Collective intelligence is a term that refers to the computational abilities of a group of agents. With collective intelligence, a group is capable of accomplishing a task, or of solving an information processing problem, that is beyond the capabilities of an individual agent.
Collective intelligence depends on more than mere numbers of agents. For a collective to be considered intelligent, the whole must be greater than the sum of its parts. This idea has been used to identify the presence of collective intelligence by relating the amount of work done by a collective to the number of agents in the collection (Beni & Wang, 1991). If there is a linear increase in amount of work done as a function of the number of agents, then collective intelligence is not evident. However, if there is a nonlinear increase (e.g., an exponential increase) in the amount of work done as a function of the number of agents, then Beni and Wang argue that this is evidence that the collective is intelligent.
Collective intelligence is of interest in cognitive science because many colonies of social insects appear to exhibit this kind of intelligence, and this has inspired researchers to explore "porting" such processing to robot collectives. As far as robots are concerned, collective intelligence is exciting because it offers the possiblity of developing systems that are scalable (they don't get disrupted when more agents are added) and flexible (they don't get disrupted when some agents are damaged or fail) (Sharkey, 2006).
- Beni, G., & Wang, J. (1991, April 9-11). Theoretical problems for the realization of distributed robotic systems. Paper presented at the IEEE International Conference on Robotics and Automation, Sacramento, CA.
- Sharkey, A. J. C. (2006). Robots, insects and swarm intelligence. Artificial Intelligence Review, 26(4), 255-268.