Re: Replicating What?

From: Michael Ragland (ragland37_at_webtv.net)
Date: 09/21/04 

Date: Tue, 21 Sep 2004 00:34:56 +0000 (UTC)

Guy Hoelzer <hoelzer@unr.edu> wrote or quoted:
If you find my view to be implausible, then I would ask you why natural
selection would result in the unnecessary encoding of information in the
genome, which would be relatively inflexible in the face of
unpredictable environmental variation, when it would be far more
efficient to take advantage of the opportunities for automatic
information generation through physical tendencies for
self-organization? I expect that fighting those tendencies through rigid
and expansive genomic encoding would be highly maladaptive.

Tim Tyler:
When nature builds things in, the things is builds in are usually not
strongly dependent on the environment.
It builds them in because there are costs associated with the
flexibility of learning things after birth.
The trial-and-error process involved in learning can be time consuming -
and mistakes made while it is occurring can be expensive.

Michael Ragland:
Where is your evidence when "nature builds things in", the "things it
builds are usually not strongly dependent on the environment"? Natural
selection is the principle mechanism of evolutionary change described by
Darwin in 1859; the mechanism whereby those individuals best "adapted to
the environment" contribute more offspring to succeeding generations
than do the remainder, so that as their characteristics are inherited,
the composition of the population is changed.
www.chias.org/www/edu/cse/wdnglo.html
I'm willing to say Darwinian evolution is very slow and that currently
individuals are not best adapted to their environment in certain ways
although in caveman days they would have been adapted to their
environment. But this isn't the same as stating when "nature builds
thing in, the things it builds are usually not strongly dependent on the
environment. On the contrary, when natural selection builds things in
the environment is critical in whether the organism is adaptive or not.
Just because natural selection selects individuals best adapted to the
environment does not mean that will always be the case. In terms of
evolutionary time scale the environment could change so radically that
natural selection hadn't had time to adapt to this new environment.

You state, "When nature builds things in, the things is builds in are
usually not strongly dependent on the environment.
It builds them in because there are costs associated with the
flexibility of learning things after birth. The trial-and-error process
involved in learning can be time consuming - and mistakes made while it
is occurring can be expensive." Well, I would certainly agree with you
as an organism we are generally not highly developed when it comes to
the flexibility of learning things after birth but that depends on your
definition of highly developed and flexibility of learning things. Most
human toddlers can be toilet trained, taught how to walk, speak a few
words (eventually learning a language), and as they grow older learn how
to read. Illiteracy and many other ills to society [world] are partially
a result of a lack of education, necessary infrastructure and not to
mention social, political and economic turmoil. Does nature build these
things in because there are costs associated with the flexibility of
learning things after birth?
You write, "The trial-and-error process involved in learning can be time
consuming - and mistakes made while it is occurring can be expensive."
Of course, learning is time consuming. Medical doctors go to school a
total of 12 years. Are mistakes made? Sure. The most important thing is
one learn from a mistake but of course there are incompetents in every
profession.
 

"It's uncertain whether intelligence has any long term survival value.
Bacteria do quite well without it."
 Stephen Hawking

Relevant Pages

  • Re: Reinforce learn this
    ... controller, which works as well for the robot, that our brain works for us. ... give us a few years of exposure to an environment with ... Only some basic learning hardware was there at ... This is where reinforcement learning comes in. ...
    (comp.ai.philosophy)
  • Re: Biochemistry of Genetic Mechanisms
    ... Sometimes, the environmental selection process ... conditions (in a confined environment). ... Evolution, insists that natural selection cannot account for the ... He maintains that the workings of life ...
    (sci.bio.evolution)
  • Re: The Brain
    ... aka the correct behavior to produce, then the learning ... It has no creativity. ... be entirely due to the environment. ... it must learn to apply that reward to other future situations ...
    (comp.ai.philosophy)
  • Re: Robotics, AI, and Ethics
    ... giving it negative rewards for illegal moves. ... is not how you make the learning algorithm stronger. ... the learner not in the environment. ...
    (comp.ai.philosophy)
  • Re: Robotics, AI, and Ethics
    ... because the one you wrote to play TTT isn't ... I think the "strength" of a learning system is limited ... And if the only reward ... environment we are dealing with, ...
    (comp.ai.philosophy)