The question beneath management practice
The Capability Science trilogy is the culmination of a way of thinking about organisations that developed over many years of study, observation and practice.
Long before the books were conceived, the study of management science, organisational theory, strategy, economics, leadership and related fields was valuable. Yet I repeatedly observed a dissonance between the attractiveness of management ideas and their adoption within actual organisations.
Ideas that appeared persuasive in theory were often rejected, adopted only symbolically, implemented incompletely or abandoned when they encountered the history, incentives, relationships, constraints and established practices of a particular organisation. What was commonly described as an implementation failure often appeared to be a deeper theoretical problem. The organisation was being treated as an empty setting into which a preferred practice could be inserted, rather than as an accumulated system with its own capabilities, limitations and path dependencies.
From prescription to context
This led me to become increasingly sceptical of prescriptive “how-to” propositions. Such work often begins with a successful organisation, leader or method and converts a particular experience into a general prescription: adopt this structure, imitate this culture, follow these steps or introduce this management practice.
These approaches can be useful, but they are often too prescriptive, too limiting and insufficiently contextual. A practice that succeeds in one organisation may fail in another because the two organisations differ in their people, knowledge, technology, authority, incentives, governance, relationships, history and purpose. The more fundamental question wasn’t, “What should an organisation do?” It was, “What must this particular organised system become capable of doing, given its purpose, context, existing capacity and constraints?”
That question became one of the foundations of Capability Science.
Organisations as capability systems
The work also emerged from a career spent attempting to create and improve organisations that were both efficient and effective: economical in their use of resources, but oriented towards the achievement of intended outcomes.
Although I did not initially use the formal language developed in these books, the underlying concern was consistent. An organisation had to be capable of doing what its purpose required. Its people, knowledge, systems, technologies, resources, relationships and governance had to operate as an integrated productive system rather than merely coexist.
Efficiency without effectiveness can produce an organisation that performs the wrong activities cheaply. Effectiveness without efficiency can produce outcomes that are costly, fragile or unsustainable. The practical objective was therefore to build organisations that were capable, outcome-oriented and able to reproduce their performance over time.
Experience repeatedly showed that organisational performance could not be explained by isolated resources, individual competence or formal structure alone. Good people could be rendered ineffective by poor systems. Strong technologies could fail because knowledge, authority and workflow had not been reorganised around them. Strategies could remain unrealised because the organisation had not formed the capabilities required to execute them.
Conversely, organisations sometimes achieved exceptional results with apparently ordinary resources because those resources had been combined into a coherent system. Over time, I came to see organisations not principally as legal entities, hierarchies, collections of assets or bundles of contracts, but as historically accumulated systems of capability.
Continuity, fashion and human nature
Capability Science is also shaped by the belief that human nature changes far less rapidly than the language used to describe it.
Technologies, institutions, structures and social conventions can change greatly. The underlying human tendencies with which organised systems must contend—self-interest, cooperation, ambition, loyalty, fear, trust, opportunism, habit, conformity and resistance—are more persistent.
This does not mean that human behaviour is fixed. Context, incentives, culture, authority and institutions strongly influence how these tendencies are expressed. But the basic human materials from which organisations are constructed do not transform each time a new management movement appears.
For that reason, I have generally been cautious about claims that an emerging practice represents a complete break with the past. The more strongly an idea depends upon the proposition that everything is now different, the more carefully it should be examined.
Some changes are unquestionably profound. Digital networks, new institutional forms and artificial intelligence can materially alter what organised systems are able to do. But genuine transformation should eventually be visible in changed productive capacity: altered decision rights, accumulated knowledge, redesigned workflows, improved organisational memory, different relationships between people and technology, stronger repeatability of outcomes and greater resilience under varying conditions.
Where those changes do not occur, an organisation may have adopted the language and symbols of transformation without forming the capability the language implies.
Risk, innovation and renewal
A related belief is that organisations are as much managers of risk as they are agents of innovation and change.
Much contemporary management writing privileges disruption, agility, speed and continuous transformation. These ideas can be valuable, particularly where organisations have become complacent or their environments are changing rapidly. But the emphasis is incomplete.
Organisations also exist to manage uncertainty, preserve knowledge, allocate responsibility, maintain standards, protect resources and produce outcomes more reliably than unstructured individual action. Their routines, controls, hierarchies, professional standards and institutional memories are not simply barriers to change. They are also mechanisms for managing risk.
