Reframing the health workforce as a hybrid human–AI capability asset: the proposal of Workforce Science

It proposes Workforce Science as a discipline for measuring, analysing and stewarding health and care capability, treating the health workforce as a strategic national capability asset rather than merely a human resource or manpower planning object. The article moves the policy discussion from manpower or human resource concerns and health labour market supply-and-demand framing toward a stewardship framing in which public policy governs complex relationships among people, organisations and technologies to achieve equity in population health. This is a conceptual argument in an Opinion article, grounded in the evolution of health workforce policy and economic thinking since 2000 (the World Health Report 2006, Workforce 2030, health labour market analysis, and the theoretical lines of Becker, Sen, Nussbaum and Bueno) rather than in newly collected empirical data.

It introduces the construct of effective capability (C*), understood as a function of education financing (Ef), education quality (Eq), labour (L), the Human–AI dynamic (H-AI), governance capacity (G), and the conditions, rights and realities of work (R). The article states that capability can no longer be proxied by employment of licensed health professionals, their full-time equivalents, or density per 10,000 population, and instead expands earlier workforce planning logic on availability, accessibility, acceptability and quality by adding analytical layers on human–AI capability, governance and decision integrity, and the conditions of work. The construct is presented in Table 1 of the article and is a conceptual framework-building effort; it rests on an argument extending existing workforce planning models, and the article reports no empirical measurement testing that function.

It proposes that Workforce Science progresses through five layers: defining what exists, determining what is measured, how evidence is analysed, how evidence informs decisions, and how those decisions are translated into practice. The article states that this definition → measurement → analysis → governance → stewardship progression distinguishes Workforce Science from traditional workforce approaches by making lexical, ontological, numerical, analytical and decision disciplines explicit and cumulative, and that it enables structured appraisal of evidence in published and grey literature. The five-layer structure is presented in Table 2 and is a conceptual methodological architecture; the article also claims utility for health journal editorial boards, academic commissions and government reviews on how AI will affect the future health and care workforce, but provides no application cases or validation data.

It frames the object of analysis as the Human–AI Economy of Health and Care, holding that care capability increasingly emerges through complex interactions of workers, workforce teams and intelligent technologies, so the stewardship question shifts from how many workers to educate and employ to how hybrid human–AI capability is created, distributed and stewarded. The article explicitly resists reducing this change to a debate on task automation and the substitution or replacement of humans with AI, arguing instead for a higher degree of scientific enquiry and, in the health setting, greater attention to epistemic resilience. This is the article's central normative claim, built on the argument that labour in ISIC Sector Q (Human Health and Social Work Activities), predominantly female, has far greater added value for societal progress and national wellbeing frameworks, supported by cited literature; it is an opinion argument rather than an empirical test.

The article positions Workforce Science as a next stage in the evolution of the development and management of human resources for health, aimed at governments and all relevant stakeholders, and intended for the setting of the Human–AI Economy of Health and Care, where capability is co-produced by workers, workforce teams and intelligent technologies across homes, communities, primary care, hospitals and virtual environments. It proposes that the discipline can integrate complementary approaches on AI, demography, epidemiology, labour economics, statistical systems and workforce planning, and can be used for structured appraisal of published and grey literature as well as by health journal editorial boards, academic commissions and government reviews on how AI will affect the future health and care workforce. The article also states that existing workforce and research frameworks remain indispensable, having been developed primarily to understand units of labour within analogue health systems.

The text loaded here is a version without the figure and table content, so the specific entries of Table 1 (the C* construct) and Table 2 (the five-layer path) are not available and cannot be restated here; this is an open point for readers to check against the original. In addition, how each component of C* would be operationalised and measured, how the five layers would be implemented in a specific country or institution, and how epistemic resilience would be assessed in human–AI care collaboration are all put forward as concepts and claims awaiting further research and policy practice. The references to OECD employment shares and the Scotland and UK examples serve as background, and their sources and years should be verified against the original citations.