FW: REPLY Wisdom of Crowds (SD5319)

I'm forwarding a message from the system dynamics list because it's ringing
a memory bell in the dim recesses of my increasingly foggy mind.

Isn't there a PCT demo related to crowd behavior somewhere? Might that not
inform the "wisdom of crowds" discussion occurring on the system dynamics
list?

Regards,

Fred Nickols
www.nickols.us
nickols@att.net

···

-----Original Message-----
From: system dynamics listserve [mailto:SDMAIL@listserv.albany.edu] On
Behalf Of sheldon.friedman comcast.net
Sent: Friday, June 10, 2005 7:08 AM
To: SDMAIL@listserv.albany.edu
Subject: REPLY Wisdom of Crowds (SD5319)

Posted by sheldon.friedman@comcast.net
Hi
I noticed the discussion on the wisdom of crowds and the use of groups.
Part of this wisdom may come from what has ben called transactive memory.
People sharing bits if memory of information they all experienced. On the
other hand there is the concept of hiddent profiles where members do not
share information because they assume the information they have is not
important to a problem. Group members talk about what they all share as
knowledge , not unique information.

Shelly Friedman
Posted by sheldon.friedman@comcast.net
posting date Thu, 09 Jun 2005 15:15:16 +0000

Hi, Fred --

There is a Crowd demo, but it needs a writeup before I send it to you. I'll do that when I get back from Phoenix late next week. Have you thought of coming to the CSG meeting in Toronto? I would be nice to meet you.

Bill

[Martin Taylor 2005.06.10.16.31]

Fred Nickols (apparently Fri, 10 Jun 2005 12:40:03 -0400)

I'm forwarding a message from the system dynamics list because it's ringing
a memory bell in the dim recesses of my increasingly foggy mind.

Isn't there a PCT demo related to crowd behavior somewhere? Might that not
inform the "wisdom of crowds" discussion occurring on the system dynamics
list?

Yes, and it served as the basis for one of the article in the PCT issue of the International Journal of Human-Computer Studies.

But from that system dynamics list thread, most of the contributions had more to do with (what seems to be) the fact that the concensus opinion of a group is often better informed than the best individual in the group. One contributor asked at what point the "Wisdom of Crowds" becomes the "Madness of Crowds" -- another phenomenon that is perhaps better known.

The contribution that comes closest to the PCT Crowd demo is this one, though it has more to do with stigmergy than with interacting control systems:

···

===================================

Sender: system dynamics listserve <SDMAIL@listserv.albany.edu>
From: "Tim Scheffmann Tim.Scheffmann web.de" <system-dynamics@VENSIM.COM>
Subject: REPLY Wisdom of Crowds (SD5316)
To: SDMAIL@listserv.albany.edu

Posted by "Tim Scheffmann" <Tim.Scheffmann@web.de>

Hi everyone,

I read a very interesting article about swarm intelligence. These creepy
little armies are very robust and their organization structure is nearly
unbreakable. There might be the link to the system dynamics methodology to
strengthen system structures (rather than to predict the future). I attached
an abstract from my master thesis about "Enterprise Risk Dynamics"(please
find this further down).

On the Boston conference I read about a workshop with the title: Agent-based
Models for Crowd Dynamics
(#484)".(http://www.systemdynamics.org/conf2005/WorkshopSummary.htm)
  Here might be another link to how to model such behavior.

Cheers, Tim

Tim Scheffmann
IBM BCS Consulting
+49-172 - 7929242
Tim@Scheffmann.de

The abstract from my master thesis:
"Swarm intelligence", a concept introduced to business optimisation from
scientific observa-tion, can give further insight into how an organisation
can be strengthened and how complex matters can be handled. Swarm
intelligence is based on the system of order that prevails in social insect
colonies, for example ant colonies. There is no one ant in charge and
individual ants seems to act on their own initiative, not in synchrony with
the rest of the colony. How-ever, "ant highways," which can run over
hundreds of feet, represent highly coordinated forms of collective
behaviour. Humans, on the other hand, and especially organisations, suf-fer
from a "centralised mindset"; they would like to assign the coordination of
activities to a central command. Self-organisation, by contrast, feeds upon
errors to provide the colony with flexibility and robustness. The colony can
adapt to a changing environment, and even when one or more individuals fail,
the group can still perform its tasks. "The simple rules by which
individuals interact can generate complex group behaviour. Indeed, the
emergence of such collective behaviour out of simple rules is one the great
lessons of swarm intelligence." If applied well, self-organisation endows
the swarm with the ability to adapt to situations that are immediate and
were not known before. This approach has proven itself in a number of
situations, ranging from network routing to supply chain optimisation.
However, it certainly requires a drastic shift in the human mindset to
become widely used.

A bank which promotes the personal responsibility of every employee can
establish a robust organisation similar to that of the ant colony mentioned
above. In the end, business is about people, and the human factor has a
strong impact on change and risk management as peo-ple execute and take the
right (or wrong) decisions. As Charles Darwin brought out:"It is not the
strongest species that survive, nor the most intelligent, but the ones most
responsive to change.."

The source is:
"Swarm Intelligence: An Interview with Eric Bonabeau", 2003, available at
http://www.openp2p.com/pub/a/p2p/2003/02/21/bonabeau.html, accessed 1 March
2005.

Posted by "Tim Scheffmann" <Tim.Scheffmann@web.de>
posting date Thu, 9 Jun 2005 13:46:50 +0100

===========================

Martin

Bill,

How about the writeup you (and I) put together a year ago?

Specifically for Crowd:

http://www.livingcontrolsystems.com/demo_programs/crowd_delphi_demo.pdf
and
http://www.livingcontrolsystems.com/demo_programs/crowd_dos_demo.pdf

See http://www.livingcontrolsystems.com/files/tutor_pct.html
for all the tutorials, demos and simulations that you and I gathered/updated/documented at that time.

I expected this work to propagate to other PCT sites, but that has not happened. The PCT demos listed in most places are still the 1994/5 diskettes I put togetger back then.

Have you updated the Crowd demo further than what I have?

Best, Dag

···

At 01:46 PM 6/10/2005 -0600, you wrote:

Hi, Fred --

There is a Crowd demo, but it needs a writeup before I send it to you. I'll do that when I get back from Phoenix late next week. Have you thought of coming to the CSG meeting in Toronto? I would be nice to meet you.

Bill

[From Bill Powers (2005.06.10.1615 MDT)]

How about the writeup you (and I) put together a year ago?

Specifically for Crowd:

http://www.livingcontrolsystems.com/demo_programs/crowd_delphi_demo.pdf
and
http://www.livingcontrolsystems.com/demo_programs/crowd_dos_demo.pdf

See http://www.livingcontrolsystems.com/files/tutor_pct.html
for all the tutorials, demos and simulations that you and I gathered/updated/documented at that time.

I expected this work to propagate to other PCT sites, but that has not happened. The PCT demos listed in most places are still the 1994/5 diskettes I put together back then.

Have you updated the Crowd demo further than what I have?

No -- I'd forgotten about all that. Leaky memory. That's obviously where Fred should look. But see the post from Martin -- probably not relevant to what the SD people are talking about.

I'm off first thing in the morning for a four-day visit with Ed Ford. Back at home next Thursday, what with travel.

Best,

Bill P.