🚀 go-pugleaf

>[Article copied from DejaNews.  For some reason many posts in
>this thread never make it to earthlink's server.]
>
>In <618uoq$3fs@ux.cs.niu.edu> rickert@cs.niu.edu (Neil Rickert) writes:
>In <34352c7c.1516199@news.earthlink.net> savainl@earthlink.net
>(Louis Savain) writes:
>
>>  Thanks.  I think you have a somewhat pessimistic view of the power
>>of inductive learning.  Sure, it is a trial and error processes that
>>can easily give rise to superstitions, as you pointed out, but this
>>trial and error process should not be underestimated.  It can also be
>>used to send men to the moon and invent intelligent, conditionable
>>robots.
>
>Where is the evidence that induction can be used to send men to the
>moon?  I don't see anything convincing, either in the philosophy
>literature, or in the literature on machine learning, which would
>support such a claim.  Carnap admitted that he could not explain
>physical law on the basis of induction, and that in the terminology
>of confirmationism, such physical laws had been confirmed only to
>degree 0.  (You can find those comments in Carnap's "Logical
>foundations of probability).

  Sorry for not making myself clear.  I don't disagree with the above.
I was speaking of the fundamental processes of human learning.  I
believe that these processes are inductive.  Everything else (such as
the knowledge needed to send men to the moon), including logical
deduction, arithmetic, calculus, the scientific method, etc., is
learned inductively.  These things are not fundamental prerequisites
to intelligence and learning.  They are learned methods (or tools) of
discovery.  I agree with you (and David Longley) that induction alone
is inadequate to do research in the hard sciences.  Nevertheless
induction is all that is needed for an intelligence to learn.  It
allowed us humans to discover the difference between the scientific
method and superstition (Longley's term).  IOW, we discover deduction
through induction.  Do correct me if you think I'm wrong.  Unlike
Longley however, I believe the power of induction should be
emphasized, not vilified, because I don't think learning can exist
otherwise.  Everything that human beings need to know to survive and
lead a happy life can be and is learned inductively.
  Since the grand old goal of AI is to create a general learning
machine, this machine will have to be given the power of inductive
learning.  My company, Marengo Media, is promoting the thesis that the
only inductive learning that is needed, and maybe even possible, is
causal or temporal in nature, even in areas such as vision and
hearing.  We believe that biological neurons are really associative
causal event detectors.  The latest findings (see references below)
from neurophysiology seem to strongly support our stance.

Regards,

Louis Savain,

President, Marengo Media, Inc.


A few relevant sites:

http://www.hhmi.org/communic/bulletin/oct95/crack.htm
http://sanjuan.cs.wwu.edu/mobus/
http://www.cs.stir.ac.uk/~lss/AntiDromic/AntiDromic.html
http://www.sciam.com/explorations/020397brain/020397explorations.html
http://www.sat.t.u-tokyo.ac.jp/javier/wsc1.html
http://www.klab.caltech.edu/
http://www.cnl.salk.edu/CNL/