An organisation must therefore perform two functions simultaneously. It must be capable of changing when its purpose, environment or technology requires change. It must also preserve the knowledge, discipline, relationships and controls on which reliable performance depends.
Innovation without risk management can destroy valuable accumulated capability. Risk management without adaptation can preserve an organisation until its existing capability becomes obsolete. The problem is not to choose change over continuity, or continuity over change, but to govern their interaction.
Using AI to examine cognition
The more recent programme of study and theoretical development did not create this way of thinking. It allowed it to be made explicit, disciplined, challenged and organised into a coherent body of work.
Artificial intelligence played an extensive role in that process. AI was used in research, comparison, drafting, criticism, restructuring, editing and production. It assisted in surveying adjacent bodies of thought, proposing alternative formulations, testing definitions, identifying inconsistencies, generating rival explanations and translating abstract ideas into different forms.
Its use was also deliberate for a more substantive reason. One of the defining claims of the present period is that artificial intelligence will replace, or radically displace, substantial parts of human cognition. It is increasingly expected to perform research, analysis, synthesis, writing, decision support and other work previously associated with professional and managerial judgement.
I wanted to evaluate that proposition through use rather than speculation.
The development of Capability Science provided a demanding setting in which to do so. The project required integration across management science, organisational theory, economics, governance, engineering, artificial intelligence, investment and professional practice. It required definitions to remain consistent across several volumes, theoretical boundaries to be maintained, contradictions to be identified and abstract propositions to be translated into practical methods.
The use of AI therefore became an experiment in the division of cognitive labour between a person and an artificial system.
What the process revealed
AI proved to be an extraordinarily powerful cognitive resource. It could compare large bodies of material, generate alternatives, expose inconsistencies and support repeated cycles of criticism and revision at a speed and scale that would otherwise have required substantially more time.
It also revealed important limitations. AI could generate many plausible ideas, but plausibility was not the same as importance. It could propose concepts without independently knowing which belonged in the governing theory and which were merely attractive extensions. It could produce local coherence while gradually introducing conceptual drift across a larger body of work. It tended naturally towards expansion, whereas intellectual progress often required restraint, subtraction, consolidation and rejection.
It could discuss organisational experience, but it had not lived that experience. It could analyse risk, innovation, adoption and resistance, but it did not possess the accumulated observations that caused those matters to be treated as central. It could propose answers, but it did not determine why these particular questions were worth asking.
The process reinforced the view that cognition is not a single, undifferentiated activity that will simply pass from humans to machines. AI was highly capable in information retrieval, comparison, pattern recognition, alternative generation, drafting and formal consistency. In some of these activities, it greatly exceeded what an individual could efficiently perform alone.
The human contribution remained decisive in purpose, context, selection, theoretical restraint, judgement, governance and responsibility.
Where the generation of ideas becomes inexpensive, selection becomes more valuable. Where persuasive prose becomes abundant, judgement becomes more important. Where alternatives can be produced instantly, the ability to determine purpose, context and acceptable risk becomes more consequential.
The trilogy as culmination
The production of the trilogy can itself be understood as an application of Capability Science.
AI was neither an autonomous substitute for the author nor a merely passive instrument. It was a powerful cognitive resource integrated into a larger productive system that included accumulated study, organisational experience, governing questions, prior beliefs, theoretical controls, repeated review, rejection, revision and final human responsibility.
AI materially increased the productive capacity of that system. But its value depended upon how it was organised, governed and directed.
The trilogy therefore represents more than a set of books about capability. It is also a practical demonstration of how human experience, purpose, judgement, governance and artificial intelligence can be organised into a productive capability system.
Capability Science ultimately reflects the way I had come to understand organised systems: as historically accumulated capability systems that must achieve outcomes, manage risk, preserve what remains valuable, adapt when necessary and distinguish genuine productive transformation from the temporary adoption of fashionable ideas.
AI helped me formalise, test and express that understanding. The resulting work is the culmination of the intellectual and practical trajectory from which the trilogy arose.
General information
Celerity publishes general research and commentary only. Nothing in this publication constitutes financial advice, investment advice, personal advice, an offer, solicitation or recommendation to buy or sell any financial product or security.