singularity

Posted by Jochen Fromm-3 on Jul 19, 2006; 1:16pm
URL: http://friam.383.s1.nabble.com/no-subject-tp522127p522155.html


I don't agree with you on this topic.
In your blog you say "just imagine that you have
all that computational power right now [..] how
the hell do you program a human mind in there"
It is perhaps easier than we think: the brain
delegates its own construction largely to the
environment. All you need is therefore a suitable
environment (which is complex enough) and an
adaptive system with a high capability to learn,
including advanced learning rules (something as
simple as the delta-rule for neural networks,
Oja's learning rule or Hebb's rule). Then connect
it to the real world (through a robot) or a realistic
virtual world (through an agent) and start learning.
The drawback is of course then you might end up
with an artificial mind you don't understand
anymore. It is for instance very hard to understand
even simple neural networks that have been
trained with backpropagation and the
delta-rule.

-J.
________________________________

From: Carlos Gershenson
Sent: Wednesday, July 19, 2006 2:08 PM
To: The Friday Morning Applied Complexity Coffee Group
Subject: Re: [FRIAM] singularity

"How the hell do you program a human mind in there???
It takes us several years just to learn to talk!"