Re: Emotions: The Easy Part

Steve Witham (sw@tiac.net)
Mon, 18 Aug 1997 22:59:56 -0400


I know why I don't believe emotions are the easy part.

To me, "emotions" are all the energy-allocation and value-calculating
heuristics and signals, in particular those that direct thought and
color (or colorize) memory.

In this light, "raw intelligence" would be all the brute-force algorithms
with no knowlege of efficiency, importance, other representations, context,
learning, etc. Emotion is all the heuristics, meta-heuristics, and so on.

So, emotion is everything but the easiest part. Everything except
carrying out what you've already learned how to do by rote. Anytime
there are previously-unreconciled goals or aspects of a situation that
come up simultaneously, the process that recognizes that and finds ways
to balance or resolve it is emotion. It does this by having *somewhat*
more general knowledge of kinds of conflict--but there are many
kinds of inner conflicts and many skills for dealing with them. And
there are conflicts between the conflict-resolution strategies which
require resolution, and so on.

Was it Anders who said emotion's easy? He seems to mean pre-wired
pleasure and pain. If that's all it means, then one-celled creatures
have "emotions," and we'll only have to design something as complex
as an E. Coli to have "emotion." Lewis Thomas proposed modestly that
we not start any nuclear wars until we understand one single-celled
animal completely.

Or, are we talking about emotions as complex as a worm has? A lizard?
How many billions of years of evolution of "easy part" are we trying
to duplicate?

--Steve

--
sw@tiac.net           Steve Witham          web page under deconstruction
"...when activated, it pops a message off the bag
    and recurs with the tail of the bag."
             --Vijay Saraswat and Patrick Lincoln