Home Future of Work Work Readiness Functional + AI Sovereign Capability Build With Us About Contact

THE FUTURE OF WORK

Work has never remained in one form.

The fixed job will remain. But it will no longer contain all useful work.

A LONGER VIEW

Every economic era reorganised capability.

In the agricultural era, work was closely tied to land, family, seasons and local knowledge. People learned by observing and working beside others.

The age of guilds and trades placed apprenticeship and reputation at the centre of economic life. A person was known by the quality of the work.

The industrial era divided production into standardised roles. The modern job emerged: a defined position, a set of tasks, a salary and a place within an organisation.

The information era changed how work was performed, but the job description remained its principal container.

The AI era is beginning to alter that structure again.

WHAT IS CHANGING

A job is a bundle of tasks. AI is beginning to unbundle it.

Some tasks are repetitive. Some depend on analysis. Some require conversation, trust, interpretation or professional responsibility.

AI will reach these tasks at different speeds.

Work that is repeated, rules-based or burdened by avoidable handovers will change first. Other work will be assisted, recombined or elevated.

Jobs will remain, but many will become less fixed. The boundaries between employee, specialist, independent practitioner and project contributor will become more fluid.

THE ECONOMICS BENEATH THE CHANGE

AI lowers the cost of execution. Craft determines whether the work creates value.

A capable individual can already complete work that once required several handovers. A small team can research, design, test and communicate at a level that previously required a larger organisation.

The cost of producing a first answer is falling. The cost of a poor decision is not.

A wrong filing, unsuitable recommendation, careless customer interaction or poorly judged automated action may cost far more than it saves.

As execution becomes less expensive, value moves towards choosing the right problem, understanding its context, maintaining quality and accepting responsibility for the result.

THE ENDURING CAPABILITY

Craft is judgment carried from one problem to the next.

Craft is not limited to traditional trades. It belongs equally to the doctor, accountant, agriculturalist, lawyer, designer, hotel manager, psychologist and logistics professional.

It is the ability to notice what others miss, recognise what good work looks like, understand where something may fail and adapt knowledge to a real situation.

Craft appears in the questions a person learns to ask: What matters here? What is unusual? Who will be affected? What could go wrong? What should remain human? How will we know whether the work succeeded?

AI can generate an answer. Craft determines whether that answer deserves to be used.

HOW CAPABILITY DEEPENS

Experts do not only know more. They see differently.

The expert eye

Formed by responsibility, not information

A clinician sees a signal where another sees a symptom.

An agriculturalist reads soil, season and behaviour together.

An accountant notices the transaction that does not fit.

A designer senses when an idea is technically complete but culturally wrong.

This expert eye is not formed by information alone. It grows through observation, application, failure, correction and repeated responsibility for the quality of the work.

ECONOMIC SIGNAL

Work is being reorganised while execution becomes cheaper

39%

Of workers' core skills expected by employers to change by 2030

170m / 92m

Roles projected to be created / displaced by 2030

>280x

Reduction in the cost of GPT-3.5-level inference between Nov 2022 and Oct 2024

The evidence points to reorganisation rather than a single direction of travel — a net addition of 78 million roles alongside deep churn. At the same time, the cost of producing competent first-pass output is falling sharply. Together, these changes increase the premium on judgment, context and responsibility.

AITH calculations: Net role change = 170m − 92m = +78m. Cost compression = $20.00 / $0.07 per million tokens = 285.7x between November 2022 and October 2024; Stanford AI Index reports the rounded result as more than 280x.

Source: World Economic Forum, Future of Jobs Report 2025; Stanford HAI, AI Index Report 2025.

Data note: WEF figures are surveyed employer expectations, not labour-market census results. The Stanford figure compares models at a specified benchmark level and does not imply equal quality across all tasks.