Across conference stages from Washington to Davos, powerful tech executives are calmly announcing that paid work, as we know it, is heading for the exit. Yet once microphones turn to the question, “So what replaces it?”, answers quickly grow vague.
From labour shortage to “who needs workers?”
In the United States, a sharp fall in immigration has arrived at the exact moment AI systems are rolling through offices, warehouses and call centres. According to recent Census Bureau figures, net immigration is projected to plunge from 2.7 million people in 2024 to about 320,000 in 2026 if current trends continue.
For most of the last century, such a collapse would have triggered political panic about missing workers, ageing populations and slower growth. This time, a different narrative is gaining traction among Silicon Valley elites: maybe the economy will not need as many workers anyway.
In boardrooms and on panels, a new idea is spreading: AI will do the jobs immigrants once filled, and then some.
That stance dovetails neatly with the Trump administration’s restrictionist approach to immigration, which has favoured headline job numbers over long‑term demographic health. Fewer arrivals mask weak employment creation and give space to argue that robots and algorithms will take up the slack.
The Davos doctrine: optional work and robot factories
At the World Economic Forum in Davos, the argument was laid out with unusual bluntness. Alex Karp, chief executive of data analytics giant Palantir, questioned the very premise of large‑scale migration in a world of advanced automation. Routine jobs in retail, logistics, clerical work and even journalism, he suggested, are being targeted by increasingly capable AI systems.
In that world view, immigration should be limited to those with “very specialised” skills that machines cannot yet replicate. Everyone else, the implication goes, is on borrowed time in the workforce.
Elon Musk has pushed the idea furthest. The Tesla and SpaceX boss has said repeatedly that artificial intelligence and humanoid robots will “replace all jobs” and make work a lifestyle choice, not an economic necessity. At one point he likened future employment to growing your own vegetables even though supermarkets exist: a hobby, not a requirement.
His companies are aligning with that script. Parts of a California Tesla plant are being converted to produce humanoid robots designed to handle warehouse tasks, factory work and eventually more delicate operations in care and services.
➡️ Cancer Uses Tiny Bubbles To Prepare Its Spread Through The Body
➡️ Toilet debate settled: should the seat stay up or down and what hygiene experts actually recommend
➡️ In Bulgaria, a strange rock stumbled upon in a forest may be the first star map in human history
➡️ Stop washing your hair this often dermatologist warns we have been doing it all wrong
➡️ 14 Yoga Poses That Help Open Tight Hips and Improve Mobility
➡️ Goodbye Balayage: The New Technique That Eliminates Grey Hair for Good
For Musk, humans will keep working because they want purpose or fun, not because they need a pay cheque to survive.
Bill Gates, less theatrical but equally radical in his predictions, has argued that AI will trigger a transformation even deeper than the personal computer revolution he helped launch. Software agents, he says, will let societies produce “far more goods and services with far less labour”, potentially reshaping everything from taxation to healthcare.
The missing piece: a believable life after full‑time jobs
Where these visions falter is in the “what next?” question. If AI wipes out a large share of existing jobs, who pays for people’s lives, and on what political terms?
Pressed on such questions, tech leaders tend to gesture toward vague concepts: universal basic income, “abundance”, new creative roles. None of those are detailed policy plans. They are more like slogans, offered in passing before the conversation returns to product demos.
The grand promise is a life of leisure; the concrete roadmap is an empty slide.
That gap matters. Economists warn that the transition between today’s labour‑centric capitalism and any post‑work arrangement could be chaotic. Without buffers, rapid automation could deepen inequality, polarise politics and hollow out communities that rely on stable, if modestly paid, jobs.
Which jobs are on the firing line first?
While “all jobs” makes for a dramatic quote, the impact of AI will not be evenly spread. Some roles are already under heavy pressure, others remain relatively safe, at least in the medium term.
- High risk: cashiers, telemarketers, basic data entry clerks, warehouse pickers, routine legal assistants.
- Medium risk: accountants, paralegals, junior journalists, graphic designers, call‑centre agents.
- Lower near‑term risk: nurses, electricians, plumbers, teachers of young children, care workers.
AI systems excel at pattern recognition, language generation and repetitive decision‑making. That makes them strong substitutes for office tasks and back‑office functions long thought “safe” compared with factory work.
