The long tail does not drop off into nothingness (which would be the ‘exponential’ rather than ‘power-law’ distribution), where there are a few hubs and every other node has almost no connections. Instead, the long tail is extensive, consisting of small groups of dense connections, going ever onwards. In fact, the vast majority of the connections in the network are not in the hub, but in the long tail. One clear example is that of book-selling in the 21st century: the majority of Amazon.com’s book sales are not in the best-seller list, but in those millions of titles in the long tail that only a few people order. Every successful movement must be built on dense local connections. It is these dense local connections that support the dynamic creation of hubs.

In a perfect world, every node would be a hub – we would all easily connect with any other person and be able to communicate. However, creating connections takes time and energy, so nodes that are more long-standing or just have more spare time will naturally become hubs

The Construction of Collective Intelligence Hubs tend to evolve naturally in well-functioning networks – but we can accelerate the process of network development

Unfortunately people can’t become hubs without largely re-inventing the wheel. It might be irritating for existing hubs, but it’s true. Being a hub requires more than just introductions, it requires information, skills, knowledge, and a memory of the past. However, we can accelerate this process by decentring as much of the connections and knowledge as possible away from individual humans and onto the environment, whether this environment be books, websites, songs, maps, videos, and a myriad of yet un-thought-of representational forms. A useful example is the pheromone trace of the ant, reinforced as more ants use a particular trail. The mere act of ‘leaving a trail’ shows how individuals with limited memory can use the shaping of the environment as an external memory.

You can imagine this on an individual level: a person using their mobile phone to remember the phone numbers of their friends. With easy access and reliability, the phone almost seems part of your intelligence. Just extend this so that the part of your mind that is extended into the environment is accessible and even modifiable by other people, and collective intelligence begins.

This use of the environment to store collective intelligence allows for the easier creation of hubs.

Collective intelligence allows highly organised successful actions to be performed by individuals who, with limited memory and knowledge, would otherwise be unable to become hubs.

Collective intelligence requires a commons of collective representations and memory accessible to the network, and so digital representations on the internet are idea

A key focus for improving our collective intelligence would be a few central websites compiling analyses of social movements and events, alongside practical pieces from key hubs and organisers on how particular events were pulled off. A collective ratings approach would allow people to quickly find needles in the electronic haystack, via Digg-It-style ‘I like this article’ tags, or collaborative bookmarking, allowing different users to see each other’s bookmarked webpages. Of course some of these types of things exist, with tagging systems well developed on sites of magazines, newspapers and blogs. However, no current website performs the function of an analysis and learning hub

If we are to act swiftly and sustain momentum we will need to create collective intelligence – the ability to create accurate records of events, distribute them widely, analyse success and failure, and to pass on skills and knowledge.