![[Home]](http://meatballwiki.org/meatball.gif)
CollectiveIntelligence
MeatballWiki | RecentChanges | Random Page | Indices | CategoriesAs opposed to a HiveMind which is top-down, a CollectiveIntelligence is bottom-up. Instead of one thinker controlling many agents, many agents create one thought.
The crucial difference may be that agents must be capable of autonomous individual thought to achieve a CollectiveIntelligence because there is no other source of ideas to generate the CollectiveIdea. Compare to a HiveMind, where the HiveMind is the source of all ideas.
In general, even if there is no direct desire to form a stable low-energy system of convergent ideas (i.e. everyone agrees), one will form anyway. However, sometimes this doesn't happen and you end up with a highly energetic but stable system of extremely divergent ideas (i.e. everyone disagrees vehemently, such as on abortion). Whenever the system stabilizes, no new ideas are generated and the entire system threatens to become stale (new, exotic problems cannot be solved). Fortunately, random mutations on existing ideas and totally random new ideas get generated fairly consistently, sometimes causing an entire evolutionary swing in the idea genome. Sometimes this is called a ParadigmShift.
Most of the techniques that create a CollectiveIntelligence individually may be perceived as oppressive and negative; however, usually CollectiveIdeas are considered positive. This may be an example where the ends justify the means, or the concept that CollectiveIdeas are positive may itself be a CollectiveIdea. It's impossible to know from within the system, and no one talks to anyone outside the system (by definition).
CollectiveIdeas do, however, create problems because they are self-reinforcing to the exclusion of other ideas. Unfortunately, this GroupThink from TheCollective can cause good ideas to vapourize.
Good or bad, interesting applications of a CollectiveIntelligence do exist in the form of blackboards in ArtificialIntelligence, democratic capitalism, the GlobalBrain, and the Linux kernel (an example of a forced convergence).
-- SunirShah
[This has been lifted in part from Wiki:CollectiveIntelligence ]
- The [Institute for Co-Intelligence] has a visionary article on [CollectiveIntelligence]
- [Blog of Collective Intelligence] is an insightful resource
References
Surowiecki, J. (2004) The wisdom of crowds: Why the many are smarter than the few and how collective wisdom shapes, business, economies, societies and nations. Doubleday. ISBN 0385503865 (alternate, search)
According to Kuhn's "Structure of Scientific Revolution", the "stale" periods of widespread agreement are when most of the productive work gets done. People have a common foundation and common aims; they don't have to spend their time explaining themselves to each other. They get on with the BarnRaising instead of bickering about what kind of barn to raise.
This productive work is what exposes flaws in the current paradigm. To call the process "random" is a bit misleading. It's not arbitrary, although it may be too complex to predict in advance.
This is for humans doing science. I am not sure how much applies to other kinds of collective intelligence.
You don't know that all of the participants at Meatball are human. Some of us could be AIs. And I don't mean to be facetious here. When I play internet chess, I don't know whether my opponent is one human, two humans, a human plus a computer, or just one computer. If they play badly, I suspect humans, but even then I don't know for sure.
The Wiki:WikiMind model supports this idea. -- FridemarPache
See WikiPedia:collective_intelligence