2-FEB-2003 YON

See Also: Book Notes,

Swarm Intelligence

by James Kennedy, Russell C. Eberhart, with Yuhui Shi

Table of Contents

The is a very beefy book that has a ton of summaries of papers and concepts and is a rich read, even for a layperson such as myself.

The thesis: "Thinking is Social"


There is one minor disappointment - no mention of Kevin Kelly and Out Of Control, which was talking about swarms and the "hive mind" close to a decade ago.

The other more serious detriment is that intelligence itself is not properly defined. There is no mention of Multiple Intelligences. The thesis that there are several ways to be intelligent: Linguistic, Logical-mathematical, Spatial, Bodily-Kinesthetic, Musical, Interpersonal, Intrapersonal, Naturalist, and perhaps even Emotional Intelligence (book: Daniel Goleman). A casual glimpse of the net selects this as a good starting point for multiple intelligence:
http://www.thomasarmstrong.com/multiple_intelligences.htm

On to the good things. The book spend the first few chapters reviewing basic genetic and adaptive algorithms in enough detail to shed light on how they will be useful. This is the whole "complex behavior from simple rules" thing.

It then follows the work of practitioners to layout the thesis. That the things we are smart about develop because we exchange information in a social context.

On page 93, there are several profound statements:

"it may be impossible for humans to really extract ourselves from our egoistic perspective, to see ourselves as disposable stepping stones for some bigger or littler entities progress and survival."

"Our role in nature, whether it is to pass on genes or mitochondria, to support our group, to maintain the Gaia-sphere, or even to provide for the survival of the individual organism, involves our ability to minimize error in our adaptations to the environment."

Perhaps it is just because I was reading the Upanishads when I read this section, but add Gaia and I'm there. It reminds me of how  in Stephenson's Snow Crash, he talks about religion being a virus who's sole purpose is self perpetuation.

Here are some other really cool things:
Kurt Lewin's field theory and "life space" - a geometry of the individual in relation to its surroundings. I'm always a sucker for that - check out Gärdenfors Conceptual Spaces.

http://fates.cns.muskingum.edu/~psych/psycweb/history/lewin.htm

Albert Bandura's social learning in which there is vicarious learning. That stimuli and response are not what they seem. In fact, in social situation, the behavior causes the stimuli - take that Skinner.

Robert Boyd and Peter Richardson's evolution cultural model. This is a mathematical model for how cultures adapt based on social learning, individual learning. Both are needed, but, evolution will favor those who can learn from other's mistakes. Also, Latane talks about the strength of influence of groups by distance and size. Modeling of this results in a society which maintains some minorities.

Kaiping Peng & Richard Nisbett review notions of culture and the history of science and come up with these differences between Chinese/Eastern cultural beliefs:

* Principle of Change (Reality is changing 24/7) (Process Theology)
* Principle of Contradiction - opposites exist, get used to it.
* Holism - sum of the parts, etc. It is not en ought to understand something in isolation.

Western beliefs:

* Principle of Identity - I am what I am
* Law of noncontradiction - A cannot equal NOT A
* Law of the Excluded Middle - between two members of a contradiction there can be no middle term.

Conclusion

The book seems packed with useful stuff that would help implement a nontrivial model of social groups, such as the MASSIVE model used for the Lord of the Rings battle scenes. From WIRED news:
http://www.wired.com/news/digiwood/0,1412,56778,00.html

In Massive, agents' brains -- which look like intricate flow charts -- define how they see and hear, how fast they run and how slowly they die. For the films, stunt actors' movements were recorded in the studio to enable the agents to wield weapons realistically, duck to avoid a sword, charge an enemy and fall off tower walls flailing. [later "We set off the simulation, and in the distance you could see several guys running for the hills."]

This year's GEEK book - The Cognitive Neuroscience of Vision , by Martha J. Farah