• High-level principles of organizations

    • An organization is an information-processing entity—just like a computer, a cell, or an organism. In AI jargon: it is an intelligent agent. In military jargon: it follows an OODA loop:
      1. It receives information from its environment (Observe)
      2. It integrates that information into a model of the environment, and of itself (Orient)
      3. It uses that model to make decisions—toward accomplishing some task, or maximizing some objective (Decide)
      4. It executes its decisions—which, in turn, affect the environment (Act)
    • An organization’s intelligence is distributed among its members.
      • Each member is capable of receiving information from the environment.
      • The organization’s collective model is composed of its members’ individual models.
        • The collective model may be detailed, consistent, and extensive if appropriate channels of communication are enabled within the organization.
        • If channels of communication are broken or stifled, then the organization’s parts will operate independently, using incomplete or inconsistent models.
        • If communications are too open, then the members’ limited processing power may be overwhelmed by unnecessary demands on their attention.
      • An important part of organizational intelligence is collective memory.
        • Does the organization learn from experience?
        • Does it have a system to catalog lessons from the past?
        • Does it have a system for drawing upon those lessons from the past?
      • Decision-making and execution will also be distributed, to some extent.
    • Human nature must be accounted for, in organizations composed of humans.
      • When people get involved in organizations, it’s because they want something.
        • This could be any number of things:
          • meaning
          • to be part of something important
          • responsibility; ownership; autonomy
          • struggle and growth
          • community
          • transcendence?
          • money (in the case of employees and workplaces)
          • (parenthetically: some people avoid responsibility, ownership, struggle and growth—but they’re probably not the kind of people you want in your organization.)
        • If the organization doesn’t satisfy its members’ needs, then it probably won’t exist for very long.
      • Humans have finite capacities for processing information. This capacity needs to be carefully allocated between:
        • attending to details;
        • maintaining a sound understanding of their environment;
        • communicating;
        • deliberating over decisions;
        • taking action.
      • Humans learn from experience—or, at least, they should.
        • If a human is given well-defined responsibilities and prompt feedback on their performance, then they will grow quickly in competence.
        • This ought to be balanced with humans’ need for growth.
      • Humans are social animals.
        • Sometimes they get along. Sometimes they don’t. We won’t get into why that is. There need to be processes in place which make it possible for people to get things done, even if they don’t like each other. (There also need to be systems in place for identifying and resolving conflict.)
        • Humans haven’t really evolved out of their tribal natures. The establishment of shared myths and values can help them cooperate beyond Dunbar’s number.
        • Humans are highly motivated by status within their social group. Of course, status means different things in different groups.
      • Bottom line: when you build an organization your materials are Homo Sapiens, with all of the attendant complexity, challenge, and opportunity. Forget this at your peril.
    • The organization usually exists to accomplish some task.
      • This could be practically anything, though some tasks are more glamorous than others.
      • It’s easier to satisfy the human needs listed above if the task at least seems important.


  • The role of the manager:

    • A manager must strive to answer the following questions:
      • What task or tasks are we trying to accomplish?
      • How can those tasks be decomposed into sub-tasks?
      • How should people be allocated to those subtasks?
      • What information needs to be shared between those sub-tasks?
      • What processes will reliably lead to the successful completion of those tasks?
      • What hierarchy of responsibility will ensure the reliable execution of those processes?
        • how much autonomy should be granted to the different hierarchical levels?
      • What systems can be established, in order to reduce the role of luck in success?
        • systems for making decisions?
        • systems of collective memory?
        • channels of communication?
      • Can we answer these questions in a fashion that respects the idiosyncracies of human nature?
    • A manager must be prepared to update their answers to those questions, in response to evolving circumstances.
    • If the manager has done their job well, then…
      • …most of their time will be spent:
        • inspecting the organization for dysfunction
        • making minor tweaks to the organization
        • formulating big-picture strategy
        • making high-level decisions
      • instead of:
        • getting into the weeds of little decisions (and, hence, forming a bottleneck in the organization’s collective OODA loop).
        • bossing people around, like an aspiring dictator.
        • rushing around, frantically, as though everything depends on them personally.
      • I.e., good management benefits the organization, and the manager’s sanity.

\( \blacksquare\)