From: Mike Dougherty (email@example.com)
Date: Tue Aug 08 2006 - 19:06:02 MDT
What is the purpose of discussing this concept? Is the idea to better
defend against any of the proposed attacks? Is there some way friendly AI
could think of every target before we do and direct resources to prevent
it? I suggest that unlike the blog, that attack plans here might include
plausible preventative countermeasures - else why bother?
On 8/8/06, Michael Anissimov <firstname.lastname@example.org> wrote:
> Use the <i> and </i> tags to put someone else's quote in italics and
> respond below.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:57 MDT