01 Meaning 9 min read
Meaning Was Never the Job
If usefulness goes, does meaning go with it — or was the link always a historical accident?
The fear underneath every conversation about artificial intelligence is not really about jobs. It is about mattering. If a machine can write the report, diagnose the patient, draft the contract and compose the score, then the question that arrives at three in the morning is not how will I earn but what am I for.
That question deserves a better answer than reassurance. It also deserves a better answer than despair. The honest answer begins by noticing that the link between meaning and economic usefulness is much younger, and much weaker, than it feels.
The bundle
For most of human history, what you did to survive and what made your life significant were not the same thing and nobody expected them to be. Meaning came from kinship, ritual, land, faith, war, craft, story. Work was mostly what you had to get through. The idea that your occupation should be the primary source of your identity — that “what do you do?” is the correct first question to ask a stranger — is roughly two hundred years old. It arrives with industrial capitalism, gets a theological chassis from the Protestant work ethic, and hardens in the twentieth century into something close to a moral law.
That bundling was enormously productive. It was also contingent. And it is now being unbundled by force, which is a painful way to discover that two things were only ever taped together.
What the evidence actually says
The strongest data point is nearly a century old and still the most clarifying.
In 1930 the Austrian village of Marienthal lost its textile factory, and with it essentially all employment. Marie Jahoda and her colleagues went to live there and document what happened. They expected radicalisation. What they found was worse: apathy. People stopped reading the library books that were free. They stopped attending the political clubs they had built. Most eerily, they lost time itself — men who had nothing whatsoever to do took longer and longer to walk down a single street, and could not account for their afternoons.
Jahoda’s conclusion became the latent deprivation model: paid work supplies five things beyond money — time structure, social contact, collective purpose, status and identity, and regular activity. Take the job away and the wage is the most replaceable of the losses.
This is why a universal basic income, on its own, is an incomplete answer. It addresses the manifest deprivation and leaves the latent one intact. It is also why the retirement literature is so strange-looking at first glance: voluntary retirement into a dense social life is associated with improved wellbeing, while involuntary job loss at the same age is associated with sharp declines. Same amount of free time. Opposite outcomes. The variable is not idleness. It is whether the structure was chosen and whether anything still depends on you.
The three roads to meaning, and how many AI blocks
Viktor Frankl, writing out of the camps, argued that meaning is available by three routes: through creating a work or doing a deed; through encountering something or someone — beauty, nature, another person loved in their particularity; and through the stance one takes toward unavoidable suffering.
Notice the accounting. Advanced AI puts real pressure on the first road. It does essentially nothing to the second and third. If your entire meaning portfolio is allocated to road one, and inside road one it is concentrated in cognitive output that a model can now approximate, you are exposed. That is a portfolio problem, and portfolio problems have portfolio solutions.
That objection is largely correct and should not be waved off. Two things can be true: the meaning–work link is historically contingent, and dismantling it inside a single lifetime will hurt badly. The argument here is not that the transition is painless. It is that the destination is habitable, and that whether we arrive there depends on decisions about structure and status that we are making now, mostly without noticing.
Where meaning sits on the ladder
Meaning is the one thing on this site that cannot be delegated at all — not because machines are forbidden from touching it, but because a meaning you did not construct is not yours. A model can generate an excellent essay about your purpose. Reading it will change nothing, in the same way that reading about exercise does not make you strong.
- 1
Cede Machine
The machine does it. You never see the work.
- 2
Approve Machine-led
The machine drafts. You review and sign.
Letting a model narrate your life back to you produces fluent, borrowed meaning that evaporates on contact with difficulty.
- 3
Collaborate Shared
You and the machine work the problem together.
- 4
Critique Human-led
You do the work first, unaided. Then ask the machine to attack it.
- 5
Reserve Human
You do it alone, on purpose, knowing help exists.
Constructing an account of what your life is for — and revising it under pressure — is non-transferable by definition.
Write down, honestly, where your sense of significance currently comes from, and what share each source holds. If more than half sits in professional output, you are concentrated in the one asset class that AI is repricing. Diversification here is not a metaphor.
Time structure, social contact, collective purpose, status, activity. For each, name the specific thing in your week that supplies it. Any latent supplied only by your job is a single point of failure.
Not something you enjoy — something that would visibly get worse if you stopped. A garden, a class, an ageing parent, a club’s accounts, a neighbourhood. Consequence is the ingredient that leisure cannot fake.
Stop asking new people what they do. Ask what they are working on, or what they have changed their mind about. You will notice how hard the old question is to dislodge, which is the point.