An excerpt from Chapter 5 of
Copyright © 2018 by Princeton University Press
Five crucial but nonobvious factors make all the difference. These factors may sound abstract. But they quickly become practical, giving shape to the knowledge commons that holds the group together, and they are relevant to any large-scale group wanting to think, act, and learn coherently and successfully. These are the organizing principles for collective intelligence.
The first is the extent of what I call the autonomous commons of the intelligence in the system. By this I mean how much the elements of intelligence are allowed free rein, and not subordinated too easily to ego, hierarchy, assumption, or ownership. Autonomy means allowing arguments to grow and become more refined. It requires a dialectical approach to intelligence — seeking out alternatives and refutations as a way of sharpening understanding. A group where people quickly become attached to their assertions, where secrets are guarded, or where too much weight is put on the speaker instead of what they say will tend to be collectively less intelligent. So will one that narrows options too quickly.
The second factor will be a contextually proportionate balance: how balanced the intelligence is between its different elements, and how well-suited the balance is to the tasks at hand. Intelligence combines many distinct elements, from observation and focus to memory and creativity. Groups, like individuals, need to keep these in balance, and a high proportion of the cases where collective intelligence goes wrong reflect problems of imbalance, such as where groups are rich in data but poor in judgment, or rich in memory and poor in creativity, and vice versa. Knowing how to orchestrate these different elements of intelligence in a coherent way is one of the fundamental tasks facing any group and leader.
The third factor will be how well the group can focus. Focus means attending to what really matters and not being distracted. Knowing what to ignore matters as much as knowing what to attend to. That may not be so obvious. For the group stranded at an airport, there will certainly need to be a focus on getting through to someone out there. But if they are to be stuck for a long period, then holding the group intact and preventing conflict may matter even more. Focus also has a subtler meaning since it introduces granularity — knowing what is relevant on different scales.
The fourth factor will be the group’s capacity to be reflexive — to be intelligent about itself and recursive. Knowledge needs knowledge about the knowledge, and this requires loops — what I describe as the three loops of active intelligence: thinking about things, changing the categories with which we think about things, and changing how we think. The more reflexive any group is, the more intelligent it is in the long run. As I will show, this reflexiveness works best when it is most visible — for example, with predictions made explicitly, and explicit learning when the anticipated doesn’t happen, all feeding into a shared knowledge commons. And it works best when it is helped by what I call self-suspicion — the ability to question the patterns that make most sense.
Finally, the fifth factor will be the group’s ability to integrate for action, drawing on different types of data and ways of thinking to make a decision. It’s not enough to think great thoughts and host glorious arguments. Life depends on action. So this type of integrative thinking is what marks out the most sophisticated civilizations. It’s much of what we and our ancestors call wisdom, and it tends to develop through experience rather than only logic. It’s where thought and action come together. We complicate to understand, but simplify to act, and search for a simplicity that lies on the far side of complexity. What I call high-dimensional choices — complex in terms of cognitive tools, social relationships, and time — require more work and loops to arrive at a composite picture that can guide action, whether that action is physical in nature or communicational.
Together, these five organizing principles help any group to think more clearly about the past (the relevant collective memories), present (the facts of what is happening), and future (the options for resolving the situation). They help the group to imagine possible future options, discover them, and then realize them.
These dimensions of intelligence sound simple. But they are remarkably difficult to sustain. This is why intelligence is fragile and rare, and runs as much against nature as with it. Collective intelligence can easily regress — as has happened time and again in many places throughout human history, from Tasmanian aborigines to the once-great cities of Mohenjo Daro or Machu Picchu. The world is full of places where people live amid the ruins of superior past civilizations and full of institutions that were once far more competent than they are today. The direction of travel in our world is toward more complexity, more integration, and more intelligence — yet this is by no means a given.
Powerful enemies threaten intelligence all the time, such as purveyors of lies, distortions, rumors, and distractions, and in the digital environment, threats like trolling, spamming, cyberattacks, and denials of service. All can disrupt clear communication and thought.
The virtues that underpin collective intelligence are also rare and difficult, because each one clashes with other basic features of human interaction and other virtues. The autonomy of intelligence challenges the social order (which rests on an agreement not to see certain things while suppressing the views and voices of the powerless). It also runs directly against accountability in many instances (and too much accountability, like no accountability, can make institutions surprisingly stupid).
Balance challenges the status of the groups or professions associated with particular elements of intelligence, such as the guardians of memory. Reflexiveness challenges practicality and the pressure of events — the need to act now. It takes time to think, and time is scarce, so life rewards shortcuts.
Focus fights against curiosity, and it is particularly hard for clever people (and intelligent groups and civilizations have repeatedly been defeated by ones that are less intelligent but more focused on what really matters in a particular situation). Appropriate focus is even harder for machines that, like human brains, struggle to concentrate in a granular way, recognizing the scale of different tasks and contexts.
Finally, integrative thinking fights against our tendency to latch on to one way of thinking (to the person with a hammer, every problem looks like a nail).
This framework helps to explain why some fields act foolishly. The most common pattern is a failure to sustain the autonomy of intelligence. This can happen because power subordinates truth, but it can also happen because of incentives (as happens often in finance) or excessive loyalty.
The five principles for organizing collective intelligence are rarely found together in an ideal form. Yet most human groups follow some of them, however imperfectly. They provide a theoretical basis for understanding any kind of intelligence that exists at the level of a group, organization, network, or family. They mimic some of the properties of natural evolution and development. They involve the multiplication of options, selection, and replication (and like DNA, rarely mutate in wholly random ways, but tend to mutate more where there have already been mutations, or where there is stress and pressure). Nevertheless, the patterns of intelligence are also unlike the natural world, primarily because of awareness and freedom; we can choose whether to give greater weight to collective intelligence and can turn this into a moral choice — a commitment to be part of the intelligence of larger wholes.