Swarm intelligence in social insects is based on self-organization; no one is in charge, but social insects successfully solve complex tasks (p.20). According to Bonabeau, self-organization has four properties:
- positive feedback reinforces desired behavior;
- negative feedback counterbalances positive feedback;
- amplification of randomness leads to positive reinforcement;
- amplification of interactivity has a positive outcome.
Errors and randomness contribute very strongly to the success of social insects by enabling them to discover, explore, and exploit. Errors feed self-organization, creating flexibility so the colony can adapt to a changing environment with robustness, ensuring that-even when one or more individuals fail-the group can still perform its tasks. Swarm intelligence offers an alternative way of designing "intelligent" systems in which autonomy, emergence, and the ability to distribute tasks replace control, pre-programming, and centralization (p.21).
The obvious advantages in accomplishing complex tasks through swarm intelligence include no central control, errors being okay, flexibility, robustness, and self-repair. It is difficult for conventional managers to accept the idea that solutions are emergent rather than predefined and pre-programmed (p. 21).
People working with the innovator are not working for her or him because they have been ordered to do so, but because they want the innovation to succeed. They all share the same vision and goals (in a sense, the same "genes"); they want to succeed, and they want to see their innovation spread and be accepted by the outside world (p. 22).
If you and I swap a dollar, you and I still each have a dollar. If you can I swap an idea, you and I have two ideas each.
The free flow of ideas and thoughts is essential to the success of creative teams . It is entirely self-organizing and self-selecting. Roles and responsibilities of each member are clear to all, with no need for lengthy coordination meetings (p. 24).