How times have changed [was Sr. Developer ]

Fred puissante at lrc.puissante.com
Thu Feb 16 10:23:01 EST 2006


Ah, finally got Kontact (Kmail) to not crash on me. Too many email accounts 
for it, I think.

On Wednesday 15 February 2006 12:16, Ben Scott wrote:
...
>   Group's are neither clueless nor clueful.  People are.  Groups do
> not do things; individuals do.  A group may contain many clueful
> people doing good things (or clueless people doing evil things, or
> whatever), but that does not mean one can know everything about $GROUP
> or any arbitrary member of it  simply by knowing something about one
> or some of its members.
>
>   This is part of what I call this "the myth of the organization".
> Organizations are a myth; they don't exist in reality.

I would have to somewhat disagree with this. Whenever a group of nodes 
interact in concert, that entire group can be considered a node in its own 
right, with its own peculiar set of dynamics.  It does not matter whether 
the "nodes" are ants, computers, or people.

When the number of interacting nodes are small, one may easily distinguish 
them as separate, though they are acting as a group. When the number of 
nodes are large, it's much more difficult to cull them apart.

This is all about emergence of behavior and also a new mathematics I am 
quietly (for now) working on.

So, to differ, "organizations" DO exist -- but their efficiency to act is 
inversely proportional to the number of participants. That is to say, the 
"collective IQ" of the group, if there is such a thing, will always be 
*less* than the IQ of individuals. The reason for this is simple: The 
individuals are *not free* to act as the group can. The group itself is by 
design forced to act as a unit; therefore it is a node in its own right.

This scares the willies out of me because group nodes where it involves 
humans typically become *less humane* than the individuals themselves.

> -- Ben "This statement is false" Scott

-Fred
 Be afraid, be very afraid...



More information about the gnhlug-discuss mailing list