Demise of the Super Star

Military thinkers are moving away from reliance on individual expertise, and developing strategies that depend on collective expertise and organizational intelligence.

In the organizational equivalent of natural selection, leaders in industry and senior commanders in the military have long been selected for their superior intuition. ... Experts use intuition rather than rational decisionmaking to make sense of situations. These experts discern patterns and relate these patterns to their prior experience and/or knowledge to determine the nature of the situation and the appropriate response. ...

It is becoming increasingly clear that the complexity of the situations faced and the responses needed have outpaced not only decision theoretic approaches, but have also outpaced the ability of even the best of experts (super stars) to deal with the complexities involved.

• First, the sources of complexity are accelerating. These sources of complexity include the variety of events and entities that are connected, the density of the interactions, and the speed of interactions that make it difficult to relate a cause to an effect and almost impossible to predict cascading effects.
• Second, it takes a long time for individuals to become experts and senior decisionmakers in industry and the military, spending decades to arrive in positions of leadership. This means that the bulk of their experience is well aged, increasingly out-of-date, and of questionable relevance.

Intuition is not only unlikely to help (in the face of growing complexity and volatility), but may often be misleading. Rather than rely on individual genius, Information Age processes tap collective knowledge and collaboration.

from Power to the Edge, pp 88-89

Complexity and attenuation

When events and relationships are complex, the information generated is correspondingly complex. However, management information systems are often intolerant of complex information, and seek to simplify or flatten the information by removing apparent inconsistencies. Complex information is framed as misfit, and valued negatively.

1. At the business level, misfit is regarded as inconsistency or noise, and suppressed or ignored. In systems jargon, this is known as attenuation. We can see many examples of this: from CRM systems that refuse to acknowledge (and attempt to flatten) the rich and complex signals received from or about customers, to governance systems that ignore early warning signals of misfeasance. This is perhaps most conveniently studied retrospectively; once a scandal is known about, it is often easy to see the warning signals that should have been recognized at the time.

Alternatively, in some contexts, customers whose data don’t fit the norm are automatically regarded as suspicious: this is encouraged by bureaucratic regulations (or bureaucratic interpretations of regulations) covering fraud, money laundering, and so on.

2. At the engineering level, information misfit acts as a metacommunication: interpreted as saying more about the appalling state of the systems, and of the relationship between business and systems, than about the intended content of the information.

This leads to an epistemological dilemma, which we can characterize in topological terms. An agent (e.g. person or organization) receiving complex information has a choice: open(-minded) or closed(-minded).

Closure represents the elimination of contradiction, ambiguity or uncertainty. In some contexts, some degree of closure is appropriate. However, there are powerful cultural and psychological forces that lead people and organizations into premature foreclosure.

The alternative to premature foreclosure is to hold open some (positive) space for ambiguity and uncertainty. (Note: this should not be regarded as a mandate for muddle and procrastination. There is a state of indecision represented by many fictional characters including Hamlet. However, this may be interpreted not as a positive space of uncertainty and freedom, but as a space in which the character is imprisoned in a fixed Either/Or frame.)

There are several practical implications of this argument, which I can only hint at here.
SOA. See SOAPbox: Beyond Binary Logic
Governance. Internal audit systems must be capable of holding complex information for multiple interpretation.
Trust. Always provisional. Transitive trust (friends of friends) is widely used as a social mechanism, but has well-known failure modes. (Who killed Harry Potter’s parents?)
Business Geometry / Design. Services must be capable of use in multiple contexts (provisional ontology, semantics, pragmatics).