Design around capability, not headcount
Discussions about artificial intelligence often collapse into a single question: which jobs will it replace? That framing is simple, dramatic and frequently unhelpful. Organisations are better served by asking which parts of work can be improved and which human capabilities become more valuable as a result.
A job is a bundle of tasks, relationships, decisions and responsibilities. Some activities are repetitive and well suited to automation. Others depend on context, empathy, negotiation, professional judgement or accountability. Treating the entire role as one unit obscures the real design opportunity.
Automation and augmentation are choices
AI can summarise information, identify patterns, prepare a first draft or explore alternatives. A person can test assumptions, understand consequences, make trade-offs and own the final decision. The useful system is often the combination, not either part in isolation.
This does not mean avoiding automation. It means automating deliberately. Low-risk, repeatable work can move quickly. High-impact decisions need appropriate review. The boundary should reflect the consequence of being wrong, not enthusiasm for the technology.
Human-centred systems build trust
People are more likely to adopt AI when they understand its role and retain meaningful agency. A hidden system that changes decisions without explanation creates resistance. A transparent assistant that shows its sources, communicates uncertainty and makes review easy can strengthen professional confidence.
Good design also creates feedback. Users should be able to correct an output and improve the workflow. Their expertise is not an obstacle to automation; it is the mechanism through which an AI-enabled process becomes safer and more useful.
The leadership question
Leaders should judge AI investments by more than labour reduction. Does the system improve decision quality? Does it make scarce expertise available to more people? Does it reduce administrative friction and create time for customers, research or problem-solving? Does it help people do work they can be proud of?
The future of work should not be a contest between people and machines. It should be a design challenge: how can technology amplify the judgement, creativity and purpose that organisations already depend on?