicate. They are skilled, they are necessary, and suddenly they are scarce in ways that white collar roles increasingly are not.

The trades, for decades talked about as a lesser path, as what you did if you couldn't make it to university, are quietly becoming the most AI-proof careers available. The irony is not subtle.

Meanwhile, the most vulnerable roles are the ones that came with the most status. The junior analyst. The entry-level lawyer. The graduate trainee. The first-rung consulting associate. These roles were always, in part, about learning, about accumulating experience and judgment through years of doing the work. AI is now doing much of that work. Which raises a question nobody has cleanly answered yet: how do you develop the senior judgment you need if you skip the junior experience that builds it?

The Vacuum Nobody Planned For

One of the more quietly alarming findings in recent workforce research is this: workers report saving an average of two hours per day using AI tools. Two hours. That is a significant chunk of a working day, recovered and handed back.

But most companies have no idea what to do with it. There is no plan for the freed time. No guidance on how it should be used. Workers are saving two hours a day and largely spending it on more low-level tasks, or feeling vaguely guilty about the extra capacity, or quietly wondering if they will soon be asked to do the work of two people in the same number of hours they used to do one person's work.

The productivity gain is real. The question of what it is for has barely been asked.

One way to read this is optimistic: companies are sitting on enormous reserves of human time and attention that could be redirected toward the creative, relational, strategic work that AI cannot do well. The work of building culture. The work of genuinely understanding customers. The work of mentoring younger colleagues. The work of making judgment calls in ambiguous situations where there is no clean data to run.

The other way to read it is less comfortable: that most organizations will simply use the efficiency gains to reduce headcount, and the workers who remain will be left in leaner, faster, more automated environments, doing jobs that feel increasingly like oversight of machines rather than work in the fullest sense of the word.

Both things are probably true, in different places, at different speeds.

What You Can Actually Do With This

I want to be careful here not to tip into the genre of "five tips to thrive in the AI economy," because I think that genre is mostly useless and a little insulting. The scale of what is changing is not addressable by a productivity hack.

But there are things worth thinking about, honestly, before the decisions get made for you.

The first is to be clear-eyed about what your job actually consists of. Not the job description, but the actual daily work. How much of it is pattern recognition and information processing? How much of it is judgment in genuinely ambiguous situations? How much of it is human relationship, trust, and the particular way you know how to read a room? The first category is the most vulnerable. The second two are where human value is most durable, for now.

The second is to take seriously the question of what gives your work meaning, and whether that source of meaning is changing. Not as a threat, but as information. If the thing that made your role feel worthwhile is being automated away, that is worth knowing and naming rather than quietly absorbing.

The third is to resist the narrative that says learning to use AI tools is the whole answer. Learning to use AI tools is necessary and you should do it. But the workers who will matter most in the years ahead are not the ones who are best at prompting a model. They are the ones who bring something that the model cannot replicate, judgment, trust, creativity, genuine human connection, and who use AI to amplify that rather than replace it.

The goal is not to become a better AI user. The goal is to become more deeply, irreducibly human at work. Which, in a world increasingly mediated by machines, turns out to be the rarest and most valuable thing you can be.

The Question Worth Sitting With

There is a version of the future where AI handles the tedious and the repetitive, and humans are freed to do the work that is most meaningful, most creative, most genuinely human. People who work on these systems will tell you that is the goal. It might even be true.

But meaning is not a thing you are handed. It is built through effort, difficulty, repetition, relationship, and the long slow accumulation of doing something hard and getting better at it. If AI removes the effort and the difficulty before you've had the chance to build meaning through them, what do you have left?

That is not a question the workforce reports answer. It is not a question that the productivity metrics can reach. It is the question that sits underneath all the numbers, the one that workers are already feeling in their chests at their desks on quiet Tuesday afternoons when the work is done and the day is only half over.

Your job might be safe. But something about it has already changed. And the sooner we start talking honestly about what that is, the better our chances of building something worthwhile on the other side of it.