Physical jobs involving unpredictability, social nuance or intimate care are harder to automate fully, although robotics is advancing rapidly. The rise of humanoid systems in factories, like those Musk is betting on, signals that even these niches may erode over time.
Immigration policy collides with the AI transition
Against this backdrop, slashing immigration amounts to running a huge social experiment with little preparation. Ageing societies like the US and much of Europe have depended on younger migrants to sustain pension systems, staff hospitals and care homes, and keep construction projects moving.
If fewer migrants arrive just as AI displaces domestic workers, there is a risk of awkward mismatches. Hospitals might still struggle to find nurses while chatbots replace receptionists. Logistics firms could lack drivers even as algorithms take over scheduling and management.
| Area | Role migrants often fill | AI/robot trend |
|---|---|---|
| Healthcare | Nurses, care assistants | Automation of admin tasks, not bedside care |
| Construction | Labourers, specialists | Early-stage robotics, 3D printing of components |
| Retail & services | Cashiers, cleaners, kitchen staff | Self‑checkout, cleaning robots, automated kitchens |
| Logistics | Drivers, packers | Warehouse robotics, progress towards self‑driving |
For governments, the twin challenges are timing and coordination. Restricting migration while betting on automation assumes new technologies will arrive exactly where and when they are needed. History suggests that rollouts are bumpy, uneven and shaped by corporate incentives rather than social ones.
Who owns the machines, and who gets the gains?
Another unresolved question sits at the heart of the “end of work” pitch: if AI and robots generate most of the value, who owns them? Today, the answer is simple: mostly a small group of large firms and their investors.
Unless that structure changes, a heavily automated economy could funnel an even larger share of income to capital owners while sidelining wages. Politicians and policy thinkers are already floating ideas to counterbalance that trend:
- Taxing robots or highly automated firms to fund social benefits.
- Giving citizens “data dividends” based on the information used to train AI models.
- Encouraging worker ownership stakes in companies deploying automation.
- Expanding public services – healthcare, housing, education – so people rely less on wages alone.
None of these ideas is straightforward. Robot taxes might slow innovation; data dividends are hard to measure; broader public ownership can run into governance problems. Yet without some version of them, the post‑work future looks less like utopian leisure and more like a small AI‑owning elite surrounded by insecure gig work.
What a post‑work week could actually look like
Beyond economic design, there is a cultural question: how do humans structure their days when paid labour is no longer the default adult activity?
One plausible scenario does not eliminate work so much as change its shape. Instead of a 40‑hour job, people might cycle between short bursts of project‑based paid tasks and longer periods of unpaid but socially valuable activities: caring for relatives, volunteering, creative work, community projects.
Governments could support that model with shorter standard work weeks, strong safety nets and retraining schemes that treat mid‑life career shifts as normal. Local authorities might fund community labs or workshops where people use AI tools for art, education or neighbourhood problem‑solving rather than strictly commercial output.
A post‑work society is not a society without effort; it is one where effort is less tightly chained to survival.
Terms that shape the debate
Universal basic income
Universal basic income, or UBI, refers to a regular cash payment made to every resident, regardless of employment status or wealth. Supporters see it as a simple way to cushion people against automation shocks and give them freedom to refuse exploitative jobs. Critics worry about cost, inflation and political backlash from taxpayers who feel they are funding others’ leisure.
Technological unemployment
Technological unemployment describes job loss caused by new technology. The phrase is old: economists used it during the mechanisation of farming and the rise of factories. Historically, new sectors eventually emerged to absorb displaced workers. The fear this time is that AI might erode both manual and cognitive work at once, shrinking the space for humans to reinvent themselves.
Practical risks and opportunities for workers now
For today’s workers, the debate about a distant post‑work future can feel abstract. Yet choices made in the next five to ten years will shape how painful the transition becomes.
On the risk side, people in repetitive office roles face creeping automation without yet seeing clear new career paths. Whole regions reliant on logistics hubs or basic manufacturing lines may be exposed to sudden job cuts when firms install robots or automated software.
There are also opportunities. Workers who learn to manage, supervise or complement AI systems can increase their value. Skills in care, education, trades and hands‑on maintenance – often underpaid today – could become more central if societies realise that not everything meaningful can or should be automated.
What remains missing is an honest public conversation that matches the ambition of the tech forecasts. If the end of work is really on the horizon, leaving the details to a handful of CEOs and quarterly earnings calls will not be enough.