The organizations getting real value from artificial intelligence are not the ones spending the most on upskilling. They are the ones that redesigned the work itself before deciding what the machine should do. That distinction — between a training gap and a design gap — is the central lesson emerging from HR leaders, workforce researchers, and leadership-development studies, and it can be organized into four actionable principles: redesign work before deploying AI, shift from static roles to dynamic skills, prioritize uniquely human capabilities, and cultivate an adaptive culture.
The AI Era: A Design Gap, Not Just a Training Gap
Expert commentary on AI adoption converges on an uncomfortable observation: many organizations have treated adoption as an end in itself, leaving AI's productivity promise largely unmet — a point Katie Williams, People and Culture Leader at Sky New Zealand, has made about companies that pursue the technology without rethinking what surrounds it. The response, according to reporting on organizational readiness, is a different kind of investment than most leaders are making. Workforce design is not a step that follows AI adoption; as one analysis of the readiness gap puts it, it is the step that makes AI adoption worth doing.
This reframing matters for how budgets and attention get allocated. If the problem is a skills shortage, the answer is more courses. If the problem is structural — jobs, roles, and workflows built for a pre-AI era — then no volume of training closes it. The organizations closing the gap share a common discipline: they designed the work first, then matched the machine to that design.
