AI leaders are asking what humans will do when more jobs can be automated.
It is a fair question. Demis Hassabis recently looked ahead to a future in which AGI has arrived, AI has helped compress drug discovery from years to weeks, and the remaining items on the list include “more philosophical topics” around “what it means to be human.”
So the rough sequence is: build AGI, cure all disease, solve human meaning.
Hard to say if that is how the next few years will go. Still. Quite a list.
But work has not felt like purpose and meaning for many people for a while now.
So maybe losing the jobs is not what creates the crisis. Maybe it is what exposes it. Or, less bleakly, maybe it is the first chance many people get to reconnect with the things that mattered before the economy taught them to call margin “impact.”
There is a particular emptiness that arrives when the day has been full, everything felt urgent, and still you cannot quite say who was better off because you showed up.
You did the work. You protected a margin. You helped the business.
But that is not the same as helping a person.
Useful work is not hard to understand. A child learns to read. A patient gets care. A pipe stops leaking. Food reaches a shelf. You know useful work by what breaks when it stops.
Covid made this clear for a while. We called people “essential workers,” which was a polite way of admitting that some jobs mattered more than we had been paying them. Society did not survive because someone improved a brand deck from a spare bedroom. It survived because some people still had to show up.
Then the moment passed, because learning from it would have been expensive.
The old work came back. The full day. The tired body. The odd shame of being worn down by something that, if described honestly, did not sound important enough to hurt you.
The problem is not hard work. Meaningful work is often hard. The problem is work that is demanding and serious, but not clearly linked to making anyone’s life better.
That is where the word “value” starts to matter.
A business can create value by making something better for people. It can also create value for itself by finding a better way to charge people for something they need. The economy often treats these as the same thing.
They are not the same thing.
A home can be made warmer and safer. That is value. But the same home can also become a better asset while the person inside learns to live with damp. That is value too, at least to the owner.
A care service can give someone time, attention and dignity. That is value. But it can also work out how little dignity can be given before the contract fails. That is value too, somewhere in the accounts.
Software can remove a problem from someone’s day. That is value. But it can also turn a solution into a subscription, and a subscription into a trap. That is value too, if the renewal rate looks good.
This is how the economy gets away with so much.
The mould is still there. The care visit is still short. The support is still useless. But somewhere, the graph looks better.
Some jobs feel valuable and empty at the same time because they are useful to something. The question is whether that thing is a person, or a number that improves when a person has fewer choices.
A lot of people know the other side of this. You pay more. You wait longer. You get less. The system apologises and does nothing.
Then you go to work and find the same logic there. A price rise becomes a strategy. A trapped customer becomes retention. A refusal becomes policy.
That is the trap. People need salaries. They have rent to pay. So they end up inside an economy that asks them to make extraction smoother, cheaper, and easier to defend, while calling the result work.
You might hope the public sector would be different.
That should be the place where society protects real value: decent homes, good schools, health, transport, care, clean air, help before a crisis becomes a disaster.
Instead, people pay high taxes, watch services get worse, and are told there is no money.
There is money. Just not usually when it would stop ordinary life getting worse. There is money when banks fail, when war arrives, or when a neglected public service becomes an emergency.
Now assume the AI optimists are right. The machines get very good. They make useful things cheaper. They speed up science, medicine, logistics, education, and admin.
Fine.
Now change nothing else.
Keep the same owners. The same landlords. The same platforms. The same debt. The same weak public services. The same governments explaining why the public cannot have nice things.
The machines produce more. The owners capture more. Everyone else gets a better interface.
This is not freedom from work. It is a world where businesses need fewer workers, but people still need wages to live.
Human contact becomes the premium plan. Poor people get automation. Rich people get people. Everyone gets told the system is more personalised now.
That is the real horror. Not that AI fails. That it works, and the gains go exactly where gains already go.
AI is not creating the meaning crisis. It is walking into one already happening, taking notes, and offering to make the paperwork faster.
If machines replace work that helps people, something real may be lost: skill, care, judgement, trust, and the kind of human knowledge that never fits cleanly into a process.
But if machines replace the work of smoothing hostile systems, something else becomes clear.
Not that humans had nothing to offer.
That too much of what they were asked to offer was never worthy of them.
There is no shortage of meaningful work. People need care. Homes need repairing. Children need teaching. Public services need rebuilding. The climate needs attention.
The crisis of meaning is not that AI may leave humans with nothing useful to do.
It is that humans already have plenty worth doing, and the economy keeps paying them to do something else.