The new tool gets demoed on a Thursday. It drafts the report in forty seconds. Everyone in the room nods.
One person doesn't say anything. They are doing a quiet calculation that nobody else in the room is doing out loud: if the tool can do it, what am I still here for?
Somewhere else, someone whose job is perfectly safe feels a version of the same question on a Sunday evening. Not because a machine is coming for the work. Because the work was never quite what they hoped it would be, and they aren't sure where else they would find what they were looking for in it.
Those two people are asking the same thing.
Why am I here?
Almost nothing written about AI and the future of work is answering it.
A dividend can replace a paycheck. It cannot replace being needed.
Most of the public conversation about AI job loss is an income conversation. Which jobs go, how fast, and who pays people when they do. Lower prices, shorter working weeks, some kind of citizen dividend. Those are serious questions, and economists are working on them.
But work has never only been a paycheck. It gives us structure: a reason to get up, a shape to the week. It gives us identity: the answer to "what do you do?" It gives us relationships, mastery, status, recognition, purpose, and something quieter than all of those, the experience of being needed.
None of those things belong to work. Work just happens to be where most of us get them today. If work becomes a smaller part of our lives, we lose them unless we have already started building them somewhere else. That preparation can't wait for the jobs to go. It starts while we still have them.
Which means underneath "will AI take my job" sits a bigger question, about agency: the experience that I can make choices, develop capabilities, influence what happens, contribute something of value and make a difference. Agency is one of the foundations that purpose, belonging and meaning grow from. And unlike a job, nobody can automate it away from you, as long as you have somewhere to practice it.
Why the current conversation misses it
Almost every plan for AI is about what machines can do. We are investing extraordinary resources in increasing the capabilities of machines. We are investing very little, deliberately, in increasing the agency of humans.
Part of the reason is that agency gets treated as something people either have or lack. Some people are "self-starters." Some people "take initiative." That is too simple. Agency is built or blocked by the environment around us: health, money, time, relationships, transport, access to technology, and whether there is anywhere nearby where our contribution is wanted.
So the transition will not land evenly. It is worth asking:
- Who gets most of their identity, relationships and purpose from their job right now?
- Who already has a meaningful role outside work, as a carer, a coach, a neighbor people call?
- Who has a community where they can contribute, and who has nowhere?
- Who has the health, money and networks to go and find one?
So the central question becomes: who gets agency in an AI-enabled world, and who doesn't?
If you have tied a lot of your worth to your work, that is not a personal failing. Most of us were taught to. For a long time, "what do you do for a living?" has been the first thing we ask a stranger, and the answer has done a lot of work for us.
Five shifts worth starting now
Agency sits on five foundations. Each one has a shift in it.
For the individual. From "I am what I achieve" to "I can learn, choose, contribute and make a difference." Identity that rests on more than one thing is harder to take away.
For leadership. From leadership as a position to leadership as agency: the capacity to bring yourself and other people together around something that matters, with or without a title.
For community. From community as somewhere I live to community as somewhere I have relationships, something to contribute, and people who would notice if I stopped.
For organizations. From preparing people to perform their current job to developing their capacity to shape what comes next.
For recognition. From "what do you do for a living?" to "what are you learning, creating, caring for, changing and contributing to?"
The second and third are where most people can start this month. They are also the ones almost nobody is being taught.
Learn to build community, and you have something no rollout can take away
The skills that move change are not taught in most jobs. They come from community organizing: listening, bringing people together, turning an idea into something people can actually do, and then helping someone else lead it. People who learn them in a neighborhood or a school or a park end up far more effective at work too. The reverse is rarely true.
Each step below is small on purpose.
1. Pick one thing near you that bothers you. Instead of: waiting until you have the time, the title or the perfect cause. Try: the crossing outside the school, the empty lot, the neighbor who hasn't been out since the winter. Something you can see from where you already are.
2. Listen before you propose. Instead of: arriving with the solution. Try: asking five people what they notice about it, what they have tried, and what they wish someone would do. You are looking for the people who already care, not trying to convince the ones who don't.
3. Bring two or three of them together. Instead of: calling a public meeting and waiting to see who turns up. Try: a coffee with the two or three who lit up when you asked. A small group is where people find out they are not the only one, and it is the thing that keeps an effort alive between good intentions.
4. Turn the idea into a small experiment with a date. Instead of: a plan for the whole problem. Try: one Saturday morning, one walk, one table outside the library. Small, social and scheduled. People say yes to something they can picture.
5. Make the contribution visible. Instead of: a quiet thank-you at the end. Try: a photo, a count, a few words about what each person brought. People keep coming back to things where they can see they were needed.
6. Help someone else lead the next one. Instead of: becoming the person it all depends on. Try: asking the most engaged person in the group to run the next one with you, then without you. A leader who creates more leaders is the whole point.
That is where this playbook stops. If the thing that bothers you turns out to need a policy changed, grassroots advocacy has its own sequence, and it starts from exactly this kind of group.
A person who can do these six things has a source of purpose that does not depend on an employer. Whether work shrinks, disappears, or simply turns out not to be what they hoped, they know how to find the people who care about something and get it moving.
If you lead people through AI at work
The same skills, pointed inward. If you are the one rolling out the tool, you have more influence over whether people keep their agency than any policy paper.
- Involve people in deciding how AI changes their work. Not a survey afterwards. A seat in the decision. People protect what they helped design.
- Use AI to remove drudgery, and protect the parts that build mastery. Ask your team which tasks make them better at their job, and keep those human on purpose.
- Give people a change to lead. Every rollout is full of small problems nobody owns. Handing one to someone is a leadership opportunity most organizations waste.
- Treat transitions as something you support, not something that happens to individuals. If roles change, how you handle it tells everyone still there what they are worth.
- Build bridges to the community. The facilitation, listening and problem-solving your people practice at work are exactly what their neighborhoods need. Make it easy for them to use those skills in both places.
If your team is already pushing back on the tool, start with what their resistance is protecting. If you are building the rollout itself, the people-first adoption framework covers the steps.
A place to think it through
Starting something on your street and helping a team through a new tool at work have one thing in common: you are leading change you don't control. If you would like to talk through what you want to start, and who you need with you, before you begin, the Real Change Partner is built for that conversation. Three sessions free at realchangepartner.com.
We don't need to know when
Nobody knows whether a largely post-work society arrives in ten years, thirty years or at all. We don't need to know.
Broader identities, stronger relationships, somewhere to contribute, communities people feel part of: all of that is worth having now. Preparing for an uncertain future turns out to improve the present.
AI should not only expand what machines can do. It should expand what people are able, and invited, to do with their lives.
So, not what do you do for a living. What are you learning, creating, caring for, changing and contributing to right now?
And who near you would notice if you started?