From: Michael Vassar (firstname.lastname@example.org)
Date: Fri Dec 16 2005 - 05:25:50 MST
Richard Loosemore said
>So when you say "Friendly AIs [...] act according to decision theory." you
>mean "The particular interpretation of how to build a Friendly AI that is
>common on this list, acts according to decision theory."
Well, since Friendly was capitalized in my post, Friendly did in fact mean,
explicitly, the sort of AI described here
http://www.intelligence.org/CFAI/index.html and elsewhere.
I'm perfectly aware that Aibo is friendly, and imagine that the upload of
any congenial person would be as well, but don't expect Aibos or uploads to
recursively self-improve safely.
On a more congenial note, I like what Richard, Chris Capel, and Jeff have
been suggesting for improving discussion quality. You might all want to
look into Chris Phoenix's appearently stalled "Wise Nano" project and
discuss with him the question of "what worked".
I don't agree that reading the complete archives before contributing is an
unreasonable standard. I did just that, and the site explicitly asks people
to, or at least used to do so. However, it's an unattained standard. If
people were just asked to read the SL4 wiki there might be more compliance.
Unlike David Picon Alvarez, I would not dare to say anything about what the
majority of posters on this list think about qualia.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:54 MDT