Smart Swarm

Some quotes from Perter Miller’s Smart Swarm.  A very interesting book, and highly recommended:

…a large number of individuals without supervision can accomplish difficult tasks by following simple rules when they meet and interact.- p.262

…groups can reliably make good decisions in a timely fashion as long as they seek a diversity of knowledge and perspectives, encourage a friendly competition of ideas, and narrow their choices through a mechanism like voting. – p. 263

…even small contributions to a shared project can create something useful and impressive when large numbers of individuals build upon one another’s efforts. – p.263

…without direction from a single leader, members of a group can coordinate their behavior with amazing precision simply by paying close attention to their nearest neighbors […] but it can also tempt us to follow the crowd uncritically… – pp.263-4

To vastly oversimplify our dilemma, we’re torn between belonging to a community and maximizing our personal welfare. – p.265

And some outlines from the text:

Swarms in nature have taught us two lessons:

  1. By working together in smart groups, we too can lessen the impact of uncertainty, complexity, and change. – p.267
  2. As members of smart groups, we don’t have to surrender our individuality. In nature, good decision-making comes from competition as much as compromise, from disagreement as much as from consensus. – p.268

Natural mechanisms of interactions between individuals (p.267):

  1. Reliance on local knowledge (maintain diversity of information)
  2. Simple rules of thumb (minimise computational needs)
  3. Repeated interactions between group members (amplify faint but important signals and speed up decision-making)
  4. Use of quorum thresholds (improve accuracy of of decisions)
  5. A healthy dose of randomness in individual behavior (prevent getting stuck in problem-solving ruts).

Networks

Visualization of the various routes through a ...

Image via Wikipedia

The broader lesson… is that it doesn’t make much sense to discuss properties of network structures – such as small-world or scale-free types – without also discussing what it is you want to do on them. Different types of networks are better at different types of problems. Depending on the challenges an organization might be facing, for example, greater connectivity among workers or departments might be a good thing or a bad thing. Some organizations function best as a loose collection of tribes, each with its own specialists and experts, while others work better if they involve more collaboration. Tribal networks make it easier to differentiate one group from another, while greater connectivity makes it easier for an organization to reach consensus.

Peter Miller, Smart Swarm, p. 149.

It’s a fascinating book, pulling examples from animal swarms (ants, bees, termites, birds etc) and describing how they apply to our world’s dazzling array of networks (social, information, communication etc.). Then again, I’m an INTJ, so I love stuff like this. Patterns and networks and flowcharts – oh my!

What I find particularly fascinating is the distinction between small-world (think village) and scale-free networks (think Internet) that Miller makes. Not something that’d readily come to mind.