Intelligence, and, Goal Systems

From: Ben Goertzel (ben@goertzel.org)
Date: Wed Mar 06 2002 - 22:48:22 MST


Two other things that came up on the chat session:

1) Some people thought my paper on intelligence & levels of
self-modifiability was hard to understand due to the use of math notation.

I believe I can write a version that does not involve any math notation. In
a sense there is no loss in doing this sense I did not use the math notation
to do any proofs or calculations. The reason I like using mathematical
concepts for such things is that it is the only way to unambiguously define
things. With verbal descriptions of things like "intelligence" one too
often seems to get bogged down in disputes over the meanings of words ...
and gets lost in the web of interlocking meanings.

2) Eli and I continued our debate on hierarchical vs. heterarchical goal
systems.

Recently Eli asked me to write something systematic on my ideas about goal
systems and Friendliness. I replied that it was hard for me to do so,
outside of the context of the overall Novamente design. However, I've been
thinking about this a bit, and though it's hard, I don't think it's
impossible.

So, I'll write these two documents at some point in the next month or so....
(I always think I don't have time for such things because of having so much
"serious work" on my hands, but yet the fun conceptual documents always seem
to weasel their way into my schedule anyway ;)

I'm afraid that, in spite of me being one of the biggest goofballs on the
face of God's green Earth, my presence seems to have made this SL4 chat a
little serious than the previous one.... Next time I'll try to bring up
iridescent chinchillas or some such ;> Actually, the ongoing meme at
Webmind Inc. was about goats -- but, well... this is a G-rated list ;.>

-- Ben G



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT