AI & the Workforce

Start with this

More Valuable Work, Not Less Work

The people side of AI adoption is where the structural evidence is least subtle: talent, skills, redeployment, and what happens to the employment relationship as AI takes on more of the work. PwC’s 2025 Global AI Jobs Barometer puts the wage premium for AI skills at 56%, while Harvard research published in 2025 found that workers targeting high-AI-exposure roles without genuine capability development face a 29% earnings penalty. The same roles, opposite outcomes. The difference is the quality of the capability investment, not access to the tools.

Most of the Boards I meet approach this from the wrong end. The AI business case arrives framed as headcount reduction, the training budget buys tool familiarity and completion certificates, and workforce resistance is assumed rather than tested. The evidence contradicts all three. Workers are readier for AI than the resistance assumption allows, token retraining delays displacement rather than preventing it, and the organisations capturing the largest gains are those that redesigned work around AI rather than dropping tools into unchanged processes.

The position these articles take is that AI’s primary workforce value is releasing the intellectual capital trapped in undifferentiated work, and that the right measure of success is redeployment, not reduction. That claim carries obligations with it: capability development that builds verification and judgement rather than credentials, expertise pipelines that are augmented rather than replaced, and a transition managed as partnership rather than extraction. Not every skill deserves preservation, but the atrophy has to be chosen, not stumbled into.

There is a Boardroom dimension too. The Institute of Directors now positions AI competence as a core NED responsibility, and the divide opening in the workforce has an exact equivalent around the Board table: directors who can evaluate AI strategy independently, and directors who ratify management narratives they cannot assess.

Read this briefing and you should be able to make one judgement with confidence: whether your organisation’s workforce investment is building the capability that earns premiums or collecting the credentials that incur penalties, and whether your Board could currently tell the difference.

Then read this

From the Divide to the Design

Take the articles in the order given. The core sequence opens with the talent bifurcation, because the premium-and-penalty evidence is the sharpest way into the whole subject: it establishes that the workforce question turns on the quality of capability investment, not the quantity of it. From there, the upskilling article sets out how genuine capability is built, segment by segment, and the redeployment dividend answers the question that follows: once AI frees your people’s time, where should that capacity flow? Governing the Redeployment Dividend then shows what happens when that capacity is left unowned: the hours are saved and then reabsorbed, and capturing them is a Board discipline.

The sequence then widens the lens. The three-hour work week is the earliest article here and deliberately speculative: it asks what happens to hours and compensation when the value of human work concentrates into judgement, and what Hollywood’s AI agreements suggest about sharing the benefits. The future of AI expertise then brings the argument back to organisational design: who builds and manages AI-capable teams, and why the questioning mindset matters as much as technical depth.

The further reading extends the argument in several directions: what happens when employees own the most capable AI in the building, how the remaking of the four dimensions of work runs at different speeds, why culture rather than capital determines adoption, where accountability sits when work is delegated to machines, what the oversight labour AI adds costs when no business case has priced it, and why the verification premium means that delegating code to AI consumes expertise rather than dispensing with it, starting with the junior pipeline that produces it. None is required to follow the core sequence; each rewards the reader who wants the edges of the subject.

When the reading is done, Remake holds the mechanisms that turn the thinking into apparatus for a Board agenda, and the questions are the ones I would put to any executive team proposing workforce AI investment. If you can answer them comfortably, you are ahead of most organisations I see.