Some jobs will disappear because of artificial intelligence. Pretending otherwise does not help anyone.
But neither does imagining every occupation as a single task that can simply be handed from a person to a machine. Most jobs are collections of very different activities, and AI will not affect all of them equally.
The more interesting question is what happens to a job when some of those activities become dramatically easier.
A job is really a bundle of tasks
An engineer investigates, calculates, checks, writes, communicates, negotiates, interprets evidence, makes decisions and accepts responsibility. A lawyer, accountant, project manager or business owner has a similarly varied collection of activities.
AI may become exceptionally good at some of those tasks while remaining dependent on human judgement, accountability, relationships and domain understanding for others.
Employment therefore may not change as a simple contest between a human worker and an AI replacement. Jobs themselves may be reorganised.
What remains may become more valuable
If machines become better at searching, summarising, drafting, calculating and monitoring, the relative value of other human capabilities may increase.
Judgement. Responsibility. Leadership. Creativity. Trust. Communication. Empathy. Experience.
These are not sentimental additions to professional work. In many situations they are the reason clients trust a professional with an important decision in the first place.
The uncomfortable transition
There will still be disruption. Some organisations will need fewer people for particular processes. Some entry-level tasks may change substantially. Some skills that were once valuable may become less scarce.
That deserves serious attention because technological transitions are experienced by individuals, families and communities, not just productivity charts.
The answer should not be to deny technological progress. It should be to help people build the skills required to direct, verify and work alongside increasingly capable systems.
One professional, many agents
There is another side to the employment story that receives less attention. A capable professional supported by several specialist AI agents could eventually possess capabilities that once required an entire team.
A small business might gain research, marketing, operations, analysis and administrative capabilities without building a large corporate structure. A specialist could turn expertise into a business with far less overhead.
AI could therefore remove some jobs while simultaneously lowering the barrier to creating entirely new businesses and services.
The question worth asking
Instead of asking only whether AI will take our jobs, we should ask what kind of work we want humans to spend their lives doing.
If technology can remove repetitive work while amplifying judgement, creativity and entrepreneurship, then the transition can create opportunity as well as disruption.
The people and organisations that learn how to work with intelligent systems will help determine what the next definition of a job becomes.