[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

RE: [ba-unrev-talk] Edge: Howard Rheingold on Smart Mobs

Possibly only peripherally related but still fascinating is this article on
the use of agents in simulation to try to understand how societies operate.
http://www.theatlantic.com/issues/2002/04/rauch.htm    (01)

A summarizing quote:
"To me, the early results of this work suggest that social engineering can
never be as effective as liberals hope, but also that it need not be as
clumsy as conservatives insist."    (02)

This article references the work of the Santa Fe Institute on mechanisms of
order evolving out of chaos.    (03)

While agent simulations cannot predict societal behavior, the elements used
do give some insights. When agents are endowed with plausible differences
and their knowledge is limited to what they can see and what their local
networks can tell them, macro behavior occurs which would lead one to
believe that the "motivation" of the agents is dramatically different than
that which was programmed. The simulations that show that a preference for
neighbors of the same color that is only 1 out of 4 leads to segregated
clusters that, on casual inspection, appear to be the result of rampant
racism.    (04)

Combined with some of the ideas from Smart Mobs which are clearly instances
of independent agents in cooperative interact, there is definite food for
thought.    (05)

One implication for our efforts is that it is very unlikely that we can
determine any global feat of social engineering that will solve any of the
"complex problems" that are of interest; we cannot even determine with any
degree of certainty what impact any augmentation tool or tools will have on
those who use it. Given that, we can expect that the best we can do is to
take our best shot at creating useful tools, make them available to people
who will use them to cooperate with other individuals and groups, which will
produce results over which we have no control and which we cannot predict.    (06)

In the simulations, global knowledge and predictive capability on the part
of the agents dramatically affects the way the simulations evolve - still in
unpredictable ways. Even so, the experiments on artificial genocides, the
difference in the actions of peacekeepers makes definite differences in
outcomes - passive reactivity proves uniformly ineffective in preventing the
"tipping" phenomenon while proactive agents can prevent the "tipping". This
doesn't imply that everything should be "action before the fact" but that,
perhaps, the "one size fits all" approach is less workable than treating the
areas that are really problems.    (07)

I hope you find as much food for thought in the combination of this article
and the one on mart mobs as I did.    (08)

Thanks,    (09)

Garold (Gary) L. Johnson    (010)

-----Original Message-----
From: owner-ba-unrev-talk@bootstrap.org
[mailto:owner-ba-unrev-talk@bootstrap.org]On Behalf Of Grant Bowman
Sent: Sunday, July 21, 2002 11:27 PM
To: ba-unrev-talk@bootstrap.org
Subject: [ba-unrev-talk] Edge: Howard Rheingold on Smart Mobs    (011)

Here's another item that describes the core ideas behind Howard
Rheingold's new book.  His book is shipping yet, but it should be soon.
He also has a thriving online community.    (012)

http://www.edge.org/3rd_culture/rheingold/rheingold_print.html    (013)

http://www.rheingold.com/community.html    (014)

-- Grant Bowman                                <grantbow@grantbow.com>    (015)