Part 3 of a five-part series
There is a word that appears in almost every AI transition framework, every government workforce policy, and every HR strategy document produced in the last three years. That word is skills. The World Economic Forum Future of Jobs Report 2026 leads with it. The UK government’s AI Opportunities Action Plan is structured around it. McKinsey’s emerging talent profile research – which identifies M-shaped managers, T-shaped specialists, and AI-enabled frontline workers – frames its entire argument through it. Skills is the wrong unit of analysis. And the consequences of that error are compounding at scale.
Skills frameworks have a legitimate function inside a stable system where tasks are broadly known, the workforce is human, and the gap between current and required capability can be closed through training. That describes a world that no longer exists at the pace the current transition demands. A skills framework is a snapshot instrument applied to a moving system. It answers what a person can do at a task level today. It does not answer what value that person generates as a component within a system whose architecture is structurally changing beneath them.
Three distinctions the skills framework collapses
The unit of analysis the workforce conversation now requires is three related but distinct dimensions: knowledge, capability, and capacity.
Knowledge is the accumulated body of understanding, pattern recognition, contextual judgement, and relationship intelligence that a person carries. It is strategic, not transactional. It compounds over time as a productive asset compounds, and it depreciates when the context in which it was built changes, or when the human carrier leaves the organisation. Most organisations have no instrument for measuring knowledge as an asset. Knowledge is currently being extracted daily – through redundancy programmes and through Digital Worker training processes where a human teaches the system that subsequently displaces them — without recognition of its value or consideration for its transfer.
BLOOR RESEARCH DATA POINT
26–65%
Layoff concentration in knowledge-intensive roles
Knowledge embedded in these roles – built over careers – is being extracted into digital systems or lost entirely. This constitutes Labour Debt transferred onto a reduced human workforce without addressing the liability.
Source note: Bloor Research observation across organisations with advanced AI maturity (AI deployment at scale, significant offshoring, and technology-led operating models), 2024–2026. Concentration in the £50,270–£125,140+ annual salary range. Methodology: cross-functional analysis of published redundancy programmes and Bloor primary research interviews.
Capability is the architecture of how a person thinks, decides, and applies knowledge across variable contexts. The M-shaped, T-shaped, and N-shaped profile frameworks currently circulating are attempts to describe capability without a framework for building or governing it – shape descriptions without an architectural specification. Capacity is the total productive resource available to the organisation. In every organisation deploying AI systems and digital workers alongside human ones, total productive capacity includes non-human components. Workforce planning that counts only human headcount is measuring a fraction of the system.
FusionWork™ and the category the market has not named
FusionWork™ is the category Bloor Research has established to govern this complete picture. It is the first category designed specifically for a workforce that is simultaneously human and digital – the design, governance, and measurement discipline for a system in which human intelligence and digital workers each play a defined architectural role. The M-shaped manager McKinsey describes operates within a FusionWork™ architecture whether that architecture has been designed or not. The difference is whether the organisation governing that manager has designed the system intentionally or is running on inherited assumptions that no longer fit the productive reality.
The implications for the HR function are direct. The mandate is not to manage human resources. It is to govern a productive workforce whose capacity is distributed across human and digital workers – with the measurement instruments, governance frameworks, and architectural design discipline that a hybrid productive system requires.
Stop measuring skills. Start governing knowledge, capability, and capacity across the full workforce – human and digital. The organisations that make that shift in the next eighteen months will have a structural advantage that skills frameworks alone cannot close.