Scribner, 2002
Emergence is the ability of low-level components to self-organize into a higher-level system of sophistication and intelligence. Known by many names - collective phenomenon, bottom-up behavior, self-organization, and decentralization - it is a fascinating phenomenon that Steven Johnson approaches from numerous angles in his 2001 book, Emergence: The Connected Lives of Ants, Brains, Cities, and Software.
Johnson begins with the amoeba-like slime mold cell, which in certain circumstances coalesces with thousands of neighbors to create a creature of greater complexity and even - yes - intelligence. In 2000, Japanese scientists created conditions in which a slime mold found its way through a maze. From this introduction to emergent behavior, Johnson takes the reader on a whirlwind tour of the phenomenon's study: Alan Turing's work with complexity theory, Jane Jacob's study of the emergent properties of cities, Oliver Selfridge and John Holland's experiments at the nascence of artificial intelligence, and E.O. Wilson's study of ants and theories of sociobiology are all touched upon in this survey course in bottom-up intelligence.
Ant colonies, in which every ant operates based on set of low level rules and feedback from its neighbors, serve as an excellent example of the success of emergence in nature with which Johnson illustrates five fundamental principals for building bottom-up systems:
Emergence is the ability of low-level components to self-organize into a higher-level system of sophistication and intelligence. Known by many names - collective phenomenon, bottom-up behavior, self-organization, and decentralization - it is a fascinating phenomenon that Steven Johnson approaches from numerous angles in his 2001 book, Emergence: The Connected Lives of Ants, Brains, Cities, and Software.
Johnson begins with the amoeba-like slime mold cell, which in certain circumstances coalesces with thousands of neighbors to create a creature of greater complexity and even - yes - intelligence. In 2000, Japanese scientists created conditions in which a slime mold found its way through a maze. From this introduction to emergent behavior, Johnson takes the reader on a whirlwind tour of the phenomenon's study: Alan Turing's work with complexity theory, Jane Jacob's study of the emergent properties of cities, Oliver Selfridge and John Holland's experiments at the nascence of artificial intelligence, and E.O. Wilson's study of ants and theories of sociobiology are all touched upon in this survey course in bottom-up intelligence.
Ant colonies, in which every ant operates based on set of low level rules and feedback from its neighbors, serve as an excellent example of the success of emergence in nature with which Johnson illustrates five fundamental principals for building bottom-up systems:
More is different: A critical mass of ants is necessary for useful statistical averages to emerge. One or two ants bumping against each other does not a colony make.
Ignorance is Useful: Simplicity of the individual components (ie: the ants) is beneficial. There is no need for each ant to have imprinted a map of what is in the colony's best interests, and in fact such ideas would be a detriment to the colony as a whole.
Encourage random encounters: Johnson illustrates how ants use the feedback from encountering the activities of other ants to usefully modify their behavior. Similarly, Jane Jacobs, in The Death and Life of Great American Cities, shows how humans in urban areas positively effect the emergence of cities by their encounters in public areas.
Look for Patterns in the Signs: Ants follow trails of pheromones left by other ants. Programmers have modeled this behavior when creating computer software that emerges from very simple programs that leave behind digital pheromone trails of their own.
Pay Attention to Your Neighbors: "Local information leads to global wisdom." When an ant notices a large number of his fellow ants are foraging, he will alter his behavior to another activity. Likewise, in the development of a human embryo, individual cells are able to get information from their neighbors that will guide them in their own formation, whether that be as skin cells, bone cells, muscle cells...
Ignorance is Useful: Simplicity of the individual components (ie: the ants) is beneficial. There is no need for each ant to have imprinted a map of what is in the colony's best interests, and in fact such ideas would be a detriment to the colony as a whole.
Encourage random encounters: Johnson illustrates how ants use the feedback from encountering the activities of other ants to usefully modify their behavior. Similarly, Jane Jacobs, in The Death and Life of Great American Cities, shows how humans in urban areas positively effect the emergence of cities by their encounters in public areas.
Look for Patterns in the Signs: Ants follow trails of pheromones left by other ants. Programmers have modeled this behavior when creating computer software that emerges from very simple programs that leave behind digital pheromone trails of their own.
Pay Attention to Your Neighbors: "Local information leads to global wisdom." When an ant notices a large number of his fellow ants are foraging, he will alter his behavior to another activity. Likewise, in the development of a human embryo, individual cells are able to get information from their neighbors that will guide them in their own formation, whether that be as skin cells, bone cells, muscle cells...
From ants, Johnson jumps straight into the world wide web. He cites Slashdot.org , the news aggregation web site that directly engages its users to control the quality of its content and has done so since its inception in the late 1990's. With its users as the "ants" Slashdot has all the elements of an emergent system: positive and negative feedback, structured randomness, neighbor interactions, and decentralized control. As Johnson points out, user-ratings provide exactly the sort of decentralized feedback necessary for emergence. And there are far more examples today than there were in 2001 when Emergence was first published: Digg , Reddit , Amazon , and Technorati are all employing bottom-up theory in building their own versions of macro-intelligence.
Emergence was in fact a harbinger of the social networking explosion we've seen on the web since just after the book was published. Friendster was launched in 2002 with MySpace following in 2003 and Google's Orkut in 2004. Today, there are social networking sites for everyone from teenagers to professionals to people who love their cats. Combine social-networking with other aspects of the social web, and we're seeing a disorganized entity organize rapidly into a bottom-up intelligence.
Emergence was in fact a harbinger of the social networking explosion we've seen on the web since just after the book was published. Friendster was launched in 2002 with MySpace following in 2003 and Google's Orkut in 2004. Today, there are social networking sites for everyone from teenagers to professionals to people who love their cats. Combine social-networking with other aspects of the social web, and we're seeing a disorganized entity organize rapidly into a bottom-up intelligence.
The world wide web is particularly well built for collective phenomenon and by the book's end, Johnson predicts much of what is now considered part of web 2.0 including the web's social aspects and the viral spread of software able to recognize our habits, tastes, and desires based upon the feedback that we provide (Amazon , Pandora , Del.icio.us ...).
Will bottom-up behavior find a home in the corporate world or politics? Johnson is no Nostradamus and doesn't claim to be. What he delivers in Emergence is an historical and natural understanding of the sort of the collective phenomenon that has exploded virally in our wired world today. Such cognizance will not only generate awareness of a phenomenon in which we are already enveloped, but open our minds to its possibilities in other systems that might benefit from it.
Will bottom-up behavior find a home in the corporate world or politics? Johnson is no Nostradamus and doesn't claim to be. What he delivers in Emergence is an historical and natural understanding of the sort of the collective phenomenon that has exploded virally in our wired world today. Such cognizance will not only generate awareness of a phenomenon in which we are already enveloped, but open our minds to its possibilities in other systems that might benefit from it.
