From: Jordan Dimov (jdimov@cis.clarion.edu)
Date: Sun Nov 25 2001 - 18:27:40 MST
Ok Guys, here's my thoughts about this.
Bear with me through my 'NLP overview' and take the AI-related
speculations at the end as just that.
The way I see it, NLP completely dispenses with any notions of truth and
scientific correctness and attempts to create useful models of the world.
Useful in the sense that by following these models one can achieve a
desired outcome with high reliability multiple times.
One of the premises behind NLP is that because humans have similar
neurology, ANYTHING that YOU can do, I can do too. All I need to do is
observe how you perform the thing that you are skilled at, build a model
based on my observations, and implement it myself. The classic example is
Richard Bandler and John Grinder's model of Milton Erickson's hypnotic
techniques. Erickson was (and still is) considered one of the most
effective psychotherapist of all times. Bandler and Grinder studied his
techniques, found a structure and patterns behind what he was doing, and
were able to replicate his success and teach others how to do this.
What is noteworthy in this process is that they did NOT try to find out
what he was actually doing, neither did they listen to Erickson's own
explanations of what he was doing. Instead, they observed what he SEEMED
TO be doing, and ended up with a model that even Erickson once proclaimed
more complete than his own idea of his actions.
A similar process has since been used to mirror the skills of effective
communicators, educators, street fighters, chiefs, seducers, and all sorts
of other oddities... And of course there's all the things that Anthony
Robbins has done, like Simon mentioned. All this NOT by understanding
what's actually going on inside those people's minds and bodies, but by
observing what they seem to be doing.
It's interesting to note that a lot of what NLP'ers do just plain does
not make sense at first. If you listen to Bandler talk for example, a lot
of the time his statements thoroughly violate any laws of logic and are
seemingly unrelated if you stop to think about it, but your mind just
kindda follows along anyway, because it "makes sense" in a different
sense... And the communication that has taken place is, from Bandler's
perspective, meaningful and successful, because for him the meaning of any
communication between humans is the end result that it accomplishes.
What he does is he induces various feelings, sounds, and pictures in
your imagination by using language and tonality that are bound to do so.
And the underlying principle is that whatever you can get a person to
imagine is perceived by that person as being their own thought, and
therefore they don't resist it.
Well, I see two connections to AI at this point.
On one hand, Simon's remarks about the utilization of some of these
principles by an AI to persuade humans seem very relevant to me. After
seeing some heavy NLP in action and studying it for a while, I now firmly
believe that it is absolutely feasible to persuade a human being of just
about anything, solely by using spoken language and observing that
person's reactions. By virtue of neurology, people are bound to respond
to certain linguistic stimuli in predictible ways, and whether you are
aware of what's going on or not doesn't mean a thing (just try not to
think of a beautiful large butterfly flapping it's wings right in front
of you... You just can't do it! In order to understand what I am
saying, you have no choice but to imagine a butterfly.) This obviously
has tremendous potential for benefitting and enhancing humans as well as
for causing harm beyond belief.
On the other hand, a systematic method to develop NLP-style models of
the world (meta-NLP?) seems like a subtly novel approach to machine
learning.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT