100224 - emergence
My notes from Steven Johnson's Emergence:
Introduction
Slime mold: oscillation between a single creature and a swarm.
Morphogenesis: the development of ever more complex structures out of simple beginnings without any ‘master planner’ calling the shots. “Bottom-up behavior.”
Simple agents follow simple rules to generate complex structures. They operate according local conditions, not a knowledge of the whole.
Positive feedback loops encourage particular behaviors to take shape.
Behaviors (or qualities) identified in emergent systems are only recognizable at the collective scale, not at the scale of an individual agent.
Emergent systems are operative in varied fields. The systems are similar, but the medium in which they operate is different.
Emergent systems get their intelligence from “masses of relatively stupid elements rather than a single, intelligent ‘executive branch.’”
Emergence is movement from low-level rules to higher-level sophistication, however a system is not emergent until it displays some type of macro-behavior.
Adaptive emergent systems adjust themselves until a productive or useful macro-behavior is produced. Emergence without adaptation is like snowflakes, beautiful but useless.
Tuning the system. Given a stated goal, how do you make an emergent system adaptive?
Control Artist
Cannot predict results just by looking at the rules. The system must live before it can be understood.
Our tendency is to think of systems such as flocking birds as having a leader rather than a set of the simple rules that each bird follows.
Emergent systems obey rules defined in advance; the rules govern micro-motives. Macro-behaviors are controlled indirectly. “All you do is set up the conditions you think will make that behavior possible. Then you press play and see what happens.”
New form of programming, software that is “grown” rather than “engineered.” Programming that is ‘more like baking a cake’ than ‘engineering a machine.’
In the fitness landscape, there are local maximums. Finding global maximums is a process of trial and error.
‘Fitness’ implies that there is a gauge for success.
The rules of the game and the world of the game can be explored simultaneously. As a society we are becoming more tolerant of being somewhat out of control. We are more tolerant of the phase where the rules don’t all make sense.
Emergent systems are controlled “from the margins,” therefore the unexpected is possible.
“Rules give games their structure, and without that structure, there’s no game: every move is a checkmate, and every toss of the dice lands you on Park Place.”
A game where anything can happen is, by definition, not a game.
Emphasizing rules may seem antithetical to an open-ended, exploratory system, but this is not the case. The capacity for growth and experimentation relies on low-level rules.
“Emergent behaviors, like games, are all about living within the boundaries defined by rules, but also using that space to create something greater than the sum of its parts.”
In game design where a player has oblique control, it is up to the game designer to determine how far to the margin the player’s control will be located. Too much or too little control results in a poor game.
Designers have a feel for the middle ground between too much control and too little.
Introduction
Slime mold: oscillation between a single creature and a swarm.
Morphogenesis: the development of ever more complex structures out of simple beginnings without any ‘master planner’ calling the shots. “Bottom-up behavior.”
Simple agents follow simple rules to generate complex structures. They operate according local conditions, not a knowledge of the whole.
Positive feedback loops encourage particular behaviors to take shape.
Behaviors (or qualities) identified in emergent systems are only recognizable at the collective scale, not at the scale of an individual agent.
Emergent systems are operative in varied fields. The systems are similar, but the medium in which they operate is different.
Emergent systems get their intelligence from “masses of relatively stupid elements rather than a single, intelligent ‘executive branch.’”
Emergence is movement from low-level rules to higher-level sophistication, however a system is not emergent until it displays some type of macro-behavior.
Adaptive emergent systems adjust themselves until a productive or useful macro-behavior is produced. Emergence without adaptation is like snowflakes, beautiful but useless.
Tuning the system. Given a stated goal, how do you make an emergent system adaptive?
Control Artist
Cannot predict results just by looking at the rules. The system must live before it can be understood.
Our tendency is to think of systems such as flocking birds as having a leader rather than a set of the simple rules that each bird follows.
Emergent systems obey rules defined in advance; the rules govern micro-motives. Macro-behaviors are controlled indirectly. “All you do is set up the conditions you think will make that behavior possible. Then you press play and see what happens.”
New form of programming, software that is “grown” rather than “engineered.” Programming that is ‘more like baking a cake’ than ‘engineering a machine.’
In the fitness landscape, there are local maximums. Finding global maximums is a process of trial and error.
‘Fitness’ implies that there is a gauge for success.
The rules of the game and the world of the game can be explored simultaneously. As a society we are becoming more tolerant of being somewhat out of control. We are more tolerant of the phase where the rules don’t all make sense.
Emergent systems are controlled “from the margins,” therefore the unexpected is possible.
“Rules give games their structure, and without that structure, there’s no game: every move is a checkmate, and every toss of the dice lands you on Park Place.”
A game where anything can happen is, by definition, not a game.
Emphasizing rules may seem antithetical to an open-ended, exploratory system, but this is not the case. The capacity for growth and experimentation relies on low-level rules.
“Emergent behaviors, like games, are all about living within the boundaries defined by rules, but also using that space to create something greater than the sum of its parts.”
In game design where a player has oblique control, it is up to the game designer to determine how far to the margin the player’s control will be located. Too much or too little control results in a poor game.
Designers have a feel for the middle ground between too much control and too little.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment