The Committee of Public Safety

Hierarchical template memory is a theory of how the human mind works created by Palm Pilot creator Jeff Hawkins. In this theory, the brain does four things:

1. Discover causes in the world
2. Infer causes of novel input
3. Make predictions
4. Direct behavior

This is close to a concept created by theorist John Boyd: the OODA loop. At its simplest level, the OODA loop is a repeating sequence of Observe – Orient – Decide – Act. This can be represented simply for Wall Street types:

OODA for the proles

Or you can use the Boyd version for adults:

OODA for adults

• Observation is discovering causes in the world through the reception of input.
• Orientation is inferring causes of novel input.
• Decision is implicitly making predictions about the future by choosing one future behavior over other potential behaviors.
• Action is directed behavior.

The loop proceeds in a repeating cycle. The OODA loop is a cycle of adaptation. Information comes in. Orientation conducts analysis and deduction on the evidence, which involves tearing information apart, going from the outside in and filtering the results through genetic, cultural, and experiential lenses. Then it synthesizes the information in a new configuration through induction. This synthesis is compressed into a prediction, the hypothesis, which is subjected to testing through action. The results of the test are then fed back into observation and the cycle continues. The cycle of adaptions eventually conforms to the reality of the external environment. This is the scientific method in Boyd’s model.

The OODA is a good way to model iterative adaptive processes. The OODA loop is conventionally viewed as a matter of adaptation through speed, “getting inside an opponent’s OODA loop”, a race between opposing OODA loops with the winner being the one who cycles through his OODA loop faster. Boyd saw it as more of a matter of disrupting the opponent’s OODA loop, cutting them off from the external environment, creating a closed system that will be consumed by its own entropy.

A key part of an iterative process is the decision, the prediction. This is the center of planning, making a family of predictions about the future, usually arranged in a particular sequence. However, the important part, according to Boyd’s model, isn’t decisions or predictions. The most important step in the sequence is orientation, where information is filtered through an analysis/synthesis process, previous experience, culture, and genetic predisposition. Orientation is what produces the results that decision rules on. Not only that, Boyd’s version can bypass decision and go straight from orientation to observation in a short-circuited OO loop.

A key part of using OODA loops is realizing that there are usually multiple OODA loops operating in parallel at different levels of adaption. I find a 5-level adaption cycle model useful:

Multilevel adaption cycles

The five adaption cycles in this model are:

1. Cultural – adaption through the prioritization of goals.
2. Political - adaption through the division of power between goals.
3. Strategic – adaption through the integration of goals and power.
4. Operational – adaption through the arrangement of goals and power in time and space.
5. Tactical – adaptation through the direct interaction of goals and power with the external environment consisting of friction (anything non-living that obstructs adaption) and The Other (anything living that obstructs adaptation).

And these are just adaption cycles on human timescales. An additional cycle is the biological cycle, the OODA loop of evolutionary adaptation.

Keeping each level separate is a good way for the purpose of analysis and synthesis to ensure that the result of a lower priority loop does not compromise the operation of a higher level loop and to properly prioritize what goes where. This difference and interaction is best demonstrated by Thomas Kuhn’s idea of a paradigm shift. Most adaptions within a current paradigm are small scale adaptions that optimize (better adapt) the paradigm. Eventually, these small scale adaptions build up until they add up to a total systemic adaptation and the paradigm undergoes a total transformation. The pattern of minor iterative adaptions then resumes. Similarly, changes in lower priority levels of adaptations can add up and cascade up the stack until they change all higher levels. This can also happen with higher levels trickling down as well.

The devil lies in the details.