From: William Pearson (wil.pearson@gmail.com)
Date: Fri Feb 24 2006 - 10:03:53 MST
On 24/02/06, Keith Henson <hkhenson@rogers.com> wrote:
> At 11:01 PM 2/23/2006 -0800, you wrote:
> >And a human baby cannot recursively self improve.
>
> It takes outside help (consider badly neglected or extremely rare wild
> raised children) but most people would consider the progress between a
> newborn and a PhD self improvement. How is your see AI going to acquire its
> knowledge of the world?
I don't think humans are self-improving in a Seed AI sense. It is much
weaker and probably better thought of as self-specialisation. For
example during human development we lose our ability to absorb
languages like sponge early on, so we can devote resources (energy or
computational) to other processes.
What the weaker form of self-transformation is, is obviously an open
question. My theory at the moment say local (rather than the global
understanding of Seed style transformations), specialised, and bounded
so that one section of the brain cannot become the mental equivalent
of cancer if it makes a mistake in the transformation.
All this means is that we can expect that Seed AI sense
self-improvement will be of different length to humans.
Will Pearson
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT