From: Anthony Berglas (anthony@berglas.org)
Date: Sun Jun 15 2008 - 19:20:24 MDT
Thanks for these, I'll add them.
Although without a sharp understanding of Evolution I think there is
little justification for statements like
"Self-improvement causes systems to allocate their physical and
computational resources according to a universal principle. It also causes
systems to exhibit four natural drives: 1) efficiency, 2)
self-preservation, 3)
resource acquisition, and 4) creativity."
He talks about deliberation vs Evolution (eg. when choosing a mate),
without realizing that deliberation comes from Evolution.
The paper is also a bit rambly and not sharp, IMHO.
I like the idea of split personalities. A possible source of goals.
"Self-improving systems do not yet exist". Of course they do -- from
the beginning. It is all a matter of extent. Even dumb chess
programs try to improve the "book" of openings etc. in their spare
time, thus improving themselves. Neural nets in Conway's life.
Anthony
At 12:45 AM 16/06/2008, Joshua Fox wrote:
>Note also
>- Stephen M. Omohundro, "The Basic AI Drives"
>http://selfawaresystems.files.wordpress.com/2008/01/ai_drives_final.pdf
>
>- Stephen M. Omohundro, ""The Nature of Self-Improving Artificial
>Intelligence"
>http://selfawaresystems.files.wordpress.com/2008/01/nature_of_self_improving_ai.pdf,
>Singularity Summit 2007
>
>Joshua
>
>
>On Sat, Jun 14, 2008 at 5:11 AM, Anthony Berglas <anthony@berglas.org> wrote:
> >
> > Having scanned the literature, I decided to write a paper on the
> dangers of intelligence. I have tried to keep it short, sharp and
> very focused.
> >
> > I took the trouble to write it because I could not find any other
> paper that put it all together succinctly without philosophical,
> technical, egotistical and other distractions. There are a few
> ideas in it that I have not seen in Singularity community such as
> DNA size and brain size/speech understanding. But the main purpose
> of the paper is to be succinct and convincing.
> >
> > It mainly addresses issues raised in discussions with "ordinary"
> people and software engineers -- that is the target audience. In
> particular, "computers obviously can never be intelligent". "They
> would just do what we tell them". "They would be just like us but
> smarter". And "but what about global warming, biotechnology,
> nanotechnology and other distractions".
> >
> > So all comments most welcome, especially as to what the paper
> does not need to say.
> >
> > http://berglas.org/Articles/AIKillGrandchildren/AIKillGrandchildren.html
> >
> > Anthony
> >
> >
> > Dr Anthony Berglas, anthony@berglas.org Mobile: +61 4 4838 8874
> > Just because it is possible to push twigs along the ground with ones nose
> > does not necessarily mean that is the best way to collect firewood.
> >
> >
> >
> > __________________________________________________
> > D O T E A S Y - "Join the web hosting revolution!"
> > http://www.doteasy.com
Dr Anthony Berglas, anthony@berglas.org Mobile: +61 4 4838 8874
Just because it is possible to push twigs along the ground with ones nose
does not necessarily mean that is the best way to collect firewood.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:03 MDT