On 14-Feb-11, at 5:03 AM, Anand Manikutty wrote:
Eugen:
--- In [email protected], Eugen Leitl <eugen@...> wrote:
>
> On Sun, Feb 13, 2011 at 06:57:54PM -0000, Anand Manikutty wrote:
> > I assume that by GA, you mean Genetic Algorithms. Genetic
Algorithms are
> > a search heuristic. GA doesn't change the conclusions of my
analysis.
>
> The Darwinian evolution is a "search heuristic", too.
Yes, but the theoretical argument from the elasticity of
substitution obviously takes individual's preferences with respect
to procreation, reproduction, behavior with respect to relatives, et
cetera into account. You should have taken this into account before
framing your reply.
You asked for a mechanism by which a machine might seek self-
improvement, and I've provided one that is in common use. Talking
about individual's preferences in this context is meaningless unless
you are demanding complete anthropomorphisation of this machine and
its motivations. There is no reason to assume that a post-singularity
intelligence will subscribe to Maslow.
-Taj.
